Surgery for Ovarian Cancer: Principles and Practice

Citation preview

Prelims final.qxd

12/18/2006

15:07

Page i

Surgery for Ovarian Cancer Principles and Practice

Prelims final.qxd

12/18/2006

15:07

Page iii

Surgery for Ovarian Cancer Principles and Practice

Edited by Robert E Bristow

MD

Associate Professor and Director The Kelly Gynecologic Oncology Service The Johns Hopkins Medical Institutions, Baltimore, Maryland, USA

and Beth Y Karlan

MD

Director Women’s Cancer Research Institute Division of Gynecologic Oncology Cedars-Sinai Medical Center; Professor, Obstetrics and Gynecology Geffen School of Medicine at UCLA, Los Angeles, California, USA

Prelims final.qxd

12/18/2006

15:07

Page iv

© 2006 Taylor & Francis, an imprint of the Taylor & Francis Group First published in the United Kingdom in 2006 by Taylor & Francis, an imprint of the Taylor & Francis Group, 2 Park Square, Milton Park Abingdon, Oxon OX14 4RN, UK Second printing 2007 Tel: +44 (0)20 7017 6000 Fax: +44 (0)20 7017 6699 Email: [email protected] Website: http://www.tandf.co.uk/medicine All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of the publisher or in accordance with the provisions of the Copyright, Designs and Patents Act 1988 or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 0LP. Although every effort has been made to ensure that all owners of copyright material have been acknowledged in this publication, we would be glad to acknowledge in subsequent reprints or editions any omissions brought to our attention. British Library Cataloguing in Publication Data Data available on application Library of Congress Cataloging-in-Publication Data Data available on application ISBN10: 1-84214-165-1 ISBN13: 9-78-1-84214-165-6 Distributed in North and South America by Taylor & Francis 2000 NW Corporate Blvd Boca Raton, FL 33431, USA Within Continental USA Tel: 800 272 7737; Fax: 800 374 3401 Outside Continental USA Tel: 561 994 0555; Fax: 561 361 6018 E-mail: [email protected] Distributed in the rest of the world by Thomson Publishing Services Cheriton House North Way Andover, Hampshire SP10 5BE, UK Tel: +44 (0) 1264 332424 E-mail: [email protected] Composition by Parthenon Publishing Printed and bound by Replika Press Pvt. Ltd.

Prelims final.qxd

12/18/2006

15:07

Page v

Contents

List of contributors

vii

Chapter 1

Epidemiology, staging and clinical characteristics Christina S Chu, Stephen C Rubin

1

Chapter 2

Preoperative preparation and surgical instrumentation Marcela G del Carmen, Robert E Bristow, Linda R Duska

39

Chapter 3

Management of early-stage ovarian cancer Anil K Sood, David M Gershenson

57

Chapter 4

Cytoreductive surgery: principles and rationale Christine H Holschneider, Jonathan S Berek

87

Chapter 5

Cytoreductive surgery: pelvis Robert E Bristow, Leo D Lagasse

127

Chapter 6

Cytoreductive surgery: abdominal retroperitoneum and adenopathy Jason D Wright, Thomas J Herzog

171

Chapter 7

Cytoreductive surgery: intestinal tract and omentum Yukio Sonoda, Dennis S Chi, Richard R Barakat

195

Chapter 8

Cytoreductive surgery: right upper abdomen Robert L Giuntoli II, Robert E Bristow, Richard D Schulick

227

Chapter 9

Cytoreductive surgery: left upper abdomen Krishnansu S Tewari, Michael L Berman

261

Chapter 10

Second-look surgery Ilana Cass, Beth Y Karlan

289

v

Prelims final.qxd

12/18/2006

15:07

Page vi

SURGERY FOR OVARIAN CANCER

Chapter 11

Secondary cytoreductive surgery Robert L Coleman, Adnan Munkarah, Robert E Bristow

305

Chapter 12

Laparoscopic surgery David E Cohn, Inbar Ben-Shachar, Jeffrey M Fowler

331

Chapter 13

Palliative surgery for ovarian cancer Laura J Havrilesky, Daniel L Clarke-Pearson

351

Chapter 14

Postoperative management Lee-may Chen

375

Index

413

vi

Prelims final.qxd

12/18/2006

15:07

Page vii

List of contributors

Richard R Barakat Gynecology Service Department of Surgery Memorial Sloan-Kettering Cancer Center 1275 York Avenue New York, NY 10021 USA

Robert E Bristow The Kelly Gynecologic Oncology Service Department of Gynecology and Obstetrics The Johns Hopkins Medical Institutions 600 North Wolfe Street, Phipps 281 Baltimore, MD 21287 USA

Inbar Ben-Shachar Division of Gynecologic Oncology James Cancer Hospital and Solove Research Institute The Ohio State University College of Medicine and Public Health 320 West 10th Avenue M-210 Starling Loving Hall Columbus, OH 43210 USA

Ilana Cass Division of Gynecologic Oncology Department of Obstetrics and Gynecology Cedars-Sinai Medical Center 8700 Beverly Boulevard, Suite 290W Los Angeles, CA 90048–1865 USA

Jonathan S Berek Department of Obstetrics and Gynecology Division of Gynecologic Oncology David Geffen School of Medicine at UCLA 10833 LeConte Avenue Los Angeles, CA 90095–1740 USA Michael L Berman Division of Gynecologic Oncology Chao Family Comprehensive Cancer Center University of California, Irvine Medical Center 101 The City Drive Orange, CA 92868 USA

Lee-may Chen Division of Gynecologic Oncology UCSF Comprehensive Cancer Center 1600 Divisadero Street 4th Floor San Francisco, CA 94115–1702 USA Dennis S Chi Gynecology Service Department of Surgery Memorial Sloan-Kettering Cancer Center 1275 York Avenue New York, NY 10021 USA

vii

Prelims final.qxd

12/18/2006

15:07

Page viii

SURGERY FOR OVARIAN CANCER

Christina S Chu Department of Obstetrics and Gynecology Division of Gynecologic Oncology University of Pennsylvania Health System 3400 Spruce Street Philadelphia, PA 19104 USA

Linda R Duska Division of Gynecologic Oncology Wayne State University/Karmanos Cancer Institute Harper Hospital 3990 John R Street, 2 Webber Detroit, MI 48201 USA

Daniel L Clarke-Pearson Division of Gynecologic Oncology Duke University Medical Center 3713 Benson Drive Durham, NC 27710 USA

Jeffrey M Fowler Division of Gynecologic Oncology James Cancer Hospital and Solove Research Institute The Ohio State University College of Medicine and Public Health 320 West 10th Avenue M-210 Starling Loving Hall Columbus, OH 43210 USA

David E Cohn Division of Gynecologic Oncology James Cancer Hospital and Solove Research Institute The Ohio State University College of Medicine and Public Health 320 West 10th Avenue M-210 Starling Loving Hall Columbus, OH 43210 USA Robert L Coleman Department of Obstetrics and Gynecology University of Texas, Southwestern Medical Center 5323 Harry Hines Boulevard, J7.124 Dallas, TX 75390–9032 USA Marcela G del Carmen Division of Gynecologic Oncology Vincent Memorial Obstetrics and Gynecology Massachusetts General Hospital Harvard Medical School Boston, MA 02114 USA

viii

David M Gershenson Department of Gynecologic Oncology The University of Texas MD Anderson Cancer Center 1515 Holcombe Boulevard, Unit 440 Houston, TX 77030 USA Robert L Giuntoli II The Kelly Gynecologic Oncology Service Department of Gynecology and Obstetrics The Johns Hopkins Medical Institutions 600 North Wolfe Street, Phipps 281 Baltimore, MD 21287 USA Laura J Havrilesky Division of Gynecologic Oncology Duke University Medical Center 3713 Benson Drive Durham, NC 27710 USA

Prelims final.qxd

12/18/2006

15:07

Page ix

LIST OF CONTRIBUTORS

Thomas J Herzog Division of Gynecologic Oncology Department of Obstetrics and Gynecology Columbia University 622 West 168th Street, PH 16 New York, NY 10032 USA

Richard D Schulick The Kelly Gynecologic Oncology Service Department of Gynecology and Obstetrics The Johns Hopkins Medical Institutions 600 North Wolfe Street, Phipps 281 Baltimore, MD 21287 USA

Christine H Holschneider Department of Obstetrics and Gynecology Division of Gynecologic Oncology David Geffen School of Medicine at UCLA 10833 LeConte Avenue Los Angeles, CA 90095–1740 USA

Yukio Sonoda Gynecology Service Department of Surgery Memorial Sloan-Kettering Cancer Center 1275 York Avenue New York, NY 10021 USA

Beth Y Karlan Women’s Cancer Research Institute Division of Gynecologic Oncology Cedars-Sinai Medical Center 8700 Beverly Boulevard, Suite 290W Los Angeles, CA 90048–1865 USA

Anil K Sood Department of Gynecologic Oncology and Cancer Biology The University of Texas MD Anderson Cancer Center 1515 Holcombe Boulevard, Unit 1362 Houston, TX 77030–1439 USA

Leo D Lagasse UCLA School of Medicine Division of Gymecologic Oncology Cedars-Sinai Medical Center 8635 West 3rd Street 160-W Los Angeles, CA 90048 USA Adnan Munkarah Division of Gynecologic Oncology Wayne State University/Karmanos Cancer Institute Harper Hospital 3990 John R Street, 2 Webber Detroit, MI 48201 USA

Krishnansu S Tewari Division of Gynecologic Oncology The Chao Family Comprehensive Cancer Center University of California, Irvine Medical Center 101 The City Drive Orange, CA 92868 USA Jason D Wright Division of Gynecologic Oncology Department of Obstetrics and Gynecology Washington University School of Medicine 4911 Barnes Hospital Plaza St Louis, MO 63110 USA

Stephen C Rubin Department of Obstetrics and Gynecology Division of Gynecologic Oncology University of Pennsylvania Health System 3400 Spruce Street Philadelphia, PA 19104 USA

ix

Prelims final.qxd

12/18/2006

15:07

Page x

Dedication

This book is dedicated to our spouses, Michelle and Scott; our children, Jackson and Chloe Bristow, and Matthew and Jocelyn Karlan; and our colleagues and mentors from whom we have been fortunate enough to learn and alongside whom we have faced the many challenges of gynecologic oncology and life – you have our deepest gratitude for your patience, support and guidance. Also, to our friend, colleague and mentor Rick Montz with a special note of remembrance. This book was inspired by our respect and admiration for all women with ovarian cancer, whose courage and determination will always motivate us to improve care until cures are discovered. Robert E Bristow Beth Y Karlan

x

Prelims final.qxd

12/18/2006

15:07

Page xi

Preface

Ovarian cancer is a significant cause of morbidity and mortality, and represents one of the most pressing problems in women’s health care today. Perhaps more than any other gynecologic malignancy, effective treatment of ovarian cancer requires an integrative model of multidisciplinary care that includes gynecologic, medical, surgical and radiation oncologists, pathologists, oncology nurses, psychologists and basic scientists among others. Recent advances in molecular biology, diagnostic imaging, chemotherapy, supportive care and targeted treatment strategies have contributed to an improved overall outcome for women with ovarian cancer. Nevertheless, complete and effective surgical intervention remains the cornerstone of ovarian cancer diagnosis and treatment. As surgical instrumentation and operative techniques have evolved, it has become increasingly apparent that special training is necessary in order to practice surgery for ovarian cancer beyond the fundamentals that are typically learned during postgraduate residency education. Formal fellowship training in gynecologic oncology is the preferred method of learning the myriad of individualized surgical approaches to women with ovarian cancer. In reality, however, many surgeons must acquire such expertise under less favorable circumstances. Admittedly, postgraduate training can only provide a foundation. Proficient ovarian cancer surgical skills are further developed through continued experience, challenge and self-instruction. Many surgical texts and atlases address the complexities of gynecologic and pelvic cancer surgery as a general topic, yet no single work has been devoted exclusively to the rational and effective operative management of ovarian cancer. The authors have attempted to fill this void with a

text that emphasizes the practical applicability of important procedures. The basic topics of ovarian cancer epidemiology and preoperative preparation are covered initially, followed by a section addressing the operative management of early-stage disease. The majority of the text, however, is devoted to the technical aspects of cytoreductive surgery, since this is where we perceive that the greatest need for extended training exists. The chapters on cytoreductive surgery are divided according to anatomic regions and include the relevant anatomical considerations, the surgical challenges specific to each region and the operative approaches and techniques favored by the authors. Chapters on laparoscopy, second-look surgery, secondary cytoreduction and palliative surgery follow. The final section discusses the unique postoperative care issues common to ovarian cancer surgical patients. This text is intended for all clinicians caring for women with ovarian cancer including attending surgeons, fellows and residents in the disciplines of gynecologic oncology, pelvic surgery, surgical oncology and general surgery. Ultimately, the optimal surgical management of the ovarian cancer patient is dependent on multiple factors including her age and general medical condition, the extent and biological aggressiveness of her disease, available access to an appropriately skilled multidisciplinary care team, and the preparedness of the surgeon to put forth the maximum operative effort. This text was conceived and prepared with this latter factor in mind.

Robert E Bristow Beth Y Karlan

xi

Ch 01 final.qxd

12/18/2006

15:08

Page 1

CHAPTER 1

Epidemiology, staging and clinical characteristics Christina S Chu, Stephen C Rubin

EPIDEMIOLOGY Overview Epithelial ovarian cancer is responsible for the majority of gynecologic cancer deaths, though it is only the second most common gynecologic neoplasm, accounting for approximately one-quarter of gynecologic cancer diagnoses.1,2 In 2003, the American Cancer Society estimated that 25 400 women would be diagnosed with ovarian cancer in the USA, and that 14 300 would die of their disease.3 A woman’s lifetime risk of ovarian cancer is about 1 in 70. Ovarian cancer is the fifth leading cause of cancer-related deaths among women in the USA, following cancer of the lung, breast, colon and pancreas.3 Most women are diagnosed in advanced stages and eventually die of progressive, chemotherapy-resistant disease. Overall 5-year survival is approximately 53%.3 Malignant tumors of the ovary occur at all ages. Malignant germ cell tumors are most commonly found in adolescents and young women under the age of 20. In contrast, epithelial ovarian cancer is seen primarily in patients over the age of 50, with most women diagnosed between the ages of 60 and 64. Older patients have a significantly worse prognosis. Yancik et al. noted that those over the age of 65 were more likely to be diagnosed with advanced disease.4 Similarly, the National Cancer Institute reported that, while younger patients had a mortality rate of 1.0 per 100 000, those over the age of 50 demonstrated a mortality rate of 25.5 per 100 000.5

Etiologic theories and risk factors While the exact cause of epithelial ovarian cancer is unknown, several theories have been proposed. The most common theory of ‘incessant ovulation’ postulates that ovulatory trauma to the surface epithelium of the ovary may predispose to malignant transformation. In support of this theory, women who ovulate regularly, such as nulliparous patients,6 or women with a higher number of lifetime ovulatory events, such as those undergoing early menarche or late menopause,7 appear to have a higher risk for the development of epithelial ovarian cancer. Others have noted that ovulation-inducing drugs such as clomiphene citrate may also increase the risk of ovarian cancer.8,9 In contrast, multiparous women, and those with a history of breastfeeding or oral contraceptive use, appear to have a lower risk. Casagrande et al.10 reported that the use of oral contraceptives may decrease the risk by approximately 40%. This protective effect is maximized with 10 years of use and persists up to 15 years after discontinuation. Another theory suggests that chemical carcinogens introduced via the vulva and vagina with subsequent migration to the peritoneum via the uterus and Fallopian tubes may promote development of ovarian cancer. Both Keal11 and Newhouse et al.12 noted that women with occupational exposure to asbestos had a higher rate of intra-abdominal carcinomatosis resembling ovarian cancer. In animal models, other investigators noted that abnormal changes in the ovarian epithelium could be induced after exposure to asbestos.13,14 Talc particles have also been found in

1

Ch 01 final.qxd

12/18/2006

15:08

Page 2

SURGERY FOR OVARIAN CANCER

benign and malignant ovarian tumors.13,14 In 1979, Longo and Young15 suggested that talc could be introduced to the upper genital tract via dusting on sanitary napkins, and 10 years later, Venter16 used technetium-99m-labeled human albumin microspheres to prove that particles in the lower genital tract could in fact migrate to reach the ovaries. Despite this finding, the effect of talc on ovarian cancer risk is controversial. Several studies have failed to find an association between talc or asbestos and the development of ovarian cancer.17–19 However, Cramer et al.20 examined 215 women with epithelial ovarian cancer compared to 215 controls and noted that 43% of the women with cancer used talc as a dusting powder on the perineum or sanitary napkins compared to only 28% of controls, yielding a relative risk of 1.9. Similarly, Cook et al.21 noted a relative risk of 1.6 for women using talc on the perineum. Racial trends

The incidence of ovarian cancer varies with country of origin, as well as race. Residents of industrialized and affluent areas, such as North America and Western Europe, appear to incur the highest rates of occurrence. The incidence of ovarian malignancies has been reported to be 14.9 per 100 000 for residents of Sweden and 13.3 per 100 000 for residents of the USA compared to only 4.6 per 100 000 for residents of India and 2.7 per 100 000 for residents of Japan.22 Some have speculated that differences in parity may account for variations in incidence rates between industrialized and non-industrialized nations,23 while others suggest that changes in family size due to the introduction of oral contraceptives with their inherent protective effect may be responsible for the differences.24 Within the USA, ovarian cancer appears more frequently among White women than Black women (14.2 per 100 000 vs. 9.3 per 100 000, respectively).25 Of note, although the incidence among native-born Japanese is only 3.2 per 100 000, after one to two generations of residence in the USA, Japanese immi-

2

grants develop ovarian cancer at the same rate as other native-born Americans.26 These racial differences appear to affect survival as well. For the period 1973–77, the National Cancer Institute reported the age-adjusted mortality rate to be 8.7 per 100 000 for White women versus only 6.9 for Black women.25 Similarly, in more recently gathered data, other non-White women also displayed lower mortality rates – 7.3 per 100 000 for American Indians, 4.8 for Hispanics and 3.4 for Filipinos.27 Radiation exposure, viral infection and dietary factors

Investigators have examined several other risk factors associated with the development of epithelial ovarian cancer, although the evidence is mixed (see Table 1.1). The effect of radiation exposure on subsequent ovarian cancer risk is controversial. Annegers et al. have reported a relative risk of 1.8 for those individuals exposed to radiation,28 although others have noted no significant difference in risk.12 Viruses have been purported to affect the incidence of ovarian cancer. Although rubella and influenza infection6 have been occasionally reported to exert an effect, most speculation has centered on mumps infection. Some investigators have noted a protective effect for mumps,29 although this phenomenon may be related to reproductive factors: women exposed to mumps are more likely to come from families with more children and, in turn, produce more children themselves.30,31 In contrast, reports by Menczer et al.32 and Cramer et al.33 found that mumps infection may be associated with an increased risk for ovarian cancer. These authors noted that patients with ovarian cancer often demonstrated subclinical mumps infection, which they theorized might lead to early ovarian failure with subsequent elevation of gonadotropins, stimulating ovarian epithelial growth, ultimately leading to an elevated risk for cancer. Dietary factors have also been reported to affect the risk for developing ovarian cancer. In particular, a diet high in fat and meat has been noted to increase the risk for malignancy.30 Accordingly, obesity has also been reported to be associated with increased risk.30

Ch 01 final.qxd

12/18/2006

15:08

Page 3

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Table 1.1 Suggested risk factors for the development of epithelial ovarian cancer Race White European Jewish descent Age > 50 years Residence in industrialized nation (except Japan) Diet high fat high coffee low fiber low vitamin A Environmental exposure talc asbestos radiation Reproductive early menarche late menopause nulliparity Viral mumps rubella Familial family history BRCA1, BRCA2 mutation DNA mismatch repair gene mutation

However, some have linked dietary risk with residence in industrialized nations.34 Cramer et al.35 examined the relationship between a high fat–high galactose diet and ovarian cancer risk in those populations lacking the enzyme galactose-1-phosphate uridyltransferase, which converts galactose to glucose. In these populations, increased levels of galactose are associated with increased gonadotropin levels, which are hypothesized to lead to elevated ovarian cancer risk. The authors reported a significantly lower level of enzyme activity in patients with ovarian cancer. Other dietary factors such as coffee consumption have been examined. Based on the observation that residents of Sweden have both the highest per capita

consumption of coffee and the highest incidence of ovarian cancer,36 Trichopoulos et al.37 conducted a case–control study confirming that patients with ovarian cancer did indeed consume more coffee than controls. However, other authors38,39 found no such correlation. The effect of other dietary factors is also controversial. Byers et al.38 noted an increased risk for patients with diets low in vitamin A and fiber. In contrast, a case–control study from Slattery et al.40 did not detect an association between ovarian cancer risk and intake of protein, fat, fiber, vitamin A, vitamin C, or total calories. In the largest prospective study to date, Fairfield et al. followed more than 80 000 women in the Nurses’ Health Study and found no association between intake of vitamins A, C and E and specific carotenoids, as well as fruit and vegetable intake, in relation to ovarian cancer risk.41

Familial syndromes The greatest single risk factor for the development of epithelial ovarian cancer is a family history of breast and/or ovarian cancer. In 1989, case–control studies by Koch et al.42 and Hartge et al.43 noted an increase in the risk for ovarian cancer in the setting of a family history of ovarian cancer, as well as a personal history of breast cancer. While the lifetime risk for developing ovarian cancer in the general population is between 1 and 2%, women with one family member suffering from ovarian cancer have their risk increased to 4–5%. With two affected family members, the risk increases to 7%.10,44 Approximately 5–10% of ovarian cancers occur in patients with a familial history of breast cancer, ovarian cancer or other adenocarcinoma.45–49 In many of these families, cancer risk appears to be transmitted in an autosomal dominant fashion.31 Two major syndromes of familial ovarian carcinoma have been described: hereditary breast–ovarian cancer and hereditary non-polyposis colon cancer (HNPCC).31,50 Specific genetic mutations involved in each syndrome are delineated in Table 1.2. Hereditary breast–ovarian cancer syndrome is usually seen in families with a history of cancer affecting

3

Ch 01 final.qxd

12/18/2006

15:08

Page 4

SURGERY FOR OVARIAN CANCER

first- and second-degree relatives. Women from these families tend to develop breast and/or ovarian cancer at a young age, and breast cancer may occur in both breasts. Over 90% of familial ovarian cancer cases are caused by germline mutations in the BRCA1 gene, with mutations in the BRCA2 gene being linked to most of the remaining cases.51 BRCA1, located on the long arm of chromosome 17, was initially cloned in 1994,52 and a year later, BRCA2 was isolated on the long arm of chromosome 13.53 Although the exact function of these genes is not known, BRCA1 and BRCA2 are felt to act as tumor suppressor genes, possibly involved in the repair of DNA damage54,55 and in regulation of gene expression.56–60 BRCA1 and 2 appear to be important genes in familial ovarian cancer, but not sporadic cancers. Mutations are transmitted in an autosomal dominant fashion with high penetrance. Mutations in BRCA1 have been reported to occur in 3–6% of all patients with epithelial ovarian carcinoma.61,62 Although only 1 in 800 people in the general population carries a mutation in BRCA1, the prevalence of mutations may be as high as 2.5% in individuals of Ashkenazi Jewish heritage,63,64 where three specific founder mutations, including the 185delAG and 5382insC mutations in BRCA1 and the 6174delT mutation in BRCA2, account for 90% of the cases of hereditary breast–ovarian cancer. Not all patients carrying mutations in BRCA1 and BRCA2 will develop cancer. Penetrance varies and

may be affected by age and the specific mutation inherited. In large families carrying mutations in BRCA1 studied by the Breast Cancer Linkage Consortium, the risk of developing breast cancer was approximately 82–87% by the age of 70, compared to only 11% in the general population.65 In these families, the risk for developing ovarian cancer by age 70 was approximately 44–63%, in contrast to the risk in the general population of only 1.4%. Overall, high penetrance of BRCA1 mutations in these families conveyed a greater than 90% lifetime risk of developing either breast or ovarian cancer. In similar studies of BRCA2 families, the risk of developing breast or ovarian cancer by age 70 was noted to be 84% and 27%, respectively.66 Overall, by the age of 70, the risk of developing either breast or ovarian cancer was estimated to be 88%. Mutation carriers in the general population may demonstrate lower penetrance. In a study of 120 Ashkenazi Jewish patients unselected for family history carrying one of the three founder mutations in BRCA1 and BRCA2, risks for developing breast cancer or ovarian cancer by age 70 were estimated to be 56% and 16%, respectively.67 BRCA1 and BRCA2 mutation carriers appear to exhibit a distinct clinical course. Rubin et al.68 reported on 53 ovarian cancer patients with germline mutations in BRCA1, and noted significantly improved survival compared to patients with sporadic cancers. The actuarial median survival for the 43 patients with advanced disease was 77 months compared to only 29

Table 1.2 Genes associated with hereditary ovarian cancer syndromes Gene

Syndrome

Location

Percentage of ovarian cancer

BRCA1

Breast and ovarian cancer

17q21

4.1

BRCA2

Breast and ovarian cancer

13q12

3.3

Mismatch repair genes MSH2 MLH1 MSH6 PMS2 PMS1

Hereditary non-polyposis colorectal cancer

4

2.9 2p22-p21 3p21 2p16-p15 7p22 2q31-q33

Ch 01 final.qxd

12/18/2006

15:08

Page 5

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

months for age- and stage-matched controls (p < 0.001). In 2000, Boyd et al.69 confirmed these findings. They examined 88 Jewish ovarian cancer patients with BRCA1 or BRCA2 mutations and noted that, compared to controls, mutation carriers had both longer median time to recurrence (7 months vs. 14 months, p < 0.001) and increased survival (p = 0.004). Among patients with stage III tumors, after adjustment for age and residual disease, BRCA mutation status was noted to be an independent prognostic factor, conferring a 25% reduction in the relative risk of death when compared to sporadic cancers. HNPCC, or Lynch II syndrome, occurs in families affected by a combination of early-onset colon cancer and cancers of the ovary, endometrium and other gastrointestinal sites. This syndrome is a less common cause of familial ovarian cancer, accounting for only about 2% of cases. HNPCC has been linked to multiple genetic mutations in the DNA mismatch repair (MMR) genes.70,71 Mutations in these genes result in the loss of the ability to correct mismatched DNA nucleotides, causing microsatellite instability. Accumulation of genetic replication errors in tumor suppressor genes and oncogenes ultimately leads to carcinogenesis. While five different mutations in the MMR genes have been noted to cause HNPCC, mutations in MLH1 on chromosome 3 and MSH2 on chromosome 2 account for 45% and 49% of cases, respectively.72 Most of the remainder occur in PMS2 on chromosome 7, while mutations in PSM1 and MSH6 are noted only sporadically. Although little is known about the prevalence of MMR gene mutations, one large study conducted in Finland estimated the overall carrier frequency of mutations in MLH1 and MSH2 to be 1 in 660.73 Penetrance for HNPCC carriers also varies. By age 65, approximately 70% of carriers will develop colorectal cancer.74,75 The frequency of non-colonic cancers varies among families. The majority of HNPCC carriers who develop colon cancer will also develop a second primary, usually a synchronous or metachronous colon cancer or endometrial cancer. Endometrial, gastric and biliary tract carcinoma are

the most common extracolonic primaries, with lifetime risks of 43–60%, 13–19% and 18%, respectively.74 The cumulative lifetime risk for the development of ovarian cancer is approximately 9–12%,74 and is usually associated with mutations in MSH2. Screening

Because of the low prevalence of mutations, genetic screening is not appropriate for the general population. The American Society of Clinical Oncology recommends testing for individuals with a strong family history of early-onset disease, when results of testing will affect medical management.76 Obtaining an accurate family history is crucial to identify patients at risk for genetic mutation. A careful three-generation family history should be initially assessed and updated at each visit. Based on history, certain patients may benefit from genetic testing for BRCA1 and BRCA2 mutations. These include individuals having a personal or family history of breast cancer diagnosed before age 50 and ovarian cancer diagnosed at any age; a first-degree relative with a known mutation; a family history of two or more cases of breast cancer diagnosed before age 50; a family history of two or more cases of ovarian cancer; a personal or family history of bilateral breast cancer; a family history of male breast cancer; or Ashkenazi Jewish ancestry in the setting of a personal or family history of breast or ovarian cancer.76 For individuals at risk for HNPCC mutations, the Amsterdam and Bethesda criteria are used to identify those likely to benefit from genetic screening.75 The Amsterdam criteria recommend testing for individuals with a family history of at least two successive generations affected by colorectal cancer; diagnosis of a family member before the age of 50; colon cancer diagnosed in at least three relatives; or an increased incidence of other cancers (such as ovarian, endometrial, gastric, urinary tract and biliary). The Bethesda criteria designate that individuals with either two cases of colon cancer (in very small families) or two first-degree relatives with colon cancer, combined with a third relative with either early-onset colon

5

Ch 01 final.qxd

12/18/2006

15:08

Page 6

SURGERY FOR OVARIAN CANCER

cancer or endometrial cancer should undergo genetic testing. Prevention strategies

No method exists which can reliably identify patients with early ovarian cancer,77 but attempts have been made to screen populations of women at elevated risk. In 1994, the National Cancer Institute Consensus Statement on Ovarian Cancer78 recommended annual or semiannual bimanual pelvic examination, transvaginal ultrasonography (TVUS) and CA-125 measurement to screen high-risk women. Similarly, in 1997, the Cancer Genetics Studies Consortium recommended annual or semiannual TVUS and CA-125 testing starting at the age of 25–35 in women known to be BRCA1 or BRCA2 mutation-positive.75 Despite the recommendations, these methods have limitations. While relatively sensitive in its ability to detect small potentially curable ovarian tumors,79 TVUS may miss ovarian cancers in the setting of normal-sized ovaries or primary peritoneal cancers, and cannot necessarily distinguish between benign and malignant tumors, which may lead to unnecessary invasive procedures. While CA-125 is the most commonly used tumor marker to screen for ovarian cancer, the value is elevated in only 80% of cases. For patients with clinically diagnosed stage I disease, potentially the most curable patients, only half exhibit elevations.80 Again, false positives may lead to unneeded surgical intervention. The National Cancer Institute acknowledged that no definitive evidence exists to show that these tests can improve the survival of women ultimately diagnosed with the disease.78 In regards to increased breast cancer risk, the Cancer Genetics Studies Consortium recommends monthly self-examinations of the breast, beginning at the age of 18, and annual mammography beginning at age 25–35 for BRCA mutation carriers. Because of the difficulty in reading mammograms in younger women, annual or semiannual clinical breast examinations are also advised.75 For HNPCC mutation carriers, the Cancer Genetics Studies Consortium81 also

6

recommends screening for colon and endometrial cancers. Recommendations include colonoscopy every 1–3 years beginning at age 25, and either annual endometrial biopsy and/or TVUS beginning at age 25–35. In addition to increased vigilance, medical intervention may decrease the risk for ovarian cancer. Oral contraceptive use was noted to reduce the risk of ovarian cancer in the general population by as much as 40% as early as 1979.10 Use of chemoprophylaxis for BRCA1 and BRCA2 mutation carriers was not examined until 1998, when Narod et al.82 conducted a case–control study comparing 207 women with BRCA1 and BRCA2 mutation-related ovarian cancer with 161 of their cancer-free sisters. The authors found that any past use of oral contraceptives conferred a 50% reduction in risk for both BRCA1 and BRCA2 mutation carriers. No difference in breast cancer risk was noted. However, other researchers have found conflicting results. Modan et al.83 conducted a large population-based case–control study examining 840 Jewish women in Israel with ovarian cancer and 751 controls. All women were tested for the three founder mutations in BRCA1 and BRCA2, and the effect of oral contraceptive use on ovarian cancer risk was assessed. The use of oral contraceptives only appeared to reduce the risk in women without mutations. Among mutation carriers, the reduction in risk was only 0.2% per year of use. Ness et al.84 examined 727 women with ovarian cancer and 1360 controls in relation to use of various types of contraception including oral contraceptive pills, intrauterine devices, barrier methods, tubal ligation and vasectomy. The authors noted that nulligravid women did not derive reduction in cancer risk from any method, whereas multigravid women improved their risk with all contraceptive methods. They concluded that the risk reduction afforded by oral contraceptives may be due to mechanisms unrelated to hormonal or ovulatory status. Narod et al.85 confirmed that, among BRCA1 mutation carriers, tubal ligation decreased risk, as did the combination of tubal ligation and oral

Ch 01 final.qxd

12/18/2006

15:08

Page 7

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

contraceptive use. The same effects were not seen for BRCA2 mutation carriers. The results of prospective chemoprevention trials are necessary to resolve the question of the effectiveness of oral contraceptives for ovarian cancer risk reduction. However, it appears that mutation carriers may use oral contraceptives without increase in their breast cancer risk. The most definitive method for preventing ovarian cancer in high-risk women is surgical extirpation of the ovaries. Because prophylactic oophorectomy may usually be accomplished laparoscopically on an outpatient basis, the potential morbidity and mortality of the procedure is low enough to counter the elevated risk of developing ovarian cancer in mutation carriers. However, aside from surgical risks and issues associated with early menopause, it is important that the physician counsel prospective patients that the procedure is not entirely protective. Streuwing et al.86 concluded that prophylactic oophorectomy may confer only a 50% reduction in risk of ovarian or primary peritoneal cancer. Primary peritoneal carcinoma (PPC), which is histologically indistinguishable from ovarian cancer, has been documented to occur with increased frequency in women susceptible to hereditary ovarian cancer. Tobacman et al.87 reported on three cases (11%) of intra-abdominal carcinomatosis resembling metastatic ovarian cancer among 28 women from high-risk families who had undergone prophylactic oophorectomy. Similarly, Piver et al.31 examined data from the Gilda Radner Familial Ovarian Cancer Registry and identified six cases (1.85%) of primary peritoneal cancer in 324 high-risk women who had undergone previous prophylactic oophorectomy from 1 to 27 years earlier. Because of these risks, both thorough exploration of the abdominopelvic cavity at the time of prophylactic surgery and meticulous histologic examination of even normal-appearing ovaries must be performed in order to exclude the presence of an occult malignancy. Care should also be taken to remove the fallopian tubes entirely, in order to decrease the risk of tubal carcinoma.

PATTERNS OF SPREAD AND STAGING BY THE INTERNATIONAL FEDERATION OF GYNECOLOGY AND OBSTETRICS Overview of anatomy The pelvis is divided by a fold of peritoneum named the broad ligament, which runs transversely across the cavity. The uterus and fallopian tubes are situated between the anterior and posterior leaves of the broad ligament, with the free ends of the fallopian tubes opening into the peritoneal cavity. The round ligament runs within the leaves of the broad ligament, extending from a point anterior to the uterotubal junction to the pelvic sidewall, through the internal inguinal ring, to merge with the subcutaneous tissues of the labia majora. The ovaries are suspended from the posterior surface of the uterus by the uteroovarian ligaments and are attached to the broad ligament via the mesovarium. Blood and lymphatic vessels supplying the ovary course through a peritoneal fold (called the infundibulopelvic ligament, or suspensory ligament of the ovary) from the ovary to the pelvic wall. The ovaries are not covered by the peritoneum, but are suspended freely within the peritoneal cavity (Figure 1.1). The ovaries are supplied by the ovarian arteries, which arise from the aorta near the level of the L2 vertebra. The ovarian arteries then course along the posterior abdominal wall and cross over the external iliac vessels at the pelvic brim to enter the infundibulopelvic ligament. The ovarian artery supplies the ovary through the mesovarium and continues to course along the fallopian tube and uterus, eventually to anastomose with the uterine artery. The ovarian veins typically follow the course of the ovarian arteries, with the right ovarian vein usually draining directly into the inferior vena cava to the renal vein, and the left ovarian vein draining into the left renal vein. Most lymphatic drainage from the ovary, as well as some from the fallopian tube and uterus, forms a plexus in the mesovarium. The lymphatic channels then course through the infundibulopelvic ligament

7

Ch 01 final.qxd

12/18/2006

15:08

Page 8

SURGERY FOR OVARIAN CANCER

Urinary bladder Deep inguinal ring Round ligament Fundus of uterus Utero-ovarian ligament Broad ligament

Fallopian tube Body (corpus) of uterus Cervix of uterus Sigmoid colon

Mesovarium Ovary Rectouterine pouch (cul-de-sac of Douglas) External iliac vessels Uterosacral ligament Infundibule pelvic ligament (contains ovarian vessels)

Sacral promontory Middle sacral vessels

Abdominal aorta

Figure 1.1 Anatomy of the pelvis

with the ovarian vessels out of the pelvis. At the level of the lower pole of the kidney, the lymphatics join the aortic lymphatic chain. Some direct lymphatic channels also drain from the ovary to the common iliac chain. In addition, some lymphatic fluid may travel through the leaves of the broad ligament to the pelvic sidewall, and eventually drain to the external iliac, internal iliac and obturator lymph nodes.88

Routes of spread Epithelial ovarian cancers generally arise in cysts from the germinal epithelium of the ovary. When the tumor’s growth penetrates the ovarian capsule, spread occurs mainly through peritoneal and lymphatic dissemination (Table 1.3). Once the tumor reaches the surface of the ovary, malignant cells may exfoliate to float freely in the peritoneal cavity. As these malignant cells are circulated by the normal flow of peritoneal fluid up the right paracolic gutter to the under-

8

Table 1.3 Routes of spread Peritoneal dissemination Direct extension Lymphatic spread Hematogenous spread

surface of the right hemidiaphragm, implantation on any abdominopelvic surface may occur, causing diffuse seeding of the peritoneal cavity, which may manifest as either microscopic or macroscopic tumor nodules. Commonly, implants may occur on the uterus, adnexae, bowel, omentum, hemidiaphragms and liver capsule (Figure 1.2). It is important to note that peritoneal spread may occur even in the absence of gross capsular infiltration.89 Formation of ascites occurs in more than two-thirds of patients. Ascites may result from both decreased plasma oncotic pressure, which

Ch 01 final.qxd

12/18/2006

15:08

Page 9

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Diaphragm

Liver Serosal bowel implants Colon

Pleura

Stomach

Omentum

Nodes

Ovaries Pelvic peritoneal implant

Figure 1.2 Routes of spread

favors third-spacing of fluid into the pertioneal cavity, and the increased production of fluid by damaged peritoneal surfaces. Decreased absorption of fluid may occur as a result of tumor obstruction of lymphatic channels.90,91 Another common route of spread is lymphatic dissemination. As discussed previously, ovarian lymphatics predominantly drain to the para-aortic, iliac and obturator lymph nodes. Autopsy studies have demonstrated lymph node metastases in up to 80% of ovarian cancer patients.92,93 At the time of diagnosis, Burghardt et al.94 examined 180 patients of various stages and noted that, in patients with apparent stage I disease, 24% had positive lymph nodes. This number increased to 50% for those with apparent stage II disease, and to more than 70% for those with apparent stage III or IV disease. Of 105 patients who underwent both pelvic and para-aortic lymphadenectomy, 12% had positive pelvic but negative para-aortic nodes, while 9% had positive para-aortic but negative pelvic nodes. In addition to these routes, ovarian cancer may spread by direct extension, as ovarian tumor masses grow to engulf and invade the adnexa and surrounding structures. Hematogenous spread may also

occur, and manifest as brain or parenchymal lung and liver metastases.

Staging by the International Federation of Gynecology and Obstetrics Staging is performed to define the extent of tumor spread. Staging for ovarian cancer is accomplished surgically, with laparotomy generally serving the dual purpose of staging and primary treatment via surgical cytoreduction. Table 1.4 defines the International Federation of Gynecology and Obstetrics (FIGO) staging schema. Assignment of the patient’s stage of disease is based on findings at the time of surgery. An adequate staging procedure (Table 1.5) begins with a vertical midline incision large enough to allow full examination of the pelvis and abdomen. Unfortunately, most patients present in late stages (Table 1.6) with diffuse abdominal disease, and stage is ascertained by gross examination rather than by surgical sampling. In the absence of widespread disease, pelvic and abdominal washings are first obtained for cytologic analysis. A thorough examination of all pelvic and abdominal surfaces is then carried out. In addition to

9

Ch 01 final.qxd

12/18/2006

15:08

Page 10

SURGERY FOR OVARIAN CANCER

Table 1.4 FIGO staging of ovarian cancer Stage

Definition

Stage I IA

Growth limited to the ovaries Growth limited to one ovary; no ascites present containing malignant cells; no tumor on the external surfaces; capsule intact Growth limited to both ovaries; no ascites present containing malignant cells; no tumor on the external surfaces; capsules intact Tumor stage IA or IB, but with tumor on the surface of one or both ovaries; or with capsule ruptured; or with ascites present containing malignant cells, or with positive peritoneal washings

IB IC Stage II IIA IIB IIC

Growth involving one or both ovaries with pelvic extension Extension and/or metastases to the uterus and/or tubes Extension to other pelvic tissues Tumor stage IIA or IIB but with tumor on the surface of one or both ovaries; or with capsule(s) ruptured; or with ascites present containing malignant cells, or with positive peritoneal washings

Stage III

Tumor involving one or both ovaries with peritoneal implants outside the pelvis and/or positive retroperitoneal or inguinal nodes; superficial liver metastasis equals stage III; tumor is limited to the true pelvis but with histologically verified malignant extension to small bowel or omentum Tumor grossly limited to the true pelvis with negative nodes with histologically confirmed microscopic seeding or abdominal peritoneal surfaces Tumor of one or both ovaries; histologically confirmed implants of abdominal peritoneal surfaces, none exceeding 2 cm in diameter; nodes negative Tumor of one or both ovaries; histologically confirmed implants of abdominal peritoneal surfaces, none exceeding 2 cm in diameter; nodes negative

IIIA IIIB IIIC Stage IV

Growth involving one or both ovaries with distant metastasis; if pleural effusion is present, there must be positive cytologic test results to allot a case to stage IV; parenchymal liver metastasis equals stage IV

Table 1.5 Elements of ovarian cancer staging Adequate vertical incision Abdominal and pelvic washings Inspection and palpation of all abdominal and pelvic surfaces Random peritoneal biopsies Biopsy/resection of peritoneal adhesions Removal of affected ovary Removal of remaining ovary, uterus and tubes* Total abdominal hysterectomy and bilateral salpingooophorectomy Omentectomy Appendectomy Pelvic and para-aortic lymph node sampling * May be forgone in selected patients

10

the pelvic organs, the entire abdominal cavity must be adequately inspected, including the bowel, liver, mesentery, omentum and undersurfaces of the diaphragm. Biopsies are taken from multiple random sites, as well as from any adhesions or suspicious locations. Omentectomy is also performed to rule out occult metastases. In the absence of gross upper abdominal disease greater than 2 cm in diameter (stage IIIC), sampling of the pelvic and para-aortic lymph nodes to rule out nodal metastases is indicated. In general, a total abdominal hysterectomy with bilateral salpingo-oophorectomy is performed. When disease is limited to the pelvis, great care must be taken to avoid rupture of the neoplasm during removal to prevent possible intraoperative spread of tumor. Certain patients may forgo hysterectomy, depending on age, reproductive desires, tumor type and the

Ch 01 final.qxd

12/18/2006

15:08

Page 11

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Table 1.6 Stage at initial diagnosis and survival by stage (all histologic classes). From reference 226

rare neoplasms of uncertain origin that may involve the ovary (Table 1.8). A summary of the incidence of tumor bilaterality is found in Table 1.9.

Patients Stage

n

%

5-Year survival (%)

I

1639

27

79

II

748

12

60

III

2440

40

22

IV

1291

21

14

Total

6118

100

31

extent of disease. Young patients who desire future childbearing with certain unilateral germ cell, sex cord–stromal, or suspected borderline tumors may undergo surgical staging with removal of only the affected ovary and preservation of the contralateral ovary and uterus. Accurate assignment of stage is crucial to assessment of prognosis. The need for thorough surgical staging cannot be overstated: evidence suggests that, in cases where disease appears limited to the ovary (apparent stage I disease), 30% of patients may have occult nodal metastases (actual stage IIIC disease).94 Upon examination of multiple series of apparent early-stage ovarian cancer, occult metastases may occur in 7–10% of peritoneal biopsies.95 While the overall 5-year survival of patients with ovarian cancer is approximately 50%, survival varies by stage (Table 1.6), ranging from 79% for those with stage I disease, to 14% for those with stage IV disease.

CLASSIFICATION AND CLINICAL CHARACTERISTICS OF OVARIAN CANCER The current classification of ovarian tumors is based on the schema delineated by the World Health Organization (WHO) in 1999 (Table 1.7).96 The system is based on the histogenesis of the normal ovary and includes tumors of epithelial, germ cell, sex cord–stromal type, as well as metastases and other

Epithelial ovarian cancer Epithelial ovarian tumors are the most common type of ovarian cancer, accounting for approximately 90% of all malignancies in women. These lesions are thought to arise from the surface epithelium of the ovary, which is derived from the mesothelial lining of the celomic cavity. During embryonic development, this mesothelial lining also gives rise to the Müllerian ducts which form the uterus, fallopian tubes and upper vagina. Epithelial inclusion cysts of the ovary may undergo neoplastic transformation to display different types of Müllerian differentiation (Table 1.10). Serous tumors are the most common, followed by mucinous, endometrioid and clear cell tumors. Brenner tumors, which resemble transitional cell epithelium, are another potential type of differentiation. Epithelial ovarian tumors may be benign, malignant, or of low malignant potential. Malignant epithelial ovarian cancer is treated by surgical extirpation, followed by platinum- and taxane-based chemotherapy. Patients with stage IA grade 1 or 2 tumors may be treated with surgery alone. Each subtype will be briefly discussed. Serous

Serous adenocarcinomas comprise the majority of ovarian epithelial malignancies (Figure 1.3). The mean age at presentation is 56 years.97 About onethird of stage I tumors are bilateral, while more advanced lesions are bilateral in about two-thirds of cases. Grossly, these tumors are composed of both cystic and solid components. Microscopically, serous malignancies display cells resembling fallopian tube epithelium. Psammoma bodies are present in 30% of cases.98 Serous tumors may also produce extracellular mucin. Well-differentiated, grade 1 lesions may display well-formed glands and papillary fronds, occasionally showing ciliated cells. Moderately differentiated, grade 2 lesions display sheets of tumor cells in

11

Ch 01 final.qxd

12/18/2006

15:08

Page 12

SURGERY FOR OVARIAN CANCER

Table 1.7 WHO histologic classification of ovarian neoplasms. From reference 96

I.

E. Unclassified F. Steroid (lipid) cell tumors 1. Stromal luteoma 2. Leydig cell tumor 3. Steroid cell tumor, not otherwise specified

Surface epithelial tumors A. Serous tumors B. Mucinous tumors 1. Endocervical-like 2. Intestinal type C. Endometrioid tumors D. Clear cell tumors E. Transitional cell tumors F. Squamous cell tumors G. Mixed epithelial tumors H. Undifferentiated and unclassified carcinoma

II. Sex cord–stromal tumors A. Granulosa–stromal cell tumors 1. Granulosa cell tumors (adult and juvenile) 2. Tumors in the thecoma–fibroma group a. Thecoma b. Fibroma c. Cellular fibroma d. Fibrosarcoma e. Sclerosing stromal tumor B. Sertoli–stromal cell tumors 1. Sertoli cell tumors 2. Sertoli–Leydig cell tumors C. Sex cord tumors with annular tubules D. Gynandroblastoma

III. Germ cell tumors A. Dysgerminoma B. Yolk sac tumor (endodermal sinus tumor) C. Embryonal carcinoma D. Polyembryoma E. Choriocarcinoma F. Teratoma 1. Immature 2. Mature 3. Struma ovarii 4. Carcinoid tumors G. Mixed germ cell tumors IV. Gonadoblastoma V. Miscellaneous A. Small cell carcinoma B. Malignant lymphomas, leukemias, plasmacytomas C. Unclassified D. Metastatic tumors

Table 1.8 Distribution of types of ovarian malignancies. From

Table 1.9 Incidence of bilaterality of various ovarian cancers.

reference 97

From references 115, 227 and 228

Histologic type

Frequency (%)

Epithelial

86

Germ cell

6

Sex cord–stromal

4

Metastatic

3

addition to areas with defined papillae. Poorly differentiated, grade 3 lesions predominantly exhibit sheets of malignant cells. Approximately half of serous carcinomas are grade 3. Grade of the tumor has been found to correlate with stage and prognosis.

12

Tumor type

% Bilateral

Epithelial serous mucinous endometrioid clear cell

33–66 10–20 13–30 12–39

Germ cell immature teratoma dysgerminoma other malignant germ cell tumors

2–5 10–15 rare

Sex cord–stromal thecoma Sertoli–Leydig cell tumor granulosa–theca cell tumor

rare rare rare

Ch 01 final.qxd

12/18/2006

15:08

Page 13

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

a

b

Figure 1.3 High-grade serous carcinoma. (a) Gross appearance. Cystic and solid components with extensive hemorrhage. (b) Microscopic appearance. Papillary structures interspersed with slit-like spaces. Extensive nuclear atypia and frequent mitoses present

Table 1.10 Frequency (%) of malignant epithelial histologies. Modified from reference 97 Histologic type

Malignant

LMP

Total

Serous

37.2

16.3

53.5

Mucinous

3.5

11.6

15.1

Endometrioid

15.1

3.5

18.6

Clear cell

7.0

rare

7.0

Brenner

rare

rare

rare

Mixed

2.3

2.3

Undifferentiated

3.5

3.5

Total

68.6

31.4

100

Figure 1.4 Psammocarcinoma with extensive psammoma body formation

LMP, low malignant potential

When extensive psammoma bodies are noted in conjunction with only low to moderate nuclear atypia, tumors are termed psammocarcinoma (Figure 1.4).96,99 These neoplasms demonstrate microscopic invasion, and are typically diagnosed in advanced stages. However, the clinical course may more closely resemble that of lesions with low malignant potential and favorable outcome than typical serous carcinomas.99

Mucinous

Most mucinous ovarian neoplasms are benign, and approximately 15% are of low malignant potential. Only 5% are frankly malignant. Patients present at a mean age of 52 years.97 Grossly, mucinous tumors may be very large. Although malignant tumors may be bilateral in 20% of cases, benign growths are only rarely bilateral. Microscopically, mucinous carcinomas are composed of cells resembling the lining of the

13

Ch 01 final.qxd

12/18/2006

15:08

Page 14

SURGERY FOR OVARIAN CANCER

endocervix or colon (Figure 1.5), and the possibility of metastasis from a gastrointestinal primary must be excluded. Pseudomyxoma peritonei is the presence of mucinous ascites throughout the peritoneal cavity. Most cases of pseudomyxoma peritonei are thought to be due to a primary appendiceal lesion metastatic to the ovary and peritoneum.100–102 Endometrioid

The majority of endometrioid neoplasms are malignant (Figure 1.6). The mean age at presentation is 57 years.97 On gross examination, these tumors are usu-

Figure 1.5 Mucinous carcinoma, intestinal type

a

ally 12–25 cm in size,103,104 and have cystic and solid components. Microscopically, endometrioid carcinomas resemble tumors commonly found in the uterine endometrium. Extracellular mucin may be present. Squamous differentiation is common, and may be benign or malignant. In approximately 10% of cases, endometrioid carcinoma is associated with endometriosis,105 and may arise as a result of malignant transformation. Ovarian endometrioid carcinomas may be associated with endometrial cancers of the uterus in 10–25% of cases.106–108 Differentiation between synchronous primaries and metastatic lesions from one site to another may be problematic. Most cases involving well-differentiated lesions appear to represent synchronous primaries in the uterus and ovary. However, Ulbright and Roth109 defined cases as endometrial primary tumors with ovarian metastases in the presence of either a multinodular ovarian pattern, or two or more of the following criteria: ovaries less than 5 cm in size; bilateral ovarian involvement; deep myometrial invasion; vascular invasion; and involvement of the fallopian tube lumen. Malignant mixed mesodermal (Müllerian) tumors (MMMT) of the ovary are rare, aggressive neoplasms accounting for less than 1% of all primary ovarian tumors.110 These tumors are classified with endometrioid lesions of the ovary because of similarity to

b

Figure 1.6 Endometrioid carcinoma. (a) Gross appearance. Predominantly solid with cystic spaces. (b) Microscopic appearance. Endometrioid glands infiltrating the stroma

14

Ch 01 final.qxd

12/18/2006

15:08

Page 15

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

their uterine counterpart. The average age at diagnosis is 66 years, approximately one decade older than for patients with most other epithelial subtypes.110 MMMTs are almost always diagnosed in perimenopausal or postmenopausal women. These tumors may be homologous (also called carcinosarcomas),111 or heterologous, containing elements not found in the ovary such as skeletal muscle, cartilage, or bone. Patients are usually diagnosed with advanced disease, and the prognosis is grim. DiSilvestro et al.110 reviewed 246 cases in the literature and noted a median survival of only 6–12 months, with more than 70% of patients dead 1 year from diagnosis despite treatment. Clear cell

Clear cell tumors comprise only about 7% of epithelial ovarian malignancies. The mean age at diagnosis is 53 years.112 Grossly, clear cell carcinomas may be predominantly solid or cystic, and contain white, yellow, or pale brown polypoid masses protruding into the lumen of cysts.113 Clear cell carcinomas are usually 5–20 cm in size,114 and are bilateral in about 12% of cases.115 Microscopically, the tumors are composed of clear cells and hobnail cells, though some may contain cuboidal or signet-ring cells (Figure 1.7). The typical clear cell appearance is due to the presence of copious cytoplasmic glycogen. Tumor cells may be

arranged in solid, glandular, tubular, or papillary patterns. A strong association exists between clear cell carcinoma and endometriosis, with ovarian endometriosis documented in up to 67% of cases.116 Clear cell carcinoma has also been reported to arise in extraovarian endometriosis.117 Transitional cell

Transitional cell tumors are composed of cells similar in appearance to transitional cells of the urinary tract. The term malignant Brenner tumor describes those lesions containing both invasive transitional cell groups as well as benign nests, while transitional cell carcinoma refers to those tumors lacking a benign Brenner component (see Figure 1.8).96 Brenner tumors comprise only 2% of all ovarian tumors, and over 99% are benign.118 Patients with malignant Brenner tumors present at a mean age of 64 years. Roth and Czernobilsky119 reported that those with well-differentiated lesions were usually diagnosed with stage I disease and incurred a good prognosis. Patients with poorly differentiated lesions were also noted to present with early-stage disease (80% stage IA), although 5-year survival was only 60%. Despite histologic similarities, transitional cell carcinomas display different clinical behavior. Austin and Norris120 reported that patients with transitional cell carcinoma presented with stage II–IV disease in 69% of cases,

Figure 1.7 Clear cell carcinoma. Solid pattern with cells

Figure 1.8 Transitional cell carcinoma. Irregular nests of

containing clear cytoplasm

transitional cells

15

Ch 01 final.qxd

12/18/2006

15:08

Page 16

SURGERY FOR OVARIAN CANCER

compared to only 19% of those with malignant Brenner tumors. In addition, the prognosis appeared to be worse. Among patients with stage IA transitional cell carcinoma, only 43% were alive at last contact, compared to 88% of patients with stage IA malignant Brenner tumor. Squamous

The vast majority of squamous carcinomas of the ovary arise within teratomas, or as a consequence of metastatic spread, and are not of true epithelial origin. The WHO stipulates that squamous carcinomas of the epithelial–stromal category must arise in association with either endometriosis121,122 or a Brenner tumor, or else occur in a pure form without teratomatous elements.96 Pins et al.122 reported seven cases of squamous carcinoma associated with endometriosis. Patients were diagnosed at a mean age of 49 years. Only one patient had stage I disease, and all tumors were grade 3. The authors cited a poor median survival of only 5 months. Of 11 patients with pure squamous carcinoma, the mean age at diagnosis was 56 years. Again, only one case was stage I at diagnosis, and all but one were grade 3. Grossly, the pure squamous carcinomas ranged in size from 6 to 26 cm, and were usually solid with focal areas of necrosis. Mixed epithelial types

Approximately 2–3% of epithelial ovarian cancers contain a mixture of different elements. When more than 10% of a tumor is composed of a different type of epithelium, the neoplasm qualifies for the diagnosis of a mixed epithelial tumor.96 Common mixed carcinomas include clear cell–endometrioid tumors, and serous–endometrioid tumors. Undifferentiated carcinoma

The WHO defines undifferentiated carcinoma as a malignant epithelial tumor that is too poorly differentiated to warrant classification into any of the categories previously described. Fewer than 5% of epithelial malignancies are undifferentiated. Small foci of distinguishable features such as gland formation,

16

psammoma bodies, or mucin production do not preclude the diagnosis of undifferentiated carcinoma.96 Some tumors may display neuroendocrine features, but should be differentiated from pure small cell tumors. Most tumors are solid with foci of hemorrhage and necrosis present on a cut surface. At diagnosis, about 75% are diagnosed in stages II–IV, and prognosis is generally poor with a 5-year survival rate of about 15%.123 Tumors of low malignant potential

Tumors of low malignant potential (LMP), also called borderline malignancies, were first included in the categorization of epithelial tumors of the ovary in 1971 by FIGO.124 The WHO soon followed suit in 1973.125 Approximately 15% of all ovarian malignancies are of LMP. Diagnosis of LMP tumors is made on a histologic basis. To ensure an accurate diagnosis, at least one histologic section should be examined per 1–2 cm of the greatest diameter of the ovarian cyst to ensure adequate tissue sampling. LMP tumors display epithelial budding, multilayering of the epithelium, frequent mitoses and cellular atypia.126 These neoplasms are generally non-invasive, although approximately 10% may show evidence of microinvasion, which does not appear to affect prognosis adversely.127,128 LMP tumors of all epithelial histologic types have been reported, although serous (Figure 1.9) and mucinous subtypes (Figure 1.10) are the most common. In a review of several studies comprising 352 patients with LMP lesions, 51.7% were mucinous, 41.8% serous, 3.7% endometrioid, 2.6% mixed and 0.2% Brenner cell (Table 1.11).129 LMP tumors of clear cell histology are rare, with only nine cases reported to date.130 Most patients with LMP tumors present before the age of 50 years, with a mean age of approximately 45 years, in contrast to approximately 52 years for patients with adenocarcinoma.123 The tumors occur more frequently in White women as opposed to those of African or Asian heritage.131 As with frankly invasive cancers, pregnancy, oral contraceptive use and breast feeding appear to exert a protective effect on

Ch 01 final.qxd

12/18/2006

15:08

Page 17

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

the risk for LMP tumors,132,133 although no significant correlation has been noted with age at menarche or menopause. Unlike invasive epithelial cancers, LMP tumors predominantly present in early stages, although meticulous staging may reveal occult metastases.134 Sutton135 reviewed 12 studies encompassing 946 patients with LMP tumors and noted that 80% presented with stage I disease. Other smaller studies have reported the incidence of stage II lesions to be 4–7%, and stage III lesions to be 11–14%.136,137 Stage IV

tumors are rare. Examining data from eight series with a total of 677 patients, only nine cases of stage IV disease (1.3%) were reported138–145 (Table 1.12). Therapy for LMP tumors is primarily surgical. In the usual case, total abdominal hysterectomy with bilateral salpingo-oophorectomy is performed, although patients desirous of future childbearing with disease limited to the ovary may be candidates for conservative management. Prognosis for patients with LMP tumors is generally excellent, particularly for patients with stage I disease. Barnhill et al.146 Table 1.11 Histologic types of tumors of low malignant potential from 352 patients. From reference 129 Histology

Percentage

Serous

41.8

Mucinous

51.7

Endometrioid

3.7

Mixed

2.6

Brenner cell

0.2

Clear cell

0

Figure 1.9 Serous tumor of low malignant potential. Cysts are lined by stratified cells with budding

a

b

Figure 1.10 Mucinous tumor of low malignant potential. (a) Intestinal type. Goblet cells present with nuclear stratification. (b) Endocervical type. Cells display budding and cytoplasmic mucin

17

Ch 01 final.qxd

12/18/2006

15:08

Page 18

SURGERY FOR OVARIAN CANCER

reviewed 27 studies encompassing 988 patients with stage I tumors. With follow-up ranging from 2.9 to 11.7 years, 99.3% of patients showed no evidence of disease. Only 0.7% died of disease. Trimble and Trimble126 examined 415 patients with 7 years’ median follow-up and noted 5-year survival rates of 99% for stage I, and 92% for stages II and III. There exists no definitive evidence that the use of adjuvant therapy after primary surgery improves survival for patients with LMP tumors. While survival is generally excellent regardless of stage, approximately 10% of patients will go on to develop recurrent disease. Aside from stage at presentation, attempts have been made to find risk factors for disease recurrence in order to identify a subset of patients who may benefit most from adjuvant therapy. Burks et al.147 initially defined micropapillary serous carcinoma as being characterized by a ‘filigree pattern of highly complex

Table 1.12 Stage at presentation of low-malignant-potential tumors of the ovary. From references 138–145 and 229 Stage

Percentage

I

80

II

4–7

III

11–14

IV

1.3

micropapillae arising directly from large, bulbous papillary structures’. Several authors have noted higher mortality rates among patients with micropapillary tumors, as well as those with invasive peritoneal implants.147–150 Primary peritoneal cancer

In 1959, Swerdlow first reported a tumor of the pelvic peritoneum histologically resembling a papillary serous carcinoma of the ovary.151 Primary peritoneal carcinoma (PPC) consists of a tumor with light microscopic, histochemical and immunohistochemical features indistinguishable from primary epithelial ovarian cancer.152–154 To clarify the difference between ovarian cancer and PPC, in 1993 the Gynecologic Oncology Group (GOG) specified that ‘extraovarian peritoneal serous papillary carcinoma’ should be histologically identical to epithelial ovarian cancer, with involvement of the ovary limited to less than 5 mm of cortical invasion (Table 1.13). By these criteria, approximately 10% of patients initially diagnosed with ovarian cancer may instead be reclassified as having PPC.155 Grossly, the tumor causes diffuse studding of the pelvic and abdominal peritoneum, with almost universal involvement of the omentum.153,156,157 Microscopically, PPCs most often display papillary serous differentiation (99% vs. 36% for epithelial ovarian cancer); however, other types have been reported including mucinous, endometrioid, clear cell

Table 1.13 Gynecologic Oncology Group classification of primary peritoneal carcinoma. From reference 230 1. Both ovaries must be either physiologically normal in size or enlarged by a benign process 2. Involvement in the extraovarian sites must be greater than the involvement on the surface of either ovary 3. Microscopically, the ovarian component must be one of the following: a. nonexistent b. confined to ovarian surface epithelium with no evidence of cortical invasion c. involving ovarian surface epithelium and underlying cortical stroma but with any given tumor size less than 5 × 5 mm d. tumor less than 5 × 5 mm within ovarian substance with or without surface disease 4. Histological and cytological characteristics of the tumor must be predominantly of the serous type that is similar or identical to ovarian serous papillary adenocarcinoma, any grade

18

Ch 01 final.qxd

12/18/2006

15:08

Page 19

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

and MMMT.155 As with ovarian cancer, psammoma bodies, desmoplasia and tumor necrosis are common. The grade of PPC is moderate to high grade in over 80% of cases.158 Despite the histologic similarities, differences exist between patients with epithelial ovarian cancer and PPC (Table 1.14). Patients with PPC are typically older, with mean age of approximately 61 years, compared to 56 years for epithelial ovarian cancer.153,154,156–168 Eltabbakh et al.164 reported on a series of 50 patients with PPC compared to 503 women with epithelial ovarian cancer. For the patients with PPC, the authors noted a statistically significantly higher mean age at diagnosis (63.8 vs. 55.0 years, respectively, p < 0.001), later menarche (13.3 vs. 12.8 years, p = 0.024) and less talc use (26% vs. 48.1%, p = 0.003). Molecular differences are also apparent. While epithelial ovarian cancer has been demonstrated to arise from a monoclonal proliferation of tumor cells,169–174 most investigators have reported at least some cases of PPC to be polyclonal in origin.175–178 PPC is staged and treated in an identical fashion to epithelial ovarian cancer. Although by definition there can be no stage I PPC, the vast majority of patients present in advanced stages. Chu et al.155 reviewed nine reported series of patients with PPC

Table 1.14 Comparison of ovarian and primary peritoneal cancer. Modified from reference 155 Epithelial ovarian carcinoma

Primary peritoneal carcinoma

56

61

14–92

4–92

12.8

13.3

Median parity

2

3–4

Talc exposure

48%

26%

27–30

12–24

Mean age at diagnosis (years) Age range (years) Mean age at menarche (years)

Median survival (months)

and noted that 2% presented in stage II, 73% in stage III and 25% in stage IV.

Germ cell tumors Germ cell tumors comprise approximately 20% of all ovarian tumors. Most germ cell tumors present in young women of reproductive age. In fact, for women under the age of 30, germ cell tumors are the most common ovarian neoplasm. Although some lesions are asymptomatic, most cause abdominal pain. Occasionally, torsion, rupture or intracapsular hemorrhage may lead to an acute abdomen. Fortunately, most germ cell tumors are benign, with only 2–3% composed of malignant primitive germ cell tumors. The majority of benign tumors are mature cystic teratomas. Of the malignant germ cell tumors, the most common are the dysgerminoma, endodermal sinus tumor and immature teratoma. The majority of these tumors are curable with appropriate chemotherapy and conservative surgery, which is especially important for young women desirous of future childbearing. Most germ cell tumors secrete chemical markers that may be measured in the blood and used to monitor response to therapy and potential disease recurrence. Teratomas

Teratomas are germ cell tumors composed of cells resembling products of the three embryonic layers of ectoderm, mesoderm and endoderm. The tissues present may be immature or mature, and occasionally may be monodermal with predominant representation of one particular tissue type. In general, teratomas comprised of immature elements are malignant and those composed of mature elements are benign, although any element within a benign tumor may itself undergo malignant transformation. Mature cystic teratoma The mature cystic teratoma,

or dermoid cyst, is a benign neoplasm accounting for nearly one-third of all ovarian tumors. In 10–15% of cases, these cysts may occur bilaterally. Most dermoid cysts are diagnosed in women of reproductive age, although they may be found at any age. The cyst

19

Ch 01 final.qxd

12/18/2006

15:08

Page 20

SURGERY FOR OVARIAN CANCER

lining is predominantly composed of epidermis and associated skin appendages. Grossly, the cysts are filled with sebaceous material and hair. Bone and teeth may also be apparent. Patients usually experience no symptoms, but may present with acute pain secondary to torsion of the affected ovary. Rupture of the cysts is uncommon, but is more likely to occur during pregnancy. When these tumors are diagnosed in women over the age of 40, malignant transformation of one of the cyst components is more likely, and may appear as a firm nodule on the wall of the cyst. Squamous cell carcinomas are the most common secondary tumor, occurring in about 1–2% of mature cystic teratomas. Treatment of dermoid cysts consists of either cystectomy or oophorectomy. Care should be taken to minimize spillage of cyst contents intraoperatively, as chemical peritonitis may ensue. Immature teratoma Immature teratomas are malig-

nant neoplasms composed of embryonic elements with admixed mature tissues. These tumors are bilateral in less than 5% of patients, although a benign dermoid may be found in the contralateral ovary in 10% of cases. Immature teratomas generally present before the age of 20. Grossly, the tumors may be either cystic or solid. Upon microscopic examination, the tumor is usually composed of immature neuroectoderm (Figure 1.11). Patients may also present with

Figure 1.11 Immature teratoma with neural elements

20

peritoneal implants of mature glial tissue, which does not affect prognosis. Tumors are graded from 1 to 3 based upon the amount of immature neural tissue present. Grade is correlated with prognosis and determines therapy. Initial treatment consists of removal of the affected ovary and thorough surgical staging. Only patients with stage IA grade 1 tumors may forgo adjuvant chemotherapy. Patients with stage IA disease have a 10-year survival of 70%.179 Monodermal tumors Struma ovarii is a teratoma com-

posed almost exclusively of thyroid tissue, which may be hormonally active and lead to clinical hyperthyroidism in about 30% of patients. Simple removal is usually curative, although struma may occasionally undergo malignant transformation. Metastatic disease may be treated with radioactive 131I therapy. Carcinoid tumors are rare in the ovary and histologically resemble carcinoid tumors of the gastrointestinal tract. These neoplasms are unilateral, and occur primarily in older women. Approximately 30% of patients may present with carcinoid syndrome. Dysgerminoma

The most common ovarian germ cell malignancy is the dysgerminoma, which accounts for 50% of all germ cell malignancies, but comprises only about 1% of all ovarian tumors. These lesions are grossly bilateral in 10–15% of cases, although occult disease may be present in the contralateral ovary in an additional 10% of cases. Occasionally, patients may present with an associated gonadoblastoma. Women are usually diagnosed in the first three decades of life, and sometimes during pregnancy. Grossly, the tumor appears lobulated and solid, with a pink or tan interior. Microscopically, the tumor is composed of large round clear cells, resembling primordial germ cells, which contain glycogen (Figure 1.12). Nuclei are central and contain prominent nucleoli. Tumor cells are either diffuse, or arranged in islands or cords, and are often infiltrated by lymphocytes. Up to 10% of dysgerminomas contain syncytiotrophoblast, which may secrete human chorionic gonadotropin (hCG). Lactate

Ch 01 final.qxd

12/18/2006

15:08

Page 21

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Figure 1.12 Dysgerminoma. Cells display clear cytoplasm and

Figure 1.13 Yolk sac tumor. Communicating spaces lined by

large round nuclei with prominent nucleoli

clear cells merge with solid areas

dehydrogenase (LDH) may also be detected and used as a serum tumor marker.180 Dysgerminomas have a tendency to lymphatic spread, and a thorough surgical staging procedure is indicated. Fortunately, two-thirds of patients present with stage I disease, and may be managed with unilateral salpingo-oophorectomy. Biopsy of the contralateral ovary is not necessary if it appears grossly normal. Cytoreduction is recommended for patients with advanced disease. Patients with stage IA disease have an excellent prognosis with 10-year survival in excess of 90%, and may forgo adjuvant chemotherapy. Although up to 25% of these patients experience disease recurrence, treatment with chemotherapy at that time may still be curative. Patients with advanced disease should undergo adjuvant treatment with either chemotherapy or radiation.

termed hepatoid yolk sac tumors. On gross examination, these tumors are typically yellow solid masses with focal areas of hemorrhage and necrosis. Microscopically, the reticular pattern is most common, displaying networks of spaces lined by primitive cells that secrete α-fetoprotein (AFP). Classic Schiller–Duval bodies may be present, and are composed of single rounded or elongated papillae containing a single central blood vessel protruding into spaces. In 1976, Kurman and Norris181 reported on 71 cases of endodermal sinus tumors and noted a 3-year actuarial survival of only 13%. However, with the routine use of surgical removal followed by multi-agent chemotherapy, survival has improved dramatically. Nawa et al.182 reviewed the outcomes of 47 patients with either pure yolk sac tumors or tumors with yolk sac components and noted survival to be 95% for stage I, 75% for stage II, 30% for stage III and 25% for stage IV.

Endodermal sinus tumors

Endodermal sinus tumors, also called yolk sac tumors, are the second most common type of primitive germ cell malignancy, accounting for about 20% of cases (Figure 1.13). These tumors are infrequently bilateral (less than 5%), and present at a median of 19 years of age.181 Endodermal sinus tumors are derived either from primitive gut tissue and designated glandular yolk sac tumors, or from primitive liver tissue and

Embryonic carcinoma

Despite being a common tumor of the testis, embryonic carcinoma is extremely rare as a pure tumor of the ovary, although it may present as part of a mixed germ cell tumor. These tumors comprise only 4% of malignant germ cell tumors. Histologically, epithelial cells resembling those of the embryonic germ disc are

21

Ch 01 final.qxd

12/18/2006

15:08

Page 22

SURGERY FOR OVARIAN CANCER

noted, growing in glandular, tubular, papillary and solid patterns. Often, scattered syncytiotrophoblast cells and foci of yolk sac differentiation are present. Accordingly, serum hCG and AFP may be useful markers to follow tumor response. Kurman and Norris183 reported on a series of 15 cases in 1976 and noted that 60% of patients presented with abnormal endocrine manifestations such as precocious puberty, irregular bleeding, amenorrhea and hirsutism. Patients presented at a median age of 15 years. Actuarial survival for all patients was only 39%, although stage I patients had a slightly better survival of 50%. Improved survival has been obtained using surgery followed by multi-agent chemotherapy.

nancy; or as a metastasis from choriocarcinoma in the uterus. Most tumors are diagnosed under the age of 20. Sexual precocity and irregular bleeding are common at presentation. Treatment consists of surgery followed by multi-agent chemotherapy. Mixed germ cell tumors

Mixed tumors compose 10% of malignant germ cell tumors. Most cases are a combination of dysgerminomatous and yolk sac components. Tumors are usually unilateral, although if dysgerminoma is present as a component, 10% of cases may be bilateral. Prognosis is related to the composition of the tumor and to the volume that the most malignant component occupies.

Polyembryoma

Sex cord–stromal tumors

Also more frequently noted in the testis, polyembryoma is another extremely rare ovarian tumor composed of embryoid bodies resembling normal early embryos. In most cases, polyembryoma is associated with other malignant germ cell elements, usually immature teratoma. Chapman et al.184 reviewed 11 reported cases. Age at presentation ranged from 3 to 44 years. Patients usually presented with pain, although menstrual abnormalities were noted in four patients and precocious puberty was noted in one patient. Of the 11 patients, eight presented with stage I disease and three with stage III disease. Half of stage I patients were alive 19–132 months after diagnosis. Two patients with stage III disease died within 2 months of diagnosis, and the third was treated with surgery followed by multi-agent chemotherapy and was noted to be alive and well 20 months after diagnosis.185

Tumors derived from the sex cords and mesenchyme of the embryonic ovary compose approximately 8% of all ovarian tumors,186 and are the third most common class of ovarian neoplasms. Tumors may contain granulosa cells, theca cells, stromal cells, Sertoli cells and Leydig cells in different combinations and in various states of differentiation. Many of these tumors are hormonally active and present in children and young women. Early-stage tumors may be treated with conservative surgery, although late-stage tumors are generally treated with surgery and adjuvant multi-agent chemotherapy.

Choriocarcinoma

Choriocarcinoma is also extremely rare in the pure form and is more commonly seen as part of a mixed tumor. These malignancies are composed of both cytotrophoblast and syncytiotrophoblast, and consequently produce hCG. Ovarian choriocarcinoma may develop through one of three mechanisms: as a tumor of ovarian germ cell origin; within an ovarian preg-

22

Granulosa cell tumors

Granulosa cell tumors represent about 6% of all ovarian cancers and over half of all malignant sex cord–stromal tumors. These malignancies can be diagnosed at any age, but usually present in women after the menopause.187 Tumors are more common in women of European or American background.188 The most common presenting symptoms associated with granulosa cell tumors are abnormal uterine bleeding and abdominal pain. These tumors are the most common estrogen-producing tumors; unopposed estrogen stimulation of the uterine lining may lead to endometrial hyperplasia. A synchronous endometrial carcinoma may be present in up to 13% of patients.189–192

Ch 01 final.qxd

12/18/2006

15:08

Page 23

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Patients may experience other symptoms of excess estrogen including breast tenderness and swelling. Occasionally, granulosa cell tumors may exert virilizing effects as well. The majority of these tumors are limited to the ovary at presentation. Granulosa cell tumors are divided into two subtypes: adult and juvenile. Adult granulosa cell tumors account for 95% of all granulosa cell tumors. Grossly, tumors may be cystic or solid and are typically large, with an average diameter of 12 cm. Cyst locules may be filled with fluid or clotted blood, and solid areas commonly display foci of hemorrhage. Cysts are prone to rupture. Tumors are bilateral in fewer than 5% of cases. Histologically, cells have light round or oval nuclei with frequent grooves, arranged in diffuse patterns (Figure 1.14). Well-differentiated tumors have follicular patterns, although a mixture of patterns may exist in a single tumor. Call–Exner bodies are a characteristic finding of the microfollicular pattern. While younger women may experience amenorrhea, postmenopausal women may have vaginal bleeding. Approximately 80–90% of patients present with stage I disease.193–195 For younger women desirous of future childbearing, stage IA tumors may be treated by removal of the affected ovary, with a wedge biopsy of the contralateral ovary to check for occult disease. Sampling of the endometrium is recommended to

Figure 1.14 Granulosa cell tumor, adult type. Diffuse pattern

exclude a synchronous primary tumor of the uterus. Stage IA disease in postmenopausal women, or those not desirous of future childbearing, may be treated with total abdominal hysterectomy and bilateral salpingo-oophorectomy, although some have advocated adjuvant chemotherapy with bleomycin, etoposide and platinum for postmenopausal women.196 Patients with cyst rupture should receive adjuvant chemotherapy. Advanced disease is treated with surgical cytoreduction and combination chemotherapy. These tumors have an indolent course, and 5-year survival rates have been reported to be 90%.191,194,195 Unfortunately, granulosa cell tumors may present with late recurrences more than 5 years after initial diagnosis. Hines et al.197 reviewed 16 cases with recurrences ranging from 13 to 37 years after initial diagnosis. Advanced stage, tumor rupture, tumor size, mitotic activity and nuclear atypia are associated with worsened prognosis. Juvenile granulosa cell tumors are usually found in women below the age of 30. Young et al.198 reported on 125 cases and noted that 44% presented in girls under the age of 10, and that only 3% occurred in women over the age of 30. Up to 80% of tumors presenting in prepubertal females are associated with sexual precocity.113 Juvenile granulosa cell tumors have been reported in association with Potter syndrome,199 Ollier disease200,201 and Maffucci syndrome.198,202 Over 95% of patients present with stage I disease.198,202 Bilaterality occurs in less than 5% of cases. Microscopically, juvenile granulosa cell tumors are arranged in solid nodules, with follicles containing mucicarminophilic fluid. Call–Exner bodies are rare and theca cells may often be present. Nuclei are dark and lack the grooves observed in adult granulosa cell tumors. Survival for stage I and II tumors has been reported to be 92%.198,202 Unfortunately, the rare late-stage juvenile granulosa cell tumors are aggressive with short time to relapse and death. Unlike their adult counterparts, most recurrence is diagnosed within 3 years of initial diagnosis. If thorough staging reveals disease limited to the ovary, conservative surgery may be conducted to preserve future fertility.

23

Ch 01 final.qxd

12/18/2006

15:08

Page 24

SURGERY FOR OVARIAN CANCER

Adjuvant platinum-based chemotherapy is recommended for patients with stage IC or more advanced tumors. Thecoma–fibroma

The ovarian stroma gives rise to fibroblasts and theca cells. Neoplasms in this category display a variety of histologies, ranging from those composed of pure fibroblasts, to those containing a combination of cells, to those composed primarily of theca cells. Tumors in this group are generally benign neoplasms, treated by simple surgical removal. Thecoma Thecomas account for only 1% of ovarian neoplasms. Most patients present between the ages of 50 and 70.203 Abnormal bleeding occurs in 60% of patients, owing to estrogen production by the tumor. Evans et al.191 noted that endometrial hyperplasia was noted in 37% and endometrial adenocarcinomas in 27% of evaluable patients. Tumors are rarely extraovarian. Bilaterality occurs in 2% of cases. Histologically, the tumor displays two different types. The typical thecoma displays sheets or nodules of large cells containing abundant pale cytoplasm, and a variable portion of spindle cells that produce collagen. Typical thecomas almost always produce estrogen. Luteinized thecomas display clusters of steroid-type cells with abundant eosinophilic cytoplasm.

a

Estrogenic manifestations are present in only 50% of patients, and androgenic changes in 11%.204 Fibroma and fibrosarcoma Fibromas are the most

common sex cord–stromal tumor, comprising approximately 4% of all ovarian neoplasms. These benign tumors may present at any age, but are less common before the age of 30. The average age at presentation is 48 years.113 On a cut surface, fibromas are usually hard, flat and white, although areas of hemorrhage and edema may be present. Microscopically, these tumors are composed of spindle cells and abundant collagen. Because the separation between fibromas and thecomas is often vague, the term ‘fibrothecoma’ is sometimes used (Figure 1.15).113 When they are greater than 10 cm, fibromas are associated with ascites in 40% of cases.205 The presence of ovarian fibroma, ascites and hydrothorax comprise the components of Meigs’ syndrome.206 Although the syndrome is well known, only 1% of patients with fibromas will present with these findings. In 1981, Prat and Scully reported on a group of 17 fibromas with malignant behavior.207 They defined the cellular fibroma as a neoplasm characterized by densely packed nuclei and scarce collagen with 1–3 mitoses per 10 high-power fields and minimal nuclear atypia. Approximately 10% of fibromas may be classified as cellular fibromas. These tumors display low

b

Figure 1.15 Fibrothecoma. (a) Gross appearance with hard, flat cut surface. (b) Microscopic appearance

24

Ch 01 final.qxd

12/18/2006

15:08

Page 25

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

malignant potential, and are prone to recurrence, even 10 years from initial surgery. Recurrence is more likely in cases of rupture, incomplete excision, or adherent disease. In contrast, the authors defined fibrosarcomas of the ovary as neoplasms typically displaying marked nuclear atypia and more than 3 mitoses per 10 high-power fields. While extremely rare, these tumors are frankly malignant neoplasms that usually present as large, unilateral growths with areas of hemorrhage and necrosis. Prognosis is poor, and patients should be treated postoperatively with bleomycin, etoposide and cisplatin chemotherapy.196

Sertoli cell tumor Sertoli cell tumors comprise only

5% of this class of neoplasm. Young and Scully214 reviewed 23 cases and noted the mean age at presentation to be 27 years of age. Approximately two-thirds of patients may show signs of excess estrogen, resulting in postmenopausal bleeding, irregular menses and precocious puberty, depending on the age of the patient. Grossly, tumors are usually solid with a lobulated appearance. Microscopically, Sertoli cell tumors are composed of tubules, either solid or hollow, formed of Sertoli cells containing lipid. All reported cases to date have presented as unilateral, stage I tumors, with only one reported death.214

Sclerosing stromal tumors

Sclerosing stromal cell tumors were initially reported by Chalvardjian and Scully in 1973.208 Histologically, these benign tumors display pseudolobules, composed of a combination of fibroblasts and lipid-containing lutein cells. These pseudolobules may be vascular, and are separated by hypocellular fibrous tissue. Grossly, tumors are unilateral, occurring more commonly on the right side.209 Tumors may reach 20 cm in size, but ascites is uncommon. Sclerosing stromal tumors display clinical features distinct from other tumors in the thecoma–fibroma group. Unlike with many other sex cord–stromal tumors, most women are diagnosed under the age of 30. Also, while androgen and estrogen production have been reported on occasion,210–212 most tumors are not hormonally active. Sertoli–stromal cell tumors

This group of tumors is composed of neoplasms containing Sertoli cells, Leydig cells, cells resembling rete epithelial cells and cells resembling fibroblasts. These tumors may display only one cell type, or be a combination of different types. In the past, these neoplasms had been termed ‘androblastomas’ because of their tendency to cause masculinization. However, in 1958, Morris and Scully proposed the term ‘Sertoli–Leydig cell tumor’ to prevent inclusion of diverse androgenproducing tumors of different origins, and because some tumors may present with either excess estrogen or absent hormonal manifestions.213

Sertoli–Leydig cell tumor Sertoli–Leydig cell tumors account for fewer than 1% of all ovarian neoplasms. On average, patients present in their mid-twenties. Young and Scully215 reported on a series of 207 cases and noted 75% of the patients to be under the age of 30, with only 10% diagnosed over the age of 50. Presenting complaints may include menstrual irregularities and abdominal pain. Up to half of patients may present with some signs of androgen excess, including amenorrhea, hirsutism, deepening of the voice, malepattern baldness, clitorimegaly, or breast atrophy. Rarely, these tumors may secrete estrogen as well. Grossly, these tumors range in size from 5 to 15 cm, and are yellow and lobulated in appearance. Most are solid, or solid with cystic components. Microscopically, Leydig cells are present, in addition to Sertoli cells in various states of differentiation (Figure 1.16). Young and Scully215 noted that over 97% of cases were stage I at diagnosis, with 1.5% stage II and 1% stage III. Fewer than 2% of cases were bilateral. Among 164 patients for whom follow-up was available, 18% of tumors displayed malignant behavior. Prognosis was correlated with stage and degree of tumor differentiation. Stage IA disease may be treated with unilateral oophorectomy,216 resulting in a 5year survival of 92%. More advanced stages require total abdominal hysterectomy, bilateral salpingooophorectomy and surgical cytoreduction. Adjuvant chemotherapy is recommended in patients with significant residual tumor.

25

Ch 01 final.qxd

12/18/2006

15:08

Page 26

SURGERY FOR OVARIAN CANCER

(range 16–65 years). Patients presented with hyperandrogenism, hyperestrogenism, or no endocrine symptoms. The neoplasms were unilateral, and usually small. Microscopically, granulosa cells with Call–Exner bodies are present along with tubules lined by Sertoli cells. Only one case to date has been reported to display malignant behavior.220 Steroid-cell tumors

with nests of eosinophilic Leydig cells and aggregates of darker

Although these neoplasms have previously been called ‘lipid’ or ‘lipoid’ tumors, in fact 40% do not possess intracellular fat. This group of neoplasms includes tumors composed of cells that resemble Leydig, lutein and adrenal cortical cells.

Sertoli cells

Stromal luteoma Stromal luteomas account for

Figure 1.16 Sertoli–Leydig cell tumor. Edematous stroma

Sex cord tumors with annular tubules

Sex cord tumors with annular tubules possess features of both granulosa cell tumors and Sertoli–Leydig cell tumors.217,218 About 40% of patients present with signs of estrogen excess, and most patients present with abnormal vaginal bleeding. One-third of these neoplasms occur in women with Peutz–Jeghers syndrome. A strong association with adenoma malignum of the cervix also exists. When associated with Peutz–Jeghers syndrome, sex cord tumors with annular tubules tend to be bilateral and multi-focal, with the primary ovarian lesion measuring less than 3 cm in size. Disease in this circumstance is benign. When no association with Peutz–Jeghers syndrome is apparent, tumors are usually large and unilateral, but approximately 20% are malignant. Microscopically, simple and complex ring-shaped tubules are present. Management of these tumors parallels that of granulosa cell tumors. Gynandroblastoma

Gynandroblastomas are extremely rare ovarian tumors. Martin-Jimenez et al.219 presented a review of 17 cases gleaned from the literature. The authors reported the mean age of presentation as 29.5 years

26

approximately 20% of tumors in this category. Hayes and Scully221 reported 25 cases and noted the mean age at presentation to be 58. Sixty per cent of their patients presented with symptoms of excess estrogen production, and 12% with signs of hyperandrogenism. Grossly, the authors described the tumors to be well circumscribed, less than 3 cm in size, and completely confined to the ovarian stroma. Microscopically these tumors are composed of steroid cells without crystals completely surrounded by a rim of ovarian stroma. Stromal hyperthecosis is usually present. These tumors are benign and treated with surgical excision. Leydig cell tumors Leydig cell tumors also account for

approximately 20% of steroid-cell tumors. Like stromal luteomas, the mean age at presentation is in the late fifties. Grossly, these tumors are unilateral and usually under 3 cm in size. Roth and Sternberg222 noted two different types – hilar, and the extremely rare non-hilar variety. Histologically, Leydig cell tumors are composed of lutein or adrenocortical cells, and display Reinke crystals (Figure 1.17). Most patients present with signs of virilization, although irregular bleeding and postmenopausal bleeding may occur in some patients. These lesions are benign and treated with simple surgical removal. Signs of hyperandrogenism usually regress, but half of patients may be left with residual symptoms.

Ch 01 final.qxd

12/18/2006

15:08

Page 27

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Gonadoblastoma

Figure 1.17 Leydig cell tumor with occasional eosinophilic crystals

Steroid cell tumor, not otherwise specified Approx-

imately 60% of steroid cell tumors lack the characteristics of Leydig cell tumors or stromal luteomas, and are termed steroid cell tumors not otherwise specified. Hayes and Scully223 examined 63 cases and noted a mean age at presentation of 43 years, although reported ages ranged from 2 to 80 years. About 40% of patients presented with virilization, and 6% with hyperestrogenism. Another 6% presented with Cushing’s syndrome. At the time of diagnosis, only four tumors were bilateral. Grossly, these tumors appeared solid and yellow, ranging in size from 1 to 45 cm (mean 8.5 cm). In the series by Hayes and Scully, 81% of patients had stage I disease, 6% had stage II and 13% had stage III or IV. About 25% of these lesions display malignant behavior. Poor prognosis is correlated with Cushing’s syndrome, tumors larger than 7 cm, the presence of more than two mitoses per 10 high-power fields, nuclear atypia and the presence of necrosis or hemorrhage. No clinically malignant cases have been reported to date in patients less than 20 years of age. After surgical removal, adjuvant chemotherapy may be considered for advanced stage patients or those with risk factors for malignant behavior.

Gonadoblastoma is a rare tumor first defined by Scully in 1953.224 These tumors occur in young patients, usually in those with dysgenetic gonads. Most patients are phenotypic females who may display virilization, and the remainder are phenotypic males with female internal secondary sexual organs. Occasionally, these neoplasms may occur in normal females. Over 80% of patients have sex-chromatin-negative nuclei, and the most common karyotypes are 46,XX and 45,XO/46,XY (mosaic).225 Scully reviewed 74 cases225 and described the tumors as yellow-brown or gray and calcified, ranging from microscopic to several centimeters in size. Microscopically, the tumor is composed of large germ cells and small cells resembling immature Sertoli cells. The cells grow in solid nodules that often contain areas of calcification. About 50% of patients with gonadoblastoma may have a synchronous dysgerminoma.

Metastatic tumors to the ovary Several different malignancies may metastasize to the ovaries, including gynecologic primaries and those from more distant sites. In fact, at the time of surgical exploration for an adnexal mass, about 5–10% of ovarian tumors are found to be of metastatic origin. Common gynecologic sources of metastases include endometrial carcinoma and fallopian tube carcinoma. The most common sources of non-gynecologic metastases include the gastrointestinal tract and the breast. Krukenberg tumors are metastatic ovarian lesions composed of mucin-filled signet-ring cells contained within cellular ovarian stroma (Figure 1.18). These tumors typically spread from a primary gastric carcinoma, but may also originate from other mucinous primaries such as breast or colon cancers. In some cases, difficulty may arise in distinguishing an ovarian metastasis from a synchronous mucinous primary tumor in cases of intestinal cancer. Grossly, intestinal metastases generally result in bilateral ovarian involvement. When characteristic histologic features cannot differentiate, immunohistochemical

27

Ch 01 final.qxd

12/18/2006

15:08

Page 28

SURGERY FOR OVARIAN CANCER

a

b

Figure 1.18 Krukenberg tumor (gastric metastasis). (a) Gross appearance. (b) Microscopic appearance with signet-ring cells

staining may be helpful. Cytokeratin-7 is usually positive in primary ovarian cancers, and cytokeratin-20 is usually positive in intestinal metastases.

6.

McGowan L, Parent L, Lednar W, et al. The woman at risk for developing ovarian cancer. Gynecol Oncol 1979; 7: 325–44

7.

Parazzini F, La Vecchia C, Negri E, et al. Menstrual factors and the risk of epithelial ovarian cancer. J Clin Epidemiol 1989; 42: 443–8

8.

Whittemore AS, Harris R, Itnyre J. Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case–control studies. IV. The pathogenesis of epithelial ovarian cancer. Collaborative Ovarian Cancer Group. Am J Epidemiol 1992; 136: 1212–20

9.

Rossing MA, Daling JR, Weiss NS, et al. Ovarian tumors in a cohort of infertile women. N Engl J Med 1994; 331: 771–6

10.

Casagrande JT, Louie EW, Pike MC, et al. Incessant ovulation and ovarian cancer. Lancet 1979; 2: 170–3

11.

Keal EE. Asbestosis and abdominal neoplasms. Lancet 1960; 2: 1211–16

12.

Newhouse ML, Pearson RM, Fullerton JM, et al. A case control study of carcinoma of the ovary. Br J Prev Soc Med 1977; 31: 148–53

13.

Graham J, Graham R. Ovarian cancer and asbestos. Environ Res 1967; 1: 115–28

14.

Henderson WJ, Hamilton TC, Griffiths K. Talc in normal and malignant ovarian tissue. Lancet 1979; 1: 499

15.

Longo DL, Young RC. Cosmetic talc and ovarian cancer. Lancet 1979; 2: 1011–12

ACKNOWLEDGMENTS We would like to thank Dr Geza Acs for providing gross and histologic photographs for this manuscript.

REFERENCES 1.

Tortolero-Luna G, Mitchell MF, Rhodes-Morris HE. Epidemiology and screening of ovarian cancer. Obstet Gynecol Clin North Am 1994; 21: 1–23

2.

Baker TR, Piver MS. Etiology, biology, and epidemiology of ovarian cancer. Semin Surg Oncol 1994; 10: 242–8

3.

Jemal A, Murray T, Samuels A, et al. Cancer statistics, 2003. CA Cancer J Clin 2003; 53: 5–26

4.

Yancik R, Ries LG, Yates JW. Ovarian cancer in the elderly: an analysis of surveillance, epidemiology, and end results program data. Am J Obstet Gynecol 1986; 154: 639–47

5.

28

Ries LAG, Kosary CL, Hankey BF, et al., eds. Cancer Statistics Review 1973–1994. National Institutes of Health Publication No. 97-2789. Bethesda, MD: National Cancer Institute, 1997

Ch 01 final.qxd

12/18/2006

15:08

Page 29

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

16.

Venter PF. Ovarian epithelial cancer and chemical carcinogenesis. Gynecol Oncol 1981; 12: 281–5

17.

Hartge P, Hoover R, Lesher LP, et al. Talc and ovarian cancer. J Am Med Assoc 1983; 250: 1844

18.

Roe FJ. Controversy: cosmetic talc and ovarian cancer. Lancet 1979; 2: 744

19.

Whittemore AS, Wu ML, Paffenbarger RS Jr, et al. Personal and environmental characteristics related to epithelial ovarian cancer. II. Exposures to talcum powder, tobacco, alcohol, and coffee. Am J Epidemiol 1988; 128: 1228–40

31.

Piver MS, Baker TR, Jishi MF, et al. Familial ovarian cancer: a report of 658 families from the Gilda Radner Familial Ovarian Cancer Registry 1981–1991. Cancer 1993; 71: 582–8

32.

Menczer J, Modan M, Ranon L, et al. Possible role of mumps virus in the etiology of ovarian cancer. Cancer 1979; 43: 1375–9

33.

Cramer DW, Welch WR, Cassells S, et al. Mumps, menarche, menopause, and ovarian cancer. Am J Obstet Gynecol 1983; 147: 1–6

34.

Rose DP, Boyar AP, Wynder EL. International comparisons of mortality rates for cancer of the breast, ovary, prostate, and colon, and per capita food consumption. Cancer 1986; 58: 2363–71

20.

Cramer DW, Welch WR, Scully RE, et al. Ovarian cancer and talc: a case-control study. Cancer 1982; 50: 372–6

21.

Cook LS, Kamb ML, Weiss NS. Perineal powder exposure and the risk of ovarian cancer. Am J Epidemiol 1997; 145: 459–65

35.

Cramer DW, Harlow BL, Willett WC, et al. Galactose consumption and metabolism in relation to the risk of ovarian cancer. Lancet 1989; 2: 66–71

22.

Lingeman CH. Environmental factors in the etiology of carcinoma of the human ovary: a review. Am J Ind Med 1983; 4: 365–79

36.

Stocks P. Cancer mortality in relation to national consumption of cigarettes, solid fuel, tea and coffee. Br J Cancer 1970; 24: 215–25

23.

Beral V, Fraser P, Chilvers C. Does pregnancy protect against ovarian cancer? Lancet 1978; 1: 1083–7

37.

24.

dos Santos Silva I, Swerdlow AJ. Recent trends in incidence of and mortality from breast, ovarian and endometrial cancers in England and Wales and their relation to changing fertility and oral contraceptive use. Br J Cancer 1995; 72: 485–92

Trichopoulos D, Papapostolou M, Polychronopoulou A. Coffee and ovarian cancer. Int J Cancer 1981; 28: 691–3

38.

Byers T, Marshall J, Graham S, et al. A case–control study of dietary and nondietary factors in ovarian cancer. J Natl Cancer Inst 1983; 71: 681–6

25.

Roush GC, Holford TR, Schymura MJ, et al. Cancer Risk and Incidence Trends: the Connecticut Perspective. New York: Hemisphere, 1987

39.

Cramer DW, Welch WR, Hutchison GB, et al. Dietary animal fat in relation to ovarian cancer risk. Obstet Gynecol 1984; 63: 833–8

26.

Buell P, Dunn JE. Cancer mortality among the Japanese Issei and Nisei of California. Cancer 1965; 18: 656–64

40.

Slattery ML, Schuman KL, West DW, et al. Nutrient intake and ovarian cancer. Am J Epidemiol 1989; 130: 497–502

27.

Parker SL, Davis KJ, Wingo PA, et al. Cancer statistics by race and ethnicity. CA Cancer J Clin 1998; 48: 31–48

41.

28.

Annegers JF, Strom H, Decker DG, et al. Ovarian cancer: incidence and case–control study. Cancer 1979; 43: 723–9

Fairfield KM, Hankinson SE, Rosner BA, et al. Risk of ovarian carcinoma and consumption of vitamins A, C, and E and specific carotenoids: a prospective analysis. Cancer 2001; 92: 2318–26

42.

Koch M, Gaedke H, Jenkins H. Family history of ovarian cancer patients: a case–control study. Int J Epidemiol 1989; 18: 782–5

43.

Hartge P, Schiffman MH, Hoover R, et al. A case–control study of epithelial ovarian cancer. Am J Obstet Gynecol 1989; 161: 10–16

29.

West RO. Epidemiologic study of malignancies of the ovaries. Cancer 1966; 19: 1001–7

30.

Greene MH, Clark JW, Blayney DW. The epidemiology of ovarian cancer. Semin Oncol 1984; 11: 209–26

29

Ch 01 final.qxd

12/18/2006

15:08

Page 30

SURGERY FOR OVARIAN CANCER

57.

Monteiro AN, August A, Hanafusa H. Evidence for a transcriptional activation function of BRCA1 C-terminal region. Proc Natl Acad Sci USA 1996; 93: 13595–9

58.

Milner J, Ponder B, Hughes-Davies L, et al. Transcriptional activation functions in BRCA2. Nature 1997; 386: 772–3

59.

Houlston RS, Collins A, Slack J, et al. Genetic epidemiology of ovarian cancer: segregation analysis. Ann Hum Genet 1991; 55: 291–9

Scully R, Anderson SF, Chao DM, et al. BRCA1 is a component of the RNA polymerase II holoenzyme. Proc Natl Acad Sci USA 1997; 94: 5605–10

60.

Bewtra C, Watson P, Conway T, et al. Hereditary ovarian cancer: a clinicopathological study. Int J Gynecol Pathol 1992; 11: 180–7

Gowen LC, Avrutskaya AV, Latour AM, et al. BRCA1 required for transcription-coupled repair of oxidative DNA damage. Science 1998; 281: 1009–12

61.

Narod SA, Madlensky L, Bradley L, et al. Hereditary and familial ovarian cancer in southern Ontario. Cancer 1994; 74: 2341–6

Takahashi H, Behbakht K, McGovern PE, et al. Mutation analysis of the BRCA1 gene in ovarian cancers. Cancer Res 1995; 55: 2998–3002

62.

Schildkraut JM, Thompson WD. Familial ovarian cancer: a population-based case–control study. Am J Epidemiol 1988; 128: 456–66

63.

Couch FJ, Hartmann LC. BRCA1 testing – advances and retreats. J Am Med Assoc 1998; 279: 955–7

64.

Malone KE, Daling JR, Thompson JD, et al. BRCA1 mutations and breast cancer in the general population: analyses in women before age 35 years and in women before age 45 years with first-degree family history. J Am Med Assoc 1998; 279: 922–9

65.

Easton DF, Bishop DT, Ford D, et al. Genetic linkage analysis in familial breast and ovarian cancer: results from 214 families. The Breast Cancer Linkage Consortium. Am J Hum Genet 1993; 52: 678–701

66.

Ford D, Easton DF, Stratton M, et al. Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families. The Breast Cancer Linkage Consortium. Am J Hum Genet 1998; 62: 676–89

67.

Struewing JP, Hartge P, Wacholder S, et al. The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med 1997; 336: 1401–8

68.

Rubin SC, Benjamin I, Behbakht K, et al. Clinical and pathological features of ovarian cancer in women with germ-line mutations of BRCA1. N Engl J Med 1996; 335: 1413–16

69.

Boyd J, Sonoda Y, Federici MG, et al. Clinicopathologic features of BRCA-linked and

44.

Kerlikowske K, Brown JS, Grady DG. Should women with familial ovarian cancer undergo prophylactic oophorectomy? Obstet Gynecol 1992; 80: 700–7

45.

Schildkraut JM, Risch N, Thompson WD. Evaluating genetic association among ovarian, breast, and endometrial cancer: evidence for a breast/ovarian cancer relationship. Am J Hum Genet 1989; 45: 521–9

46.

47.

48.

49.

Lynch HT, Lynch JF, Conway TA. Hereditary ovarian cancer. In: Rubin SC, Sutton GP, eds. Ovarian Cancer. New York: McGraw-Hill, 1993: 189–217

50.

Lynch HT, Harris RE, Guirgis HA, et al. Familial association of breast/ovarian carcinoma. Cancer 1978; 41: 1543–9

51.

Berchuck A, Cirisano F, Lancaster JM, et al. Role of BRCA1 mutation screening in the management of familial ovarian cancer. Am J Obstet Gynecol 1996; 175: 738–46

52.

Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 1994; 266: 66–71

53.

Wooster R, Bignell G, Lancaster J, et al. Identification of the breast cancer susceptibility gene BRCA2. Nature 1995; 378: 789–92

54.

Brugarolas J, Jacks T. Double indemnity: p53, BRCA and cancer. p53 mutation partially rescues developmental arrest in Brca1 and Brca2 null mice, suggesting a role for familial breast cancer genes in DNA damage repair. Nat Med 1997; 3: 721–2

55.

Zhang H, Tombline G, Weber BL. BRCA1, BRCA2, and DNA damage response: collision or collusion? Cell 1998; 92: 433–6

56.

Chapman MS, Verma IM. Transcriptional activation by BRCA1. Nature 1996; 382: 678–9

30

Ch 01 final.qxd

12/18/2006

15:08

Page 31

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

sporadic ovarian cancer. J Am Med Assoc 2000; 283: 2260–5 70.

Aaltonen LA, Peltomaki P, Leach FS, et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993; 260: 812–16

71.

Leach FS, Nicolaides NC, Papadopoulos N, et al. Mutations of a mutS homolog in hereditary nonpolyposis colorectal cancer. Cell 1993; 75: 1215–25

Genetics Studies Consortium. J Am Med Assoc 1997; 277: 915–19 82.

Narod SA, Risch H, Moslehi R, et al. Oral contraceptives and the risk of hereditary ovarian cancer. Hereditary Ovarian Cancer Clinical Study Group. N Engl J Med 1998; 339: 424–8

83.

Modan B, Hartge P, Hirsh-Yechezkel G, et al. Parity, oral contraceptives, and the risk of ovarian cancer among carriers and noncarriers of a BRCA1 or BRCA2 mutation. N Engl J Med 2001; 345: 235–40

84.

Ness RB, Grisso JA, Vergona R, et al. Oral contraceptives, other methods of contraception, and risk reduction for ovarian cancer. Epidemiology 2001; 12: 307–12

85.

Narod SA, Sun P, Ghadirian P, et al. Tubal ligation and risk of ovarian cancer in carriers of BRCA1 or BRCA2 mutations: a case–control study. Lancet 2001; 357: 1467–70

86.

Streuwing JP, Watson P, Easton EF, et al. Prophylactic oophorectomy in inherited breast/ovarian cancer families. J Natl Cancer Inst Monogr 1995: 33–5

87.

Statement of the American Society of Clinical Oncology: genetic testing for cancer susceptibility, Adopted on February 20, 1996. J Clin Oncol 1996; 14: 1730–6, discussion 7–40

Tobacman JK, Greene MH, Tucker MA, et al. Intraabdominal carcinomatosis after prophylactic oophorectomy in ovarian-cancer-prone families. Lancet 1982; 2: 795–7

88.

Mangan CE, Rubin SC, Rabin DS, et al. Lymph node nomenclature in gynecologic oncology. Gynecol Oncol 1986; 23: 222–6

77.

Mackey SE, Creasman WT. Ovarian cancer screening. J Clin Oncol 1995; 13: 783–93

89.

Keettel WC, Pixley EE. Diagnostic value of peritoneal washings. Clin Obstet Gynecol 1958; 1: 592

78.

NIH consensus conference. Ovarian cancer. Screening, treatment, and follow-up. NIH Consensus Development Panel on Ovarian Cancer. J Am Med Assoc 1995; 273: 491–7

90.

79.

van Nagell JR Jr, DePriest PD, Reedy MB, et al. The efficacy of transvaginal sonographic screening in asymptomatic women at risk for ovarian cancer. Gynecol Oncol 2000; 77: 350–6

Feldman GB, Knapp RC, Order SE, et al. The role of lymphatic obstruction in the formation of ascites in a murine ovarian carcinoma. Cancer Res 1972; 32: 1663–6

91.

Holm-Nielsen P. Pathogenesis of ascites in peritoneal carcinomatosis. Acta Pathol Microbiol Scand 1953; 33: 10–21

92.

Bergman F. Carcinoma of the ovary: a clinical–pathological study of 86 autopsied cases with special reference to mode of spread. Acta Obstet Gynecol Scand 1966; 45: 211

93.

Abrams LA, Spiro R, Goldstein N. Metastases in carcinoma. Cancer 1950; 3: 74

72.

73.

Bellacosa A, Genuardi M, Anti M, et al. Hereditary nonpolyposis colorectal cancer: review of clinical, molecular genetics, and counseling aspects. Am J Med Genet 1996; 62: 353–4 Lynch HT, de la Chapelle A. Genetic susceptibility to non-polyposis colorectal cancer. J Med Genet 1999; 36: 801–18

74.

Aarnio M, Mecklin JP, Aaltonen LA, et al. Life-time risk of different cancers in hereditary non-polyposis colorectal cancer (HNPCC) syndrome. Int J Cancer 1995; 64: 430–3

75.

Burke W, Daly M, Garber J, et al. Recommendations for follow-up care of individuals with an inherited predisposition to cancer. II. BRCA1 and BRCA2. Cancer Genetics Studies Consortium. J Am Med Assoc 1997; 277: 997–1003

76.

80.

van Nagell JR Jr, Gallion HH, Pavlik EJ, et al. Ovarian cancer screening. Cancer 1995; 76: 2086–91

81.

Burke W, Petersen G, Lynch P, et al. Recommendations for follow-up care of individuals with an inherited predisposition to cancer. I. Hereditary nonpolyposis colon cancer. Cancer

31

Ch 01 final.qxd

12/18/2006

15:08

Page 32

SURGERY FOR OVARIAN CANCER

94.

Burghardt E, Girardi F, Lahousen M, et al. Patterns of pelvic and paraaortic lymph node involvement in ovarian cancer. Gynecol Oncol 1991; 40: 103–6

106. Kline RC, Wharton JT, Atkinson EN, et al. Endometrioid carcinoma of the ovary: retrospective review of 145 cases. Gynecol Oncol 1990; 39: 337–46

95.

Moore DH. Primary surgical management of early epithelial ovarian carcinoma. In Rubin SC, Sutton GP, eds. Ovarian Cancer, 2nd edn. Philadelphia: Lippincott Williams & Wilkins, 2001: 201–18

107. Czernobilsky B, Silverman BB, Mikuta JJ. Endometrioid carcinoma of the ovary. A clinicopathologic study of 75 cases. Cancer 1970; 26: 1141–52

96.

Scully RE. Histological Typing of Ovarian Tumors, 2nd edn. New York: Springer-Verlag, 1999

108. Tidy J, Mason WP. Endometrioid carcinoma of the ovary: a retrospective study. Br J Obstet Gynaecol 1988; 95: 1165–9

97.

Robboy SJ, Duggan M, Kurman RJ. The female reproductive system. In Rubin E, Farber J, eds. Pathology, 2nd edn. Philadelphia: LB Lippincott, 1988

98.

Aure JC, Hoeg K, Kolstad P. Psammoma bodies in serous carcinoma of the ovary. A prognostic study. Am J Obstet Gynecol 1971; 109: 113–18

99.

Gilks CB, Bell DA, Scully RE. Serous psammocarcinoma of the ovary and peritoneum. Int J Gynecol Pathol 1990; 9: 110–21

100. Ronnett BM, Kurman RJ, Shmookler BM, et al. The morphologic spectrum of ovarian metastases of appendiceal adenocarcinomas: a clinicopathologic and immunohistochemical analysis of tumors often misinterpreted as primary ovarian tumors or metastatic tumors from other gastrointestinal sites. Am J Surg Pathol 1997; 21: 1144–5 101. Cuatrecasas M, Matias-Guiu X, Prat J. Synchronous mucinous tumors of the appendix and the ovary associated with pseudomyxoma peritonei. A clinicopathologic study of six cases with comparative analysis of c-Ki-ras mutations. Am J Surg Pathol 1996; 20: 739–46 102. Chuaqui RF, Zhuang Z, Emmert-Buck MR, et al. Genetic analysis of synchronous mucinous tumors of the ovary and appendix. Hum Pathol 1996; 27: 165–71 103. Long ME, Taylor HCJ. Endometrioid carcinoma of the ovary. Am J Obstet Gynecol 1964; 90: 936–50 104. Schueller EF, Kirol PM. Prognosis in endometrioid carcinoma of the ovary. Obstet Gynecol 1966; 27: 850–8 105. Stern RC, Dash R, Bentley RC, et al. Malignancy in endometriosis: frequency and comparison of ovarian and extraovarian types. Int J Gynecol Pathol 2001; 20: 133–9

32

109. Ulbright TM, Roth LM. Metastatic and independent cancers of the endometrium and ovary: a clinicopathologic study of 34 cases. Hum Pathol 1985; 16: 28–34 110. DiSilvestro PA, Gajewski WH, Ludwig ME, et al. Malignant mixed mesodermal tumors of the ovary. Obstet Gynecol 1995; 86: 780–2 111. Dehner LP, Norris HJ, Taylor HB. Carcinosarcomas and mixed mesodermal tumors of the ovary. Cancer 1971; 27: 207–6 112. Behbakht K, Randall TC, Benjamin I, et al. Clinical characteristics of clear cell carcinoma of the ovary. Gynecol Oncol 1998; 70: 255–8 113. Scully RE, Young RH, Clement PB. Tumors of the ovary, maldeveloped gonads, Fallopian tube, and broad ligament. In Atlas of Tumor Pathology, Third Series, Fascicle 23. Washington, DC: Armed Forces Institute of Pathology, 1999 114. Ohkawa K, Amasaki H, Terashima Y, et al. Clear cell carcinoma of the ovary: light and electron microscopic studies. Cancer 1977; 40: 3019–29 115. Eastwood J. Mesonephroid (clear cell) carcinoma of the ovary and endometrium: a comparative prospective clinico-pathological study and review of literature. Cancer 1978; 41: 1911–28 116. Brescia RJ, Dubin N, Demopoulos RI. Endometrioid and clear cell carcinoma of the ovary. Factors affecting survival. Int J Gynecol Pathol 1989; 8: 132–8 117. Hitti IF, Glasberg SS, Lubicz S. Clear cell carcinoma arising in extraovarian endometriosis: report of three cases and review of the literature. Gynecol Oncol 1990; 39: 314–20 118. Wheeler JE. Pathology of malignant ovarian epithelial tumors and miscellaneous and rare ovarian and paraovarian neoplasms. In Rubin SC, Sutton GP, eds.

Ch 01 final.qxd

12/18/2006

15:08

Page 33

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

Ovarian Cancer, 2nd edn. Philadelphia: Lippincott Williams & Wilkins, 2001: 99–134 119. Roth LM, Czernobilsky B. Ovarian Brenner tumors. II. Malignant. Cancer 1985; 56: 592–601

132. Trimble EL, Trimble CL. Epithelial ovarian tumors of low malignant potential. In Markman M, Hoskins WJ, eds. Cancer of the Ovary. New York: Raven Press, 1993:

120. Austin RM, Norris HJ. Malignant Brenner tumor and transitional cell carcinoma of the ovary: a comparison. Int J Gynecol Pathol 1987; 6: 29–39

133. Harlow BL, Weiss NS, Roth GJ, et al. Case–control study of borderline ovarian tumors: reproductive history and exposure to exogenous female hormones. Cancer Res 1988; 48: 5849–2

121. Lele SB, Piver MS, Barlow JJ, et al. Squamous cell carcinoma arising in ovarian endometriosis. Gynecol Oncol 1978; 6: 290–3

134. Yazigi R, Sandstad J, Munoz AK. Primary staging in ovarian tumors of low malignant potential. Gynecol Oncol 1988; 31: 402–8

122. Pins MR, Young RH, Daly WJ, et al. Primary squamous cell carcinoma of the ovary. Report of 37 cases. Am J Surg Pathol 1996; 20: 823–33

135. Sutton GP. Ovarian tumors of low malignant potential. In Rubin SC, Sutton GP, eds. Ovarian Cancer, 2nd edn. Philadelphia: Lippincott Williams & Wilkins, 2001: 399–417

123. Aure JC, Hoeg K, Kolstad P. Clinical and histologic studies of ovarian carcinoma. Long-term follow-up of 990 cases. Obstet Gynecol 1971; 37: 1–9 124. Classification and staging of malignant tumours in the female pelvis. Acta Obstet Gynecol Scand 1971; 50: 1–7 125. Serov SF, Scully RE, Sobin LH. International histological classification and staging of tumors. No. 9 Histologic typing of ovarian tumors. Geneva: World Health Organization, 1973 126. Trimble CL, Trimble EL. Management of epithelial ovarian tumors of low malignant potential. Gynecol Oncol 1994; 55: S52–61 127. Nayar R, Siriaunkgul S, Robbins KM, et al. Microinvasion in low malignant potential tumors of the ovary. Hum Pathol 1996; 27: 521–7

136. Buttini M, Nicklin JL, Crandon A. Low malignant potential ovarian tumours: a review of 175 consecutive cases. Aust NZ J Obstet Gynaecol 1997; 37: 100–3 137. Nakashima N, Nagasaka T, Oiwa N, et al. Ovarian epithelial tumors of borderline malignancy in Japan. Gynecol Oncol 1990; 38: 90–8 138. Chambers JT, Merino MJ, Kohorn EI, et al. Borderline ovarian tumors. Am J Obstet Gynecol 1988; 159: 1088–94 139. Julian CG, Woodruff JD. The biologic behavior of low-grade papillary serous carcinoma of the ovary. Obstet Gynecol 1972; 40: 860–7 140. Nikrui N. Survey of clinical behavior of patients with borderline epithelial tumors of the ovary. Gynecol Oncol 1981; 12: 107–19

128. Bell DA, Scully RE. Ovarian serous borderline tumors with stromal microinvasion: a report of 21 cases. Hum Pathol 1990; 21: 397–403

141. Nation JG, Krepart GV. Ovarian carcinoma of low malignant potential: staging and treatment. Am J Obstet Gynecol 1986; 154: 290–3

129. Chu CS, Randall TC, Mikuta JJ. Low malignant Potential Tumors Of The Ovary. Baltimore, MD: Post Graduate Obstetrics & Gynecology, 1998; 18: 1–6

142. Kliman L, Rome RM, Fortune DW. Low malignant potential tumors of the ovary: a study of 76 cases. Obstet Gynecol 1986; 68: 338–44

130. Norris HJ, Mount PM. Pathology of ovarian tumors of low malignant potential. In Rutledge F, Wharton T, Gershenson D, eds. Gynecologic Cancer: Diagnosis and Treatment Strategies. Austin: University of Texas Press, 1987: 171–92

143. Bostwick DG, Tazelaar HD, Ballon SC, et al. Ovarian epithelial tumors of borderline malignancy. A clinical and pathologic study of 109 cases. Cancer 1986; 58: 2052–65

131. Harlow BL, Weiss NS, Lofton S. Epidemiology of borderline ovarian tumors. J Natl Cancer Inst 1987; 78: 71–4

144. Fort MG, Pierce VK, Saigo PE, et al. Evidence for the efficacy of adjuvant therapy in epithelial ovarian tumors of low malignant potential. Gynecol Oncol 1989; 32: 269–72

33

Ch 01 final.qxd

12/18/2006

15:08

Page 34

SURGERY FOR OVARIAN CANCER

145. Barnhill D, Heller P, Brzozowski P, et al. Epithelial ovarian carcinoma of low malignant potential. Obstet Gynecol 1985; 65: 53–9 146. Barnhill DR, Kurman RJ, Brady MF, et al. Preliminary analysis of the behavior of stage I ovarian serous tumors of low malignant potential: a Gynecologic Oncology Group study. J Clin Oncol 1995; 13: 2752–6 147. Burks RT, Sherman ME, Kurman RJ. Micropapillary serous carcinoma of the ovary. A distinctive lowgrade carcinoma related to serous borderline tumors. Am J Surg Pathol 1996; 20: 1319–30 148. Seidman JD, Kurman RJ. Subclassification of serous borderline tumors of the ovary into benign and malignant types. A clinicopathologic study of 65 advanced stage cases. Am J Surg Pathol 1996; 20: 1331–45 149. Eichhorn JH, Bell DA, Young RH, et al. Ovarian serous borderline tumors with micropapillary and cribriform patterns: a study of 40 cases and comparison with 44 cases without these patterns. Am J Surg Pathol 1999; 23: 397–409 150. Seidman JD, Kurman RJ. Ovarian serous borderline tumors: a critical review of the literature with emphasis on prognostic indicators. Hum Pathol 2000; 31: 539–57 151. Swerdlow M. Mesothelioma of the pelvic peritoneum resembling papillary cystadenocarcinoma of the ovary. Am J Obstet Gynecol 1959; 77: 197 152. Raju U, Fine G, Greenawald KA, et al. Primary papillary serous neoplasia of the peritoneum: a clinicopathologic and ultrastructural study of eight cases. Hum Pathol 1989; 20: 426–36 153. Truong LD, Maccato ML, Awalt H, et al. Serous surface carcinoma of the peritoneum: a clinicopathologic study of 22 cases. Hum Pathol 1990; 21: 99–110 154. Lele SB, Piver MS, Matharu J, et al. Peritoneal papillary carcinoma. Gynecol Oncol 1988; 31: 315–20 155. Chu CS, Menzin AW, Leonard DG, et al. Primary peritoneal carcinoma: a review of the literature. Obstet Gynecol Surv 1999; 54: 323–35 156. Fromm GL, Gershenson DM, Silva EG. Papillary serous carcinoma of the peritoneum. Obstet Gynecol 1990; 75: 89–95 157. Killackey MA, Davis AR. Papillary serous carcinoma of the peritoneal surface: matched-case comparison

34

with papillary serous ovarian carcinoma. Gynecol Oncol 1993; 51: 171–4 158. Dalrymple JC, Bannatyne P, Russell P, et al. Extraovarian peritoneal serous papillary carcinoma. A clinicopathologic study of 31 cases. Cancer 1989; 64: 110–15 159. Altaras MM, Aviram R, Cohen I, et al. Primary peritoneal papillary serous adenocarcinoma: clinical and management aspects. Gynecol Oncol 1991; 40: 230–6 160. Zhou J, Iwasa Y, Konishi I, et al. Papillary serous carcinoma of the peritoneum in women. A clinicopathologic and immunohistochemical study. Cancer 1995; 76: 429–36 161. Kannerstein M, Churg J, McCaughey WT, et al. Papillary tumors of the peritoneum in women: mesothelioma or papillary carcinoma. Am J Obstet Gynecol 1977; 127: 306–14 162. Ransom DT, Patel SR, Keeney GL, et al. Papillary serous carcinoma of the peritoneum. A review of 33 cases treated with platin-based chemotherapy. Cancer 1990; 66: 1091–4 163. August CZ, Murad TM, Newton M. Multiple focal extraovarian serous carcinoma. Int J Gynecol Pathol 1985; 4: 11–23 164. Eltabbakh GH, Piver MS, Natarajan N, et al. Epidemiologic differences between women with extraovarian primary peritoneal carcinoma and women with epithelial ovarian cancer. Obstet Gynecol 1998; 91: 254–9 165. Ben-Baruch G, Sivan E, Moran O, et al. Primary peritoneal serous papillary carcinoma: a study of 25 cases and comparison with stage III–IV ovarian papillary serous carcinoma. Gynecol Oncol 1996; 60: 393–6 166. Dvoretsky PM, Richards KA, Angel C, et al. Distribution of disease at autopsy in 100 women with ovarian cancer. Hum Pathol 1988; 19: 57–63 167. Gooneratne S, Sassone M, Blaustein A, et al. Serous surface papillary carcinoma of the ovary: a clinicopathologic study of 16 cases. Int J Gynecol Pathol 1982; 1: 258–69 168. White PF, Merino MJ, Barwick KW. Serous surface papillary carcinoma of the ovary: a clinical, pathologic, ultrastructural, and immunohistochemical study of 11 cases. Pathol Annu 1985; 20: 403–18

Ch 01 final.qxd

12/18/2006

15:08

Page 35

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

169. Kupryjanczyk J, Thor AD, Beauchamp R, et al. Ovarian, peritoneal, and endometrial serous carcinoma: clonal origin of multifocal disease. Mod Pathol 1996; 9: 166–73

182. Nawa A, Obata N, Kikkawa F, et al. Prognostic factors of patients with yolk sac tumors of the ovary. Am J Obstet Gynecol 2001; 184: 1182–8

170. Pejovic T, Heim S, Mandahl N, et al. Bilateral ovarian carcinoma: cytogenetic evidence of unicentric origin. Int J Cancer 1991; 47: 358–61

183. Kurman RJ, Norris HJ. Embryonal carcinoma of the ovary: a clinicopathologic entity distinct from endodermal sinus tumor resembling embryonal carcinoma of the adult testis. Cancer 1976; 38: 2420–33

171. Mok CH, Tsao SW, Knapp RC, et al. Unifocal origin of advanced human epithelial ovarian cancers. Cancer Res 1992; 52: 5119–22

184. Chapman DC, Grover R, Schwartz PE. Conservative management of an ovarian polyembryoma. Obstet Gynecol 1994; 83: 879–82

172. Jacobs IJ, Kohler MF, Wiseman RW, et al. Clonal origin of epithelial ovarian carcinoma: analysis by loss of heterozygosity, p53 mutation, and X-chromosome inactivation. J Natl Cancer Inst 1992; 84: 1793–8

185. Tsukahara Y, Fukuta T, Yamada T, et al. Retroperitoneal giant tumor formed by migrating polyembryoma with numerous embryoid bodies from an ovarian mixed germ cell tumor. Gynecol Obstet Invest 1991; 31: 58–60

173. Li S, Han H, Resnik E, et al. Advanced ovarian carcinoma: molecular evidence of unifocal origin. Gynecol Oncol 1993; 51: 21–5 174. Tsao SW, Mok CH, Knapp RC, et al. Molecular genetic evidence of a unifocal origin for human serous ovarian carcinomas. Gynecol Oncol 1993; 48: 5–10 175. Muto MG, Welch WR, Mok SC, et al. Evidence for a multifocal origin of papillary serous carcinoma of the peritoneum. Cancer Res 1995; 55: 490–2 176. Schorge JO, Muto MG, Welch WR, et al. Molecular evidence for multifocal papillary serous carcinoma of the peritoneum in patients with germline BRCA1 mutations. J Natl Cancer Inst 1998; 90: 841–5 177. Bandera CA, Muto MG, Welch WR, et al. Genetic imbalance on chromosome 17 in papillary serous carcinoma of the peritoneum. Oncogene 1998; 16: 3455–9 178. Moll UM, Valea F, Chumas J. Role of p53 alteration in primary peritoneal carcinoma. Int J Gynecol Pathol 1997; 16: 156–62

186. Young RH, Scully RE. Ovarian sex cord-stromal tumours: recent advances and current status. Clin Obstet Gynaecol 1984; 11: 93–134 187. Schwartz PE, Smith JP. Treatment of ovarian stromal tumors. Am J Obstet Gynecol 1976; 125: 402–11 188. Ohel G, Kaneti H, Schenker JG. Granulosa cell tumors in Israel: a study of 172 cases. Gynecol Oncol 1983; 15: 278–86 189. Bennington JL, Ferguson BR, Haber SL. Incidence and relative frequency of benign and malignant ovarian neoplasms. Obstet Gynecol 1968; 32: 627–32 190. Fox H, Agrawal K, Langley FA. A clinicopathologic study of 92 cases of granulosa cell tumor of the ovary with special reference to the factors influencing prognosis. Cancer 1975; 35: 231–41 191. Evans AT 3rd, Gaffey TA, Malkasian GD Jr, et al. Clinicopathologic review of 118 granulosa and 82 theca cell tumors. Obstet Gynecol 1980; 55: 231–8

179. Kurman RJ, Norris HJ. Malignant germ cell tumors of the ovary. Hum Pathol 1977; 8: 551–64

192. Nakashima N, Young RH, Scully RE. Androgenic granulosa cell tumors of the ovary. A clinicopathologic analysis of 17 cases and review of the literature. Arch Pathol Lab Med 1984; 108: 786–91

180. Schwartz PE, Morris JM. Serum lactic dehydrogenase: a tumor marker for dysgerminoma. Obstet Gynecol 1988; 72: 511–15

193. Bjorkholm E, Silfversward C. Prognostic factors in granulosa-cell tumors. Gynecol Oncol 1981; 11: 261–74

181. Kurman RJ, Norris HJ. Endodermal sinus tumor of the ovary: a clinical and pathologic analysis of 71 cases. Cancer 1976; 38: 2404–19

194. Malmstrom H, Hogberg T, Risberg B, et al. Granulosa cell tumors of the ovary: prognostic factors and outcome. Gynecol Oncol 1994; 52: 50–5

35

Ch 01 final.qxd

12/18/2006

15:08

Page 36

SURGERY FOR OVARIAN CANCER

195. Piura B, Nemet D, Yanai-Inbar I, et al. Granulosa cell tumor of the ovary: a study of 18 cases. J Surg Oncol 1994; 55: 71–7

207. Prat J, Scully RE. Cellular fibromas and fibrosarcomas of the ovary: a comparative clinicopathologic analysis of seventeen cases. Cancer 1981; 47: 2663–70

196. Schwartz PE, Price FV, Snyder MK. Management of ovarian stromal tumors. In Rubin SC, Sutton GP, eds. Ovarian Cancer, 2nd edn. Philadelphia: Lippincott Williams & Wilkins, 2001: 383–98

208. Chalvardjian A, Scully RE. Sclerosing stromal tumors of the ovary. Cancer 1973; 31: 664–70

197. Hines JF, Khalifa MA, Moore JL, et al. Recurrent granulosa cell tumor of the ovary 37 years after initial diagnosis: a case report and review of the literature. Gynecol Oncol 1996; 60: 484–8 198. Young RH, Dickersin GR, Scully RE. Juvenile granulosa cell tumor of the ovary. A clinicopathological analysis of 125 cases. Am J Surg Pathol 1984; 8: 575–96 199. Roth LM, Nicholas TR, Ehrlich CE. Juvenile granulosa cell tumor: a clinicopathologic study of three cases with ultrastructural observations. Cancer 1979; 44: 2194–205 200. Tamimi HK, Bolen JW. Enchondromatosis (Ollier’s disease) and ovarian juvenile granulosa cell tumor. Cancer 1984; 53: 1605–8 201. Velasco-Oses A, Alonso-Alvaro A, Blanco-Pozo A, et al. Ollier’s disease associated with ovarian juvenile granulosa cell tumor. Cancer 1988; 62: 222–5 202. Lewis RJ, Ketcham AS. Maffucci’s syndrome: functional and neoplastic significance. Case report and review of the literature. J Bone Joint Surg Am 1973; 55: 1465–79 203. Bjorkholm E, Silfversward C. Theca-cell tumors. Clinical features and prognosis. Acta Radiol Oncol 1980; 19: 241–4

209. Marelli G, Carinelli S, Mariani A, et al. Sclerosing stromal tumor of the ovary. Report of eight cases and review of the literature. Eur J Obstet Gynecol Reprod Biol 1998; 76: 85–9 210. Ho Yuen B, Robertson DI, Clement PB, et al. Sclerosing stromal tumor of the ovary. Obstet Gynecol 1982; 60: 252–6 211. Quinn MA, Oster AO, Fortune D, et al. Sclerosing stromal tumour of the ovary case report with endocrine studies. Br J Obstet Gynaecol 1981; 88: 555–8 212. Damajanov I, Drobnjak P, Grizelj V, et al. Sclerosing stromal tumor of the ovary: a hormonal and ultrastructural analysis. Obstet Gynecol 1975; 45: 675–9 213. Morris JM, Scully RE. Endocrine Pathology of the Ovary. St Louis: Mosby, 1958 214. Young RH, Scully RE. Ovarian Sertoli cell tumors: a report of 10 cases. Int J Gynecol Pathol 1984; 2: 349–63 215. Young RH, Scully RE. Ovarian Sertoli–Leydig cell tumors. A clinicopathological analysis of 207 cases. Am J Surg Pathol 1985; 9: 543–69 216. Zaloudek C, Norris HJ. Sertoli–Leydig tumors of the ovary. A clinicopathologic study of 64 intermediate and poorly differentiated neoplasms. Am J Surg Pathol 1984; 8: 405–18 217. Hart WR, Kumar N, Crissman JD. Ovarian neoplasms resembling sex cord tumors with annular tubules. Cancer 1980; 45: 2352–63

204. Zhang J, Young RH, Arseneau J, et al. Ovarian stromal tumors containing lutein or Leydig cells (luteinized thecomas and stromal Leydig cell tumors) – a clinicopathological analysis of fifty cases. Int J Gynecol Pathol 1982; 1: 270–85

218. Tavassoli FA, Norris HJ. Sertoli tumors of the ovary. A clinicopathologic study of 28 cases with ultrastructural observations. Cancer 1980; 46: 2281–97

205. Samanth KK, Black WC 3rd. Benign ovarian stromal tumors associated with free peritoneal fluid. Am J Obstet Gynecol 1970; 107: 538–45

219. Martin-Jimenez A, Condom-Munro E, Valls-Porcel M, et al. Gynandroblastoma of the ovary. Review of the literature. J Gynecol Obstet Biol Reprod (Paris) 1994; 23: 391–4

206. Meigs JV, Cass JW. Fibroma of the ovary with ascites and hydrothorax with a report of seven cases. Am J Obstet Gynecol 1937; 107: 538

220. Novak ER. Gynandroblastoma of the ovary: review of 8 cases from the ovarian tumor registry. Obstet Gynecol 1967; 30: 709–15

36

Ch 01 final.qxd

12/18/2006

15:08

Page 37

EPIDEMIOLOGY, STAGING AND CLINICAL CHARACTERISTICS

221. Hayes MC, Scully RE. Stromal luteoma of the ovary: a clinicopathological analysis of 25 cases. Int J Gynecol Pathol 1987; 6: 313–21

International Federation Obstetrics, 1994; 22

of

Gynecology

and

222. Roth LM, Sternberg WH. Ovarian stromal tumors containing Leydig cells. II. Pure Leydig cell tumor, non-hilar type. Cancer 1973; 32: 952–60

227. Herbst AL. Neoplastic Diseases of the Ovary. In Mishell DR Jr, Stenchever MA, Droegemueller W, et al., eds. Comprehensive Gynecology, 3rd edn. St Louis: Mosby-Year Book, 1997: 901–44

223. Hayes MC, Scully RE. Ovarian steroid cell tumors (not otherwise specified). A clinicopathological analysis of 63 cases. Am J Surg Pathol 1987; 11: 835–45

228. Fine G, Clarke HD, Horn RC Jr. Mesonephroma of the ovary. A clinical, morphological, and histogenetic appraisal. Cancer 1973; 31: 398–410

224. Scully RE. Gonadoblastoma: a gonadal tumor related to dysgerminoma (seminoma) and capable of sex hormone production. Cancer 1953; 6: 455–63

229. Katzenstein AL, Mazur MT, Morgan TE, et al. Proliferative serous tumors of the ovary. Histologic features and prognosis. Am J Surg Pathol 1978; 2: 339–55

225. Scully RE. Gonadoblastoma. A review of 74 cases. Cancer 1970; 25: 1340–56 226. Pettersson F. Annual Report on the Results of Treatment in Gynecological Cancer. Stockholm:

230. Bloss JD, Liao SY, Buller RE, et al. Extraovarian peritoneal serous papillary carcinoma: a case–control retrospective comparison to papillary adenocarcinoma of the ovary. Gynecol Oncol 1993; 50: 347–51

37

Ch 02 final.qxd

12/18/2006

15:11

Page 39

CHAPTER 2

Preoperative preparation and surgical instrumentation Marcela G del Carmen, Robert E Bristow, Linda R Duska

INTRODUCTION While surgery is the cornerstone of ovarian cancer diagnosis and initial treatment, it is in fact a detailed process that begins well before the patient ever reaches the operating room. It is therefore incumbent upon the surgeon to lay the foundation for a safe and successful operation and recovery through appropriate diagnostic evaluation, optimization of co-morbid medical conditions, and having a working understanding of the contemporary surgical instrumentation utilized during ovarian cancer surgery.

PREOPERATIVE EVALUATION Risk assessment In preparation for surgery for a suspected ovarian cancer all patients should undergo a comprehensive his-

tory and physical examination with a bimanual rectovaginal pelvic examination. The review of symptoms and examination should focus specifically on those areas that may indicate a reduced capacity to tolerate major surgery or place the patient at elevated risk for postoperative complications. The five-tiered classification system of the American Society of Anesthesiologists (ASA) can provide a global assessment of the risk for adverse perioperative outcomes (Table 2.1). In a prospective study of 6301 general surgical patients, Wolters et al. found that increasing ASA classification (I–IV) was significantly associated with longer operative time, increased intraoperative blood loss, a higher likelihood of postoperative mechanical ventilation requirement, longer hospital stay and time in the intensive care unit, a higher incidence of bronchopulmonary infection and other pulmonary complications, cardiac complications (arrhythmia, myocardial infarction), more frequent

Table 2.1 American Society of Anesthesiologists (ASA) classification system ASA class

Description

1

Healthy patient

2

Mild systemic disease – no functional limitation

3

Severe systemic disease – definite functional limitation

4

Severe life-threatening systemic disease that is a constant threat to life

5

Moribund, unlikely to survive 24 h with or without operation

39

Ch 02 final.qxd

12/18/2006

15:11

Page 40

SURGERY FOR OVARIAN CANCER

wound infection, a higher risk of bowel anastomosis dehiscence and a higher mortality rate.1 After controlling for confounding factors (type and extent of operation, age, medical co-morbidities), ASA classification was an independent and significant predictor of the overall risk of postoperative complications, with ASA classes II–IV being associated with 1.6-fold, 2.2-fold and 4.3-fold increased risk, respectively, relative to ASA class I patients.1 The validity of the ASA classification system as a predictor of postoperative complications specifically in gynecologic oncology patients was examined by Giannice et al. in a case–control study of 323 patients aged 70 years and older undergoing either minor or major surgical procedures.2 Compared to patients classified as ASA I–II, those patients who were ASA III–V were significantly more likely to experience any postoperative complication (48% vs. 28%), two or more postoperative complications (13% vs. 9%), or severe (grade 3 or 4) postoperative complications (17% vs. 5%). In this study, there were no significant differences in operative time, median blood loss, or transfusion rate according to the ASA classification.

Laboratory and routine testing The recommended routine preoperative laboratory testing program prior to surgery for suspected ovarian cancer should include a minimum of a complete blood count (CBC), serum electrolytes, age-appropriate health screening studies, a chest radiograph, and electrocardiogram for women aged 50 years and above (Table 2.2). The CBC may reveal anemia, which may be acute (e.g. due to gastrointestinal losses from secondary tumor involvement) or chronic (due to chronic disease or underlying hemoglobinopathy) and require further evaluation (e.g. peripheral smear, reticulocyte count, serum iron studies, or hemoglobin electrophoresis) or blood product replacement prior to surgery. Derangements of the platelet count may also be seen with ovarian cancer. Thrombocytosis (platelet count > 400 000/ml) may be present in as many as 22% of patients undergoing primary surgery for epithelial ovarian cancer and has been associated

40

with advanced-stage disease, higher tumor grade, more frequent lymph node metastasis, a larger volume of ascites and a hypercoagulable state predisposing to pulmonary embolism compared to patients with normal platelet counts.3,4 Among patients with advanced-stage disease, thrombocytosis has also been implicated as a predictor of worse overall survival outcome.3,5 Rarely, thrombocytopenia in the setting of newly diagnosed ovarian cancer may indicate the presence of a chronic disseminated intravascular coagulopathy. Routine examination of serum electrolytes is recommended to identify abnormalities in fluid balance and renal function that may be present as co-morbid conditions or caused by advancedstage ovarian cancer. Significant hyponatremia (Na < 120 mEq/l), hyperkalemia (K > 5.0 mEq/l), hypokalemia (K < 3.3 mEq/l) and renal insufficiency (creatinine > 2.5 mg/ml) require further evaluation

Table 2.2 Preoperative testing program prior to ovarian cancer surgery Routine All patients Complete blood count Serum electrolytes Chest radiograph Pap smear* Age 50 and above Electrocardiogram Screening sigmoidoscopy/colonoscopy* Mammography† Indicated for specific personal history, symptoms, or physical findings Serum chemistries Coagulation studies Echocardiogram Exercise stress test Pulmonary function tests Sigmoidoscopy/colonoscopy/barium enema/upper gastrointestinal series Urinalysis *As per recommended screening guidelines; †baseline mammography at age 40; subsequent screening studies every 1–2 years between ages 40 and 50 years

Ch 02 final.qxd

12/18/2006

15:11

Page 41

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

and correction prior to surgery. In selected patients, a serum chemistry panel may identify significant disturbances in synthetic liver function (infectious or alcoholic hepatitis) or calcium balance (hypercalcemia of malignancy). A history of poorly controlled endocrinopathies (diabetes, hyper- or hypothyroidism, adrenal insufficiency) should also be specifically elicited in the preoperative evaluation, with additional testing and medical specialist consultation obtained as necessary to optimize the patient’s overall fitness for surgery. Coagulation studies (prothrombin time, activated partial thromboplastin time) are unnecessary as routine preoperative tests and should be obtained only if indicated by symptoms (easy bruisability or bleeding) or history (family member with bleeding diathesis). Similarly, a preoperative urinalysis is unlikely to yield clinically significant results unless the patient is symptomatic. A chest radiograph is indicated for all women with suspected ovarian cancer, irrespective of age, to look for the presence of a pleural effusion or pulmonary metastases, which may impair postoperative respiratory function. Selected patients may benefit from preoperative thoracentesis or chest tube placement to optimize ventilation during and after surgery. Although severe underlying pulmonary disease is uncommon in ovarian cancer patients, those with a history of longstanding smoking, chronic obstructive pulmonary disease, or severe asthma may benefit from preoperative pulmonary function testing. Routine electrocardiography should be performed in women aged 50 years and older and in patients with a significant cardiac history. Unrecognized cardiac conditions can have devastating consequences in the perioperative period. The preoperative evaluation should therefore detail any symptoms suggestive of ischemic heart disease, impaired cardiac reserve (reduced exercise tolerance, dyspnea, chest pain, orthopnea, or unexplained syncopal episodes) or peripheral vascular disease. Positive findings may require further evaluation with an echocardiogram, exercise stress test, or angiography prior to proceeding with major surgery. In these instances, a cardiologist

should be consulted for optimization of cardiac function and preoperative clearance. Screening sigmoidoscopy or colonoscopy should be obtained in women aged 50 years and older if not done within the previous 5 years. Directed evaluation of the colon (colonoscopy, barium enema) is indicated for patients with occult blood in the stool or evidence of large-bowel obstruction to exclude the presence of a primary colorectal malignancy. An upper gastrointestinal series or upper endoscopy may be necessary to investigate symptoms of a more proximal disturbance in the gastrointestinal tract (hematemesis, bilious emesis) on an individualized basis.

Tumor markers Serum tumor markers, although frequently obtained, are not a prerequisite to undertaking surgery for suspected ovarian cancer. Nevertheless, specific abnormalities can reaffirm the clinical suspicion in many cases and may occasionally suggest another site of primary malignancy in patients with an ovarian mass. CA-125 is an antigenic determinant on a high molecular weight glycoprotein recognized by the murine monoclonal antibody OC125.6 An arbitrary cut-off value of 35 U/ml has been set as the upper limit of normal for CA-125 in clinical practice; however, levels tend to be somewhat lower in postmenopausal women and in women who have undergone hysterectomy, such that lower cut-off levels (20–26 U/ml) have been proposed for these patients.7–9 The CA125 is not specific to ovarian cancer, however, as elevated levels may be seen with non-ovarian malignancies (e.g. endometrial cancer, pancreatic cancer) or benign conditions (endometriosis, leiomyomas, pelvic inflammatory disease, congestive heart failure).7,10,11 Nevertheless, approximately 80% of all patients with non-mucinous epithelial ovarian cancer will have an elevated CA-125 level (> 35 U/ml).7 This observation is stage-dependent, with only 50% of patients with stage I disease having levels of > 35 U/ml compared to approximately 90% of patients with advanced-stage epithelial ovarian cancer.12

41

Ch 02 final.qxd

12/18/2006

15:11

Page 42

SURGERY FOR OVARIAN CANCER

In general, a preoperative serum CA-125 level is recommended in patients undergoing surgery for suspected ovarian cancer, not so much for its diagnostic value, but rather to serve as a baseline level in the event that an ovarian cancer diagnosis is confirmed pathologically.13 Whether or not a preoperative CA-125 level can be used to predict the outcome of primary cytoreductive surgery for advanced-stage disease is controversial. Using a threshold level of 500 U/ml, Chi et al. reported a positive predictive value of 78% and a negative predictive value of 73% for optimal residual disease in 100 patients with stage III epithelial tumors.14 In contrast, Memarzadeh et al. reviewed 99 patients from UCLA Medical Center with advanced-stage disease and found that, while the ideal threshold level of 912 U/ml yielded a positive predictive value of 78% for optimal cytoreduction, the negative predictive value was just 31%.15 These mixed results suggest that an isolated serum CA-125 value is not sufficiently accurate to justify deferring an attempt at primary cytoreduction when advancedstage ovarian cancer is suspected. In addition to CA-125, ovarian cancer may be associated with abnormal elevations of other tumor markers that may have clinical utility in certain circumstances. Carcinoembryonic antigen (CEA), an oncofetal antigen found in small amounts in the adult colon, is elevated in 25–50% of ovarian cancer patients and may be a useful marker for tumors of mucinous histology. Elevated CEA values are associated with cancers of the colon and pancreas, as well as benign conditions of the liver, gastrointestinal tract and lung, and in smokers. CA19-9, part of the Lewis blood group antigens, is elevated in endometrial, gastrointestinal and lung cancers in addition to ovarian cancer, but is most useful in monitoring patients with pancreatic cancer. Mucinous ovarian cancers express CA19-9 more frequently than other histologic subtypes. CA27.29 is a tumor-associated antigen in the human milk fat globule membrane that is most commonly used to follow response to therapy in patients with metastatic breast cancer, but may be elevated in approximately 60% of ovarian cancers. In patients

42

≤ 30 years of age undergoing surgery for a solid ovarian mass, it is appropriate to obtain baseline serum tumor markers for ovarian germ cell tumors, specifically lactate dehydrogenase (dysgerminoma), α-fetoprotein (endodermal sinus tumor, embryonic carcinoma) and human chorionic gonadotropin (choriocarcinoma).

Advanced radiographic imaging techniques There are two primary diagnostic objectives in radiographic imaging of the patient with suspected ovarian cancer: the determination of the presence of malignancy in an ovarian mass and the evaluation of tumor extent. Contemporary imaging techniques for the evaluation of ovarian cancer include ultrasonography, computed tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET). Transabdominal and transvaginal ultrasound imaging are frequently used to characterize an ovarian mass, with findings of solid components, thick (≥ 3 mm) septations, papillary projections or excrescences and size ≥ 9 cm suggestive of malignancy. These features can also be appreciated on CT and MRI, and are likewise predictive of ovarian cancer. Kurtz et al. reporting for the Radiology Diagnostic Oncology Group, described results of a prospective study evaluating the accuracy of Doppler ultrasound, CT and MRI in 280 patients undergoing surgery for suspected ovarian cancer.16 The accuracy of all three imaging modalities for the overall detection of ovarian cancer was similar (91%); however, for diagnosis of malignancy in the ovary, the performance of MRI was statistically significantly better (91%) compared to Doppler ultrasound (78%), but not statistically different from that of CT (85%). For diagnosis of malignancy in the extraovarian pelvis and in the abdomen, MRI yielded higher accuracy (91% and 95%, respectively) compared to both ultrasound (88% and 90%, respectively) and CT (87% and 93%, respectively); however, these differences were not statistically significant. Both CT and MRI were found to have greater sensitivity than ultrasound for detecting abdominal

Ch 02 final.qxd

12/18/2006

15:11

Page 43

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

metastasis. Because of its cost-effectiveness, ultrasound imaging remains the procedure of choice for the initial evaluation of an ovarian mass. If further characterization of suspected abdominal disease is required to plan the operative approach, either CT or MRI is appropriate. CT is widely available, relatively easy to interpret and frequently the only imaging study obtained prior to ovarian cancer surgery. Recently, attention has focused on attempts to utilize preoperative CT imaging to predict the outcome of primary cytoreductive surgery for patients with advanced-stage disease. In 1993, Nelson et al. identified CT criteria of omental attachment to the spleen, large-volume tumor (> 2 cm) on the diaphragm, liver parenchyma, pleura, bowel mesentery, involvement of the gall bladder fossa, or suprarenal para-aortic adenopathy that predicted optimal cytoreduction with a sensitivity of 92%.17 In this study, however, the positive predictive value was just 67%, indicating that one out of every three patients predicted to be unresectable would, in fact, be optimally debulked if submitted to surgical exploration. Subsequent studies have examined various combinations of similar ‘resectability’ criteria with only modest improvements in predictive accuracy.18–20 Although continued advances in CT technology provide for increasingly better image resolution, the inability to exclude patients reliably from an attempt at cytoreductive surgery is troublesome. Based on the currently available data it seems reasonable to conclude that the operability of advanced ovarian cancer should be determined on an individual basis, with perhaps the greatest consideration being given to the skill and preparedness of the surgeon to put forth a maximum effort. PET uses radiopharmaceuticals labeled with positron-emitting isotopes such as 11C, 13N, 15O and 18F. 2-[18F]fluoro-2-deoxy-D-glucose (2-[18F] FDG), a glucose analog, is commonly used in oncologic imaging and takes advantage of the high glycolytic rate characteristic of malignant tissue compared with corresponding normal tissue. A variety of clinical applications of PET imaging of ovarian cancer have been

studied in recent years. While PET imaging has shown reliable accuracy for identifying primary or recurrent ovarian cancer metastases of 1 cm or larger, the utility of PET for evaluating an adnexal mass is less clear.21–23 Rieber et al. evaluated the ability of PET to characterize suspicious adnexal masses detected by ultrasonography in 103 patients, 12 of whom ultimately had pathologically confirmed ovarian malignancies.24 While PET imaging demonstrated an overall accuracy of 76% for detecting ovarian malignancy, the sensitivity was just 58% compared to 92% sensitivity for ultrasonography. Fenchel et al. studied PET performance in 99 patients with an incidentally detected adnexal mass and reported that the falsenegative rate for stage I ovarian cancers or borderline tumors was 71%.25 Taken together, these data indicate that PET imaging has insufficient sensitivity for detecting early-stage ovarian cancer. On the other hand, PET appears to be a sensitive technique for detecting recurrent ovarian cancer, particularly when combined with CT in patients with clinically occult suspected recurrence.26

PREOPERATIVE PREPARATION Preoperative nutritional supplementation The majority of ovarian cancer patients will not have a significant component of malnutrition at the time of initial diagnosis. Nevertheless, selected patients with advanced-stage disease involving the intestinal tract or producing massive ascites will present with either a functional or a mechanical bowel obstruction and associated nutritional compromise. A thorough inquiry regarding normal body weight, oral intake and gastrointestinal symptoms (e.g. nausea, vomiting, hematochezia, change in stool caliber) should be conducted. A 5% or more weight loss in a period of a month, a 10% weight loss in 6 months, or a current weight < 85% of the ideal body weight may be indicative of significant malnutrition.20 A targeted approach to laboratory evaluation can also provide important information suggestive of malnutrition. This includes

43

Ch 02 final.qxd

12/18/2006

15:11

Page 44

SURGERY FOR OVARIAN CANCER

a serum albumin level of < 3.0 g/dl, anergy to skin testing (e.g. Candida), a serum transferrin level of < 200 mg/dl and a total lymphocyte count of < 1200 cells/µl.27 Establishing the diagnosis of malnutrition is not difficult. In contrast, determining whether or not preoperative optimization of the patient’s nutritional status will significantly reduce the likelihood of postoperative morbidity has been more problematic. The most definitive study on the subject was published in 1991 by the Veteran Affairs Total Parenteral Nutrition Cooperative Study Group, who reported on perioperative total parenteral nutrition in 395 malnourished patients undergoing laparotomy or noncardiac thoracotomy.28 In this study, patients were randomized to receive either parenteral nutrition for 7–15 days preoperatively and 3 days postoperatively or no parenteral nutritional supplementation whatsoever (control group). Overall, both groups experienced similar rates of complications and postoperative mortality; however, patients receiving parenteral nutrition had significantly more infectious complications (14.1%) compared to control patients (6.4%). In a subset analysis, it was determined that severely malnourished patients who received parenteral nutrition had fewer non-infectious complications (5%) than controls (43%), without a concomitant increase of infectious morbidity. The authors concluded that the use of preoperative parenteral nutrition should be confined to patients who are severely malnourished. More recently, Heyland et al. reported results of a meta-analysis of 26 randomized clinical trials of parenteral nutrition in 2211 patients who were either critically ill or undergoing surgery.29 Analysis of those studies reporting surgical patients revealed that perioperative parenteral nutrition was associated with significantly lower rates of complications (relative risk 0.76, 95% confidence interval 0.48–1.00) but no difference in mortality rates (relative risk 0.91, 95% confidence interval 0.68–1.21). Contemporary indications for perioperative parenteral nutritional supplementation have been sum-

44

marized by Buzby as follows: first, preoperative parenteral nutrition should be considered in the most severely malnourished surgical candidates if an operative delay is not contraindicated and is not indicated for patients with only mild to moderate malnutrition; second, postoperative parenteral nutrition should be considered for malnourished surgical patients when oral or enteral feeding is not anticipated for 7–10 days postoperatively in previously well-nourished patients or 5–7 days in previously malnourished or critically ill patients.30 It has also been suggested that, if parenteral nutrition is administered perioperatively, it should be continued for a minimum of 7 days, as the risk of complications (e.g. electrolyte abnormalities, infection) outweigh any potential benefit if administered for shorter time intervals.31

Antibiotic prophylaxis As with most patients undergoing any major gynecologic operation, surgical site infection is a leading cause of postoperative morbidity in ovarian cancer surgical patients as well.32 Risk factors associated with an increased likelihood of developing a postoperative surgical site infection include obesity, diabetes, other sites of concomitant infection, a prolonged surgical time, excessive blood loss and prolonged hospitalization. Even in the absence of risk factors, however, prophylactic antibiotics are indicated for patients with a suspected diagnosis of ovarian cancer, as the procedure is likely to include abdominal hysterectomy and possible surgery on the intestinal tract. Although the use of prophylactic antibiotics has not been studied specifically in ovarian cancer patients, analogous data were presented by Mittendorf et al. in a meta-analysis of 25 randomized studies comparing the use of prophylactic antibiotic use versus no antibiotic prophylaxis in patients undergoing an abdominal hysterectomy.33 This large study documented a significant decrease in the incidence of surgical site infections in those patients who received antibiotics prior to undergoing hysterectomy. For maximum reduction of the risk of postoperative infection, antibiotics should

Ch 02 final.qxd

12/18/2006

15:11

Page 45

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

be administered approximately 30 min prior to incision but no earlier than 2 h prior to incision.34 Acceptable antibiotic choices include: cephazolin 1 g, cefotetan 1 g to 2 g, clindamycin 800 mg, and doxycycline 100 mg. Patients at risk of developing bacterial endocarditis require extended prophylactic antibiotic coverage (Table 2.3). These patients include women with prosthetic cardiac valves, a history of bacterial endocarditis in the past, complex cyanotic congenital cardiac disease, acquired valvular dysfunction, mitral valve prolapse with regurgitation and hypertrophic cardiomyopathy.

Bowel preparation Because ovarian cancer surgery carries the possibility of bowel resection or injury, preoperative bowel preparation is generally recommended in all such cases. The potential benefits of bowel preparation for ovarian cancer patients are largely extrapolated from data in the colorectal surgical literature. It is commonly accepted that surgical breach in the integrity of the intestinal tract, and particularly the colon, is associated with an increase in the risk of complications such as wound infection and intra-abdominal abscess. In an effort to reduce infection-related complications from intestinal surgery, both mechanical and antibiotic methods of bowel preparation have been recommended. Mechanical bowel preparation is thought to reduce fecal load, and thereby the risk of bacterial contamination, and generally includes the

use of oral cathartic agents and colonic irrigation (i.e. enema). Recently, the clinical value of a mechanical bowel preparation has been called into question by several prospective randomized trials. Brownson et al. reported that, among 179 patients undergoing bowel resection and anastomosis randomized to either mechanical preparation or no preparation, mechanical bowel preparation was associated with higher rates of anastomotic leak and intra-abdominal infection.35 Several other randomized trials have found no significant difference in the incidence of anastomotic leak between patients who have undergone mechanical bowel preparation and those who have not.36,37 In addition, Zmora et al. prospectively randomized 380 patients undergoing elective colon and rectal resections to undergo a mechanical bowel preparation or not prior to surgery.38 All patients received preoperative oral and intravenous antibiotics, and the two groups were well matched with respect to demographics, surgical procedures and duration of antibiotic therapy. The results of this study showed no significant difference between the two groups in terms of the incidence of wound infection, clinically apparent anastomotic leak, or intra-abdominal abscess. The authors concluded that elective colon and rectal surgery could be safely performed without mechanical bowel preparation. Although mechanical bowel preparation may reduce the amount of feces in the colon, the problem of persistent colonic bacteria provides the rationale

Table 2.3 Preoperative antibiotic prophylaxis for patients at risk for bacterial endocarditis

Antibiotic Ampicillin plus gentamicin

Regimen Ampicillin, 2 g IM/IV plus gentamicin, 1.5 mg/kg IM/IV(not to exceed 120 mg) give within 30 min of starting procedure Ampicillin, 1 g IM/IV, or amoxicillin, 1 g orally, 6 h later

Penicillin-allergic patients: vancomycin plus gentamicin

Vancomycin, 1 g IV over 1–2 h, plus gentamicin, 1.5 mg/kg IM/IV (not to exceed 120 mg) complete injection/infusion within 30 min of starting the procedure

45

Ch 02 final.qxd

12/18/2006

15:11

Page 46

SURGERY FOR OVARIAN CANCER

for using either non-absorbable or minimally absorbable oral antibiotics, in addition to routine intravenous antibiotics, to reduce the bacterial load. Theoretically, if the bowel is entered intraoperatively, there will be less chance of contamination. Multiple studies have been conducted, with conflicting results. In a randomized trial of 350 patients undergoing colon and rectal resections, Coppa and Eng demonstrated that the wound infection rate was significantly reduced when oral neomycin and erythromycin were added to mechanical bowel preparation and intravenous antibiotics, especially for prolonged procedures (> 215 min) and rectal resections.39 Such findings have been supported by additional randomized studies in the surgical literature.40 Interestingly, others have demonstrated no difference between oral antibiotics, intravenous antibiotics and a combination of the two in preventing septic complications or wound infections in patients undergoing elective colorectal surgery.41 Amidst these conflicting data, a large survey of North American colorectal surgeons was conducted by Nichols et al., who showed that most surgeons (86.5%) add oral and parenteral antibiotics to a mechanical preparation regimen of either oral polyethylene glycol solution or sodium phosphate solution with or without bisacodyl.42 Oral neomycin and erythromycin or metronidazole were the antibiotics most commonly used. This regimen seems reasonable for patients undergoing ovarian cancer surgery as well. Prophylactic parenteral antibiotics should be given at the time of anesthesia induction and repeated if the procedure exceeds two half-lives of the antibiotic used, fecal contamination occurs, or rectal surgery is performed. For single-agent parenteral antibiotic prophylaxis, second-generation cephalosporins are recommended (e.g. cefotetan, 2 g).

Thromboembolic prophylaxis The incidence of deep vein thrombosis (DVT) in women undergoing a pelvic surgical procedure lasting

46

longer than 30 min is 30%.43 Virtually all patients undergoing a surgical procedure for suspected ovarian cancer should be considered to be at moderate or high risk of developing a postoperative DVT. Consequently, some form of thromboembolic prophylaxis is warranted for the majority of cases, as DVT prophylaxis has been documented to prevent the occurrence of a fatal pulmonary embolism. Either pharmacologic or mechanical methods of prophylaxis can be used with comparable efficacy. Low-dose unfractionated heparin (5000 units) given subcutaneously 1–2 h before surgery and every 8 h thereafter has been shown to be effective in preventing DVT formation in a randomized trial of patients undergoing gynecologic oncology surgery.44 Low-molecular-weight heparin (LMWH) (enoxaparin, 40 mg once a day) has been documented to be as efficacious in DVT prophylaxis as low-dose unfractionated heparin and may have a lower incidence of bleeding complications.45 Dalteparin, another LMWH agent, can also be given once a day at a dose of 2500 units. Mechanical methods of DVT prophylaxis include graduated compression stockings and intermittent pneumatic compression devices, which have been shown to decrease the risk of DVT in moderate-risk patients. However, the use of graduated compression stockings alone has not been sufficiently studied in high-risk patients to support their use for DVT prophylaxis in this setting. Pneumatic compression devices must be in place at induction of anesthesia and continued until the patient is fully ambulatory. When used in this fashion, pneumatic compression appears to be effective in reducing DVT risk among medium- and high-risk patients.44 For patients undergoing surgery for suspected ovarian cancer, either pharmacologic or mechanical DVT prophylaxis is appropriate. Considering the risk of postoperative bleeding complications associated with heparin therapy, however, pneumatic compression devices, with or without graduated compression stockings, would seem to have a superior risk–benefit

Ch 02 final.qxd

12/18/2006

15:11

Page 47

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

profile. In patients at extremely high risk for postoperative DVT (e.g. history of a postoperative thromboembolic event), combined therapy with both methods may be considered, although there are currently no data that support an additive benefit over either method alone.

SURGICAL INSTRUMENTATION Retractors Usually, surgery for ovarian cancer is performed through a midline abdominal incision and requires access to both pelvic and abdominal structures, often simultaneously. A self-retaining retractor is essential to optimizing exposure, maximizing patient safety and reducing surgeon fatigue. Of the available models of self-retaining retractors, those with a fixed arm attaching the retractor ring to the operating table are best suited to ovarian cancer surgery. The Bookwalter retractor is the standard self-retaining fixed-ring retractor and is versatile enough to be adapted to a

variety of operative requirements (Figure 2.1). The retractor clips that attach the blades to the ring allow for two-dimensional adjustments of the blade position in relation to the surgical field. The oval ring of the Bookwalter is most commonly used for ovarian cancer surgery, but circular and hinged rings are also available, depending on the exposure needed. For example, the hinged ring can be used to surgical advantage when operating in the upper abdomen (e.g. diaphragm, liver, spleen) by increasing the angulation of the retractor blade to provide more pronounced ventral displacement of the costal margin, improving exposure. The Omni retractor has two adjustable ‘boomerang’ shaped arms, rather than a ring, that are attached to a fixed post. Each arm can be moved in three dimensions, and finer modifications in exposure can be achieved with the adjustable retractor blades. The Omni retractor is especially helpful when operating on obese patients, since the extent of lateral retraction is not limited by the width of a retractor ring (e.g. Bookwalter). Non-fixed self-retaining retractors, such as the Balfour and O’Connnor– O’Sullivan, can also be used, but are more limited in their field of exposure and are less steady than the fixed models, since they are stabilized only by creating pressure on the opposing sides of the abdominal wall incision. With any self-retaining retractor, the surgeon must exercise particular attention when placing the blades along the lateral abdominal wall so as not to compress the psoas muscle and traumatize the underlying femoral nerve.

ELECTROSURGERY The electrosurgical unit

Figure 2.1 Bookwalter self-retaining retractor. The fixed arm attaches to the operating room bed. The retractor blades can be adjusted in two dimensions in relation to the surgical field, or rotated around the oval ring to optimize exposure. Photograph courtesy of Aesculap Incorporated, Center Vallay, PA

The electrosurgical unit (ESU), electrocautery, or Bovie (Valleylab, Boulder, CO) consists of a generator and electrodes and is probably the most commonly used instrument in ovarian cancer surgery. The ESU can be configured as either a unipolar or a bipolar device that generates an alternating current of variable frequency. The unipolar device is versatile and

47

Ch 02 final.qxd

12/18/2006

15:11

Page 48

SURGERY FOR OVARIAN CANCER

can be used for both cutting and cauterizing tissue. With cutting current, a continuous high-frequency flow leads to a rapid buildup of heat and vaporization of intracellular water, resulting in local tissue disintegration without a significant coagulative effect but with minimal lateral heat transfer. In contrast, coagulation mode employs an interrupted current of lower energy, which leads to a slower heating of intracellular water, increasing the resistance to flow and producing a more pronounced coagulative effect on small blood vessels. Often a combination (or blended) current produces the most satisfactory tissue effect. Most gynecologic oncology surgeons utilize the unipolar ESU. Generally, the lowest effective generator settings should be used in order to avoid excessive thermal damage to surrounding tissues. Customary settings for a blended current range from 30 to 40 W for coagulation and from 30 to 50 W for cutting. The bipolar ESU employs a dual paddle design that conducts current to produce a tissue coagulating effect. Recently, a group of surgical instruments referred to as ‘vessel sealers’ have demonstrated clinical utility by simultaneously cauterizing a tissue pedicle and cutting it with a self-contained surgical knife (e.g. Ligasure®, Valleylab, Boulder, CO, Figure 2.2). The bipolar current generated reforms the collagen in vessel walls and connective tissue, producing a permanent seal that can effectively cauterize vessels up to 7 mm in diameter. Vessel sealers can be adapted for both laparoscopic and open applications and are par-

ticularly useful for controlling vascular pedicles in areas that are difficult to reach.

Argon beam coagulator As compared to the standard ESU, which conducts current through air, the argon beam coagulator (ABC) (Conmed Corp, Utica, NY), conducts radiofrequency current to the target tissue through a coaxial stream of inert argon gas that is automatically regulated. ABC power settings range from 70 to 150 W and are selected according to the type of tissue being treated (e.g. 70–80 W for cauterizing small-caliber vessels, 110–120 W for treating the surface of the liver or spleen). The ABC does not come into direct contact with the tissue; rather, as the current contacts the tissue via the stream of argon gas, individual ‘arc tunnels’ are formed within the target tissue. It is the formation of these arc tunnels that is thought to account for a more uniform distribution of current within the tissue and therefore a more uniform coagulative effect with less thermal injury. The ABC can effectively cauterize vessels of up to 3 mm in diameter. The flow of argon gas serves to improve visualization of the operative field by displacing blood and debris. In addition to its utility in achieving hemostasis, the underlying coagulative necrosis generated by application of the ABC is an efficient means of destroying small-volume implants of metastatic ovarian cancer and may facilitate optimal cytoreduction of disease in sites inaccessible to conventional resection. Tumor destruction has been documented in areas such as bowel mesentery, the diaphragm, ureters, vagina, presacral space and the iliac vessels.46,47 When used for ovarian cancer tumor implant ablation, the depth of tissue destruction is dependent on both the power setting and tissue interaction time.48

Cavitron ultrasonic surgical aspirator

Figure 2.2 The Ligasure® vessel sealer. Photograph courtesy of Valleylab, Boulder, CO

48

The cavitron ultrasonic surgical aspirator (CUSA) (Valleylab, Boulder, CO) is another surgical adjunct that may be used during cytoreduction of advancedstage ovarian cancer. The CUSA handpiece encloses

Ch 02 final.qxd

12/18/2006

15:11

Page 49

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

a hollow titanium tube that vibrates at high frequency in a longitudinal axis. The variable amplitude of longitudinal vibration determines the depth of tissue disruption when the handpiece is placed in contact with tissue. The handpiece also contains an irrigation and aspiration system to remove tissue fragments and reduce heat buildup. The extent of tissue disruption is also dependent on the water content of the target tissue, the CUSA causing relatively greater damage to tissues with a high water content (visceral parenchyma, nodal tissue, tumor implants) compared to tissue composed of predominantly connective tissue (muscle, ureter, vessel wall adventitia). The CUSA can be used to resect ovarian cancer metastases on the diaphragm, bowel serosa, liver and splenic capsules, and the peritoneum. Several reports have cited the efficacy and safety of the CUSA as an adjunctive procedure in achieving optimal cytoreduction in areas inaccessible by more standard surgical techniques.49,50 In one prospective trial, the CUSA was associated with a lower perioperative blood loss, shorter hospital stay and less overall morbidity when compared to the findings in patients in whom the CUSA was not used during the cytoreductive effort.51

Automated stapling devices Advanced-stage ovarian cancer commonly involves the intestinal tract by contiguous extension or distant peritoneal metastasis. Consequently, the surgeon must be familiar with a variety of techniques of bowel resection and anastomosis. Traditionally, these procedures were performed using hand-sewn suture techniques. The introduction of automated surgical stapling devices permits the same procedures to be performed with comparable efficacy, greater simplicity and perhaps increased speed. There are multiple brands of commercially available automated stapling devices; however, all utilize the same basic principle of compressing an inverted ‘U-shaped’ staple into a ‘sideways-B’ in the closed position (Figure 2.3). The closed staple position secures the tissue contained within but does not constrict the vascular supply to

the resulting staple line, with the exception of the vascular load staplers (see below). There are three basic categories of automated stapling devices used for bowel surgery as well as other purposes. All contemporary stapling devices are single-use and disposable. The first category is the thoracoabdominal (TA) stapler, which lays down a double row of titanium staples staggered in an overlapping fashion (Figure 2.4). The TA stapler does not have a cutting component and therefore is used to close a segment of the intestinal tract distal to the point of division, or to close an enterotomy or colostomy created during one of the anastomotic techniques described in Chapter 7. The TA stapler is available in three different sizes (30 mm, 60 mm and 90 mm), depending on the width of tissue to be secured. There are two standard staple sizes for the TA stapler, the choice being dependent upon the compression thickness of the stapled tissue. The 3.5-mm staple (open position) compresses to a thickness of approximately 1.5-mm in the closed position, while the 4.8-mm staple (open position) should be used for tissue that will compress to approximately 2.0 mm in the closed position. The Roticulator stapling device is a variation of the standard TA stapler that incorporates a rotating shaft and hinged cartridge head to allow greater flexibility of application. It is particularly useful when dividing a segment of colon or rectum deep in the pelvis. The Roticulator lays down a double row of 4.8-mm titanium staples 55 mm in length.

Figure 2.3 Automated stapling device. Titanium staple in the compressed (closed) position. Photograph courtesy of United States Surgical Corporation, Norwalk, CT

49

Ch 02 final.qxd

12/18/2006

15:11

Page 50

SURGERY FOR OVARIAN CANCER

The second category of automated stapling devices is the gastrointestinal anastomosis (GIA) stapler, which lays down two double rows of staggered titanium staples and has a self-contained cutting blade that divides the tissue between (Figure 2.5).

The GIA stapler is used simultaneously to secure and divide a segment of bowel or other tissue and is available in two lengths – 60 mm and 80 mm – depending on the width of tissue. The basic staple sizes adapted for use in the GIA staplers are 3.8 mm, which

Figure 2.4 Thoracoabdominal (TA) stapling device. Photograph courtesy of United States Surgical Corporation, Norwalk, CT

Figure 2.5 Gastrointestinal anastomosis (GIA) stapling device. Photograph courtesy of United States Surgical Corporation, Norwalk, CT

50

Ch 02 final.qxd

12/18/2006

15:11

Page 51

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

Figure 2.6 Circular end-to-end anastomosis (CEEA) stapling device. Photograph courtesy of United States Surgical Corporation, Norwalk, CT

compresses to 1.5 mm in the closed position; and 4.8 mm, which compresses to 2.0 mm in the closed position. Vascular load staple cartridges are also now in use with the GIA-type staplers that have a staple size of 2.5 mm, which compresses to 1.0 mm in the closed position. The staple line thus created is hemostatic for most small-caliber vascular pedicles. The third category of automated stapling devices is the circular end-to-end anastomosis (CEEA) stapler (Figure 2.6). The CEEA stapler lays down a double row of circular staples and has a self-contained circular cutting blade that simultaneously excises the inverted internal tissue. The 4.8-mm staples compress to a tissue thickness of approximately 2 mm. The CEEA stapler is most commonly used to create endto-end anastomoses of the colon but is also applicable to small bowel–small bowel and small bowel–colon anastomoses. Both straight and curved shafts are available with the CEEA stapler, although when performing a low colorectal anastomosis, navigation of the pelvic curvature is usually easier with the curved model. A low-profile detachable anvil (Figure 2.6) is also available for the CEEA stapler, which is easier to place within the bowel lumen in some circumstances (e.g. stapled end-to-side anastomosis, see Chapters 5 and 7). The standard CEEA stapler comes in four sizes that reflect the outer diameter of the circular staple cartridge: 21 mm, 25 mm, 28 mm and 31 mm. In general, the functional lumenal diameter is approxi-

mately 10 mm smaller than the size of the stapler used to create the anastomosis. A 33-mm cartridge, yielding a 23-mm lumenal diameter, is also available from selected manufacturers (Ethicon Endosurgery, Cincinnati, OH). Successful outcomes using automated surgical staplers to perform bowel anastomoses, including colorectal anastomoses below the levator muscles, have been reported in the gynecologic literature since the late 1970s.52,53 The rate of enteric anastomoticrelated complications following trauma-related intestinal surgery has been confirmed to be similar regardless of whether an automatic stapler or handsewn technique is used.54 In one of the largest case series documenting the use of end-to-end anastomosis stapling devices in the setting of radical gynecologic surgery, the two anastomotic breakdowns reported were noted in patients who had previously undergone radiation therapy.55 Some authors have argued that the favorable outcomes associated with the use of automated stapling devices in patients with gynecologic malignancies are due, at least in part, to improved blood flow to the anastomosis, a contention that has been confirmed in an animal model.53,55,56 It must be stressed that, in all instances, the method of anastomosis elected should reflect the technique with which the surgeon is the most comfortable. The specifications of the three types of automated stapling devices are depicted in Figure 2.7.

51

Ch 02 final.qxd

12/18/2006

15:11

Page 52

SURGERY FOR OVARIAN CANCER

Stapler type

Staple line

Open

Thoracoabdominal (TA)

Staple size Compressed

3.5 mm 4.8 mm

1.5 mm 2.0 mm

2.5 mm 3.8 mm 4.8 mm

1.0 mm 1.5 mm 2.0 mm

cutting blade Gastrointestinal anastomosis (GIA)

Cartridge diameter

Lumenal diameter*

21 mm 25 mm 28 mm 31 mm 33 mm

11 mm 15 mm 18 mm 21 mm 23 mm

cutting blade

Circular end-to-end anastomosis (CEEA)

4.8 mm

2.0 mm

* approximate

Figure 2.7 Specifications of the three types of automated stapling devices

patients with ovarian carcinoma. Aust NZ J Obstet Gynaecol 1978; 18: 209–12

REFERENCES 1.

Wolters U, Wolf T, Stützer H, Schröeder T. ASA classification and perioperative variables as predictors of postoperative outcome. Br J Anaesth 1996; 77: 217–22

2.

Giannice R, Foti E, Poerio A, et al. Perioperative morbidity and mortality in elderly gynecological oncological patients (>/= 70 years) by the American Society of Anesthesiologists physical status classes. Ann Surg Oncol 2004; 11: 219–25

3.

Li AJ, Madden AC, Cass I, et al. The prognostic significance of thrombocytosis in epithelial ovarian cancer. Gynecol Oncol 2004; 92: 211–14

4.

Planner RS, O’Sullivan EF, Campbell JJ, Ball DL. The hypercoaguable state and pulmonary embolism in

52

5.

Menczer J, Schejter E, Geva D, et al. Ovarian carcinoma associated thrombocytosis. Correlation with prognostic factors and with survival. Eur J Gynaecol Oncol 1998; 19: 82–4

6.

Bast RC, Feeney M, Lazarus H, et al. Reactivity of a monoclonal antibody with human ovarian carcinoma. J Clin Invest 1981; 68: 1331–7

7.

Bast RC, Klug TL, St John E. A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 1983; 309: 883–7

8.

Alagoz T, Buller RE, Berman M, et al. What is a normal CA125 level? Gynecol Oncol 1994; 53: 93–7

Ch 02 final.qxd

12/18/2006

15:11

Page 53

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

9.

Bon GG, Kenemans P, Verstraeten R, et al. Serum tumour marker immunoassays in gynecologic oncology: establishment of reference values. Am J Obstet Gynecol 1996; 174: 107–14

10.

Ozols RF, Steele GD, Phillips TL, Chabner BA. Atlas of Clinical Oncology: Ovarian Cancer. Hamilton: BC Decker, 2003: 107–12

11.

Carlson KJ, Skates SJ, Singer DE. Screening for ovarian cancer. Ann Intern Med 1994; 121: 124

12.

Jacobs I, Bast RC Jr. The CA 125 tumour-associated antigen: a review of the literature. Hum Reprod 1989; 4: 1–12

13.

Brooks SE. Preoperative evaluation of patients with suspected ovarian cancer. Gynecol Oncol 1994; 55: S80

14.

Chi DS, Venkatraman ES, Masson V, Hoskins WJ. The ability of preoperative serum CA-125 to predict optimal primary tumor cytoreduction in stage III epithelial ovarian carcinoma. Gynecol Oncol 2000; 77: 227–31

15.

Memarzadeh S, Lee SB, Berek JS, Farias-Eisner R. CA125 levels are a weak predictor of optimal cytoreductive surgery in patients with advanced epithelial ovarian cancer. Int J Gynecol Cancer 2003; 13: 120–4

16.

Kurtz AB, Tsimikas JV, Tempany CMC, et al. Diagnosis and staging of ovarian cancer: comparative values of Doppler and conventional US, CT, and MR imaging correlated with surgey and histopathologic analysis – report of the Radiology Diagnostic Oncology Group. Radiology 1999; 212: 19–27

ty of ovarian carcinoma at primary laparotomy? Br J Obstet Gynaecol 2003; 109: 369–75 21.

Karlan BY, Hawkins R, Hoh C, et al. Whole-body positron emission tomography with 2-[18F]fluoro-2deoxy-D-glucose can detect recurrent ovarian carcinoma. Gynecol Oncol 1993; 51: 175–81

22.

Hubner KF, McDonald TW, Niethammer JG, et al. Assessment of primary and metastatic ovarian cancer by positron emission tomography (PET) using 2-18Fdeoxy-D-glucose (2-18F-FDG). Gynecol Oncol 1993; 51: 197–204

23.

Bristow RE, Simpkins F, Pannu HK, et al. Positron emission tomography for detecting clinically occult surgically respectable metastatic ovarian cancer. Gynecol Oncol 2002; 85: 196–200

24.

Rieber A, Nüssle K, Stöhr H, et al. Preoperative diagnosis of ovarian tumors with MR imaging: comparison with transvaginal sonography, positron emission tomography, and histologic findings. Am J Roentgenol 2001; 177: 123–9

25.

Fenchel S, Grab D, Nuessle K, et al. Asymptomatic adnexal masses: correlation of FDG PET and histopathologic findings. Radiology 2002; 223: 780–8

26.

Bristow RE, del Carmen MG, Pannu HK, et al. Clinically occult recurrent ovarian cancer: patient selection for secondary cytoreductive surgery using combined PET/CT. Gynecol Oncol 2003; 90: 519–28

27.

Barakat KK, Bevers MW, Gershenson DM. Hoskins WJ. Handbook of Gynecologic Oncology. London: Martin Dunitz, 2000: 111–30

17.

Nelson BE, Rosenfield AT, Schwartz PE. Preoperative abdominopelvic computed tomographic prediction of optimal cytoreduction in epithelial ovarian carcinoma. J Clin Oncol 1993; 11: 166–72

28.

Perioperative total parenteral nutrition in surgical patients. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. N Engl J Med 1991; 325: 525–32

18.

Meyer JI, Kennedy AW, Friedman R, et al. Ovarian carcinoma: value of CT in predicting success of debulking surgery. Am J Roentgenol 1995; 165: 875–8

29.

Heyland DK, MacDonald S, Keefe L, Drover JW. Total preanteral nutrition in the critically ill patient. A meta-analysis. J Am Med Assoc 1998; 280: 2013–19

19.

Bristow RE, Duska LR, Lambrou NC, et al. A model for predicting surgical outcome in patients with advanced ovarian carcinoma using computed tomography. Cancer 2000; 89: 1532–40

30.

Buzby GP. Overview of randomized clinical trials of total parenteral nutrition for malnourished surgical patients. World J Surg 1993; 17: 173–7

31.

20.

Byrom J, Widjaja E, Redman CW, et al. Can preoperative computed tomography predict respectabili-

Waitzberg DL, Plopper C, Terra RM. Postoperative total parenteral nutrition. World J Surg 1999; 23: 560–4

53

Ch 02 final.qxd

12/18/2006

15:11

Page 54

SURGERY FOR OVARIAN CANCER

32.

Hemsell DL. Prophylactic antibiotics in gynecologic and obstetric surgery. Rev Infect Dis 1991; 13: S821–41

33.

Mittendorf R, Aronson MP, Berry RE. Avoiding serious infections associated with abdominal hysterectomy: a meta-analysis of antibiotic prophylaxis. Am J Obstet Gynecol 1993; 169: 1119–24

44.

Clarke-Pearson DL, DeLong E, Synan IS, et al. A controlled trial of two low-dose heparin regimens for the prevention of post-operative deep vein thrombosis. Obstet Gynecol 1990; 75: 684–9

45.

Enoxacan Study Group. Efficacy and safety of enoxaperin versus unfractionated heparin for the prevention of deep vein thrombosis in elective cancer surgery: a double-blind randomized multicentre trial with venographic assessment. Br J Surg 1997; 84: 1099–103

46.

Brand E, Pearlman N. Electrosurgical debulking of ovarian cancer: a new technique using the argon beam coagulator. Gynecol Oncol 1990; 39: 115–18

47.

Burke P, Mealy K, Gillen P, et al. Requirement for bowel preparation in colorectal surgery. Br J Surg 1994; 81: 907–10

Eisenkop SM, Nalick RH, Wang H, Nelson T. Peritoneal implant elimination during cytoreductive surgery for ovarian cancer: impact on survival. Gynecol Oncol 1993; 51: 224–9

48.

37.

Santos JC Jr, Batista J, Sirimarco MT, et al. Prospective randomized trial of mechanical bowel preparation in patients undergoing elective colorectal surgery. Br J Surg 1994; 81:1673–6

Bristow RE, Smith Sehdev AE, Kaufman HS, Montz FJ. Ablation of metastatic ovarian carcinoma with the argon beam coagulator: pathological analysis of tumor destruction. Gynecol Oncol 2001; 83: 49–55

49.

38.

Zmora O, Mahajna A, Bar-Zakai B, et al. Colon and rectal surgery without mechanical bowel preparation: a randomized prospective trial. Ann Surg 2003; 237: 363–7

Deppe G, Malviya VK, Malone JM. Debulking surgery for ovarian cancer with the Cavitron Ultrasonic Surgical Aspirator (CUSA) – a preliminary report. Gynecol Oncol 1988; 31: 223–6

50.

39.

Coppa GF, Eng K. Factors involved in antibiotic selection in elective colon and rectal surgery. Surgery 1988; 104: 853–8

Adelson MD, Baggish MS, Seifer DB, et al. Cytoreduction of ovarian cancer with the Cavitron ultrasonic surgical aspirator. Obstet Gynecol 1988; 72: 140–3

40.

Clarke JS, Condon RE, Bartlett JG, et al. Preoperative oral antibiotics reduce septic complications of colon operations: results of prospective, randomized, double-blind clinical study. Ann Surg 1977; 186: 251–9

51.

van Dam PA, Tjalma W, Weyler J, et al. Ultraradical debulking of epithelial ovarian cancer with the ultrasonic surgical aspirator: a prospective randomized trial. Am J Obstet Gynecol 1996; 174: 943–50

41.

Stellato TA, Danziger LH, Gordon N, et al. Antibiotics in elective colon surgery. A randomized trial of oral, systemic, and oral/systemic antibiotics for prophylaxis. Am Surg 1990; 56: 251–4

52.

Wheeless CR. Avoidance of permanent colostomy in pelvic malignancy using the surgical stapler. Obstet Gynecol 1979; 54: 501–5

53.

42.

Nichols RL, Smith JW, Garcia RY, et al. Current practices of preoperative bowel preparation among North American colorectal surgeons. Clin Infect Dis 1997; 24: 609–19

Wheeless CR, Dorsey JH. Use of the automatic surgical stapler for intestinal anastomosis associated with gynecologic malignancies: review of 283 procedures. Gynecol Oncol 1981; 11: 1–7

54.

43.

Clarke-Pearson DL, Olt G. Thromboembolism in patients with gyn tumors: risk factors, natural history, and prophylaxis. Oncology 1989; 3: 39–45

Kirkpatrick AW, Baxter KA, Simons RK, et al. Intraabdominal complications after surgical repair of small bowel injuries: an international review. J Trauma 2003; 55: 399–406

34.

35.

36.

54

Classen DC, Evans RS, Pestotnik SL, et al. The timing of prophylactic administration of antibiotics and the risk of surgical wound infection. N Engl J Med 1992; 326: 281–6 Brownson P, Jenkins S, Nott D, et al. Mechanical bowel preparation before colorectal surgery: results of a prospective randomized trial [abstract]. Br J Surg 1994; 79: 461–2

Ch 02 final.qxd

12/18/2006

15:11

Page 55

PREOPERATIVE PREPARATION AND SURGICAL INSTRUMENTATION

55.

Harris WJ, Wheeless CR. Use of the end-to-end anastomosis stapling device in low colorectal anastomosis associated with radical gynecologic surgery. Gynecol Oncol 1986; 23: 350–7

56.

Wheeless CR, Smith JJ. A comparison of the flow of iodine 125 through three different intestinal anastomoses: standard, Gambee, and stapler. Obstet Gynecol 1983; 62: 513–18

55

Ch 03 final.qxd

12/18/2006

15:12

Page 57

CHAPTER 3

Management of early-stage ovarian cancer Anil K Sood, David M Gershenson

INTRODUCTION A minority of patients with epithelial ovarian cancer will have disease confined to the ovary or extraovarian pelvis, and can be expected to have a favorable longterm survival outcome. A comprehensive surgical staging procedure is critical not only for identifying this select subset of patients but also for determining the prescription for adjuvant therapy. While most patients with borderline tumors, ovarian germ cell tumors and sex cord stromal tumors will have earlystage disease, the information gleaned from surgical staging is no less relevant. This chapter will define the rationale for and techniques of the ovarian cancer staging operation for women with apparent early-stage disease.

EPITHELIAL OVARIAN CANCER Rationale for surgical staging The large number of deaths from ovarian cancer is largely due to the fact that most women go undiagnosed until extensive tumor spread has occurred, leading to poor survival. Survival of women with ovarian cancer is determined largely by stage – women with stage I or II cancer have 5-year survival rates between 50 and 90%,1 whereas survival of women with advanced-stage disease decreases to around 30%.2 Because of the lack of an acceptable screening test for ovarian cancer, early detection becomes crucial to improving survival. Historically, it was thought that

most patients with ovarian cancer did not have symptoms until advanced stages, and as a result ovarian cancer was often called the ‘silent killer’. However, recent studies suggest that most patients, even those with early-stage carcinoma, do have symptoms.3,4 Goff et al. reported that only 11% of stage I/II patients reported no symptoms before their diagnosis.3 Eltabbakh et al. performed a retrospective analysis of 72 women with stage I or II borderline or invasive ovarian cancers and found that most (78%) of the patients had presenting symptoms of either abdominal or pelvic pain, bloatedness or vaginal bleeding.4 Collectively, these studies emphasize that even women with early-stage ovarian cancer have symptoms, and individuals with non-specific symptoms should be carefully evaluated for possible ovarian pathology. The detection of a pelvic mass is relatively easy; however, the preoperative characterization of that mass as benign versus malignant has remained a challenge despite the use of numerous modalities including ultrasound and serum markers. Although surgery alone is curative in only a small number of patients, it remains the most important modality of treatment in an individual with a suspicious pelvic mass found to be ovarian cancer. The stage of disease is determined by the extent of tumor at the time of initial diagnosis. The staging classification for ovarian cancer is based on surgical and pathologic findings, and was last modified by the International Federation of Gynecology and Obstetrics (FIGO) in 1985 (Table 3.1). Earlystage epithelial ovarian carcinoma is defined pathologically as tumor limited to one or both ovaries or tumor extension confined to the pelvis. This diagnosis can be

57

Ch 03 final.qxd

12/18/2006

15:12

Page 58

SURGERY FOR OVARIAN CANCER

Table 3.1 FIGO staging for primary carcinoma of the ovary Stage

Description

I

Growth limited to the ovaries

IA

Growth limited to one ovary; no ascites present containing malignant cells; no tumor on the external surfaces; capsule intact

IB

Growth limited to both ovaries; no ascites present containing malignant cells; no tumor on the external surfaces; capsules intact

IC

Tumor either stage IA or IB, but with tumor on the surface of one or both ovaries; or with capsule ruptured; or with ascites present containing malignant cells, or with positive peritoneal washings

II

Growth involving one or both ovaries with pelvic extension

IIA

Extension and/or metastases to the uterus and/or tubes

IIB

Extension to other pelvic tissues

IIC

Tumor either stage IIA or IIB but with tumor on the surface of one or both ovaries; or with capsule(s) ruptured; or with ascites present containing malignant cells, or with positive peritoneal washings

III

Tumor involving one or both ovaries with peritoneal implants outside the pelvis and/or positive retroperitoneal or inguinal nodes; superficial liver metastasis equals stage III; tumor is limited to the true pelvis but with histologically verified malignant extension to small bowel or omentum

IIIA

Tumor grossly limited to the true pelvis with negative nodes with histologically confirmed microscopic seeding or abdominal peritoneal surfaces

IIIB

Tumor of one or both ovaries; histologically confirmed implants of abdominal peritoneal surfaces, none exceeding 2 cm in diameter; nodes negative

IIIC

Abdominal implants greater than 2 cm in diameter and/or positive retroperitoneal or inguinal nodes

IV

Growth involving one or both ovaries with distant metastases; if pleural effusion is present, there must be positive cytology to allot a case to stage IV; parenchymal liver metastasis equals stage IV

obtained only after an exhaustive surgical staging procedure.5 This chapter will focus on the implications of and technique for surgical staging of early-stage ovarian cancer. For the purpose of this chapter, we will define early-stage disease as FIGO stages I and II. A gynecologic surgeon embarking upon exploratory surgery on a patient with a pelvic mass should be prepared to perform full staging and, if needed, cytoreduction for ovarian cancer. Approximately 15–30% of patients with epithelial ovarian cancers present with disease confined to the ovaries.6 Although a high proportion of patients with stage I ovarian cancer are cured by the surgical removal of tumor, the curative value of postoperative treatment has not been proved. Thus, accurate predictors of

58

relapse might allow clinicians to separate those patients with an extremely small risk from those with an appreciable risk of relapse who might benefit from adjuvant therapy. The importance of surgical staging was further highlighted by Bagley et al. who showed a 30% discrepancy between staging by non-oncologists and reoperation by gynecologic oncologists.7 Since then, others have also demonstrated that there can be up to 30% discrepancy between reported surgical findings by non-oncologic surgeons and the definitive stage demonstrated in re-operation by a gynecologic oncologist.8–10 Piver et al. identified the metastatic sites most frequently overlooked by the non-oncologist as the diaphragm, omentum and retroperitoneal lymph

Ch 03 final.qxd

12/18/2006

15:13

Page 59

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

nodes.8,9 Young et al. systematically restaged 100 patients referred to them with apparent early-stage ovarian cancer.6 Thirty-one per cent were found to have a more advanced stage and most of these patients had stage III disease. McGowan et al. found that, among 291 patients with ovarian cancer, only 54% received proper staging.11 The most commonly missed sites included the diaphragm, biopsy of the pelvic peritoneum, peritoneal fluid sampling and omental biopsy. The likelihood of complete staging varied depending on the specialty of the operating surgeon: gynecologic oncologists, 97%; obstetrician– gynecologists, 53%; and general surgeons, 35%. Mayer et al. found a higher survival rate in patients with stage I and II disease who were comprehensively staged by a gynecologic oncologist, compared with those women who underwent more limited surgical procedures by a non-oncologic gynecologist or a general surgeon.10 Le et al. evaluated 80 patients with apparent stage I ovarian cancer with a comprehensive surgical staging procedure and 30 (38%) were upstaged after the surgical staging, most to stage III.12 Furthermore, they found that endometrioid, mucinous and clear cell carcinoma had better correlation with a true surgical stage I (76–88%) than other histologic subtypes. Serous and anaplastic tumors had about a 70% chance of being upstaged by the careful surgical procedure. In a subsequent study, Le et al. evaluated 138 patients with tumor grossly confined to the ovary at the time of laparotomy.13 Ninety-four (68%) patients had a complete surgical staging procedure and 36% of these patients were found to have extraovarian metastases. Forty-three per cent of those not having staging laparotomy were offered chemotherapy based on high-risk factors only. Among surgically proven stage I patients treated expectantly, only 10% developed a recurrence compared to 28% in the unstaged group who were treated expectantly because of lack of risk factors (p = 0.036).13 This study supports the concept that absence of surgical pathologic high-risk factors is inferior to comprehensive staging laparotomy findings with regard to guiding adjuvant therapy. Not surprisingly, patients who

were not properly staged had a high risk of developing recurrent disease despite receiving chemotherapy. Collectively, these studies suggest that there is a substantial likelihood of finding metastatic disease if a formal staging surgery is performed in women with disease apparently confined to the ovary or ovaries. Furthermore, a gynecologic oncologist should be involved in performing the staging procedure for ovarian cancer. If the operative risks are not too high, a restaging procedure must be routinely considered before the decision is made concerning an adjuvant treatment.

Incision selection and management Abdominal wall anatomy

A detailed knowledge of the abdominal and pelvic anatomy is crucial to the success of a gynecologic surgeon. The layers of the abdominal wall include the skin, superficial fascia, external oblique, internal oblique, transversus abdominis muscles and aponeuroses, the transversalis fascia, the preperitoneal fat and the parietal peritoneum (Figure 3.1). The combination of aponeuroses and muscles cross each other in such a way to strengthen the anterolateral abdominal wall. The three flat muscles of the abdominal wall originate from the lower six ribs, thoracolumbar fascia, the iliac crest and the lateral two-thirds of the inguinal ligament. The insertion of these muscles is onto the linea alba through their aponeuroses. These muscles are innervated by the inferior five intercostal nerves as well as the subcostal, iliohypogastric and ilioinguinal nerves. These muscles increase the intraabdominal pressure, which can play a role in normal processes such as defecation and micturition. They also play a role in flexion and rotation of the trunk. The rectus abdominis is a long strap muscle – the two inferior parts lie together, whereas superiorly the muscles broaden out where they are separated by the linea alba. The upper attachment of the rectus abdominis muscle is up to three times as broad as the pubic insertion. The rectus abdominis muscle is largely enclosed in the rectus sheath, formed by the aponeuroses of the

59

Ch 03 final.qxd

12/18/2006

15:13

Page 60

SURGERY FOR OVARIAN CANCER

External oblique muscle Internal oblique muscle Posterior sheath

Transversus abdominis muscle

Arcuate line

Rectus muscle

Rectus muscle

Figure 3.1 Anatomy of the anterior abdominal wall

three flat abdominal muscles. The rectus abdominis muscle originates from the symphysis pubis and the pubic crest and inserts onto the xiphoid process and fifth through seventh costal cartilages. The pyramidalis is a small muscle that originates from the body of pubis and inserts onto the linea alba. It is innervated by the subcostal nerve, but does not have much functional significance. The fibers of the rectus abdominis and pyramidalis muscles run in a vertical fashion. The rectus sheath forms by the fusion and separation of the aponeuroses of the flat abdominal muscles (Figure 3.2). At its lateral margin, the internal oblique aponeurosis splits into two layers – one passing anterior to the rectus muscle and one passing posterior to it. The anterior layer joins with the aponeurosis of the external oblique to form the anterior wall of the rectus sheath, and the posterior layer joins with the aponeurosis of the transversus abdominis muscle to form the posterior wall of the rectus sheath. The anterior and posterior walls of the rectus sheath fuse in the anterior midline to form the linea alba. The inferior portion of the rectus sheath is deficient in its posterior wall, because the internal oblique aponeurosis does not split here to enclose the rectus muscle. The infe-

60

a

1 2 3

b 1 2 3

Figure 3.2 Anatomy of the rectus sheath in cross-section. (a) Above the arcuate line; (b) below the arcuate line. 1, external oblique; 2, internal oblique; 3, transversus abdominis muscles

rior limit of the posterior wall of the rectus sheath is marked by a crescentic border called the arcuate line. The position of the arcuate line is usually midway between the umbilicus and the pubic crest. Inferior to the arcuate line, the aponeuroses of the three flat abdominal muscles pass anterior to the rectus muscle to form the anterior layer of the rectus sheath. The posterior wall of the rectus sheath is also deficient superior to the costal margin, because the transversus

Ch 03 final.qxd

12/18/2006

15:13

Page 61

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

abdominis muscle passes internal to the costal cartilages, and the internal oblique muscle is attached to the costal margin. The fibers of the external oblique, the internal oblique and the transversus abdominis muscles run transversely or diagonally. Superior to the costal margin, the rectus abdominis muscle lies directly on the thoracic wall. In addition to the rectus abdominis muscle, the rectus sheath contains the superior and inferior epigastric vessels, and the terminal parts of the inferior five intercostals and subcostal vessels and nerves. The distribution and course of the nerves and blood vessels of the abdominal wall bear directly on the postoperative healing and function of the abdominal wall. The implications of various types of abdominal incisions on postoperative healing are discussed later in this chapter. The transversalis fascia is the internal investing layer, which lines the entire abdominal wall. It is continuous from side to side, deep to the linea alba. Internal to the transversalis fascia is the peritoneum, and these two layers are separated by a variable amount of extraperitoneal fat. The blood supply to the abdominal wall comes from several sources. The main blood supply of the rectus muscles and midabdomen comes from the superior epigastric arteries and the inferior epigastric arteries. The superior epigastric artery is a continuation of the internal thoracic artery and it enters the rectus sheath from behind the seventh costal cartilage and descends posterior to the rectus. It has multiple branches in the substance of the rectus muscle and anastomosis to the inferior epigastric artery. The inferior epigastric artery arises from the external iliac artery and continues in a cephalad course along the posterolateral portion of the rectus muscle. It has anastomoses with the superior epigastric artery. The lateral abdominal wall is supplied by the lower intercostal and lumbar arteries and the superficial and deep circumflex arteries. The deep circumflex artery runs on the deep aspect of the anterior abdominal wall, parallel to the inguinal ligament, and along the iliac crest between the transversus abdominis and internal oblique muscles. The veins lie in close proximity to these arteries. The medial abdominal wall

receives blood from the epigastric arteries. Overall, the anterior abdominal wall is well vascularized except for the linea alba. The innervation of the anterior abdominal wall (skin and muscles) is derived almost entirely from the thoracoabdominal (seventh to eleventh intercostals) nerves. The inferior part of the anterolateral abdominal wall is supplied by the iliohypogastric and ilioinguinal nerves. The thoracoabdominal nerves supply the internal and external obliques, rectus abdominis and the transversus abdominis muscles. The main trunks of the thoracoabdominal nerves pass anteriorly from the intercostal spaces and then run between the internal oblique and transversus abdominis muscles. The iliohypogastric and ilioinguinal nerves are sensory in function. Both nerves supply the lower part of the internal oblique and transversus muscles. The anterior cutaneous nerves penetrate the rectus sheath a short distance from the median plane. The anterior cutaneous branches of T7 to T9 supply the skin superior to the umbilicus, T10 innervates the skin at the level of the umbilicus; and T11, T12 and L1 supply the skin inferior to the umbilicus. The nerve supply to the anterior abdominal wall can be damaged by some incisions. A transverse incision is least likely to cause injury to nerves. Similarly, a vertical incision in the midline has a low likelihood of nerve damage. However, a vertical incision that passes lateral to the rectus muscle or through the muscle will denervate the medial tissues, depending on the length of the incision. Type of incision

This chapter will focus on open surgical techniques – details regarding laparoscopic approaches are discussed in Chapter 12. It is the responsibility of the surgeon to select one of the several abdominal incisions available that is most suited to the successful performance of the planned procedure (Figure 3.3). The main considerations in selecting an incision for any operation are exposure, ability to extend the incision and minimize risk for complications. The incision can influence the relative ease with which abdominal and

61

Ch 03 final.qxd

12/18/2006

15:13

Page 62

SURGERY FOR OVARIAN CANCER

Midline

Low Maylard Cherney Pfannenstiel

Figure 3.3 Incision selection for ovarian cancer surgery; the low vertical midline incision is the most versatile when ovarian malignancy is suspected

pelvic dissection can be accomplished. Limited exposure through an inadequate incision is dangerous and can result in a poor cosmetic and therapeutic outcome. Although several types of abdominal surgical incisions are available to the gynecologic surgeon, the most common surgical approach for women with suspected ovarian cancer is by the midline vertical incision. It provides easy access to both the deep pelvis and the upper abdomen. A low vertical incision is a highly versatile incision, which can be easily extended superiorly if disease is discovered in the upper abdomen. It is a relatively simple incision that can provide rapid access to the abdomen and pelvis in emergency cases. A vertical incision has minimal associated blood loss because there are no significant blood vessels encountered during this incision. The disadvantages of a vertical incision include less cosmesis, less strength due to greater tension and higher rates of wound breakdown. The appearance of the surgical incision will be noticed by the patient and family members. Most patients are concerned about the length and final appearance of the surgical incision. It is helpful to mark the entire distance between the pubic symphysis and the xiphoid process; this allows for a straight inci-

62

sion should the need arise for extending a limited incision. If a previous scar is present, depending on its appearance, it can be either excised or incised. Entry into the peritoneal cavity is at times easier above the superior extent of the previous scar in order to avoid entry into dense adhesions, also avoiding inadvertent damage to the abdominal organs. Making a midline incision alongside a previous paramedian incision should be avoided in order to prevent devascularization of the intervening tissues. Transverse incisions have the advantage of being more cosmetic and less painful, and tend to have less effect on pulmonary function, and may be stronger than a vertical incision. Transverse incisions are appropriate for small, benign pathology or when dissection is primarily limited to the pelvis. The disadvantages of a transverse incision include more technical difficulty, greater blood loss and limited access to the upper abdomen. Thus, for ovarian cancer staging and cytoreduction, transverse incisions have a limited role. If a surgeon makes a transverse incision and needs access to the upper abdomen, either a separate vertical incision (T-incision) must be made or the transverse incision can be extended upward laterally (J-incision). Pfannenstiel incision is probably the most limiting, because the rectus muscles are retracted and it provides less access to the pelvis than other transverse incisions. The Cherney incision is made closer to the pubic symphysis to divide the tendon of the rectus abdominis at the pubic crest. Its disadvantages include the need for lower placement of the incision and difficulty in retracting the abdominal wall in an obese patient. The Maylard incision is most commonly made in the infraumbilical region. If a Pfannenstiel incision has been performed and the pelvic exposure is inadequate, the incision should not be converted to a Maylard incision, because the ends of the rectus muscle will retract beneath the previously dissected rectus fascia and will not be easily reapproximated. The Pfannenstiel incision can be converted to a Cherney incision for greater exposure. Surgical incisions are closed most commonly by primary intention (approximation of the surgical wound

Ch 03 final.qxd

12/18/2006

15:13

Page 63

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

to allow healing), and less commonly by secondary intention (the skin and subcutaneous tissues are allowed to heal spontaneously) and delayed primary closure (tissues are approximated after allowing time for granulation). There are multiple factors that can affect postoperative wound healing including tissue oxygenation, the patient’s age and other co-morbid conditions, nutritional status and other therapeutic and pharmacologic agents. The midline abdominal incision should be closed with an internal mass closure technique using monofilament delayed absorbable or permanent suture. In comparison to the rapidly absorbable suture, the delayed absorbable suture is associated with a lower incidence of incisional hernia.14 However, there are no significant differences in risk of incisional hernia between delayed absorbable and non-absorbable sutures.14 The continuous mass closure (Figure 3.4) appears to be similar to the interrupted mass closure with regard to short- and longterm wound security, but the continuous method is more cost-efficient and faster.15,16 We prefer to use the running Smead–Jones closure, with each bite 1.5–2 cm lateral to the edge of the fascial incision.

The transverse incisions are closed with a running delayed-absorbable suture in the fascia. There are limited data regarding the use of subcutaneous retention sutures; however, Soisson et al. have found that their use in obese women (subcutaneous tissue of 5 cm or greater) was associated with a lower risk of superficial wound separation in univariate and multivariate analyses.17

Surgical staging technique The standards for staging were established more than 20 years ago by the Gynecologic Oncology Group.18 Buchsbaum et al. performed a prospective study of 187 patients with malignant ovarian cancer through the Gynecologic Oncology Group to develop a standard surgical procedure for ovarian cancer patients.18 The stage was elevated in almost 13% of patients based on full surgical staging. Several steps exploring both the intra- and the retroperitoneal spaces are performed at the time of surgical management for complete staging evaluation (summarized in Table 3.2). If a perioperative diagnosis of invasive ovarian malignancy is not

1 cm

Rectus sheath Muscle Posterior rectus sheath

2 cm

a

b

Figure 3.4 (a) Continuous mass closure of vertical midline incision; (b) the closure incorporates all fascial and muscular layers of the anterior abdominal wall, with each bite taken 2 cm back from the midline incision edge

63

Ch 03 final.qxd

12/18/2006

15:13

Page 64

SURGERY FOR OVARIAN CANCER

possible, then a secondary staging procedure should be considered as soon as possible. Exploration

The type of incision is discussed above in detail. Our standard approach is to perform a midline vertical incision in cases where an ovarian malignancy is suspected. The initial incision is made from the umbilicus to the pubic bone. The incision is then extended superiorly, as needed, to complete the staging operation in the upper abdomen or perform cytoreductive surgery. Sampling of the peritoneal fluid is obtained for cytologic examination. When no ascites is present, a peritoneal washing with 100–150 ml of saline solution must be performed. Washings are obtained from the pelvis, the paracolic spaces and the level of each hemidiaphragm. Careful and thorough exploration of the entire peritoneal cavity is performed in a systematic fashion by inspection and palpation for tumor implants. The exploration is started by palpating the right paracolic space, and advancing the hand to the right kidney, suprahepatic space, the right diaphragm, right hepatic lobe, gallbladder, Morison’s pouch, left hemi-diaphragm, left hepatic lobe, spleen, stomach,

Table 3.2 Systematic surgical staging for epithelial ovarian cancer Peritoneal cytology Careful and systematic abdominal exploration – inspect and palpate all peritoneal surfaces Omentectomy Total abdominal hysterectomy and bilateral salpingooophorectomy Pelvic and para-aortic lymphadenectomy Random and directed peritoneal biopsies – posterior cul-desac, bladder reflection, both pelvic sidewalls and both paracolic spaces Biopsy or scrapings from the undersurface of both diaphragms Appendectomy in selected cases (for example, mucinous histology)

64

transverse colon, left kidney and left paracolic space. Decompression of the stomach may facilitate access to these structures. The lesser sac is entered on the left side of the gastrocolic ligament to evaluate the stomach and pancreas. The entire small bowel and colon including the mesentery are carefully examined. Both surfaces of the small bowel mesentery are inspected from the jejunum to the cecum. The retroperitoneal areas are palpated along the vascular structures. The pancreas and duodenum can be palpated transperitoneally. The results from this exploration are carefully noted with regard to tumor size, extent and location. Any other abnormalities are also recorded. Management of the primary tumor

The primary ovarian tumor and pelvis should be carefully examined. Both ovaries should be evaluated for size, presence of gross tumor, capsule rupture, external excrescences and adherence to surrounding structures. If surgical findings are suggestive of a benign mass in a young patient, then ovarian cystectomy may be indicated. Otherwise, a unilateral salpingooophorectomy should be performed and submitted for frozen section evaluation. If bilateral ovarian masses are present, the more suspicious side should be removed initially. If frozen section analysis reveals a malignant epithelial tumor, then standard surgical therapy consists of hysterectomy and bilateral salpingo-oophorectomy. In patients for whom fertility is not a concern, removal of the contralateral ovary is justified even if it appears normal. Specifically, the contralateral ovary may harbor occult metastatic disease or even a primary lesion. Hysterectomy is also recommended in most cases, as the uterus may be involved by microscopic lymphatic or serosal metastasis and is a potential site for a synchronous primary tumor of the endometrium. The possibility of conservative surgery is discussed later in this chapter. Although the issue of whether intraoperative rupture of malignant ovarian tumors influences the prognosis of patients with stage I ovarian cancer is controversial, ideally an adnexal mass should be removed

Ch 03 final.qxd

12/18/2006

15:13

Page 65

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

intact. Sainz de la Cuesta et al. evaluated 79 patients with stage I invasive epithelial ovarian cancer and found that, compared to women with stage IA disease, individuals with intraoperative rupture had a higher risk of recurrence and decreased overall survival.19 However, other studies have failed to show an adverse effect of tumor rupture or ascites on survival in patients with stage I disease.20–22 The findings of at least some of these studies are probably confounded by inadequate staging information and the effect of histologic grade. Staging biopsies

As a part of staging, surgical biopsies are taken to evaluate microscopic metastases in high-risk areas, which grossly appear normal. Adhesions and any other abnormal-appearing areas on the peritoneum should be biopsied. If no abnormalities on peritoneal surfaces are noted, then directed biopsies of the peritoneum are taken from the posterior cul-de-sac, bladder peritoneum, bilateral pelvic sidewalls, both paracolic gutters and the undersurface of the diaphragm on both sides. Some surgeons prefer to take peritoneal scrapings from the diaphragms rather than biopsies, and put the specimen on a slide, fixed immediately with a preservative.23 Omentectomy

An omentectomy should be performed as a part of the staging procedure. The vascular supply of the omentum comes from the right and left gastroepiploic arteries, which arise from the gastroduodenal artery and the splenic artery, respectively. In the absence of gross omental disease, an infracolic omentectomy appears to be adequate.18 The value of omentectomy is illustrated by the finding that clinical impression is reported to be inaccurate in 45% of cases with metastases to the omentum.18 In patients with apparent early-stage ovarian cancer, omental disease is detected in about 5% (0–10%).6,18 The omentum is elevated and the posterior reflection onto the transverse colon incised to enter the lesser sac (Figure 3.5). The lesser sac is further developed by dissecting between the posterior

layer of the gastrocolic ligament and the anterior layer of the transverse mesocolon. During this dissection, the middle colic artery must be protected from injury. The dissection is then continued bilaterally toward the hepatic and splenic flexures. On the left, hard traction should be avoided, so that the splenic capsule is not torn. After the omentum has been fully mobilized from the transverse colon, a series of clamps are placed across the omental arteries and veins, avoiding the gastroepiploic vascular arcade along the greater curvature of the stomach (Figure 3.6). The infracolic omentum is removed and the pedicles are sequentially suture ligated. Retroperitoneal lymph node dissection

In 1985, FIGO modified the staging for ovarian carcinoma to reflect the prognostic significance of metastatic spread to the pelvic or para-aortic lymph nodes.24 Assessment of the retroperitoneal nodes is an important part of the initial staging for ovarian cancer. Ovarian cancer drains preferentially to the high aortic nodes in the vicinity of the junction of the ovarian veins and the vena cava on the right side and the renal vein on the left side (Figure 3.7).25–28 Therefore, the diagnostic aortic lymph node sampling should include the high aortic nodes as well as the lower aortic and common iliac regions. Lymphatic drainage of the ovary is known to follow the gonadal blood supply. The para-aortic lymph nodes are part of the lumbar lymph node groups, which comprise three subgroups: preaortic, retroaortic and lateral aortic. The lateral group receives lymphatic drainage from the iliac lymph nodes, ovaries and other pelvic organs. The dominant lymph channels coalesce in the infundibulopelvic ligament, where they travel with the pampiniform plexus of the ovarian veins, and then drain toward the inferior pole of the kidney and move medially into the para-aortic and precaval lymph nodes. An additional pathway from the hilus of the ovary traverses the broad ligament draining into the obturator, external and common iliac lymph nodes.29–31

65

Ch 03 final.qxd

12/18/2006

15:13

Page 66

SURGERY FOR OVARIAN CANCER

Operator

Second assistant

Omentum

First assistant

Stomach

Omentum

Colon

Line of incision Posterior taenia Mesocolon Transverse colon

Mesocolon

Posterior taenia

Figure 3.5 Infracolic omentectomy. The omentum is elevated into the incision and the posterior reflection onto the transverse colon incised to provide access to the lesser sac

Operator

Anterior taenia

Figure 3.6 Infracolic omentectomy. After developing the lesser sac, the omental vascular pedicles are skeletonized, clamped, divided and ligated

66

Ch 03 final.qxd

12/18/2006

15:13

Page 67

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

The frequency of lymphatic spread in patients with apparent stage I ovarian cancer ranges from 5 to 20% (Table 3.3). Cass et al. evaluated 96 consecutive patients treated at two institutions, with disease visibly confined to one ovary.32 Forty-two patients had lymph node sampling only on the side ipsilateral to

the neoplastic ovary, four (10%) of whom had lymph node metastases. Fifty-four patients had bilateral sampling performed and ten (19%) of these patients were found to have lymph node involvement. Among the ten patients with bilateral sampling, 50% had only ipsilateral involvement. However, the remaining 50%

Para-aortic nodes

Common iliac nodes Internal iliac nodes External iliac nodes

Figure 3.7 Lymphatic drainage of the ovary. The principal drainage routes are: (1) cephalad, along the course of the ovarian vessels, to the high para-aortic nodes, and (2) through the broad ligament to the obturator, external and common iliac nodes

Table 3.3 Positive lymph nodes in ovarian cancer visibly confined to the ovary(ies) Author

Number of cases

Positive pelvic (%)

Positive para-aortic (%)

Both (%)

35

8.6

5.7

2.9

Burghardt31 – stage I/II

27

11.1

3.7

14.8

Cass32

96

7.3

5.2

2.1

Petru33 – stage I

40

17.5

2.5

2.5

Chen26

11

9.1

18.2

Benedetti–Panici25

– stage I

– stage I

– stage I

67

Ch 03 final.qxd

12/18/2006

15:13

Page 68

SURGERY FOR OVARIAN CANCER

had either contralateral (30%) alone or bilateral (20%) lymph node involvement.32 All patients with positive lymph nodes in this study had high-grade disease. This study emphasizes the importance of performing bilateral pelvic and para-aortic lymph node dissection in patients with apparent local disease. Petru et al. evaluated 40 patients with disease grossly confined to the ovary who underwent comprehensive surgical staging.33 Nine (23%) were found to have lymph node metastases. Although most patients with nodal metastases had high-grade (2 or 3) disease, none of the other clinical–morphologic factors such as ascites, adherence or extracystic excrescences were predictive of nodal metastasis. One of six patients with unilateral ovarian involvement had isolated contralateral pelvic nodal metastasis. The results of this study do not support limiting lymphadenectomy to any subset of patients with stage I ovarian cancer. Some of the variation in the nodal positivity rates may be explained by the differences in the type of lymphadenectomy performed, which can range from cursory sampling to systematic removal of all the lymphatic tissue surrounding the retroperitoneal vessels. Technique for pelvic lymph node dissection

The peritoneum overlying the pelvic wall is opened, extending from the round ligament to above the aortic bifurcation. The ureter, psoas muscle, genitofemoral nerve and iliac vessels are identified. The paravesical and pararectal spaces are developed. The paravesical space is lined medially by the bladder and vagina, laterally by the external iliac vessels and the obturator fossa, posteriorly by the cardinal ligament and anteriorly by the pubic ramus. The floor of the paravesical space is formed by the endopelvic fascia. The anterior paravesical space is opened by dissecting between the umbilical artery at the lateral border of the bladder and the external iliac vein. The umbilical artery is identified by retracting up on the lateral bladder peritoneum. The pararectal space is lined laterally by the hypogastric artery, medially by the rectum, posteriorly and superiorly by the sacrum, and inferiorly by the cardinal ligament. This space is developed, using

68

blunt dissection between the ureter and the hypogastric artery. This dissection should be performed with care to avoid dissecting lateral to the hypogastric artery, because of the risk of bleeding from the internal iliac venous system. Exposure of the operative field is maintained by using a fixed retractor – we prefer to use the Bookwalter retractor. Bleeding up to this point is usually minimal. An incision is made through the fatty tissue overlying the external iliac artery. The external and common iliac vessels are mobilized from the pelvic wall, first with scissors then digitally. A small branch to the psoas muscle is usually encountered and is clipped and divided. The obturator space is entered lateral to the vessels, and the obturator nerve is identified. Lymph node samples are removed from the external iliac, internal iliac and obturator regions (Figure 3.8). If bleeding occurs in the obturator space, it usually can be controlled by tamponade with a moist pack. Technique for para-aortic lymph node dissection

The peritoneal incision on the right begins at the infundibulopelvic ligament and crosses the vena cava and aorta and ends at the duodenum at the ligament of Treitz (Figure 3.9). The cecum and ascending colon are then mobilized from the loose underlying connective tissue. The lymph node dissection is then started on the vena cava from the mid-common iliac region extending superiorly. The nodal bundle is elevated and small vascular branches from the great vessels are clipped and divided. The next step is removal of the para-aortic lymph nodes to the left of the aorta. The left ureter is identified and retracted out of harm’s way. The sheath of the common iliac artery is incised and a plane between the lymph node-containing fat pad and the common iliac artery is developed. The dissection is carried superiorly to the level of the left renal vein. The origin of the left ovarian vein at the left renal vein is isolated, clamped and divided. The fat pad is clipped sequentially, medially and superiorly, and transected. The nodal bundle is elevated and posterior attachments are clipped and divided. The

Ch 03 final.qxd

12/18/2006

15:13

Page 69

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

Round ligament

Genitofemoral nerve Obturator fossa Psoas muscle

Ureter

External iliac artery and nerve

a

Obturator nerve

Hypogastric artery

b

Figure 3.8 Pelvic lymph node dissection. (a) The retroperitoneum has been opened, the ureter retracted medially and the para-vesicle and para-rectal spaces developed; lymph node-bearing tissue is sharply dissected from the underlying external iliac artery and vein. (b) The external iliac vessels are retracted laterally, exposing the obturator space; the lymph node-bearing tissue anterior to the obturator nerve is sharply dissected and removed

Duodenum

Superior mesenteric artery Duodenum

Pre-caval lymph nodes Right ureter

Para-aortic lymph nodes

Right renal vein Aorta

a

b

c

Inferior mesenteric artery

Figure 3.9 Para-aortic lymph node dissection. (a) The cecum and terminal ileum are mobilized by incising the peritoneum of the base of the small bowel mesentery up to the ligament of Treitz. (b) The ureter is retracted laterally and the lymph node-bearing tissue in the para-aortic and pre-caval fat pad is carefully dissected from the underlying vessels. (c) The duodenum is mobilized cephalad, exposing the high para-aortic nodes for removal

69

Ch 03 final.qxd

12/18/2006

15:13

Page 70

SURGERY FOR OVARIAN CANCER

operative field is irrigated to evaluate the level of hemostasis.

cauterized. Some gynecologic surgeons invert the appendiceal stump, while others do not.

Appendectomy

Adjuvant therapy

The appendix is a potential site for metastasis from ovarian malignancies.34 Nevertheless, routine appendectomy during primary surgical treatment for ovarian carcinoma is controversial. Malfetano identified no metastasis to the appendix in early-stage ovarian carcinomas, while it was noted in 70% of stage III and IV ovarian carcinomas.35 The high likelihood of appendiceal involvement has been reported by others as well.36,37 Bese et al. evaluated 90 patients with malignant ovarian carcinoma and found that none of the early-stage patients had metastases to the appendix.37 However, metastases to the appendix were detected in 21% of stage III and 50% of stage IV ovarian cancer patients. Others have reported metastases to the appendix in up to 70% of patients with advanced (stage III or IV) ovarian cancer. Some physicians have recommended routine appendectomy in women with mucinous cystadenocarcinomas, as these tumors can be metastatic from the gastrointestinal tract, and the appendix may harbor the primary lesion.35 Overall, the risk associated with an appendectomy during staging surgery for ovarian cancer appears to be low.34,37,38 Westermann et al. performed 233 appendectomies during extensive gynecologic procedures and reported no procedure-related morbidity.34 Therefore, appendectomy strictly for determining the extent of disease does not appear to be beneficial. However, appendectomy may have a role in patients with mucinous cancers or in those with advanced disease for cytoreduction.38 The technique for appendectomy depends on the location of the appendix. For a retrocecal appendix, the cecum may need to be mobilized by incising the peritoneum laterally. The mesoappendix and the appendiceal artery are sequentially clamped, cut and suture ligated (Figure 3.10). The base of the appendix is crushed and ligated with non-absorbable suture. The appendix is removed and the exposed mucosa is

As discussed above, patients with localized tumor have good overall survival.1 However, the specific subsets of patients likely to benefit from adjuvant chemotherapy are not clearly known. It is generally accepted that patients with stage IA or IB, grade 1 ovarian cancer comprise a ‘good prognosis’ group and can be safely treated with surgery only, and do not require adjuvant chemotherapy.5,39 Patients with

70

a

Appendiceal artery

b

Figure 3.10 Appendectomy. (a) The appendiceal artery is identified running in the base of the mesoappendix and clamped. (b) The mesoappendix is divided and ligated; the base of the appendix is ligated, divided, ligated and cauterized

Ch 03 final.qxd

12/18/2006

15:13

Page 71

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

stage IC or grade 3, and stage II ovarian cancer are generally considered to need adjuvant therapy because of the 30–40% risk of recurrence within 5 years. However, the role of adjuvant therapy in women with stage IA or IB, grade 2 ovarian cancer is unresolved, because of the small number of patients studied.1 The Gynecologic Oncology Group randomized 81 patients with stage IA or IB, grade 1 or 2 ovarian cancer to receive either no chemotherapy or melphalan.39 There were no differences in either disease-free survival or overall survival and both groups of patients had 5-year overall survival exceeding 90%.39 In a second trial, 141 patients with stage I, grade 3 or stage II patients were randomized to treatment with either melphalan or 32P.39 Again, there were no differences in the disease-free or overall survival. Bolis et al. evaluated 271 consecutive patients with stage I ovarian cancer in two randomized trials.40 Trial I compared six cycles of cisplatin chemotherapy to no further therapy in stage IA, IB, grades 2 and 3 ovarian cancer. Cisplatin significantly reduced the relapse rate by 65%, but the survival was not significantly different. In trial II, patients with stage IA (grade 2), IB and IC were randomized to receive cisplatin chemotherapy or 32P. Again, the relapse rate was decreased by 61%, but the overall survival was similar.40 Young et al. performed another randomized study through the Gynecologic Oncology Group (protocol no. 95), which compared intraperitoneal 32P versus intravenous cyclophosphamide (1 g/m2) and cisplatin (100 mg/m2) in ‘poor prognosis’ patients (stages IA and IB, grade 3 and IC, II).41 There was a better progression-free interval, and better treatment compliance with a lower complication rate in the chemotherapy arm. However, there was no difference in overall survival. In a subsequent Gynecologic Oncology Group study (protocol no. 157, patients with stage IA and IB (grade 3), IC and II epithelial ovarian cancer were randomized to receive either three or six cycles of paclitaxel (175 mg/m2) and carboplatin (AUC 7.5).42 Recurrence rates were not sig-

nificantly different between the two arms, but there was greater toxicity with six cycles of chemotherapy. Recently, two additional prospective randomized trials on early-stage ovarian cancer have been completed and published.43–45 The International Collaborative Ovarian Neoplasm (ICON1) and the Adjuvant ChemoTherapy in Ovarian Neoplasm (ACTION) trials were both designed to include a large number of patients. However, owing to slow accrual, the trials were stopped prematurely. The ICON1 trial (n = 477) enrolled all patients with early stages, grades, histologic cell types and those for whom the responsible physician was ‘uncertain whether the patients would benefit from immediate adjuvant chemotherapy’. The ACTION trial (n = 448) included patients with stage IA and IB, grades 2–3, all IC and IIA, and all I–IIA clear cell carcinomas of the ovary. Both studies recommended platinum-based chemotherapy, but the type of chemotherapy was not further specified. The results of the combined trials showed better overall survival with adjuvant chemotherapy compared to the observation arm (82% vs. 74%, p = 0.008).43,44 Recurrence-free survival was also better with adjuvant chemotherapy (76% vs. 65%, p = 0.001).43,44 However, when the trials were analyzed separately, differences in outcomes emerged. In the ACTION trial, there was improvement in recurrence-free survival with adjuvant chemotherapy, but the difference in overall survival was not significantly different.43,44 Further subset analyses revealed that stage I patients with complete staging did not benefit from chemotherapy, whereas those with incomplete staging (more likely to have higher stage disease) did have an improvement in both overall and recurrence-free survival. In the ICON1 trial, platinum-based adjuvant chemotherapy improved both overall and recurrencefree survival.45 The ICON1 trial probably included many patients with more advanced disease, since about 25% of patients were incompletely staged. Although the findings from these two trials suggest that platinum-based adjuvant chemotherapy probably has the greatest impact on incompletely staged

71

Ch 03 final.qxd

12/18/2006

15:13

Page 72

SURGERY FOR OVARIAN CANCER

patients, identification of the truly ‘high-risk’ earlystage patients is still needed. Collectively, the published data suggest that wellstaged patients with stage IA or IB, grade 1 ovarian cancer have a good prognosis and do not benefit from adjuvant chemotherapy. The role of adjuvant chemotherapy in grade 2 patients is not clearly known at present. All other early-stage and incompletely staged patients should be treated with platinum-based chemotherapy. Until further studies are completed, our practice is to treat these ‘high-risk’ early-stage patients with six cycles of paclitaxel and carboplatin chemotherapy.

SPECIAL CIRCUMSTANCES Low malignant potential (borderline) ovarian tumors Epithelial ovarian tumors of low malignant potential (LMP) were first described by Taylor in 1929 as a group of tumors with histologic features and biological behavior between benign and frankly malignant epithelial ovarian neoplasms.46 In 1971, FIGO included these tumors as a separate entity in its classification and staging system of gynecologic malignancies.47 About 3000 women are diagnosed with borderline ovarian tumors annually in the USA. Histological criteria for the diagnosis of borderline ovarian tumors include nuclear atypia, stratification of the epithelium, formation of microscopic papillary projections and the absence of stromal invasion. It is widely accepted that LMP tumors have a much better prognosis than invasive epithelial ovarian cancers. In a large retrospective study, Kaern et al. treated 370 patients with LMP ovarian tumors and reported a total mortality rate of 7.8% and total recurrence rate of 7.3% for all stages.48 Up to 90% of ovarian borderline tumors present with stage I disease and these patients have an excellent 5-year survival approaching 100%.49 Staging is of paramount importance, because it has consistently been shown to be a significant prognostic

72

factor.48,50–53 Some authors have reported that 18–20% of apparent stage I tumors are upstaged on final pathologic examination.53–55 There are many patients who are operated upon at local hospitals where a gynecologic oncologist may not be available to perform surgical staging. Lin et al. evaluated the adequacy of surgical staging of cases of serous borderline tumors referred for a second opinion.56 Most patients (78%) were treated primarily by general obstetrician–gynecologists. Only 12% had comprehensive surgical staging, as defined by biopsy samples taken from pelvic and abdominal peritoneum, omentum and retroperitoneal lymph nodes. There were substantial differences in the likelihood of staging based on the physician’s training: general surgeons performed complete staging in 0% of patients, obstetrician–gynecologists in 9% and gynecologic oncologists in 50%. Approximately 47% of patients who underwent biopsies were upstaged as a result of positive biopsies. However, surgical staging of patients with borderline ovarian tumors remains controversial. Unless there is gross residual disease, some authors feel that a restaging laparotomy is probably not needed.57 The most common histologic subtype of borderline ovarian tumors is serous. Serous LMP tumors have an overall risk of recurrence of approximately 10–15%.58–60 Patients with stage I ovarian LMP tumors have an excellent prognosis with a recurrence risk of less than 5%, and patients with stage II–IV have a risk of relapse of about 20%.57 Zanetta et al. followed 339 women with stage I–III LMP tumors prospectively and reported a 99.6% disease-free survival for stage I patients and 95.8% for stage II and III patients.57 A number of risk factors for recurrence have been identified including stage and residual disease.61–65 Intraoperative management of LMP tumors is similar to that for invasive ovarian carcinoma and should include systematic staging and upper abdominal exploration. These tumors can be large and a definitive diagnosis based on frozen section evaluation can be difficult. Under these circumstances, a frozen section diagnosis of ‘at least low malignant

Ch 03 final.qxd

12/18/2006

15:13

Page 73

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

potential’ should be adequate to proceed with surgical staging. Most studies have reported a high incidence of finding invasive foci in these tumors on final pathology.66,67 Menzin et al. studied 48 patients with a frozen section diagnosis of borderline malignancy and found that 27% had invasive carcinoma on final pathology.68 Houck et al. evaluated 140 women with borderline tumors of the ovary.67 Frozen section interpreted a benign lesion as borderline or borderline lesions as malignant when compared with permanent pathologic diagnoses in about 11%. Furthermore, malignant lesions were interpreted as borderline or borderline lesions read as benign by frozen section in 29%.67 Management of the adnexa is discussed in more detail below. The staging operation should include omentectomy, multiple peritoneal biopsies, and pelvic and para-aortic lymphadenectomy. Women with LMP tumors tend to present at younger ages and are more likely to have early-stage disease. Consequently, preservation of fertility is highly desirable by many women with LMP tumors. Several studies have examined the role of fertility preservation in these patients. Lim-Tan et al. followed 35 women with LMP tumors who underwent a unilateral or bilateral cystectomy.69 The overall recurrence rate was 12% and was mostly associated with positive margins and extent of disease. However, all patients were alive with an average follow-up of 7.5 years.69 In addition, several patients in this study had subsequent successful full-term pregnancies.69 Gotlieb et al. treated 39 patients with borderline ovarian tumors with conservative management.70 Fifteen women had 22 pregnancies after conservative treatment in this study. Morris et al. evaluated 43 patients with LMP tumors who underwent a cystectomy or oophorectomy with or without surgical staging.71 Overall recurrence was more frequent in patients with ovarian cystectomy (58%) than in those with unilateral oophorectomy (23%, p < 0.04). However, when patients with only stage I tumors are considered, the risk of recurrence was similar between the two groups. After treatment, 81% of the patients retained normal menstrual cycles

and 12 of 24 patients attempting pregnancy conceived 25 pregnancies. Collectively, these studies suggest that patients with ovarian borderline tumors have a high likelihood of achieving a successful pregnancy. The decision regarding conservative surgery in a young patient with an adnexal mass is based on several characteristics. If the mass is unilateral with surrounding normal ovarian tissue, then an ovarian cystectomy may be performed. The ovarian capsule should be inspected for any evidence of rupture, adherence, or excrescences. If no normal ovarian tissue is evident, then an oophorectomy or salpingooophorectomy is appropriate. The specimen should be submitted for frozen section analysis and the contralateral ovary is carefully inspected. Although some authors have recommended routine biopsy of a normal-appearing contralateral ovary, we prefer to leave it undisturbed, because of low yield and potential for peritoneal adhesions or inducement of ovarian failure. If bilateral adnexal masses are present, then intraoperative decision-making may be more difficult. In such cases, we recommend removing the more suspicious lesion first by conservative means (ovarian cystectomy, if possible) and sent for frozen section analysis. If an LMP tumor is diagnosed, then either ovarian cystectomy or oophorectomy should be performed on the contralateral side. As discussed above, systematic surgical staging should be performed in these patients. Ideally, the various intraoperative scenarios and possible treatment options should be discussed with the patient and her family preoperatively. Even for patients with stage II–IV LMP tumors who have not completed childbearing, conservative surgery may be appropriate. Treatment of the ovaries should be considered separately from treatment of extraovarian disease. Metastatic disease in ovarian LMP tumors tends to be of small volume and fully resectable in most patients.62 The outcome of patients with peritoneal implants remains quite good with surgical resection. In young patients for whom bilateral salpingo-

73

Ch 03 final.qxd

12/18/2006

15:13

Page 74

SURGERY FOR OVARIAN CANCER

oophorectomy cannot be avoided, uterine preservation should be considered because in vitro fertilization with donor oocytes may be an option in the future. Some authors have raised a question as to whether the uterus and the remaining ovary(ies) should be removed after childbearing has been completed for women treated conservatively for ovarian LMP tumors. There is no scientific basis for such a recommendation. For women with early-stage LMP tumors or more extensive disease, who have undergone surgical staging and resection of all gross disease, there is no known benefit for subsequent surgery. For women without full surgical staging at primary surgery, there may be potential prognostic and/or therapeutic benefits. The pros and cons of each approach should be carefully discussed with patients focusing on the risk–benefit ratio. There are limited data regarding retroperitoneal lymph node involvement in borderline tumors. Leake et al. evaluated 171 patients with borderline ovarian tumors and performed a full surgical staging on 34.55 The incidence of retroperitoneal lymphatic involvement was 21%.55 The occurrence of positive pelvic and para-aortic lymph nodes was 17% and 18%, respectively. Although the nodal status of patients did not significantly affect survival, those with positive lymph nodes did have a higher incidence of recurrence.55 However, some authors have questioned whether an observed focus of lymph node disease may be due to in situ transformation within the node, since no focus of invasion was found in the ovaries.72 Therefore, retroperitoneal lymph node sampling may provide some prognostic information regarding recurrence in patients with borderline ovarian tumors, but its impact on therapeutic intervention or survival has not been confirmed. However, because of the inaccuracy of intraoperative frozen section in predicting borderline pathology on the final pathologic diagnosis, as discussed above, it would be prudent to perform a full systematic staging including a retroperitoneal lymph node dissection.

74

Fertility preservation in patients with epithelial ovarian cancer Although most women with invasive ovarian cancer are postmenopausal, and in these women bilateral salpingo-oophorectomy and hysterectomy are standard treatment, a conservative treatment approach must be considered in young patients with stage I ovarian cancer. Conservative treatment approach is used here to denote surgery that preserves the reproductive potential without compromising the likelihood of cure. It has been estimated that 3–17% of all epithelial ovarian cancers occur in women < 40 years of age.73–77 Nevertheless, a staging procedure is necessary to confirm the early stage and to guide chemotherapy decisions. As discussed above, standard management of epithelial ovarian cancer involves abdominal hysterectomy, bilateral salpingooophorectomy and staging biopsies. However, many women with early-stage ovarian cancer wish to maintain reproductive capability. Although this has been considered by some physicians, there are limited data with regard to the safety of this approach.78–81 Schilder et al. evaluated 52 patients with stage I invasive epithelial ovarian cancer (42 with stage IA and ten with stage IC cancer), who underwent fertilitysparing surgery (conservation of the uterus, one ovary and fallopian tube).81 All patients underwent staging biopsies and 20 patients received adjuvant chemotherapy. Five patients developed tumor recurrence, but the estimated survival was 98% at 5 years and 93% at 10 years. Twenty-four patients attempted pregnancy and 17 (71%) conceived. These 17 patients had 26 term deliveries and five spontaneous abortions. Zanetta et al. performed 53 fertility-sparing procedures (unilateral salpingo-oophorectomy) and 46 hysterectomies with bilateral salpingo-oophorectomy.82 With a median follow-up of 7 years, recurrence was suffered by 9% of the conservatively managed patients and 11% of the radically managed patients. Despite the potential risk, several studies have shown no survival difference in stage IA ovarian cancer patients treated with unilateral versus bilateral salpingo-oophorectomy.81,83 Although these data are

Ch 03 final.qxd

12/18/2006

15:13

Page 75

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

encouraging, these are relatively small studies and it is difficult to draw definitive conclusions. Some authors have proposed specific criteria including young age, stage IA tumor, low grade and possibility of close follow-up78 to select low-risk patients for a conservative operation. Systematic secondary hysterectomy and adnexectomy after completion of childbearing is classically advocated when pregnancy is no longer a concern. As shown above, the risk of relapse after a conservative approach is low, especially in welldifferentiated tumors. The benefits of a secondary systematic radical operation in this setting are not clearly known. General criteria for conservative surgical treatment are listed in Table 3.4. Although strict criteria usually exclude patients with bilateral, adherent, or ruptured tumors, and individuals with ascites or positive cytologic washings, some studies suggest that such criteria can be liberalized, especially if adjuvant chemotherapy is used. The role of conservative management in patients with non-epithelial ovarian cancers is discussed later in this chapter. Emerging reproductive technologies are now allowing for pregnancies for patients who historically have not been able to maintain reproductive potential. For example, donor oocyte transfer and hormonal support allows a woman without ovaries to sustain a normal intrauterine pregnancy. Therefore, traditional guidelines for surgical management of ovarian

cancer patients may have to be modified for selected young individuals.

Germ cell tumors Malignant germ cell tumors represent about 10% of all ovarian tumors. Dysgerminoma is the most common malignant germ cell tumor. The median age of patients with ovarian germ cell tumors (OGCT) falls in the teenage years. Germ cell tumors are classified on the basis of the World Health Organization (WHO) classification introduced in 1973.84 The different types of OGCT are listed in Table 3.5. Bilateral ovarian involvement with tumor is uncommon except in pure dysgerminoma, in which 10–15% of tumors are bilateral. Bilateral ovarian involvement may also occur in cases of advanced-stage OGCT and in cases of mixed germ cell tumors that have a dysgerminoma component. Before the advent of effective chemotherapy, the prognosis of patients with these tumors was dismal, particularly for individuals with non-dysgerminoma and advanced cases.85,86 After the introduction of cisplatin-based chemotherapy, survival has improved dramatically and a reassessment of the extent of surgery has been required.

Table 3.5 World Health Organization classification of germ cell tumors Dysgerminoma

Table 3.4 Criteria for conservative surgical management of ovarian cancer patients Young patient desirous of future childbearing Patient and family consent and agreement to close follow-up No evidence of dysgenetic gonads Specific situations: any unilateral malignant germ cell tumor any unilateral stromal tumor any unilateral borderline tumor stage IA invasive epithelial tumor

Endodermal sinus tumor Embryonal carcinoma Polyembryoma Choriocarcinoma Teratomas immature mature (dermoid cyst) monodermal (struma ovarii, carcinoid) Mixed forms Gonadoblastoma

75

Ch 03 final.qxd

12/18/2006

15:13

Page 76

SURGERY FOR OVARIAN CANCER

Owing to the propensity of germ cell tumors to occur in young women, this diagnosis should be considered in women of reproductive age presenting with a pelvic mass. An algorithm for management of young patients with an ovarian mass is presented in Figure 3.11. Preoperative evaluation should include evaluation of serum tumor markers including β-human chorionic gonadotropin, α-fetoprotein, lactate dehydrogenase (LDH1 and LDH2 fractions), placental alkaline phosphatase and CA125. A chest X-ray is indicated, as germ cell tumors can metastasize to the lungs or mediastinum. Owing to the association of germ cell tumors with dysgenetic gonads (5% of all cases), a karyotype analysis should be considered for premenarchal girls with a pelvic mass of > 2 cm.87 The initial treatment of patients with germ cell tumors is surgical, which establishes the diagnosis and initiates therapy. A suggested algorithm for management of patients with an OGCT is presented in Figure 3.12. In patients suspected of having an OGCT, a laparotomy through a vertical incision should be per-

formed. The role of laparoscopy in the management of OGCT is undefined. If ascites is present, then the fluid should be evacuated and submitted for cytologic analysis. In the absence of ascites, a peritoneal washing from the abdomen and pelvis using saline should be performed and submitted for cytologic analysis. For most patients with germ cell tumors, unilateral salpingo-oophorectomy with preservation of the normal contralateral ovary, fallopian tube and uterus are appropriate, preserving the potential for fertility. Several studies have supported the principle of using conservative surgery for the vast majority of young patients with OGCT.85,88 In patients with limited disease, the role of fertility-sparing surgery is now well established.89 Peccatori et al. treated 100 patients with OGCT with fertility-sparing surgery.90 The overall survival was 96% with a mean follow-up of 55 months. Kurman and Norris evaluated 218 patients with OGCT and noted no worsening of outcome associated with conservative surgery in 182 patients

Tumor markers: AFP hCG LDH PLAP Premenarchal

Mass > 2 cm

Postmenarchal

Cystic > 8 cm Negative tumor markers

Cystic > 8 cm Solid, suspicious Positive tumor markers

Observation for 6–8 weeks

Karyotype

Decreasing size

Increasing size

Surgery

Clinical follow-up

Surgery

Figure 3.11 Algorithm for the management of the young patient with an ovarian mass. AFP, α-fetoprotein; hCG, human chorionic gonadotropin; LDH, lactate dehydrogenase; PLAP, placenta alkaline phosphatase

76

Ch 03 final.qxd

12/18/2006

15:13

Page 77

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

If the tumor appears confined to one or both ovaries, it is imperative to perform proper staging. Surgical staging is important to determine the extent of disease for prognostic reasons and for guiding postoperative adjuvant therapy. Without a careful staging operation, about 25% of women with stage IA dysgerminoma can relapse after conservative surgery.91 All areas of the abdomen and pelvis should be carefully inspected and palpated. Any suspicious areas should be resected if possible or biopsied. Peritoneal washings should be obtained for cytologic evaluation. Systematic staging biopsies including multiple peritoneal biopsies, omentum, bilateral pelvic and paraaortic lymph nodes should be performed. Systematic staging during primary surgical management is advocated by most gynecologic oncologists.89,92–94 However, it is controversial whether an individual should be restaged after inadequate first surgery. Some would argue that restaging is not necessary because OGCT are highly responsive to chemotherapy even at recurrence, and sensitive tumor markers are available for most patients.

who had tumors grossly confined to one ovary.85 These studies reveal at least equivalent sustained remission rates between conservative surgery and bilateral salpingo-oophorectomy with or without hysterectomy. If a unilateral mass suspected of being malignant is encountered at primary surgery in a young patient, a unilateral salpingo-oophorectomy with frozen section analysis should be performed. If the frozen section reveals a germ cell tumor, then the other ovary and uterus can be left in situ. Routine biopsy of the remaining ovary should be avoided, because this could lead to adhesion formation, ovarian failure and/or infertility. If bilateral ovarian masses are present, the more suspicious one should be removed for frozen section analysis. If there is a malignancy, then a bilateral salpingo-oophorectomy is indicated. Ovarian cystectomy of one or both ovaries may be considered in selected patients, although this surgical approach has not been extensively tested. Owing to the high cure rates with chemotherapy, this approach may be appealing for patients who desire to preserve fertility.

Germ cell tumor Grossly confined to ovary/ovaries

Metastatic disease

USO, inspect contralateral ovary

USO, remove readily resectable metastasis

Remove contralateral ovary if replaced by tumor or dysgenetic

BEP × 3 to 6 cycles

Surgical staging performed

? SLL if macroscopic residual with immature teratoma

Surgical staging not performed

No metastatic disease

All others

Observation

Adjuvant treatment

Stage 1A dysgerminoma or stage 1A, grade 1 immature teratoma

BEP × 3 cycles

CT or ultrasound scan of abdomen/pelvis q 3 mo × 12 mo q 6 mo × 12 mo

BEP × 3 cycles

Observation

If relapse BEP × 4 to 6 cycles

Figure 3.12 Algorithm for management of ovarian germ cell tumors. USO, unilateral salpingo-oophorectomy; BEP, bleomycin, etoposide, cisplatin chemotherapy, SLL, second-look laparotomy

77

Ch 03 final.qxd

12/18/2006

15:13

Page 78

SURGERY FOR OVARIAN CANCER

Despite the importance of complete surgical staging, inadequate staging at initial surgery remains relatively common.89 This presents a dilemma to the treating physician regarding postoperative treatment. Imaging with computed tomography should be performed. Most germ cell tumors are very chemosensitive; therefore, if chemotherapy is planned and there is no evidence of bulky residual disease, surgical exploration is probably unnecessary. Because of the increasing frequency of long-term survivors after combination treatment for OGCT, the implications of such treatment on ovarian function and reproductive potential have been addressed. Gershenson evaluated the menstrual and reproductive function of 40 patients who had been treated with conservative surgery followed by combination chemotherapy for malignant OGCT.95 Prior to treatment 83% reported regular menstrual cycles, and after treatment 68% had regular menstruation. Of the 16 patients who attempted pregnancy after completing chemotherapy, 11 delivered 22 healthy infants.95 Tangir et al. evaluated 64 patients treated with fertility-preserving surgery for OGCT.96 Thirty-eight attempted conception and 29 (76%) achieved at least one pregnancy. Among these patients, 25 had been treated with adjuvant combination chemotherapy. Thus, most women with OGCT will retain normal menstrual function and a reasonable probability of having normal offspring.

cyclophosphamide (VAC), or cisplatin, vinblastine and bleomycin (PVB) were used in the 1970s and 1980s with improved outcome. Subsequently, the combination of bleomycin, etoposide and cisplatin (BEP) was found to be more active.97,98 Gershenson et al. reported that 96% of OGCT patients treated with BEP remained in sustained remission.97 Similarly, Williams et al. reported a Gynecologic Oncology Group study in which 89 of 93 patients (96%) remained continuously disease free.98 Patients with dysgerminoma were historically treated with radiation; although dysgerminoma is very radiosensitive, fertility could not be preserved with such treatment. However, several studies over the past 12–15 years have demonstrated that BEP is a highly effective regimen in treatment of ovarian dysgerminoma.97,99,100 Based on these studies, chemotherapy has essentially replaced radiation therapy for treatment of ovarian dysgerminoma patients and 3–4 cycles of BEP appear to be adequate.

Sex cord–stromal tumors Ovarian sex cord–stromal tumors (OSCST) are diverse and relatively uncommon, and their classification is based on the WHO system (Table 3.6). A hallmark of their biologic behavior is the production of

Table 3.6 World Health Organization classification of sex cord–stromal tumors

Postoperative therapy

Prior to combination chemotherapy, the prognosis of patients with an OGCT was dismal. However, starting around 1970, combination chemotherapy was found to improve the outcome of OGCT patients. Currently, the only patients with an OGCT who should be considered for treatment with surgery alone are those with well-documented stage IA, grade 1 pure immature teratoma or stage I pure dysgerminoma. For all other patients with non-dysgerminoma tumors, postoperative chemotherapy is currently the standard treatment. Combination chemotherapy regimens including vincristine, dactinomycin and

78

A.

Granulosa stromal cell 1. Granulosa cell 2. Thecoma–fibroma

B.

Androblastomas; Sertoli–Leydig cell tumors 1. Well-differentiated (Pick’s adenoma, Sertoli cell tumor 2. Intermediate differentiation 3. Poorly differentiated 4. With heterologous elements

C.

Lipid cell tumors

D.

Gynandroblastoma

E.

Unclassified

Ch 03 final.qxd

12/18/2006

15:13

Page 79

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

steroid hormones including estrogens, androgens, progestins and cortisol. This hormonal production can result in clinical changes. Over one-third of patients with Sertoli–Leydig cell tumors will present with definite evidence of androgen excess in the form of virilization.101 Similarly, a high proportion of young patients with juvenile granulosa cell tumor of the ovary will present with isosexual pseudo-precocity manifested by various combinations of uterine bleeding, vaginal discharge, enlargement of breasts or labia, or the appearance of pubic hair, axillary hair, or both.102 Preoperative evaluation of patients with suspected sex cord–stromal tumors should include evaluation of potential serum tumor markers such as androstenedione, testosterone and inhibin levels. Granulosa cell tumor of the ovary is the most common type of sex cord–stromal tumor and accounts for 70% of the tumors in this category. This neoplasm is divided into adult and juvenile types based on different clinical and histopathologic features. Juvenile granulosa cell tumors (JGCTs) represent only about 5% in prepubertal girls and women younger than 30 years.103 JGCT patients typically present at an early stage and have a favorable prognosis, although those with more advanced disease may experience an aggressive clinical course.104 Most patients with adult granulosa cell tumors also present with early-stage disease, but relapses tend to occur in a more indolent

fashion. Surgery is the mainstay of initial management for patients with a suspected sex cord–stromal tumor of the ovary. A suggested algorithm for management of patients with sex cord–stromal tumors of the ovary is presented in Figure 3.13. Based on multiple studies of patients with granulosa cell tumors and Sertoli–Leydig cell tumors, about 95% of these tumors are unilateral and most are confined to one ovary.101,105–111 Therefore, fertility preservation with unilateral salpingo-oophorectomy seems to be appropriate management for young individuals with disease confined to the ovary. However, for older patients or those with bilateral ovarian involvement, abdominal hysterectomy and bilateral salpingo-oophorectomy is the treatment of choice. The surgical principles are generally similar to those used in management of other ovarian malignancies. The tumor should be submitted for frozen section analysis. If this is consistent with a sex cord–stromal tumor, the entire abdominal cavity and pelvis should be carefully inspected and palpated – any suspicious areas should be resected. If there is no gross extraovarian disease, then systematic staging biopsies, including multiple random peritoneal biopsies from the abdomen and pelvis, omentum, and pelvic and para-aortic lymph nodes, should be performed. In a patient of reproductive age with granulosa cell tumor of the ovary, if the uterus is not removed to maintain

Sex cord–stromal tumor

Grossly confined to ovary

Metastatic disease

USO, staging

TAH-BSO, surgical cytoreduction

Stage 1A (grade 1, 2)

Stage 1, grade 3 or ≥ stage II

Observation

BEP or other platinum-based chemotherapy or adjuvant radiotherapy

BEP or other platinum-based chemotherapy

Figure 3.13 Algorithm for management of ovarian sex cord–stromal tumors. USO, unilateral salpingo-oophorectomy, TAH-BSO, total abdominal hysterectomy–bilateral salpingo-oophorectomy; BEP, bleomycin, etoposide, cisplatin chemotherapy

79

Ch 03 final.qxd

12/18/2006

15:13

Page 80

SURGERY FOR OVARIAN CANCER

fertility, then a dilatation and curettage should be performed to look for possible concurrent endometrial adenocarcinoma.112 Granulosa cell tumors can secrete estrogen and, as a result of this hormonal secretion, an association between granulosa cell tumors with endometrial hyperplasia and adenocarcinoma has been reported.105,106,113–115 Endometrial carcinoma has been reported in 2–27% of patients with granulosa cell ovarian tumors.105,106,115,116 Although there is general agreement that patients with metastatic sex cord–stromal tumors should receive postoperative treatment, other criteria for postoperative treatment are not as well defined. The role for adjuvant chemo- or radiotherapy in stage I–II or completely resected stage III disease has not been defined. For granulosa cell tumors, prognostic factors may include stage, tumor rupture, age at diagnosis, tumor size, mitotic index, cellular atypia and DNA ploidy.105–107,116–126 Determination of these factors has been based on retrospective reviews, mostly based on univariate analyses. Therefore, the importance of individual factors is unknown. Candidates for postoperative treatment among patients with OSCST include patients with stage I poorly differentiated Sertoli–Leydig cell tumors, Sertoli–Leydig cell tumors with heterologous elements, or metastatic tumors of any histology. There are limited data regarding chemotherapeutic management of patients with OSCST. Nonplatinum-based chemotherapy regimens have limited activity against OSCST.88 However, platinumcontaining regimens appear to have the best response rates for patients with OSCST. Overall response rates approaching 63% have been reported with combination chemotherapy consisting of cisplatin, doxorubicin and cyclophosphamide (PAC) in patients with metastatic OSCST.127 Combination chemotherapy regimen of cisplatin, vinblastine and bleomycin (PVB) has been reported to result in response rates ranging from 57 to 92%; however, durable remissions occur in only about 50% of patients with OSCST.128–130 Eventually, PVB therapy was replaced with BEP due to a better toxicity profile. Gershenson

80

et al. reported the first use of this regimen for OSCST. There was an overall response rate of 83%, but only 14% had a durable remission.131 Homesley et al. reported on a phase II Gynecologic Oncology Group study of 75 patients with OSCST.132 Fifty-five patients were assessable for response, and 38 of these patients underwent second-look surgery. Fourteen (37%) of these 38 achieved a pathologic complete response at the time of second-look surgery. There were 17 patients who did not undergo a second-look operation but were clinically free of disease at the end of the study. The authors reported that 69% of patients with advanced tumors and 51% of those with recurrent disease remained progression-free, with a median follow-up of 3 years. Other treatment modalities such as radiotherapy and hormonal therapy with progestins or gonadotropin releasing hormone analogs have shown some activity in small studies, but their overall role in the treatment of OSCST is unclear. Patients with sex cord–stromal ovarian tumors require long-term follow-up because the median time to relapse can be over 4 years after diagnosis.103 The clinical follow-up of these patients should include pelvic examination and tumor marker studies that were elevated at original diagnosis.

CONCLUSION Surgery remains the cornerstone of initial diagnosis and treatment for women with suspected ovarian cancer. The majority of patients will benefit from a thorough staging evaluation, in terms of both determining prognosis and guiding subsequent therapy. The role of conservative surgical treatment for reproductive-age women with cancer confined to the ovary continues to evolve; however, current data suggest that fertility preservation is safe in appropriately selected patients. For the surgeon, an intimate knowledge of the surgical staging techniques and their clinical applications described in the preceding pages are prerequisites for the safe and successful management of patients with apparent early-stage ovarian cancer.

Ch 03 final.qxd

12/18/2006

15:13

Page 81

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

14.

van ‘t Riet M, Steyerberg EW, Nellensteyn J, et al. Meta-analysis of techniques for closure of midline abdominal incisions. Br J Surg 2002; 89: 1350–6

15.

Colombo M, Maggioni A, Parma G, et al. A randomized comparison of continuous versus interrupted mass closure of midline incisions in patients with gynecologic cancer. Obstet Gynecol 1997; 89: 684–9

16.

Gallup DG, Talledo OE, King LA. Primary mass closure of midline incisions with a continuous running monofilament suture in gynecologic patients. Obstet Gynecol 1989; 73: 675–7

17.

National Institute of Health Consensus Conference on Ovarian Cancer. Screening, treatment, and followup. NIH Consensus Development Panel of Ovarian Cancer. JAMA 1995; 273: 491–7

Soisson AP, Olt G, Soper JT, et al. Prevention of superficial wound separation with subcutaneous retention sutures. Gynecol Oncol 1993; 51: 330–4

18.

Buchsbaum HJ, Brady MF, Delgado G, et al. Surgical staging of carcinoma of the ovaries. Surg Gynecol Obstet 1989; 169: 226–32

Young RC, Decker DG, Wharton JT, et al. Staging laparotomy in early ovarian cancer. JAMA 1983; 250: 3072–6

19.

Bagley CM Jr, Young RC, Schein PS, et al. Ovarian carcinoma metastatic to the diaphragm – frequently undiagnosed at laparotomy. A preliminary report. Am J Obstet Gynecol 1973; 116: 397–400

Sainz de la Cuesta R, Goff BA, Fuller AF Jr, et al. Prognostic importance of intraoperative rupture of malignant ovarian epithelial neoplasms. Obstet Gynecol 1994; 84: 1–7

20.

Hart WR, Norris HJ. Borderline and malignant mucinous tumors of the ovary. Histologic criteria and clinical behavior. Cancer 1973; 31: 1031–45

8.

Piver SM, Barlow J, Lele SB. Incidence of subclinical metastasis in stage I and II ovarian carcinoma. Obstet Gynecol 1978; 52: 100–4

21.

9.

Piver SM. Optimal surgical therapy in stage I and II ovarian malignancies. Int J Radiat Oncol Biol Phys 1982; 8: 247–9

Smith JP, Day TG Jr. Review of ovarian cancer at the University of Texas Systems Cancer Center, M.D. Anderson Hospital and Tumor Institute. Am J Obstet Gynecol 1979; 135: 984–93

22.

Sigurdsson K, Alm P, Gullberg B. Prognostic factors in malignant epithelial ovarian tumors. Gynecol Oncol 1983; 15: 370–80

23.

Jadhon ME, Morgan MA, Kelsten ML, et al. Cytologic smears of peritoneal surfaces as a sampling technique in epithelial ovarian carcinoma. Obstet Gynecol 1990; 75: 102–5

REFERENCES 1.

Young RC, Pecorelli S. Management of early ovarian cancer. Semin Oncol 1998; 25: 335–9

2.

Jemal A, Murray T, Ward E, et al. Cancer statistics, 2004. CA Cancer J Clin 2005; 55: 10–30

3.

Goff BA, Mandel L, Muntz HG, Melancon CH. Ovarian carcinoma diagnosis. Cancer 2000; 89: 2068–75

4.

5.

6.

7.

10.

Eltabbakh GH, Yadav PR, Morgan A, et al. Clinical picture of women with early stage ovarian cancer. Gynecol Oncol 1999; 75: 476–9

Mayer AR, Chambers SK, Graves E, et al. Ovarian cancer staging: does it require a gynecologic oncologist? Gynecol Oncol 1992; 47: 223–7

11.

McGowan L, Lesher LP, Norris HJ, et al. Misstaging of ovarian cancer. Obstet Gynecol 1985; 65: 568–72

12.

Le T, Krepart GV, Lotocki RJ, et al. Clinically apparent early stage invasive epithelial ovarian carcinoma: should all be treated similarly? Gynecol Oncol 1999; 74: 252–4

24.

International Federation of Gynecology and Obstetrics. Changes in gynecologic cancer staging by the International Federation of Gynecology and Obstetrics. Am J Obstet Gynecol 1990; 162: 610–11

13.

Le T, Adolph A, Krepart GV, et al. The benefits of comprehensive surgical staging in the management of early-stage epithelial ovarian carcinoma. Gynecol Oncol 2002; 85: 351–5

25.

Benedetti-Panici P, Greggi S, Maneschi F, et al. Anatomical and pathological study of retroperitoneal nodes in epithelial ovarian cancer. Gynecol Oncol 1993; 51: 150–4

81

Ch 03 final.qxd

12/18/2006

15:13

Page 82

SURGERY FOR OVARIAN CANCER

26.

Chen SS, Lee L. Incidence of para-aortic and pelvic lymph node metastases in epithelial carcinoma of the ovary. Gynecol Oncol 1983; 16: 95–100

27.

Knapp RC, Friedman EA. Aortic lymph node metastases in early ovarian cancer. Am J Obstet Gynecol 1974; 119: 1013–17

28.

Lanza A, D’Addato F, Valli M, et al. Pelvic and paraaortic lymph nodal positivity in the ovarian carcinoma: its prognostic significance. Eur J Gynaecol Oncol 1988; 9: 36–9

29.

Walter AJ, Magrina JF. Contralateral pelvic and aortic lymph node metastasis in clinical stage I epithelial ovarian cancer. Gynecol Oncol 1999; 74: 128–9

30.

Burghardt E, Pickel H, Lahousen M, et al. Pelvic lymphadenectomy in operative treatment of ovarian cancer. Am J Obstet Gynecol 1986; 155: 315–19

31.

Burghardt E, Girardi F, Lahousen M, et al. Patterns of pelvic and paraaortic lymph node involvement in ovarian cancer. Gynecol Oncol 1991; 40: 103–6

32.

Cass I, Li AJ, Runowicz CD, et al. Pattern of lymph node metastases in clinically unilateral stage I invasive epithelial ovarian carcinomas. Gynecol Oncol 2001; 80: 56–61

33.

Petru E, Lahousen M, Tamussino K, et al. Lymphadenectomy in stage I ovarian cancer. Am J Obstet Gynecol 1994; 170: 656–62

34.

Westermann C, Mann WJ, Chumas J, et al. Routine appendectomy in extensive gynecologic operations. Surg Gynecol Obstet 1986; 162: 307–12

35.

Malfetano JH. The appendix and its metastatic potential in epithelial ovarian cancer. Obstet Gynecol 1987; 69: 396–8

36.

Sonnendecker EW. Is appendicectomy mandatory in patients with ovarian carcinoma? S Afr Med J 1982; 62: 978–80

37.

Bese T, Kosebay D, Kaleli S, et al. Appendectomy in the surgical staging of ovarian carcinoma. Int J Gynaecol Obstet 1996; 53: 249–52

38.

Rose PG, Reale FR, Fisher A, et al. Appendectomy in primary and secondary staging operations for ovarian malignancy. Obstet Gynecol 1991; 77: 116–18

39.

Young RC, Walton LA, Ellenberg SS, et al. Adjuvant therapy in stage I and stage II epithelial ovarian can-

82

cer. Results of two prospective randomized trials. N Engl J Med 1990; 322: 1021–7 40.

Bolis G, Colombo N, Pecorelli S, et al. Adjuvant treatment for early epithelial ovarian cancer: results of two randomized clinical trials comparing cisplatin to no further treatment or chromic phosphate (32P). Am J Clin Oncol 1995; 6: 887–93

41.

Young RC, Brady MF, Nierberg RM, et al. Randomized clinical trial of adjuvant treatment of women with early (FIGO I-IIA high risk) ovarian cancer – GOG n.95. Int J Gynecol Oncol 1997; 7: 17(Abstr)

42.

Bell J, Brady M, Lage J, et al. A randomized phase III trial of three versus six cycles of carboplatin and paclitaxel as adjuvant treatment in early stage ovarian epithelial carcinoma: a Gynecologic Oncology Group Study. Gynecol Oncol 2003; 88: 156

43.

Trimbos JB, Parmar M, Vergote I, et al. International Collaborative Ovarian Neoplasm trial 1 and Adjuvant ChemoTherapy In Ovarian Neoplasm trial: two parallel randomized phase III trials of adjuvant chemotherapy in patients with early-stage ovarian carcinoma. J Natl Cancer Inst 2003; 95: 105–12

44.

Trimbos JB, Vergote I, Bolis G, et al. Impact of adjuvant chemotherapy and surgical staging in early-stage ovarian carcinoma: European Organization for Research and Treatment of Cancer-Adjuvant Chemotherapy in Ovarian Neoplasm trial. J Natl Cancer Inst 2003; 95: 113–25

45.

Colombo N, Guthrie D, Chiari S, et al. International Collaborative Ovarian Neoplasm trial 1: a randomized trial of adjuvant chemotherapy in women with early-stage ovarian cancer. J Natl Cancer Inst 2003; 95: 125–32

46.

Taylor HC. Malignant and semi-malignant tumors of the ovary. Surg Gynecol Obstet 1929; 48: 701–12

47.

International Federation of Gynecology and Obstetrics. Classification and staging of malignant tumors in the female pelvis. Acta Obstet Gynecol Scand 1971; 50: 1–12

48.

Kaern J, Trope CG, Abeler VM. A retrospective study of 370 borderline tumors of the ovary treated at the Norwegian Radium Hospital from 1970 to 1982. A review of clinicopathologic features and treatment modalities. Cancer 1993; 71: 1810–20

Ch 03 final.qxd

12/18/2006

15:13

Page 83

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

49.

Barnhill DR, Kurman RJ, Brady MF, et al. Preliminary analysis of the behavior of stage I ovarian serous tumors of low malignant potential: a Gynecologic Oncology Group study. J Clin Oncol 1995; 13: 2752–6

62.

Gershenson DM, Silva EG, Levy L, et al. Ovarian serous borderline tumors with invasive peritoneal implants. Cancer 1998; 82: 1096–103

63.

Hopkins MP, Kumar NB, Morley GW. An assessment of pathologic features and treatment modalities in ovarian tumors of low malignant potential. Obstet Gynecol 1987; 70: 923–9

64.

Buttin BM, Herzog TJ, Powell MA, et al. Epithelial ovarian tumors of low malignant potential: the role of microinvasion. Obstet Gynecol 2002; 99: 11–17

65.

Bell DA, Weinstock MA, Scully RE. Peritoneal implants of ovarian serous borderline tumors. Histologic features and prognosis. Cancer 1988; 62: 2212–22

50.

Link CJ Jr, Reed E, Sarosy G, et al. Borderline ovarian tumors. Am J Med 1996; 101: 217–25

51.

Bostwick DG, Tazelaar HD, Ballon SC, et al. Ovarian epithelial tumors of borderline malignancy. A clinical and pathologic study of 109 cases. Cancer 1986; 58: 2052–5

52.

Massad LS Jr, Hunter VJ, Szpak CA, et al. Epithelial ovarian tumors of low malignant potential. Obstet Gynecol 1991; 78: 1027–32

53.

Trope C, Kaern J. Management of borderline tumors of the ovary: state of the art. Semin Oncol 1998; 25: 372–80

66.

Leake JF, Currie JL, Rosenshein NB, et al. Long-term follow-up of serous ovarian tumors of low malignant potential. Gynecol Oncol 1992; 47: 150–8

Haas JS, Mansfield CM, Hartman GV, et al. Results of radiation therapy in the treatment of epithelial carcinoma of the ovary. Cancer 1980; 46: 1950–6

67.

Houck K, Nikrui N, Duska L, et al. Borderline tumors of the ovary: correlation of frozen and permanent histopathologic diagnosis. Obstet Gynecol 2000; 95: 839–43

68.

Menzin AW, Rubin SC, Noumoff JS, et al. The accuracy of a frozen section diagnosis of borderline ovarian malignancy. Gynecol Oncol 1995; 59: 183–5

69.

Lim-Tan SK, Cajigas HE, Scully RE. Ovarian cystectomy for serous borderline tumors: a follow-up study of 35 cases. Obstet Gynecol 1988; 72: 775–81

70.

Gotlieb WH, Flikker S, Davidson B, et al. Borderline tumors of the ovary: fertility treatment, conservative management, and pregnancy outcome. Cancer 1998; 82: 141–6

71.

Morris RT, Gershenson DM, Silva EG, et al. Outcome and reproductive function after conservative surgery for borderline ovarian tumors. Obstet Gynecol 2000; 95: 541–7

72.

Shiraki M, Otis CN, Donovan JT, et al. Ovarian serous borderline epithelial tumors with multiple retroperitoneal nodal involvement: metastasis or malignant transformation of epithelial glandular inclusions? Gynecol Oncol 1992; 46: 255–8

73.

Duska LR, Chang YC, Flynn CE, et al. Epithelial ovarian carcinoma in the reproductive age group. Cancer 1999; 85: 2623–9

54.

55.

56.

57.

58.

Leake JF, Rader JS, Woodruff JD, et al. Retroperitoneal lymphatic involvement with epithelial ovarian tumors of low malignant potential. Gynecol Oncol 1991; 42: 124–30 Lin PS, Gershenson DM, Bevers MW, et al. The current status of surgical staging of ovarian serous borderline tumors. Cancer 1999; 85: 905–11 Zanetta G, Rota S, Chiari S, et al. Behavior of borderline tumors with particular interest to persistence, recurrence, and progression to invasive carcinoma: a prospective study. J Clin Oncol 2001; 19: 2658–64 Nikrui N. Survey of clinical behavior of patients with borderline epithelial tumors of the ovary. Gynecol Oncol 1981; 12: 107–19

59.

Gershenson DM, Silva EG. Serous ovarian tumors of low malignant potential with peritoneal implants. Cancer 1990; 65: 578–85

60.

Manchul LA, Simm J, Levin W, et al. Borderline epithelial ovarian tumors: a review of 81 cases with an assessment of the impact of treatment. Int J Radiat Oncol Biol Phys 1992; 22: 867–74

61.

Crispens MA, Bodurka D, Deavers M, et al. Response and survival in patients with progressive or recurrent serous ovarian tumors of low malignant potential. Obstet Gynecol 2002; 99: 3–10

83

Ch 03 final.qxd

12/18/2006

15:13

Page 84

SURGERY FOR OVARIAN CANCER

87.

Manuel M, Katayama PK, Jones HW, Jr. The age of occurrence of gonadal tumors in intersex patients with a Y chromosome. Am J Obstet Gynecol 1976; 124: 293–300

88.

Gershenson DM. Management of early ovarian cancer: germ cell and sex cord–stromal tumors. Gynecol Oncol 1994; 55: S62–72

89.

Gershenson DM. Update on malignant ovarian germ cell tumors. Cancer 1993; 71: 1581–90

90.

Peccatori F, Bonazzi C, Chiari S, et al. Surgical management of malignant ovarian germ-cell tumors: 10 years’ experience of 129 patients. Obstet Gynecol 1995; 86: 367–72

91.

Thomas GM, Dembo AJ, Hacker NF, et al. Current therapy for dysgerminoma of the ovary. Obstet Gynecol 1987; 70: 268–75

92.

Creasman WT, Soper JT. Assessment of the contemporary management of germ cell malignancies of the ovary. Am J Obstet Gynecol 1985; 153: 828–34

93.

Gonzalez-Lira G, Escudero-De Los Rios P, SalazarMartinez E, et al. Conservative surgery for ovarian cancer and effect on fertility. Int J Gynaecol Obstet 1997; 56: 155–62

Schwartz PE, Chambers SK, Chambers JT, et al. Ovarian germ cell malignancies: the Yale University experience. Gynecol Oncol 1992; 45: 26–31

94.

Schilder JM, Thompson AM, DePriest PD, et al. Outcome of reproductive age women with stage IA or IC invasive epithelial ovarian cancer treated with fertility-sparing therapy. Gynecol Oncol 2002; 87: 1–7

Curtin JP, Morrow CP, D’Ablaing G, et al. Malignant germ cell tumors of the ovary: 20-year report of LACUSC Women’s Hospital. Int J Gynecol Cancer 1994; 4: 29–35

95.

Gershenson DM. Menstrual and reproductive function after treatment with combination chemotherapy for malignant ovarian germ cell tumors. J Clin Oncol 1988; 6: 270–5

96.

Tangir J, Zelterman D, Ma W, et al. Reproductive function after conservative surgery and chemotherapy for malignant germ cell tumors of the ovary. Obstet Gynecol 2003; 101: 251–7

97.

Gershenson DM, Morris M, Cangir A, et al. Treatment of malignant germ cell tumors of the ovary with bleomycin, etoposide, and cisplatin. J Clin Oncol 1990; 8: 715–20

98.

Williams S, Blessing JA, Liao SY, et al. Adjuvant therapy of ovarian germ cell tumors with cisplatin, etoposide, and bleomycin: a trial of the Gynecologic Oncology Group. J Clin Oncol 1994; 12: 701–6

99.

Gershenson DM, Wharton JT, Kline RC, et al. Chemotherapeutic complete remission in patients with metastatic ovarian dysgerminoma. Potential for

74.

Plaxe SC, Braly PS, Freddo JL, et al. Profiles of women age 30–39 and age less than 30 with epithelial ovarian cancer. Obstet Gynecol 1993; 81: 651–4

75.

Rodriguez M, Nguyen HN, Averette HE, et al. National survey of ovarian carcinoma XII. Epithelial ovarian malignancies in women less than or equal to 25 years of age. Cancer 1994; 73: 1245–50

76.

Swenerton KD, Hislop TG, Spinelli J, et al. Ovarian carcinoma: a multivariate analysis of prognostic factors. Obstet Gynecol 1985; 65: 264–70

77.

Smedley H, Sikora K. Age as a prognostic factor in epithelial ovarian carcinoma. Br J Obstet Gynaecol 1985; 92: 839–42

78.

Raspagliesi F, Fontanelli R, Paladini D, et al. Conservative surgery in high-risk epithelial ovarian carcinoma. J Am Coll Surg 1997; 185: 457–60

79.

Miyazaki T, Tomoda Y, Ohta M, et al. Preservation of ovarian function and reproductive ability in patients with malignant ovarian tumors. Gynecol Oncol 1988; 30: 329–41

80.

81.

82.

83.

84.

Zanetta G, Chiari S, Rota S, et al. Conservative surgery for stage I ovarian carcinoma in women of childbearing age. Br J Obstet Gynaecol 1997; 104: 1030–5 Munnell EW. Is conservative therapy ever justified in stage I (IA) cancer of the ovary? Am J Obstet Gynecol 1969; 103: 641–53 Scully RE. World Health Organization classification and nomenclature of ovarian cancer. Natl Cancer Inst Monogr 1975; 42: 5–7

85.

Kurman RJ, Norris HJ. Malignant germ cell tumors of the ovary. Hum Pathol 1977; 8: 551–64

86.

Kurman RJ, Norris HJ. Endodermal sinus tumor of the ovary: a clinical and pathologic analysis of 71 cases. Cancer 1976; 38: 2404–19

84

Ch 03 final.qxd

12/18/2006

15:13

Page 85

MANAGEMENT OF EARLY-STAGE OVARIAN CANCER

cure and preservation of reproductive capacity. Cancer 1986; 58: 2594–9 100. Williams SD, Blessing JA, Hatch KD, et al. Chemotherapy of advanced dysgerminoma: trials of the Gynecologic Oncology Group. J Clin Oncol 1991; 9: 1950–5 101. Young RH, Scully RE. Ovarian Sertoli–Leydig cell tumors. A clinicopathological analysis of 207 cases. Am J Surg Pathol 1985; 9: 543–69

112. Segal R, DePetrillo AD, Thomas G. Clinical review of adult granulosa cell tumors of the ovary. Gynecol Oncol 1995; 56: 338–44 113. Salerno LJ. Feminizing mesenchymomas of the ovary. An analysis of 28 granulosa–theca cell tumors and their relationship to coexistent carcinoma. Am J Obstet Gynecol 1962; 84: 731–8 114. Dockerty MB, Mussey E. Malignant lesions of the uterus associated with estrogen-producing ovarian tumors. Am J Obstet Gynecol 1951; 61: 147–53

102. Young RH, Dickersin GR, Scully RE. Juvenile granulosa cell tumor of the ovary. A clinicopathological analysis of 125 cases. Am J Surg Pathol 1984; 8: 575–96

115. Gusberg SB, Kardon P. Proliferative endometrial response to theca–granulosa cell tumors. Am J Obstet Gynecol 1971; 111: 633–43

103. Schumer ST, Cannistra SA. Granulosa cell tumor of the ovary. J Clin Oncol 2003; 21: 1180–9

116. Ohel G, Kaneti H, Schenker JG. Granulosa cell tumors in Israel: a study of 172 cases. Gynecol Oncol 1983; 15: 278–86

104. Frausto SD, Geisler JP, Fletcher MS, et al. Late recurrence of juvenile granulosa cell tumor of the ovary. Am J Obstet Gynecol 2004; 191: 366–7

117. Pankratz E, Boyes DA, White GW, et al. Granulosa cell tumors. A clinical review of 61 cases. Obstet Gynecol 1978; 52: 718–23

105. Fox H, Agrawal K, Langley FA. A clinicopathologic study of 92 cases of granulosa cell tumor of the ovary with special reference to the factors influencing prognosis. Cancer 1975; 35: 231–41

118. Bjorkholm E, Silfversward C. Granulosa- and thecacell tumors. Incidence and occurrence of second primary tumors. Acta Radiol Oncol 1980; 19: 161–7

106. Stenwig JT, Hazekamp JT, Beecham JB. Granulosa cell tumors of the ovary. A clinicopathological study of 118 cases with long-term follow-up. Gynecol Oncol 1979; 7: 136–52 107. Dinnerstein AJ, O’Leary JA. Granulosa–theca cell tumors. A clinical review of 102 patients. Obstet Gynecol 1968; 31: 654–8 108. Evans AT, Gaffey TA, Malkasian GD, et al. Clinicopathologic review of 118 granulosa and 82 theca cell tumors. Obstet Gynecol 1980; 55: 231–8 109. Roth LM, Anderson MC, Govan AD, et al. SertoliLeydig cell tumors: a clinicopathologic study of 34 cases. Cancer 1981; 48: 187–97 110. Zaloudek C, Norris HJ. Sertoli–Leydig tumors of the ovary. A clinicopathologic study of 64 intermediate and poorly differentiated neoplasms. Am J Surg Pathol 1984; 8: 405–18 111. Young RH, Scully RE. Well-differentiated ovarian Sertoli-Leydig cell tumors: a clinicopathological analysis of 23 cases. Int J Gynecol Pathol 1984; 3: 277–90

119. Bjorkholm E, Silfversward C. Prognostic factors in granulosa-cell tumors. Gynecol Oncol 1981; 11: 261–74 120. Schwartz PE, Smith JP. Treatment of ovarian stromal tumors. Am J Obstet Gynecol 1976; 125: 402–11 121. Norris HJ, Taylor HB. Prognosis of granulosa–theca tumors of the ovary. Cancer 1968; 21: 255–63 122. Malmstrom H, Hogberg T, Risberg B, et al. Granulosa cell tumors of the ovary: prognostic factors and outcome. Gynecol Oncol 1994; 52: 50–5 123. Klemi PJ, Joensuu H, Salmi T. Prognostic value of flow cytometric DNA content analysis in granulosa cell tumor of the ovary. Cancer 1990; 65: 1189–93 124. Chadha S, Cornelisse CJ, Schaberg A. Flow cytometric DNA ploidy analysis of ovarian granulosa cell tumors. Gynecol Oncol 1990; 36: 240–5 125. Hitchcock CL, Norris HJ, Khalifa MA, et al. Flow cytometric analysis of granulosa tumors. Cancer 1989; 64: 2127–32 126. Swanson SA, Norris HJ, Kelsten ML, et al. DNA content of juvenile granulosa tumors determined by flow cytometry. Int J Gynecol Pathol 1990; 9: 101–9

85

Ch 03 final.qxd

12/18/2006

15:13

Page 86

SURGERY FOR OVARIAN CANCER

127. Gershenson DM, Copeland LJ, Kavanagh JJ, et al. Treatment of metastatic stromal tumors of the ovary with cisplatin, doxorubicin, and cyclophosphamide. Obstet Gynecol 1987; 70: 765–9

130. Zambetti M, Escobedo A, Pilotti S, et al. Cisplatinum/vinblastine/bleomycin combination chemotherapy in advanced or recurrent granulosa cell tumors of the ovary. Gynecol Oncol 1990; 36: 317–20

128. Colombo N, Sessa C, Landoni F, et al. Cisplatin, vinblastine, and bleomycin combination chemotherapy in metastatic granulosa cell tumor of the ovary. Obstet Gynecol 1986; 67: 265–8

131. Gershenson DM, Morris M, Burke TW, et al. Treatment of poor-prognosis sex cord–stromal tumors of the ovary with the combination of bleomycin, etoposide, and cisplatin. Obstet Gynecol 1996; 87: 527–31

129. Pecorelli S, Wagener P, Bonazzi C, et al. Cisplatin (P), vinblastine (V), and bleomycin (B) combination chemotherapy in recurrent or advanced granulosa cell tumor of the ovary (GCTO): an EORTC Gynecologic Cancer Cooperative Group Study. Proc Am Soc Clin Oncol 1988; 17: 147

86

132. Homesley HD, Bundy BN, Hurteau JA, et al. Bleomycin, etoposide, and cisplatin combination therapy of ovarian granulosa cell tumors and other stromal malignancies: a Gynecologic Oncology Group study. Gynecol Oncol 1999; 72: 131–7

Ch 04 final.qxd

12/18/2006

15:14

Page 87

CHAPTER 4

Cytoreductive surgery: principles and rationale Christine H Holschneider, Jonathan S Berek

PRIMARY CYTOREDUCTIVE SURGERY Historical perspective The history of primary surgery for ovarian neoplasms dates back to the 18th century. Until then, ovarian cysts were considered incurable. In 1701, a large cystic tumor had been evacuated successfully via laparotomy by Robert Houstoun of Scotland, but the ovary was left in situ.1 In 1775, William Hunter recommended aspiration of ovarian cysts, but did not perform the procedure.1,2 Removal of an ovarian cyst was performed by Johannes Theden and by Samuel Hartman d’Escher, in the years 1771 and 1807, respectively.2 The first published ovariotomy for ovarian cancer was performed in 1809 by Ephraim McDowell in Kentucky and reported in 1817 along with two other cases.3 Despite significant initial skepticism towards surgical extirpation of ovarian tumors voiced by physicians in Europe and the USA, oophorectomy for the treatment of ovarian tumors gained popularity and acceptance. By 1856, 212 oophorectomies had been reported, 8 years later 787.1 The modern approach to primary surgery for ovarian cancer evolved in the 20th century. Cytoreductive surgery for carcinoma of the ovary was first championed by Meigs in 1934.4 The value of excision of the omentum as a major site of tumor spread was recognized Pemberton in 1940.5 The concept that maximal surgical effort may improve survival goes back to the 1960s. In 1968, Christopher Hudson described a technique of radical oophorectomy for tumors fixed in the pelvis.6 In the same year, Munnell, who introduced the

concept of ‘maximum surgical effort’, reported improved survival in patients who had a ‘definitive operation’ during which most of the tumor was removed, versus those who had ‘partial removal’ or ‘biopsy only’.7 In 1975, a landmark study was published by Griffiths, that quantified residual disease and demonstrated an inverse relationship between residual tumor diameter and patient survival.8 This observation has since been confirmed by numerous other retrospective and prospective studies.9–17 The cancer commission of the League of Nations initiated talks on the formulation of an international classification in the late 1920s.2 This led to the staging for ovarian cancer being developed by the International Federation of Gynecology and Obstetrics (FIGO), which was last modified in 1988.18 Despite the substantial changes that have characterized the medical therapy for ovarian cancer in the past four decades, surgery continues to be the cornerstone in all treatment modalities. The theoretical basis and main rationale for performing tumor debulking in ovarian cancer rests on the kinetics of tumor growth, on the spread pattern of ovarian cancer and on the observation that epithelial ovarian cancer is fairly sensitive to chemotherapy.

Reasons for primary surgery in patients with suspected ovarian cancer Patients with suspected ovarian cancer should undergo primary surgery with four main objectives: diagnosis, staging, palliation and cytoreduction.

87

Ch 04 final.qxd

12/18/2006

15:14

Page 88

SURGERY FOR OVARIAN CANCER

Diagnosis

A definitive diagnosis of ovarian cancer is rarely made prior to surgery. Histological and molecular-biological information on the tumor tissue may help direct adjuvant therapy. Staging

The FIGO stage is a major prognostic factor in ovarian cancer (Table 4.1).18 Therefore, exact surgicopathological assessment of spread of the disease is important for counseling the patient regarding her prognosis and choosing the adjuvant therapy. Palliation of symptoms and short-term increase of quality of life

Ovarian cancer often remains undiagnosed until symptoms develop from metastatic disease. The bulk of tumor can cause pain, early satiety, paralytic ileus

and urinary symptoms. Ascites may lead to significant abdominal distension and shortness of breath. Possible palliative effects of primary surgery include temporary reduction of the ascites after removal of the primary tumor and omental cake, improved intestinal function with alleviation of nausea and early satiety, improved nutritional status and resolution of shortness of breath.19,20 Cytoreduction

The oncologically ideal, as potentially curative surgical approach to cancer, is the en-bloc resection of the tumor with wide margins of normal tissue. In many patients with ovarian cancer this is not attainable, owing to the existence of diffuse metastases to vital structures at the time of diagnosis. In these patients, the goal is to reduce the tumor burden as much as possible. This concept of maximal tumor resection

Table 4.1 International Federation of Obstetrics and Gynecology (FIGO) staging for ovarian cancer. From reference 18

Stage

Description

I IA IB IC*

Growth limited to the ovaries Growth limited to one ovary; no tumor on external surface; capsule intact; no malignant ascites Growth limited to both ovaries; no tumor on external surfaces; capsules intact; no malignant ascites Growth limited to one or both ovaries; but with tumor on external surfaces; or with capsule ruptured or with cytologically malignant ascites or washings

IIA IIB IIC*

Growth involving one or both ovaries with pelvic extension Extension and/or metastases to the uterus and/or tubes Extension to other pelvic tissue Extension to uterus and/or tubes and/or other pelvic tissue with tumor on the surface of one or both ovaries; or with capsule ruptured or with cytologically malignant ascites or washings

II

III IIIA IIIB IIIC IV

Tumor involving one or both ovaries with histologically confirmed peritoneal implants outside the pelvis and/or positive retroperitoneal or inguinal lymph nodes; metastases to the liver surface equal stage III Tumor is grossly limited to the true pelvis; with histologically confirmed microscopic seeding to abdominal peritoneal surfaces; lymph nodes are negative Abdominal peritoneal implants not exceeding 2 cm in diameter; lymph nodes are negative Abdominal peritoneal implants exceeding 2 cm in diameter and/or positive retroperitoneal or inguinal lymph nodes Presence of distant metastases including parenchymal liver metastases; if pleural effusions are present, there must be positive cytology to allot the case to stage IV

*FIGO recommends for stage IC and IIC to document if (1) source of the malignant cells was ascites or peritoneal washings and (2) if rupture of the capsule was spontaneous or iatrogenic

88

Ch 04 final.qxd

12/18/2006

15:14

Page 89

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

without expectation of cure by the surgical intervention alone but rather with the goal of enhancing the effectiveness of subsequent chemotherapy is called cytoreductive surgery or tumor debulking.

Theoretical rationale for cytoreductive surgery There are numerous theoretical biological advantages to cytoreductive surgery. Reduction in tumor mass prior to chemotherapy

Based on the fractional cell kill hypothesis of Skipper,21 which postulates that the proportion of tumor cells destroyed with each chemotherapy treatment is constant, fewer chemotherapy cycles would be needed to eradicate the cancer if the absolute number of cancer cells were less, provided that the growth fraction and phenotype were the same, and the cells were not resistant to chemotherapy. The larger the initial tumor burden, the longer the necessary exposure to the chemotherapeutic agent and the greater the risk of cancer cells developing acquired drug resistance. The spontaneous mutation rate of a malignancy is an inherent characteristic of genetically unstable tumor cells, and increases as the tumor size increases. Thus, the likelihood of developing phenotypic drug resistance is a function of both tumor size and mutational frequency.22 The larger a tumor the higher the chance that phenotypic drug resistance may have already developed before any exposure to chemotherapy. Drug exposure then merely provides a selective procedure, allowing the resistant cells to outgrow the sensitive tumor cell population.23 The role of cytoreductive surgery is to reduce the number of cells to a level where chemotherapy has a maximal chance of curing according to an optimal, generally multi-agent schedule while minimizing the chance of inducing drug resistance. In addition, clones of phenotypically chemotherapy-resistant cells may possibly be removed during the cytoreductive surgery, particularly if all gross macroscopic disease can be eliminated.

Improved tumor perfusion

For solid tumors, adequate drug diffusion to all areas is of paramount importance for effective cytotoxicity of chemotherapy. Large bulky tumors may have central areas that are hypoperfused, which tend to be exposed to suboptimal concentrations of the chemotherapeutic agent and are the most likely to demonstrate relative resistance to chemotherapy.24 Surgical resection of poorly vascularized tumors may eliminate those pharmacologic sanctuaries, thus increasing the likelihood of response to adjuvant therapy. Increased growth fraction

Cancer growth kinetics have been studied extensively in animal models and to a limited extent in humans. During initial cell divisions, growth approximates an exponential pattern. As the cancer grows larger, the rate of growth slows. This growth pattern of exponential tumor growth with exponential growth retardation is known as the Gompertzian growth curve (Figure 4.1). Large tumor masses have a decreased growth rate and are composed of a higher proportion of cells that are non-dividing or in the G0 phase of the cell cycle, which are essentially resistant to chemotherapy.25–27 Cytoreductive surgery resulting in smaller residual tumor masses with a relatively higher growth fraction may increase the likelihood of response to chemotherapy. A well-known stimulus for G0 cells to re-enter the cell cycle in normal tissue can be a sudden loss of actively cycling cells, owing either to surgical removal or to chemotherapy.26 Enhanced immunological competence of the patient

Large tumor masses appear to be more immunosuppressive than small tumors and may be less amenable to control by the host defense mechanisms. It appears that the normal mechanisms for abnormal antigen recognition may be overwhelmed and abrogated by the relatively large number of cancer cells, as excess antigen can block the function of cytotoxic lymphocytes. Large tumors may result in the production of

89

Ch 04 final.qxd

12/18/2006

15:14

Page 90

SURGERY FOR OVARIAN CANCER

Tumor volume

immunologically suppressive substances and in the induction of suppressor lymphocyte activity.28 A contribution of cytoreductive surgery could be to lower the level of immune suppression by resecting the bulk of the tumor.

Impact of residual disease Theoretical rationale for maximal cytoreduction

The following theoretical example conceptualizes the effect that debulking surgery may have on a 1-kg stage III ovarian cancer. Such a mass contains an estimated 1012 cells, a size that takes about 40 tumor doublings of the original tumor cell to achieve (Figure 4.2). Suboptimal resection of, for example, only half the tumor mass would achieve less than one log cell kill and it would take only one tumor doubling time to grow back to its former size, leaving no significant impact for patient survival. If the tumor is, however,

Time

Figure 4.1 Hypothetical Gompertzian tumor growth curve. Exponential

tumor

growth

with

exponential

growth

retardation. With permission from reference 25

1014 Number of cells 1012

1012 1010 Number of cells

Tumor mass 1 kg

10

10 109

108

1.3 × 108

6

10 g 1g 130 mg

6

10

10

Tumor first palpable (30 doublings)

1 mg

104 Tumor first visualized on X-ray (27 doublings)

2

10

1 5

10

15

20

25

30

35 40

50

Number of doublings

Figure 4.2 Theoretical tumor doubling curve assuming exponential tumor growth. With permission from reference 27

90

Ch 04 final.qxd

12/18/2006

15:14

Page 91

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

cytoreduced to 1 cm residual disease (109 cells), one has achieved a 3 log cell kill and the tumor would take 10 doublings to grow back to its original size.29 Considering Skipper’s fractional cell kill hypothesis,21 each chemotherapy course that produces a 90% cell kill would reduce the tumor cell population by one log. Thus, the combination of maximal cytoreduction to 1–2 cm residual disease followed by six cycles of chemotherapy would reduce the cancer cell population into the range between 101 and 104 cells, where prolonged survival can be measurably achieved.25 Considering re-growth of tumor cells occurring between cycles of chemotherapy, the theoretical benefit of cytoreductive surgery should be clinically more valuable in patients with less than 1 g (1 cm) residual disease, and most valuable in those with microscopic residual disease only. Terminology of residual disease

On theoretical grounds, survival should be related to the initial tumor mass, the total tumor mass left behind after primary surgery and the diameter of each individual tumor nodule left after tumor debulking. The amount of tumor left behind is prognostically more important for survival than the amount of tumor removed,30 and numerous studies have described the benefits of reducing the tumor burden to an ‘optimal status’. The definition of what constitutes optimal has varied over the decades as the philosophy about cytoreductive surgery has evolved. To this day, general agreement on the amount of residual tumor that constitutes optimal cytoreduction is lacking. Griffiths and Fuller had initially proposed that tumor and metastatic nodules should be reduced to ≤ 1.5 cm in diameter, and showed that survival was better in such patients (27 vs. 11 months’ median survival).31 Subsequently, Hacker et al.9 as well as van Lindert et al.32 showed that patients whose largest residual lesions were ≤ 5 mm following cytoreductive surgery had superior survival. The median survival with < 5 mm residual disease was 40 months, compared to 18 months with 5–15 mm residual disease,

and 6 months with > 15 mm residual disease.9 Principally, most studies denote optimal debulking as residual disease no larger than 1 or 2 cm and suboptimal debulking as residual disease exceeding 1 or 2 cm in diameter.33 The prognostic importance of only microscopic residual disease deserves further discussion. Gynecologic Oncology Group (GOG) data serve as one example illustrating that patients in whom the initial cytoreductive surgery leaves microscopic residual disease only, have better survival than patients with ≤ 1 cm, ≤ 2 cm, or > 2 cm residual disease (Figure 4.3).34 In this context, one area where any current definition of optimal cytoreduction is lacking is that of widespread small-volume peritoneal disease. Eisenkop et al. demonstrated that the removal of all small metastatic tumor nodules using laser, cavitron ultrasonic surgical aspirator (CUSA), argon beam coagulator, or sharp dissection can significantly improve survival over those patients who were cytoreduced to ≤ 1 cm residual disease, but whose small lesions were not resected.17 The survival impact of meticulously converting small-volume (≤ 1 cm) visible disease to microscopic residual disease has been confirmed in a subsequent study by Bristow and Montz, where median progression-free survival was 22.2 months versus 12.3 months for microscopic compared to 0.1–1.0 cm residual disease.35 It is difficult to know how to quantify confluent, thin, plaque-like disease; its correlation with survival continues to be uncertain and requires further study. Furthermore, there are data to suggest that even the number of tiny residual lesions is an important prognostic factor for survival, as demonstrated by Farias-Eisner et al. Extensive carcinomatosis was associated with a 17 months’ median survival, compared to 31 months and 57 months for minimal residual nodules and microscopic residual disease, respectively.36 While the best tumor debulking surgery is that to no visible residual disease and any numeric cut-off for ‘optimal’ residual disease is somewhat arbitrary, and the ultimate definition of optimal cytoreduction may need to be further refined, clearly defined criteria for

91

Ch 04 final.qxd

12/18/2006

15:14

Page 92

SURGERY FOR OVARIAN CANCER

optimal cytoreduction are critical for clinical research trials. In the 1970s, the GOG defined optimal cytoreduction as ≤ 2 cm; in the early 1980s, this was redefined as ≤ 3 cm for GOG protocol 47. For all clinical trials that followed, the GOG established ≤ 1 cm residual disease as the criterion for optimal cytoreduction for their trials.37

understood. Prospective data are insufficient to resolve that question and the prospect for randomized clinical studies is poor. In fact, the GOG attempted to study this question with a randomized trial comparing primary debulking surgery to six courses of chemotherapy in stage III ovarian cancer (GOG protocol 80). The study was closed, owing to low accrual of patients. Based on retrospective data, predictors of prognosis and predictors of optimal cytoreduction tend to be the same.10 Even though the amount of tumor left behind is prognostically more important for survival than the amount of tumor removed,30 the ability of cytoreductive surgery to impact on survival is still influenced by the extent of metastases prior to surgery. In 1983, Hacker et al. reported that patients with large metastatic disease had a worse prognosis than those with small initial tumor burden, even after accounting for the extent of cytoreduction.9 Another adverse prognostic factor related to tumor biology is the presence of peritoneal carcinomatosis, as shown

Surgery versus biology

One of the long-debated questions regarding the efficacy of primary cytoreductive surgery remains whether improved survival is related to the biology of the tumor itself or to the skill and philosophy of the surgeon. In other words, are the tumors, for which the surgeon is able to achieve optimal cytoreduction, of a different biology? Are they less infiltrative and thus more amenable to surgical resection? Are they less aggressive and thus associated with a better prognosis? Most experts currently assume that both surgery and biology have an impact on survival, yet the relative contribution of each of these factors is poorly

1.0 0.9

Proportion surviving

0.8 0.7 0.6 0.5 0.4 Protocol and 0.3

size of residual

0.2 0.1 0.0 0

Alive

Dead Total

PR 52, microscopic 41

56

97

PR 52, ≤ 1 cm

62

184

246

PR 97, < 2 cm

12

19

31

PR 97, ≥ 2 cm

55

208

283

6

12

18

24

30

36

42

48

Months from entry into study

Figure 4.3 Survival by residual disease based on patients in Gynecologic Oncology Group protocols (PR) 52 and 97. With permission from reference 34

92

Ch 04 final.qxd

12/18/2006

15:14

Page 93

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

by Heintz et al. in 1988 and later by Farias-Eisner et al., which may, even in the setting of metastatic lesions measuring < 5 mm, worsen survival.36,38 Indirectly to address the question of surgery versus biology further, Hoskins et al. re-analyzed data from 348 patients treated on GOG protocol 52, a randomized trial of cisplatin and cyclophosphamide with or without doxorubicin, which failed to demonstrate a significant difference in outcome. The survival of two groups of patients with optimal (≤ 1 cm) residual disease after primary surgery was compared. Patients with intra-abdominal metastases of 1 cm or less at the beginning of exploration were compared to patients whose larger intra-abdominal metastases were cytoreduced to ≤ 1 cm residual disease. If cytoreduction were the sole factor, both groups would be expected to have similar survival. However, patients who had only small-volume metastases at exploration had a better survival than those whose larger metastases were optimally cytoreduced to ≤ 1 cm (Figure 4.4). In addition

to patient age, other prognostic factors were the number of nodules and the tumor grade.14 Recent molecular biological data using gene expression arrays further suggest that the favorable prognosis associated with optimal debulking of ovarian cancer may be, at least in part, a function of the underlying biologic characteristics of these cancers.39 Although these observations point to the importance of tumor biology, they do not negate the effectiveness of cytoreductive surgery. There is little doubt that patients with small residual disease after surgery have a better prognosis than patients with bulky residual disease. Multiple retrospective studies have documented improved response rates to chemotherapy, longer progression-free survival and improved overall survival with optimal cytoreductive surgery. Thus, while surgery may not be able to change the inherent behavior of the tumor, it can, within the given biological tumor characteristics, improve prognosis and result in maximum benefit for patients.

1.0 0.9 0.8

Per cent surviving

0.7 0.6 0.5 0.4 0.3 Abdomen Alive

0.2 0.1

Dead Total

≤ 1 cm

57

91

148

> 1 cm

52

148

200

0.0 0

6

12

18

24

30

36

42

48

54

60

Months on study

Figure 4.4 Survival by initial maximum diameter of abdominal metastases (including omentum) in patients with optimal (≤ 1 cm) residual disease (Gynecologic Oncology Group 52; with permission from reference 14

93

Ch 04 final.qxd

12/18/2006

15:14

Page 94

SURGERY FOR OVARIAN CANCER

Response to chemotherapy

Numerous reports have documented improved response rates to chemotherapy with optimal cytoreduction. Table 4.2 summarizes studies40–48 evaluating complete pathological response rates to platinumbased chemotherapy as documented at second-look surgery in patients with no residual disease, optimally cytoreduced disease or suboptimally resected disease. The higher response rate to chemotherapy in patients with optimal (≤ 1 cm) residual disease after cytoreductive surgery continues to be observed with current first-line chemotherapy using a platinum compound and a taxane, as can be exemplified by a comparison of GOG 111 and GOG 158.49,50 Patients with suboptimally cytoreduced (> 1 cm residual disease) advanced ovarian cancer (GOG 111),49 who were treated with cisplatin and paclitaxel, demonstrated a 73% clinical response rate (51% complete clinical response and 22% partial clinical response). Complete pathological response as assessed at second-look surgery was 26%. This is in contrast to optimally resected advanced ovarian cancer treated with either

carboplatin and paclitaxel, or cisplatin and paclitaxel (GOG 158),50 where the complete pathological response rate at second look was almost twice as high (49%); a difference that translated into a median survival difference of 15 months (38 vs. 53 months, respectively).49,50 The importance of maximal primary cytoreduction is further underscored by the fact that even patients who achieve a complete pathologic response after primary therapy as established at second-look surgery are at increased risk of recurrence if they had large residual disease after primary surgery.51,52 Survival

Some authors have questioned the ability of cytoreductive surgery to improve the overall outcome of ovarian cancer and have expressed concern that these operations are excessively morbid and that modern chemotherapies may be sufficient.53,54 While no randomized studies exist that compare aggressive primary debulking with non-debulking of advanced ovarian cancer, there is a large body of retrospective and non-

Table 4.2 Complete pathological response rates to platinum-based chemotherapy as documented at second-look surgery in patients with microscopic, optimal and suboptimal residual disease Complete pathological response by degree of cytoreduction at primary surgery (%) Author, year

n

Microscopic residual disease

Optimal cytoreduction

Cohen, 198340

67

73

35

34

Podratz, 198541

118

82

44

39

Dauplat, 198642

51

85

73

19

Free 198743

89

75

36

9

Carmichael, 198744

146

62

30

25

Bertelsen, 198845

150

67

56

29

Sonnendecker, 198846

24

—

69

38

Ayhan, 199147

49

92

61

15

Katsoulis, 199748

115

—

76

45

Total

809

70

52

31

94

Suboptimal cytoreduction

Ch 04 final.qxd

12/18/2006

15:14

Page 95

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

randomized prospective data and two large metaanalyses that consistently show an inverse correlation between survival and amount of postoperative residual disease.33,55 This has been observed for both progression-free (Table 4.3)56–59 and overall survival (Table 4.4).13,15,30,34,56–67 Composite overall survival curves of GOG protocols 52 and 97 illustrate this significant inverse relationship between residual disease and overall survival (Figure 4.3).34 Data on progression-free survival from GOG protocol 158 demonstrate the significance of microscopic residual disease versus gross residual disease, even if optimally cytoreduced to ≤ 1 cm in diameter (Figure 4.5).50 The significance of no visible residual disease after primary cytoreduction has been further stressed in a recent prospective study of 408 patients by Eisenkop et al., where the median survival associated with microscopic residual disease was 76 months, compared to 32 months for visible ≤ 1 cm residual disease and 19 months for suboptimal > 1 cm residual tumor.30 The impact of maximal cytoreductive surgery on survival in the era of platinum-based chemotherapy has recently been quantified in a meta-analysis by Bristow et al. of 53 studies including 6885 patients.33 The authors confirmed a significant correlation between the per cent maximal cytoreduction and median survival after controlling for dose-intensity of Table 4.3 Progression-free survival (PFS) in patients with optimal versus suboptimal residual disease, receiving platinum-

the platinum compound administered, proportion of patients with stage IV disease, median age and year of publication. Each 10% increase in maximal cytoreduction was associated with a 5.5% increase in median survival time. Study cohorts in which there was a high proportion of maximal cytoreduction (> 75%) had a significantly better median survival compared to study cohorts with a ≤ 25% maximal cytoreduction rate (33.9 vs. 22.7 months). Maximal cytoreduction was one of the most powerful determinants of survival among patients with stage III or IV ovarian cancer.

Table 4.4 Overall survival (OS) in patients with optimal versus suboptimal residual disease, receiving platinum-based chemotherapy Median overall survival (months) Author, year

Cut-off (cm)

n

Vogl, 198356

2

12

> 40

26

15

3

34

38

52

26

2

60

> 40

65

16

1

62

> 42

127

22

Hainsworth, 198862 3

20

72

35

13

198958

3

17

45

39

23

1

91

48

258

20

1

75

> 60

88

25

1

137

34

113

16

1

343

40

294

22

1

55

57

81

32

2

25

> 38

66

22

Redman,

Conte, 198660 Neijt,

198761

Sutton,

Bertelsen, 199013 Potter,

based chemotherapy

198657

199163

Eisenkop, 199215 Median progression-free survival (months) Author, year

Cut-off (cm)

Vogl, 198356

2

Redman,

198657

Sutton, 198958 Baker, Total

199459

Optimal n PFS

Suboptimal n PFS

12

26

32

Hoskins,

199434

Baker, 199459 Del Campo,

199464

Suboptimal n OS

9

Makar, 199565

2

123

56

332

18

199766

2

105

> 60

191

18

2

138

> 30

14

14

1

392

72

16

19

1689

51

1797

20

3

34

23

52

16

Le,

3

17

33

39

16

Michel, 199767

1

Optimal OS

55

38

81

26

Eisenkop,

118

32

198

19

Total

200330

95

Ch 04 final.qxd

12/18/2006

15:14

Page 96

SURGERY FOR OVARIAN CANCER

1.0 Treat.-Res. Dis.

Proportion progression-free

0.9 0.8 0.7

PF

Failed

Total

Cisp/Taxol

-MRD

48

96

144

Cisp/Taxol

-GRD

36

220

256

Carbop/Taxol

-MRD

57

80

137

Carbop/Taxol

-GRD

38

217

255

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

6

12

18

24

30

36

42

48

54

60

Months on study

Figure 4.5 Progression-free (PF) survival in patients with stage III ovarian cancer, optimally cytoreduced to ≤ 1 cm residual disease, with microscopic residual disease (MRD) or gross residual disease (GRD) treated with paclitaxel (Taxol) and cisplatin (Cisp) or paclitaxel and carboplatin (Carbop) (Gynecologic Oncology Group 158; with permission from reference 50

Operative sequence of cytoreductive surgery: surgical approach The main goal of cytoreductive surgery is to resect all visible tumor. Primary surgical cytoreduction for ovarian cancer typically includes the performance of a total abdominal or supracervical hysterectomy, bilateral salpingo-oophorectomy, omentectomy and resection of any metastatic lesion. Most of the tumor is usually located in the pelvis. The cecum, terminal ileum, omentum, rectosigmoid and pelvic peritoneum are often involved, and there may be ascites. Multiple small tumor implants on the peritoneal surfaces and intestinal serosa may cause a segmental paralytic ileus. The supine position is adequate for most patients who will undergo cytoreductive surgery for ovarian cancer. However, for patients with extensive pelvic disease, for whom a low resection of the colon may be necessary, the low lithotomy position should be used to facilitate a rectosigmoid reanastomosis if necessary.

96

In order to gain adequate access to the pelvis and upper abdomen, a vertical midline incision should be used. This can be extended as needed from the symphysis pubis to the processus xyphoideus. After the peritoneal cavity is entered, ascites, if present, should be evacuated. The fluid should be sent for cytological evaluation if there is no overt evidence of metastatic disease. If there is no ascites and no evidence of metastases, abdominopelvic washings are performed and sent for cytology. A thorough inspection and palpation of the entire peritoneal cavity and retroperitoneum is carried out in order to assess the extent of the primary and metastatic disease. The localization and diameter of the primary tumor and its extension into surrounding organs is noted, as are the metastatic pattern throughout the whole abdominal cavity and the diameter of the largest metastases. The presence of an omental cake is noted. The bowel and its mesentery and all

Ch 04 final.qxd

12/18/2006

15:14

Page 97

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

other abdominal viscera are inspected and palpated to assess the extent of disease involvement and the possibility that the ovarian disease is metastatic, particularly from the stomach, colon or pancreas. The existence and extent of diaphragmatic disease involvement and peritoneal carcinomatosis is noted. The pelvic and para-aortic lymph nodes are palpated if possible. Sometimes these regions cannot be accessed before larger tumor masses are removed. If the disease appears to be of primary Müllerian origin, a careful assessment is then made of whether optimal cytoreduction is considered achievable. The resectability of metastatic tumor is usually determined by the location of the disease. Optimal cytoreduction may be difficult to achieve in the presence of bulky suprarenal lymphadenopathy, extensive disease in the liver parenchyma, along the root of the small bowel mesentery, close to the origin of the superior mesenteric artery, or in the porta hepatis. If optimal cytoreduction is deemed not to be achievable, extensive tumor debulking with bowel resection, urologic resection, splenectomy or diaphragm resection is not justified, unless the bowel resection is performed to overcome a bowel obstruction. However, even if optimal cytoreduction is not achievable, resection of the primary tumor and omental cake is generally both feasible and desirable. Advanced epithelial ovarian cancer often completely replaces the greater omentum, forming an ‘omental cake’, which may be adherent to the anterior abdominal wall, liver, spleen, intestines and pelvic tumor. The omental cake is often the first tumor encountered upon entering the peritoneal cavity. It is generally relatively easily mobilized and provides easily accessible tissue for rapid-frozen section if intraoperative pathologic consultation is deemed helpful. After mobilization of the omental edges and lysis of omental adhesions, the infracolic omentum is separated from the transverse colon and resected. For patients with only limited omental metastases, resection of the infracolic omentum is often sufficient to achieve optimal cytoreduction. If the omental metastases involve the gastrocolic ligament as well, it is

resected. Since epithelial ovarian tumors tend to spread along the gastrocolic ligament to the hilum of the spleen, special attention has to be paid to this region. The next step is to remove the primary tumor in the pelvis along with the other adnexa and uterus. If the primary tumor is limited to the ovary without extension to other organs with only limited metastases to the pelvic peritoneum, hysterectomy and bilateral salpingo-oophorectomy can be performed in the usual fashion. However, advanced ovarian cancer often results in a tumor mass involving both adnexae, the uterus, rectosigmoid, cecum, ileum and bladder. Metastases to the pelvic peritoneum often completely obliterate the anterior and posterior cul-de-sac, yet it is rare that the pelvic tumor cannot be optimally cytoreduced. The essential principle to removal of a more extensive pelvic tumor lies in the retroperitoneal approach.68 This approach takes advantage of the fact that even large bulky ovarian carcinomas usually do not infiltrate deeply through the peritoneum of the pelvic side wall or through the walls of the rectum or bladder. The retroperitoneal approach allows for mobilization of the pelvic mass using the peritoneum as a pseudo-capsule while identifying vital retroperitoneal structures such as the iliac vessels and ureter. The ovarian and uterine vessels should be ligated as early as possible during the dissection to prevent unnecessary blood loss. Depending on the accessibility of the posterior cul de sac, the hysterectomy can be accomplished using the usual antegrade or a retrograde approach. Because epithelial ovarian tumors, unlike gastrointestinal cancers, typically spread along the surface of the peritoneum and tend not to invade the lumina of the bowel or bladder, it is frequently feasible to remove the tumor without having to resect portions of the intestine or urinary tract.69–71 If the disease directly involves the bowel and its mesentery, most commonly encountered at the rectosigmoid colon, the cecum and terminal ileum, resection of the involved bowel segments may be necessary and is thus justified to achieve optimal tumor cytoreduction,

97

Ch 04 final.qxd

12/18/2006

15:14

Page 98

SURGERY FOR OVARIAN CANCER

either with primary reanastomosis or ostomy. In most cases where rectosigmoid resection is necessary, a low anterior resection of the colon en bloc with the tumor can be accomplished.72–77 The use of stapling devices, particularly the end-to-end anastomosis (EEA) stapler, has resulted in primary bowel reanastomosis being achieved in the vast majority of cases. Partial lower urinary tract resection involving a partial cystectomy, ureteroneocystotomy or ureteroureteral anastomosis may occasionally be necessary and indicated if optimal cytoreduction is to be achieved.71,78 Tumor spread to the hilum of the spleen may be carefully dissected off this site. Splenectomy may be indicated to achieve maximal tumor debulking in up to 10% of cytoreductive surgeries.79–82 Any other peritoneal tumor implants should be removed, particularly if they are large, isolated masses and their removal will render the patient optimally cytoreduced. If at all feasible, complete removal of all visible disease should be attempted. Use of the CUSA and the argon beam coagulator may help facilitate resection of small tumor nodules.35,83,84 As much tumor as possible is removed utilizing such techniques. Kapnick et al. showed that diaphragmatic tumor implants larger than 5 cm tend to penetrate the muscle.85 Thus, diaphragm resection may become necessary to achieve optimal cytoreduction.86,87 For staging purposes, pelvic and para-aortic lymphadenectomy is indicated in all those patients with apparent stage IIIB disease or less. In patients with more advanced disease, the lymph nodes should be carefully palpated and any enlarged node should be resected whenever possible to achieve optimal cytoreduction. For patients in whom intraperitoneal disease has been completely resected and the patient’s status is stable for additional surgery, some authors recommend performance of a complete pelvic and paraaortic lymphadenectomy. It is controversial whether systematic pelvic and para-aortic lymphadenectomy in this setting conveys a survival benefit. Lymph node involvement is seen in as many as 65–75% of patients with stage III or IV ovarian cancer.88–90 Retroperitoneal lymph nodes may be pharmacologic

98

sanctuaries not well reached by chemotherapy.91,92 Some small studies suggest a therapeutic value, but its magnitude is difficult to assess, owing to a number of confounders.93,94 Others found little therapeutic value in the removal of macroscopically negative lymph nodes in otherwise optimally debulked advanced ovarian cancer.89 An international randomized study is ongoing which compares systematic lymphadenectomy with debulking of grossly enlarged lymph nodes only for patients with optimally cytoreduced ovarian cancer (intraperitoneal disease < 1 cm). Preliminary results on 215 patients demonstrate lymph node metastases in 58% and a 2-year survival advantage for the systematic lymphadenectomy arm (85% vs. 70%).95 Patients with advanced ovarian cancer are at increased risk for wound dehiscence. Mass closure using the Smead–Jones technique as described by Morrow et al.96 or a modification thereof is advised. The operative report should describe in great detail the extent of disease at the beginning of the surgery, the respective FIGO stage, the procedures performed and most importantly the extent of residual disease at completion of the cytoreductive surgery, including its volume, number and distribution.

Degree of radicality to achieve cytoreduction Judging the degree to which an extensive surgical procedure for cytoreduction is feasible in any individual patient is one of the most difficult decisions a gynecologic oncologist must make, taking into account the probability of improving survival, the risk and magnitude of morbidity of the procedures necessary, and the effects on quality of life and cost. In general, every effort should be made to remove as much tumor as possible, since survival correlates inversely with extent of residual disease. In order to complete optimal cytoreduction, bowel resection may be needed in 14–33% of primary cytoreductive surgeries for ovarian cancer, splenectomy in 7–10% and urinary tract resection in up to 3%.30,69,70,79–82 While achieving maximum tumor reduction is an important principle in the surgical management of

Ch 04 final.qxd

12/18/2006

15:14

Page 99

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

advanced ovarian cancer, the extent to which the surgeon should go has not been addressed satisfactorily. In 1981, Wharton and Herson distinguished simple tumor removal (total abdominal hysterectomy, bilateral salpingo-oophorectomy, resection of peritoneal implants and omentum) and radical tumor reductive surgery, which may include resection of portions of the gastrointestinal or urinary tracts or extensive nodal metastases.97 Morrow and Curtin extended those definitions, as summarized in Table 4.5.98 Criteria as to when radical surgery, which might include a diaphragm resection, partial hepatectomy, subtotal colectomy or multiple segmental resections of the small bowel to achieve an optimal status of tumor reduction is warranted, have not been established. Since there is little evidence that cytoreduction with more than 2 cm residual disease has significant impact on survival (Figure 4.6), it appears indicated to perform extended procedures only if such will result in optimal cytoreduction of all tumor to less than 1–2 cm residual disease.

Morbidity In spite of several considerations in favor of cytoreductive surgery, its application in patients with advanced ovarian cancer who are frequently quite ill

even prior to going to the operating room continues to stir controversy. The criticism revolves around two major issues: the potentially high morbidity of the procedure and the feasibility of achieving optimal cytoreduction with metastatic disease. Overall, the morbidity of primary cytoreductive surgery for ovarian cancer depends not only on the extent of the surgical procedure but also on the patient’s overall medical condition and performance status, the experience and skill of the operating surgeon, as well as the quality of the supportive care, especially anesthesia and critical care. Complication rates vary considerably and do not appear to be systematically increased in series with high optimal cytoreduction rates. At least one complication is reported to occur in 32–67% of patients.99,100 Blood loss greater than 1000 ml is encountered in more than 20% of patients undergoing cytoreduction for ovarian cancer.99 Of the complications following primary cytoreductive surgery, pyrexia has been the most common, being reported in 9–53% of patients,100,101 followed by urinary tract infection in 3–23%,59,100 and paralytic ileus in 4–21%.10,15,59,89 One-third of women have more than one complication. Major morbidity rates are up to 15% and mortality rates average 1.8% (Table 4.6). Intestinal resection or lymphadenectomy

Table 4.5 Procedures considered extended tumor reductive surgery versus radical tumor reductive surgery. Modified from reference 98 Extended tumor reductive surgery

Radical tumor reductive surgery

Pelvis

Total or supracervical hysterectomy Bilateral salpingo-oophorectomy Sigmoid resection Resection of peritoneal nodules Peritoneal stripping Pelvic lymph node excision

Rectosigmoid resection Extensive pelvic lymphadenectomy Resection of ureteral/bladder segment Radical oophorectomy

Abdomen

Omentectomy Segmental bowel resection Splenectomy Resection of peritoneal nodules Para-aortic lymph node excision

Diaphragm stripping Extensive para-aortic lymphadenectomy Multiple or extensive bowel resection Resection of liver, kidney, diaphragm

99

Ch 04 final.qxd

12/18/2006

15:14

Page 100

SURGERY FOR OVARIAN CANCER

Size of residual

1.0 0.9

Proportion surviving

0.8 0.7

Alive Dead Total

< 2 cm

12

19

31

2.0–3.9

20

70

90

4.0–5.9

17

63

80

6.0–9.9

11

47

58

≥ 1 cm

7

28

35

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

6

12

18

24

30

36

42

48

Months from entry into study

Figure 4.6 Survival rates of patients with advanced-stage ovarian cancer by maximum diameter of residual disease (Gynecologic Oncology Group 97; with permission from reference 34

Table 4.6 Morbidity associated with primary cytoreductive surgery for advanced ovarian cancer based on cumulative data from ten studies10,11,15,59,67,89,99–102 encompassing 1411 patients Complication

Average rate (%)

Range (%)

Minor infectious morbidity (wound or urinary tract infection, cuff cellulitis)

17

7–42

Severe infectious morbidity (pneumonia, sepsis)

5

5–15

Cardiac morbidity (congestive heart failure, myocardial infarction)

1.1

0.7–7

Pulmonary (ARDS, intubation > 48 h)

1.4

1–6

Thromboembolic (deep venous thrombosis, pulmonary embolism)

2

0–4

Coagulopathy

2

1.9–2.6

Gastrointestinal (prolonged ileus, small bowel obstruction not requiring re-operation)

9

6–22

Renal failure

0.5

0–2

Central nervous system (cerebrovascular accident, prolonged delirium)

0.8

0.7–4

Re-laparotomy (bleeding, bowel obstruction, fascial dehiscence, fistula)

3

0–25

1.8

0–6.1

Mortality (within 28 days postoperatively) ARDS, adult respiratory distress syndrome

100

Ch 04 final.qxd

12/18/2006

15:14

Page 101

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

in these patients does not appear to increase the overall morbidity.72,73,77,89,99 Similarly, series that focus on splenectomy as part of cytoreductive surgery for ovarian cancer do not report increased overall morbidity rates. However, certain specific postoperative complications are more frequent in these patients; these include left-sided pleural effusion, transient leukocytosis and thrombocytosis, as well as pancreatitis, pancreatic tail injury and pseudocyst formation.79,81,103 To reduce the risk of overwhelming post-splenectomy infection with encapsulated bacteria, a rare but frequently lethal complication, vaccination against Streptococcus pneumoniae, Haemophilus influenza and Neisseria meningitidis is recommended. While these studies do give a general indication of the associated risks, it is important to recognize that these numbers are derived from retrospective, often small series of a great diversity of patients with variable medical conditions and performance status undergoing a variety of concomitant procedures, and lacking a control population. Generally speaking, an acceptable rate of major complications should be less than 10–15% and should not delay chemotherapy.

Feasibility/achievability The detection of a pelvic or adnexal mass, especially in combination with an omental cake or ascites, is suspect for ovarian cancer. The preoperative diagnostic evaluation of a patient should focus on detecting the extent of disease, on the exclusion of metastases from non-Müllerian neoplasms, such as of the gastrointestinal tract, pancreas or breast, and should assess the patient’s fitness for surgery. If imaging studies reveal, for example, extensive parenchymal liver or pulmonary metastases, optimal cytoreduction may be determined not to be feasible even prior to exploratory celiotomy. For those patients who are fit to undergo surgery and who are candidates for cytoreduction, analysis of retrospective or prospective cohort data shows a great variability of optimal tumor reduction rates (Table 4.7), ranging from 25 to 99%. The feasibility of optimal cytoreduction depends on the disease distribu-

tion, the patient’s overall medical condition and performance status, and the operating surgeon’s philosophy, experience and skill. The highest success rates (> 70%) for optimal tumor debulking to less than 1–2 cm residual disease are observed if the surgery is performed by a gynecologic oncologist. Cytoreduction to the true optimal residual disease, namely no visible residual disease, is also achieved more frequently if the surgeon is specialized in gynecologic oncology (48% vs. 12% in series that include variable surgeons). If the two series by Eisenkop that report exceptionally frequent achievement of no visible residual disease are excluded,17,30 the reported microscopic residual disease rate in the hand of gynecologic oncologists averages 29%, compared to 12% in series including obstetrician gynecologists or general surgeons. Nevertheless, even in the most expert hands, there remains a subgroup of patients who cannot be optimally primarily cytoreduced and many of these patients may be better served by neoadjuvant chemotherapy and interval cytoreduction if there is clinical evidence of the tumor’s response to chemotherapy. The challenge lies in predicting in whom optimal cytoreduction is likely to be feasible (see below, section on neoadjuvant chemotherapy and predictors for optimal versus suboptimal cytoreduction, page 111).

Cytoreductive surgery in stage IV disease FIGO stage IV ovarian cancer encompasses a wide spectrum of disease and should be managed on an individualized basis when considering the question of resectability. For example, patients who have stage IV disease due to pleural effusions or an isolated fully resectable anterior abdominal wall or peripheral parenchymal liver mass may be excellent candidates for optimal cytoreduction. This has been substantiated by several studies indicating that patients with stage IV disease who undergo optimal cytoreduction have more than double the survival of patients with suboptimal disease. Median survival for patients with optimally cytoreduced stage IV disease is 34 months,

101

Ch 04 final.qxd

12/18/2006

15:14

Page 102

SURGERY FOR OVARIAN CANCER

Table 4.7 Feasibility of optimal primary cytoreduction in advanced-stage ovarian cancer Cytoreduction to Author, year

n

Optimal residual disease (%)

Variable surgeons (including general gynecologists and general surgeons) Neijt, 198761 191 49 Heintz, 198838 65 54 Bertelsen, 199013 360 25 Marsoni, 1990104 914 27 Potter, 199163 163 46 Eisenkop, 199215 263 54 Del Campo, 199464 91 27 Venesmaa, 1994105 264 36 Baker, 199459 136 40 Makar, 199565 455 27 Junor, 1999107 768 27 Olaitan, 2001106 281 55 Total (range)

3951

Surgeons specialized in gynecologic oncology Chen, 1985100 47 Heintz, 198610 70 Michel, 199767 152 Eisenkop, 199817 163 Vergote, 2000108 112 Eisenkop, 200330 408 Total (range)

952

11 14 8 — 27 12 8 — — 10 — —

34 (25–55)

12 (8–27)

98 70 91 99 89 96

— 4 30 85 43 86

93 (70–99)

65 (4–86)

compared to 14 months for those with suboptimal residual disease (Table 4.8). Several series have found similar optimal cytoreduction rates for patients who had pleural effusions only versus those who had stage IV disease due to other metastases.110, 113 The survival advantage for optimally cytoreduced patients appears to be independent of the metastatic site that led to the stage IV designation.109,110 For example, Munkarah et al. compared patients with stage IV due to pleural effusion only with those due to other sites and found that median survival for the optimally cytoreduced patients did not differ (25 vs. 23 months).109 Even patients with localized resectable or multiple small (< 1–2 cm) unresectable liver metastases may be good candidates for optimal primary surgical cytoreduction. By contrast, the patient with multiple

102

No residual disease (%)

Cut-off (cm) for optimal 2 1.5 1 2 1 1 2 2 1 2 2 2

1.5 1.5 2 1 1.5 1

large parenchymal lung or liver metastases may be unlikely to undergo optimal primary tumor debulking and may be better served by treatment with neoadjuvant chemotherapy followed by interval cytoreductive surgery, provided the metastases respond to chemotherapy. However, even the latter point is not unequivocal and the presence of hepatic metastases should not exclude the patient a priori from exploration for cytoreduction. For example, Bristow et al. demonstrated that even in the setting of parenchymal liver metastases, optimal extrahepatic cytoreduction was still associated with a survival advantage. In their series of 84 patients with stage IV disease, individuals with optimal hepatic and extrahepatic cytoreduction had the best median survival of 50 months, compared to 27 months for those with optimal extrahepatic debulking but suboptimal intrahepatic disease, and 8

Ch 04 final.qxd

12/18/2006

15:14

Page 103

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

Table 4.8 Cytoreductive surgery in stage IV ovarian cancer Optimal debulking

Median survival (months)

Author, year

n

n

%

Optimal

Suboptimal

Cut-off (cm) for optimal

Munkarah, 1997109

92

31

34

25

15

2

Curtin, 1997110

92

41

45

40

18

2

Lui, 1997111

47

14

30

37

17

2

Bristow, 1999112

84

25

30

38

10

1

Naik, 2000113

37

16

43

25

8

2

352

127

36

34

14

Total

months for those patients who were left with suboptimal disease both within and outside the liver.112

Access to care and the contemporary ovarian cancer team Surgical evaluation of a pelvic mass is one of the most common indications for gynecologic surgery and therefore it is unlikely that all patients with an adnexal mass will be referred to a gynecologic oncologist. The obvious principle reason for such referral would be the risk of ovarian cancer, the surgical management of which significantly benefits from the expertise of a trained gynecologic oncologic surgeon. Existing clinical data led to the 1995 National Institutes of Health (NIH) consensus panel opinion recommending that preoperative consultation with a gynecologic oncologist should be offered to all women with a suspected ovarian malignancy, as ‘aggressive attempts at cytoreductive surgery as the primary management of ovarian cancer will improve the patient’s opportunity for long-term survival.’114 Failure to refer a patient with suspected ovarian carcinoma to a gynecologic oncologist is one of the most controversial issues about the initial surgical management of the disease and in many regions referral patterns to gynecologic oncologists remain poor. Hospital audits have shown that women are frequently not properly staged and do not undergo max-

imal primary debulking surgery. A US patterns of care study of 785 women diagnosed with ovarian cancer and identified through the National Cancer Institute’s (NCI’s) Surveillance, Epidemiology, and End Results (SEER) Program was published in 1997 and indicated that only 9% and 14% of patients with stage I and II ovarian cancer, respectively, were treated appropriately according to NIH recommendations. Only 71% of women with stage III and 51% of women with stage IV received the recommended surgery and chemotherapy.115 Based on a national survey on ovarian cancer conducted by the American College of Surgeons Commission on Cancer, comprehensive surgical staging of ovarian cancer is completed routinely by gynecologic oncologists (97%), less frequently by gynecologists (52%) and uncommonly by general surgeons (36%), probably resulting in an underestimation of the extent of disease in a substantial proportion of patients.116 A similar trend is seen in pattern of care studies when metastatic ovarian cancer is found. Gynecologic oncologists are the most likely to achieve optimal surgical cytoreduction (Table 4.7) as compared to obstetrician gynecologists and general surgeons.15,106,107 General surgeons are the least likely to perform complete debulking when compared to gynecologists and gynecologic oncologists (25% vs. 45%).117

103

Ch 04 final.qxd

12/18/2006

15:14

Page 104

SURGERY FOR OVARIAN CANCER

The fact that fewer than one-half of patients with ovarian cancer undergo optimal primary cytoreduction may reflect the continued doubt by some oncologists and non-specialist gynecologists as to the value of maximal debulking surgery. A Canadian study further illustrates the issue, where 260 obstetrician gynecologists responded to five theoretical case scenarios of ovarian neoplasms as to whether they would operate on these patients themselves or refer to a gynecologic cancer center; 43% of gynecologists indicated that they would operate on a patient whose history was highly suggestive of frank ovarian cancer.118 A recent study on volume-based access to ovarian cancer surgery in Maryland underscores the issue even further: 91% of the pool of surgeons who performed ovarian cancer surgeries between 1990 and 2000, performed on average one such procedure per year. Less than 40% of surgeries for ovarian cancer were performed by surgeons who performed at least ten such procedures annually.119 These data indicate that, despite the fact that optimization of the care for the patient with ovarian cancer depends on the diligence of her primary care physician or general gynecologist to refer her to a gynecologic oncologist for surgery, such referral often does not occur. Based on the American College of Surgeons Commission on Cancer study published in 1993, obstetrician–gynecologists provided surgical care to 45%, general surgeons to 21% and gynecologic oncologists to 21% of the 5156 patients with ovarian cancer surveyed.117 Despite some improvements in primary access to gynecologic oncologists, subsequent patterns of care studies have confirmed that, during the past decade, the number of patients with ovarian cancer initially operated upon by a gynecologic oncologist remains below 50%.106,119,120 Numerous studies have demonstrated that care provided by physicians specialized in gynecologic oncology is associated with enhanced survival for patients with ovarian cancer.15,107,120,121 A study from the UK with data from 1866 patients demonstrated, after adjusting for differential use of chemotherapy, a 25% reduction in death rate when women with stage

104

III ovarian cancer were operated upon by gynecologic oncologists rather than general gynecologists.107 A similar survival advantage for stage III patients treated by gynecologic oncologists compared to general surgeons was found in the American Cancer Society (ACS) Commission on Cancer Study.117 Centralization of primary ovarian cancer surgery in Norway has improved median survival for patients with advanced ovarian cancer from 12 months to 21 months.122 In a survey of patients with stage IIIC ovarian cancer treated at 14 US hospitals, optimal cytoreduction at the time of primary surgery and overall survival were both significantly associated with gynecologic oncology specialty training of the operating surgeon (Figure 4.7).15 Bristow et al. performed a meta-analysis of maximal cytoreduction and survival, which provided further indirect insights into the impact of care by a specialist team on survival of women with suspected ovarian cancer. Study cohorts in which there was a high proportion of maximal cytoreduction (> 75%), a surrogate for care by a subspecialist trained in gynecologic oncologic surgery, had a 50% increase in median survival compared to study cohorts with a ≤ 25% maximal cytoreduction rate (33.9 vs. 22.7 months; Figure 4.8).33 A multifaceted approach to improving access to specialist care for women with ovarian cancer is underway and includes the establishment of practice guidelines and criteria for appropriate referrals, as well as educating professionals and the lay public, elected officials and managed care executives about the best care for women with ovarian cancer. To assist in the referral process and in the selection of cases for which a gynecologic oncologist should at least be on stand-by, the Society of Gynecologic Oncologists has developed guidelines for the referral of patients with a pelvic mass or suspected ovarian cancer.123 Table 4.9 summarizes the specific clinical characteristics that suggest a higher risk of malignancy in which pretreatment referral to or consultation with a gynecologic oncologist is recommended.

Ch 04 final.qxd

12/18/2006

15:14

Page 105

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

1.0

Cumulative survival rate (%)

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

20

Time (months)

40

60

Figure 4.7 Cumulative proportion of survival subdivided by subspecialty training of surgeon present. –––, gynecologic oncologist present; …., gynecologic oncologist absent. With permission from reference 15

40 38

Median survival time (months)

36 34 32 30 28 26 24 22 20 0

20

40

60

80

100

Maximum cytoreductive surgery (%)

Figure 4.8 Meta-analysis of 81 patient cohorts reported in the literature. Median survival time versus per cent maximal cytoreduction in patient cohort: light gray area, maximal cytoreductive surgery accomplished in ≤ 25% and > 75%; dark gray area, corresponding range of median survival times. With permission from reference 33

105

Ch 04 final.qxd

12/18/2006

15:14

Page 106

SURGERY FOR OVARIAN CANCER

Consensus statements on the management of ovarian cancer have been put forth in Europe and the US.114,124 The National Comprehensive Cancer Network (NCCN), a consortium of NCI-designated cancer centers, and the Society of Gynecologic Oncologists have published detailed guidelines for the management of ovarian cancer.125,126 To maximize access to expert care for women with ovarian cancer, comprehensive systems of care need to be established, with a regional density and distribution appropriate for the respective population. While the specifics may differ for various regions or countries, the principle characteristics will be those of a model of excellence for cancer care as described by Smith et al. in 1999 and summarized by Trimble (Table 4.10).127 The treatment of patients with ovarian cancer in a cancer center has the advantage of the presence of an expert multidisciplinary team. Advances in periopera-

tive care have made aggressive debulking safer and the progress made in chemotherapy has led to further increases in survival times. Increased efforts to provide access to such specialized care are desirable so that as many women as possible are seen by an expert multidisciplinary care team providing treatment based on the best available evidence and offering access to research protocols. The expert team may consist of a gynecologic oncologist, who is centrally coordinating the care, is the operating surgeon and directs chemotherapy; a pathologist, a radiologist, a medical oncologist and a radiation oncologist, all specialized in gynecologic oncology; a clinical nurse specialist as a key member of the gynecologic oncology team providing support and information to the patient and acting as a liaison between the cancer center and other support services. Additional services available include pain management, nutritionist, social worker, psychologist, stoma and wound care specialist, physical

Table 4.9 Summary of guidelines by the Society of Gynecologic Oncologists (SGO) for referral of a patient with a pelvic mass or suspected ovarian cancer to a gynecologic oncologist. From reference 123 Characteristic

Referral, comment

Pelvic mass with ascites, omental caking, pleural effusion

All

Complex mass, > 10 cm, fixed, nodular, with solid components or excrescences, bilateral

All

Premenarchal girls with ovarian mass requiring surgery

All

Young patients with a pelvic mass and elevated tumor markers (CA-125, AFP, hCG)

All

Elevated tumor markers in peri- or postmenopausal women

All; ovarian mass and elevated CA-125 of 35–65 U/ml is associated with a cancer risk of 50–60%. An ovarian mass with a CA-125 level of > 65 U/ml in a woman 50 years of age or older is virtually diagnostic of malignancy with a specificity of 98%

Postmenopausal patients with suspicious ovarian masses

All; in postmenopausal patients, the risk of malignancy with a unilocular cyst is < 1%. This risk of malignancy increases to 8% with a multilocular cyst and 70% in a mass with solid components

Adnexal mass in women with significant family or personal history of ovarian, breast or other cancers

All

AFP, α-fetoprotein; hCG, human chorionic gonadotropin

106

Ch 04 final.qxd

12/18/2006

15:14

Page 107

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

Table 4.10 Characteristics of a model of excellence for cancer care. From reference 127 1.

Coordinated care with one person in charge

2.

Ease of access

3.

Ready access to information, answers to questions, psychosocial support

4.

Multidisciplinary care with transparency for the patient among the disciplines

5.

Guidelines for patient management of all common problems

6.

Full range of services from prevention to survivor follow-up to hospice care

7.

Measurement of patient processes and outcomes to ensure good care

8.

Accountability of health-care providers to measured outcomes

9.

Acceptable predetermined cost

therapist, a palliative care team and at times involvement of a general surgeon or surgical oncologist. A regularly convening multidisciplinary tumor board is an excellent venue to formulate jointly the interdisciplinary treatment approach. Research functions that are more efficiently fulfilled by a multidisciplinary team in a cancer center include an accurate prospective data collection and the administration of study protocols. Although data on and insight into quality of life of patients with ovarian cancer are limited, it appears that both surgery followed by platinum-based chemotherapy and neoadjuvant chemotherapy with interval cytoreduction can improve quality of life.128,129 In addition to effective and efficient treatment, psychological counseling, palliative and home care, nutritional support and pain relief are the most important areas for improving quality of life for patients with ovarian cancer. The patterns of care described above strongly suggest that care by a multidisciplinary expert team can substantially improve survival and quality of life for women with ovarian cancer. Current recommenda-

tions for enhancing the care of women with ovarian cancer include: (1) Educating physicians and the lay public on the importance of cancer staging, tumor debulking and care by an expert team; (2) Maximizing access for women with ovarian cancer to effective primary surgery; (3) Offering neoadjuvant chemotherapy followed by interval cytoreduction to those patients for whom poor performance status or severe comorbidities prohibited initial cytoreduction; (4) Assuring that appropriate chemotherapy is given based on best available evidence and that the patient is cared for by a multidisciplinary team with expertise in ovarian cancer; (5) Assuring the establishment of a firm relationship between the patient and her multidisciplinary cancer care team to assure appropriate cancer care for the rest of her life; (6) Developing systems for more effective outcome data collection and patient enrollment into clinical trials.

INTERVAL CYTOREDUCTIVE SURGERY Interval cytoreductive surgery after a short course of chemotherapy is a management option for patients with advanced ovarian cancer in principally two clinical scenarios: first, the patient who underwent upfront surgical exploration for primary tumor resection but whose disease was not optimally cytoreduced; and second, the patient who was deemed not suitable for primary cytoreductive surgery, owing to poor performance status, medical co-morbidities or the extent and distribution of disease. The major advantage of the first approach is that all patients who are medically fit for surgery are given the opportunity for primary optimal cytoreduction, at the cost of patients with suboptimally resected disease potentially undergoing a second surgical procedure with the associated

107

Ch 04 final.qxd

12/18/2006

15:14

Page 108

SURGERY FOR OVARIAN CANCER

morbidity. Upfront, neoadjuvant chemotherapy avoids the morbidity of two major surgical procedures, but at the risk of denying some patients the potentially greater survival benefit of primary cytoreduction. The appropriate selection of patients to this approach is crucial and needs further investigation. Retrospective data suggest median survival rates of 26 months with neoadjuvant chemotherapy and interval cytoreduction (Table 4.11), which is superior to the 20 months’ median survival of patients with primarily suboptimally cytoreduced disease, but inferior to the 51 months’ median survival reached in patients who underwent optimal primary cytoreduc-

tion followed by chemotherapy (Table 4.4). The European Organization for Research and Treatment of Cancer (EORTC) has a currently ongoing randomized controlled trial (EORTC-55971) of primary maximal cytoreduction followed by chemotherapy versus neoadjuvant chemotherapy with interval cytoreduction (Figure 4.9).143 Until these data become available, neoadjuvant chemotherapy with planned interval debulking should not be offered electively to any group of patients outside a clinical trial without solid indication as, based on the retrospective data, it does not appear to be associated with overall survival outcomes equivalent to those of primary cytoreduction.

Table 4.11 Non-randomized studies on interval debulking surgery after platinum-based neoadjuvant chemotherapy in patients with advanced-stage ovarian cancer

Median no. of Author, year

n

chemo cycles

No. optimal

Survival (months) of entire cohort

No. ICR

(% of ICR)

(optimal vs. suboptimal)

24 (48%)

18 (75%)

Studies with some primary surgery for ovarian cancer* Wils, 1986130 1989131

50

3

> 41 (> 48 vs. 20)

36

3

28 (78%)

26 (93%)

26

Ng, 1990132

25

2

25 (100%)

19 (76%)

n/a

Jacob, 1991133

22

3

22 (100%)

17 (77%)

16 (18 vs. 8)

Lim,1993134

30

3

11 (37%)

9 (82%)

10 (23 vs. 6)

49

3

31 (63%)

26 (84%)

n/a

31

3

31 (100%)

26 (84%)

42

Lawton,

Vergote,

1998135

Kuhn, 2001137 2001138

54

4

46 (85%)

39 (85%)

22

Shibata, 2003140

23

3

16 (70%)

13 (81%)

19 (21 vs. 10)

Morice, 2003141

34

3

34 (100%)

32 (94%)

26

Fanfani, 2003142

73

3

62 (85%)

52 (84%)

> 25 (> 27 vs. 14)

Ansquer,

Studies with no primary surgery for ovarian cancer** Surwit, 1996149

29

2

29 (100%)

16 (55%)

23 (32 vs. 18)

Schwartz, 1999136

59

5

41 (69%)

n/a

13

2001139

45

3

45 (100%)

34 (76%)

34

560

3–5

445 (79%)

327 (81%)

> 26 (> 28 vs. 13)

Kayikciog, Total

ICR, interval cytoreduction Studies include a broad spectrum of patients ranging from those deemed unresectable by CT or medically unfit for surgery (**) to those who underwent suboptimal primary surgery (*). The latter again varied greatly within and between studies ranging from open and close biopsy or laparoscopic assessment of unresectable disease to failed cytoreduction in the hands of a specialist trained in gynecologic oncologic surgery

108

Ch 04 final.qxd

12/18/2006

15:14

Page 109

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

Interval surgery following suboptimal primary cytoreduction Rationale

Optimal primary tumor debulking remains the cornerstone of ovarian cancer treatment and one of the most powerful predictors of prognosis. Nevertheless, not every patient can successfully undergo optimal primary cytoreduction and the prognosis of patients with suboptimal debulking remains poor, with, according to 2003 FIGO statistics, a 21% 5-year survival for stage IIIC patients with > 2 cm residual disease.18 In efforts to improve survival for patients with suboptimal residual disease after primary cytoreduction, interval cytoreduction after a prescribed short course of chemotherapy, typically 2–3 cycles, has been postulated as an alternative means of achieving optimal disease status in an effort to improve response to

subsequent chemotherapy and to improve survival. The feasibility of achieving optimal tumor debulking in 64–83% of such patients at interval cytoreduction has been demonstrated in several retrospective and prospective cohort studies (Table 4.11) and three randomized controlled trials.144–146

Selection criteria Response to chemotherapy

The indication for interval cytoreductive surgery is generally based on the response to chemotherapy. Patients with chemotherapy-resistant tumors, i.e. those 5–15% whose disease progresses on chemotherapy, are extremely unlikely to derive any survival benefit from interval cytoreductive surgery. Progressive disease on chemotherapy has been an exclusion criterion in the two large randomized controlled trials on

Stage IIIC or IV epithelial ovarian, peritoneal or tubal cancer

Randomization

Upfront maximal cytoreductive surgery followed by cisplatin/paclitaxel or carboplatin/paclitaxel every 3 weeks for three courses

Neoadjuvant chemotherapy with cisplatin/paclitaxel or carboplatin/paclitaxel every 3 weeks for three courses

Optimal primary debulking

Suboptimal primary debulking

Disease is stable or responding

Three additional courses of the same chemotherapy

Interval debulking surgery followed by three additional courses of the same chemotherapy

Interval debulking surgery followed by three additional courses of the same chemotherapy

Disease progression

Offer chemotherapy off protocol

Figure 4.9 Ongoing phase III randomized trial (EORTC-55971) of neoadjuvant chemotherapy and primary cytoreductive surgery in advanced ovarian cancer (modified with permission from Vergote et al. 2000108,143)

109

Ch 04 final.qxd

12/18/2006

15:14

Page 110

SURGERY FOR OVARIAN CANCER

interval cytoreduction, EORTC-GCG 55865 and GOG 152.144,146 The EORTC data further underscore the significance of chemosensitivity of the tumor as survival was substantially better in patients who were found at exploration for interval debulking to have achieved optimal tumor reduction to < 1 cm with chemotherapy only compared to those reaching the same residual disease based on the surgical intervention (42 vs. 27 months’ median survival).144 Potential for optimal tumor cytoreduction

Based on the same EORTC data, it also appears that, if the disease cannot be maximally resected to < 1 cm residual disease at interval cytoreduction, no survival benefit is derived from this second surgical procedure.144 Aggregate data from non-randomized studies confirm the poor outcome for patients whose tumor cannot be optimally cytoreduced at interval surgery. Therefore, the potential for optimal resection of the disease at interval debulking must exist. Medical condition

The patient’s general medical condition and performance status need to be such that she is fit to undergo interval cytoreductive surgery. In the EORTC and GOG 152 studies, 9% and 6% of enrolled patients, respectively, did not make it to the interval cytoreductive surgery after three cycles of chemotherapy, owing to early death or poor overall medical condition. If disease progression is included, 19% and 11% of patients, respectively, were ineligible for interval cytoreduction.144,146 Clinical outcome

The majority of non-randomized prospective and retrospective studies have demonstrated a positive survival impact of optimal interval cytoreduction (Table 4.11). Most of the trials are small, and have varied inclusion criteria and chemotherapy regimens, making comparative interpretation difficult. The studies include a broad spectrum of primary surgeries for ovarian cancer within and between series ranging from open-and-close biopsy or laparoscopic assess-

110

ment of unresectable disease to failed cytoreduction in the hands of specialists trained in gynecologic oncologic surgery. Overall, 78% of the patients treated responded to chemotherapy and met the other selection criteria for interval cytoreduction. If interval laparotomy was performed it led to optimal tumor debulking in 75–94% of patients. One of the largest retrospective series with stringent inclusion criteria was a recent study of 184 patients with stage IIIC ovarian cancer,142 in which Fanfani et al. demonstrated that survival outcomes of patients with stage IIIC ovarian cancer judged unresectable at initial laparotomy by gynecologic oncologists were improved with neoadjuvant chemotherapy followed by successful interval cytoreductive surgery; yet the results were inferior to those achieved with optimal primary cytoreduction and were more similar to those of suboptimal cytoreduction. In 1995, the Gynecological Cancer Cooperative Group (GCG) of the EORTC showed for the first time in a randomized controlled trial (EORTC/GCG55865) that interval debulking significantly lengthened progression-free and overall survival.144 The study enrolled 425 patients with suboptimally resected tumors. If there was no clinical evidence of progressive disease after three cycles of chemotherapy, patients were randomized to undergo interval cytoreduction or not. After randomization, patients received at least an additional three cycles of chemotherapy. In this study, interval debulking was not associated with death, severe morbidity or intraoperative complications. Overall and progression-free survival were significantly increased in favor of the patients who underwent interval cytoreductive surgery. After a median follow-up period of 6.3 years, the 5-year survival rate was 24% in the interval cytoreduction group and 13% in the no surgery group. After adjusting for other prognostic factors, the reduction in risk of death attributable to surgery was 49%.147 This survival advantage was seen only in patients who achieved optimal disease status (< 1 cm residual disease) after induction chemotherapy and interval cytoreduction. Patients who remained with

Ch 04 final.qxd

12/18/2006

15:14

Page 111

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

of patients had disease larger than 5 cm at study entry. Thus, in the EORTC study unresected tumors predominated, whereas unresectable tumors made up the majority of GOG 152, which may indicate that more patients with intrinsic poor-prognosis tumor characteristics were enrolled in GOG 152 than in the EORTC trial. These data suggest that patients with the worst prognostic factors, i.e. those who had undergone extensive, but suboptimal primary cytoreductive surgery by a gynecologic oncologist followed by platinum–taxane-based chemotherapy, may be less likely to benefit from interval cytoreductive surgery than patients whose initial surgery was less extensive or an open-and-close biopsy only.

Neoadjuvant chemotherapy followed by delayed primary cytoreduction Rationale

In reviewing data from GOG 52 and GOG 97, Hoskins et al. showed that only initial cytoreduction to a residual disease volume of < 2 cm had a significant effect on survival. There was no benefit for

100

Surgery

100

< 1 cm

90

No surgery

90

Optimal

80

Suboptimal

70

No surgery

80 70

Overall survival (%)

Overall survival (%)

residual tumors > 1 cm in diameter after the interval cytoreduction had overall and progression-free survivals literally identical to those treated with chemotherapy only (18.6 vs. 19.2 months and 11.5 vs. 11.5 months, respectively; Figure 4.10).148 A followup study by the Gynecologic Oncology Group (GOG 152), for which preliminary results were presented at the American Society for Clinical Oncology (ASCO) in 2002, did not show a survival benefit for interval cytoreduction. While both studies had comparable objective study designs, some salient differences between the EORTC study and GOG 152 exist, which may have contributed to the differences in results (Table 4.12). The chemotherapy used in the EORTC study was cisplatin and cyclophosphamide, in GOG 152 patients were treated with cisplatin and paclitaxel. In the EORTC study, many patients underwent ‘open and close’ procedures as their primary surgery. In GOG 152, 95% of the primary debulking surgeries were performed by gynecologic oncologists. This presumed maximal effort at cytoreduction resulted in 68% of cases with < 5 cm maximal disease diameter at enrollment, whereas in the EORTC study 64%

p = 0.0032

60 50 40 30

60 50 40 30

20

20

10

10

0

2

4 6 Time (years)

8

10

0

2

4 6 Time (years)

8

10

Figure 4.10 Overall survival of patients in the EORTC/GCG-55865 study of interval debulking for advanced ovarian cancer (from reference 148). Surgery, interval cytoreduction; no surgery, no interval cytoreduction; < 1 cm, largest disease diameter of < 1 cm at beginning of exploration for interval cytoreduction; optimal, largest residual disease of < 1 cm at completion of interval cytoreduction; suboptimal, residual disease after interval cytoreduction of > 1 cm; no surgery, no interval cytoreduction

111

Ch 04 final.qxd

12/18/2006

15:14

Page 112

SURGERY FOR OVARIAN CANCER

patients who underwent cytoreduction but who were left with residual disease greater than 2 cm. This implies that, if initial cytoreductive surgery to optimal residual disease is not possible, the patient in terms of survival is unlikely to benefit from the operation.34 The logical conclusion from these GOG data in combination with the aggregate data on interval cytoreduction summarized in Table 4.11 would be that patients deemed initially not optimally cytoreducible might benefit from a short course of initial neoadjuvant chemotherapy followed by delayed primary cytoreduction. The goal is for the neoadjuvant chemotherapy to reduce tumor masses to an extent that would allow for successful surgical debulking during delayed primary cytoreduction, thus improving the patient’s prognosis. For patients with a poor performance status due to massive ascites or pleural effusions, neoadjuvant chemotherapy may lead to reduction or resolution of ascites and pleural effusions, an enhanced performance status and improved likelihood that the patient with other severe co-morbidities will now become fit to undergo cytoreductive surgery. Furthermore, in an occasional patient, neoadjuvant chemotherapy allows for treatment of the cancer while the patient is undergoing the necessary therapy for an acute medical co-morbidity, which would have greatly enhanced her short-term surgical risk, such as an acute thromboembolic event or an acute myocardial infarction. After two or three cycles of chemotherapy, many of these patients will reach a medical condition and performance status that will permit surgery. Furthermore, following neoadjuvant chemotherapy, the surgery needed to achieve optimal cytoreduction is often less extensive and generally thought to be attended by fewer complications.136,150

Table 4.12 Results of two large randomized controlled trials of interval cytoreductive surgery (EORTC-GCG 55865 and GOG 152). From references 144, 146, 147 EORTC-GCG 55865

GOG 152

No. enrolled No. randomized

425 319 (75%)

550 424 (77%)

Chemotherapy

cisplatin cyclophosphamide

cisplatin paclitaxel

Stage IV

21%

6%*

WHO performance status 2

17%

7%

Residual disease after PCR < 5 cm > 5 cm

36% 64%

56% 44%

After 3 cycles chemo progressive disease early death medically unfit for surgery

10% 3% 6%

5% 3% 4%

Optimal ICR rate

64%

Overall response rate ICR no ICR

84% 70%

Complete clinical response ICR no ICR

70% 35%

Median OS ICR optimal suboptimal no ICR

26.4 months 31.2 months 18.6 months 19.2 months

35.7 months

Median PFS ICR no ICR

14.5 months 11.5 months

12.5 months 12.7 months

36.2 months

*Patients with stage IV disease were excluded. PCR, primary cytoreduction; ICR, interval cytoreduction; OS, overall survival; PFS, progression-free survival

Selection criteria Disease appears not optimally resectable at time of diagnosis

One of the most critical questions surrounding the issue of neoadjuvant chemotherapy is how to identify the patient whose disease is not optimally cytore-

112

ducible at primary surgery without subjecting her to a surgical exploration. Factors that may be identified in the preoperative evaluation of a patient with presumed ovarian cancer that would cause most clinicians to recommend against primary surgery may include bulky parenchymal lung metastases, extensive

Ch 04 final.qxd

12/18/2006

15:14

Page 113

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

Several investigators have studied the value of computed tomography (CT) for predicting surgical outcome in patients with ovarian cancer and have suggested that CT scans can predict suboptimal cytoreduction with an overall sensitivity of 71% and specificity of 86% (Table 4.14). Nelson was able to predict suboptimal versus optimal cytoreduction with 79% accuracy using the following criteria: attachment of the omentum to the spleen, tumors > 2 cm in the mesentery, liver, diaphragm, pleura, porta hepatis, or suprarenal para-aortic lymph nodes.158 Bristow et al. created a model including 13 radiographic features and performance status to calculate predictive index scores with a 93% accuracy for optimal versus suboptimal primary debulking.160 Dowdy et al. found on multivariate analysis only diffuse peritoneal thickening to be an independent predictor of suboptimal surgical cytoreduction.156 These data supporting the use of preoperative indices to predict suboptimal cytoreduction need to be interpreted with extreme caution, as numerous variables contribute to the ‘optimal resectability’ of a tumor, ranging from tumor stage, distribution and biology to the surgeon’s technical adeptness and philosophy. Thus, these findings are probably only reproducible in settings where practice patterns and optimal cytoreduction rates are similar. Furthermore, even with the predictive imaging index scores optimized to the unique settings of each of

intraparenchymal liver metastases, or bulky suprarenal adenopathy as some examples of frankly not optimally resectable disease. Those scenarios, however, are rare. Most patients with ovarian cancer have disease limited to the abdominal cavity, and preoperative assessment of resectability becomes more difficult. Attempts have been made for two decades to identify means of predicting preoperatively in whom optimal cytoreduction is feasible and in whom it is not. Accuracy of the methods used to date continues to be only modest. In 1986, Heintz et al. published the UCLA experience, where a preoperative diameter of the largest metastasis of > 5 cm or ascites of > 1000 ml were associated with reduced rates of optimal cytoreduction. However, even among patients with large metastases or large ascites, more than 50% achieved optimal cytoreduction.10 A number of studies have demonstrated an association between the preoperative CA-125 level and the ability to achieve optimal cytoreduction (Table 4.13), yet the overall accuracy rates at predicting surgical outcome (optimal versus suboptimal) are only 50–78% with sensitivity and specificity rates of 45–80% and 52–83%, respectively. The two series with high optimal cytoreduction rates (> 70%) found preoperative CA-125 levels completely lacking as a predictor of surgical outcome.155,156

Table 4.13 Preoperative predictors of suboptimal cytoreduction in advanced-stage ovarian cancer: CA-125 Author, year

n

% Optimal

CA-125 cut-off (U/ml)

Sensitivity (%)

Specificity (%)

Accuracy (%)

100

45 (45%)

500

78

73

76

Gemer, 2001152

40

24 (60%)

500

62

83

75

2002153

92

48 (52%)

500

73

77

75

Cooper, 2002154

112

65 (58%)

500

75

59

64

99

72 (73%)

912

54

58

57

87

62 (71%)

500

45

52

50

77

32 (42%)

586

80

75

78

607

348 (57%)

500–912

67

66

67

Chi,

2000151

Saygili,

Memarzadeh,

2003155

Dowdy, 2004156 Brockbank, Total

2004157

113

Ch 04 final.qxd

12/18/2006

15:14

Page 114

SURGERY FOR OVARIAN CANCER

Table 4.14 Preoperative predictors of suboptimal cytoreduction in advanced-stage ovarian cancer: computed tomography (CT) using various predictive imaging index scores Author, year

n

% Optimal

Sensitivity (%)

Nelson, 1993158

34

21 (62%)

92

71

79

Meyer, 1995159

18

6 (33%)

58

100

72

Bristow, 2000*160

41

20 (49%)

100

85

93

Dowdy, 2004156

87

62 (71%)

52

90

79

180

109 (61%)

71

86

81

Total

Specificity (%)

Accuracy (%)

*Includes performance status in predictive index score

these studies, up to 45% of patients were unnecessarily explored (i.e. underwent suboptimal debulking only)156,159 and up to 29% would have been inappropriately left unexplored (i.e. underwent optimal debulking despite preoperative imaging suggesting non-debulkable disease).158 Given the significant survival impact of optimal cytoreduction, which is unparalleled by neoadjuvant chemotherapy and interval cytoreduction, it seems that, if in doubt, every patient except the one who is medically unfit for surgery should be offered exploration for optimal cytoreduction. In other words, as stated by Hacker: ‘clearly not all patients will derive equal benefit from aggressive cytoreduction, but to deny it … is to deny it to those for whom it will undoubtedly prolong survival.’161 Medically compromised patient

Medically severely compromised patients, in whom the medical status precludes a proper initial operation, may be excellent candidates for neoadjuvant chemotherapy. Frequently two or three cycles of chemotherapy lead to resolution of pleural effusion and ascites and a measurable improvement in performance status and fitness for surgery. There are inadequate data to determine what degree of medical compromise would identify the patient who would be better served by neoadjuvant chemotherapy than by primary cytoreduction. Certainly a patient with a

114

severely compromised performance status or a severe systemic disorder that is already life-threatening and may not be correctable by the operation, for example decompensated congestive heart failure, unstable angina or an acute pulmonary embolus, would be best treated with neoadjuvant chemotherapy, delaying surgery until the medical condition has been optimized. However, for patients with moderate systemic diseases that are well controlled, such as chronic obstructive pulmonary disease or chronic stable angina, or with some reduction in performance or nutritional status due to the tumor, the answer is less clear. Some experts recommend neoadjuvant chemotherapy to all patients with large ascites, severe malnutrition (serum albumin < 2.8 g/dl, weight loss > 10–15% of total body weight), and concurrent serious medical problems, such as chronic obstructive pulmonary disease, myocardial ischemia and age over 75 years.150 However, while national patterns of care data indicate that older patients are less likely to undergo maximal cytoreductive surgery,162 chronologic age per se should not be a contraindication to cytoreductive surgery. In fact, with appropriate patient selection, optimal cytoreduction can be carried out in older patients as successfully as in younger patients and with no higher morbidity or mortality.101 Similarly, the association between reduced nutritional status and the inability to undergo optimal cytoreduction is not well established. In addition, the volume of preoperative ascites

Ch 04 final.qxd

12/18/2006

15:14

Page 115

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

has been found to be a poor predictor of suboptimal cytoreduction.156,160 With judicious intra- and postoperative fluid management, patients without major underlying cardiovascular or renal pathology can generally handle the perioperative fluid shifts without significant sequelae. Prerequisites

Prior to undergoing neoadjuvant chemotherapy, pathologic confirmation of an ovarian cancer needs to be obtained, which can be done by fine-needle aspiration, CT-guided or laparoscopic biopsy. In carefully selected cases, a positive cytology on the ascites or pleural effusion fluid suffices for making the diagnosis as long as tumor markers and disease distribution on clinical examination and imaging are typical for ovarian cancer, and other malignancies, such as breast or gastrointestinal have been reasonably excluded. Selection criteria for delayed primary cytoreduction

These are essentially identical to those discussed above for interval cytoreduction, including an overall medical condition permitting surgery and the prospect for optimally resectable disease. Most importantly, there needs to be evidence of response to chemotherapy. Aggressive surgical intervention is not indicated in patients whose tumor shows little or no response to preoperative chemotherapy. The latter predicts a very poor outcome that is not altered by surgical debulking. Clinical outcome

The interpretation of outcomes and survival data of studies on neoadjuvant chemotherapy is difficult. As discussed previously, many of the series are small and retrospective. Inclusion criteria are highly variable, leading to a very heterogeneous patient population (Table 4.11). Some include only patients judged inoperable at primary surgical exploration by a gynecologic oncologist;142 others include patients who were considered inoperable because of medical conditions.136 In the strictest sense applicable to the question of neoadjuvant chemotherapy followed by

delayed primary cytoreductive surgery, are only the studies by Surwit et al.,149 Schwartz et al.,136 and Kayikciog Lu et al.,139 where primary unresectability was predicted clinically on the basis of CT findings without performing primary surgery for ovarian cancer. In one of the most frequently cited retrospective studies on this subject, Schwartz et al. reported longterm outcomes of 59 patients who received primary neoadjuvant chemotherapy, 41 of whom subsequently underwent cytoreductive surgery, and compared them with 209 patients with stage IIIC/IV ovarian cancer treated with primary surgery followed by chemotherapy. The median overall survival of the conventionally treated patients was 2.18 years, whereas that of the neoadjuvant chemotherapy-treated patients was 1.07 years, a statistically non-significant difference.136 However, the survival outcomes for the conventionally treated patients in this study did not approach those reached in many other centers. This may at least in part be due to the low optimal cytoreduction rate to residual disease of < 1 cm of only 30.5% with only 7.3% having microscopic residual disease in this series. In a similar study of 45 patients treated with neoadjuvant chemotherapy and 158 patients who underwent primary cytoreduction, Kayikciog Lu et al. also found no significant difference in median survival when the two patient groups overall were compared (34 vs. 38 months).139 Survival data were unfortunately not stratified by amount of residual disease. The perioperative morbidity after neoadjuvant chemotherapy is not increased. In fact, Schwartz et al. found less intraoperative blood loss and shorter hospital stays when comparing interval cytoreduction with primary surgery.136

Chemotherapy regimen and timing of interval cytoreduction Based on the available evidence, the agents used for neoadjuvant chemotherapy are the same drugs used in postoperative first-line chemotherapy for ovarian cancer: a platinum compound in combination with a

115

Ch 04 final.qxd

12/18/2006

15:14

Page 116

SURGERY FOR OVARIAN CANCER

taxane, most commonly carboplatin and paclitaxel. If a patient is medically significantly compromised, the single agent carboplatin may be preferred. The number of chemotherapy cycles that should be administered prior to interval cytoreduction is uncertain and data that directly address this question are very limited. In a co-operative, non-randomized study from The Netherlands, interval surgery was planned based upon the response to chemotherapy after at least two cycles and the median number of chemotherapy cycles given was three (range 2–6).130 Based on the theoretical rationale for cytoreductive surgery discussed at the beginning of this chapter, it appears that the timing of cytoreductive surgery is important. It should be performed after enough chemotherapy cycles have been given to document sensitivity to the treatment and allow for improvement of the patient’s performance status and overall medical condition, and yet early enough to avoid chemotherapy resistance and maximize the therapeutic impact of the tumor debulking. The two large prospective interval cytoreduction trials by the EORTC and the GOG both required that three cycles of chemotherapy be given prior to interval cytoreduction.144,146 Extrapolating from second-look data, delay of cytoreduction for 6–8 cycles of chemotherapy or more appears less likely to be beneficial. In the assessment of planned secondary cytoreductive surgery at second look of patients with suboptimal primary cytoreduction followed by eight cycles of platinumbased chemotherapy with or without bacillus Calmette Guérin (BCG) (GOG 60), only 25% of the 216 evaluable patients had successful secondary debulking surgery in the sense that there was tumor that could be optimally cytoreduced to microscopic residual disease in 17% and to < 1 cm residual disease in 8% with an associated improvement in survival. The remaining 75% of patients were explored but were not candidates for cytoreduction, because either there was no or only microscopic residual disease (43%) or there was significant unresectable disease (32%).163 In contrast, almost twice as many patients (48%) had the benefit of successful interval cytore-

116

duction after three cycles of chemotherapy on the EORTC/GCG 55865 protocol, with 19% undergoing cytoreduction to microscopic residual disease and 29% to optimal visible residual tumor.147 Median survival for patients optimally cytoreduced after three cycles of chemotherapy in the EORTC study was 31 months, compared to 24 months for patients optimally cytoreduced at second look after eight cycles of chemotherapy on GOG 60.147,163 Both studies used platinum-based chemotherapy without a taxane. The number of chemotherapy cycles to be given after interval cytoreduction has also not been studied in any systematic fashion and the clinician is left extrapolating from data on chemotherapy after primary cytoreduction. Three randomized trials assessing the duration of platinum-based chemotherapy in advanced ovarian cancer have failed to demonstrate an improvement in median survival with more cycles of chemotherapy, but longer duration was associated with more toxicity.164–166 These studies support the current standard of six cycles of platinum-based chemotherapy following primary cytoreduction. The EORTC and GOG trials both followed this standard and prescribed three cycles before and at least three additional cycles of chemotherapy after the interval tumor debulking.144,146 Nonetheless, especially patients with significant disease present at interval cytoreduction that was optimally resectable, may benefit from more than three additional cycles of chemotherapy.

Timing of interval surgery and post-interval cytoreduction chemotherapy Ideally, interval cytoreduction should be performed 2–4 weeks after the preceding chemotherapy course. Prior to surgery, the patient must be evaluated clinically and with laboratory testing and imaging as indicated to assess disease response. Patients with stable disease or objective response are candidates for interval cytoreduction. Surgery should be performed as soon as nadir counts allow and if possible all visible disease should be resected. The subsequent course of chemotherapy should be administered as soon as

Ch 04 final.qxd

12/18/2006

15:14

Page 117

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

possible after secondary cytoreductive surgery, which frequently can be accomplished prior to discharging the patient postoperatively. In conclusion, aggressive primary cytoreductive surgery followed by chemotherapy continues to be the hallmark of initial treatment for advanced ovarian

cancer. In selected cases, neoadjuvant chemotherapy with delayed primary cytoreduction may be appropriate. Figure 4.11 summarizes a suggested management approach to the patient with advanced ovarian cancer.

Suspected advanced-stage ovarian cancer

Co-morbidities, performance status and disease distribution permit primary surgery

Yes

No

Primary cytoreductive surgery

Histologic/cytologic diagnosis

Optimal

Suboptimal

Chemotherapy with platinum compound and taxane

Neoadjuvant chemotherapy with platinum compound and taxane

Assessment of response after 2–4 cycles of chemotherapy

Response or stable disease

Progressive disease

Interval cytoreduction

Alternate chemotherapy Clinical trial Supportive care

Additional chemotherapy

Figure 4.11 Treatment algorithm for patients with advanced stage ovarian cancer

117

Ch 04 final.qxd

12/18/2006

15:14

Page 118

SURGERY FOR OVARIAN CANCER

REFERENCES 1.

Speert H. Ephraim McDowell and Ovariotomy. In Speert H, ed. Obstetric & Gynecologic Milestones Illustrated. New York: Parthenon Publishing, 1996: 552–3

2.

O’Dowd MJ, Philipp EE. Cancer of the Ovary. In O’Dowd MJ, Philipp EE, eds. The History of Obstetrics and Gynecology. New York: Parthenon Publishing, 1994: 581–92

3.

McDowell E. Three cases of extirpation of diseased ovarii. Eclectic Repertory Anal Rev 1817; 7: 242–4

4.

Meigs JV. Tumors of the Female Pelvic Organs. New York: MacMillan, 1934.

5.

Pemberton FA. Carcinoma of the ovary. Am J Obstet Gynecol 1940; 40: 751–63

6.

Hudson CN. A radical operation for fixed ovarian tumours. J Obstet Gynaecol Br Commonw 1968; 75: 1155–60

7.

8.

9.

Munnell EW. The changing prognosis and treatment in cancer of the ovary: a report of 235 patients with primary ovarian carcinoma 1952–61. Am J Obstet Gynecol 1968; 100: 790–805 Griffiths CT. Surgical resection of tumor bulk in the primary treatment of ovarian carcinoma. Natl Cancer Inst Monogr 1975; 42: 101–4 Hacker NF, Berek JS, Lagasse LD, et al. Primary cytoreductive surgery for epithelial ovarian cancer. Obstet Gynecol 1983; 61: 413–20

10.

Heintz APM, Hacker NF, Berek JS, et al. Cytoreductive surgery in ovarian carcinoma: feasibility and morbidity. Obstet Gynecol 1986; 67: 783–8

11.

Piver MS, Baker T. The potential for optimal (≤ 2 cm) cytoreductive surgery in advanced ovarian carcinoma at a tertiary medical center: a prospective study. Gynecol Oncol 1986; 24: 1–8

12.

13.

118

Piver MS, Lele SB, Marchetti DL, et al. The impact of aggressive debulking surgery and cisplatin-based chemotherapy on progression-free survival in stage III and IV ovarian carcinoma. J Clin Oncol 1988; 6: 983–9 Bertelsen K. Tumor reduction surgery and long-term survival in advanced ovarian cancer: a DACOVA study. Gynecol Oncol 1990; 38: 203–9

14.

Hoskins WJ, Bundy BN, Thigpen JT, et al. The influence of cytoreductive surgery on recurrence-free interval and survival in small-volume stage III epithelial ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1992; 47: 159–66

15.

Eisenkop SM, Spirtos NM, Montag TW, et al. The impact of subspecialty training on the management of advanced ovarian cancer. Gynecol Oncol 1992; 47: 203–9

16.

Guidozzi F, Ball JHS. Extensive primary cytoreductive surgery for advanced epithelial ovarian cancer. Gynecol Oncol 1994; 53: 326–30

17.

Eisenkop SM, Friedman RL, Wang HJ. Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced epithelial ovarian cancer: a prospective study. Gynecol Oncol 1998; 69: 103–8

18.

Heintz APM, Odicino F, Maisonneuve P, et al. Carcinoma of the ovary. FIGO 25th annual report on the results of treatment in gynecological cancer. Int J Gynecol Obstet 2003; 83: 135–66

19.

Blythe JG, Wahl TP. debulking surgery: does it increase the quality of survival? Gynecol Oncol 1982; 14: 396–408

20.

Rubin SC, Hoskins WJ, Benjamin I, et al. Palliative surgery for intestinal obstruction in advanced ovarian cancer. Gynecol Oncol 1989; 34: 16–19

21.

Skipper HE. Adjuvant chemotherapy. Cancer 1978; 41: 936–40

22.

Goldie JH, Coldman AJ. A mathematical model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat Rep 1979; 63: 1727–33

23.

Curt GA, Clendeninn NJ, Chabner BA. Drug resistance in cancer. Cancer Treat Rep 1984; 68: 87–99

24.

Skipper HE. Thoughts on cancer chemotherapy and combination modality therapy. J Am Med Assoc 1974; 230: 1033–5

25.

Bookman MA, Young RC. Principles of chemotherapy in gynecologic cancer. In Hoskins WJ, Perez CA, Young RC, eds. Principles and Practice of Gynecologic Oncology. Philadelphia: Lippincott Williams & Wilkins, 2000: 404–7

Ch 04 final.qxd

12/18/2006

15:14

Page 119

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

26.

Gunduz N, Fisher B, Saffer EA. Effect of surgical removal on the growth and kinetics of residual tumor. Cancer Res 1979; 39: 3861–5

37.

Berman ML. Future directions in the surgical management of ovarian cancer. Gynecol Oncol 2003; 90: S33–9

27.

DeVita VT. The relationship between tumor mass and resistance to chemotherapy. Implications for surgical adjuvant treatment of cancer. Cancer 1983; 51: 1209–20

38.

Heintz APM, van Oosterom AT, Trimbos JBMC, et al. The treatment of advanced ovarian carcinoma (I): clinical variables associated with prognosis. Gynecol Oncol 1988; 30: 347–58

28.

Bookman MA, Berek JS. Biologic and immunologic therapy of ovarian cancer. Hematol Oncol Clin North Am 1992; 6: 941–65

39.

29.

Hoskins WJ, Lawton FG. Primary cytoreductive surgery for advanced-stage disease. In Gershenson DM, McGuire WP, eds. Ovarian Cancer Controversies in Management. New York: Churchill Livingstone, 1998: 65–84

Berchuck A, Iversen ES, Lancaster JM, et al. Prediction of optimal versus suboptimal cytoreduction of advanced-stage serous ovarian cancer with the use of microarrays. Am J Obstet Gynecol 2004; 190: 910–25

40.

Cohen CJ, Goldberg JD, Holland JF, et al. Improved therapy with cisplatin regimens for patients with ovarian carcinoma (FIGO stages III and IV) as measured by surgical end-staging (second-look operation). Am J Obstet Gynecol 1983; 145: 955–67

41.

Podratz K, Malkasian GD Jr, Hilton JF, et al. Secondlook laparotomy in ovarian cancer: evaluation of pathologic variables. Am J Obstet Gynecol 1985; 152: 230–8

42.

Dauplat J, Ferriere JP, Gorbinet M, et al. Second-look laparotomy in managing epithelial ovarian carcinoma. Cancer 1986; 57: 1627–31

43.

Free KE, Webb MJ. Second-look laparotomy: clinical correlations. Gynecol Oncol 1987; 26: 290–7

44.

Carmichael JA, Shelley WE, Brown LB, et al. A predictive index of cure versus no cure in advanced ovarian carcinoma patients – replacement of secondlook laparotomy as a diagnostic test. Gynecol Oncol 1987; 27: 269–78

30.

31.

32.

Eisenkop SM, Spirtos NM, Friedman RL, et al. Relative influences of tumor volume before surgery and the cytoreductive outcome on survival for patients with advanced ovarian cancer: a prospective study. Gynecol Oncol 2003; 90: 390–6 Griffiths CT, Fuller AF. Intensive surgical and chemotherapeutic management of advanced ovarian cancer. Surg Clin North Am 1978; 58: 131–42 van Lindert ACM, Alsbach GPJ, Barents JW, et al. The role of the abdominal radical tumor reduction procedure (A.R.Tu.R.) in the treatment of ovarian cancer. In Heintz APM, Griffiths CT, Trimbos JB, eds. Surgery in Gynecologic Oncology. Boston: Martinus Nijhoff Publishers, 1984: 275–87

33.

Bristow RE, Tomacruz RS, Armstrong DK, et al. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: a meta-analysis. J Clin Oncol 2002; 20: 1248–59

45.

34.

Hoskins WJ, McGuire WP, Brady MF, et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J Obstet Gynecol 1994; 170: 974–9

Bertelsen K, Hansen MK, Pedersen PH, et al. The prognostic and therapeutic value of second-look laparotomy in advanced ovarian cancer. Br J Obstet Gynaecol 1988; 95: 1231–6

46.

Bristow RE, Montz FJ. Complete surgical cytoreduction of advanced ovarian carcinoma using the argon beam coagulator. Gynecol Oncol 2001; 83: 39–48

Sonnendecker EWW. Is routine second-look laparotomy for ovarian cancer justified? Gynecol Oncol 1988; 31: 249–55

47.

Ayhan A, Yarali H, Develioglu O, et al. Prognosticators of second-look laparotomy findings in patients with epithelial ovarian cancer. J Surg Oncol 1991; 46: 222–5

48.

Katsoulis M, Vorgias G, Panagiotides J, et al. The prognostic significance of second-look laparotomy in

35.

36.

Farias-Eisner R, Teng F, Oliveira M, et al. The influence of tumor grade, distribution and extent of carcinomatosis in minimal residual stage III epithelial ovarian cancer after optimal primary cytoreductive surgery. Gynecol Oncol 1994; 55: 108–10

119

Ch 04 final.qxd

12/18/2006

15:14

Page 120

SURGERY FOR OVARIAN CANCER

59.

Baker TR, Piver MS, Hempling RE. Longterm survival by cytoreductive surgery to less than 1 cm, induction weekly cisplatin and monthly cisplatin, doxorubicin, and cyclophosphamide therapy in advanced ovarian adenocarcinoma. Cancer 1994; 74: 656–63

60.

Conte PF, Bruzzone M, Chiara S, et al. A randomized trial comparing cisplatin plus cyclophosphamide versus cisplatin, doxorubicin, and cyclophosphamide in advanced ovarian cancer. J Clin Oncol 1986; 4: 965–71

61.

Neijt JP, ten Bokkel Huinink WW, van der Burg ME, et al. Randomized trial comparing two combination chemotherapy regimens (CHAP-5 v CP) in advanced ovarian carcinoma. J Clin Oncol 1987; 5: 1157–68

62.

Hainsworth JD, Grosh WW, Burnett LS, et al. Advanced ovarian cancer: long-term results of treatment with intensive cisplatin-based chemotherapy of brief duration. Ann Intern Med 1988; 108: 165–70

63.

Potter ME, Partridge EE, Hatch KD, et al. Primary surgical therapy of ovarian cancer: how much and when. Gynecol Oncol 1991; 40: 195–200

64.

Covens AL. A critique of surgical cytoreduction in advanced ovarian cancer. Gynecol Oncol 2000; 78: 269–74

Del Campo JM, Felip E, Rubio D, et al. Long-term survival in advanced ovarian cancer after cytoreduction and chemotherapy treatment. Gynecol Oncol 1994; 53: 27–32

65.

Voest EE, van Houwelingen JC, Neijt JP. A metaanalysis of prognostic factors in advanced ovarian cancer with median survival and overall survival (measured with the log (relative risk)) as main objectives. Eur J Cancer Clin Oncol 1989; 25: 711–20

Makar AP, Baekelandt M, Trope CG, et al. The prognostic significance of residual disease, FIGO substage, tumor histology, and grade in patients with FIGO stage III ovarian cancer. Gynecol Oncol 1995; 56: 175–80

66.

Le T, Krepart GV, Lotocki RJ, et al. Does debulking surgery improve survival in biologically aggressive ovarian carcinoma? Gynecol Oncol 1997; 67: 208–14

67.

Michel G, De Iaco P, Castaigne D, et al. Extensive cytoreductive surgery in advanced ovarian carcinoma. Eur J Gynecol Oncol 1997; 18: 9–15

68.

Hudson CN, Chir M. Surgical treatment of ovarian cancer. Gynecol Oncol 1973; 1: 370–8

69.

Weber AM, Kennedy AW. The role of bowel resection in the primary surgical debulking of carcinoma of the ovary. J Am Coll Surg 1994; 179: 465–70

70.

Tamussino KF, Lim PC, Webb MJ, et al. Gastrointestinal surgery in patients with ovarian cancer. Gynecol Oncol 2001; 80: 79–84

advanced ovarian cancer. Eur J Gynaecol Oncol 1997; 18: 200–2 49.

McGuire WP, Hoskins WJ, Brady MF, et al. Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 1996; 334: 1–6

50.

Ozols RF, Bundy BN, Greer BE, et al. Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol 2003; 21: 3194–200

51.

Gadducci A, Sartori E, Maggino T, et al. Analysis of failures after negative second-look in patients with advanced ovarian cancer: an Italian multicenter study. Gynecol Oncol 1998; 68: 150–5

52.

Rubin SC, Randall TC, Armstrong KA, et al. Tenyear follow-up of ovarian cancer patients after second-look laparotomy with negative findings. Obstet Gynecol 1999; 93: 21–4

53.

Hunter RW, Alexander NDE, Soutter WP. Metaanalysis of surgery in advanced ovarian carcinoma: Is maximum cytoreductive surgery an independent determinant of prognosis? Am J Obstet Gynecol 1992; 166: 504–11

54.

55.

56.

57.

58.

120

Vogl SE, Pagano M, Kaplan BH, et al. Cis-platin based combination chemotherapy for advanced ovarian cancer: high overall response rate with curative potential only in women with small tumor burdens. Cancer 1983; 51: 2024–30 Redman JR, Petroni GR, Saigo PE, et al. Prognostic factors in advanced ovarian carcinoma. J Clin Oncol 1986; 4: 515–23 Sutton GP, Stehman FB, Einhorn LH, et al. Ten-year follow-up of patients receiving cisplatin, doxorubicin, and cyclophosphamide chemotherapy for advanced epithelial ovarian carcinoma. J Clin Oncol 1989; 7: 223–9

Ch 04 final.qxd

12/18/2006

15:14

Page 121

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

71.

Berek JS, Hacker NF, Lagasse LD, et al. Lower urinary tract resection as part of cytoreductive surgery for ovarian cancer. Gynecol Oncol 1982; 13: 87–92

72.

Berek JS, Hacker NF, Lagasse LD. Rectosigmoid colectomy and reanastomosis to facilitate resection of primary and recurrent gynecologic cancer. Obstet Gynecol 1984; 64: 715–20

73.

Bridges JE, Leung Y, Hammond IG, et al. En bloc resection of epithelial ovarian tumors with concomitant rectosigmoid colectomy: the KEMH experience. Int J Gynecol Cancer 1993; 3: 199–202

74.

Barnes W, Johnson J, Waggoner S, et al. Reverse hysterocolposigmoidectomy (RHCS) for resection of panpelvic tumors. Gynecol Oncol 1991; 42: 151–5

75.

Scarabelli C, Gallo A, Franceschi S, et al. Primary cytoreductive surgery with rectosigmoid colon resection for patients with advanced epithelial ovarian carcinoma. Cancer 2000; 88: 389–97

76.

77.

Obermair A, Hagenauer S, Tamandl D, et al. Safety and efficacy of low anterior en bloc resection as part of cytoreductive surgery for patients with ovarian cancer. Gynecol Oncol 2001; 83: 115–20 Bristow RE, del Carmen MG, Kaufman HS, et al. Radical oophorectomy with primary stapled colorectal anastomosis for resection of locally advanced epithelial ovarian cancer. J Am Coll Surg 2003; 197: 565–74

78.

Malviya VK, Malone JM Jr, Deppe G. Advanced ovarian cancer: urinary tract resection as a part of cytoreductive surgery. Eur J Gynaecol Oncol 1989; 10: 69–72

79.

Nicklin JL, Copeland LJ, O’Toole RV, et al. Splenectomy as part of cytoreductive surgery for ovarian carcinoma. Gynecol Oncol 1995; 58: 244–7

80.

Sonnendecker EWW, Guidozzi F, Margolius KA. Splenectomy during primary maximal cytoreductive surgery for epithelial ovarian cancer. Gynecol Oncol 1989; 35: 301–6

81.

Scarabelli C, Gallo A, Campagnutta E, et al. Splenectomy during primary and secondary cytoreductive surgery for epithelial ovarian carcinoma. Int J Gynecol Cancer 1998; 8: 215–21

82.

Bilgin T, Ozerkan K, Ozan H. Splenectomy in cytoreductive surgery for advanced ovarian cancer. Arch Gynecol Obstet 2005; 27: 329–31

83.

Eisenkop SM, Nalick RH, Wang HJ, et al. Peritoneal implant elimination during cytoreductive surgery for ovarian cancer: impact on survival. Gynecol Oncol 1993; 51: 224–9

84.

Van Dam PA, Tjalma W, Weyler J, et al. Ultraradical debulking of epithelial ovarian cancer with the ultrasonic surgical aspirator: a prospective randomized trial. Am J Obstet Gynecol 1996; 174: 943–50

85.

Kapnick SJ, Griffiths CT, Finkler NJ. Occult pleural involvement in stage III ovarian carcinoma: role of diaphragm resection. Gynecol Oncol 1990; 39: 135–8

86.

Montz FJ, Schlaerth J, Berek JS. Resection of diaphragmatic peritoneum and muscle: role in cytoreductive surgery for ovarian carcinoma. Gynecol Oncol 1989; 35: 338–40

87.

Eisenkop SM, Spirtos NM. Procedures required to accomplish complete cytoreduction of ovarian cancer: is there a correlation with ‘biological aggressiveness’ and survival? Gynecol Oncol 2001; 82: 435–41

88.

Burghardt E, Girardi F, Lahousen M, et al. Patterns of pelvic and paraaortic lymph node involvement in ovarian cancer. Gynecol Oncol 1991; 40: 103–6

89.

Spirtos NM, Gross GM, Freddo JL, et al. Cytoreductive surgery in advanced epithelial cancer of the ovary: the impact of aortic and pelvic lymphadenectomy. Gynecol Oncol 1995; 56: 345–52

90.

Onda T, Yoshikawa H, Yokota H, et al. Assessment of metastases to aortic and pelvic lymph nodes in epithelial ovarian carcinoma: a proposal for essential sites for lymph node biopsy. Cancer 1996; 78: 803–8

91.

Burghardt E, Winter R. The effect of chemotherapy on lymph node metastases in ovarian cancer. Baillière’s Clin Obstet Gynaecol 1989; 3: 167–71

92.

Morice P, Joulie F, Rey A, et al. Are nodal metastases in ovarian cancer chemoresistant lesions? Analysis of nodal involvement in 105 patients treated with preoperative chemotherapy. Eur J Gynaecol Oncol 2004; 25: 169–74

93.

Burghardt E, Pickel H, Lahousen M, et al. Pelvic lymphadenectomy in operative treatment of ovarian cancer. Am J Obstet Gynecol 1986; 155: 315–9

94.

Scarabelli C, Gallo A, Visentin MC, et al. Systematic pelvic and para-aortic lymphadenectomy in advanced ovarian cancer patients with no residual intraperitoneal disease. Int J Gynecol Cancer 1997; 7: 18–26

121

Ch 04 final.qxd

12/18/2006

15:14

Page 122

SURGERY FOR OVARIAN CANCER

95.

96.

Benedetti Panici P, Landoni F, Scarabelli C, et al. Systematic aortic and pelvic lymphadenectomy (SAPL) vs resection of any bulky nodes (RBN) only for optimally debulked advanced ovarian cancer (AOC): preliminary report from an international randomized trial [abstract 132]. Int J Gynecol Cancer 1999; 9 (Suppl 1): 44 Morrow CP, Hernandez WL, Townsend DE, et al. Pelvic celiotomy in the obese patient. Am J Obstet Gynecol 1977; 127: 335–9

108. Vergote I, De Wever T, Decloedt J, et al. Neoadjuvant chemotherapy versus primary debulking surgery in advanced ovarian cancer. Semin Oncol 2000; 27: 31–6 109. Munkarah AR, Hallum AV III, Morris M, et al. Prognostic significance of residual disease in patients with stage IV epithelial ovarian cancer. Gynecol Oncol 1997; 64: 13–17 110. Curtin JP, Malik R, Venkatraman ES, et al. Stage IV ovarian cancer: impact of surgical debulking. Gynecol Oncol 1997; 64: 9–12

97.

Wharton JT, Herson J. Surgery for common epithelial tumors of the ovary. Cancer 1981; 48: 582–9

98.

Morrow CP, Curtin JP. Surgery for ovarian neoplasia. In Gynecologic Cancer Surgery. New York: Churchill Livingstone, 1996: 627–716

111. Lui PC, Benjamin I, Morgan MA, et al. Effect of surgical debulking on survival in stage IV ovarian cancer. Gynecol Oncol 1997; 64: 4–8

99.

Venesmaa P, Ylikorkala O. Morbidity and mortality associated with primary and repeat operations for ovarian cancer. Obstet Gynecol 1992; 79: 168–72

112. Bristow RE, Montz FJ, Lagasse LD, et al. Survival impact of surgical cytoreduction in stage IV epithelial ovarian cancer. Gynecol Oncol 1999; 72: 278–87

100. Chen SS, Bochner R. Assessment of morbidity and mortality in primary cytoreductive surgery for advanced ovarian carcinoma. Gynecol Oncol 1985; 20: 190–5

113. Naik R, Nordin A, Cross PA, et al. Optimal cytoreductive surgery is an independent prognostic indicator in stage IV epithelial ovarian cancer with hepatic metastases. Gynecol Oncol 2000; 78: 171–5

101. Wright JD, Herzog TJ, Powell MA. Morbidity of cytoreductive surgery in the elderly. Am J Obstet Gynecol 2004; 190: 1398–400

114. NIH Consensus Development Panel on Ovarian Cancer. Ovarian cancer: screening, treatment, and follow-up. JAMA 1995; 273: 491–7

102. Trimbos JB, Schueler, JA, van Lent M, et al. Reasons for incomplete surgical staging in early ovarian carcinoma. Gynecol Oncol 1990; 37: 374–7

115. Munoz KA, Harlan LC, Trimble EL. Patterns of care for women with ovarian cancer in the United States. J Clin Oncol 1997; 15: 3408–15

103. Chen LM, Leuchter RS, Lagasse LD, et al. Splenectomy and surgical cytoreduction for ovarian cancer. Gynecol Oncol 2000; 77: 362–8

116. McGowan L. Patterns of care in carcinoma of the ovary. Cancer 1993; 71: 628–33

104. Marsoni S, Torri V, Valsecchi MG, et al. Prognostic factors in advanced epithelial ovarian cancer. Br J Cancer 1990; 62: 444–50

117. Nguyen HN, Averette HE, Hoskins W, et al. National survey of ovarian carcinoma part V: the impact of physician’s specialty on patient’s survival. Cancer 1993; 72: 3663–70

105. Venesmaa P. Epithelial ovarian cancer: impact of surgery and chemotherapy on survival during 1977–1990. Obstet Gynecol 1994; 84: 8–11

118. Prefontaine M, Gruslin A. Referral patterns for suspected ovarian cancer: a survey of practicing gynecologists. Int J Gynecol Cancer 1995; 5: 381–5

106. Olaitan A, Weeks J, Mocroft A, et al. The surgical management of women with ovarian cancer in the south west of England. Br J Cancer 2001; 85: 1824–30

119. Bristow RE, Zahurak ML, del Carmen MG, et al. Ovarian cancer surgery in Maryland: volume based access to care. Gynecol Oncol 2004; 93: 353–60

107. Junor EJ, Hole DJ, McNulty L, et al. Specialist gynecologists and survival outcome in ovarian cancer: a Scottish national study of 1866 patients. Br J Obstet Gynaecol 1999; 106: 1130–6

120. Carney ME, Lancaster JM, Ford C, et al. A population-based study of patterns of care for ovarian cancer: who is seen by a gynecologic oncologist and who is not? Gynecol Oncol 2002; 84: 36–42

122

Ch 04 final.qxd

12/18/2006

15:14

Page 123

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

121. Woodman C, Baghdady A, Collins S, et al. What changes in the organization of cancer services will improve the outcome for women with ovarian cancer? Br J Obstet Gynaecol 1997; 104: 135–9 122. Tingulstad S, Skjeldestad FE, Hagen B. The effect of centralization of primary surgery on survival in ovarian cancer patients. Obstet Gynecol 2003; 102: 499–505 123. Guidelines for referral to a gynecologic oncologist: rationale and benefits. A special issue by the Society of Gynecologic Oncologists. Gynecol Oncol 2000; 78: S5–S8 124. Allen DG, Baak J, Belpomme D, et al. Advanced epithelial ovarian cancer: 1993 consensus statements. Ann Oncol 1993; 4 (Suppl 4): 83–8 125. National Comprehensive Cancer Network (NCCN). Clinical practice guidelines in oncology: ovarian cancer – v.1.2003. http://www.nccn.org/professionals/ physician_gls/PDF/ovarian.pdf 126. Practice guidelines: ovarian cancer. Society of Gynecologic Oncologists Medical Practice and Ethics Committee. Oncology (Huntingt) 1998; 12: 129–33 127. Trimble EL. Patters of care in ovarian cancer. In Jacobs IJ, Shepherd JH, Oram DH, et al., eds. Ovarian Cancer. Oxford: Oxford University Press 2002: 323–9 128. Montazeri A, McEwen J, Gillis CR. Quality of life in patients with ovarian cancer: current state of research. Support Care Cancer 1996; 4: 169–79 129. Chan YM, Ng TY, Ngan HY, et al. Quality of life in women treated with neoadjuvant chemotherapy for advanced ovarian cancer: a prospective longitudinal study. Gynecol Oncol 2003; 88: 9–16 130. Wils J, Blijham G, Naus A, et al. Primary or delayed debulking surgery and chemotherapy consisting of cisplatin, doxorubicin, and cyclophosphamide in stage III–IV epithelial ovarian carcinoma. J Clin Oncol 1986; 4: 1068–73 131. Lawton FG, Redman CW, Luesley DM, et al. Neoadjuvant (cytoreductive) chemotherapy combined with intervention debulking surgery in advanced, unresected epithelial ovarian cancer. Gynecol Oncol 1989; 73: 61–5

132. Ng LW, Rubin SC, Hoskins WJ, et al. Aggressive chemosurgical debulking in patients with advanced ovarian cancer. Gynecol Oncol 1990; 38: 358–63 133. Jacob JH, Gershenson DM, Morris M, et al. Neoadjuvant chemotherapy and interval debulking for advanced epithelial ovarian cancer. Gynecol Oncol 1991; 42: 146–50 134. Lim JT, Green JA. Neoadjuvant carboplatin and ifosfamide chemotherapy for inoperable FIGO stage III and IV ovarian carcinoma. Clin Oncol (R Coll Radiol) 1993; 5: 198–202 135. Vergote I, De Wever I, Tjalma W, et al. Neoadjuvant chemotherapy or primary debulking surgery in advanced ovarian carcinoma: a retrospective analysis of 285 patients. Gynecol Oncol 1998; 71: 431–6 136. Schwartz PE, Rutherford TJ, Chambers JT, et al. Neoadjuvant chemotherapy for advanced ovarian cancer: long-term survival. Gynecol Oncol 1999; 72: 93–9 137. Kuhn W, Rutke S, Spathe K, et al. Neoadjuvant chemotherapy followed by tumor debulking prolongs survival for patients with poor prognosis in International Federation of Gynecology and Obstetrics stage IIIC ovarian carcinoma. Cancer 2001; 92: 2585–91 138. Ansquer Y, Leblanc E, Clough K, et al. Neoadjuvant chemotherapy for unresectable ovarian carcinoma: a French multicenter study. Cancer 2001; 91: 2329–34 139. Kayikciog Lu F, Kose MF, Boran N, et al. Neoadjuvant chemotherapy or surgery in advanced epithelial ovarian carcinoma. Int J Gynecol Cancer 2001; 11: 466–70 140. Shibata K, Kikkawa F, Mika M, et al. Neoadjuvant chemotherapy for FIGO stage III or IV ovarian cancer: survival benefit and prognostic factors. Int J Gynecol Cancer 2003; 13: 587–92 141. Morice P, Brehier-Ollive D, Rey A, et al. Results of interval debulking surgery in advanced stage ovarian cancer: an exposed–non-exposed study. Ann Oncol 2003; 14: 74–7 142. Fanfani F, Ferrandina G, Corrado G, et al. Impact of interval debulking surgery on clinical outcome in primary unresectable FIGO stage IIIC ovarian cancer patients. Oncology 2003; 65: 316–22

123

Ch 04 final.qxd

12/18/2006

15:14

Page 124

SURGERY FOR OVARIAN CANCER

143. EORTC-55971: http://www.eortc.be/protoc/listopen. asp 144. van der Burg MEL, van Lent M, Buyse M, et al. The effect of debulking surgery after induction chemotherapy on prognosis in advanced epithelial ovarian cancer. N Engl J Med 1995; 332: 629–34 145. Redman CW, Warwick J, Luesley DM, et al. Intervention debulking surgery in advanced epithelial ovarian cancer. Br J Obstet Gynaecol 1994; 101:142–6 146. Rose PG, Nerenstone S, Brady MF, et al. Secondary surgical cytoreduction for advanced ovarian carcinoma. N Engl J Med 2004; 351: 2489–97 147. Van der Burg MEL, van Lent M, Kobierska A, et al. Interval debulking surgery significantly increases the survival and progression-free survival in advanced epithelial ovarian cancer patients: an EORTC Gynecologic Cancer Co-operative Group study. In Jacobs IJ, Shepherd JH, Oram DH, et al., eds. Ovarian Cancer. Oxford: Oxford University Press, 2002: 299–303 148. Van der Burg MEL, Vergote I. The role of interval debulking surgery in ovarian cancer. Curr Oncol Rep 2003; 5: 473–81 149. Surwit E, Childers J, Atlas I, et al. Neoadjuvant chemotherapy for advanced ovarian cancer. Int J Gynecol Cancer 1996; 6: 356–61 150. Morrow CP, Curtin JP. Tumors of the ovary: neoplasms derived from the celomic epithelium. In Morrow CP, Curtin JP, eds. Synopsis of Gynecologic Oncology. New York: Churchill Livingstone, 1998: 233–80

154. Cooper BC, Sood AK, Davis CS, et al. Preoperative CA-125 levels: an independent prognostic factor for epithelial ovarian cancer. Obstet Gynecol 2002; 100: 59–64 155. Memarzadeh S, Lee SB, Berek JS, et al. CA-125 levels are a weak predictor of optimal cytoreduction surgery in patients with advanced epithelial ovarian cancer. Int J Gynecol Cancer 2003; 13: 120–4 156. Dowdy SC, Mullany SA, Brandt KR, et al. The utility of computed tomography scans in predicting suboptimal cytoreductive surgery in women with advanced ovarian carcinoma. Cancer 2004; 101: 346–52 157. Brockbank EC, Ind TEJ, Barton DPJ, et al. Preoperative predictors of suboptimal primary surgical cytoreduction in women with clinical evidence of advanced primary epithelial ovarian cancer. Int J Gynecol Cancer 2004; 14: 42–50 158. Nelson BE, Rosenfield AT, Schwartz PE. Preoperative abdominopelvic computed tomographic prediction of optimal cytoreduction in epithelial ovarian carcinoma. J Clin Oncol 1993; 11: 166–72 159. Meyer JI, Kennedy AW, Friedman R, et al. Ovarian carcinoma: value of CT in predicting success of debulking surgery. Am J Roentgenol 1995; 165: 875–8 160. Bristow RE, Duska LR, Lambrou NC, et al. A model for predicting surgical outcome in patients with advanced ovarian carcinoma using computed tomography. Cancer 2000; 89: 1532–40 161. Hacker NF. Cytoreduction for advanced ovarian cancer in perspective. Int J Gynecol Cancer 1996; 6: 159–60

151. Chi DS, Venkatraman ES, Masson V, et al. The ability of preoperative CA-125 to predict optimal primary tumor cytoreduction in stage III epithelial ovarian carcinoma. Gynecol Oncol 2000; 77: 227–31

162. Hightower RD, Nguyen HN, Averette HE, et al. National survey of ovarian carcinoma IV: patterns of care and related survival for older patients. Cancer 1994; 73:377–83

152. Gemer O, Segal S, Kopmar A. Preoperative CA-125 level as a predictor of non optimal cytoreduction of advanced epithelial ovarian cancer. Acta Obstet Gynecol Scand 2001; 80: 583–5

163. Williams L, Brunetto VL, Yordan E, et al. Secondary cytoreductive surgery at second-look laparotomy in advanced ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1997; 66: 171–8

153. Saygili U, Guclu S, Uslu T, et al. Can serum CA-125 levels predict the optimal primary cytoreduction in patients with advanced ovarian carcinoma? Gynecol Oncol 2002; 86: 57–61

164. Bertelsen K, Jakobsen A, Stroyer J, et al. A prospective randomized comparison of 6 and 12 cycles of cyclophosphamide, adriamycin and cisplatin in advanced epithelial ovarian cancer: a Danish

124

Ch 04 final.qxd

12/18/2006

15:14

Page 125

CYTOREDUCTIVE SURGERY: PRINCIPLES AND RATIONALE

Ovarian Study Group trial (DACOVA). Gynecol Oncol 1993; 49: 30–6 165. Hakes TB, Chalas E, Hoskins WJ, et al. Randomized prospective trial of 5 versus 10 cycles of cyclophosphamide, doxorubicin and cisplatin in advanced ovarian cancer. Gynecol Oncol 1992; 45: 284–9

166. Lambert HE, Rustin GJS, Gregory WM, et al. A randomized prospective trial of five versus eight courses of cisplatin or carboplatin in advanced epithelial ovarian carcinoma: a North Thames Ovary Group study. Ann Oncol 1997; 8: 327–33

125

Ch 05 final.qxd

12/18/2006

15:15

Page 127

CHAPTER 5

Cytoreductive surgery: pelvis Robert E Bristow, Leo D Lagasse

INTRODUCTION

REGIONAL ANATOMY

Ovarian cancer is the most lethal of all gynecologic malignancies. The American Cancer Society estimated that 25 580 new cases would be diagnosed in the USA in 2004, with 16 090 deaths directly attributable to this disease.1 Metastatic spread of ovarian cancer to local pelvic structures is a common occurrence, with International Federation of Gynecology and Obstetrics (FIGO) stage IIB–IV disease representing a majority (71.7%) of all patients newly diagnosed with epithelial ovarian cancer.2 In this setting, survival determinants are multifactorial; however, the strongest clinician-driven predictors of clinical outcome are the administration of platinum-based chemotherapy and the amount of residual tumor following primary surgery.3–9 Resection of the primary tumor mass is a key component of the initial cytoreductive surgical effort. In spite of this, the tendency of advanced ovarian cancer to obliterate the normal anatomy of the pelvis may lead to an abbreviated debulking procedure or abandonment of primary surgery altogether.10,11 In some reports, as many as 47% of patients with advanced ovarian cancer may be left with suboptimal large-volume residual pelvic disease.12–14 It is therefore incumbent upon the surgeon operating on women with locally extensive ovarian cancer to be intimately familiar with the relevant pelvic anatomy and skilled in the techniques of radical pelvic cytoreduction that are addressed in the following pages.

Pelvic viscera The visceral anatomy within the pelvis relevant to the ovarian cancer surgeon includes the ovaries and fallopian tubes, the uterus, the bladder, the ureters and the rectosigmoid colon (Figure 5.1). Depending on the patient’s body habitus, abdominal viscera (small bowel, omentum, transverse colon) may also lie within the pelvis. The ovaries are attached to the posterior and lateral pelvic wall via the infundibulopelvic ligaments and the ovarian arteries and veins contained therein. Medially, the ovaries are attached to the uterus through the utero-ovarian ligament, which contains vascular anastomosis between the uterine and ovarian vessels. The fallopian tubes arise from the uterine fundus anterior to utero-ovarian ligaments. The uterus is centrally located within the pelvis, lying ventral to the rectosigmoid colon and dorsal to the bladder, and is composed of two portions: an upper muscular corpus and a lower fibrous cervix. The cervix is covered by the bladder anteriorly, and therefore has no serosa. The posterior cervix does have a serosal layer of peritoneum and, with the proximal vagina, forms the anterior wall of the posterior cul-de-sac (pouch of Douglas). Anteriorly and laterally, the proximal vagina is bounded by the bladder and ureters and the cardinal ligaments, respectively. The bladder lies against the abdominal wall and pubic bones anteriorly, and abuts the obturator internus and levator ani muscles on its lateral and inferior surface. Posteriorly, the bladder is bounded by the proximal vagina and cervix.

127

Ch 05 final.qxd

12/18/2006

15:15

Page 128

SURGERY FOR OVARIAN CANCER

Median umbilical plica and urachus

Inferior epigastric vessels

Round ligament Paravesical fossa External iliac vessels Ovary Fallopian tube Cut edge of peritoneum Sigmoid colon

Aorta Inferior vena cava

Medial umbilical plica and ligament Bladder

Ovarian ligament

Uterus

Uterosacral ligament

Rectum

Ureteric fold Paragenital fossa Infundibulopelvic (suspensory) ligament Cecum Sacral promontory Ureter

a

Ovarian vessels

Sigmoid colon Sacral promontory Fallopian tube Ovary External iliac vessels

Ureter Uterus Uterosacral ligament Posterior cul-de-sac Rectum Vagina Puborectalis muscle External anal sphincter

Round ligament Anterior cul-de-sac Bladder Anterior vaginal fornix Urethra Urogenital diaphragm

b

Figure 5.1 Normal visceral anatomy of the pelvis. (a) Operative viewpoint; (b) sagittal section

The ureter crosses the pelvic brim at the level of the common iliac artery bifurcation, just medial to the ovarian vessels, before descending into the pelvis. Within the pelvis, the ureter lies within a connective tissue sheath and is attached to the medial surface of the lateral pelvic wall peritoneum. The sigmoid colon follows an ‘S’-shaped curvature that begins at the level of the pelvic brim and iliac wing of the left pelvis. While the descending colon is retroperitoneal, the sigmoid colon has a well-defined mesentery, particularly at its central portion. The transition from sig-

128

moid colon to the rectum is marked by an absence of appendices epiploicae and attenuation of the longitudinal muscular bands (taenia coli).

Ligaments and potential spaces The ligamentous structures of the pelvis are not true ligaments in the pure sense (e.g. associated with skeletal joints), but are visceral ligaments and contain varying amounts of smooth muscle, vessels, nerves and other connective tissues. The paired Müllerian ducts and ovaries are derived from the lateral

Ch 05 final.qxd

12/18/2006

15:15

Page 129

CYTOREDUCTIVE SURGERY: PELVIS

abdominopelvic walls. During embryonic development, these structures migrate toward the midline and in the process pull a mesentery of peritoneum out from the pelvic wall. This layer of peritoneum, the broad ligament, covers the central pelvic structures with the exception of the anterior and lateral cervix. The entire parietal subperitoneal surface represents a potential space that can be used as a surgical cleavage plane, given the tendency of ovarian cancer to respect peritoneal lines of demarcation. The round ligaments are extensions of the uterine musculature and represent the female homolog of the gubernaculum testis. They extend from the anterior uterine fundus to the pelvic retroperitoneum just lateral to the deep inferior epigastric vessels and pass into the inguinal canal. The cardinal ligaments and uterosacral ligaments lie at the inferior border of the broad ligament. The cardinal ligaments begin just caudal to the uterine arteries, are relatively discrete structures as they attach to the cervix below the isthmus, and fan out in a somewhat ill-defined manner as they attach to the lateral pelvic

walls. The cardinal ligament contains the uterine, vaginal, inferior vesicle, and middle rectal (hemorrhoidal) arteries and veins, in addition to associated lymphatics and nodal tissue, and separate the paravesicle spaces from the pararectal spaces (see below). The ureter crosses through the cardinal ligament 1–2 cm lateral to the isthmus of the uterus and runs beneath the uterine artery. The posterior component of the cardinal ligament (rectal stalk) contains a major component of the autonomic nerve supply to the bladder and rectum. The uterosacral ligaments originate on the posterolateral surface of the cervix and are seen as paired thickenings of the pelvic peritoneum that straddle the posterior cul-de-sac (pouch of Douglas). At their insertion on the rectal wall, the lower portion of the uterosacral ligaments are contiguous with the rectal pillars (or stalks). The pelvic viscera are separated from one another and the pelvic walls by eight potential spaces (Figure 5.2). These potential spaces are filled with fatty or areolar connective tissue and are two-dimensional

Symphysis Space of Retzius

Inguinal ligament Femoral artery and vein Umbilical artery

Vesicovaginal space

Paravesical space

Vesicouterine ligament

External iliac artery and vein

Cervix Cardinal ligament

Uterine artery

Uterosacral ligament/rectal pillar

Internal iliac artery

Rectovaginal space Rectum

Pararectal space Ureter

Retrorectal space with Waldeyer's fascia

Sacrum

Common iliac artery and vein

Figure 5.2 Eight potential spaces of the pelvis: retropubic space of Retzius, paravesicle spaces (two), vesicovaginal space, pararectal spaces (two), rectovaginal space and presacral space

129

Ch 05 final.qxd

12/18/2006

15:15

Page 130

SURGERY FOR OVARIAN CANCER

until surgically developed, thereby serving as natural cleavage planes and allowing relatively bloodless isolation of diseased tissue or viscera. The retropubic space of Retzius is bounded anteriorly by the symphysis pubis and superior pubic rami, and posteriorly by the bladder. Laterally, the retropubic space is continuous with the paired paravesicle spaces, the point of demarcation being the obliterated umbilical artery. The paired paravesicle spaces are bounded anteriorly by the superior pubic rami and obturator internus fascia, medially by the bladder and vagina, laterally by the external iliac vessels and obturator fossa, and posteriorly by the cardinal ligament. The vesicovaginal space is defined laterally by the vesicouterine ligaments (bladder pillars), the bladder anteriorly, and the vagina and cervix posteriorly. The paired pararectal spaces are defined anteriorly by the cardinal ligament, posteriorly by the lateral sacrum, and bounded laterally and medially by the hypogastric vessels and the rectum, respectively. Within the anteromedial portion of the pararectal space is a recess between the ureter and uterosacral ligament, which can be used to surgical advantage during the course of resecting extensive cul-de-sac tumor. The rectovaginal space is bounded by the vagina anteriorly, the rectum posteriorly and the rectal pillars, contiguous with the uterosacral ligaments, laterally. The presacral or retrorectal space is bounded by the distal sigmoid mesentery and posterior rectal fascia anteriorly, the fascia of the anterior sacrum (Waldeyer’s fascia) posteriorly, and the pararectal spaces laterally.

Vascular anatomy The ovarian arteries arise from the anterior surface of the aorta in the region of the second lumbar vertebra (L2) (Figure 5.3). The right ovarian vein empties directly into the vena cava in this region, while the left ovarian vein empties into the left renal vein. More caudally, the inferior mesenteric artery arises from the anterior surface of the aorta at the level of the third lumbar vertebra (L3), several centimeters below the renal vessels (located at L1/L2) and just caudal to the third portion of the duodenum. The inferior mesen-

130

teric artery gives rise to ascending and descending branches of the left colic artery and continues caudally, giving off three or four sigmoid arteries to the sigmoid colon before becoming the superior hemorrhoidal (or rectal) artery supplying the proximal rectum. The middle sacral artery and vein originate from the dorsal aspect of the terminal aorta and vena cava, respectively, are located within the presacral space, and lie directly on the sacrum. These vessels may be encountered as the presacral space is developed during rectosigmoid colectomy. The aorta bifurcates at the level of L4/L5 into the paired common iliac arteries, which run for a distance of approximately 5 cm before dividing into the external iliac and internal iliac (or hypogastric) arteries. The external iliac artery gives off the inferior epigastric artery before crossing under the inguinal ligament to become the femoral artery and supply the lower extremity. The hypogastric artery provides the primary blood supply to the viscera of the pelvis and musculature of the pelvic wall and gluteal region. The hypogastric artery divides into an anterior and posterior division 3–4 cm after the common iliac bifurcation and thereafter has a variable branching pattern. The posterior division normally gives rise to the iliolumbar, lateral sacral and superior gluteal arteries. Although individual anatomy varies, the anterior division of the hypogastric artery gives off four visceral arterial branches (uterine, superior vesicle, vaginal or inferior vesicle, and middle hemorrhoidal or rectal arteries) and three distinct parietal arteries (obturator, internal pudendal and inferior gluteal). The middle hemorrhoidal arteries supply the proximal rectum below the peritoneal reflection of the posterior culde-sac. The pelvic ureter receives small arterial twigs from the common iliac, hypogastric, uterine and vesicle arteries. The venous drainage of the external iliac system generally follows that of the artery. The tributaries to the hypogastric vein, however, are much more variable and tend to form a large and intricate plexus within the deep lateral pelvis. An accessory obturator vein, draining into the external iliac vein, is present in 25% of patients and is a clinically relevant

Ch 05 final.qxd

12/18/2006

15:15

Page 131

CYTOREDUCTIVE SURGERY: PELVIS

Renal artery

Right ovarian artery Inferior mesenteric artery

Left ovarian artery Left colic artery

Sigmoid arteries Internal iliac artery

Superior rectal artery

External iliac artery Superior vesical artery

Inferior vesical artery

Figure 5.3 Vascular anatomy relevant to pelvic surgery. The venous drainage follows that of the corresponding arterial system with the exceptions of the inferior mesenteric vein, which combines with the splenic vein to form the portal vein, and the left ovarian vein, which drains into the left renal vein

anatomic variant, as this vein may be injured during resection of the obturator lymph nodes.

Lymphatic drainage There are three principle routes of efferent lymphatic drainage from the sub-ovarian plexus15–17 (Figure 5.4). The first route accompanies the ovarian blood vessels to the para-aortic nodal basins after following the ovarian vessels out of the pelvis to the lower poles of the kidneys, turning medially to join the aortic lymphatics. The second pathway of ovarian lymphatic drainage passes from the ovarian plexus through the lymphatics of the broad ligament to the obturator lymph nodes.18 There is a rich anastomotic network

of lymphatics between the obturator nodes and the external iliac, hypogastric, common iliac and paraaortic nodal basins.15,19 Although less clinically significant, the third route of lymphatic drainage from the ovary runs through the round ligament to the external iliac and inguinal nodal basins.16,19 With involvement of the external iliac nodes, there may also be retrograde metastatic flow to the inguinal lymph nodes secondary to obstructed proximal drainage. Because locally advanced ovarian cancer involves other pelvic viscera (uterus, rectosigmoid colon, peritoneum), retroperitoneal lymph node involvement may follow the primary drainage routes of any and all of these structures after secondary involvement. In particular,

131

Ch 05 final.qxd

12/18/2006

15:15

Page 132

SURGERY FOR OVARIAN CANCER

Aortic nodes Ureter

Superior gluteal (hypogastric) nodes

Subaortic (presacral) nodes Common iliac nodes Tumor

External iliac node Obturator nodes Hypogastric nodes Obturator nerve Inferior epigastric artery Deep circumflex iliac artery

Figure 5.4 Lymphatic drainage of the ovary. The three principle routes are to: (1) the para-aortic nodal basins accompanying the ovarian vessels; (2) the obturator and iliac nodal basins through the broad ligament; and (3) the external iliac and inguinal nodal basins via the round ligament

the paracolic nodes drain the rectosigmoid colon above the peritoneal reflection and follow the course of the superior hemorrhoidal artery.

Neural anatomy The nerves of the lumbar plexus are the most commonly encountered neural structures during pelvic tumor reductive surgery (Figure 5.5). The lateral femoral cutaneous nerve (L2–L3) emerges from the lateral border of the psoas muscle and passes under the inguinal ligament just medial to the anterior superior iliac spine. The genitofemoral nerve (L1–L2) pierces the psoas fascia more medially and divides into two branches; the femoral branch follows the course of the external iliac artery and supplies cutaneous sensation to the anterior thigh, while the genital branch follows the course of the round ligament as it passes

132

lateral to the inferior epigastric vessels and traverses the inguinal canal before supplying sensation to the groin and lateral external genitalia. The obturator nerve (L2–L4) is located deeper within the lateral pelvis, emerging from the medial border of the psoas muscle, traversing the obturator fossa ventral to the obturator artery and vein, and exiting the pelvis through the obturator canal in the obturator internus muscle. The obturator nerve innervates the adductor muscles of the thigh (gracilis, adductor longus, adductor brevis, adductor magnus and obturator internus), providing adduction, flexion and lateral rotation. An accessory obturator nerve will be present in 5–25% of patients and follows the medial surface of the psoas muscle to exit the pelvis by passing over the superior pubic ramus. The femoral nerve (L2–L4) emerges from the lateral border of the psoas muscle and

Ch 05 final.qxd

12/18/2006

15:15

Page 133

CYTOREDUCTIVE SURGERY: PELVIS

Lateral femoral cutaneous nerve 4th lumbar nerve

Genitofemoral nerve

5th lumbar nerve Obturator nerve Lumbosacral trunk Sacral sympathetic trunk 1st sacral nerve (ventral or anterior ramus) 2nd sacral nerve 3rd sacral nerve 4th sacral nerve 5th sacral nerve Coccygeal nerve Nerves to coccygeus and levator ani

Femoral nerve Superior gluteal artery Superior gluteal nerve Sciatic nerve Inferior gluteal artery Internal pudendal artery Pelvic splanchnic (visceral) nerves (Internal) pudendal nerve

Figure 5.5 Relevant neural anatomy of the pelvis: lumbar plexus and sacral plexus

continues distally between the psoas and iliacus muscles, deep to the psoas fascia, before passing under the inguinal ligament. The femoral nerve supplies the four components of the quadriceps muscle (rectus femoris, vastus lateralis, vastus medialis and vastus intermedius) in addition to the iliopsoas muscle and sartorius muscle, providing flexion of the thigh and extension of the leg. The sacral plexus of nerves, lying deep within the pelvic sidewall and formed by the lumbosacral trunk (L4–L5) and the ventral rami of first three sacral nerves (S1–S3), is rarely encountered during pelvic surgery for ovarian cancer (Figure 5.5). The sciatic nerve (L4–S3) exits the pelvis through the greater sciatic foramen, and forms a tibial division and common peroneal division, supplying the muscles of the posterior thigh and leg. The pudendal nerve (S2–S4) exits the pelvis through the greater sciatic foramen, passes around the ischial spine and re-enters the pelvis through the lesser sciatic foramen before accompanying the pudendal vessels into the pudendal canal. The pudendal nerve supplies the muscles of the perineum, including the external anal sphincter.

The autonomic nervous system contributes both sympathetic and parasympathetic innervation to the pelvis. The sympathetic component is formed in part by the superior hypogastric plexus, which is located just below the aortic bifurcation and gives rise to right and left hypogastric nerves. The hypogastric nerves descend into the pelvis anterior to the sacrum and onto the sidewalls, where they are joined by the parasympathetic preganglionic pelvic splanchnic nerves (S2–S4) to form the right and left inferior hypogastric plexuses (Figure 5.6). These nerves supply the descending colon, pelvic colon, rectum and anus. In addition, each inferior hypogastric plexus surrounds the corresponding hypogastric artery, distributing branches containing both sympathetic and parasympathetic fibers to the pelvic viscera (e.g. vesicle plexus, uterovaginal plexus).

PELVIC CYTOREDUCTIVE SURGERY Ovarian cancer spreads by four principle routes: direct extension, peritoneal dissemination, retroperitoneal

133

Ch 05 final.qxd

12/18/2006

15:15

Page 134

SURGERY FOR OVARIAN CANCER

Superior hypogastric plexus (presacral nerve) Inferior hypogastric (pelvic) plexus Right hypogastric nerve

Vesical plexus Parasympathetic nerves or nervi erigentes Rectal plexus

Uterovaginal plexus

Figure 5.6 Autonomic nervous system of the pelvis

lymphatics and hematogenous dissemination. The first three spread patterns may be present as isolated findings or occur simultaneously within the pelvis. The primary tumor mass may involve contiguous structures via local expansion with compression or invasion of pelvic viscera (rectosigmoid colon, genitourinary tract) with absent or only a minor component of carcinomatosis. Alternatively, individual tumor implants may be numerous with extensive coverage of virtually all visceral and parietal pelvic peritoneal surfaces or progress to the extent that individual tumor nodules coalesce into a confluent sheet of tumor encasing the pelvic viscera. Retroperitoneal lymph node metastases have been documented in as many as 78.6% of patients with advanced-stage ovarian cancer and may, on occasion, represent the bulk of extraovarian disease.20 In most cases, successful cytoreduction of even extensive pelvic disease can be accomplished using the techniques described in the following sections. The details of the preoperative preparation program for ovarian cancer surgery, including throm-

134

boembolic prophylaxis and bowel preparation, presented in Chapter 2, are particularly important for patients undergoing radical pelvic tumor cytoreduction. The patient may be positioned in the dorsal lowlithotomy (perineolithotomy) position using Allen Universal Stirrups (Allen Medical Systems, Cleveland, OH) or supine on the operating table. The low-lithotomy position is preferable when extensive pelvic surgery for ovarian cancer is anticipated, as it permits intraoperative bimanual examination to ascertain the extent of cul-de-sac tumor involvement accurately, and allows access to the perineum should resection and reanastomsis of the rectosigmoid colon be required. Abdominal entry and exposure are best achieved through a midline xyphopubic incision with placement of a self-retaining retractor (e.g. Bookwalter, Codman Division, Johnson & Johnson, Piscatawy, NJ). A preliminary assessment is made of the extent of disease, with particular attention to the feasibility of resecting upper abdominal disease. Directing initial cytoreductive efforts toward bulky upper abdominal disease and exploring the abdominal

Ch 05 final.qxd

12/18/2006

15:15

Page 135

CYTOREDUCTIVE SURGERY: PELVIS

retroperitoneum will facilitate exposure to the pelvis and ensure a reasonable likelihood of achieving an optimal (< 1 cm) or complete (no macroscopic residual) overall resection prior to undertaking the radical pelvic dissection. Such an approach allows the surgeon to plan the procedure more effectively by determining the number of major procedures (e.g. bowel resections) that will be required and developing a logical and sequential strategy for the overall operation.

RADICAL OOPHORECTOMY Radical extirpative procedures for locally advanced ovarian cancer have evolved since 1965, when Barber and Brunschwig first reported on 22 patients undergoing pelvic exenteration.21 In this series, the postoperative mortality rate was 23% and there were only two long-term survivors. In 1968 and 1973, Hudson and Chir published two reports describing a technique they termed ‘radical oophorectomy’, specifically designed for the intact removal of a fixed ovarian tumor en bloc with attached peritoneum and surrounding structures.22,23 These authors advocated a

a

retroperitoneal approach, using the ‘false capsule’ of the ovarian tumor within the pouch of Douglas to effect en bloc excision. During the past three decades, varying terminology has been used to describe modifications of this procedure, including: en bloc rectosigmoid colectomy,24–29 reverse hysterocolposigmoidectomy,30 complete parietal and visceral peritonectomy,31 en bloc pelvic peritoneal resection of the intrapelvic viscera,32 modified posterior exenteration33 and low anterior resection. Terminology aside, the cardinal feature of the radical oophorectomy procedure is the retroperitoneal approach to ovarian cancer encasing the pelvic viscera, utilizing the tendency of epithelial ovarian cancer to respect peritoneal planes of demarcation to surgical advantage, particularly if the pelvic organs can no longer be clearly identified (Figure 5.7). In this fashion, the retroperitoneal spaces, uninvolved by extensive intraperitoneal tumor, can be used to develop the dissection in a centripetal fashion with maximum safety to surrounding vital structures. The indications for radical oophorectomy have been summarized by Eisenkop et al. (Table 5.1).33 Relative contraindications to the procedure included

b

Figure 5.7 Locally advanced epithelial ovarian cancer with (a) confluent extension to reproductive organs obscuring normal tissue planes; (b) pelvic peritoneal carcinomatosis of the anterior pelvis and sidewalls. The circumscribing peritoneal incisions have been extended into the pelvis

135

Ch 05 final.qxd

12/18/2006

15:15

Page 136

SURGERY FOR OVARIAN CANCER

Table 5.1 Criteria for radical oophorectomy. From reference 33 1.

Gross evidence of ovarian cancer supported by frozen section biopsy

2.

Extensive confluent tumor involvement of one or both adnexae and their adjacent peritoneum, cul-de-sac, posterior uterine serosa (if present) and the sigmoid colon

3.

The surgeon subjectively judges that complete removal of disease could not be effected by simple hysterectomy and salpingo-oophorectomy and piecemeal dissection or resection/ablation of serosal and peritoneal metastases

4.

An overall optimal resection would be otherwise achievable

5.

The procedure is not medically contraindicated

a Gynecologic Oncology Group (GOG) performance status score of ≥ 3 (Karnofsky score ≤ 30–40) and/or tumor distribution precluding an attempt at optimal resection: extensive tumor infiltration of small bowel mesenteric root, celiac axis nodal involvement, unresectable involvement of the porta hepatis, largevolume (≥ 1 cm) unresectable extra-abdominal metastasis (e.g. pulmonary), or multiple unresectable parenchymal liver metastases. Rarely, a radical oophorectomy procedure may be performed in the setting of overall suboptimal residual disease as a means of relieving a concomitant or impending large bowel obstruction.

Classification system To define the scope of surgical resection with uniform terminology, a descriptive classification system has been applied to the radical oophorectomy procedure.34 Briefly, a type I radical oophorectomy consists of a retrograde modified radical hysterectomy (resection of medial parametria and proximal vagina) with en bloc resection of the adnexae, pelvic cul-de-sac tumor and involved pelvic peritoneum. The procedure may also include excision of the peritoneum and/or serosa of the anterior sigmoid colon or a limit-

136

ed full-thickness segmental wedge-shaped resection of the anterior wall of the sigmoid colon. A type II radical oophorectomy is broadened to include an en bloc resection of the rectosigmoid colon below the peritoneal reflection with complete parietal and visceral pelvic peritonectomy. Finally, the type III radical oophorectomy is an extension of the type I or II procedure incorporating a portion of urinary bladder and/or pelvic ureter. In a large contemporary series of patients undergoing the pelvic retroperitoneal approach to locally advanced ovarian cancer by Benedetti-Panici et al., the distribution of procedures according to this classification system was: type I, 18%; type II, 74%; type III, 8%.14

Surgical technique The radical oophorectomy procedure is initiated by incising the paracolic gutters bilaterally and mobilizing the cecum, terminal ileum and sigmoid colon. The paracolic gutter incisions are extended caudally into the pelvis, along the psoas muscles, moving ventromedially along the posterior margin of the symphysis pubis. All pan-pelvic disease is circumscribed and included within this peritoneal incision (Figure 5.8). The pelvic dissection proceeds in a centripetal fashion. Extensive tumor infiltration of the pelvis can obscure the round ligaments, in which case they can be located retroperitoneally, ligated and divided as laterally as possible. The pararectal and paravesicle spaces are developed using a combination of sharp and blunt dissection, exposing the cardinal ligament. The ureters are identified within the pararectal space and mobilized from their attachments to the medial leaf of the broad ligament, moving from the pelvic brim to the tunnel of Wertheim, and held for traction with vasa-loops. The central pelvic tumor mass should be devascularized early in the course of the operation by securing the infundibulopelvic ligaments (containing the ovarian vessels) with suture ligatures and dividing them at or above the pelvic brim. The anterior pelvic peritoneum is grasped and placed on traction with Allis clamps and the retropubic space of Retzuis developed. A plane of dissection

Ch 05 final.qxd

12/18/2006

15:15

Page 137

CYTOREDUCTIVE SURGERY: PELVIS

Tumor mass

Uterus

Sigmoid colon

Round ligament Ureter preserved Bladder

External iliac artery

Uterine vessels

Figure 5.8 Radical oophorectomy: a circumscribing peritoneal incision extends from the paracolic gutters into the pelvis to encompass the pan-pelvic disease; the round ligaments are divided as far laterally as possible; the anterior pelvic peritoneum with associated tumor is dissected from the bladder muscularis; the ovarian vessels are divided at or above the pelvic brim; the ureters are widely mobilized

is established between the anterior pelvic peritoneum and the bladder dome muscularis using the electrosurgical unit (ESU) or argon beam coagulator. The anterior pelvis is then deperitonealized, moving ventral to dorsal and lateral to medial toward the uterus until the pubo–vesico–cervical fascia is reached (Figure 5.9). The uterine vascular pedicles are skeletonized, doubly ligated and divided at the level of the ureters, allowing additional lateral displacement of the ureters from the central specimen (Figure 5.10a). The ureters are extricated from within the bladder pillars by developing the ureteral tunnel using a rightangle clamp and suture, ligating and dividing both the anterior and posterior leaves of the ureteral tunnel, allowing them to be reflected laterally out of the field of dissection (Figure 5.10b). The bladder is sharply

Figure 5.9 Radical oophorectomy: anterior pelvic peritonectomy. The anterior pelvis is deperitonealized by dissecting in the subperitoneal plane using the electrosurgical unit (the round ligaments have not yet been divided)

137

Ch 05 final.qxd

12/18/2006

15:15

Page 138

SURGERY FOR OVARIAN CANCER

Uterus

Tumor mass

Bladder a

b

Figure 5.10 Radical oophorectomy: modified radical hysterectomy ureteral dissection. (a) The uterine vascular pedicle is developed with a right-angle clamp, doubly suture ligated and divided at the level of the ureter. (b) The ureter is extricated by developing the ureteral tunnel, ligating and dividing both anterior and posterior portions of the bladder pillar

mobilized ventrocaudally to expose the proximal 2–3 cm of the vagina. If the anterior pelvic peritoneal tumor is densely adherent to the cervix, the paravesicle spaces can be further developed in a lateral to medial direction below the level of the cervix until the vesicovaginal space is reached. The vesicovaginal space is virtually always free of disease and can be used to define the proper plane of dissection between the adherent tumor and the bladder wall. Occasionally, an intentional cystotomy in the bladder dome will be necessary to identify the proper plane.35 The tumor-laden peritoneum is retracted cephalad while the bladder is mobilized inferiorly off the anterior vaginal wall. The hysterectomy is completed in a retrograde fashion by first creating an anterior colpotomy. An intraoperative bimanual examination or placement of a spongestick into the anterior vaginal fornix will facilitate selecting the proper site for incising the anterior vaginal wall transversely with the ESU 1–2 cm below the cervicovaginal junction, exposing

138

the inner vagina (Figure 5.11). Heaney clamps are used to circumscribe the anterior and lateral vagina, dividing and securing each pedicle with a suture ligature in sequence. Transabdominal placement of a narrow malleable retractor in the vagina will help to displace the bladder and distal ureters anterolaterally during this portion of the dissection. The posterior vaginal wall is incised with the ESU and the rectovaginal space developed caudally, until the lowermost extent of the cul-de-sac tumor has been reached and bypassed for a distance of 2–3 cm (Figure 5.11, inset). The retrograde approach is continued by retracting the cul-de-sac tumor mass sharply upward, exposing the remaining cardinal ligament attachments medial to the ureters, the uterosacral ligaments and the rectal pillars, which are sequentially divided between clamps and secured with suture ligatures (Figure 5.12). The maximum amount of ventral displacement of the cul-de-sac tumor is achieved by sharp dissection from the anterior wall of the rectum, moving superiorly toward the rectosigmoid junction

Ch 05 final.qxd

12/18/2006

15:15

Page 139

CYTOREDUCTIVE SURGERY: PELVIS

Tumor mass and uterus retracted

Bladder

Figure 5.11 Radical oophorectomy: the retrograde approach to hysterectomy. An anterior colpotomy is created using the electrosurgical unit, exposing the inner vagina, and the vaginal tube circumscribed with clamps, each pedicle divided and

Figure 5.12 Radical oophorectomy: the retrograde approach

secured in sequence; the posterior vaginal wall is incised and

to hysterectomy. The remaining attachments of the cardinal

the rectovaginal space developed (inset)

ligaments, uterosacral ligaments and rectal pillars are divided between clamps, and suture ligated

(Figure 5.13). The radical oophorectomy is then completed, using either the type I or type II modification, depending on the extent of involvement of the anterior rectal wall and sigmoid colon. Peritoneal edge

Type I modification The type I procedure is appropriate for cul-de-sac disease with no or only limited involvement of the rectosigmoid colon such that the posterior broad ligament peritoneum is resected in its entirety, en bloc with the central specimen, down to the lowermost extent of pelvic tumor. With the ureters reflected laterally, bilateral incisions are created from the previously ligated infundibulopelvic ligament stumps along the lateral pelvic gutters to the cul-de-sac, incorporating the posterior broad ligament peritoneum and paraovarian fossae (Figure 5.14). With only superficial involvement, relatively large areas of the cul-de-sac peritoneum can be sharply dissected from the anterior surface of the rectum and distal sigmoid colon, making bowel resection unnecessary. If the tumor

Vaginal cuff

Figure 5.13 Radical oophorectomy: the retrograde approach to hysterectomy. The cul-de-sac tumor is sharply dissected from the

anterior

rectal

wall,

achieving

maximal

upward

displacement of the pelvic tumor mass

139

Ch 05 final.qxd

12/18/2006

15:15

Page 140

SURGERY FOR OVARIAN CANCER

penetrates into the muscularis of the colon but is of limited (≤ 2 cm) longitudinal extent, a full-thickness, ‘wedge-shaped’ segment of anterior rectal wall can be sharply excised and the defect repaired using a singlelayer closure of fine monofilament (3-0 or 4-0 polypropylene) suture in interrupted inverting stitches, incorporating minimal mucosa, placed perpendicular to the long axis of the bowel (Figure 5.15). The resulting defect may also be closed with a modification of the triangulation stapling technique (see below) using two firings of the TA automated stapling device (4.8 mm) placed at a 60° angle to one another.

Type II modification The type II modification is indicated for pan-pelvic disease with obliteration of posterior pelvic tissue planes and extensive involvement of the cul-de-sac and rectosigmoid colon (i.e. frozen pelvis). It is an extension of the type I procedure with an en bloc resection of the rectosigmoid colon, which is typically classified according to the distance between the anal verge and the distal margin of resection: high, greater

than 11 cm; low, 7–11 cm; very low, less than 7 cm. The type II modification is the most frequently performed variation of the radical oophorectomy procedure.25 In most experienced centers, 25–40% of cases of primary surgery for advanced-stage ovarian cancer include an en bloc resection of the rectosigmoid colon.14,25,36–39 Division of the proximal sigmoid colon can be performed whenever it is most convenient during the operation, once it has been determined that bowel resection is necessary to achieve an optimal surgical result. The paracolic gutter incisions are extended upward with wide mobilization of the cecum, ascending colon, descending colon and proximal sigmoid colon. The sigmoid colon is divided 2–3 cm above the most proximal extent of gross tumor in an area that is free of diverticuli. Depending on the planned anastomosis, a variety of methods can be used to divide the bowel; however, the GIA automated stapling device (4.8 mm) is the most expedient and has the advantage of placing two rows of staples on either side of the divided bowel, thus controlling both proximal and

Figure 5.14 Radical oophorectomy: type I modification. The peritoneal ‘wings’ of the posterior broad ligament have been

Figure 5.15 Radical oophorectomy: type I modification. The

fully developed and extended down to the cul-de-sac

resection may include a full-thickness ‘wedge’-shaped portion

peritoneum, which is sharply dissected from the anterior rectal

of the anterior rectal wall with a single-layer closure of

wall

inverting interrupted fine monofilament suture

140

Ch 05 final.qxd

12/18/2006

15:15

Page 141

CYTOREDUCTIVE SURGERY: PELVIS

distal fecal contamination. Alternatively, bowel clamps, or a combination of a bowel clamp and pursestring device (automated stapler or manual suture), can also be employed. Both ureters are reflected laterally out of the field of dissection and the sigmoid colon separated from its attachments to the iliac fossa on the left. To ensure an adequate resection of mesocolon, the peritoneal incision is extended along the sigmoid mesentery from the point of proximal bowel division medially toward the right sacroiliac joint to join the right-sided circumscribing pelvic peritoneal incision, thus incorporating a ‘wedge’ of colonic mesentery with the central pelvic tumor. Available data indicate that secondary spread to the mesocolic lymph nodes occurs in 37.9–70% of cases in which ovarian cancer locally invades the sigmoid colon, resembling the metastatic spread of primary malignancies of the colon and rectum.37,40,41 Consequently, regional resection of the sigmoid mesentery to include the paracolic and intermediate nodal basins (similar to procedures performed for

colorectal cancer) may be indicated, if it will contribute to an overall macroscopically disease-free surgical result (Figure 5.16). The deep dissection proceeds from the pararectal spaces posteriorly and medially, behind the sigmoid vessels and inferior mesenteric artery (continuing as the superior hemorrhoidal artery) toward the entrance of the presacral space just caudal to the sacral promontory. Individual vessels within the sigmoid mesentery are isolated, clamped and suture ligated as they are encountered. If the inferior mesenteric artery and vein are sacrificed, care must be taken to preserve the left colic artery with its blood supply to the descending colon. From a functional standpoint, the pararectal and presacral potential spaces are unified into one large posterior pelvic retroperitoneal space (Figure 5.17). The posterior pelvis is further mobilized by developing the presacral space caudally to the level of pelvic floor musculature, where the pre-sacral ligament is sharply divided in the midline with electrocautery just below the sacral curvature. The dissection should remain anterior to the presacral fascia, as the presacral veins lie just beneath and may produce troublesome bleeding if injured. Once the

Intermediate nodes

Paracolic nodes

Figure 5.17 Radical oophorectomy: type II modification. The Figure 5.16 Lymphatic drainage of the rectosigmoid colon,

proximal sigmoid colon and inferior mesenteric vessels are

showing paracolic and intermediate nodes that may be included

divided and the sigmoid mesentery mobilized anteriorly; the

within

pararectal spaces are developed in continuity with the presacral

the

scope

of

resection

to

achieve

cytoreduction of subclinical mesenteric adenopathy

complete

space and the dissection developed down to the pelvic floor

141

Ch 05 final.qxd

12/18/2006

15:15

Page 142

SURGERY FOR OVARIAN CANCER

presacral space has been adequately developed, the lateral ligaments (or stalks) of the rectum, containing the middle hemorrhoidal vessels, are divided between clamps and ligated, achieving additional mobilization by detaching the rectosigmoid colon from its attachments to the lateral pelvic wall. Any remaining mesorectal attachments can be taken down with electrocautery or divided between clamps. At this level, preservation of the inferior hypogastric plexus, seen as a dense plaque of nerve tissue that comes close to the rectum at the level of the upper vagina and levator muscles, is essential for anal continence. The cul-de-sac tumor mass is mobilized by exposing the rectovaginal space and dissecting retrograde (proximally) to the point at which the tumor no longer follows peritoneal planes of demarcation and directly invades the bowel wall. It is often possible to gain significant length of the distal bowel by lifting the tumor mass into the incision and straightening the rectum (Figure 5.18). The bowel wall should be adequately cleared of surrounding fat and any remaining mesorectal attachments. The rectum or rectosigmoid

junction is then divided 2–3 cm distal to the lowermost extent of tumor between a TA (4.8 mm) stapler and a proximal bowel clamp (Figure 5.19). The GIA automated stapling device (4.8 mm) can also be used for this purpose, but can be cumbersome when operating deep in the pelvis. Alternatively, bowel clamps can be placed proximal and distal to the level of resection if a hand-sewn anastomosis is planned. The central pelvic tumor mass is removed en bloc with the rectosigmoid colon (Figure 5.20). The resulting operative site is macroscopically tumor free (Figure 5.21). The type II radical oophorectomy is the procedure of choice for complete removal of extensive cul-de-sac disease, creating a ‘false capsule’ of tumor consisting of parietal and visceral pelvic peritoneum delineated by the posterior broad ligament and pelvic sidewall peritoneum laterally, the uterus and proximal vaginal wall anteriorly, and the rectosigmoid colon posteriorly (Figure 5.22). Following resection of the central pelvic tumor, consideration should be given to performing any additional pelvic debulking procedures and reconstructing the urinary tract, if

Figure 5.18 Radical oophorectomy: type II modification after prior hysterectomy. A spongestick has been placed into the vagina to facilitate anterior colpotomy and proximal vaginectomy (held with a Kocher clamp); the lateral ligaments of the rectum and mesorectum have been divided; the

Figure 5.19 Radical oophorectomy: type II modification. The

specimen is lifted sharply upward, straightening the rectum to

rectum is divided between a TA stapler and proximal bowel

preserve maximal length prior to resection

clamp to complete the resection

142

Ch 05 final.qxd

12/18/2006

15:15

Page 143

CYTOREDUCTIVE SURGERY: PELVIS

necessary, prior to re-establishing intestinal continuity to reduce the risk of disrupting the intestinal anastomosis. Intestinal continuity can be re-established by a variety of methods using either automated stapling devices or hand-sewn techniques. The use of automated stapling devices is at least as safe as a conven-

tional hand-sewn anastomosis (depending on operator skill and experience) and in many circumstances offers distinct advantages. By and large, a stapled low rectal anastomosis is faster and has comparable complication rates, and the device is easy to use given the limited access to suturing in the deep pelvis, even after removal of the internal genitalia and rectosigmoid

Figure 5.20 Radical oophorectomy: type II modification. En bloc specimen including uterus, adnexae, anterior pelvic peritoneal tumor, cul-de-sac tumor mass and rectosigmoid

Figure 5.21 Radical oophorectomy: type II modification. The

colon

pelvis is macroscopically tumor-free following en bloc resection

a

b

Figure 5.22 Radical oophorectomy: type II modification. (a) and (b) Extensive cul-de-sac disease is encompassed by the tumor’s ‘false capsule’ of parietal and visceral pelvic peritoneum

143

Ch 05 final.qxd

12/18/2006

15:15

Page 144

SURGERY FOR OVARIAN CANCER

colon. Although the end-to-end anastomosis is most commonly employed, an end-to-side or side-to-side anastomosis may be more desirable under certain circumstances. The anastomosis selected is determined largely by the fit and surgeon preference. Irrespective of the type of procedure, a successful colorectal anastomosis requires an adequate blood supply based on the premise that the lower/middle rectum can be sustained by the inferior hemorrhoidal vessels and that the descending colon can be sustained by the left colic branch of the inferior mesenteric artery or the middle colic artery via the marginal artery of Drummond. The anastomosis should be watertight and have meticulous hemostasis, fecal contamination should be minimized and the absence of tension on the anastomotic staple or suture line should be assured. The divided end of the proximal colon is brought down to the rectal stump to confirm that the anastomosis will be tension-free. Several maneuvers can be employed if additional mobilization of the sigmoid and descending colon is required:31 (1) Any remaining peritoneal attachments of the infracolic omentum to the transverse colon and splenic flexure are divided. (2) The lateral peritoneal attachments of the descending colon are divided, including the splenico-colic ligament, with elevation and medial mobilization of the entire left mesocolon. (3) The lesser sac is opened and the gastrocolic ligament is divided from the greater curvature of the stomach, allowing the splenic flexure to ‘drop’. (4) The mesentery of the descending colon is incised in a direction parallel to the marginal artery of Drummond, ligating and dividing individual arteries, including the left colic artery, up to the splenic flexure and extending medially, again parallel to the marginal artery, up to the middle colic artery, allowing the descending mesocolon to expand or ‘fan’ out. This maneuver assumes a competent marginal artery, which

144

must be preserved so as not to interrupt the anastomotic vascular arcade. (5) The inferior mesenteric vein is ligated and divided just below the inferior border of the pancreas. Transient vascular congestion of the left colon is commonly observed after this maneuver. Once the above conditions have been satisfied, the anastomosis is completed by one of the following techniques. End-to-end anastomosis: stapled techniques

The transanal double-stapling technique using the circular end-to-end anastomosis (CEEA) is a safe and efficient method of re-establishing intestinal continuity.42–46 This technique has yielded excellent functional results in patients undergoing low anterior resection for rectal cancer, with anastomostic leak rates ranging from 2.8% to 8%.44,45,47 With this technique, the rectum is divided and closed by means of a linear stapler and the need for a distal purse-string suture is eliminated. The largest circular stapler that will be comfortably accommodated by both bowel segments should be used, as these anastomoses have a predisposition toward stricture and stenosis. In general, at least a 28-mm or 31-mm stapler should be used. Blunt rectal sizing instruments are gently passed into the proximal and distal bowel segments. The proximal sigmoid colon is the most common site of anastomosis and may be of significantly smaller caliber than the rectal stump. Several maneuvers can be used to accommodate a discrepancy in lumenal diameter: the rectal sizing instruments or a Foley catheter balloon can be used for gentle dilatation; relaxation of intestinal spasm can be induced by topical application of 1% lidocaine to the cut edges of the bowel wall, or by intravenous administration of glucagon (2 mg) or papaverine (30–65 mg); the proximal bowel segment can be divided at an oblique angle using the GIA stapler, thereby increasing the functional surface area for anastomosis; or additional sigmoid colon can be resected, such that the anastomosis is performed using the larger capacity lumen of the descending colon.

Ch 05 final.qxd

12/18/2006

15:15

Page 145

CYTOREDUCTIVE SURGERY: PELVIS

The proximal colon is prepared by inserting a 2-0 polypropylene purse-string suture using the automated purse-string stapling device, the manual suture purse-string clamp, or placing it by hand. If the manual suture purse-string clamp is used, a 2-0 prolene suture on a Keith needle is passed through the openings on either end of the clamp to create the pursestring. If the purse-string is placed by hand, a throughand-through stitch placed 2–3 mm from the bowel edge (weave stitch) is preferable to the whip stitch, so as to ensure a tight approximation of the bowel edges to the anvil shaft. The CEEA anvil is inserted and the purse-string suture tied within the notch on the anvil shaft (Figure 5.23). It is imperative that the bowel wall be cleared of fat for a distance of 2–3 cm proximal to the planned site of anastomosis, as this area will be enclosed within the stapler anvil and inverted so as to become the proximal (or upper) ‘donut’. The disposable trocar is inserted into the shaft of the cartridge head of the main CEEA instrument and

withdrawn into the cartridge by counterclockwise rotation of the wing nut. The main CEEA instrument is lubricated and inserted into the rectum transanally, with depression of the handle toward the floor to navigate the cartridge head past the coccyx, and approximated against the staple line of the rectal stump. The wing nut is rotated clockwise until the trocar pierces the closed rectum adjacent to or through the staple line, such that when the anastomosis is complete, the linear row of staples will be partially excised. The knife will bend the staple at the intersection of the linear and circular staple lines rather than cut it, and the intersecting staple is usually transferred to the removed donut.43 If there is concern that the colonic tissue between the circular and linear rows of staples may become ischemic, the trocar can be directed 2 cm anterior or posterior to the linear staple line, thereby avoiding a staple-to-staple intersection in the anastomotic ring. The disposable trocar is removed from the shaft once it has been completing extended. The anvil

Figure 5.23 Circular end-to-end anastomosis (CEEA) using the automated CEEA stapler. The CEEA anvil is introduced into the proximal colon and the purse-string suture tied within the notch on the anvil shaft; the main CEEA instrument is passed transanally and the trocar advanced through the rectal stump; the trocar is removed and the anvil shaft inserted into the cartridge shaft of the main CEEA instrument (inset)

145

Ch 05 final.qxd

12/18/2006

15:15

Page 146

SURGERY FOR OVARIAN CANCER

shaft is inserted into the cartridge shaft of the main CEEA instrument until it engages with an audible click (Figure 5.23, inset). The wing nut is rotated clockwise, the bowel being kept on stretch and care being taken not to incorporate excess tissue in the anastomosis, until the color-bar indicator on the handle is in the proper position, indicating adequate compression of the bowel wall. The wing nut should not be over-rotated. Displacement of the anastomosis ventrally and cephalad into the pelvis will help to attenuate the rectal wall and prevent a redundant fold of tissue from being incorporated into the staple line. The staple line is inspected by rotation of the main CEEA instrument to the right and to the left. If the staple line is unsatisfactory, the instrument should be released and the purse-string suture revised as necessary. To complete the anastomosis, the safety catch on the main CEEA instrument is released and the stapler fired by squeezing the handles together as far as they will go until an audible crunch of the staples being engaged is heard, thus releasing two circular rows of staples and making a circumferential cut inside the innermost staple ring. The handles are released and the wing nut rotated counterclockwise no more than 360°, or one full turn. The manufacturer’s instructions for releasing the CEEA stapler should be followed carefully to avoid disrupting the staple line during removal. The handpiece of the main CEEA instrument is then rotated 90° and gently moved ventrally (upward), dorsally (downward), to the right and to the left to release the stapled anastomosis from the surrounding colon soft tissue and withdrawn. Placement of a suture in the anterior staple line may facilitate lifting the staple line over the anvil as it is removed. The security of the anastomosis is confirmed by several means. First, the resection rings around the cartridge shaft should be inspected to ensure there are two complete donuts of colonic tissue. The orientation of the donuts should be maintained so that the location of any defect can be related to the anastomosis. Second, the ‘water test’ or ‘bubble test’ is per-

146

formed by filling the pelvis with sterile water or saline and manually obstructing the proximal colon prior to insufflating the rectum with 200–300 ml of air through a rigid sigmoidoscope or Asepto syringe (Figure 5.24). The presence of air bubbles indicates an anastomotic leak, which should be repaired by oversewing the defect with interrupted stitches of 3-0 delayed absorbable or silk suture. If the leak is inaccessible for repair, the anastomosis must be taken down and repeated. Possibly a diverting colostomy or ileostomy may be required if the condition of the anastomosis is questionable. Finally, some surgeons also recommend direct inspection of the anastomotic staple line for defects using a rigid sigmoidoscope. An alternative technique for end-to-end anastomosis using the CEEA stapler is to place the anvil in the rectal stump with a purse-string closure and direct the handpiece of the main CEEA instrument through a longitudinal colotomy in the proximal colonic segment (Figure 5.25). This procedure is particularly useful if transanal access to the rectum is limited (e.g. supine position, anal stenosis). The anastomosis is performed using standard technique, and the proximal colotomy is closed using a TA (4.8 mm) stapler placed perpendicular to the long axis of the bowel, converting the longitudinal incision into a transverse closure, and the security of the anastomosis confirmed. A stapled end-to-end anastomosis can also be created using the triangulation technique, although this method is only applicable following a high anterior resection of the rectosigmoid colon or subtotal sigmoid colectomy (Figure 5.26). Stay sutures are placed at the anti-mesenteric border and approximately twothirds of the way toward the mesenteric border, dividing the circumference of the bowel segments into three approximately equal lengths. The TA linear stapler (4.8 mm) is applied three times, once along each anastomotic margin, with the staple lines overlapping to ensure complete closure. Note that the posterior row of staples is inverted, while the two anterior staple line limbs are everted. Although an everting anastomosis is not advocated when performing a

Ch 05 final.qxd

12/18/2006

15:15

Page 147

CYTOREDUCTIVE SURGERY: PELVIS

Figure 5.24 The ‘water test’ to confirm the security of

Figure 5.25 Alternative technique of stapled circular end-to-

colorectal anastomosis; with the proximal colon occluded, the

end anastomosis (CEEA) with the main CEEA placed through

pelvis is filled with saline and air insufflated into the rectum

a longitudinal proximal colotomy and the anvil placed in the

transanally

rectal stump

a

b

c

Figure 5.26 Triangulation technique of end-to-end stapled colorectal anastomosis requiring three applications of a TA stapler (30 mm or 60 mm) placed at 60° angles to one another

147

Ch 05 final.qxd

12/18/2006

15:15

Page 148

SURGERY FOR OVARIAN CANCER

conventional hand-sewn anastomosis, it has not been associated with an increased complication rate using the stapled technique.48 End-to-end anastomosis: hand-sewn techniques

The hand-sewn technique of end-to-end anastomosis can be performed using a single- or double-layer closure. The standard two-layer closure consists of an outer layer of interrupted permanent silk suture and an inner layer of running delayed absorbable suture. Following adequate mobilization, the bowel ends are

Posterior serosal sutures

approximated using bowel clamps and traction sutures placed at each angle of both distal and proximal bowel segments. A posterior serosal layer of 3-0 silk interrupted vertical mattress sutures is placed with the knots located on the outside of the bowel wall (Figure 5.27a). All posterior mucosal sutures should be placed first before they are tied. These sutures are cut except for those at the angles, which are retained for traction. The field is walled off with gauze and an enterostomy clamp placed proximally to avoid fecal contamination. The bowel clamps are

Scudder clamp A

Angle suture Crushed margin c

a

Posterior mucosal suture

B

b

d

f

e

A B

Anterior mucosal suture

g

Figure 5.27 Hand-sewn double-layer end-to-end colorectal anastomosis. (a) Posterior serosal layer of interrupted 3-0 silk sutures; (b) excision of crushed margin (or staple line) of bowel wall; (c) inner posterior mucosal layer closure with running delayed absorbable 3-0 suture; (d) Connell stitch to navigate the angles; (e) and (f) inverting Connell stitch to close the anterior mucosal layer; (g) anastomosis completed with anterior seromuscular layer closure of interrupted 3-0 silk suture using an inverting Lembert stitch

148

Ch 05 final.qxd

12/18/2006

15:15

Page 149

CYTOREDUCTIVE SURGERY: PELVIS

removed and the crushed margins of bowel excised (Figure 5.27b). The inner posterior mucosal layers are approximated by placing two 3-0 delayed absorbable sutures (or a double-armed single suture) in the midline of the posterior wall, with each suture line proceeding laterally in a running, locking stitch until the angles are reached (Figure 5.27c). An inverting submucosal Connell stitch is used to navigate each corner and continued to the midline to close the anterior inner mucosal layer (Figure 5.27d–f). Finally, a row of interrupted 3-0 silk sutures is placed using an inverting Lembert stitch along in the anterior wall of the anastomosis (Figure 5.27g). The anastomosis is inspected for patency, and the absence of tension on the suture line confirmed. The single-layer closure technique of end-to-end anastomosis, using permanent monofilament suture, is safe and effective, and has the advantages of producing a wider lumen (as less tissue is turned in), distributing tension on the suture line more evenly and being faster than the two-layer technique.49–51 With the single-layer technique, the lateral traction sutures are placed using an inside-out outside-in technique to

facilitate inversion of the posterior bowel edges. A double-armed suture of 4-0 polypropylene is tied at the angle away from the surgeon and the posterior layer closed from the mucosal side moving from one stay suture to the other. Each bite should be placed 2–3 mm apart and incorporate the edge of the mucosa as it passed tangentially through the submucosal and seromuscular layers. The second arm of the suture is used to close the anterior layer from the serosal side, again catching the edge of the mucosa and passing tangentially through the submucosal and seromuscular layers (Figure 5.28). When the anterior layer has advanced to the stay suture, the sutures are then tied securely but not too tightly, to avoid a purse-string effect. End-to-side anastomosis: stapled techniques

An end-to-side (ESA) low rectal anastomosis may be advisable when there is a significant discrepancy in lumenal diameter between the proximal resected colon and the rectal stump, when an excess of mesenteric fat is present that would encroach unduly on the lumen of an end-to-end anastomosis, or when the

Figure 5.28 Hand-sewn single-layer end-to-end colorectal anastomosis: continuous closure with 4-0 polypropylene suture passing through the seromuscular layer 3 mm from the bowel margin and tangentially through the mucosa

149

Ch 05 final.qxd

12/18/2006

15:15

Page 150

SURGERY FOR OVARIAN CANCER

proximal colon cannot be straightened sufficiently without sacrificing intestinal length to ensure a tension-free anastomosis (i.e. the ESA conforms better to the natural curvature of the proximal sigmoid colon). Another proposed advantage of the ESA is the assurance of a larger and more secure anastomosis with a better blood supply than may be possible by the EEA method. In the double stapling technique of ESA colorectal anastomosis, the rectum has been closed with a linear stapler, the handpiece of the main CEEA instrument is passed transanally, and the trocar advanced through or adjacent to the rectal staple line, as previously described. The proximal colon segment is prepared by partially excising the linear staple line (GIA) and introducing the CEEA anvil into the lumen. The low-profile anvil is usually easier to manipulate into the correct position. Using electro-

cautery, a defect is created in the antimesenteric tenia 3–4 cm proximal to the staple line and the anvil cartridge shaft brought through and secured with a 2-0 monofilament purse-string suture. To avoid creating a blind pouch of bowel distal to the anastomotic site, the distal end of the proximal colon segment, including the initial linear staple line, is resected using the TA (4.8 mm) stapler so as to leave no more than 2–3 cm of bowel beyond the planned anastomosis. The anvil shaft is then engaged with the receptacle of the main CEEA instrument and the two segments brought carefully together, ensuring close approximation of tissue without redundant bowel wall (Figure 5.29a). The instrument is fired, released and carefully removed transanally. In the event that transanal access is limited by anatomy or patient positioning, a stapled ESA can be

Sigmoid

Anvil

Trocar

Purse-string suture

Trocar Staple line Posterior rectum

a

Anvil schaft Anterior rectum

b

Figure 5.29 Stapled end-to-side colorectal anastomosis. (a) The main circular end-to-end anastomosis (CEEA) instrument is introduced through the rectal stump and the trocar advanced; the anvil is passed through the end of the proximal colon segment and advanced through the antimesenteric border 3–4 cm distal to the line of resection; (b) alternative approach to stapled end-to-side anastomosis with placement of the anvil in the rectal stump with a purse-string suture and the main CEEA instrument through the cut end of the proximal colon segment

150

Ch 05 final.qxd

12/18/2006

15:15

Page 151

CYTOREDUCTIVE SURGERY: PELVIS

performed by reversing the placement of the anvil and the main CEEA instrument. In this case, a pursestring stitch of 2-0 monofilament suture (polypropylene) is inserted in the rectal stump, the CEEA anvil is placed into the lumen of the rectum, and the pursestring suture secured. The proximal colon is prepared by partially resecting the linear staple line and introducing the handpiece of the main CEEA instrument into the lumen of the distal end, bringing the disposable trocar out through the antimesenteric taenia 3–4 cm proximal to the staple line (Figure 5.29b). The trocar is removed, the anvil shaft mated with the main CEEA instrument and the anastomosis completed. The distal end of the proximal colonic segment is then closed with a TA (4.8 mm) stapler 2–3 cm distal to the anastomosis and any redundant colonic tissue excised. The J-pouch is a variation of the ESA that uses an extended length (80 mm) GIA to create a colonic reservoir and the CEEA double-stapling technique to complete the anastomosis. The primary indications are to provide a reservoir for stool when most of the rectum has been resected (very low anastomosis) and the anastomosis is performed to the proximal sigmoid colon. This technique is unnecessary if the entire sigmoid colon has been removed, as the descending colon has sufficient lumenal capacity for storage function. The descending colon is mobilized so that the sigmoid colon can be folded back on itself, opposing the antimesenteric taenia to create a J-shape within the pelvis. Stay sutures are placed to stabilize the bowel in this position and a 1-cm colotomy created in the antimesenteric border of the apex of the ‘J’ using the ESU. The colonic reservoir is constructed by placing one prong of the extended GIA (4.8 mm) stapler in each bowel loop, opposing the staple line along the antimesenteric border of the two limbs of bowel, engaging and firing the stapler (Figure 5.30). A pursestring suture is placed around the colotomy at the apex of the J-pouch and used to secure the CEEA anvil within the lumen of the pouch. The handpiece of the main CEEA instrument is passed transanally and approximated against the rectal stump staple line;

a

b

Figure 5.30 Colon reservoir (J-pouch): (a) the colonic reservoir is constructed by placing one prong of the GIA stapler in each loop of the sigmoid colon folded back on itself opposing the antimesenteric borders, (b) the circular end-to-end anastomosis (CEEA) anvil is placed in the colostomy, the main CEEA instrument passed into the rectal stump and the CEEA anastomosis completed using standard technique

the instrument is engaged and fired in the usual manner. In patients with rectal cancer undergoing very low colorectal anastomosis, the J-pouch has been associated with a reduction in the number of daily stools and incidence of fecal incontinence, but the completeness of evacuation may be unpredictable.52–54 End-to-side anastomosis: hand-sewn technique

A hand-sewn ESA requires application of a noncrushing bowel clamp across the rectal margin of resection. The proximal sigmoid colon is mobilized into the pelvis and the taenia along the inferior surface (most dependent) secured at either end of the planned anastomosis with traction sutures, leaving no more than 2 cm of colon distally. A posterior serosal

151

Ch 05 final.qxd

12/18/2006

15:15

Page 152

SURGERY FOR OVARIAN CANCER

Symphysis

A

A

A

Traction suture C B

B

B

a

b

Sacral promontory A

A A

c

A

d

Figure 5.31 Hand-sewn end-to-side anastomosis. (a) The proximal colon is aligned to the rectal stump with stay sutures at the angles to define the planned anastomosis and a posterior serosal layer of interrupted 2-0 or 3-0 silk sutures placed (dashed line); (b) the bowel clamp is removed and a full-thickness closure of the posterior wall completed with a full-thickness, running, locking stitch of 3-0 delayed absorbable suture; (c) an inverting Connell stitch is used to navigate each angle; (d) the outer anterior wall is closed with interrupted 2-0 or 3-0 silk sutures in an inverting submucosal mattress stitch

row of interrupted 2-0 or 3-0 silk sutures is placed 2–3 mm apart. The non-crushing bowel clamp is removed from the rectal stump and an incision created for the anastomosis between the traction sutures in the proximal colon segment along the taenia. Two sutures of 2-0 or 3-0 delayed absorbable suture are tied in the midline of the posterior wall, with each suture proceeding laterally in a full-thickness, running, locking stitch (Figure 5.31). Each angle is navigated with an inverting Connell stitch, and the anterior inner layer closure completed using an inverting submucosal Connell stitch. Finally, a row of interrupt-

152

ed 2-0 or 3-0 silk sutures are placed along in the outer anterior wall of the anastomosis in an inverting submucosal mattress stitch. Side-to-side functional end-to-end anastomosis

The side-to-side functional end-to-end anastomosis (FEEA) has the advantage of providing a larger lumen than the CEEA, reducing the chance for stenosis, and is suited to proximal (high) rectal anastomosis, but is not generally applicable to the low or very low rectal anastomosis. This anastomosis may also be indicated on anatomic grounds if it conforms better to the

Ch 05 final.qxd

12/18/2006

15:15

Page 153

CYTOREDUCTIVE SURGERY: PELVIS

natural curve of the sigmoid colon and may have a somewhat better blood supply than the CEEA. In the standard closed technique of stapled FEEA, the rectum and proximal sigmoid colon have been closed with a linear stapler during resection of the rectosigmoid colon. The antimesenteric borders of the two limbs are aligned side-by-side and secured with proximal and distal stay sutures of 3-0 silk placed 6–8 cm apart. The inside corner of the linear staple lines is excised and one prong of the GIA stapler (4.8 mm) inserted into each lumen of bowel. The stapler fork is aligned along the antimesenteric borders of the planned anastomosis (Figure 5.32a). The two halves of the GIA stapler are engaged, the lock lever is closed, and the instrument is fired, thereby creating a colocolostomy. The staple line should be carefully inspected and any bleeding sites controlled with 3-0 delayed absorbable sutures or electrocautery. To prevent intralumenal adhesion formation, the staple lines of the colocolostomy are offset slightly and the common defect closed with the TA stapler (4.8 mm) (Figure 5.32b).

Type III modification Although uncommon, resection of a portion of the lower urinary tract may be required as part of the primary cytoreductive surgical effort. In the original report of radical oophorectomy by Hudson and Chir, resection of a portion of the bladder or ureter, the most commonly performed procedures, was required in 20% and 12% of cases, respectively.23 In more contemporary series, the frequency of urinary tract resection ranges from 8 to 33%.14,55 Although preoperative obstructive uropathy predicts a diminished overall prognosis among patients with ovarian cancer, optimal cytoreduction is still associated with a survival benefit compared to patients with urinary obstruction and left with bulky residual disease.56 Ultimately, the need for lower urinary tract resection is determined by the contribution that such a procedure will make to the overall volume of residual disease. Partial cystectomy

Partial cystectomy is indicated for extensive involvement of the vesicouterine peritoneal reflection with

a

b

Figure 5.32 Side-to-side functional end-to-end stapled anastomosis. (a) The antimesenteric borders of the proximal sigmoid colon and rectum are aligned and one prong of the GIA stapler inserted into each lumen of bowel; (b) the colocolostomy thus created is closed by a single application of the TA stapler across the common defect

153

Ch 05 final.qxd

12/18/2006

15:15

Page 154

SURGERY FOR OVARIAN CANCER

tumor densely adherent to or infiltrating the muscularis of the bladder dome. The vesicouterine peritoneum and associated tumor should be mobilized from the underlying bladder wall by dissecting in the subperitoneal plane, proceeding in a centripetal fashion, to the point at which the tumor no longer respects peritoneal planes of demarcation. The bladder dome is opened in the sagittal direction with the ESU and the incision extended up to the point of dense tumor attachment. The ureteral orifices should be located, and it may be advisable to place ureteral stents under direct vision to facilitate identification of the ureters during resection and reconstruction. Allis clamps are placed on the cut edges of the bladder wall for traction, and the cystotomy incision is extended to either side of the involved portion of the bladder dome, using the cutting current of the ESU, circumscribing the involved area of the bladder dome with a 1 cm margin (Figure 5.33). Resection of up to 25–35% of the bladder wall can be accomplished without a significant impact on voiding function or bladder capacity. The resected portion of bladder is left attached to the vesicouterine peritoneum and the central tumor specimen retracted upward while the bladder base is mobilized downward, sharply dissecting the vesicovaginal space and exposing the anterior vaginal wall for colpotomy. Resection of the central tumor mass should be completed prior to repairing the bladder defect. The bladder is reconstructed in the sagittal plane using a one-layer or two-layer imbricating closure of 2-0 or 3-0 delayed absorbable suture in a running, non-locking stitch. A transurethral Foley catheter is left in place and removed 7–10 days postoperatively. Ureteral stents can be removed immediately if the scope of resection is far removed from the bladder trigone or extracted cystoscopically in 4–6 weeks if the ureteral orifices are in close proximity to the bladder dome repair. Partial ureterectomy

Obstruction of the pelvic ureter due to ovarian cancer is most commonly due to compression from the central tumor mass against the pelvic sidewall. In the

154

Figure 5.33 Radical oophorectomy: type III modification with partial cystectomy: the bladder is opened in the sagittal plane and the incision extended up to the point of direct tumor invasion,

the

vesicouterine

reflection

tumor

mass

is

circumscribed using the electrosurgical unit with a 1-cm margin

event that tumor directly invades the ureter, or the ureter cannot be separated from the overlying peritoneum without compromising the completeness of resection, partial ureterectomy is indicated. The surgical approach begins by mobilizing the ureter from its attachments to the pelvic peritoneum down to the point of tumor involvement; the ureter is held with a vasa-loop for traction. If not done previously, the uterine artery pedicle should be developed, doubly suture ligated and divided to allow mobilization of the ureter distal to the point of obstruction. The involved segment of pelvic ureter is suture ligated and divided proximal and distal to the point of obstruction and left attached to the central pelvic tumor specimen. The radical oophorectomy procedure is then completed as described previously. The technique of reconstructing the pelvic ureter is determined by the available length of ureter and

Ch 05 final.qxd

12/18/2006

15:15

Page 155

CYTOREDUCTIVE SURGERY: PELVIS

local tissue conditions. Ureteroureterostomy is applicable when a very limited portion of ureter is resected and both proximal and distal ends are well vascularized and free of disease; however, this technique is associated with a higher risk of stricture than a reimplantation procedure, and therefore should be undertaken with caution. Adequate mobilization of the ureteral segments to be re-attached is essential, with care taken to preserve the longitudinal blood supply running in the adventitial sheath. The anastomosis is prepared by excising the proximal and distal margins obliquely at opposing 45° angles (to provide a 1.5-cm opening) or by spatulating each lumen for a distance of 1 cm, thereby enlarging the anastomotic circumference and minimizing the risk of stricture. Repair is performed over a ureteral stent (6-Fr or 7-Fr double-J) using five or six interrupted seromuscular stitches of 3-0 or 4-0 delayed absorbable suture (Figure 5.34). The ureteral stent is removed cystoscopically 8–12 weeks postoperatively. The re-implantation technique of ureteroneocystotomy (UNC) with psoas hitch is recommended by most authorities as the procedure of choice following resection of the pelvic ureter and can accommodate replacement of the ureter up to the pelvic brim. The proximal ureter is widely mobilized above the pelvic brim and any devitalized tissue excised. In the pelvis, the retropubic space of Retzius is fully developed and the false lateral vesicle ligaments (obliterated umbilical arteries) divided using electrocautery, creating continuity with the para-vesicle spaces. Water is instilled through the transurethral Foley catheter to distend the bladder and a curvilinear cystotomy created in the bladder dome, with the base directed toward the planned site of fixation on the ipsilateral psoas muscle. The bladder is transposed to the ipsilateral psoas muscle (over the external iliac vessels) and fixed in place using several interrupted 0 delayed absorbable sutures (psoas hitch) (Figure 5.35a). Provided there is sufficient length, a 2-cm submucosal tunnel can be created from the site of implantation near the psoas hitch running toward the bladder base by injecting a dilute solution of epinephrine (adrena-

Figure 5.34 Ureteroureterostomy. The proximal and distal ureteral margins are prepared by excising the ends at opposing oblique angles. The repair is completed over a ureteral stent with interrupted delayed absorbable sutures

line) (1 : 200 000) beneath the mucosa, raising a plane of dissection. The bladder mucosa at the distal end of the tunnel is incised transversely and the submucosal plane developed through to the back wall of the bladder. A 4-0 delayed absorbable suture is placed through the distal ureter and used to guide the ureter through the back wall of the bladder, along the course of the submucosal tunnel, and through the mucosal incision without twisting or kinking (Figure 5.35b). A ureteral stent (6-Fr or 7-Fr, double-J) is placed under direct vision. The ureter is spatulated (1 cm) and a mucosato-mucosa anastomosis completed using 4-0 delayed absorbable sutures placed 2–3 mm apart (Figure 5.35c). If additional length is required, the submucosal tunnel can be omitted. The remaining defect in the bladder dome is closed along the long axis, running from the site of psoas fixation toward the bladder base, using a single full-thickness closure or doublelayer imbricating closure of continuous 2-0 or 3-0

155

Ch 05 final.qxd

12/18/2006

15:15

Page 156

SURGERY FOR OVARIAN CANCER

a

b

c

Figure 5.35 Ureteroneocystotomy. (a) The bladder is distended with saline and a curvilinear cystotomy created with the electrosurgical unit (inset). The bladder is transposed and fixed to the ipsilateral psoas muscle; (b) the ureter is brought through the back wall of the bladder using a tunneled or direct implantation technique; (c) the distal ureter is spatulated and a mucosa-to-mucosa anastomosis completed with interrupted sutures

delayed absorbable suture. The ureteral stent is removed cytoscopically 8–12 weeks postoperatively. Normal voiding function should be expected. When the gap between proximal ureter and bladder is too large to be bridged by the psoas hitch technique alone, the Boari bladder flap is the preferred surgical option. The bladder is distended with water and the flap raised by creating a U-shaped incision in the bladder dome that is approximately 4 cm across at the base (oriented toward the ipsilateral psoas muscle) and 3 cm wide at the tip (Figure 5.36). This flap may be up to 8 cm in length, depending on the amount of available bladder. The flap is rotated into position and secured to the psoas muscle using interrupted 0 delayed absorbable sutures. The submucosal tunnel is created as described above, starting at the cut edge of the flap. Alternatively, the tunnel can be omitted and the ureter implanted directly into the flap 1–2 cm distal to the cut edge of the flap and a direct mucosa-to-mucosa anastomosis performed. The flap is closed in the form of a tube using a single

156

or double layer of delayed absorbable suture, incorporating the cystotomy incision in the repair. A larger gap between the proximal ureter and bladder can be accommodated by uretero–ileo–neocystotomy (UINC) (or ileal interposition), which utilizes a segment of ileum that is uninvolved by tumor, as a ureteral conduit.57 A satisfactory length of ileum that will comfortably span the distance is divided proximally and distally using the GIA (3.8 mm) stapling device. An independent mesenteric blood supply must be carefully maintained; however, if additional mobility is required, the proximal and distal mesenteric margins can be ‘back-cut’ to the level of the dominant arterial arcade for the ileal segment (usually 3–5 cm). Intestinal continuity is reestablished via stapled side-to-side functional end-toend anastomosis (described in Chapter 7) anterior to the isolated ileal segment, and the mesenteric defect closed with interrupted sutures. The ileal segment is irrigated and placed in an isoperistaltic direction to the urinary stream. The distal end of the ureter is cut

Ch 05 final.qxd

12/18/2006

15:15

Page 157

CYTOREDUCTIVE SURGERY: PELVIS

Figure 5.36 Ureteroneocystotomy with Boari bladder flap: the bladder is distended with saline and the flap raised with the electrosurgical unit (inset); the flap is secured to the ipsilateral psoas muscle, the ureter is brought through the back wall of the flap using either a tunneled or direct implantation technique and a mucosa-to-mucosa anastomosis completed with interrupted sutures; the bladder defect is then closed in the form of a tube

obliquely at a 45° angle or spatulated for a distance of 1cm and an end-to-side mucosa-to-mucosa ureteroileal anastomosis created over a 6-Fr or 7-Fr double-J ureteral stent using five or six interrupted stitches of 3-0 delayed absorbable suture. The bladder is prepared by excising a small (1 cm) full-thickness disc of posterolateral bladder dome. The distal end of the ileal segment is opened, the ureteral stent guided into

the bladder, and the ileoneocystotomy completed by placing five or six mucosa-to-mucosa full-thickness stitches of 3-0 delayed absorbable suture over the ureteral stent. A closed suction drain is placed, and the bladder is left to straight drainage for 7–10 days. Intravenous pyelography is obtained at 6 weeks postoperatively to confirm the integrity of the anastomosis, and the ureteric stent is removed cystoscopically at that time. Transureteroureterostomy (TUU) is an alternative to UINC if there is extensive but small volume (miliary) tumor involvement of the ileum that would preclude its use as a urinary conduit. The primary disadvantage of TUU is that it places both urinary collecting systems at risk in the event of chronic ascending infection. For TUU, the abdominal retroperitoneum is exposed and the donor ureter mobilized proximally for a distance of 9–12 cm. Depending on the available length, the donor ureter may be brought over (preferred) or under the inferior mesenteric artery, but sharp angulation must be avoided. The distal end of the donor ureter is incised obliquely at a 45° angle, and the recipient ureter opened along its medial edge for a distance of 1.5 cm. A 6-Fr or 7-Fr double-J ureteral stent is placed into the recipient ureter distally and the donor ureter proximally; the end-to-side mucosa-to-mucosa anastomosis is completed using six or eight 3-0 or 4-0 delayed absorbable sutures. The integrity of the anastomosis is confirmed by intravenous pyelography 6 weeks postoperatively followed by cystoscopic removal of the ureteral stent. In rare instances, immediate reconstruction of the urinary tract may be impractical for technical reasons or temporal limitations brought about by an unstable medical condition of the patient intraoperatively. If necessary, the proximal end of the resected ureter can be ligated as a temporizing measure and the procedure terminated. After 8–12 h, the ureter and collecting system proximal to the ligature will become sufficiently dilated to allow for placement of a percutaneous nephrostomy tube with ultrasound guidance. Definitive reconstruction of the urinary tract can be

157

Ch 05 final.qxd

12/18/2006

15:15

Page 158

SURGERY FOR OVARIAN CANCER

completed at a later time when circumstances are more favorable. Radical oophorectomy after prior hysterectomy

The cervix and uterus can serve as physical anatomic barriers to tumor involvement of the anterior cul-desac and preserve the central tissue planes of dissection (vesicovaginal space, rectovaginal space). In the absence of the cervix, these anatomic relationships may be distorted, with displacement of the bladder posteriorly or rectosigmoid colon anteriorly overlying the vaginal cuff. In such cases, the anterior pelvis should be deperitonealized as previously described, moving in a centripetal fashion from the retropubic space of Retzius and paravesicle spaces toward the central pelvis. Placement of a rectal sizing instrument or spongestick into the vagina, with elevation of the vaginal apex, will aid identification of the proper plane of dissection into the vesicovaginal space. Instilling 200 ml of saline into the bladder through the transurethral Foley catheter may assist in delineating the dissection plane between the bladder base and anterior vaginal wall so that the bladder can be mobilized to expose the proximal vagina. Incorporating a proximal vaginectomy (2–3 cm) within the scope of resection allows controlled entry into the rectovaginal space and completion of the posterior pelvic dissection, usually requiring resection of the rectosigmoid colon (type II modification) for complete clearance of extensive cul-de-sac tumor. Morbidity of radical oophorectomy

En bloc resection of locally advanced ovarian cancer is associated with a predictably high but acceptable risk of perioperative morbidity. The operative mortality following radical oophorectomy ranges from 0 to 8%, with contemporary series reporting rates of 4% or lower.14,24,25,27,28,32,33,38,58 The median estimated blood loss during radical oophorectomy ranges from 800 to 2900 ml, with 52.3–87.6% of patients requiring blood product transfusion.14,24–28,30,33,55,58 The overall incidence of significant morbidity ranges from 12 to

158

49%.14,27,29,30,32,33,37 In larger series, the most commonly observed complications are: prolonged ileus (7.8–40.4%), wound infection (6.1–21.5%), non-specific febrile morbidity (19.4–28%), pulmonary embolism (1.5–16.7%), and pneumonia (3.2–18.3%).24,28–30,33,37,55,58 Dehiscence of a colorectal anastomosis accompanied by leakage of bowel contents can have devastating consequences including abscess, fistula formation, sepsis and emergency return to the operating room for re-exploration. In general, the lower the anastomosis, the higher the rate of anastomotic leak.59,60 For example, Vignali et al. reported that, among patients undergoing stapled anastomosis for rectal cancer, the rate of anastomotic leak was 7.7% after very low stapling and just 1% after stapled anastomosis greater than 7 cm from the anal verge.59 Historically, there have been concerns about re-establishing intestinal continuity in patients undergoing a radical debulking operation for ovarian cancer requiring resection of the rectosigmoid colon. Such patients often have large-volume ascites, may be nutritionally compromised from the metabolic effects of an extensive tumor burden, or have evidence of early bowel obstruction. Temporary or permanent intestinal diversion was typically performed in 12–59% of such cases.26–28,30,33,38,58 More contemporary studies, however, have reported anastomotic breakdown in 0–8% of cases undergoing sigmoid colectomy for ovarian cancer.24,25,27–30,34,58 Furthermore, among patients with large-volume ascites (≥ 500 ml), the incidence of anastomotic dehiscence is just 2.1–3.1%.28,33 These data suggest that protective intestinal diversion is unnecessary and that terminal excretory function can be preserved in the majority of patients undergoing radical oophorectomy. Rare instances in which a protective colostomy may be indicated include prior pelvic radiation, an unsatisfactory bowel preparation, or a technically suboptimal anastomosis. Some authors have recommended a protective colostomy for anastomoses situated less than 7 cm from the anal verge, particularly in

Ch 05 final.qxd

12/18/2006

15:15

Page 159

CYTOREDUCTIVE SURGERY: PELVIS

obese patients.60,61 It is important to appreciate, however, that a protective colostomy is not associated with a reduction in the rate of anastomotic dehiscence but may minimize the clinical sequelae if a leak does occur. Many surgeons recommend placing a closed suction drainage system within the pelvis and presacral space for evacuation of serous fluid produced by the denuded pelvic surface, ostensibly to reduce the risk of abscess formation and anastomotic leak. Although somewhat controversial, the preponderance of contemporary well-designed (randomized, prospective) studies of bowel resection and anastomosis for colon or rectal cancer indicate that postoperative drainage is not associated with a significant reduction in the incidence of anastomotic leak, wound infection, or major complication rate.62–64 There are few data specifically addressing this issue among ovarian cancer patients, who can be expected to have more advanced disease and are subject to larger postoperative fluid shifts than patients undergoing surgical resection for localized colon cancer. Re-accumulation of large-volume ascites in the immediate postoperative period can result in an abdominal compartment syndrome, which can be pre-empted by intraoperative placement of a closed suction drainage system to serve as an ‘outlet’ valve for decompression. Postoperative pelvic drainage is also recommended in all cases involving resection of the urinary tract to facilitate early identification of a deficient surgical repair and avoid urinoma formation. Clinical outcomes following radical oophorectomy

The clinical utility of radical oophorectomy is predicated on its demonstrated efficacy for clearance of pan-pelvic disease and the extent to which this contributes to the overall volume of residual disease. It is worth noting that the observed rates of optimal cytoreduction in published series of radical oophorectomy may not be reflective of the surgical outcome for all patients with advanced ovarian cancer, as only

those patients in whom a reasonable attempt at resection of upper abdominal disease could be undertaken are usually submitted to this procedure. While this necessarily introduces a component of selection bias, overall optimal residual disease (≤ 1 cm or 2 cm) is typically achievable in 74–100% of patients undergoing radical oophorectomy, with almost universal clearance of advanced pelvic tumor.14,24,28–34,37,38,55 Several studies have documented favorably low rates of pelvic failure following radical oophorectomy, given the locally advanced extent of disease necessitating an en bloc resection. Eisenkop et al. reported persistent pelvic disease at second-look surgery in 15% of cases.33 Similarly, Spirtos et al. described their series of 151 patients undergoing second-look laparotomy after radical oophorectomy and found persistent or progressive pelvic tumor in just 5.3% of cases.65 Among studies with extended follow-up, the rate of pelvic failure at first relapse ranges from 4.5 to 18%.27,34,37,38 Although Eisenkop et al. reported that the presence of a ‘frozen pelvis’ was an independent predictor of poor overall survival among patients with advanced-stage ovarian cancer, radical oophorectomy, combined with contemporary platinum-based combination chemotherapy, has been associated with median survival times ranging from 30.6 to 39.5 months and 5-year overall survival rates as high as 42%.10,24,25,34,37

PELVIC PERITONEAL IMPLANTS Metastases to the pelvic parietal and visceral peritoneum are almost ubiquitous among patients with stage II or greater ovarian cancer; however, not all cases require, or will be encompassed by, an en bloc resection. There are several techniques to approach individual implants in a macronodular (≥ 5 mm) distribution, larger areas of pelvic peritoneum involved by multiple micronodular (< 5 mm) implants, or a coalescent growth of implants that have progressed into tumor plaques.

159

Ch 05 final.qxd

12/18/2006

15:16

Page 160

SURGERY FOR OVARIAN CANCER

Surgical techniques Local excision and peritonectomy

a

b

c

Figure 5.37 Peritoneal tumor implant excision. (a)–(c) The electrosurgical unit is used to raise an edge of peritoneum, and dissection proceeds in the subperitoneal plane, circumscribing the entire implant or group of implants

160

Local excision by sharp dissection is most applicable to macronodular disease on pelvic sidewalls, the vesicle peritoneum, the anterior wall of the cul-de-sac of Douglas, and even the anterior surface of the rectosigmoid colon. Excision can be performed with fine dissecting scissors or the ESU. The ESU has the advantage of cauterizing small bleeding points during dissection and doubling as a mechanical dissector without current application to develop the subperitoneal plane. The ESU is used to raise an edge of peritoneum 0.5–1.0 cm to the side of an individual implant, which is then reflected and placed on traction. The involved peritoneum is completely circumscribed using the ESU and separated from the subperitoneal tissue using sharp and blunt dissection (Figure 5.37). For disease located deep within the culde-sac of Douglas or on the anterior surface of the rectosigmoid colon, exposure is facilitated by placing a rectal sizing instrument into the rectum or distal sigmoid colon to elevate the target lesion into the operative field. Resection of serosal tumor implants from the rectosigmoid colon is more safely accomplished with dissecting scissors to prevent underlying thermal damage to the bowel wall. The dissection is carried in a plane superficial to the colonic muscularis beneath the visceral peritoneum (serosa). If the tumor penetrates more deeply into the bowel wall, a subtotal sigmoid colectomy or a wedge resection of the anterior colonic wall with primary closure (type I or type II procedures) is more appropriate. Peritonectomy, or peritoneal ‘stripping’, can be employed for macronodular, micronodular, or confluent disease involving the parietal or visceral pelvic peritoneum. The same surgical principles for implant excision are utilized during peritonectomy, except that a broad front of dissection is raised in the subperitoneal plane that encompasses the involved area of peritoneum (Figure 5.38). This technique is applicable to most pelvic surfaces, including the peritoneum of the sigmoid colon mesentery. In resecting sidewall

Ch 05 final.qxd

12/18/2006

15:16

Page 161

CYTOREDUCTIVE SURGERY: PELVIS

peritoneum, the surgeon must exercise caution with respect to wide mobilization of the peritoneum from the underlying ureters and vasculature (iliac vessels) in order to avoid injuring these structures. Occasionally, multiple tumor implants, or ‘drop metastases’ will replace the posterior cul-de-sac peritoneum and remain superficial to the bowel muscularis, such that formal resection of the rectosigmoid colon is not required. Cul-de-sac peritonectomy is approached by circumscribing the peritoneum surrounding the lesion with the ESU. The subperitoneal

Figure 5.38 Pelvic peritonectomy: the electrosurgical unit is used to raise a broad front of peritoneum in the right anterior pelvis, which is reflected off the underlying areolar tissue as dissection proceeds

a

b

plane of dissection is then developed in a centripetal fashion, working medially from the deep pararectal spaces and downward from the proximal rectovaginal space. In this manner, the cul-de-sac tumor mass or plaque can be safely excised from the posterior vaginal wall, anterior surface of the rectum and deep pelvic sidewalls. A rectal sizing instrument placed transanally can be used to elevate the cul-de-sac into the operative field and improve exposure. Peritoneal implant ablation and aspiration

Peritoneal tumor implant ablation using the argon beam coagulator (ABC) is most applicable to micronodular or small-volume disease on parietal pelvic peritoneal surfaces where peritonectomy or local excision are impractical. Electrosurgical tumor destruction with the ABC is achieved through the conduction of unipolar current via a beam of inert argon gas, which spreads out on the tissue surface in a more homogeneous distribution than standard electrocautery. The current generates individual ‘arc tunnels’ within the tissue, and it is the interconnection of these arc tunnels that produces both tissue coagulation and tumor ablation effects. The ABC can be used in a series of short, ‘burst-like’ applications or a ‘painting’ technique over the involved peritoneum using a power setting of 80–100 W, intermittently removing the resulting carbonized eschar with a moist laparotomy sponge (Figure 5.39). When used in this fashion

c

Figure 5.39 Peritoneal tumor implant ablation using the argon beam coagulator. (a) Pelvic peritoneal implants; (b) unipolar current is conducted via the beam of argon gas, generating individual arc tunnels within the implant; (c) remaining carbonized eschar

161

Ch 05 final.qxd

12/18/2006

15:16

Page 162

SURGERY FOR OVARIAN CANCER

for patients with advanced-stage ovarian cancer, the ABC has been associated with an increased rate of complete pelvic cytoreduction (86%) compared to patients undergoing surgery without the aid of the ABC (50%).66 The ABC can also be used to ablate small-volume disease on the surface of the rectosigmoid colon at a power setting of 60 W; however, caution should be taken not to induce a full-thickness bowel wall injury. This may be somewhat difficult to predict given that an underlying degree of coagulative necrosis accompanies the visibly apparent carbonized eschar.67 It may be advisable to over-sew sites of bowel implant ablation with several imbricating stitches of 3-0 silk or delayed absorbable suture to ensure that the integrity of the bowel muscularis is maintained. The cavitational ultrasonic surgical aspirator (CUSA) has also been used as an effective adjunct to conventional surgical techniques of cytoreductive surgery for advanced ovarian cancer.68,69 The CUSA employs a high-frequency ultrasonic vibrator, which destroys tissue by cavitation, and an irrigation and aspiration system, which cleans the operative field and cools the tip of the instrument. The CUSA induces selective fragmentation of tissues with a high water content (fat, muscle, cancer), while tissues with a high content of collagen and elastic fibers (blood vessel adventitia, nerves, ureter) are damaged with more difficulty. The amplitude setting most commonly used for tumor resection is 0.7–0.8. Several authors have advocated the CUSA for removal of ovarian cancer tumor nodules from the bowel and bladder serosa and pelvic sidewall, and for fixed retroperitoneal adenopathy attached to major blood vessels.68,69 For implants on the surface of the sigmoid colon, the CUSA offers a more predictable degree of tissue destruction compared to the ABC and is probably the safer surgical option. Use of the CUSA for cytoreductive surgery has been associated with an increased risk of intraoperative disseminated intravascular coagulation, although this has not been confirmed in all studies.68,70 Another potential hazard of this instrument is that the mist produced by the

162

CUSA has been reported to contain intact and potentially viable tumor cells.71

PELVIC ADENOPATHY Dissemination of epithelial ovarian cancer through the retroperitoneal lymph nodes represents a significant route of spread, with 34–78.6% of patients with advanced-stage (FIGO stage III–IV) disease having nodal metastases documented by systematic lymphadenectomy.20,72–77 Extensive nodal involvement may, on occasion, represent the predominant site of extraovarian spread. In the pelvis, the external iliac and obturator are the most frequently involved nodal basins.20,77 Despite the high rate of nodal involvement, the role of therapeutic lymphadenectomy in the context of a maximal cytoreductive surgical effort has yet to be precisely defined. Several retrospective studies have suggested that complete pelvic lymphadenectomy may confer a survival advantage for patients with optimally debulked stage III disease.76,78 However, there appears to be little benefit to systematic lymphadenectomy in the setting of suboptimal residual disease.75 In view of the increased operating time and postoperative morbidity, most authorities still consider systematic lymphadenectomy to be an experimental procedure until more definitive evidence of an associated survival benefit becomes available.79 The debate over therapeutic lymphadenectomy notwithstanding, all of the arguments in support of aggressive cytoreduction of intraperitoneal ovarian cancer metastases are also applicable to the removal of grossly involved retroperitoneal lymph nodes. Eisenkop and Spirtos reported that, among 100 patients with advanced-stage ovarian cancer undergoing retroperitoneal dissection, 61% had macroscopically positive lymph nodes.80 Of those patients with macroscopically positive nodes, 80.3% had a maximal nodal diameter of ≥ 11 mm, suggesting that removal of such nodes contributed significantly to the overall residual disease status. Furthermore, in a previous

Ch 05 final.qxd

12/18/2006

15:16

Page 163

CYTOREDUCTIVE SURGERY: PELVIS

series, Spirtos et al. reported that 64.3% of 56 patients with advanced-stage (stage IIIA–IV) ovarian cancer had positive retroperitoneal lymph nodes.74 Survival analysis revealed no statistically significant differences in long-term survival between patients with negative lymph nodes (50%), microscopically positive lymph nodes (46%) and those with macroscopically positive but completely resected nodes (43%). All groups experienced superior long-term survival compared to patients left with residual disease of ≥ 1 cm (10%). These studies suggest that resection of clinically involved nodes appreciated at the time of thorough retroperitoneal exploration should be routinely attempted, contingent on the probability that the procedure will contribute significantly to an optimal residual diseased state, and that it can be performed safely.

Surgical techniques The surgical principles to be maintained during cytoreduction of bulky pelvic adenopathy include achieving adequate exposure, defining the boundaries of nodal tissue to be resected, and protecting vital structures (vessels, ureter) that may be injured during the course of dissection. The specific surgical approach is dictated by the location of disease and the growth pattern in relation to surrounding structures. An intact capsular resection of enlarged nodal disease is the most straightforward and preferred technique by which complete excision of the nodal mass can be achieved. The dissection begins by raising a margin of normal-appearing nodal tissue or grasping the node capsule itself and developing a plane beneath the capsule and superficial to the adventitia of the vessel wall. Maintaining the integrity of the capsule, while mobilizing the nodal mass laterally and medially, will facilitate establishing the proper plane (Figure 5.40). This technique is applicable to disease along the external iliac vessels or mobile adenopathy within the obturator fossa, which will usually displace the nerve laterally or medially. The obturator nerve should always be identified proximally, as it enters the pelvis lateral to the bifurcation of the common iliac vein,

and distally, as it exits through the obturator foramen, prior to extensive sharp dissection within the obturator fossa. When a matted confluence of multiple nodes is fixed to underlying vessels or surrounds the obturator nerve, it should first be mobilized to the maximum extent from the psoas muscles and/or obturator internus muscle and fascia to improve exposure. A conglomerate of enlarged nodes not amenable to intact resection can usually be successfully resected by one of the following alternative approaches. In progressive subtotal enucleation, the fibrous pseudocapsule is incised using electrocautery along the faint surface depressions delineating the dominant node or nodes. Individual nodes are resected in this fashion until the underlying vessel wall or nerve is identified and the proper anatomic surgical plane developed. Alternatively, electrocautery can be used to incise or split the nodal mass longitudinally until the underlying external iliac vessels or obturator nerve are identified. Each half of the bisected nodal mass is then gently rolled (split and roll technique) laterally and medially, establishing the correct surgical plane of dissection in the process (Figure 5.41). As a final option, the CUSA can be used to resect enlarged pelvic lymph nodes that are especially tenacious in their fixation to vessel walls. Retrograde lymphatic dissemination can occur along the external iliac chain or accompanying the round ligament, beneath the inguinal ligament, and into the femoral triangle. Autopsy studies have demonstrated that ovarian cancer can metastasize to the inguinal lymph node basin in approximately 3% of cases.81 Resection of bulky inguinal adenopathy can be approached through the transabdominal incision and obviate the need for a second (or third) incision directly overlying the groin. In this approach, the caudal 8–10 cm of the medial margin of the rectus fascia is grasped with clamps and retracted medially while electrocautery dissection proceeds laterally and caudally toward the inguinal ligament, developing the plane between the rectus fascia and the overlying Scarpa’s fascia and creating a tissue advancement

163

Ch 05 final.qxd

12/18/2006

15:16

Page 164

SURGERY FOR OVARIAN CANCER

a

b

c

d

Figure 5.40 Intact capsular resection. (a) Bulky right external and common iliac adenopathy (ureter is held with vasa-loop); (b) the distal margin of the node capsule is grasped and used for counter traction; (c) the dissection plane is developed between the node capsule and the underlying vessel adventitia; (d) complete resection of the nodal basin

flap. The dissection is carried over the inguinal ligament into the femoral triangle. Enlarged inguinal lymph nodes are then selectively removed using careful sharp dissection (Figure 5.42).

CONTROL OF PELVIC HEMORRHAGE Resection of locally advanced ovarian cancer can be complicated by vascular injury and significant hemorrhage. Diffuse bleeding from the central pelvis or parametria due to small-vessel injury can be effectively temporized by bilateral ligation of the hypogastric arteries. A right-angle clamp is used to gently

164

dissect the artery from the underlying hypogastric vein and one of two suture ligatures of 0 silk placed distal to the posterior division, usually 2.5–3.0 cm distal to the bifurcation of the common iliac artery. Control of more localized vascular injury requires wide mobilization of surrounding structures to permit adequate visualization and control of both proximal and distal blood flow to the site of injury. Small puncture injuries to the major pelvic arteries (common iliac and external iliac) can be controlled initially by digital occlusion or application of vasa-loops, and repaired using several interrupted sutures of fine monofilament suture (5-0 or 6-0 polypropylene) on a vascular needle. Manipulation of the external iliac or

Ch 05 final.qxd

12/18/2006

15:16

Page 165

CYTOREDUCTIVE SURGERY: PELVIS

Psoas muscle

External iliac artery

External iliac vein Obturator nerve

Figure 5.42 Transabdominal

approach

to

left

inguinal

adenopathy: the rectus fascia is placed on medial traction and Figure 5.41 Split and roll technique for resection of bulky

a lower abdominal wall advancement flap raised down to and

adenopathy fixed to the external iliac vessels. The node capsule

beyond the inguinal ligament, the femoral triangle is developed

is incised longitudinally using the electrosurgical unit and the

and the enlarged nodal mass grasped and dissected from

nodal mass fractured down to the underlying vessel wall; the

surrounding tissue. The nodal mass is bounded by the

bisected halves are rolled laterally, facilitating resection in the

superficial circumflex iliac vessels laterally and the external

proper plane superficial to the vessel adventitia

inguinal ring medially

common iliac artery may result in dissection of an atherosclerotic plaque or intimal hematoma leading to lower extremity ischemia. These complications, as well as significant transmural arterial injuries, are best managed by a vascular surgeon. The most commonly encountered venous injuries during pelvic cytoreductive surgery are to the external iliac vein, hypogastric vein, obturator vein and presacral venous plexus. Injury to the ventral or medial surface of the external iliac vein can be temporarily controlled with digital pressure, spongesticks, one or more Allis clamps, or a peripheral vascular clamp (e.g. Satinsky) while adequate suction and exposure are established. The defect is then repaired with 4-0 or 5-0 monofilament suture on a vascular needle in a figure-of-eight stitch or running, non-locking closure. Injury to the inferior surface of the external iliac vein can occur at the entry point of the accessory obturator vein, near the superior pubic ramus, requiring lat-

eral rotation prior to suture repair or application of one or two vascular hemoclips. Injury to the hypogastric vein can be particularly troublesome, given the complex venous drainage of the lower pelvis. Placement of crushing clamps should be avoided. Rather, the precise bleeding site or sites should be identified and repaired with figure-of-eight stitches of fine monofilament suture. The hypogastric vein is particularly prone to injury as it enters the juncture of the external iliac and common iliac veins, such that successful repair is dependent upon achieving distal occlusion of the hypogastric and external iliac veins and proximal occlusion of the common iliac vein. Prior to sutured closure of the defect, the obturator nerve should be identified as it emerges into the pelvis from behind the bifurcation of the common iliac vein. Injury to the obturator vein most commonly occurs at its entrance into the obturator canal, into which the distal end of the vein may retract.

165

Ch 05 final.qxd

12/18/2006

15:16

Page 166

SURGERY FOR OVARIAN CANCER

Bleeding is controlled by placing a figure-of-eight stitch of 4-0 or 5-0 monofilament suture at the opening of the canal, again taking care to avoid the obturator nerve. Hemorrhage from the venous plexus in the presacral space can be effectively temporized by packing.82 Injury to individual veins within the presacral space should be repaired with a figure-of-eight suture technique or application of a sterile thumbtack. Vascular hemoclips and electrocautery are generally ineffective and may make definitive suture repair more difficult. Topical procoagulants may be of some benefit. Floseal® hemostatic matrix (Baxter, Deerfield, IL) combines collagen-derived particles and topical thrombin that conform to irregular bleeding surfaces such as the sacral foramina.

logic cancer: carcinoma of the ovary. J Epidemiol Biostat 1998; 3: 75–102 3.

Hoskins WJ, Bundy BN, Thigpen JT, Omura GA. The influence of cytoreductive surgery on recurrence-free interval and survival in small-volume Stage III epithelial ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1992; 47: 159–66

4.

Piver MS, Lele SB, Marchetti DL, et al. The impact of aggressive debulking surgery and cisplatin-based chemotherapy on progression-free survival in stage III and IV ovarian carcinoma. J Clin Oncol 1988; 6: 983–9

5.

Bertelson K. Tumor reduction surgery and long-term survival in advanced ovarian cancer: a DACOVA study. Gynecol Oncol 1990; 38: 203–9

6.

Eisenkop SM, Friedman RL, Wang HJ. Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced epithelial ovarian cancer: a prospective study. Gynecol Oncol 1998; 69: 103–8

7.

Chi DS, Liao JB, Leon LF, et al. Identification of prognostic factors in advanced epithelial ovarian carcinoma. Gynecol Oncol 2001; 82: 532–7

8.

Marsden DE, Friedlander M, Hacker NF. Current management of epithelial ovarian carcinoma: a review. Semin Surg Oncol 2000; 19: 11–19

9.

Dauplat J, LeBouëdec G, Pomel C, Scherer C. Cytoreductive surgery for advanced stages of ovarian cancer. Semin Surg Oncol 2000; 19: 42–8

10.

Eisenkop SM, Spirtos NM, Montag TW, et al. The impact of subspecialty training on the management of advanced ovarian cancer. Gynecol Oncol 1992; 47: 203–9

11.

Devi KU, Banfa UD, Ahuja V, Ramesh C. Induction chemotherapy and interval debulking surgery in advanced epithelial ovarian cancer. Int J Gynecol Cancer 2002; 12: 521 (abstr OV8)

12.

Webb MJ. Cytoreduction in ovarian cancer: achievability and results. Baillière’s Clin Obstet Gynaecol 1989; 3: 83–94

13.

Makar AP, Baekelandt M, Tropé CG, Kristensen GB. The prognostic significance of residual disease, FIGO substage, tumor histology and grade in patients with FIGO Stage III ovarian cancer. Gynecol Oncol 1995; 56: 175–80

CONCLUSION Locally advanced ovarian cancer can be one of the most challenging clinical problems facing the gynecologic surgeon. Resection of the primary tumor mass, pelvic peritoneal implants and adenopathy are integral components of the initial cytoreductive surgical effort and will often contribute significantly to the overall volume of residual disease. The cytoreductive procedures outlined in this chapter, including the retroperitoneal approach of radical oophorectomy, will facilitate extirpation of extensive pelvic disease in the majority of patients with ovarian cancer. Ultimately, sound surgical judgment, technical skill and meticulous attention to detail are the basic principles inherent to the safe and successful completion of radical pelvic surgery for ovarian cancer.

REFERENCES 1.

Jemal A, Tiwari RC, Murray T, et al. Cancer statistics 2004. CA Cancer J Clin 2004; 54: 8–29

2.

Pecorelli S, Odicino F, Maisonneuve P, et al. FIGO annual report on the results of treatment in gynaeco-

166

Ch 05 final.qxd

12/18/2006

15:16

Page 167

CYTOREDUCTIVE SURGERY: PELVIS

14.

Benedetti-Panici P, Maneschi F, Scambia G, et al. The pelvic retroperitoneal approach in the treatment of advanced ovarian carcinoma. Obstet Gynecol 1996; 87: 532–8

primary and recurrent gynecologic cancer. Obstet Gynecol 1984; 64: 715–20 27.

Soper JT, Couchman G, Berchuk A, Clarke-Pearson D. The role of partial sigmoid colectomy for debulking epithelial ovarian carcinoma. Gynecol Oncol 1991; 41: 239–44

28.

Obermair A, Hagenauer S, Tamandl D, et al. Safety and efficacy of low anterior en bloc resection as part of cytoreductive surgery for patients with ovarian cancer. Gynecol Oncol 2001; 83: 115–20

29.

Bridges JE, Leung Y, Hammond IG, McCartney AJ. En bloc resection of epithelial ovarian tumors with concomitant rectosigmoid colectomy: the KEMH experience. Int J Gynecol Cancer 1993; 3: 199–202

30.

Barnes W, Johnson J, Waggoner S, et al. Reverse hysterocolposigmoidectomy (RHCS) for resection of panpelvic tumors. Gynecol Oncol 1991; 42: 151–5

31.

Sugarbaker PH. Complete parietal and visceral peritonectomy of the pelvis for advanced primary and recurrent ovarian cancer. Cancer Treat Res 1996; 81: 75–87

32.

Sainz de la Cuesta R, Goodman A, Halverson SS, Fuller AF. En bloc pelvic peritoneal resection of the intraperitoneal pelvic viscera in patients with advanced epithelial ovarian cancer. Cancer J Sci Am 1996; 2: 152–7

15.

Barber HRK. Spread and treatment of advanced ovarian cancer. Baillière’s Clin Obstet Gynaecol 1989: 3: 23–9

16.

Fuks Z. Patterns of spread of ovarian carcinoma: relation to therapeutic strategies. In Newman CE, Ford C, Jordan H, eds. Ovarian Cancer. Oxford: Pergamon, 1989: 39–51

17.

Plentl AA, Friedman EA. Lymphatic system of the female genitalia. In Friedman EA, ed. Major Problems in Obstetrics and Gynecology. Philadelphia: Saunders, 1973: 173

18.

Pickel H. Epithelial ovarian cancer: spread. In Burghardt E, ed. Surgical Gynecologic Oncology. New York: Thieme, 1993

19.

Plentl AA, Friedman EA. Lymphatic system of the female genitalia. The morphologic basis of oncologic diagnosis and therapy. Major Probl Obstet Gynecol 1971; 2: 1–223

20.

Burghardt E, Pickel H, Lahousen M, Stettner H. Pelvic lymphadenectomy in the operative treatment of ovarian cancer. Am J Obstet Gynecol 1986; 155: 315–19

21.

Barber HRK, Brunschwig A. Pelvic exenteration for locally advanced and recurrent ovarian cancer. Surgery 1965; 58: 935–7

33.

Eisenkop SM, Nalick RH, Teng NNH. Modified posterior exenteration for ovarian cancer. Obstet Gynecol 1991; 78: 879–85

22.

Hudson CN. A radical operation for fixed ovarian tumors. J Obstet Gynaecol Br Commonw 1968; 75: 1155–60

34.

23.

Hudson CN, Chir M. Surgical treatment of ovarian cancer. Gynecol Oncol 1973; 1: 370–8

Bristow RE, del Carmen MG, Kaufman HS, Montz FJ. Radical oophorectomy with primary stapled colorectal anastomosis for resection of locally advanced epithelial ovarian cancer. J Am Coll Surg 2003; 197: 565–74

24.

Clayton RD, Obemair A, Hammond IG, et al. The western Australian experience of the use of en bloc resection of ovarian cancer with concomitant rectosigmoid colectomy. Gynecol Oncol 2002; 84: 53–7

35.

Ben-Baruch G, Schiff E, Sivan E, Menczer J. Cystotomy – for facilitation of optimal resection of adherent pelvic tumors. Gynecol Oncol 1991; 41: 139–40

25.

Sonnendecker EWW, Beale PG. Rectosigmoid resection without colostomy during primary cytoreductive surgery for ovarian carcinoma. Int Surg 1989; 74: 10–12

36.

Wu PC, Lang JH, Huang RL, et al. Intestinal metastasis and operation in ovarian cancer: a report on 62 cases. Baillière’s Clin Obstet Gynaecol 1989: 3: 95–108

26.

Berek JS, Hacker NF, Lagasse LD. Rectosigmoid colectomy and reanastomosis to facilitate resection of

37.

Scarabelli C, Gallo A, Franceschi S, et al. Primary cytoreductive surgery with rectosigmoid colon

167

Ch 05 final.qxd

12/18/2006

15:16

Page 168

SURGERY FOR OVARIAN CANCER

resection for patients with advanced epithelial ovarian carcinoma. Cancer 2000; 88: 389–97 38.

39.

40.

41.

42.

43.

Hertel H, Diebolder H, Herrmann J, et al. Is the decision for colorectal resection justified by histopathologic findings: a prospective study of 100 patients with advanced ovarian cancer. Gynecol Oncol 2001; 83: 481–4 Eisenkop SM, Spirtos NM. Procedures required to accomplish complete cytoreduction of ovarian cancer: is there a correlation with ‘biological aggressiveness’ and survival? Gynecol Oncol 2001; 82: 435–41 Lax SF, Petru E, Holzer E, et al. Mesenteric and mesocolic lymph node metastases from ovarian carcinoma: a clinicopathologic analysis. Int J Gynecol Cancer 1998; 8: 119–23 O’Hanlan KA, Kargas S, Schreiber M, et al. Ovarian carcinoma metastases to gastrointestinal tract appear to spread like colon carcinoma: implications for surgical resection. Gynecol Oncol 1995; 59: 200–6 Knight CD, Griffen FD. An improved technique for low anterior resection of the rectum using the EEA stapler. Surgery 1980; 88: 710–14 Ravitch MM. Intersecting staple lines in intestinal anastomosis. Surgery 1985; 97: 8–15

44.

Feinberg SM, Parker F, Cohen Z, et al. The double stapling technique for low anterior resection of rectal carcinoma. Dis Colon Rectum 1986; 12: 885–90

45.

Baran JJ, Goldstein SD, Resnik AM. The doublestaple technique in colorectal anastomosis: a critical review. Am Surg 1992; 58: 270–2

46.

Griffen FD, Knight CD Sr, Whitaker JM, Knight CD Jr. The double stapling technique for low anterior resection. Results, modifications, and observations. Ann Surg 1990; 211: 745–51

47.

Cohen Z, Myers E, Langer B, et al. Double stapling technique for low anterior resection. Dis Colon Rectum 1983; 26: 231–5

48.

Venkatesh KS, Morrison N, Larson DM, Ramanujam P. Triangulation stapling technique: an alternative approach to colorectal anastomosis. Dis Colon Rectum 1993; 36: 73–6

49.

Bailey HR, LaVoo JW, Max E, et al. Single-layer polypropylene colorectal anastomosis: experience with 100 cases. Dis Colon Rectum 1984; 27: 19–23

168

50.

Harder F, Vogelbach P. Single-layer end-on continuous suture of colonic anastomosis. Am J Surg 1988; 155: 611–14

51.

Max E, Sweeney WB, Bailey HR, et al. Results of 1,000 single-layer continuous polypropylene intestinal anastomosis. Am J Surg 1991; 162: 461–7

52.

Lazorthes F, Fages P, Chiotasso P, et al. Resection of the rectum with construction of a colonic reservoir and colo-anal anastomosis for carcinoma of the rectum. Br J Surg 1986; 73: 136–8

53.

Parc R, Tiret E, Frileux P, et al. Resection and coloanal anastomosis with colonic reservoir for rectal carcinoma. Br J Surg 1986; 73: 139–41

54.

Leo E, Belli F, Baldino MT, et al. New perspective in the treatment of low rectal cancer: total rectal resection and coloendoanal anastomosis. Dis Colon Rectum 1994; 37: S62–8

55.

Guidozzi F, Ball JH. Extensive primary cytoreductive surgery for advanced epithelial ovarian cancer. Gynecol Oncol 1994; 53: 326–30

56.

Berek J, Hacker N, Lagasse L, Leuchter R. Lower urinary tract resection as part of cytoreductive surgery for ovarian cancer. Gynecol Oncol 1987; 13: 87–92

57.

Manolitsas TP, Copeland LJ, Cohn DE, et al. Ureteroileoneocystotomy: the use of an ileal segment for ureteral substitution in gynecologic oncology. Gynecol Oncol 2002; 84: 110–14

58.

Tamussino KF, Lim PC, Webb MJ, et al. Gastrointestinal surgery in patients with ovarian cancer. Gynecol Oncol 2001; 80: 79–84

59.

Vignali A, Fazio VW, Lavery IC, et al. Factors associated with the occurrence of leaks in stapled rectal anastomosis: a review of 1,014 patients. J Am Coll Surg 1997; 185: 105–13

60.

Schmidt O, Merkel S, Hohenberger W. Anastomotic leakage after low rectal stapler anasomtosis: significance of intraoperative anastomotic testing. Eur J Surg Oncol 2003; 29: 239–43

61.

Rullier E, Laurent C, Garrelon JL, et al. Risk factors for anastomotic leakage after resection of rectal cancer. Br J Surg 1998; 85: 355–8

62.

Galandiuk S, Fazio VW. Postoperative irrigationsuction drainage after pelvic colonic surgery: a prospective randomized trial. Dis Col Rectum 1991; 34: 223–8

Ch 05 final.qxd

12/18/2006

15:16

Page 169

CYTOREDUCTIVE SURGERY: PELVIS

63.

Merad F, Hay JM, Fingerhut A, et al. Is prophylactic pelvic drainage useful after elective rectal or anal anastomosis? a multicenter controlled randomized trial. French Association for Surgical Research. Surgery 1999; 125: 529–35

73.

Scarabelli C, Gallo A, Visentin MC, et al. Systematic pelvic and para-aortic lymphadenectomy in advanced ovarian cancer patients with no residual intraperitoneal disease. Int J Gynecol Cancer 1997; 7: 18–26

74.

Spirtos NM, Gross GM, Freddo JL, Ballon SC. Cytoreductive surgery in advanced epithelial cancer of the ovary: the impact of aortic and pelvic lymphadenectomy. Gynecol Oncol 1995; 56: 345–52

64.

Urbach D, Kennedy ED, Cohen MM. Colon and rectal anastomoses do not require routine drainage. Ann Surg 1999; 229: 174–80

65.

Spirtos NM, Eisenkop SM, Schlaerth JB, Ballon SC. Second-look laparotomy after modified posterior exenteration: patterns of persistence and recurrence in patients with Stage III and Stage IV ovarian cancer. Am J Obstet Gynecol 2000; 182: 1321–7

75.

Saygili U, Guclu S, Uslu T, et al. Does systematic lymphadenectomy have a benefit on survival of suboptimally debulked patients with Stage III ovarian carcinoma? A DECOG study. J Surg Oncol 2002; 81: 132–7

66.

Bristow RE, Montz FJ. Complete surgical cytoreduction of advanced ovarian carcinoma using the argon beam coagulator. Gynecol Oncol 2001; 83: 39–48

76.

Burghardt E, Girardi F, Lahousen M, et al. Patterns of pelvic and paraaortic lymph node involvement in ovarian cancer. Gynecol Oncol 1991; 40: 103–6

67.

Bristow RE, Smith-Sehdev AE, Kaufman HS, Montz FJ. Ablation of metastatic ovarian carcinoma with the argon beam coagulator: pathologic analysis of tumor destruction. Gynecol Oncol 2001; 83: 49–55

77.

Morice P, Joulie F, Camatte S, et al. Lymph node involvement in epithelial ovarian cancer: analysis of 276 pelvic and para-aortic lymphadeneectomies and surgical implications. J Am Coll Surg 2003; 197: 198–205

68.

van Damm PA, Tjalma W, Weyler J, et al. Ultraradical debulking of epithelial ovarian cancer with the ultrasonic surgical aspirator: a prospective trial. Am J Obstet Gynecol 1996; 174: 943–50

78.

69.

Rose PG. The cavitational ultrasonic surgical aspirator for cytoreduction in advanced ovarian cancer. Am J Obstet Gynecol 1992; 166: 843–6

Scarabelli C, Gallo A, Zarrelli A, et al. Systematic pelvic and para-aortic lymphadenectomy during cytoreductive surgery in advanced ovarian cancer: potential benefit on survival. Gynecol Oncol 1995; 56: 328–37

79.

Donovan JT, Veronikis DK, Powell JL, et al. Cytoreductive surgery for ovarian cancer with the cavitron ultrasonic surgical aspirator and the development of disseminated intravascular coagulation. Obstet Gynecol 1994; 83: 1011–14

Hacker NF. Systematic pelvic and paraaortic lymphadenectomy for advanced ovarian cancer therapeutic advance or surgical folly? Gynecol Oncol 1995; 56: 325–7

80.

Nahhas WA. A potential hazard of the use of the surgical ultrasonic aspirator in tumor reductive surgery. Gynecol Oncol 1991; 40: 81–3

Eisenkop SM, Spirtos NM. The clinical significance of occult macroscopically positive retroperitoneal nodes in patients with epithelial ovarian cancer. Gynecol Oncol 2001; 82: 143–9

81.

Abrams HR, Spiro R, Goldstein N. Metastases in carcinoma: analysis of 1000 autopsied cases. Cancer 1950; 3: 74–85

82.

Metzger PP. Modified packing technique for control of presacral pelvic bleeding. Dis Colon Rectum 1988; 31: 981–2

70.

71.

72.

Wu PC, Lang JH, Huang RL, et al. Lymph-node metastasis and retroperitoneal lymphadenopathy in ovarian cancer. Baillière’s Clin Obstet Gynaecol 1989: 3: 143–55

169

Ch 06 final.qxd

12/18/2006

15:19

Page 171

CHAPTER 6

Cytoreductive surgery: abdominal retroperitoneum and adenopathy Jason D Wright, Thomas J Herzog

INTRODUCTION Ovarian cancer is commonly characterized as an intraperitoneal process; however, the retroperitoneal lymph nodes are a common site of metastatic involvement in both primary and recurrent disease. Consequently, evaluation of the retroperitoneum is an important part of the comprehensive staging of women with epithelial ovarian cancer. In 1986 the International Federation of Gynecology and Obstetrics (FIGO) introduced a new staging system for ovarian cancer. In the revised system patients with metastatic disease in the retroperitoneal lymph nodes are classified as stage IIIC.1 While this change highlights the diagnostic importance of nodal evaluation, the therapeutic role of lymphadenectomy is still evolving. In this chapter the pattern of nodal involvement and technique for dissection of the retroperitoneum are presented.

INCIDENCE AND PATTERN OF INVOLVEMENT Retroperitoneal lymphatic involvement by epithelial ovarian cancer is relatively common and should be considered one of the primary routes of spread. Each ovary is drained by three principal lymphatic networks. The major lymphatic drainage network from the ovaries travels cephalad, following the course of the ovarian vessels through the infundibulopelvic lig-

ament, to the level of the inferior pole of the kidney and emptying into the para-aortic and para-caval nodal basins. A second, principal, lymphatic system travels from the ovary through the broad ligament to the pelvic (external iliac and obturator) basins.2–4 A third, less common, route of lymphatic drainage can occur along the course of the round ligament to the inguinal lymph nodes. The rates of lymphatic involvement by epithelial ovarian cancer reported in the literature vary widely and are dependent on the clinical stage of disease and the extent of dissection performed. Because systematic retroperitoneal lymphadenectomy is not routinely required in the staging of ovarian cancer, the precise incidence of lymph node spread remains somewhat illdefined. Autopsy series have found metastatic ovarian cancer in the retroperitoneal nodes of 64–80% of patients examined.5,6 Large studies of ovarian cancer patients who have undergone lymph node sampling or dissection have reported lymphatic involvement in 30–66% of patients.7–16 For women with otherwise early-stage disease, lymphatic involvement occurs in 3–23% of patients with stage I and in 10–50% with stage II neoplasms.7,8,11,12,14,17 Lymphatic involvement is more common with advanced-stage intraperitoneal or extra-abdominal disease, with nodal metastases being documented in 13–74% of stage III and in 33–88% of stage IV patients.7,8,11,12,14 The most frequent pattern of tumor involvement is spread to both the pelvic and para-aortic nodal chains, which occurs in 53–73% of all patients with positive

171

Ch 06 final.qxd

12/18/2006

15:19

Page 172

SURGERY FOR OVARIAN CANCER

nodes.7,8,10 It is not uncommon, however, to have isolated spread to either the pelvic or para-aortic chains. Isolated para-aortic metastases may be present in 15–33% of all patients with positive nodes. Similarly, retroperitoneal disease may be confined to the pelvic nodes in 8–28% of such patients.3,7,8,10,14,18 The most common location for metastatic spread is the paraaortic lymph nodes between the inferior mesenteric and renal arteries.7,19 Onda et al. carefully evaluated the location of lymph node spread in 110 patients undergoing nodal sampling. The authors noted metastases to the aortic nodes between the inferior mesenteric artery (IMA) and renal artery in 79% of the subjects with positive nodes.7

ANATOMY OF THE RETROPERITONEUM Successful surgical evaluation of the retroperitoneum requires a thorough knowledge of the regional anatomy

Hiatus for vena cava Hiatus for esophagus Inferior phrenic artery Hepatic artery Middle suprarenal artery Right lumbar arteries (3, 4) Common iliac artery 5th lumbar artery Middle sacral artery

as well as complete operative exposure. Retroperitoneal anatomical variants are relatively common. In a series of 309 patients undergoing retroperitoneal exploration, anatomic abnormalities were noted in 15% of patients. This included anomalies of the urinary tract and vascular structures in 1.6% and 14% of patients, respectively.20

Vasculature Abdominal aorta

The aorta enters the abdominal cavity from the thorax through the aortic hiatus of the diaphragm. Upon entering the abdomen, the aorta descends in a midline position and bifurcates into the common iliac arteries at the level of the L4–L5 interspace (Figure 6.1). During its course the aorta gives off several branches. The major vascular supply to the gastrointestinal tract consists of three unpaired vessels that arise from the ventral surface of the aorta. The celiac

Left gastric artery Celiac trunk Splenic artery Left renal artery Superior mesenteric artery Ovarian artery Inferior mesenteric artery Iliolumbar artery Internal iliac artery External iliac artery Obturator artery

Figure 6.1 Anatomy of the aorta, iliac arteries and their relationships to surrounding structures

172

Ch 06 final.qxd

12/18/2006

15:19

Page 173

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

trunk is the first branch and arises at the level of T12. The superior mesenteric artery (SMA) arises above the renal arteries and is the major arterial supply for the small intestine, right colon and transverse colon. The IMA arises 3–4 cm proximal to the aortic bifurcation and supplies the descending colon and rectum. In addition to these unpaired vessels, five paired tributaries arise from the abdominal aorta. The inferior phrenic arteries arise cephalad to the celiac trunk or directly from the celiac trunk to supply the esophagus, liver and adrenal glands. The middle adrenal arteries are found just above the renal vessels and supply a portion of the blood supply to the adrenal glands. The renal arteries begin just below the adrenal arteries at approximately the level of L2. The right renal artery usually courses posterior (dorsal) to the vena cava in its course to the right kidney. The paired ovarian vessels arise 2–3 cm below the level of the renal arteries. The ovarian arteries course over the ureters as they descend toward the pelvis. Anomalies of the ovarian arteries are relatively frequent. These vessels may arise from the renal vessels, from a single trunk or from an aberrant location on the aorta. Several pairs of lumbar arteries arise from the posterolateral surface of the aorta (Figure 6.2). These vessels each give off several branches to supply the spinal column and

Ventral ramus lumbar nerve Ventral branch lumbar artery Transversus abdominis muscle External and internal oblique muscle Psoas muscle Petit's triangle Latissimus dorsi muscle Quadratus lumborum muscle

muscle groups of the back. Finally, the middle sacral artery arises from the posterior side of the aorta just before the aortic bifurcation and supplies the rectum and anus. Inferior vena cava

The inferior vena cava (IVC) enters the abdominal cavity through the central tendon of the diaphragm. The IVC descends through the abdomen to the right of the aorta. As the IVC approaches the pelvis it assumes a more posterior position as the right common iliac artery passes over it. Superiorly, the first vessels to enter the IVC are the three hepatic veins (Figure 6.3). Below the hepatic veins the right adrenal vein usually enters the IVC. The left adrenal vein empties into the left renal vein; occasionally the right adrenal vein drains into the right renal vein. The next major vessels to enter the IVC are the renal veins. The left renal vein courses between the SMA and the aorta on its way to entering the IVC (Figure 6.4). The right ovarian vein drains directly into the IVC or the right renal vein (3–22% of cases), while the left ovarian vein empties into the left renal vein.20 Finally, several sets of lumbar veins also enter the IVC. These veins are highly variable and often drain into the left renal vein. The lumbar veins are typically connected

Sympathetic trunk Ascending colon

IVC

Aorta

Lumbar vertebra

Lateral branches of dorsal rami of nerve and artery Erector spinae muscle

Figure 6.2 Anatomy of the lumbar arteries and their branches to the spinal column and the muscles of the back. IVC, inferior vena cava

173

Ch 06 final.qxd

12/18/2006

15:19

Page 174

SURGERY FOR OVARIAN CANCER

Hepatic veins

Inferior phrenic vein

Inferior vena cava Hemiazygos vein Left adrenal vein Right adrenal vein Ovarian vein

Left renal vein

Ascending lumbar vein Lumbar vein Iliolumbar vein Ovarian vein

Middle sacral vein

Left internal iliac vein

Figure 6.3 Anatomy of the inferior vena cava, the common iliac veins and their relationships to surrounding structures

Right adrenal artery and vein Inferior vena cava

Celiac trunk Superior mesenteric artery Left renal artery and vein Aorta

Lumbar artery and vein Inferior mesenteric artery

Left common iliac artery and vein

Figure 6.4 Anatomic relationship of the left renal vein to the superior mesenteric artery and aorta

174

Ch 06 final.qxd

12/18/2006

15:19

Page 175

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

by ascending lumbar veins that run parallel to the IVC (Figure 6.3). The IVC bifurcates into the common iliac veins, which enter the pelvis. Iliac vessels

The paired common iliac arteries branch from the aorta, continue for approximately 5 cm, and bifurcate into the internal and external iliac arteries. The bifurcation of the common iliac vein commonly occurs distal to the bifurcation of the common iliac arteries. The only branches of the common iliac arteries are the internal and external iliac vessels. The ureters enter the pelvis by crossing over the bifurcation of the ipsilateral common iliac artery (Figure 6.1). The right common iliac vein lies behind the artery and courses from the lateral to the medial aspect of the artery as the common iliac vein bifurcates. The left common iliac vein runs on the medial, posterior aspect of the left common iliac artery. Each common iliac vein gives rise to an internal and external iliac vein as well as an iliolumbar vein and an ascending lumbar vein. The middle sacral vein is a tributary to the left common iliac vein and is a coalescence of multiple smaller veins, all of which lie within the presacral connective tissue. Damage to the presacral venous plexus can lead to bleeding that is often difficult to control. The paired external iliac artery travels along the pelvic sidewall from the bifurcation of the common iliac artery to the inguinal ligament, where it becomes the femoral artery. Branches of the external iliac artery include a small vessel to the psoas, the deep circumflex iliac and the inferior epigastric arteries. Both external iliac veins lie posteriorly and medially to the corresponding arteries. Occasionally (25% of cases), an accessory obturator vein drains into the external iliac vein. The paired internal iliac artery divides into an anterior and posterior division approximately 3–4 cm from its origin. The posterior division of the iliac artery gives rise to four collaterals: the iliolumbar artery, superior lateral sacral artery, inferior lateral sacral artery and the superior gluteal arterior. These vessels supply the pelvic wall and gluteal muscles. The

anterior division of the internal iliac artery gives rise to multiple branches and is highly variable. A common branch supplies the umbilical, uterine, vaginal and superior vesical branches. The obturator artery courses under the obturator nerve and above the obturator vein through the obturator fossa en route to the obturator canal. Other direct branches of the anterior division include the inferior vesical, middle rectal, internal pudendal and inferior gluteal arteries. In general, the venous drainage of the pelvis follows the arterial supply and coalesces in the internal iliac vein. However, the venous drainage system of the pelvis is highly variable and contains many collateral branches.

Kidney, adrenal gland and ureter Kidney

The kidneys are retroperitoneal organs that lie at the level of T12 to L3. Each kidney is approximately 10 cm in length and surrounded by perinephric fat. A fibroareolar renal fascia (Gerota’s fascia) surrounds the kidney, adrenal gland and perinephric fat. The arterial supply to the kidneys is via the renal arteries. Upon entering the hilum of the kidney, each renal artery bifurcates and eventually splits into several segments that supply the renal parenchyma. Abnormalities of the renal vasculature at the hilum of the kidney are the most common anatomic abnormalities of the retroperitoneum.20 Accessory renal arteries are frequent and most commonly arise from the aorta. The venous drainage of the kidneys is by way of the renal veins. Accessory renal veins may be encountered, and are more frequent on the right. The renal vein most commonly is dorsal to the artery, although this relationship is highly variable. Occasionally, the left renal vein or tributary courses behind the aorta. Ureter

The ureters are muscular conduits that arise from the renal pelvis to convey urine to the bladder. As the ureters descend through the retroperitoneum to enter the pelvis, they pass dorsal to the ovarian vessels. The

175

Ch 06 final.qxd

12/18/2006

15:19

Page 176

SURGERY FOR OVARIAN CANCER

ureters then enter the pelvis and cross the common iliac artery near its bifurcation. Within the pelvis the ureters course near the pelvic wall, then pass under the uterine arteries, approximately 1.5 cm lateral to the internal os of the cervix. The ureters pass through the ureteric tunnel within the cardinal ligament en route to the bladder. The vascular supply of the ureters lies within the adventitia. Within the abdomen, the ureter receives its longitudinal blood supply from the medial surface (aorta, renal, ovarian, iliac arteries), while the pelvic ureter receives its blood supply laterally from the vasculature along the pelvic sidewall (uterine and superior and inferior vesicular arteries) (Figure 6.5). The most common anatomic abnormality of the ureter is ureteral dupli-

cation. Unilateral duplication occurs six times more commonly than bilateral duplication.20 Adrenal glands

The adrenal glands are structures 3–5 cm long and located superior to the kidney. Each gland is enveloped within the renal fascia. The right adrenal gland is pyramidal in shape. The medial portion of the gland lies posterior to the vena cava, while the superior aspect of the organ abuts the bare area of the liver. The left adrenal gland lies posterior to the stomach and pancreas and abuts the diaphragm superiorly. The adrenal glands are supplied by the superior, middle and inferior adrenal arteries. The venous drainage is by way of a single, adrenal vein.

Lymphatics

Renal artery

Ovarian artery

Aorta

Internal iliac artery

Superior vesicular artery Inferior vesicular artery

Figure 6.5 Anatomy of the ureter and its blood supply

176

The lymphatic drainage of the pelvis and lower abdomen can be classified into several functional groups. This includes the common iliac, external iliac, internal iliac, obturator and para-aortic lymph nodes (Figure 6.6). The external iliac lymph nodes are located lateral to the artery, medial to the vein and between the two vessels. The nodes of the internal iliac chain lie within the adipose tissue around the branches of the internal iliac vessels. The obturator nodes include the node-bearing tissue within the obturator fossa. The common iliac nodes encompass those nodes between the external iliac and para-aortic node groups. Most of the common iliac nodes lie on the lateral surface of the corresponding vessel. The para-aortic nodes lie on the anterior and lateral aspects of the aorta. Additionally, lymph nodes lie between the aorta and vena cava and on the surface of the vena cava.

Diagnostic lymphadenectomy Diagnostic lymphadenectomy, or lymph node sampling, is an important component of the staging procedure in women with epithelial ovarian cancer apparently confined to the ovary. Approximately 30% of patients with apparent early-stage ovarian carcinoma

Ch 06 final.qxd

12/18/2006

15:19

Page 177

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

Left renal vein Right renal vein Vena cava

Aorta

Ureter Para-aortic lymph nodes

Common iliac lymph nodes Hypogastric lymph nodes

Ovarian vessels

Common iliac artery and vein Hypogastric artery and vein External iliac artery and vein

External iliac lymph nodes

Obturator lymph nodes

Figure 6.6 Lymphatic drainage routes relevant to ovarian cancer including the pelvic and para-aortic nodal basins

are upstaged when comprehensive surgical staging is performed.21 Up to 23% of women with apparent stage I ovarian cancer have occult nodal disease.22 The incidence of nodal disease increases with substage. Lymph node metastases are found in 2–13% of IA, 15–56% of IB and 14–38% of apparent IC tumors.14,23,24 Likewise, lymphatic spread appears to increase with tumor grade.14,23,25 Finally, several studies have reported that the propensity for nodal spread appears to be related to histologic subtype, with the risk of lymphatic metastasis being lowest in women with mucinous tumors.14,22,24 It has been well documented that inspection and palpation are insensitive for the detection of metastatic nodal disease.10,14,26 In one series, Wu et al. noted that one-third of grossly normal-appearing lymph nodes were microscopically involved by metastatic tumor.10 Not surprisingly, 87.5% of nodes appearing suspicious were histologically positive. Older reports

recommended ipsilateral lymphadenectomy for tumors that appeared to be unilateral.27 However, the literature is now replete with cases of contralateral lymphatic dissemination, and it now seems that the staging procedure of choice for women with apparent stage I ovarian cancer should include bilateral sampling of the pelvic and para-aortic nodes up to the level of the renal veins. Cass et al. noted that 30% of the subjects in their series had isolated, contralateral lymph node metastases. One patient had pelvic, one patient had para-aortic, and one patient had combined pelvic and para-aortic contralateral nodal disease.25 Based on these data, bilateral nodal evaluation appears to yield important diagnostic information and should be performed whenever feasible. Diagnostic lymph node sampling is also commonly recommended as part of the staging procedure for women with suspected tumors of low malignant potential (borderline tumors). Retroperitoneal

177

Ch 06 final.qxd

12/18/2006

15:19

Page 178

SURGERY FOR OVARIAN CANCER

lymphatic involvement occurs in 17–22% of patients with tumors of low malignant potential (LMP).28,29 The principal argument for nodal evaluation for suspected LMP tumors rests in the uncertainty of frozensection diagnosis reliably to exclude a frankly invasive neoplasm. Retrospective reviews have documented that approximately 3–27% of patients with a frozen section diagnosis of an LMP tumor will be upgraded to invasive cancer on final pathology.28,30,31 In those found with invasive malignancies, the results of lymphadenectomy provide important prognostic information and help to guide further therapy. For women with confirmed LMP tumors, surgical staging has not conferred a survival advantage.28,30

Systematic lymphadenectomy While the therapeutic potential of systematic lymphadenectomy for advanced-stage ovarian cancer remains uncertain, several reports have described a survival advantage for those patients who undergo the procedure.9,15,32–35 Burghardt et al. initially described their experience in patients who underwent systematic lymphadenectomy. Among stage III patients, the 5year survival was 53% in the group who underwent lymphadenectomy compared to 13% for those who did not undergo nodal dissection.9 di Re et al. reported the results of lymphadenectomy in 310 patients. Lymph node status was an important predictor of survival, even after adjusting for other known prognostic factors. In addition, lymphadenectomy was associated with an improved 5-year survival for both optimally and suboptimally cytoreduced patients.15 The therapeutic value of lymphadenectomy is further supported by the finding that patients with macroscopically positive nodes that are completely debulked have a similar survival to those patients with negative, or microscopically positive nodes.32 It has been suggested that the retroperitoneal nodes provide a pharmacologic sanctuary against the effects of systemic chemotherapy, which may account for the apparent survival benefit associated with systematic lymphadenectomy. This concept is supported by the high rate of tumor-containing lymph nodes

178

found in patients undergoing second-look surgery after cytotoxic therapy. Nodal metastases have been documented in 26–77% of patients after chemotherapy.36 A biologic basis for this phenomenon is provided by the observation that nodal metastases in ovarian cancer patients are often diploid and contain a low S-phase fraction, characteristics of a relatively chemoresistant phenotype.37 While the above studies are provocative, the currently available data are all based upon retrospective studies. A prospective evaluation of the role of lymphadenectomy is clearly warranted.38 Given the uncertainty of the value of therapeutic lymphadenectomy, it is not currently considered the standard of care. However, at a minimum it does appear prudent to resect any bulky nodal disease in patients who are otherwise free of macroscopic residual disease.

RETROPERITONEAL LYMPHADENECTOMY: SURGICAL PRINCIPLES AND COMPLICATIONS Systematic pelvic lymphadenectomy The technique for performing systematic pelvic lymphadenectomy is displayed in Figure 6.7.39 To begin the pelvic lymphadenectomy the posterior leaf of the peritoneum is incised to expose the retroperitoneum. This incision must extend from the round ligament to the pelvic brim to allow adequate exposure to the retroperitoneum. Upon entering the retroperitoneal space, one should identify the ureter as it crosses over the bifurcation of the common iliac artery. Opening the potential spaces of the pelvis is a key initial step that helps to define the anatomy of the region and to facilitate nodal dissection (Figure 6.7b). The pararectal space is bordered medially by the rectum, laterally by the internal iliac artery, anteriorly by the cardinal ligament and posteriorly by the sacrum. The pararectal space is opened with blunt dissection. After the pararectal space has been opened, the ureter is reflected medially. The paravesical space bordered by the superior vesicle artery medially, the external iliac

Ch 06 final.qxd

12/18/2006

15:19

Page 179

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

Ureter Aorta

Ovarian vessels

Sigmoid colon Ligated ovarian vessels

Appendix

Hypogastric artery

Ureter External iliac artery and vein

External iliac artery and vein Genitofemoral nerve Bladder

Obturator artery and vein Obturator nerve

Round ligament Round ligament

Genitofemoral nerve Deep circumflex vein a

b

c

Figure 6.7 Technique of systematic pelvic lymphadenectomy. (a) The lateral pelvic peritoneum is opened along the psoas muscles and extended to the level of the common iliac vessels. (b) A combination of sharp and blunt dissection is used to open the retroperitoneal spaces (para-rectal and para-vesicle) and the ureter reflected medially. (c) The dissection begins with the lymph-bearing tissue along the external iliac vessels

artery laterally and the pubic ramus anteriorly is also opened with blunt dissection. Establishment of these spaces and therefore the anatomy provides the proper framework for the safe and efficacious performance of pelvic lymphadenectomy. The lymphadenectomy begins with the external iliac nodes (Figure 6.7c). The genitofemoral nerve is visible along the pelvic wall, lateral to the external iliac vessels. Care should be taken to preserve this nerve if possible. The node-bearing fat pad over the external iliac is then gently grasped with Singley forceps and elevated. The nodes are then swept from cephalad to caudad and lateral to medial over the iliac vessels. Metzenbaum scissors are utilized to free the nodal tissue from the underlying vasculature as needed. Small vessels are clipped with vascular clips, then transected. The nodal tissue around the external iliac artery and vein is carefully removed as traction is

applied anteriorly and medially. The dissection is performed from the common iliac vessels down to the level of the deep circumflex iliac vein. Occasionally an anomalous obturator vessel arises near the distal aspect of the external iliac vessels. These vessels should be carefully identified and ligated. After completion of the external iliac node dissection, attention is turned to the obturator fossa (Figure 6.8). A vein retractor is placed to reflect the external iliac vessels away from the psoas muscle. The obturator space is then entered. With a combination of blunt and sharp dissection, the obturator nerve is identified. The lymphatics between the obturator nerve and external iliac vessels are cleared and the node-bearing tissue between the obturator nerve and obturator vessels gently released. The obturator artery and vein lie below the nerve. The vessels and nerve should be visualized throughout the dissection. Disruption of

179

Ch 06 final.qxd

12/18/2006

15:19

Page 180

SURGERY FOR OVARIAN CANCER

Figure 6.8 Lymph node dissection within the obturator fossa

the obturator vein can lead to bleeding that is difficult to control. Bleeding from the obturator vein is best controlled with direct visualization and ligation of the vessel when possible. When the bleeding source is not readily visible, pressure and packing is generally effective in the obturator space. Dissection of the remainder of the internal iliac nodes and nodes from the obturator space to the bifurcation of the common iliac vessels is then performed. The surgeon should be vigilant when dissecting the node-bearing tissue near the bifurcation of the common iliac vein; this is a frequent source of major bleeding. Finally, the remainder of the common iliac nodes are removed to complete the pelvic nodal dissection.

Systematic infrarenal para-aortic lymphadenectomy Aortocaval lymphadenectomy requires a generous midline abdominal incision to provide adequate exposure to the great vessels and upper abdomen. The description of the technique for extirpating the nodal tissue of the retroperitoneum is arbitrarily divided into infrarenal and suprarenal sections. The dissections begin in a similar fashion with reflection of the abdominal contents to allow maximal exposure. The posterior peritoneum of the right paracolic gutter is

180

opened with sharp dissection. The incision is begun at the junction of the cecum and terminal ileum and continued along the length of the right colon, above the hepatic flexure to the foramen of Winslow (Figure 6.9). A second peritoneal incision is made, again at the level of the ileocecal junction, and carried through the peritoneum of the small bowel mesentery along the inferior aspect of the third portion of the duodenum to the ligament of Treitz. The right colon and small bowel are then reflected off the renal fascia. The ureter and right ovarian vessel should be carefully identified and mobilized with the large bowel. Retractors are gently placed to optimize exposure to the operative field. The infrarenal lymphadenectomy will remove the node-bearing tissue of the lower paraaortic and common iliac areas. The lateral extent of the dissection is the fat plane on the lateral border of the cava (Figure 6.10). The fat pad overlying the right common iliac artery is identified and elevated. The adventitial sheath of the right common iliac artery is then gently incised and mobilized. The incision is carried superiorly along the ventral surface of the aorta to the inferior mesenteric artery. Sharp dissection is used to mobilize the sheath laterally, along the right side of the aorta. The mobilization is continued along the lateral surface of the aorta to the point where the lumbar arteries arise. Care should be taken to avoid traumatizing these vessels, which can result in bleeding that is difficult to control. If large-volume disease is present, the lumbar vessels can be ligated and divided. The ovarian arteries are identified and transected at their origins. Attention is then focused again on the common iliac vessels. The sheath over the common iliac artery is further freed to expose the common iliac vein. In a manner similar to the artery, the fat pad is elevated, the sheath of the vessel incised, and the incision carried along the anterior–lateral surface of the vessel toward the IVC. The ascending lumbar and iliolumbar veins may be encountered at this phase of the dissection and may be ligated and divided if necessary (Figure 6.10, inset). This dissection is carried up the IVC to the origin of the renal vein. The node-

Ch 06 final.qxd

12/18/2006

15:19

Page 181

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

Foramen of Winslow

Splenic vein Inferior duodenal fold

Inferior mesenteric vein

Figure 6.9 Lines of peritoneal incisions for adequate exposure of the retroperitoneum during aortic node dissection

Superior mesenteric artery

Right renal artery and vein

Left ovarian vein Left ovarian artery Inferior mesenteric artery

C

B

VC

Aorta

A

Lymph nodes D

Vertebral body

Psoas muscle

Figure 6.10 Nodal basins excised during infrarenal para-aortic lymphadenectomy. Inset: cross-section: A, para-caval nodes; B, precaval nodes; C, pre-aortic nodes; D, para-aortic nodes; VC, vena cava

181

Ch 06 final.qxd

12/18/2006

15:19

Page 182

SURGERY FOR OVARIAN CANCER

bearing specimen over the common iliac vein and IVC is now rolled laterally and removed. Small perforating vessels are secured with hemoclips. The final phase of the right-sided lymphadenectomy is the extirpation of the aortocaval nodes that lie between the IVC and lower abdominal aorta. The previous mobilization of the nodal bundles off the IVC and aorta will allow for effective isolation of the aortocaval tissue. The tissue is gently elevated and removed by sharp dissection, again providing hemostasis with vascular clips. During the aortocaval dissection caution should again be exercised to avoid the lumbar vessels if these were not previously secured. The right renal artery lies posterior to the IVC. Care should be taken to avoid disrupting this channel. Several approaches are available to gain exposure for the left-sided lymphadenectomy. In a manner similar to the right-sided dissection, the peritoneal reflection of the left paracolic gutter can be opened through the level of the splenocolic ligament. This will allow for medial reflection of the left colon. The ureter and the lateral extent of the fatty nodal tissue are identified and the dissection initiated. Alternatively, the bowel contents can be reflected laterally. If this approach is chosen the origin of the IMA from the aorta is identified. The vessel is retracted and the node-bearing tissue on the lateral aspect of the aorta identified. If necessary, the IMA can be ligated and divided to enhance exposure. The procedure again begins with the fat pad and sheath of the common iliac artery. The adventitial sheath on the anterior surface is incised and this incision continued up the aorta to the level of the left renal vein. If a right-sided infrarenal lymphadenectomy has already been performed, the nodal tissue overlying the aorta has already been partially mobilized. The left renal vein and ovarian vein should be carefully identified. Anatomic variants frequently arise from this anastomotic drainage system. The ovarian vein is normally sacrificed at this point in the dissection. The sheath over the aorta is gently dissected laterally. The nodal tissue is then bluntly dissected off the left side of the aorta and removed. The

182

lumbar vessels are sacrificed if necessary. The nodebearing tissue is removed from the level of the common iliac to the renal vein. The iliolumbar, ascending lumbar and middle sacral veins arise from the left common iliac vein and should be identified and secured if they obstruct the field of dissection.

Systematic suprarenal para-aortic lymphadenectomy Dissection of the suprarenal para-aortic nodal chain follows the same general principles as that of the infrarenal dissection. The adventitial sheaths of the vessels are incised and the node-bearing complexes mobilized laterally. Adequate intraoperative exposure is imperative. In the majority of cases the dissection can be performed through a midline abdominal incision. In rare cases where bulky nodal disease impinges on the crura of the diaphragm or extends into the mediastinum, a thoracoabdominal incision may be necessary. Access for the suprarenal lymphadenectomy is gained in a manner similar to that described for the infrarenal dissection. On the right side, the pancreas and duodenum can be carefully separated from the right renal fascia and rolled medially to gain access to the hilum of the right kidney (Figure 6.11). A similar maneuver is performed to gain access to the left suprarenal region. In this case, the descending colon and the tail of the pancreas are mobilized medially, exposing the left kidney and suprarenal aorta. The dissection begins with the left renal vein (Figure 6.12). The venous tributaries to the adrenal gland and ovary should be identified, ligated and transected. The vein is then carefully inspected for other vascular channels, such as the lumbar vein, which are isolated and divided. The adventitial sheath of the left renal vein is then incised at the vessel’s junction with the kidney and opened toward the renal hilus. The nodal tissue is gently mobilized to the cranial side of the vessel. The sheath of the renal artery is then incised at the aortorenal junction. The renal vein is retracted and the incision extended along the renal artery toward the kidney. The node pad is freed along the upper margin of the vessel. The node-bearing

Ch 06 final.qxd

12/18/2006

15:19

Page 183

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

Gallbladder Hepatoduodenal ligament Foramen of Winslow Kidney Duodenum Hepatic flexure of colon

a

Hepatoduodenal ligament Foramen of Winslow Common bile duct Portal vein Right renal vein Posterior pancreaticoduodenal artery Ascending colon Hepatic flexure

b

Figure 6.11 Retroperitneal access for suprarenal lymphadenectomy. (a) Line of peritoneal incision extending along the right paracolic gutter, across the hepatic flexure, and up to the foramen of Winslow. (b) The ascending colon, head of the pancreas and duodenum are carefully separated from their retroperitoneal attachments and mobilized medially

tissue over the left renal vessels is now completely mobilized cranially and can be extirpated with sharp dissection, clipping any perforating vessels. The dissection is continued above the renal vessels if adenopathy along the SMA or celiac trunk is identified. The incision in the adventitial sheath of the aorta is continued proximally and the node-bearing tissue is mobilized to the lateral side of the aorta. The nodal tissue between the aorta and left adrenal gland is then removed. The right suprarenal dissection allows for removal of the right perihilar nodes as well as the node-bearing tissue between the aorta and the vena cava. The pro-

cedure for excision of the right perihilar nodes is similar to that described above. Attention is then turned to removal of the aortocaval nodes. This nodebearing tissue is bounded by the IVC and aorta laterally, the crus of the diaphragm superiorly and left renal vein inferiorly. The incision in the adventitial sheath of the left renal vein is extended superiorly on the IVC and the nodal tissue mobilized medially. The aortocaval nodes are then exposed with lateral retraction of the IVC and inferior retraction of the left renal vein. The node pad is further freed from the left side of the cava and right side of the aorta. The attachments of the nodal tissue to the crus of the diaphragm

183

Ch 06 final.qxd

12/18/2006

15:19

Page 184

SURGERY FOR OVARIAN CANCER

Crus of diaphragm Right suprahilar nodes

Left suprahilar nodes

Figure 6.12 Boundaries of suprarenal aortic node dissection. The dissection begins in the region of the left renal vein (see text)

are sharply dissected and the node pad removed. The inferior end of the thoracic duct, the cisterna chili, lies to the right of the aorta at the level of T12. Damage to this lymphatic channel should be prevented, but if injury is unavoidable, both proximal and distal ends should be ligated.

Retroaortic lymphadenectomy Access to the operative field for the removal of retroaortic and retrocaval nodes is gained by opening the peritoneum as described above. The lymphatics of interest lie behind the great vessels and ventral to the vertebral bodies and psoas muscles. Retroaortic dissection usually requires identification and ligation of the lumbar and middle sacral vessels utilizing the classic description of the ‘split and roll’ technique (Figure 6.13). The procedure is initiated by incising the fascial sheaths overlying the vena cava and aorta. Each sheath is sharply dissected and mobilized to the lateral sides of the vessels. This effectively separates the nodal tissue off the anterior surface of the vessels to the lateral sides of the aorta and cava and into an

184

aortocaval bundle. Attention is then turned to the lumbar vessels. The lumbar arteries arising from the left posterolateral surface of the aorta are then secured at their origin from the aorta and near their entrance into the vertebral foramina. The vessels are then divided. If the lymphadenectomy is performed near the aortic bifurcation, the unpaired middle sacral artery, which arises from the posterior aspect of the aorta proximal to the bifurcation, should be secured and divided. The lumbar veins from the right posterolateral aspect of the IVC are then ligated and transected in a manner similar to that for the arteries. The middle sacral vein most commonly empties into the left common iliac vein. It can be sacrificed if it is in the field of dissection. The node pads are then elevated off the psoas and vertebrae and removed. Small perforating vessels should be clipped as needed. The final phase of the dissection consists of removal of the aortocaval lymphatics. The node pad is effectively isolated by mobilizing the adventitial sheaths of the aorta and cava medially. The aorta is reflected laterally while the left lumbar veins are identified, ligated and divided. The IVC is then retracted

Ch 06 final.qxd

12/18/2006

15:19

Page 185

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

Nodes

Anterior split

Lateral roll

Vena cava

a

b Lumbars divided

d

Great vessels isolated and controlled

e

c Specimen rolled off

f

g

Figure 6.13 Axial view of the aorta and inferior vena cava illustrating the classic ‘split and roll’ approach to retroaortic and retrocaval lymph nodes. (a) Undissected nodal basin. (b) An anterior split of the nodal bundle is created on the ventral surface of both the vena cava and the aorta. (c) Each nodal bundle is rolled laterally to expose the ipsilateral lumbar vessel. (d) The central nodal bundle is rolled medially, off each vessel, exposing the medial surface and its associated lumbar vessel. (e) After division of the lumbar vessels, the aorta and vena cava are mobilized anteriorly, exposing the retroaortic and retrocaval nodes. (f) The nodal bundles are removed from the anterior spinous ligaments and lumbar foramina. (g) Completed dissection

laterally to allow for the division of the right lumbar arteries. The nodal bundle is then elevated and removed.

Cytoreduction of bulky retroperitoneal nodal disease Occasionally, a bulky mass of enlarged, matted retroperitoneal nodes will be encountered at the time of primary or secondary cytoreductive surgery (Figure 6.14). The initial approach to gain exposure to the retroperitoneum is similar to that described above. Several general principles apply to the cytoreduction of a bulky retroperitoneal mass. First, the regional vessels and adherent vital structures must be identified. The most common area, a bulky nodal mass, is encountered below the level of the renal veins. The course of both the ureters should be identified. The dissection is begun by identifying and freeing the com-

mon iliac vessels at the distal edge of the mass. The renal veins should then be identified, as the dissection proceeds from caudad to cephalad (Figure 6.15). The nodal mass should be separated from the vessel’s adventitial sheath and any smaller tributaries secured with sutures or clips. If any loops of large or small bowel are adherent to the mass, these should be mobilized and cleared from the field. Infrequently the mass will encase the lower portions of the great vessels (Figure 6.16a). Prior to extirpating the retroaortic, retrocaval or aortocaval nodes, the portion of the mass anterior to the vessels should be resected to facilitate exposure. Within the tumor, smaller nodal masses can often be identified. The capsule of the tumor mass can be incised and these areas enucleated, similar to the performance of a myomectomy. If surgical planes cannot be identified for enucleation, the mass can be divided and mobilized off the vessels.

185

Ch 06 final.qxd

12/18/2006

15:19

Page 186

SURGERY FOR OVARIAN CANCER

a

b

Figure 6.14 Bulky para-aortic adenopathy from metastatic ovarian cancer. (a) Left-sided infrarenal para-aortic node (from top to bottom vasa-loops identify the left renal vein, inferior mesenteric artery and the left ureter). (b) Completed resection of nodal mass from (a)

Next, the nodal mass should be mobilized laterally off the aorta and vena cava. As described above, the adventitial sheaths of these vessels can be entered (‘split’), a plane identified and the nodal masses mobilized (‘rolled’) laterally (Figure 6.16b). The dissection can begin on either side, whichever provides the most satisfactory exposure, and is continued to the contralateral side. This will often require division of the lumbar vessels as they arise from the posterolateral surfaces of the great vessels. The final phase of the dissection involves sweeping the nodal mass from the lateral surface of the aorta or vena cava and the associated anterior spinous ligaments. Division of the lumbar vessels, as they arise from the posterolateral surfaces of the great vessels, will often be required for sufficient mobilization of the nodal mass for removal. If necessary, the lumbar vessels should be doubly ligated and transected (Figure 6.16d). Very rarely, tumor invades the wall of the cava or aorta. The approach to this situation is described below.

Perioperative and postoperative morbidity Combining data from several reports, di Re and Baiocchi found that lymphadenectomy for ovarian

186

cancer was relatively well tolerated.36 The most common perioperative complication was vascular injury, occurring in 3.9% of the patients. The perioperative mortality rate was 0.6%, occurring in three of 488 patients. Less frequent complications included nerve injuries, ureteral injury and intestinal injury.36 The inclusion of lymphadenectomy to cytoreductive surgery increased the operative time by 100 min in one report.33 The most frequent postoperative complication after systematic lymphadenectomy is lymphocyst formation, seen in 13.5% of cases. Thromboembolic phenomena – deep vein thrombosis and pulmonary embolism – occurred in 5% and 2.8% of the patients, respectively.15

Vascular complications Retroperitoneal lymphadenectomy for ovarian cancer requires operating in close proximity to multiple vascular structures, and therefore the threat of intraoperative hemorrhage is ever-present. The potential for vascular damage can be minimized by providing adequate operative exposure and carefully defining the vascular anatomy of the region. As described above, anatomic variations of the retroperitoneum,

Ch 06 final.qxd

12/18/2006

15:19

Page 187

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

a

b

c

d

Figure 6.15 Technique of cytoreduction of enlarged infrarenal para-aortic adenopathy. (a) The nodal mass is widely exposed by reflecting the duodenum cephalad, with identification of the vena cava and left renal vein (pickups). (b) The nodal capsule is ‘split’ laterally and the mass ‘rolled’ medially. (c) The medial surface of the nodal capsule is incised medially and the mass ‘rolled’ laterally off the aorta. (d) The nodal mass is dissected from the adventitia of the aorta, moving in a cephalad direction over the left renal vein

particularly near the hilus of the kidney, are frequent. If these genitourinary and vascular anomalies are not recognized, the likelihood of an intraoperative complication is greatly increased. The most frequent locations for vascular injuries to occur are the common iliac veins and the lower vena cava. If a major vascular injury occurs, the initial step is the application of pressure to the area. This can be done with the surgeon’s finger or a spongestick. After pressure is applied, the field should be manipulated so that maximal exposure to the region can be achieved. Prior to proceeding, the surgeon should verify that

sufficient suction is available to keep the operative field clear and that intravenous access is adequate for the transfusion of blood products if needed. If a small (< 5 mm) injury occurs to a major artery or vein, the area can be oversewn with interrupted or figure-ofeight stitches of fine (5-0) monofilament (polypropylene) suture on a vascular needle. The repair should be directed perpendicular to the long axis of the vessel so as to avoid constricting the lumenal diameter. A variety of vascular clamps are available and can be applied proximally and distally to keep the operative field clear and facilitate the repair (Figure 6.17). Large

187

Ch 06 final.qxd

12/18/2006

15:19

Page 188

SURGERY FOR OVARIAN CANCER

a

b

c

d

Figure 6.16 Cytoreduction of para-aortic adenopathy. (a) Bulky nodal disease may envelop the aorta and inferior vena cava. (b) The nodal mass is rolled laterally and a fine-tipped clamp is introduced into the adventitial space around the vena cava and ‘split’ with electrocautery. (c) The nodal mass is similarly rolled off the aorta, dissecting and ‘splitting’ the adventitial sheath moving medial to lateral. (d) The nodal mass is dissected from the anterior spinous ligament and the lumbar vessels ligated and divided as necessary

arterial or venous injuries should be repaired by a vascular surgeon and may require grafting. Injury to the vena cava during dissection can lead to massive hemorrhage. Pressure should be applied to the defect to control bleeding and enhance exposure to the area. Vascular clamps or vasa-loops can then be applied proximally and distally and to facilitate suturing of the defect. Rarely, the vena cava will be encased with tumor or invaded by tumor thrombus. In this situation consideration can be given to vena caval resec-

188

tion.40,41 In a series of 500 patients with testicular carcinoma undergoing retroperitoneal lymphadenectomy, resection of the vena cava was required in 45 patients.40 The lumbar vessels below the renal veins are secured and a vascular stapling clamp is fired across the cava and the segment removed. Postoperatively, ascites is the most common morbidity.40 The management of pelvic vascular injury is addressed in Chapter 5.

Ch 06 final.qxd

12/18/2006

15:19

Page 189

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

Extra-abdominal lymphadenopathy

Inguinal adenopathy

Metastatic spread of ovarian cancer to extra-abdominal lymph nodes is rare. These patients often present with large-volume intra-abdominal disease. While data are limited, surgical resection should be considered if the patient is symptomatic or if the extraabdominal adenopathy is the patient’s only site of residual disease. Because of the rarity of extra-abdominal nodal spread, surgical management must be individualized.

Metastasis to the inguinal lymph nodes from ovarian cancer is rare. Inguinal metastases were noted in 3% of patients in an autopsy series.5 Several reports of women with inguinal nodal disease have been reported, including a patient who presented with an inguinal mass and was found to have negative pelvic and para-aortic nodes.42–44 Figure 6.18 displays the anatomy of the inguinal region. The area of dissection is enclosed within the femoral triangle. The femoral triangle is bounded by the inguinal ligament, the sartorius and medially by the adductor longus. The superficial inguinal nodes lie ventral to the cribriform fascia while the femoral nodes lie deep to the fascia. The femoral artery courses laterally to the femoral vein. The femoral nerve lies on the lateral side of the artery while the majority of the lymphatics are medial to the vein. Operative exposure to the inguinofemoral nodes is gained through a linear incision between the anterior superior iliac spine and the pubis. There is no indication for a complete inguinofemoral lymphadenectomy as is performed for vulvar cancer. Instead, any grossly enlarged nodes should be excised. The saphenous

Figure 6.17 Vascular clamps. Top to bottom: Cooley clamp, Satinsky clamp, DeBakey bulldog clamps (straight and curved)

Inguinal ligament

Iliopsoas muscle Superficial inguinal nodes Skin and subcutaneous tissues Fascia lata

Femoral nerve Femoral vein and artery Adductor longus muscle

Fossa ovalis

Pectineus muscle Femoral lymph nodes

Figure 6.18 Anatomy of the inguinofemoral lymph node basins

189

Ch 06 final.qxd

12/18/2006

15:19

Page 190

SURGERY FOR OVARIAN CANCER

vein arises from the femoral vein and can be sacrificed if necessary. Consideration should be given to the placement of a closed suction drain. The most common complication from inguinal lymphadenectomy is lower extremity lymphedema. Postoperative lymphedema can be reduced by limiting the extent of nodal dissection and preservation of the saphenous vein.

Supraclavicular adenopathy Bulky supraclavicular adenopathy is occasionally seen in patients with ovarian cancer. Petru et al. performed pretherapeutic scalene biopsies in 32 patients with advanced-stage ovarian cancer. The scalene nodes were positive in 22% of the patients. Only one subject had a palpable mass. The para-aortic nodes were involved by disease in 75% of the subjects with positive scalene nodes who underwent retroperitoneal node evaluation.45 The anatomy of the supraclavicular region is displayed in Figures 6.19 and 6.20. The posterior triangle of the neck is bounded by the sternocleidomastoid, the trapezius and the clavicle. The external jugular vein courses through the triangle. The triangle is entered through a 5–6 cm incision made one finger breadth above the clavicle. The space is entered and

Sternocleidomastoid muscle sternal head clavicular head Clavicle Line of incision

the sternocleidomastoid retracted medially to expose the internal jugular vein. The omohyoid muscle is identified as it crosses the internal jugular. The fat pad below the omohyoid is gently cleared. After exposing the area, any bulky nodes can be extirpated. The nodes are elevated off the omohyoid and freed toward the clavicle. Care should be taken to avoid the brachial plexus at the lateral border of the omohyoid and the phrenic nerve, which lies on the scalene muscle. The thoracic duct enters the internal jugular vein above the clavicle and should be avoided. The transverse scapular vessels can be sacrificed as needed. The nodes are then removed off the major veins of the area. Potential complications include vascular or lymphatic injury which can lead to a hematoma or chylous fistula. Other complications include nerve injuries and wound complications.

CONCLUSION The retroperitoneum is a vital anatomic site in the management of ovarian cancer. Thorough understanding of the anatomy is necessary to facilitate full surgical staging in apparent early-stage disease including adequate resection of the para-aortic lymph nodes

Trapezius muscle External jugular vein Subclavian triangle Transverse cervical vein Subscapular vein Anterior jugular vein

Figure 6.19 Anatomy of supraclavicular node dissection. The incision is made one finger breadth above the clavicle

190

Ch 06 final.qxd

12/18/2006

15:20

Page 191

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

7.

Onda T, Yoshikawa H, Yokota H, et al. Assessment of metastases to aortic and pelvic lymph nodes in epithelial ovarian carcinoma. A proposal for essential sites for lymph node biopsy. Cancer 1996; 78: 803–8

8.

Burghardt E, Girardi F, Lahousen M, et al. Patterns of pelvic and paraaortic lymph node involvement in ovarian cancer. Gynecol Oncol 1991; 40: 103–6

9.

Burghardt E, Pickel H, Lahousen M, Stettner H. Pelvic lymphadenectomy in operative treatment of ovarian cancer. Am J Obstet Gynecol 1986; 155: 315–19

10.

Wu PC, Qu JY, Lang JH, et al. Lymph node metastasis of ovarian cancer: a preliminary survey of 74 cases of lymphadenectomy. Am J Obstet Gynecol 1986; 155: 1103–8

11.

Carnino F, Fuda G, Ciccone G, et al. Significance of lymph node sampling in epithelial carcinoma of the ovary. Gynecol Oncol 1997; 65: 467–72

12.

Chen SS, Lee L. Incidence of para-aortic and pelvic lymph node metastases in epithelial carcinoma of the ovary. Gynecol Oncol 1983; 16: 95–100

13.

Flanagan CW, Mannel RS, Walker JL, Johnson GA. Incidence and location of para-aortic lymph node metastases in gynecologic malignancies. J Am Coll Surg 1995; 181: 72–4

14.

Morice P, Joulie F, Camatte S, et al. Lymph node involvement in epithelial ovarian cancer: analysis of 276 pelvic and paraaortic lymphadenectomies and surgical implications. J Am Coll Surg 2003; 197: 198–205

15.

di Re F, Baiocchi G, Fontanelli R, et al. Systematic pelvic and paraaortic lymphadenectomy for advanced ovarian cancer: prognostic significance of node metastases. Gynecol Oncol 1996; 62: 360–5

16.

Eisenkop SM, Spirtos NM. The clinical significance of occult macroscopically positive retroperitoneal nodes in patients with epithelial ovarian cancer. Gynecol Oncol 2001; 82: 143–9

17.

Zanetta G, Chiari S, Barigozzi P, et al. Limited invasiveness to assess retroperitoneal spread in stage I–II ovarian carcinoma. Int J Gynaecol Obstet 1995; 51: 133–40

18.

Tsuruchi N, Kamura T, Tsukamoto N, et al. Relationship between paraaortic lymph node involvement and intraperitoneal spread in patients with

Figure 6.20 Excision of metastatic ovarian cancer in a supraclavicular lymph node

bilaterally. In advanced-stage ovarian cancer, complete resection of retroperitoneal structures may be necessary to confer an optimally cytoreduced state, which is associated with an improved overall survival.

REFERENCES 1.

Staging announcement: FIGO cancer committee. Gynecol Oncol 1986; 25: 383

2.

Knipscheer RJ. Para-aortal lymph nodes dissection in 20 cases of primary epithelial ovary carcinoma stage I (FIGO): influence on staging. Eur J Obstet Gynecol Reprod Biol 1982; 13: 303–7

3.

Knapp RC, Friedman EA. Aortic lymph node metastases in early ovarian cancer. Am J Obstet Gynecol 1974; 119: 1013–17

4.

Walter AJ, Magrina JF. Contralateral pelvic and aortic lymph node metastasis in clinical stage I epithelial ovarian cancer. Gynecol Oncol 1999; 74: 128–9

5.

Abrams HL, Goldstein N. Metastases in carcinoma. Analysis of 1000 autopsied cases. Cancer 1950; 3: 74–85

6.

Bergman F. Carcinoma of the ovary. A clinicopathological study of 86 autopsied cases with special reference to mode of spread. Acta Obstet Gynecol Scand 1966; 45: 211–31

191

Ch 06 final.qxd

12/18/2006

15:20

Page 192

SURGERY FOR OVARIAN CANCER

study of 229 patients. Proc Soc Gynecol Oncol 2004: abstr 48

ovarian cancer – a multivariate analysis. Gynecol Oncol 1993; 49: 51–5 19.

Tsumura N, Sakuragi N, Hareyama H, et al. Distribution pattern and risk factors of pelvic and para-aortic lymph node metastasis in epithelial ovarian carcinoma. Int J Cancer 1998; 79: 526–30

20.

Benedetti-Panici P, Maneschi F, Scambia G, et al. Anatomic abnormalities of the retroperitoneum encountered during aortic and pelvic lymphadenectomy. Am J Obstet Gynecol 1994; 170: 111–16

21.

Young RC, Decker DG, Wharton JT, et al. Staging laparotomy in early ovarian cancer. JAMA 1983; 250: 3072–6

22.

Petru E, Lahousen M, Tamussino K, et al. Lymphadenectomy in stage I ovarian cancer. Am J Obstet Gynecol 1994; 170: 656–62

23.

Sakurai SY, Utsugi K, Umezawa S, et al. Validity of complete paraaortic and pelvic lymphadenectomy in apparent stage I (pT1) ovarian cancer. Proc Am Soc Clin Oncol 2002: abstr 801

31.

Lin PS, Gershenson DM, Bevers MW, et al. The current status of surgical staging of ovarian serous borderline tumors. Cancer 1999; 85: 905–11

32.

Spirtos NM, Gross GM, Freddo JL, Ballon SC. Cytoreductive surgery in advanced epithelial cancer of the ovary: the impact of aortic and pelvic lymphadenectomy. Gynecol Oncol 1995; 56: 345–52

33.

Scarabelli C, Gallo A, Zarrelli A, et al. Systematic pelvic and para-aortic lymphadenectomy during cytoreductive surgery in advanced ovarian cancer: potential benefit on survival. Gynecol Oncol 1995; 56: 328–37

34.

Kigawa J, Minagawa Y, Itamochi H, et al. Retroperitoneal lymphadenectomy, including the para-aortic nodes in patients with stage III ovarian cancer. Am J Clin Oncol 1994; 17: 230–3

35.

Saygili U, Guclu S, Uslu T, et al. Does systematic lymphadenectomy have a benefit on survival of suboptimally debulked patients with stage III ovarian carcinoma? A DEGOG* Study. J Surg Oncol 2002; 81: 132–7

24.

Suzuki M, Ohwada M, Yamada T, et al. Lymph node metastasis in stage I epithelial ovarian cancer. Gynecol Oncol 2000; 79: 305–8

25.

Cass I, Li AJ, Runowicz CD, et al. Pattern of lymph node metastases in clinically unilateral stage I invasive epithelial ovarian carcinomas. Gynecol Oncol 2001; 80: 56–61

36.

di Re F, Baiocchi G. Value of lymph node assessment in ovarian cancer: status of the art at the end of the second millennium. Int J Gynecol Cancer 2000; 10: 435–42

26.

Tangjitgamol S, Manusirivithaya S, Sheanakul C, et al. Can we rely on the size of the lymph node in determining nodal metastasis in ovarian carcinoma? Int J Gynecol Cancer 2003; 13: 297–302

37.

27.

Benedetti-Panici P, Greggi S, Maneschi F, et al. Anatomical and pathological study of retroperitoneal nodes in epithelial ovarian cancer. Gynecol Oncol 1993; 51: 150–4

Kimball RE, Schlaerth JB, Kute TE, et al. Flow cytometric analysis of lymph node metastases in advanced ovarian cancer: clinical and biologic significance. Am J Obstet Gynecol 1997; 176: 1319–26, discussion 26–7

38.

Hacker NF. Systematic pelvic and paraaortic lymphadenectomy for advanced ovarian cancer – therapeutic advance or surgical folly? Gynecol Oncol 1995; 56: 325–7

39.

Benedetti-Panici P, Maneschi F, Cutillo G. Pelvic and aortic lymphadenectomy. Surg Clin North Am 2001; 81: 841–58

40.

Donohue J. Surgical management of testicular cancer. In Ernstoff M, Heaney JA, Peschel RE, eds. Urologic Cancer. Cambridge: Blackwell Science, 1997: 554–81

41.

Pantuck AJ, Barone JG, Cummings KB. Vena cava resection during post chemotherapy lymphadenectomy for testis tumor. Am Surg 1995; 61: 424–6

28.

Winter WE 3rd, Kucera PR, Rodgers W, et al. Surgical staging in patients with ovarian tumors of low malignant potential. Obstet Gynecol 2002; 100: 671–6

29.

Leake JF, Rader JS, Woodruff JD, Rosenshein NB. Retroperitoneal lymphatic involvement with epithelial ovarian tumors of low malignant potential. Gynecol Oncol 1991; 42: 124–30

30.

Rao GS, Skinner E, Gehrig PA, et al. Is staging of ovarian LMP tumors worth the effort? A multicenter

192

Ch 06 final.qxd

12/18/2006

15:20

Page 193

CYTOREDUCTIVE SURGERY: ABDOMINAL RETROPERITONEUM AND ADENOPATHY

42.

Scholz HS, Lax S, Tamussino KF, Petru E. Inguinal lymph node metastasis as the only manifestation of lymphatic spread in ovarian cancer: a case report. Gynecol Oncol 1999; 75: 517–18

44.

McGonigle KF, Dudzinski MR. Endometrioid carcinoma of the ovary presenting with an enlarged inguinal lymph node without evidence of abdominal carcinomatosis. Gynecol Oncol 1992; 45: 225–8

43.

Kehoe S, Luesley D, Rollason T. Ovarian carcinoma presenting with inguinal metastatic lymphadenopathy 33 months prior to intraabdominal disease. Gynecol Oncol 1993; 50: 128–30

45.

Petru E, Pickel H, Tamussino K, et al. Pretherapeutic scalene lymph node biopsy in ovarian cancer. Gynecol Oncol 1991; 43: 262–4

193

Ch 07 final.qxd

12/18/2006

15:24

Page 195

CHAPTER 7

Cytoreductive surgery: intestinal tract and omentum Yukio Sonoda, Dennis S Chi, Richard R Barakat

INTRODUCTION In 1931, Sampson described the peritoneal dissemination of ovarian cancer as follows: (1) escape of the cancer cells from the primary ovarian tumor into the free peritoneal cavity; (2) migration of these cells to their site of implantation; (3) reaction of the peritoneal surface injured by the cancer cells so that fibrin and organization of this fibrin occur; and (4) progression of the cancerous implant at that site.1 The clockwise motion of the peritoneal fluid in the abdominal cavity, driven by intestinal peristalsis and diaphragmatic excursion during respiration, accounts for the characteristic distribution pattern of ovarian cancer. Exfoliated tumor cells have a tendency to implant on fixed structures and in areas where resorption of peritoneal fluid occurs, such as the under-surfaces of the diaphragms and omentum. These cells tend to follow the circulatory path of the peritoneal fluid and implant along the surfaces of the peritoneal cavity, especially those of the intestines and their mesenteries as well as the omentum. Intestinal involvement by ovarian cancer is not uncommon, and although there is a tendency to spread along mesenteric peritoneal and bowel serosal surfaces, it can also invade directly into the visceral wall. Wu et al. reported that ovarian cancer invaded into the intestinal wall in 38% of specimens examined, and 21% of specimens actually had full-thickness invasion to the mucosa.2 This type of spread pattern frequently results in the need for bowel surgery as part

of the initial cytoreductive effort. In a review of intestinal resections performed on a gynecologic oncology service, ovarian cancer patients accounted for nearly 40% of cases.3 This chapter will review the intra-abdominal intestinal procedures that the surgeon may be called upon to perform when operating on a patient with ovarian cancer.

Indications for omentectomy and intestinal resection Omentectomy is typically performed at initial surgery, since it is a routine part of the staging procedure for this disease and is a common site of tumor implantation due to its high fat content and rich blood supply. In the usual case of advanced-stage ovarian cancer, innumerable omental tumor implants coalesce into a bulky tumor mass – the so-called omental cake – removal of which is an integral component of optimal cytoreductive surgery. There are two main indications for performing an intestinal resection for ovarian cancer: (1) to achieve an optimal volume of residual disease; and (2) to relieve obstruction. The management of bowel obstruction is addressed in Chapter 13. Intestinal resection performed as part of tumor reductive surgery is by far the most common indication for bowel resection at the time of primary ovarian cancer surgery.4 The concept of ‘optimal’ cytoreduction stems from Griffiths’ sentinel paper, which was the first to demonstrate conclusively a survival advantage for patients left with small-volume residual disease.5

195

Ch 07 final.qxd

12/18/2006

15:24

Page 196

SURGERY FOR OVARIAN CANCER

Although some have questioned the impact of bowel resection on survival in patients with ovarian cancer, the survival advantage after optimal debulking appears to be evident despite the fact that patients require a bowel resection.4,6,7 The Mayo Clinic experience with gastrointestinal surgery performed on patients with ovarian cancer, reported by Tamussino et al., is representative of the body of literature on this topic.8 At primary surgery, tumor cytoreduction was the indication for 87% of bowel resections; the remaining resections were performed for bowel obstruction. The most common complications were febrile morbidity and ileus, which occurred in 28% and 21% of cases, respectively. Anastomotic leak and abscess requiring laparotomy each occurred in less than 1% of patients. In general, approximately 20% of patients undergoing initial cytoreductive surgery for ovarian cancer will require an extrapelvic bowel resection.9

REGIONAL ANATOMY Small intestine The small intestine as a whole extends from the pylorus to the cecum and consists of the duodenum, jejunum and ileum. This chapter will focus on the latter two components of the small intestine, as surgery on the duodenum is uncommonly required as part of primary tumor cytoreduction. Together, the jejunum and ileum measure approximately 6–7 m in length, with the jejunum comprising the proximal 40% and the ileum accounting for the distal 60%. There is no sharp morphological demarcation between the jejunum and ileum; however, there are certain anatomic characteristics that aid the surgeon in distinguishing between the two: (1) The jejunum has a thicker wall, owing to the fact that the circular folds, or plicae circulares, are larger and more well developed in the proximal end of the small bowel; these folds are small

196

in the superior part of the ileum and absent in the terminal ileum. (2) The jejunum is of greater diameter. (3) The jejunal mesentery contains less fat than does that of the ileum, and the arterial arcades are easier to visualize than in the ileum. (4) The jejunal arterial arcades are simple or double with long vasa recti compared with the ileum that has four or five arcades and shorter vasa recti. Vasculature of the small intestine

The superior mesenteric artery (SMA) provides the entire blood supply to the jejunum and ileum (Figure 7.1). It arises from the aorta at the level of the first lumbar vertebra and lies behind the neck of the pancreas; it then passes over the uncinate process and anterior to the third part of the duodenum, before entering the root of the small bowel mesentery. The common inferior pancreaticoduodenal artery is the first branch given off and supplies the retroperitoneal duodenum. The remaining small intestine is supplied by the jejunal–ileal arteries. These consist of 10–16 branches which arise from the concave side of the SMA. These branches extend into the mesentery, where they form arcades. From these arcades, the vasa recti arise and pass to the mesenteric border of the bowel without anastomosing with one another. The avascular spaces are called the windows of Deaver. The vasa recti continue to form the subserosal plexus. These are sufficient to supply 6–8 cm of small intestine if the adjacent vasa recti have been occluded or ligated. The SMA also gives rise to the middle colic, right colic and ileocolic arteries, arising from the right side of the SMA, which supply the transverse colon, ascending colon and cecum, respectively. The space of Treves is an avascular space between the SMA and ileocolic artery, which may result in an inconsistent blood supply to the terminal ileum (Figure 7.1). Venous drainage of the small intestine is composed of direct tributaries that correspond to the branches of

Ch 07 final.qxd

12/18/2006

15:24

Page 197

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

Superior mesenteric artery Common inferior panocreaticoduodenal artery (with anterior and posterior branches) Middle colic artery Vasa recta Right colic artery Ileocolic artery Ascending branch of ileocolic artery

Avascular space of Treves

Anterior and posterior cecal arteries

Appendiceal artery

Figure 7.1 Anatomy and arterial supply of the small intestine. The superior mesenteric artery and its branches supply the entire jejunum and ileum, and a portion of the colon

the SMA to form the superior mesenteric vein. This joins the splenic vein to form the portal vein. Lymphatic drainage of the small intestine begins in the lacteals of the mucosa villi. These join in the bowel wall and drain through lymphatic channels that follow the veins. The drainage route begins with the mesenteric lymph nodes, which drain to the superior mesenteric lymph nodes and the left lumbar lymphatic trunk, ending in the cisterna chyli.

Large intestine The large intestine is approximately 1.5 m in length, beginning at the cecum and ending with the anus at the perineum (Figure 7.2). This chapter will primarily focus on the abdominal colon, including the cecum, ascending colon, transverse colon and descending colon. Surgery on the pelvic colon is addressed in Chapter 5. The large bowel is characterized by fullthickness infoldings of the bowel wall, which form sacculations called haustra. These infoldings correspond to transverse folds in the bowel lumen called

plicae semilunares. The large bowel has three thickened bands of longitudinal muscle that run its length from the appendix to the rectum called taenia coli (taenia omentalis, taenia libera and taenia mesocolica). Contraction of the taeniae causes the haustra to become more prominent. The large bowel has small pouches of peritoneum filled with fat called appendices epiploicae, which are most prominent on the descending and sigmoid colon. The cecum is the most proximal portion of the large intestine and is a blind pouch that is 5–7 cm in length and projects caudal to the ileocecal junction in the iliac fossa of the right lower quadrant. Usually, the cecum is entirely enveloped by peritoneum and does not have a mesentery. It is the widest part of the large intestine but also has the thinnest wall; thus, it is at risk for perforation in cases of large-bowel obstruction. The appendix arises 2–3 cm inferior to the ileocecal junction. The appendix has its own short mesentery called the mesoappendix, which connects it to the inferior part of the mesentery of the ileum.

197

Ch 07 final.qxd

12/18/2006

15:24

Page 198

SURGERY FOR OVARIAN CANCER

Transverse mesocolon

Taenia omentalis Taenia libera Ascending colon Plicae semilunares

Transverse colon Descending colon

Taenia mesocolica Haustra

Sigmoid mesocolon Appendices epiploicae

Cecum

Sigmoid colon

Appendix Rectum

Figure 7.2 Anatomy of the abdominal colon

The appendiceal artery is a branch off the ileocolic artery and runs in the lateral margin of the mesoappendix. The ascending colon is approximately 15–20 cm long and extends from the ileocecal valve to the hepatic flexure. It ascends as a retroperitoneal structure, covered by peritoneum only on its anterior and lateral surfaces. It lies anterior to the quadratus lumborum, psoas and transversus abdominus muscles, inferior pole of the right kidney, and descending portion of the duodenum. Lateral to the ascending colon is the white line of Toldt, which represents the fusion of the colonic mesentery with the parietal peritoneum. There may be congenital adhesions between the anterior aspect of the ascending colon and the right abdominal wall (Jackson’s membrane). The hepatic flexure may have several attachments to the liver and gallbladder (hepatocystocolic ligament). The ascending colon and hepatic flexure are supplied by the ileocolic and right colic arteries, and the venous drainage is through the corresponding ileocolic and right colic veins, which drain into the superior mesenteric vein. The lymphatic drainage of the

198

ascending colon is via the paracolic and epicolic lymph nodes, which empty into the superior mesenteric lymph nodes. The transverse colon lies between the hepatic and splenic flexures and is the longest portion of the large bowel, usually measuring 30–60 cm in length. Occasionally, a redundant transverse colon will reach into the pelvis. Unlike the ascending and descending colon, the transverse colon has its own mesentery, which is longest in the center, and is therefore considered an intraperitoneal structure. The root of the transverse colon mesentery covers the descending part of the duodenum, the pancreas and a portion of the left kidney. At the hepatic and splenic flexures, the mesentery is very short and may place the transverse colon in contact with the duodenum and the head of the pancreas, which may be injured during mobilization of the hepatic flexure. The splenic flexure is connected to the diaphragm by the phrenocolic ligament and to the spleen by the lienocolic ligament. The transverse colon is attached to the greater curvature of the stomach by the cephalic portion of the greater omentum or gastrocolic ligament.

Ch 07 final.qxd

12/18/2006

15:24

Page 199

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

The descending colon is approximately 20–25 cm in length and begins at the splenic flexure and ends at the pelvic brim with the start of the sigmoid colon, which is demarcated by its intraperitoneal mesentery. Like the ascending colon, the descending colon is a retroperitoneal structure that is covered by peritoneum only on its anterior and lateral surfaces. Lateral to the descending colon is the white line of Toldt, which demarcates the correct plane to enter the retroperitoneal space when mobilizing the descending colon. The proximal part of the descending colon is attached to the peritoneum overlying the left kidney by the phrenocolic ligament. Vasculature of the large intestine

The entire colon is supplied by the superior mesenteric and inferior mesenteric arteries (Figure 7.3). Variations in the distribution pattern may exist. The ileocolic artery is the last branch of the SMA as it runs towards the cecum. It in turn divides into an ascending (colic) branch, which anastomoses with branches from the right colic artery, and the descending (ileal)

Marginal artery of Drummond Middel colic artery Right colic artery Ileocolic artery colic branch ileal branch

Appendiceal artery

branch, that anastomoses with the terminal SMA. The origin of the right colic artery may vary from person to person, arising from the SMA, middle colic artery, or ileocolic artery. The right colic artery may also be absent. It generally divides into an ascending branch, which anastomoses with the middle colic artery, and a descending branch, which anastomoses with the colic branch of the ileocolic artery. The SMA also gives rise to the middle colic artery which eventually divides into a right and left branch. These branches eventually anastomose with branches from the right and left colic arteries, respectively. Variations of the middle colic artery exist, and in a small number of cases the middle colic artery can be absent. The second major arterial supply to the colon is the inferior mesenteric artery (IMA), which arises from the abdominal aorta approximately 3–4 cm above the aortic bifurcation (Figure 7.3). The IMA gives rise to the left colic artery which in turn bifurcates to give rise to an ascending branch, which anastomoses with the left branch of the middle colic

Avascular area of Riolan Marginal artery of Drummond Superior mesenteric artery Left colic artery, ascending branch Inferior mesenteric artery Left colic artery, descending branch Left colic artery Sigmoid arteries

Superior rectal artery

Figure 7.3 The arterial supply to the colon. The superior mesenteric artery (SMA) provides the blood supply to the right half of the colon, and the inferior mesenteric artery supplies the left. The right colic artery may arise directly from the SMA, ileocolic, or middle colic artery, or it may be absent altogether

199

Ch 07 final.qxd

12/18/2006

15:24

Page 200

SURGERY FOR OVARIAN CANCER

artery, and a descending branch, which anastomoses with the sigmoid arteries. The sigmoid arteries also originate from the IMA and supply the sigmoid colon. The IMA continues into the pelvis as the superior rectal artery. The arterial supply to the abdominal colon has a collateral circulation provided by the marginal artery of Drummond. This artery consists of a series of arcades along the mesenteric border of the entire colon beginning at the ileocolic artery and running to the sigmoid arteries; thus, it connects the vasculature of the SMA and IMA. The marginal artery gives rise to the vasa recta which enter the colon wall and form intramural anastomoses. These intramural anastomoses are not as extensive as in the small bowel, and they can supply only 2–3 cm of colon compared with the 6–8 cm in the small intestine. In approximately 5% of patients the marginal artery is deficient in one or more of the following areas: the last 6–8 cm of the

Portal vein

terminal ileum, between the ileocolic and right colic arteries, between the middle and left colic arteries (termed the avascular area of Riolan), and between the superior rectal and last sigmoid arteries (Figure 7.3). The surgeon should be aware of these deficiencies, as a discontinuous blood supply may compromise the integrity of an intestinal anastomosis following resection. Identification of the major arteries prior to resection of a portion of the colon can help avoid the unfavorable situation of unrecognized vascular insufficiency. The venous drainage of the colon follows the course of the arterial system and is composed of the superior mesenteric vein and the inferior mesenteric vein (Figure 7.4). The inferior mesenteric vein empties into the splenic vein, which then joins the superior mesenteric vein to form the portal vein. The portal vein ascends to the liver behind the bile duct and hepatic artery at the free edge of the lesser omentum.

Splenic vein

Left gastroepiploic vein Middle colic vein Inferior pancreaticoduodenal vein Right colic vein Ileocolic vein

Superior mesenteric vein Left colic vein Inferior mesenteric vein Ileal and jejunal veins Sigmoid veins

Superior rectal vein

Figure 7.4 The venous drainage of the intestine. The ileum, jejunum and right portion of the colon drain into the superior mesenteric vein, which joins the splenic vein to form the portal vein. The left colon is drained by the inferior mesenteric vein, which ends in the splenic vein

200

Ch 07 final.qxd

12/18/2006

15:24

Page 201

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

At the porta hepatis, the portal vein divides into the left and right branches, which eventually empty into the hepatic sinusoids. The portal system has several anastomoses with the systemic venous system, which can compensate for obstructed portal venous return. The lymphatic system of the colon also follows the arterial supply previously described. Intramural lymphatic channels drain into extramural lymph vessels, which empty into the colonic lymph nodes.

Omentum The omentum can be categorized as the lesser omentum and the greater omentum (Figure 7.5). The lesser omentum connects the liver to the lesser curvature of the stomach and to the first 2 cm of the duodenum. It lies posterior to the left lobe of the liver and is attached to the liver in the fissure for the ligamentum venosum and to the porta hepatis. The lesser omentum is composed of a double-layered sheet of peri-

Hepatic artery

toneum, which may also be referred to by its attachments (i.e. the hepatogastric ligament and the hepatoduodenal ligament). The portal vein, hepatic artery and bile duct run between the layers of the hepatogastric ligament near its free edge. The portal vein typically lies posterior to the hepatic artery and common bile duct. The greater omentum is a double-layered fatty and vascular apron that hangs off the greater curvature of the stomach and the transverse colon. Each fatty layer has two peritoneal surfaces; thus, the greater omentum has four peritoneal surfaces, within which a potential space can exist. The greater omentum connects the greater curvature of the stomach to the transverse colon, spleen and diaphragm, but varies in length and fatty content from individual to individual. The gastrocolic ligament is the portion of the greater omentum between the stomach and the transverse colon. The gastrosplenic ligament lies to the left of

Lesser omentum

Liver

Left gastric artery

Portal vein Common bile duct in gastrohepatic ligament Gallbladder Epiploic foramen of Winslow Gastroduodenal artery

Stomach

Spleen Gastrosplenic ligament Right gastric artery Left gastroepiploic artery Gastrocolic ligament

Right gastroepiploic artery Right omental artery Middle omental artery Left omental artery Greater omentum

Figure 7.5 Blood supply for the omentum. The left and right gastroepiploic arteries provide the arterial supply for the greater omentum. The right and left gastric arteries supply the lesser omentum

201

Ch 07 final.qxd

12/18/2006

15:24

Page 202

SURGERY FOR OVARIAN CANCER

the gastrocolic ligament and connects the stomach to the spleen. Both the greater and lesser omentum, along with the stomach, caudate lobe of the liver and transverse mesocolon, form the anterior wall of the lesser sac. This is an anatomical space that lies anterior to the pancreas and the retroperitoneal portion of the duodenum. The epiploic foramen of Winslow is an anatomic window from Morison’s pouch to the lesser sac. The borders of this foramen are the edge of the hepatoduodenal ligament, the reflection of peritoneum off the liver onto the vena cava, the peritoneum overlying the vena cava and the duodenum. Entering the lesser sac is typically required to access the pancreas or to perform a total omentectomy. This may be most easily performed by freeing the gastrocolic ligament from the stomach or freeing the posterior leaf of the omentum from the transverse colon. Vasculature of the omentum

The blood supply to the greater omentum is from the multiple branches off the gastroepiploic arcade, which is formed by right and left gastroepiploic arteries. The right and left gastroepiploic arteries arise from the gastroduodenal and splenic arteries, respectively. The gastroepiploic arcade also gives rise to the right, middle and left omental arteries which run down the posterior leaf of the omentum.

(1) The intestine to be anastomosed should be healthy and have an adequate blood supply. (2) The anastomosis should be tension free. (3) The new lumen of the anastomosis should adequately allow for the passage of intestinal contents. (4) The anastomosis should be complete, secure and hemostatic. Hand-sewn anastomoses

Hand-sewn anastomoses can be divided into two categories: one-layer and two-layer. Variations in technique exist, including interrupted versus continuous closure, and the selection of delayed absorbable or permanent suture. The choice is primarily dependent on the surgeon’s preference. The traditional handsewn small-bowel anastomosis is performed using a two-layer inverting technique (Figure 7.6). This technique employs an inner layer using an absorbable suture (e.g. 3-0 polyglycolic acid suture) and an outer row of permanent sutures (e.g. 3-0 silk or polypropylene suture). The sequence of layer closure depends on the mobility of the portions of intestine to be anastomosed.

INTESTINAL ANASTOMOSIS Techniques of bowel anastomosis Intestinal anastomoses can be performed using one of two general techniques: hand-sewn or stapled. Although the introduction of intestinal staplers has simplified the creation of a bowel anastomosis, there are times when the hand-sewn technique may be indicated or useful, and the surgeon should be familiar with both techniques. Independently of the technique employed, there are several general principles that are necessary to ensure a successful anastomosis.

202

Figure 7.6 Hand-sewn two-layer closure

Ch 07 final.qxd

12/18/2006

15:24

Page 203

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

If the bowel cannot be rotated 180° to expose both the posterior and anterior surfaces, the sequence of intestinal re-approximation should be the posterior outer layer, posterior inner layer, anterior inner layer and finally the anterior outer layer. The bowel ends are approximated with bowel clamps and two stay sutures, incorporating the seromuscular layer, placed mid-way between the mesenteric and antimesenteric borders to aid in alignment. To facilitate inverting the bowel edges during closure, the stay sutures are places inside-out to outside-in. The two-layered anastomosis begins with a placement of a single row of interrupted imbricating (Lembert) stitches on the posterior wall through the seromuscular layer using 3-0 silk or polypropylene suture (Figure 7.7a). A technique of

serial bisection, in which sutures are placed before being tied, is the best approach. Once all the posterior sutures have been placed and tied, they can be cut. If present, the staple lines on the ends of divided bowel are excised. The bowel clamps are removed, and the inner mucosal layer is then completely reapproximated using a 3-0 delayed absorbable suture placed as a continuous locking stitch on the posterior mucosal edges and an inverting Connell stitch on the anterior mucosal edges (Figure 7.7b–c). Since this is a running stitch, narrowing of the bowel lumen should be prevented by avoiding undue tension on the suture line. Alternatively, the mucosal layer can be closed with inverting interrupted stitches with the knots placed on the mucosal side. The anastomosis is then

Continuous locking suture

Interrupted imbricating vertical mattress sutures

b

a

Outer layer imbricating sutures reinforcing anastomosis

c Continuation of posterior row as Connell suture for anterior anastomosis

d

Figure 7.7 The two-layer end-to-end anastomosis. (a) Interrupted Lembert seromuscular sutures are placed to form the posterior outer wall. (b) An inner running suture is placed on the posterior wall. (c) The inner running suture is continued on the anterior wall using a Connell stitch. (d) Outer layer sutures are placed on the anterior wall

203

Ch 07 final.qxd

12/18/2006

15:24

Page 204

SURGERY FOR OVARIAN CANCER

completed by placing a row of imbricating (Lembert) stitches on the anterior wall, using the same serial bisection technique to place all of the sutures prior to tying them (Figure 7.7d). If the bowel can be rotated 180° to expose both the anterior and posterior sides, the bowel segments are stabilized between non-crushing clamps and a stay suture placed at the mesenteric edge (Figure 7.8a). The inner layer is completely reapproximated around its entire circumference using interrupted 3-0 delayed absorbable sutures and taking full-thickness bites, tying the knots on the mucosal side (Figure 7.8b–c). A continuous stitch can also be used for the inner mucosal closure. The bowel is rotated as necessary to

a

c

improve exposure, and the outer layer is reapproximated using an interrupted Lembert stitch of either 3-0 silk or polypropylene suture (Figure 7.8d). Once the anastomosis is completed, the integrity of the lumen is checked by squeezing a small amount of air and liquid intestinal content through it while examining for leaks. The patency of the lumen is confirmed by pinching the anastomosis ring between the thumb and index finger to delineate the lumenal ring. Lastly, the mesentery must be re-approximated to prevent an internal herniation of bowel. This can be done with interrupted sutures or a continuous suture, with care taken not to injure any mesenteric vessels lying near the transected edges.

b

Mucosa

d

Serosa

Figure 7.8 Two-layer closure when bowel can be freely rotated. Small-bowel anastomosis, hand-sewn. (a) The bowel edges are approximated with a stay suture placed at the mesenteric border. (b) The back wall mucosal layer is closed using an interrupted stitch with knots placed in the bowel lumen. (c) The mucosal closure is continued onto the anterior wall. (d) The outer layer closure is completed using a seromuscular imbrication stitch. Inset, The mucosa–muscularis closure with 3-0 polyglycolic acid suture and knots placed in the bowel lumen is shown. The outer seromuscular layer of 2-0 or 3-0 silk imbricates the bowel wall with knots tied on the serosal surface

204

Ch 07 final.qxd

12/18/2006

15:24

Page 205

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

The one-layer closure has the advantages of decreasing the time required to form the anastomosis and producing a wider lumen, since less of the bowel edge is inverted. A permanent monofilament suture is generally used on the large bowel, while a delayed absorbable suture is preferable for a small-bowel anastomosis. First, the mesenteries must be aligned and two traction sutures placed midway between the mesenteric and antimesenteric borders on each side. These traction sutures are placed from inside-out to outside-in to help invert the intestinal edges. The traditional Gambee technique of one-layered closure consists of a series of interrupted inverting stitches of 3-0 suture starting on the posterior bowel wall and proceeding circumferentially. The suture picks up a small amount of mucosa and then a substantial portion of the seromuscular layer, which causes the edges to invert (Figure 7.9). Using this technique, all of the knots are tied in the lumen of the bowel, except for the final suture, which is a single inverting mattress suture placed with the knot on the serosal side of the bowel wall. Alternatively, a one-layer closure can be performed using a continuous stitch of 3-0 suture on

a double-armed needle (or two single-armed needles). The repair is initiated on the posterior wall at the mesenteric border by taking a full-thickness bite on each side; the suture is tied at equal length to prevent slippage (Figure 7.10). Each subsequent bite is advanced for a distance of 3 mm, incorporating a small edge of mucosa while passing tangentially through the submucosa and seromuscular layers to include 4–6 mm of the seromuscular wall. Each half suture proceeds as a continuous, non-locking stitch to close half of the posterior wall, and an inverting Connell stitch to navigate the angle and reapproximate half of the anterior wall. The two sutures are then tied in the midline of the antimesenteric border. Stapled anastomosis

In general, there are several advantages and disadvantages for a stapled anastomosis compared to the conventional hand-sewn closure, although the same surgical principles apply to both techniques. Staplers have been reported to decrease operative time and cause less fecal contamination.10 In addition, Wheeless and Smith have reported that staples may

a

Continuous suture in posterior row of anastomosis

b Connell suture in anterior anastomosis

Figure 7.10 One-layer continuous closure. If the bowel cannot be rotated, it may be easier to use a double-armed suture tied with equal length in the middle of the posterior wall. Gambee

(a) Each half is then used to close half the posterior wall with a

interrupted stitch incorporates the full wall thickness with the

continuous stitch. (b) Each suture is continued in an inverting

knot tied inside the lumen to invert the bowel edges

Connell stitch to close the half of the anterior wall

Figure 7.9 Gambee

one-layer

closure.

The

205

Ch 07 final.qxd

12/18/2006

15:24

Page 206

SURGERY FOR OVARIAN CANCER

be associated with a better blood supply to the anastomosis.11 These authors measured the flow of iodine125 through intestinal anastomoses in a canine model using one of three techniques (two-layer hand-sewn, single-layer hand-sewn and stapled). On postoperative day 4, the animals were re-explored and the iodine-125 flow was measured across the anastomosis. The stapled anastomoses had a significantly higher flow than the single- and two-layer anastomoses. Advantages to the hand-sewn anastomosis may be greater strength, reduced risk of stricture and more complete healing.12

Types of bowel anastomosis There are several methods to re-establish bowel continuity following resection of a portion of the intestinal tract. The procedures of end-to-end, end-to-side and side-to-side anastomosis are applicable to both the small intestine and the colon, and can be performed using either stapled or hand-sewn techniques. The general principles of each technique will be described briefly below, followed by a review of their specific clinical applications to ovarian cancer cytoreductive surgery.

teric border (Figure 7.8). Where there is disparity between the lumen size of the two segments to be reanastamosed, such as when approximating the ileum and the ascending colon, a Cheatle incision may be created on the antimesenteric surface of the smaller lumen (Figure 7.11). This will provide an increased diameter of lumen and will also allow a more even approximation of the two segments of bowel. The hand-sewn anastomosis can now be performed using a one- or two-layer closure as described above. Alternatively, the circular EEA stapler is a safe and expedient technique for end-to-end anastomosis (Figure 7.12). After the intestine has been divided, the anvil is placed in one end of the intestine and secured with a purse-string suture of 2-0 polypropylene placed by hand or using one of the automated purse-string application devices. The other limb of bowel should be closed with a linear stapler and the EEA stapler introduced through an enterotomy or colotomy made on the antimesenteric side several centimeters proximal to the linear staple line. The EEA trocar is advanced through the midportion of the linear staple line, the trocar removed from the

End-to-end anastomosis

The hand-sewn end-to-end anastomosis can be performed using an open or closed technique. The closed technique is rarely used in modern-day surgery. The open technique requires placement of crushing Kocher clamps immediately proximal and distal to the line of resection. Both bowel limbs are occluded by non-crushing bowel clamps, to prevent spillage of bowel contents, applied several centimeters away from the ends to be anastomosed. The mesentery beneath the area to be resected should be inspected to ensure that a dominant vascular pedicle is supplying the distal and proximal portions of the remaining small bowel. The bowel is divided and the Kocher clamps on the transected edges of the proximal and distal limbs removed. After careful alignment of the bowel to avoid twisting, the edges are approximated with a 3-0 silk stay suture placed at the antimesen-

206

a

b

Figure 7.11 Cheatle incision to accommodate a discrepancy in lumenal diameters To anastomose two bowel lumens of disparate sizes, an incision is made in the antimesenteric surface of the bowel with a smaller lumen (a). This enlarges the bowel lumen, which results in tension-free anastomosis (b)

Ch 07 final.qxd

12/18/2006

15:24

Page 207

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

shaft, and the main EEA instrument mated with the anvil. The EEA stapler is fired and removed, and the enterotomy or colotomy is closed using a TA stapler placed perpendicular to the long axis of the bowel. End-to-side anastomosis

a

b

c

d

The end-to-side anastomosis is typically used in ovarian cancer surgery after an ileocecal resection or right hemicolectomy in which the ileum is joined to the large bowel. The end-to-side anastomosis is especially useful when joining two portions of intestine with different lumenal diameters and can be performed using either a sutured or a stapled technique. A stapled end-to-side anastomosis can be performed using the circular EEA stapler (Figure 7.13a). In this technique, the anvil is placed in the distal terminal ileum and secured with a purse-string suture (Figure 7.13b). The main EEA stapler instrument is introduced into the open end of the colon and the trocar advanced through the antimesenteric wall. After the trocar is removed and the anvil is mated to the shaft of the main EEA instrument, the stapler is closed and fired, creating a double row of circular staples and simultaneously excising an internal ring of tissue from each bowel limb to complete the anastomosis (Figure 7.13c). The instrument is removed and inspected to confirm that two complete ‘donuts’ of bowel wall have been excised, ensuring a full-thickness anastomosis. The open end of the colon is then closed off using a TA stapler (Figure 7.13d). After an adequate lumen is confirmed, the mesenteric defect is closed.

Figure 7.12 Stapled end-to-end anastomosis. (a) With a purse-string on one end and the other stapled closed, the bowel is opened proximal to the staple line. (b) The EEA stapler without the anvil is introduced. The antimesenteric corner is excised or opened and a purse-string placed around this. (c) The center rod is advanced through the midportion of the linear staple line and the anvil attached. (d) The anvil is introduced into the other intestinal limb and the purse-strings are tied. (e) The EEA is fired to create the anastomosis, and the

e

stapler is removed. The remaining defect can be stapled closed with a TA stapler

207

Ch 07 final.qxd

12/18/2006

15:24

Page 208

SURGERY FOR OVARIAN CANCER

In the hand-sewn technique, the narrow-caliber bowel end is aligned perpendicular to the largercaliber bowel in an end-to-side fashion and secured with stay sutures. Proximal and distal bowel clamps are recommended to minimize spillage of intestinal contents. An incision is created on the antimesenteric

border of the larger-caliber bowel segment (i.e. colon) that will accommodate the circumference of the smaller segment (i.e. ileum). The anastomosis is completed using a one- or two-layered hand-sewn technique with 3-0 delayed absorbable or permanent monofilament suture, as previously described.

a

b

c

d

Figure 7.13 Stapled end-to-side anastomosis (for joining small to large bowel). (a) The EEA stapler is introduced through the open end of the larger-caliber bowel (colon) and the center rod advanced through the antimesenteric side. (b) The anvil is placed, and is introduced through the end of the small-caliber bowel limb (ileum) with a preplaced purse-string suture. (c) The purse-string suture is tied, and the stapler is fired and removed. (d) The open end of the colon is then closed with a TA linear stapler

208

Ch 07 final.qxd

12/18/2006

15:24

Page 209

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

Side-to-side functional end-to-end anastomosis

A side-to-side functional end-to-end anastomosis is commonly used to re-establish intestinal continuity following segmental resection (e.g. ileum to ileum, ileum to ascending colon). A true side-to-side anastomosis is usually only performed as a bypass procedure for relief of a bowel obstruction and is addressed in the chapter on palliative surgery (Chapter 13). The stapled technique of functional end-to-end anastomosis, using both the GIA and TA staplers, is as secure as the hand-sewn technique but probably faster and easier to perform for most surgeons. Stay sutures are placed to delineate the common lumen of the two segments to be joined, the ends of which have been previously closed with the linear stapler (Figure 7.14a). The two blind ends of intestine to be connected are again aligned side-by-side along their antimesenteric borders with stay sutures. The antimesenteric corners of the linear staple lines are trimmed and one leg of the GIA stapler is introduced into each lumen (Figure 14b). The stapler is locked and fired. After the suture line is inspected, the staple line is offset slightly to avoid intralumenal adhesion formation, and the remaining defect is closed with a TA stapler (Figure 7.14c–d). In the hand-sewn technique of functional end-toend anastomosis, the two segments of bowel are aligned side-by-side between stay sutures placed 8–10 cm apart along the antimesenteric borders to form an isoperistaltic intestinal cul-de-sac. Parallel linear incisions are created in each bowel segment between the stay sutures, and these will form the anastomotic lumen. The two-layer closure technique is standard and starts with an outer posterior layer of interrupted seromuscular stitches of 3-0 silk. The inner posterior and anterior layers are re-approximated with a continuous, non-locking stitch of 3-0 delayed absorbable suture. Finally, the outer anterior layer of interrupted seromuscular stitches completes the closure. After any anastomosis, the new lumen should always be checked for adequacy by invaginating the two limbs of intestine between the thumb and index

finger. The mesenteric defect should be closed using a 2-0 or 3-0 silk or delayed absorbable suture to prevent an internal herniation through the defect.

CYTOREDUCTIVE SURGERY ON THE INTESTINAL TRACT AND OMENTUM Exposure Surgical exposure is one of the most fundamental principles of ovarian cancer surgery. The first step to maximizing exposure is patient positioning. Any patient with suspected advanced-stage ovarian cancer should be positioned in the modified lithotomy position, since extensive pelvic dissections may require perineal access. The next element in obtaining adequate surgical exposure is selection of the incision, which should be versatile enough to provide for extended access to the upper abdomen if necessary, have adequate strength upon closure, and avoid unnecessary injury to underlying structures. Other considerations that must be accounted for are prior incisions, possible stoma placement, cosmesis and individual body habitus. The incision should never be the limiting factor to the operative dissection. Since surgery for ovarian cancer may require exposure from the deep pelvis to the diaphragm, with stoma placement always a possibility, a midline vertical incision is preferred by most surgeons. Abdominal entry is straightforward, fast and hemostatic; however, the vertical midline incision also predisposes the patient to certain complications in the early postoperative period (e.g. pain, burst abdomen, pulmonary morbidity) and may be associated with a higher incidence of late incisional hernia.13

Surgical approach The surgical approach to any debulking procedure should begin with a comprehensive evaluation of the abdomen and pelvis for accurate identification of the extent of disease. The exact sequence of procedures for cytoreductive surgery may vary depending on the

209

Ch 07 final.qxd

12/18/2006

15:24

Page 210

SURGERY FOR OVARIAN CANCER

a

b

c

d

Figure 7.14 Side-to-side (functional end-to-end) stapled anastomosis. (a) The two segments of bowel are approximated at their antimesenteric borders with proximal and distal stay sutures. The antimesenteric corners are trimmed. (b) The GIA stapler is introduced and fired. (c) The stapler is withdrawn and the staple lines off-set. (d) The remaining defect is closed with the TA stapler

individual situation; however, certain fundamental concepts should be kept in mind. In particular, after a thorough assessment of the extent of disease and confirmation that the tumor is of Müllerian origin, the surgeon must decide whether complete gross or optimal cytoreduction is feasible, what procedures will be required, and whether the potential morbidity associated with the operation is justified. Although the ideal threshold for ‘optimal’ versus ‘suboptimal’ residual disease remains controversial, most authorities

210

would agree that, unless tumor cytoreduction can be accomplished to ≤ 2 cm, an aggressive surgical attempt may not be warranted.14 If, on initial inspection, the surgeon identifies a portion of the operation that has the potential to be particularly challenging, we favor performing this procedure first, as the rate-limiting step, because if it is not technically feasible, then additional time has not been squandered performing other procedures that may not impact on the patient’s overall outcome. At

Ch 07 final.qxd

12/18/2006

15:24

Page 211

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

Memorial Sloan-Kettering Cancer Center, we tend to perform the omentectomy first, since it provides a sufficient sample of tumor for frozen-section analysis if the diagnosis is in question, and improves exposure to the remainder of the abdomen and pelvis (Figure 7.15). The sequence of the remaining procedures will depend on the individual situation; however, unnecessary bowel resection can be avoided by thoroughly assessing the extent of disease and leaving this to the latter portion of the operation. This also avoids the potential for disrupting the anastomosis with excessive manipulation while performing the other portions of the procedure. If extensive upper abdominal disease is unresectable, the potential morbidity of a bowel resection may not be justified. The only procedure that should be performed in the face of ‘unresectable’ disease is a bowel bypass or diversion procedure to relieve obstruction.

SURGERY ON THE SMALL BOWEL Advanced-stage ovarian cancer commonly involves the serosal surfaces of the small bowel and its mesentery. Localized/limited superficial disease may be amenable to sharp excision; however, segmental resection of the small bowel is often necessary as part of a maximal cytoreductive surgical effort. The dependent location of the terminal ileum and cecum makes this region particularly prone to involvement by locally advanced disease. In such circumstances, an en bloc resection with the primary tumor mass may be required.

Small-bowel resection The specific technique of small-bowel resection for tumor cytoreduction will vary according to the clinical situation, but, in general, it can be broken down into five basic elements: (1) Identifying the portion of small bowel to be resected; (2) Dividing the bowel; (3) Dividing the mesentery; (4) The anastomosis; (5) Closure of the mesentery.

a

b

Figure 7.15 (a) Initial removal of an omental cake improves exposure to the remainder of the abdomen and pelvis. (b) The completely resected omental specimen

Before dividing the bowel, the remainder of the intestinal tract should be carefully inspected for additional areas of disease or impending obstruction that would require surgical correction. This ensures that the planned resection will accomplish its intended purpose and that the anastomosis will be technically feasible. The segment of the small bowel to be removed should be clearly demarcated at the proximal and distal points, leaving approximately 5-cm segments of tumor-free bowel on either side to ensure an adequate margin of resection. An adequate blood supply to the remaining bowel should be confirmed by transilluminating the mesentery. To prevent spillage of bowel contents directly into the abdominal cavity,

211

Ch 07 final.qxd

12/18/2006

15:24

Page 212

SURGERY FOR OVARIAN CANCER

warm moist lap packs are used to isolate the specimen. Mesenterotomies are created through windows of Deaver at the corresponding proximal and distal points of resection. The small bowel is divided either between two applications of the linear GIA stapling device or between bowel clamps (Figure 7.16). Whether using a linear stapler or traditional clamps, the bowel should be transected at an oblique angle so that the greater portion is removed from the antimesenteric side, assuring adequate vascular perfusion of the entire transected edge. Attention is then directed toward the associated mesentery of the divided smallbowel segment. The incidence of ovarian cancer spread to the regional mesenteric lymph nodes may be as high as 70% when lymphovascular space involvement is present in the involved bowel wall.15 Consequently, consideration should be given to removing a wedge of mesentery, as is done for primary intestinal cancers, if the surgical objective is a complete macroscopic resection of disease. The knife, scissors, or cautery is used to delineate the V-shaped segment to be removed by scoring the bowel mesentery (Figure 7.16a–c). The jejunal and/or ileal vessels are skeletonized and individually ligated with 3-0 silk or delayed absorbable suture; however, care must be taken not to injure the superior mesenteric artery. Alternatively, the mesenteric vessels can be controlled with the use of a ligating dividing stapler (LDS) or vessel sealer (e.g. Ligasure® Valleylab, Colorado), which simultaneously transect and secure hemostasis. A side-to-side functional end-to-end anastomosis using the GIA and TA staplers (Figure 7.14) or hand-sewn end-to-end closure (Figure 7.8) is then performed. Because of an unpredictable blood supply, the distal 8–10 cm of terminal ileum should not be incorporated into the anastomosis. Rather, the distal ileum should be included in the resection and the anastomosis performed to the cecum or ascending colon (see below).

212

a

b

c

Figure 7.16 Small-bowel resection using linear GIA staples with functional end-to-end stapled anastomosis. (a) A mesenteric defect is made just beneath the bowel lumen and the bowel is divided using the GIA stapling device. (b) The bowel mesentery is divided and the blood supply is ligated. (c) The two portions of small bowel to be re-anastomosed are approximated at their antimesenteric surfaces and intestinal continuity re-established using either a stapled side-to-side functional end-to-end anastomotic technique (Figure 7.14) or a hand-sewn end-to-end technique (Figure 7.8)

Ch 07 final.qxd

12/18/2006

15:24

Page 213

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

Ileocecal resection Extensive tumor involvement of the ileocecal region may mandate resection of the terminal ileum in conjunction with a portion of the ascending colon. The tumor-involved intestine is mobilized by incising the parietal peritoneum from the terminal ileum, around the cecum, and along the white line of Toldt up to the hepatic flexure (Figure 7.17). The hepatocystocolic ligament is divided with electrocautery at the hepatic flexure to permit additional mobility of the ascending colon as it is separated from its attachments to the retroperitoneum by a combination of careful blunt

and sharp dissection. The ileum is then mobilized by incising along the base of the small-bowel mesentery toward the ligament of Treitz. The terminal ileum, ascending colon and proximal transverse colon are reflected medially, taking care to identify the right ureter, ovarian vessels, duodenum and head of the pancreas (Figure 7.18). The extent of resection required can now be delineated and the ileum and ascending colon divided between proximal and distal Kocher clamps or by using two applications of the linear GIA stapling device (Figure 7.19). Generally, the distal 8–10 cm of ileum should be removed with the

Liver

Gallbladder

Stomach

Hepatocystocolic ligament Right lateral paracolic gutter

Greater omentum

Third part of duodenum Right colic artery Ileocolic artery

Ureter Ovarian vessels

Figure 7.17 Mobilization of the right colon. The right colon is detached from the peritoneum along Toldt’s fascia (dashed line). The incision extends around the cecum. Careful attention is paid to the right ureter, ovarian vessels, genitofemoral nerve and iliac vessels as the cecum is mobilized. The gallbladder and duodenum lie in close proximity to the hepatic flexure as the hepatocystocolic ligament is divided

213

Ch 07 final.qxd

12/18/2006

15:24

Page 214

SURGERY FOR OVARIAN CANCER

cecum, since this area can have an inconsistent blood supply and result in a poorly vascularized anastomosis. A wedge of the ileocecal mesentery, incorporating the regional lymph nodes that may harbor subclinical disease, is delineated by scoring of the peritoneum. Particular caution must be exercised to preserve the terminal branches of the SMA to the remaining colon (right colic and middle colic arteries). Individual mesenteric vessels are suture ligated and divided and the specimen removed en bloc. A side-to-side functional end-to-end ileo–ascending colon anastomosis provides for a widely patent lumen and is the method of choice for re-establishing intestinal continuity. Either a hand-sewn or a stapled technique is applicable, but the stapled anastomosis is faster to perform. In the event that the ascending colon is not fully mobilized, an end-to-side anastomosis may be techni-

cally the more satisfactory option for re-establishing continuity between the ileum and colon. Finally, the mesenteric defect is closed.

SURGERY ON THE OMENTUM AND ABDOMINAL COLON Removing a portion of or the entire omentum is usually part of any primary ovarian cancer surgical procedure. Advanced-stage ovarian cancer tends to involve the omentum, and a large omental cake is not an unusual finding. Many times, the omental cake may be adherent to the transverse colon. An attempt should be made to peel the omentum off the transverse colon, since this will be feasible in the majority of cases.

Liver

Gallbladder Lesser omentum Morison's pouch

Stomach

Hepatic flexure

Transverse colon Head of pancreas Duodenum Lower pole of kidney

Inferior vena cava Ovarian vessels Ureter Psoas muscle Iliacus muscle

Right common iliac artery and vein

Figure 7.18 Mobilized right colon. After the right colon has been reflected medially, the vital underlying structures are exposed

214

Ch 07 final.qxd

12/18/2006

15:24

Page 215

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

Omentectomy If only the infracolic omentum is involved with tumor, it can be removed by first incising the posterior leaf at its reflection onto the transverse colon (Figure 7.20). The transverse colon can then be rolled in a caudal direction to expose the inferior gastrocolic ligament. The infracolic omentum is separated from the transverse colon in this fashion from the hepatic flexure to the splenic flexure. The right and left omental vascular pedicles are divided between clamps and ligated. The intervening epiploic vascular pedicles and the middle omental artery and vein are sequentially clamped, divided and ligated to complete the resection. Vascular pedicle control can also be achieved with the LDS stapler or vessel-sealing device (e.g.

Colon divided at hepatic flexure with GIA stapler

Terminal ileum divided with GIA stapler

Figure 7.19 Ileocectomy and right hemicolectomy. The amount of ascending colon to be removed will depend on the extent of tumor. If only the cecum is involved, the right colic artery and the corresponding segment of colon can be spared. If a more extensive resection is required, the middle colic artery should be identified and spared

Ligasure). If the omentum is adherent to the anterior abdominal wall, the posterior rectus fascial sheath can be circumscribed around the area of tumor involvement with electrocautery, dissected from the overlying rectus muscle, and resected en bloc with the omentum. For extensive infiltration of the omentum by ovarian cancer, a total omentectomy, with resection of the gastrocolic ligament, should be performed. The procedure is started in similar fashion to the infracolic omentectomy by incising the posterior leaf along the posterior border of the transverse colon (Figure 7.21). The lesser sac is entered and widely developed, completely mobilizing the omentum from the transverse colon extending from the hepatic flexure to the splenic flexure. The gastrocolic ligament must be carefully dissected from the underlying transverse colon mesentery containing the middle colic artery (Figure 7.22). The correct plane of dissection is most easily developed from the left side, beginning underneath the greater curvature of the stomach after mobilizing the splenic flexure, and taking down the phrenocolic and lienocolic ligaments. Once the gastrocolic ligament has been entirely freed from the transverse mesocolon, and the lesser sac adequately exposed, the omentum is divided from the greater curvature of the stomach by skeletonizing, clamping, dividing and suture ligating the individual vascular pedicles arising from the gastroepiploic arcade (Figure 7.23). If the omental tumor encroaches onto the greater curvature of the stomach, the gastroepiploic arcade can be sacrificed, as the left gastric artery will provide adequate blood supply to the greater curvature through the intramural vascular anastomostic network within the wall of the stomach. Ligatures encroaching upon the stomach wall should be secured in place with a transfixion stitch. The stomach should be decompressed during the immediate postoperative period after total omentectomy to prevent gastric distension and the dislodging of one of the suture ligatures along the greater curvature. On occasion, a large omental cake will be inseparable from the transverse, ascending, or descending

215

Ch 07 final.qxd

12/18/2006

15:24

Page 216

SURGERY FOR OVARIAN CANCER

Lesser sac Omentum Middle omental artery Transverse colon Taenia Appendices epiploica Transverse mesocolon

Figure 7.20 Omentectomy. The posterior leaf of the omentum is incised and freed off the transverse colon

Figure 7.21 Total omentectomy. The omentum is drawn into

Figure 7.22 Total omentectomy. The lesser sac has been

the incision, the transverse colon mobilized and the point of

entered and the omentum and gastrocolic ligament completely

incision selected along the dorsal peritoneal reflection of the

mobilized away from the underlying transverse colon

omentum and the transverse colon

mesentery. The stomach is located cephalad to the gastrocolic ligament and the middle colic vessels are seen running within the transverse colon mesentery

216

Ch 07 final.qxd

12/18/2006

15:24

Page 217

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

colon despite all attempts to identify a plane of dissection. In these instances, an en bloc resection of the omentum and involved colon is appropriate.

Subtotal transverse colectomy in conjunction with omental disease The transverse colon and its mesentery may be directly infiltrated by bulky omental disease, or the associated fibrotic inflammatory tissue reaction may make separation along anatomic planes impossible. The omentum is divided from the lateral aspects of the transverse colon and the lesser sac entered and developed. The transverse colon is then completely mobilized by dividing the gastrocolic ligament from the greater curvature of the stomach and taking down the hepatocystocolic, phrencocolic and lienocolic ligaments to free the hepatic flexure and splenic flexure, respectively. The en bloc tumor specimen to be resected is clearly delineated and the vascular supply to the remaining proximal and distal ends of transverse colon inspected (Figure 7.24). The surgeon must ensure that the marginal artery of Drummond is intact and will provide a sufficient blood supply to

both ends of the planned anastomosis. A wedgeshaped section of transverse colon mesentery is demarcated by scoring the overlying peritoneum, and the middle colic artery identified, ligated and divided distal to its origin from the SMA. Any associated arterial arcades are similarly individually suture ligated and divided. If the marginal artery of Drummond is found to be discontinuous at the splenic flexure, the distal transverse colon and proximal descending colon are included in the scope of resection. A colocolostomy is created to re-establish intestinal continuity via either an end-to-end or a functional end-to-end stapled or hand-sewn anastomosis

Middle colic artery

Marginal artery of Drummond Omental tumor inseparable from transverse colon Gastrocolic ligament divided from greater curvature of stomach

a Stapled end-to-end anastomosis

b Functional end-to-end anastomosis

Figure 7.24 Subtotal transverse colectomy en bloc with omental disease. The middle colic artery is normally sacrificed; Figure 7.23 Total omentectomy. The right gastroepiploic

however, the integrity of the marginal artery of Drummond

vessels have been divided. The gastrocolic ligament is resected

must be confirmed to ensure adequate vascularization to

from the greater curvature of the stomach by skeletonizing,

proximal and distal ends for anastomosis. Inset a, stapled end-

clamping, dividing and suture ligating individual vascular

to-end anastomosis. Inset b, stapled functional end-to-end

pedicles arising from the gastroepiploic arcade

anastomosis

217

Ch 07 final.qxd

12/18/2006

15:24

Page 218

SURGERY FOR OVARIAN CANCER

(Figure 7.24). Avoidance of tension on the staple or suture line is critical to creating a viable anastomosis, and additional mobilization of the hepatic flexure and/or splenic flexure may be required. The mesenteric defect is closed; however, the duodenojejunal junction should be carefully inspected to ensure that re-approximating the colonic mesentery does produce a functional stricture at this point.

Right hemicolectomy in conjunction with omental disease Extensive omental disease involving the ascending colon or hepatic flexure warrants an en bloc resection with right hemicolectomy. The terminal ileum, cecum and ascending colon, including the hepatic flexure, are completely mobilized as described earlier. Any attachments to the gallbladder should be carefully taken down. The omentum is dissected from the transverse colon, the lesser sac entered, and the gastrocolic ligament divided from the greater curvature of the stomach. The segment of colon to be removed is demarcated, extending from the terminal ileum proximally to just beyond the hepatic flexure of the transverse colon distally (Figure 7.25). The peritoneum overlying the mesentery should be incised in a wedge-shaped fashion to expose the vascular supply to the right colon. The ileocolic and right colic arteries are carefully isolated, ligated and divided, as they are included within the scope of resection. In most cases, the right-sided branch of the middle colic artery will also need to be sacrificed, but the middle colic artery itself should be preserved. The intestine is then divided using two applications of the linear GIA stapling device or between clamps. The anastomosis can be completed by one of several techniques using either automated staplers or a hand-sewn closure. A stapled functional end-to-end anastomosis between the terminal ileum and proximal transverse colon is most commonly selected (Figure 7.25). However, an end-to-end anastomosis is also applicable and usually requires a Cheatle incision in the terminal ileum.

218

Left hemicolectomy in conjunction with omental disease The technique for left hemicolectomy and en bloc resection of omental disease is similar, in that complete mobilization of the omentum from the transverse colon and greater curvature of the stomach is required. The descending colon is retracted medially after incising along the white line of Toldt and the splenic flexure taken down by dividing the phrenocolic and lienocolic ligaments. When dividing the lienocolic ligament, excessive downward traction must be avoided to prevent possible capsular tear to the spleen. The segment of colon to be removed is demarcated, extending from the midportion of the transverse colon, to account for the unpredictable blood supply at the splenic flexure, to the juncture of the descending colon and sigmoid colon (Figure 7.26). The peritoneum overlying the mesentery is incised to expose the left colic artery, which is carefully ligated and divided distal to its origin from the IMA. Proximally, the middle colic artery should be clearly identified and preserved; however, the left-sided

a

b

Figure 7.25 Right hemicolectomy. (a) Extent of resection for extensive omental disease involving the ascending colon requiring en bloc right hemicolectomy. (b) Stapled functional end-to-end anastomosis between the terminal ileum and proximal transverse colon

Ch 07 final.qxd

12/18/2006

15:24

Page 219

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

branch may be incorporated within the scope of resection. An end-to-end anastomosis is completed between the proximal transverse colon and the sigmoid colon using either a stapled or a hand-sewn technique (Figure 7.26). Depending on the available length of the transverse colon, it may be necessary to mobilize the hepatic flexure to obtain a tension-free anastomosis of transverse to sigmoid colon. Finally, the mesenteric defect is closed.

Total colectomy Unlike the small bowel, the majority of the colon can be removed with a subtotal colectomy and intestinal continuity re-established via an ileosigmoid or ileorectal anastomosis. Although this extensive procedure is rarely, if ever, indicated for patients with ovarian cancer, it does have the advantage of only one anastomosis being performed if multiple portions of the colon are involved with tumor. However, removal of the majority of the colon may predispose the patient to diarrhea. In a series of 136 patients who underwent subtotal colectomy for colon cancer, the

a

length of the remaining colon and resected terminal ileum had a significant effect on postoperative diarrhea. If less than 10 cm of terminal ileum was resected and more than 10 cm of colon was left above the peritoneal reflection, there was a marked decrease in diarrhea.16

MESENTERIC DISEASE Ovarian cancer may spread to the bowel mesentery as miliary disease or tumor plaques covering the peritoneal surface or as regional lymphadenopathy (Figure 7.27). The decision to attempt cytoreduction should be based on the overall extent of tumor and the likelihood of achieving an optimal volume of residual disease. The two principal techniques for approaching small-bowel mesentery disease are resection and ablation, the choice depending on the pattern and extent of tumor. Often, a combination of both techniques will be necessary. Extensive coverage of the mesenteric surface by small-volume tumor nodules is a particularly important surgical consideration, as these patients are frequently categorized as ‘optimally’ debulked despite residual miliary tumor deposits. Although these implants are typically less

b

Figure 7.26 Left hemicolectomy. (a) Line of peritoneal incision for complete mobilization including sacrifice of left colic artery distal to origin from the inferior mesenteric artery.

Figure 7.27 Ovarian cancer with extensive involvement of

(b) Stapled or hand-sewn end-to-end anastomosis between the

the small bowel mesentery by multiple small-volume tumor

proximal transverse colon and the sigmoid colon

nodules

219

Ch 07 final.qxd

12/18/2006

15:24

Page 220

SURGERY FOR OVARIAN CANCER

than 1 cm in diameter, improved survival has been associated with removal of all gross visible disease.17 With an aggressive surgical approach, the rate of complete visible tumor removal can be in excess of 80% of cases.9 In a case–control study to evaluate the benefit of peritoneal and serosal implant elimination, using techniques such as sharp dissection, CO2 laser, electrocautery, argon beam coagulation and ultrasonic aspiration, Eisenkop et al.17 demonstrated that patients who underwent cytoreduction with and without implant elimination to no gross residual disease had similar outcomes. Additionally, patients who underwent implant elimination with no gross residual disease had significantly better outcomes than patients who did not undergo implant elimination and had disease of ≤ 1 cm on peritoneal surfaces. Rarely, metastatic ovarian cancer will extensively infiltrate the root of the small-bowel mesentery; this spread pattern is not generally amenable to cytoreduction.

Excisional techniques Mesenteric peritoneal surface disease can be managed by excision of individual tumor nodules, if of limited

a

extent, or by a peritonectomy procedure for more diffuse tumor spread. Peritonectomy of the bowel mesentery is initiated by using the electrosurgical unit (ESU) to score the involved peritoneum in a circumscribing fashion. The peritoneal edge is raised by sharp dissection, and the mesenteric peritoneum carefully ‘peeled’ or ‘stripped’ from the underlying fatty tissue, with the mesenteric vessels identified and preserved in the process (Figure 7.28). Macroscopically involved regional mesenteric lymph nodes can be resected by incising the overlying peritoneum and applying digital pressure from the posterior surface of the mesentery. Sharp dissection with scissors or the ESU is used to mobilize the neighboring vasculature as the enlarged lymph node is enucleated from within the mesenteric fat. The associated intestinal wall should be carefully inspected following excision of mesenteric nodal disease to ensure that the associated vascular supply has not been compromised. When mesenteric nodal disease is in close proximity to the bowel wall or intimately associated with vascular structures, it may be safer to perform a segmental resection of the associated bowel.

b

Figure 7.28 Peritonectomy of sigmoid colon mesentery. (a) A margin of mesenteric peritoneum is raised and the diseased tissue circumscribed. (b) The peritoneal edge is reflected off the underlying mesenteric soft tissue and vasculature

220

Ch 07 final.qxd

12/18/2006

15:24

Page 221

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

The loop electrosurgical excision procedure (LEEP) instrument can be a useful adjuvant during cytoreduction of ovarian cancer metastatic to the bowel mesentery peritoneum. Fanning and Hilgers reported their experience in 20 consecutive patients who underwent maximal cytoreduction using standard surgical techniques.18 The authors employed the LEEP to intensify cytoreduction of metastasis involving the intestines, diaphragm, liver, spleen and peritoneal surface. Seventeen (94%) of the 18 patients with intestinal metastasis had their disease completely resected using the LEEP. The median LEEP time for the entire group was 9 min, and there were no complications directly attributable to the LEEP. The authors concluded that this was an excellent means of removing plaque-like peritoneal disease. An alternative method for removing bowel mesenteric or serosal disease is the cavitron ultrasonic surgical aspirator (CUSA). This instrument combines ultrasonic waves with irrigation and suction to produce selective tissue resection. Tumor tissue is fragmented, simultaneously irrigated and evacuated, while normal surrounding tissue is preserved. Adelson et al. illustrated the usefulness of the CUSA in debulking nine of ten patients with ovarian or tubal cancers to residual disease no greater than 5 mm.19 Tumor was removed from the diaphragm, spleen, stomach, and small and large bowel using the ultrasonic aspirator. In a later series, Adelson described his experience with this technique in 37 patients with metastasis to the small intestine measuring > 5 mm.20 Using the CUSA, he was able to remove all grossly evident intestinal disease completely in 13 patients (35.1%). The remaining 24 patients (64.9%) were left with only small-volume (1–5 mm) residual tumor on the bowel surfaces. The usefulness of the CUSA may be limited in tumors that are particularly fibrotic in nature.21

Ablative techniques Extensive small-volume ovarian cancer spread over the bowel mesenteric surface can also be effectively eradicated using ablative techniques. Bristow and

Montz22 reported their experience using the argon beam coagulator in cytoreductive surgery for patients with advanced-stage ovarian cancer. Power settings of 60–80 W were typically used to ablate subcentimeter implants and tumor nodules. Higher settings of 100–110 W were used for larger nodules or tumor plaques. The use of the argon beam coagulator was associated with a greater likelihood of achieving complete cytoreduction, with no gross residual disease, compared to results in patients debulked by more traditional methods. When examining its use on the bowel mesentery specifically, 80% of patients were able to have their mesenteric disease completely removed, compared with 0% when the argon beam coagulator was not used. Tumor ablation of intestinal metastasis using a CO2 laser has also been described by Fanning et al.23 Twenty consecutive patients with epithelial ovarian cancer and metastasis to the mesentery and/or serosa of the small and/or large bowel were cytoreduced with the CO2 laser. In 19 of the 20 patients, all tumor was removed from the intestine using laser ablation. This method of intensive cytoreduction resulted in superior debulking without increasing postoperative morbidity. Caution should be exercised, however, when either the argon beam coagulator or the CO2 laser is used on bowel serosa, as both techniques induce a degree of underlying thermal damage that is difficult to predict.

COMPLICATIONS OF INTESTINAL SURGERY Ileus Prolonged ileus is the most common bowel complication after intestinal surgery for ovarian cancer, affecting 11–17% of patients who undergo bowel resection.4,24 After surgery, there is a gradual restoration of intestinal function depending on the portion of the intestinal tract; the small intestine recovers in 6–12 h; the stomach in 24–48 h, and the colon usually in 2–4 days. Signs and symptoms of ileus include decreased or absent bowel sounds, abdominal distension, nausea and vomiting, and lack of passage of stool or flatus.

221

Ch 07 final.qxd

12/18/2006

15:24

Page 222

SURGERY FOR OVARIAN CANCER

Flat and upright abdominal radiographs may reveal air in the stomach, air–fluid levels in the small bowel and dilated loops of intestine. Symptomatic management with nasogastric drainage, intravenous fluids and, in some cases, parenteral nutrition is warranted. Followup imaging studies should be obtained to make sure a bowel obstruction is not missed.

Obstruction Small-bowel obstruction may develop at any time in the postoperative period; signs and symptoms include abdominal distension, high-pitched bowel sounds, nausea and vomiting, and colicky abdominal pain. Adhesions may be the principal cause of obstruction, especially after extensive cytoreductive surgery with one or more intestinal resections. Typically, a trial of conservative management with nasogastric drainage and intravenous hydration is warranted if there are no signs of bowel strangulation, such as fever, an elevated white blood cell count, tachycardia, severe abdominal pain, or signs of peritoneal irritation (e.g. guarding, rebound tenderness). If there is a suspicion of bowel strangulation, surgical exploration is indicated. In the case of a partial small-bowel obstruction, spontaneous resolution occurs with conservative therapy in the majority of cases.25 Complete obstruction may be amenable to a brief trial of conservative management, but patients should be monitored closely and supplemental imaging ordered to look for bowel strangulation, which requires immediate surgical management. Large-bowel obstruction in the immediate postoperative period may be functional or mechanical. A functional large-bowel obstruction (colonic pseudoobstruction or Ogilvie syndrome) usually affects the elderly patient on narcotic analgesics following major surgery (especially involving the right colon) and is thought to result from uncoordinated innervation to the proximal and distal colon. This syndrome usually manifests as a dilated colon on abdominal films, which will also characteristically show air in the rectum. Treatment should be conservative, with bowel rest, nasogastric decompression, parenteral nutrition

222

and discontinuance of narcotic medications, if possible. Serial abdominal films should be obtained to assess the distension of the cecum, which, by virtue of its thin wall, is at greatest risk for perforation. A cecal diameter larger than 10–11 cm requires immediate decompression via endoscopy or tube cecostomy. A mechanical large-bowel obstruction in the immediate postoperative period is uncommon after ovarian cancer surgery, but, when present, requires surgical correction or decompression via diverting colostomy.

Fistulae Enterocutaneous fistulae are usually iatrogenic and result from unrecognized bowel wall injury, ischemia, or an anastomotic leak. They may present initially as bloody discharge, but soon progress to leakage of enteric contents. The diagnosis can be confirmed by oral administration of charcoal or Congo red, both non-absorbable markers which will be present in the fistula effluent. Alternatively, a computed tomography (CT) scan or fistulogram with water-soluble contrast medium can be performed. Rarely, a smallbowel fistula may present with generalized peritonitis. Fistulae are described according to their location and daily output. High-output fistulae are those with an output of 500 ml or greater in a 24-h period. These fistulae tend to occur in the proximal portion of the small bowel and can result in excessive fluid and electrolyte loss. Treatment of fistulae requires management of the associated symptoms. In most cases, particularly if the patient is febrile, broad-spectrum antibiotics should be started immediately. Fluid and electrolyte abnormalities should be corrected, and the fistulous drainage collected, which allows for accurate output measurement. If the fistula is associated with an abscess, percutaneous drainage is recommended. Patients are usually started on parenteral nutrition, since they may easily become malnourished.26 Octreotide has been used in the management of fistulae with mixed results; however, it does reduce the amount of fistula output and thus allows better control of volume loss.27 If output can be controlled, a trial of conservative management is indicated;

Ch 07 final.qxd

12/18/2006

15:24

Page 223

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

however, the appropriate duration of observation is unknown. Approximately 90% of fistulae that close spontaneously will do so in the first month, with fewer than 10% closing after 2 months. It is reasonable to manage the clinically stable patient with an enterocutaneous fistula conservatively for 4–6 weeks. Fistulae attributable to a surgical procedure are more likely to close spontaneously, especially if they have low output and are not associated with other complications.28 If spontaneous closure is not achieved after 2 months, surgical repair should be considered. Colonic fistula formation can occur after largebowel surgery, with a resulting communicating tract between the colon and the skin, vagina, small bowel, or bladder. The symptomatology will depend on which viscera are involved. Imaging studies, such as a CT scan, gastrografin enema, or intravenous pyelogram, may be useful to localize the involved segment of bowel. Colocutaneous fistulae can be managed expectantly with local care to the fistula site; however, resting the intestinal tract and administering parenteral nutrition is unnecessary unless there is a high-volume output. If conservative management is unsuccessful after a period of 6 weeks, primary repair should be undertaken. For more complex colonic fistulae (colovesical, coloenteric), a one-step restorative procedure, with resection of the fistulous portion of colon and associated viscera, is favored. This entails mobilizing the colon proximal and distal to the fistula. In the unusual situation in which a one-step procedure is unsafe, a diverting colostomy can be performed.

MANAGEMENT OF KRUKENBERG TUMORS Among patients with a history of colon cancer who subsequently undergo surgery for an ovarian mass, secondary involvement of the ovaries by metastatic colon cancer will be found in approximately 60% of cases.29 Metastatic disease to the ovaries, and occasionally the peritoneal cavity, may also be the initial manifestation of a previously undiagnosed malignancy of gastrointestinal origin (Figure 7.29). It is therefore

important for the surgeon operating on women with suspected ovarian cancer to be familiar with the appropriate management of these tumors. In 1896, Krukenberg30 described a primary fibrosarcoma of the ovary that had mucinous elements. In reality, he was describing a metastatic tumor with signet-ring cell features originating from a primary gastric cancer. Since that time, the term ‘Krukenberg tumor’ has been used to describe an adenocarcinoma secondarily involving the ovary that has pleomorphic signet-ring cells containing mucin surrounded by an excessive, reactive proliferation of ovarian stromal cells. These tumors originate from the stomach, colon, pancreas, or hepatobiliary tract and account for 4–8% of all carcinomas metastatic to the ovary. When a Krukenberg tumor is identified, a thorough search for an unrecognized gastrointestinal primary should be undertaken. In a series of 41 patients with ovarian involvement from extragenital cancer, the primary was identified in over 80% of cases if the tumor originated in the abdomen.31 Decisions regarding the extent of the surgical procedure should be made on the basis of the specific locale of the primary tumor. The prognosis for patients with Krukenberg tumors is generally poor.32 However, there have been reports of improved outcomes with an aggressive surgical approach. Webb et al. reported on a large series of tumors metastatic to the ovary in which nine of 169 patients (5%) with tumors of gastrointestinal origin were alive at 5 years, with six of the nine surviving at least 10 years.33 Morrow and Enker reported on 63 patients with late ovarian metastasis treated with aggressive surgery including pelvic exenteration and hepatic lobectomy.34 The single most important discriminant of survival was whether or not the patient could be surgically rendered disease-free, which was associated with an average survival time of 48 months, compared to just 8 months for patients left with residual disease. The authors concluded that, although many patients do not show a survival benefit in the presence of gross ovarian metastasis, significant palliation is achieved and long-term survival is

223

Ch 07 final.qxd

12/18/2006

15:24

Page 224

SURGERY FOR OVARIAN CANCER

a

b

Figure 7.29 Krukenberg tumor. (a) Metastatic malignant tumor mass replacing the left ovary but originating from (b) the primary tumor located in the sigmoid colon

occasionally possible. Recently, others have also recommended an aggressive surgical approach to patients with ovarian metastasis from colorectal cancers, especially if disease is limited to the pelvis and it can be completely resected.35,36 Krukenberg tumors of gastric origin may also have a better prognosis if patients can be rendered disease free. Kim et al. reported on a series of 34 patients with ovarian metastasis from gastric origin.37 Although median survival for the entire group was only 7.7 months, the survival of patients without gross residual disease was significantly longer (median 10.9 months) compared to patients with residual disease (median 7.5 months). These authors, as well as others, have suggested that resection might have a role in the management of Krukenberg tumors of gastric origin if it can render patients visibly disease free.38

Contemporary stapling devices and ablative technology have made cytoreductive surgery of the intestinal tract less time consuming and easier to master. However, a fundamental knowledge of the basic anatomy and surgical principles is essential for the safe and proper application of these new technologies. A combination of technical skill and sound surgical judgment is necessary to provide patients with the best possible operative result while minimizing the risk of complications.

REFERENCES 1.

Sampson JA. Implantation peritoneal carcinomatosis of ovarian origin. Am J Pathol 1931; VII: 423–43

2.

Wu PC, Lang JH, Huang RL, et al. Intestinal metastasis and operation in ovarian cancer: a report on 62 cases. Baillière’s Clin Obstet Gynaecol 1989; 3: 95–108

3.

Chalas E, Mann WJ Jr, Westermann CP, et al. Morbidity and mortality of stapled anastomoses on a gynecologic oncology service: a retrospective review. Gynecol Oncol 1990; 37: 82–6

CONCLUSION Effective operative treatment of advanced-stage ovarian cancer requires the surgeon to be skilled in the management of tumor involving the intestinal tract.

224

Ch 07 final.qxd

12/18/2006

15:24

Page 225

CYTOREDUCTIVE SURGERY: INTESTINAL TRACT AND OMENTUM

4.

Gillette-Cloven N, Burger RA, Monk BJ, et al. Bowel resection at the time of primary cytoreduction for epithelial ovarian cancer. J Am Coll Surg 2001; 193: 626–32

the treatment of colon cancer. J Am Coll Surg 1997; 184: 269–72 17.

Eisenkop SM, Nalick RH, Wang HJ, et al. Peritoneal implant elimination during cytoreductive surgery for ovarian cancer: impact on survival. Gynecol Oncol 1993; 51: 224–9

18.

Fanning J, Hilgers RD. Loop electrosurgical excision procedure for intensified cytoreduction of ovarian cancer. Gynecol Oncol 1995; 57: 188–90

19.

Adelson MD, Baggish MS, Seifer DB, et al. Cytoreduction of ovarian cancer with the Cavitron ultrasonic surgical aspirator. Obstet Gynecol 1988; 72: 140–3

5.

Griffiths CT. Surgical resection of tumor bulk in the primary treatment of ovarian carcinoma. Baillière’s Clin Obstet Gynaecol 1989; 3: 95–108

6.

Jaeger W, Ackermann S, Kessler H, et al. The effect of bowel resection on survival in advanced epithelial ovarian cancer. Gynecol Oncol 2001; 83: 286–91

7.

Weber AM, Kennedy AW. The role of bowel resection in the primary surgical debulking of carcinoma of the ovary. J Am Coll Surg 1994; 179: 465–70

8.

Tamussino KF, Lim PC, Webb MJ, et al. Gastrointestinal surgery in patients with ovarian cancer. Gynecol Oncol 2001; 80: 79–84

20.

Adelson MD. Cytoreduction of small intestine metastases using the Cavitron Ultrasonic Surgical Aspirator. J Gynecol Surg 1995; 11: 197–200

9.

Eisenkop SM, Friedman RL, Wang HJ. Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced epithelial ovarian cancer: a prospective study. Gynecol Oncol 1998; 69: 103–8

21.

Deppe G, Malviya VK, Malone JM Jr. Debulking surgery for ovarian cancer with the Cavitron Ultrasonic Surgical Aspirator (CUSA) – a preliminary report. Gynecol Oncol 1988; 31: 223–6

22.

10.

Delgado G. The automatic staple versus the conventional gastrointestinal anastomosis in gynecological malignancies. Gynecol Oncol 1981; 12: 302–13

Bristow RE, Montz FJ. Complete surgical cytoreduction of advanced ovarian carcinoma using the argon beam coagulator. Gynecol Oncol 2001; 83: 39–48

23.

11.

Wheeless CR Jr, Smith JJ. A comparison of the flow of iodine 125 through three different intestinal anastomoses: standard, Gambee, and stapler. Obstet Gynecol 1983; 62: 513–18

Fanning J, Hilgers RD, Richards PK, et al. Carbon dioxide laser vaporization of intestinal metastases of epithelial ovarian cancer. Int J Gynecol Cancer 1994; 4: 324–7

24.

12.

Dziki AJ, Duncan MD, Harmon JW, et al. Advantages of handsewn over stapled bowel anastomosis. Dis Colon Rectum 1991; 34: 442–8

Paladini D, Fontanelli R, Raspagliesi F, et al. Intestinal operations during surgical procedures for epithelial ovarian cancer. Int J Gynecol Cancer 1994; 4: 320–3

13.

Grantcharov TP, Rosenberg J. Vertical compared with transverse incisions in abdominal surgery. Eur J Surg 2001; 167: 260–7

25.

Fevang BT, Jensen D, Svanes K, et al. Early operation or conservative management of patients with small bowel obstruction? Eur J Surg 2002; 168: 475–81

14.

Hoskins WJ, McGuire WP, Brady MF, et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J Obstet Gynecol 1994; 170: 974–9, discussion 979–80

26.

Soeters PB, Ebeid AM, Fischer JE. Review of 404 patients with gastrointestinal fistulas: impact of parenteral nutrition. Ann Surg 1979; 190: 189–202

27.

Sancho JJ, di Costanzo J, Nubiola P, et al. Randomized double-blind placebo-controlled trial of early octreotide in patients with postoperative enterocutaneous fistula. Br J Surg 1995; 82: 638–41

28.

Campos AC, Andrade DF, Campos GM, et al. A multivariate model to determine prognostic factors in gastrointestinal fistulas. J Am Coll Surg 1999; 188: 483–90

15.

16.

O’Hanlan KA, Kargas S, Schreiber M, et al. Ovarian carcinoma metastases to gastrointestinal tract appear to spread like colon carcinoma: implications for surgical resection. Gynecol Oncol 1995; 59: 200–6 Papa MZ, Karni T, Koller M, et al. Avoiding diarrhea after subtotal colectomy with primary anastomosis in

225

Ch 07 final.qxd

12/18/2006

15:24

Page 226

SURGERY FOR OVARIAN CANCER

29.

Abu-Rustum N, Barakat RR, Curtin JP. Ovarian and uterine disease in women with colorectal cancer. Obstet Gynecol 1997; 89: 85–7

35.

Miller BE, Pittman B, Wan JY, et al. Colon cancer with metastasis to the ovary at time of initial diagnosis. Gynecol Oncol 1997; 66: 368–71

30.

Krukenberg F. Uber das Fibrosarcomma Ovarii Mucocellare (Carcinomatoides). Arch Gynaekol 1896; 50: 287

36.

Rayson D, Bouttell E, Whiston F, et al. Outcome after ovarian/adnexal metastectomy in metastatic colorectal carcinoma. J Surg Oncol 2000; 75: 186–92

31.

Yazigi R, Sandstad J. Ovarian involvement in extragenital cancer. Gynecol Oncol 1989; 34: 84–7

37.

32.

Hale RW. Krukenberg tumor of the ovaries. A review of 81 records. Obstet Gynecol 1968; 32: 221–5

Kim HK, Heo DS, Bang YJ, et al. Prognostic factors of Krukenberg’s tumor. Gynecol Oncol 2001; 82: 105–9

38.

Yada-Hashimoto N, Yamamoto T, Kamiura S, et al. Metastatic ovarian tumors: a review of 64 cases. Gynecol Oncol 2003; 89: 314–17

33.

Webb MJ, Decker DG, Mussey E. Cancer metastatic to the ovary: factors influencing survival. Obstet Gynecol 1975; 45: 391–6

34.

Morrow M, Enker WE. Late ovarian metastases in carcinoma of the colon and rectum. Arch Surg 1984; 119: 1385–8

226

Ch 08 final.qxd

12/18/2006

15:26

Page 227

CHAPTER 8

Cytoreductive surgery: right upper abdomen Robert L Giuntoli II, Robert E Bristow, Richard D Schulick

INTRODUCTION Ovarian cancer frequently involves the structures of the right upper abdomen. This is not surprising, given the high incidence of extrapelvic disease among patients with primary ovarian malignancies and the typical pattern of peritoneal tumor dissemination along the right paracolic gutter to the right upper quadrant. At the time of primary cytoreductive surgery, the right hemidiaphragm and liver surface are often noted to harbor metastatic disease in patients with advanced-stage ovarian cancer.1–3 Less commonly, the liver parenchyma, gallbladder and its associated fossa, and porta hepatis may be involved with disease. Ovarian cancer recurrence may also manifest at these sites, particularly after an incomplete primary resection. Recent studies have specifically addressed the feasibility and survival benefit of radical resection of synchronous and metachronous lesions in the right upper quadrant.1,4–6 Nevertheless, ovarian cancer metastases involving the diaphragm, liver and associated structures are frequently cited as principal impediments to achieving an overall optimal cytoreductive surgical outcome. Safe and effective operative management of such disease requires the ovarian cancer surgeon to be intimately familiar with the anatomy of the right upper quadrant and proficient in both excisional and ablative techniques of tumor extirpation.

REGIONAL ANATOMY Overview The diaphragm represents the superior border of the right upper quadrant, with the liver lying immediately

below. The gallbladder lies along the inferior surface of the right lobe of the liver within the gallbladder fossa. Anterolaterally, the hepatic flexure of the colon is located in close proximity to the right lobe of the liver. The duodenum is located retroperitoneally, inferior to the liver and posterior to the transverse colon. The greater omentum arises along the greater curvature of the stomach and extends from the hepatic to the splenic flexure of the transverse colon. The lesser omentum is composed of the hepatogastric and hepatoduodenal ligaments. The lesser peritoneal sac lies posterior to the greater omentum. The abdominal cavity communicates with the lesser peritoneal sac by way of the epiploic foramen, which is located immediately superior to the first part of the duodenum. The lesser peritoneal sac may also be entered by dissecting the greater omentum off the transverse colon or dividing the gastroduodenal ligament. The hepatoduodenal ligament runs from the porta hepatis to the second portion of the duodenum and contains the hepatic artery, portal vein, common bile duct, hepatic nerve plexus and lymphatic vessels. It is situated immediately medial and anterior to the epiploic foramen (Figure 8.1). The retroperitoneal structures that may be encountered during cytoreductive surgery for ovarian cancer in the right upper abdomen include the duodenum, the ascending colon, the pancreas, the right kidney and adrenal gland, the inferior vena cava and the aorta (Figure 8.2). The right kidney and adrenal gland lie lateral to the second part of the duodenum, while the head of the pancreas abuts its medial surface. The aorta lies anterior and slightly left of the spinal column and the inferior vena cava lies to the right of the aorta.

227

Ch 08 final.qxd

12/18/2006

15:26

Page 228

SURGERY FOR OVARIAN CANCER

Hepatoduodenal ligament Round ligament Diaphragm Gallbladder

Hepatogastric ligament (lesser omentum)

Falciform ligament

Left liver lobe Stomach Spleen

Right liver lobe

Epiploic foramen Duodenum

Hepatic flexure Greater omentum

Figure 8.1 Appearance of the right upper quadrant with the liver and gallbladder retracted superiorly

Diaphragm

Right adrenal gland

Pancreas

Right kidney Duodenum

Hepatic flexure Ascending colon

Superior mesenteric artery and vein

Inferior vena cava Aorta

Figure 8.2 Appearance of the right upper quadrant with liver, gallbladder, transverse colon, omentum and stomach removed

228

Ch 08 final.qxd

12/18/2006

15:26

Page 229

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

Diaphragm The diaphragm is a dome-shaped musculofibrous layer, which separates the thorax from the abdominal cavity. The thoracic side of the diaphragm is in contact with the pleura of the lung. Anterolaterally, the undersurface of the diaphragm is lined by the abdominal peritoneum; the posteromedial diaphragmatic surface is retroperitoneal. Muscular components

The muscle fibers of the diaphragm originate from three areas, referred to as the sternal, costal and lumbar components of the diaphragm. The sternal component arises from the xyphoid process, while the costal component originates from the lower six ribs. The two crura of the diaphragm form the lumbar component of the diaphragm in conjunction with the arcuate ligament. The crura form the aortic aperture and join together anteriorly as the median arcuate ligament. The medial and lateral arcuate ligaments sur-

Sternal origin

round the upper part of the psoas muscle and the quadratus lumborum muscle, respectively. All of the muscle fibers of the diaphragm insert into the central tendon, an aponeurosis that is located near the center of the diaphragm immediately below the pericardium (Figure 8.3). Apertures

There are three major openings within the diaphragm (Figure 8.3). The vena caval foramen is positioned to the right of the midline in the central tendon at the level of the intervertebral disc between the 8th and 9th thoracic vertebrae. In addition to the inferior vena cava, branches of the right phrenic nerve pass through this opening. Occasionally, the right hepatic vein passes through the vena caval foramen prior to joining the inferior vena cava. The esophageal hiatus is located to the left of the midline, posterior to the central tendon at the level of the 10th thoracic vertebra, and contains the anterior and posterior vagal

Anteromedian gap Anterolateral gap

Costal origin

Central tendon Vena cava opening Esophageal opening

Aortic opening Gap of psoas Vertebrocostal triangle 12th rib

Lateral arcuate ligament (lumbocostal arch)

Quadratus lumborum

Medial arcuate ligament (lumbocostal arch) Right crus Left crus

Figure 8.3 Diaphragm. Note the three components of the diaphragm – sternal, costal and lumbar – and the three major apertures – vena caval foramen, esophageal hiatus and aortic hiatus

229

Ch 08 final.qxd

12/18/2006

15:26

Page 230

SURGERY FOR OVARIAN CANCER

trunks and esophageal branches of the left gastric vessels in addition to the esophagus. The margins of the aortic hiatus include the crura, the median arcuate ligament and the 12th thoracic vertebra. The aorta passes posterior to the diaphragm, not through it, via the aortic hiatus at the level of the 12th thoracic vertebra. The thoracic duct, the azygos vein and lymphatic vessels also pass through the aortic hiatus. Vascular supply and innervation

The diaphragm is supplied by the phrenic, musculophrenic and pericardiophrenic arteries. The phrenic arteries arise from the aorta. The musculophrenic and pericardiophrenic arteries are branches of the internal thoracic artery. The diaphragm is innervated by the phrenic nerve, which arises from the ventral rami of the third, fourth and fifth cervical vertebrae (C3, C4 and C5). The phrenic nerve provides the only motor supply to the diaphragm.

and separates the left half of the liver into a left medial sector and left lateral sector (Figure 8.4a). The posterior surface of the liver comprises the majority of the area lying within the borders of the coronary (or triangular) ligaments. Owing to its lack of a peritoneal covering, this area is referred to as the bare area and lies in direct contact with the diaphragm and the inferior vena cava (Figure 8.4b). The remainder of the liver surface is covered by a fibrous capsule (Glisson’s capsule). The inferior surface of the liver lies over the stomach, duodenum, the right colic flexure, the gallbladder, the right kidney and the right adrenal gland (Figure 8.4c). The superior and inferior surfaces are separated by a sharp inferior border. The liver is traditionally divided into a right and left half. The right half of the liver contains about 60–65% of the hepatic parenchyma, and the left half of the liver contains the remainder. Ligamentous attachments including the lesser omentum

Liver The liver is the largest gland in the body, weighing approximately 1.2–1.4 kg in females and occupying the majority of the right upper abdomen. The liver is wedge shaped with the base on the right and the apex on the left. The superior surface extends to approximately the fifth rib on the right and is separated from the lung by the diaphragm. The inferior margin is sharply defined and typically lies just below the costal margin. The pleura extends to the level of the tenth rib; consequently, needle biopsy of the liver should be performed below the tenth rib to avoid the pleural cavity. The liver executes multiple functions, including the secretion of bile; the synthesis of coagulation factors, vitamin A, albumin and other essential substances; the regulation of glucose/glycogen metabolism; detoxification of blood; hemoglobin breakdown; and storage of iron and copper. Liver surfaces

The superior surface of the liver is smooth, conforming to the shape of the diaphragm. The falciform ligament attaches the liver to the anterior abdominal wall

230

The position of the liver in the abdomen is maintained by a combination of the ligamentous attachments, continuity with the vasculature and intraabdominal pressure. The liver is attached to the anterior abdominal wall, abdominal viscera and diaphragm through the falciform, the gastrohepatic and hepatoduodenal, and the coronary (triangular) ligaments, respectively. The double-layered falciform ligament attaches the liver to the anterior abdominal wall and runs along the superior surface of the liver from the superior margin of the bare area to the umbilicus. The round ligament lies in the free edge of the falciform ligament and represents the obliterated umbilical vein or ligamentum teres of the liver. At the superior termination of the falciform ligament, the two reflections separate to form the coronary ligaments which surround the bare area (Figures 8.4a). The right-sided layer of the falciform ligament becomes the anterior layer of the coronary ligament, which travels to the right and becomes the right triangular ligament. The inferior surface of the right triangular ligament extends medially and becomes the

Ch 08 final.qxd

12/18/2006

15:26

Page 231

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

Inferior vena cava Caudate lobe Left triangular ligament

Right triangular ligament Left lobe Right lobe

Gallbladder a Falciform ligament Caudate lobe Inferior vena cava Left triangular ligament

Coronary ligament, anterior layer Bare area

Gastric impression Lesser omentum, cut edges (hepatogastric ligament)

Area related to suprarenal gland Right triangular ligament

Caudate process Lesser omentium, anterior layer (hepatoduodenal ligament)

Coronary ligament, posterior layer Renal impression

Ligamentum teres Quadrate lobe Gallbladder b Quadrate lobe Ligamentum teres

Gallbladder

Left lobe Portal vein Papillary process Caudate process Gastric impression Esophageal groove

Colic impression Duodenal impression Renal impression Right lobe Suprarenal impression Non-peritoneal surface c

Hepatic artery proper

Bare area

Caudate lobe Inferior vena cava

Fissure for ligamentum venosum

Coronary ligament Bile duct

Figure 8.4 (a) Superior surface of the liver. Note the location of ligaments. The falciform ligament separates the liver into anatomic right and left lobes. (b) Posterior surface of the liver. Note the location of the ligaments including the lesser omentum. The bare area is in direct contact with the diaphragm and inferior vena cava. (c) Inferior surface of the liver. Note the ‘H’ formed by the inferior vena cava, gallbladder, ligamentum venosum, ligamentum teres and porta hepatis

231

Ch 08 final.qxd

12/18/2006

15:26

Page 232

SURGERY FOR OVARIAN CANCER

posterior layer of the coronary ligament, which is continuous with the left triangular ligament formed by the left-sided reflection of the falciform ligament (Figure 8.4b). The left triangular ligament is continuous with the lesser omentum, which travels to the stomach as the hepatogastric ligament and to the duodenum as the hepatoduodenal ligament.

Vascular system and biliary ducts Afferent blood supply to the liver is from two sources. Approximately one-quarter of the afferent flow arises from the hepatic artery, which is oxygen rich. The hepatic artery is a branch of the celiac trunk and passes along the upper portion of the pancreas to give off the gastroduodenal artery as it travels posterior and superior to the duodenum. The hepatic artery runs in the hepatoduodenal ligament, located anterior to the portal vein and to the left of the bile duct, before dividing into right and left hepatic arteries. The cystic artery to the gallbladder typically arises from the right branch of the hepatic artery. Nutrient-rich blood, draining directly from the intestines, arises from the portal vein and provides approximately 75% of the afferent blood flow to the liver. The portal vein forms at the junction of the superior mesenteric and splenic veins and passes posterior to the neck of the pancreas and the second portion of the duodenum. In the hepatoduodenal ligament, the portal vein runs posterior to the bile duct and the hepatic artery before dividing into right and left branches. The right branch has a larger diameter and typically is deep into the hepatic parenchyma, but the left branch is longer and is located more superficially until it divides into the various segmental branches. Efferent blood flow from the liver is through the hepatic venous system. The right, middle and left hepatic veins emerge from the posterior surface of the liver and empty blood directly into the inferior vena cava. Surgical control of bleeding arising from these veins can be challenging, because of the lack of a substantial extrahepatic component. Bile flows through biliary capillaries and then to interlobular bile ducts, which continue to merge to

232

eventually form the right and left hepatic ducts. The right and left hepatic ducts fuse to form the common hepatic duct, which joins the cystic duct from the gallbladder to become the common bile duct. Bile is then delivered to the small intestine to aid in the absorption of fat and other nutrients. The portal triad structures include the portal venous branches, the biliary ducts and the hepatic arterial branches, and run together within the substance of the liver. The hepatic venous drainage follows a different branching system, acting as tributaries to the inferior vena cava. Segmental anatomy

The surgical anatomy of the liver is based on a system of eight major segments (excluding the caudate lobe) as described by Couinaud (Figure 8.5a). The principal line of demarcation (Cautlie’s line) for the surgical right and left sides extends obliquely from the middle of the gallbladder fossa to the center of the inferior vena cava between the right and left main hepatic veins. The anatomical left side of the liver is divided into medial and lateral sectors along the line of the falciform ligament. Each of these sectors is also divided into a superior and an inferior area, yielding four anatomical segments (segments 2, 3, 4A and 4B). The left lateral sector contains segments 2 and 3, while the left medial sector consists of segments 4A and 4B. The anatomical right side of the liver is partitioned into anterior and posterior sectors along a line extending from the anteroinferior edge of the liver and running posteriorly and superiorly. As for the left side, the sectors on the right side of the liver are also split into superior and inferior areas, yielding four anatomical segments (segments 5, 6, 7, 8). The right posterior sector contains segments 6 and 7, while the right anterior sector contains segments 5 and 8. The caudate lobe is often referred to as segment 1. There are important differences in the distribution of the portal triad structures (portal vein, bile ducts and hepatic artery) and that of the hepatic venous drainage. The hepatic veins run between the segments (Figure 8.5b). The right hepatic vein lies in the

Ch 08 final.qxd

12/18/2006

15:26

Page 233

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

Cautlie's line Left

Right

Superior

2 8

4A Inferior

3

Superior 7

Lateral

4B 5 Inferior

Medial 6

Gallbladder fossa

Anterior a

12

Falciform ligament

Posterior Cautlie's line

11

13 18 16 14 10 17

3 4

2

3

Left lobe

9

5

2

5

4 1 10 11 12 9

6 15

6 7

8 7 Right lobe

1 20

19

b 1 Portal vein 2 Right anterior inferior portal vein 3 Right anterior superior portal vein 4 Right posterior inferior portal vein 5 Right posterior superior portal vein 6 Left branch of portal vein 7 Left medial inferior portal vein 8 Left lateral inferior portal vein 9 Left lateral superior portal vein 10 Left medial superior portal vein

8 c

11 Inferior vena cava 12 Right hepatic vein 13 Left hepatic vein 14 Middle hepatic vein 15 Left middle inferior hepatic vein 16 Left middle superior hepatic vein 17 Left lateral inferior hepatic vein 18 Left lateral superior hepatic vein 19 Round ligament 20 Gallbladder

1 Common hepatic duct 2 Right hepatic duct 3 Posterior segmental duct 4 Anterior segmental duct 5 Left hepatic duct 6 Lateral segmental duct

7 Medial segmental duct 8 Gallbladder 9 Common hepatic atery 10 Left hepatic artery 11 Right hepatic artery 12 Cystic artery

Figure 8.5 The liver is divided into eight anatomic segments that form the basis for surgical resection. (a) Segmental anatomy showing the medial–lateral sectoral demarcation of the left side of the liver and the anterior posterior sectoral demarcation of the right side. Each sector is divided into a superior and inferior segment. Cautlie’s line distinguishes the right and left sides of the liver. (b) Portal and hepatic venous circulations of the liver. (c) Hepatic arterial circulation and biliary drainage system

major cleft between the anterior and posterior sectors of the right side. The left hepatic vein primarily drains the left lateral sector, while the middle hepatic vein straddles the left medial sector and the right side of the liver. It is critically important to preserve these venous drainage channels during segmental resec-

tions of the liver parenchyma, since occlusion of the hepatic veins results in necrosis of the entire segment or area involved. In general, the portal triad structures bifurcate in a serial fashion and lead directly into the eight major hepatic segments (Figure 8.5b–c). A notable exception is the paraumbilicalis of the left

233

Ch 08 final.qxd

12/18/2006

15:26

Page 234

SURGERY FOR OVARIAN CANCER

hepatic branch of the portal vein, which straddles the division of medial and lateral sectors of the left side. The right posterior sector, right anterior sector and left lateral sector each contain two segments, which are divided according to the portal inflow (i.e. there are usually separate pedicles to each segment). Although not shown in Figure 8.5, segment 1 (caudate lobe) has portal inflow from both the right and the left sides and drains directly into the inferior vena cava.

of the gallbladder is in direct contact with the liver. The inferior surface of the body and the fundus are covered by peritoneum. The neck of the gallbladder tapers and makes an S curve as it becomes the cystic duct (Figure 8.6). The cystic artery supplies the gallbladder. In the majority of patients, the right hepatic artery supplies the cystic artery as it travels between the cystic duct and the common hepatic duct. However, variations in this relationship are common. Venous drainage from the fundus and body travels directly to the liver. Veins from the neck drain either through the portal venous system or directly to the liver. As described above, the right and left bile hepatic ducts fuse to form the common hepatic duct, which travels with the hepatic artery and portal vein in the free edge of the hepatoduodenal ligament, anterior and medial to the epiploic foramen. In the majority of patients, the common hepatic duct runs to the right of the hepatic artery, which travels anterior to the portal vein. Bile travels to and from the gallbladder

Gallbladder The gallbladder is a pear-shaped organ that stores bile. The gallbladder lies in a fossa along the inferior surface of the liver and is divided into the fundus, body and neck. The fundus is the widest portion of the gallbladder and extends beyond the anterior margin of the liver. From the fundus, the gallbladder then travels superiorly, posteriorly and to the left. Attached by connective tissue, the superior surface of the body

Gallbladder

Ligamentum teres hepatis Quadrate lobe

Cystic artery Cystic artery (superficial branch) Left hepatic duct

Cystic duct

Left portal vein

Right portal vein Common hepatic duct

Left hepatic artery Ligamentum venosum

Right hepatic artery

Inferior vena cava

Common bile duct

Caudate lobe Portal vein Hepatic artery

Figure 8.6 The gallbladder and the associated structures of the porta hepatis

234

Ch 08 final.qxd

12/18/2006

15:26

Page 235

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

through the cystic duct, which joins the common hepatic duct to form the common bile duct. Collectively, the common bile duct, common hepatic artery and portal vein form the porta hepatis. The common bile duct then travels posterior to the second portion of the duodenum before joining the pancreatic duct within the head of the pancreas to form the ampulla of Vater. The ampulla opens into the second portion of the duodenum at the major duodenal papilla. This opening is controlled by the sphincter of Oddi.

Duodenum The duodenum may be encountered during the course of mobilization of the right side of the liver after taking down the hepatic flexure of the colon and reflecting it medially, to approach hepatic resection or

posterior diaphragm disease. Therefore, the ovarian cancer surgeon should be familiar with its topography. The duodenum is approximately 25 cm in length, follows a C-shaped course around the head of the pancreas and is partitioned into four portions (Figure 8.7). The first part of the duodenum travels to the right from the pylorus toward the neck of the gallbladder. The majority of this portion is covered by peritoneum. The proximal half has a short mesentery and has attachments to the omentum and the hepatoduodenal ligament. The distal half has no mesentery. The portal vein, inferior vena cava, gastroduodenal artery and bile duct are located immediately behind the first part of the duodenum. The second part of the duodenum angles downward, running parallel to the inferior vena cava, with the right kidney, ureter and renal vessels situated posteriorly. Laterally, the second part of the duodenum abuts the right side

Portal vein Superior (first) part of duodenum Hepatic Common artery Ampulla of Vater bile duct (orifice of common bile duct and pancreatic duct)

Accessory pancreatic duct Pancreatic duct

Ascending (fourth) part of duodenum

Descending (second) part of duodenum Horizontal (third) part of duodenum

Superior mesenteric artery and vein

Figure 8.7 Anatomy of the duodenum. The duodenum may be encountered during hepatic mobilization for liver resection or removal of posteriorly located diaphragmatic disease

235

Ch 08 final.qxd

12/18/2006

15:26

Page 236

SURGERY FOR OVARIAN CANCER

of the liver. The bile duct and main pancreatic duct join to form the ampulla of Vater and perforate the medial wall of this portion of the duodenum. The third part of the duodenum is retroperitoneal and extends horizontally from right to left beneath the superior mesenteric artery. Finally, the fourth part of the duodenum ascends retroperitoneally to the duodenojejunal junction (ligament of Trietz).

Right kidney and adrenal gland Involvement of the right kidney and adrenal gland by metastatic ovarian cancer is exceedingly rare; nevertheless, these structures may be encountered within the retroperitoneum of the right upper abdomen during hepatic mobilization or resection of diaphragmatic disease (Figure 8.8). The right kidney measures approximately 11.5 cm and is located approximately 2 cm lower than the left kidney. The entire kidney is retroperitoneal and is surrounded by perinephric fat. The renal fascia (Gerota’s fascia), a layer of fibro-

Inferior vena cava; hepatic veins

Right adrenal gland Suprarenal branches of right renal artery and vein Accessory renal artery

Right renal vein and artery Right ureter Perirenal fat

Figure 8.8 Anatomy of the right kidney and adrenal gland

236

areolar tissue, encompasses the kidney, the perinephric fat and the adrenal glands. The superior pole of the kidney is protected by the 12th rib and lies closer to the midline than the inferior pole. The anterior surface of the right kidney is in contact with the right adrenal gland, the liver, the descending portion of the duodenum, the hepatic flexure of the colon and the small bowel. The peritoneal reflection separating the liver from the right kidney is known as Morison’s pouch and is a frequent site of ovarian cancer metastasis. The diaphragm separates the posterior surface of the right kidney from the pleura and the 12th rib. The right renal artery arises from the aorta at the level between the 1st and 2nd lumbar vertebrae and travels posterior to the left renal vein, the inferior vena cava and the right renal vein. The right adrenal gland is pyramidal in shape, with the base lying on the superiomedial surface of the right kidney. The diaphragm lies posteriomedially, while the inferior vena cava and the bare area of the

Ch 08 final.qxd

12/18/2006

15:26

Page 237

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

liver lie anteriorly. The adrenal gland receives its blood supply from the aorta, the inferior phrenic artery and the renal artery. The right adrenal vein drains directly into the inferior vena cava.

SURGICAL PROCEDURES Surgical incision and approach Primary surgery for ovarian cancer is typically performed through a midline vertical incision. Extension of this incision lateral to the xyphoid process to the level of the sternum facilitates exposure of the right upper quadrant. Surgery for an isolated recurrence of ovarian cancer limited to the right upper quadrant can be approached through a subcostal incision, but this approach may compromise access to other locations, especially the pelvis. The subcostal incision can be extended to the contralateral abdomen (chevron) or cephalad to one side of the xyphoid process for additional maneuverability within the right upper abdomen (Figure 8.9). A self-retaining retractor, such as a Bookwalter or Omni retractor, allows maximal upward displacement of the costal margin and greatly improves exposure to the right upper abdomen.

described superficial or full-thickness resection of diaphragm disease in 14 patients with ovarian cancer, resulting in optimal residual disease in 13 cases (93%) with acceptable morbidity.7 More recently, Feitoza et al. reported their experience at the Mayo Clinic with 41 patients who underwent diaphragm resection as part of ovarian cancer cytoreductive surgery. These investigators found that the risk of complications with excision of disease from the diaphragm was similar to that seen with other radical cytoreductive procedures.4 The effectiveness of diaphragm resection to improve survival in patients with stage III and IV ovarian cancer has not been widely addressed. Evaluating a group of 11 ovarian cancer patients with involvement of the diaphragm, Kapnick et al. suggested that, despite diaphragmatic resection, patients with primary disease penetrating the diaphragm continue to have poor survival.8 Nevertheless, the survival impact of an extirpative operation on the

Diaphragm disease Metastatic spread to the diaphragm is a common occurrence among women with advanced-stage ovarian cancer. Griffiths and Finkler reported diaphragm involvement in 18% of patients with stage III disease and 41% of patients with stage IV disease.3 In a review of 109 patients with stage I–IV disease, Sakai et al. reported that 28% of cases (30 of 109 patients) had pathologically documented diaphragmatic disease.2 The right hemidiaphragm is more often involved than the left side, owing to the propensity of ovarian cancer cells to accumulate in Morison’s pouch and the subhepatic recess. Several studies have evaluated the safety and feasibility of cytoreductive surgery for metastatic ovarian cancer involving the diaphragm. In 1989, Montz et al.

Figure 8.9 Subcostal incision. The subcostal incision may be extended to the contralateral abdomen or superiorly to one side of the xyphoid process for additional exposure

237

Ch 08 final.qxd

12/18/2006

15:26

Page 238

SURGERY FOR OVARIAN CANCER

diaphragm is most appropriately interpreted within the context of the contribution such a procedure will make to the overall surgical result (i.e. residual disease). Given the preponderance of data indicating improved survival rates associated with optimal residual disease for patients with both stage III and stage IV ovarian cancer in general, every reasonable attempt should be made to address diaphragmatic metastases by one or more of the following techniques.1,9–12 Diffuse involvement by small nodules is typically approached most successfully by diaphragm stripping, whereas isolated masses may be excised individually by peritoneal or full-thickness resection. Ablative techniques with electrocautery, the cavitron ultrasonic surgical aspirator (CUSA)13,14 or argon beam15,16 are alternative approaches. Exposure may be a significant impediment to optimal cytoreduction of diaphragmatic disease. A generous incision and adequate mobilization of the liver are the two essential components to ensure appropriate access to the right diaphragm and Morison’s pouch. Right hepatic mobilization begins by dividing the round ligament between suture ligatures, as it will occasionally contain patent vessels. The falciform ligament is then taken down with electrocautery toward its apex. The appropriate plane of dissection is immediately adjacent to the hepatic parenchyma. The bifurcation of the falciform ligament is carefully incised, carrying the dissection into the anterior layer of the coronary ligament and down the posterior surface of the liver. It is critical to maintain the dissection superficial to the right and left hepatic veins and inferior vena cava, which lie immediately below the coronary ligament in this region (Figure 8.10). Often, the main right phrenic vein can be seen draining towards the confluence of these vessels. The liver is then drawn medially, exposing the right paracolic gutter and Morison’s pouch. The right triangular ligament is identified and incised with electrocautery (Figure 8.11a). The appropriate plane of dissection is juxtaposed to the hepatic parenchyma. With continued medial traction, the bare area of the liver is mobilized off the diaphragm, the right kidney and adrenal gland,

238

providing optimal exposure for diaphragm peritonectomy or full-thickness resection (Figure 8.11b). Occasionally, hepatic mobilization will require dissecting the liver off the anterior surface of the inferior vena cava. With continued medial traction on the liver, the areolar tissue surrounding the caval wall is released with sharp dissection (Figure 8.12). Care should be taken to control venous branches directly draining from the liver into the inferior vena cava. The entire right liver can now be mobilized into the midline with full exposure of the right diaphragm. If the left diaphragm is involved with disease, a similar approach to mobilization of the left half of the liver can be performed. The round ligament and falciform ligament are divided. The dissection is taken along the left coronary ligament to the left triangular ligament, exposing the bare area of the liver. Special attention needs to be given to the left and middle hepatic veins as the dissection approaches the inferior vena cava. The left and middle hepatic veins usually join into a common trunk before emptying into

Right hepatic vein

Figure 8.10 Hepatic mobilization. The round ligament has been divided and the falciform ligament taken down to its apex. The coronary ligament is carefully incised, exposing the underlying right hepatic vein and inferior vena cava

Ch 08 final.qxd

12/18/2006

15:26

Page 239

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

a

b

Figure 8.11 Hepatic mobilization. (a) The right triangular ligament is incised by electrocautery, with the liver on medial traction. (b) The dissection is continued into Morison’s pouch, reflecting the bare area of the liver from the underlying diaphragm and optimizing exposure

Right hepatic vein Inferior vena cava Diaphragm

Hepatic veins (direct branches)

Figure 8.12 Hepatic mobilization. The bare area of the liver is mobilized from the anterior surface of the inferior vena cava, with care taken to isolate and ligate venous branches draining directly from the hepatic parenchyma to the vena cava

the vena cava (Figure 8.5b). Division of the gastrohepatic ligament allows full mobilization of the left half of the liver.

Diaphragm peritonectomy

The tendency of ovarian cancer metastasis to respect the peritoneal line of demarcation is commonly observed for disease involving the diaphragmatic surface. This is especially true over the thick muscular areas of the diaphragm. Consequently, peritonectomy (or peritoneal ‘stripping’) is the preferred technique for cytoreduction of extensive small-volume disease covering significant surface area of the diaphragm. There are variations in the surgeon’s choice of instrumentation to accomplish peritonectomy including using sharp scissor dissection, electrocautery, the argon beam coagulator and even the ultrasonic dissector to find the proper plane. It is the authors’ preference to use either the electrocautery device or the argon beam coagulator. An abdominal wall retractor blade should be secured on the self-retaining retractor so as to achieve maximal elevation of the right costal margin (Figure 8.13a). Placing the patient in reverse Trendelenburg position allows the liver to descend into the abdominal cavity (following mobilization) and may improve exposure. In most cases, it is easiest to define the subperitoneal plane by initiating the dissection in an area that is free of gross disease. The reflection of the diaphragm peritoneum along the costal margin is incised transversely, developing a

239

Ch 08 final.qxd

12/18/2006

15:26

Page 240

SURGERY FOR OVARIAN CANCER

broad front of free peritoneal edge that is placed on counter traction with Allis or Kocher clamps (Figure 8.13b). If difficulty is encountered along the costal margin, the proper plane can almost always be identified by beginning the dissection more laterally, where the right paracolic gutter merges with Morison’s pouch. Electrocautery is used to separate the diaphragm peritoneum from the underlying muscle, moving posteriorly until the margins of the divided coronary and right triangular ligaments are reached and the specimen completely excised (Figure 8.13c–d).

On occasion, ovarian cancer metastatic to the diaphragm may be densely adherent to the central tendinous portion of the muscle, and incidental entry into the pleural cavity may be unavoidable. For this reason, it is always advisable to secure mechanical ventilation via endotracheal intubation prior to undertaking resection of diaphragmatic disease. Once the pleural cavity has been entered, the anesthesiologist should be notified and the resection completed prior to repairing the defect. Defects in the diaphragm muscle measuring up to several centimeters in size can be repaired with a simple closure of full-thickness

a

b

c

d

Figure 8.13 Diaphragm peritonectomy. (a) The self-retaining retractor has been positioned to provide maximal elevation of the costal margin. (b) The diaphragm peritoneum is incised along the costal margin, developing a broad front of dissection in the subperitoneal plane. (c) The diaphragm peritoneum is placed on downward traction, exposing the plane of dissection at the interface with the muscular surface. (d) The dissection is carried posteriorly to the peritoneal reflection of the coronary and right triangular ligaments

240

Ch 08 final.qxd

12/18/2006

15:26

Page 241

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

interrupted or figure-of-eight stitches of large-caliber non-absorbable suture (1-0 polypropylene or silk). As an alternative to chest tube placement, the resulting pneumothorax can be evacuated intraoperatively prior to placing the final stitch of the diaphragmatic closure. A 14-Fr Robinson catheter, with additional fenestrations cut towards the tip, is placed directly into the pleural cavity through the partially closed diaphragmatic defect. Final closure stitches are placed on either side abutting the catheter and held taught but not yet tied (Figure 8.14). Several large-volume ventilations are performed by the anesthesiologist and a Valsalva maneuver initiated while gentle suction is maintained on the Robinson catheter. The catheter is quickly removed as the remaining closure sutures are drawn taught and tied. Alternatively, the end of the Robinson catheter can be placed under a column (20 cm) of water to provide a water seal, as the chest cavity is evacuated of air, to produce the same effect. Resection of the diaphragm

A full-thickness resection of a portion of the diaphragm may be required if metastatic ovarian can-

Figure 8.14 Evacuation of pneumothorax after incidental entry into the pleural cavity. A Robinson catheter is placed through the partially completed diaphragmatic repair and placed on suction during positive-pressure ventilation. The catheter is quickly withdrawn as the final sutures are drawn taught and tied, sealing the pleural cavity

cer is deeply invasive or there is significant involvement of the central tendon, where the peritoneum is less well defined. Mobilization of the liver and peritonectomy are carried out, as described above, until the area of diaphragm requiring full-thickness resection is adequately defined. The diaphragm muscle is incised, with entry into the pleural cavity, and the tumor mass or diseased surface circumscribed with electrocautery or scissor dissection. Care should be taken to gain adequate hemostasis from the branches of the phrenic artery and vein, which can be oversewn with figure-of-eight stitches of large caliber (1-0) suture if necessary. In most cases, diaphragmatic defects as large as 5–6 cm can be closed primarily with interrupted or figure-of-eight stitches using the technique described above. Rarely, a large defect will require placement of a prosthetic material (polytetrafluoroethylene, woven Dacron®, or Marlex® mesh) to effect a satisfactory closure. If bacterial contamination is a concern because of concurrent cytoreductive procedures (e.g. bowel resection with fecal spill), biomaterials (e.g. cadaveric fascia – Alloderm®) may be used instead. The prosthetic materials or biomaterials are sutured to the edges of the diaphragmatic defect circumferentially with interrupted stitches of largecaliber (1-0) non-absorbable sutures. Diaphragmatic resections involving the medial and central aspects carry a risk of injury to the main branch or branches of the phrenic nerve, which will result in unilateral diaphragm paralysis. Usually, accessory muscle recruitment and an intact left diaphragm will allow patients with an adequate ventilatory status preoperatively to tolerate this without difficulty. As an alternative to sharp dissection with scissors or electrocautery, automated stapling devices (e.g. gastrointestinal (GIA), thoracoabdominal (TA)) can be utilized simultaneously to resect a large tumor metastasis and to lay down a hemostatic staple line that effectively seals the pleural cavity (Figure 8.15). This technique is generally only applicable to a large tumor mass that has a relatively narrow (1–2 cm) pedunculated base. GIA-type stapling devices can be

241

Ch 08 final.qxd

12/18/2006

15:27

Page 242

SURGERY FOR OVARIAN CANCER

used both to resect deeply involved areas of the diaphragm, and to perform primary closure at the same time. Only stapling devices with ‘thick tissue’ loads (4.8 mm) should be used for this purpose. Additional security of the closure is provided by oversewing the staple line with a running, locking stitch of large-caliber (1-0) non-absorbable suture. Postoperative management and complications

a

b

c

Figure 8.15 Diaphragmatic resection. (a) Large pedunculated tumor mass on the right diaphragm. (b) The automated gastrointestinal stapling device can be used as an alternative to sharp

dissection

for

full-thickness

resection;

multiple

applications of the stapler may be required. (c) Excised tumor specimen with staple line

242

In general, patients who have undergone diaphragmatic resection and primary closure can be managed without a right chest tube. Care should be taken to evacuate as much of the pneumothorax as possible prior to full repair of the diaphragm, as described previously. A postoperative chest radiograph should be performed, which will often reveal a small apical pneumothorax. In the absence of an air leak from the lung parenchyma, spontaneous resolution of a small pneumothorax can be expected and should be monitored with daily radiographs. Additionally, administration of supplemental oxygen in the immediate postoperative period will speed the resolution of the pneumothorax via the second gas effect. The second gas effect takes advantage of the fact that air in the pneumothorax contains approximately 70% nitrogen, which is highly diffusible across the lung. If the alveolar concentration of nitrogen is kept low with supplemental oxygen, then the pneumothorax will decrease more rapidly, owing to escape of the nitrogen down the concentration gradient. On occasion, a right chest tube will be required because of the presence of a large right pleural effusion preoperatively, a major resection of the right diaphragm is undertaken, a larger right pleural effusion develops postoperatively, or the patient develops a worsening pneumothorax postoperatively. If the need for a chest tube can be anticipated, it is preferable to place it intraoperatively, while the patient is under general anesthesia. For right chest tube placement, the right hemithorax should be adequately prepped and draped. If the patient is not under general anesthesia, a local anesthestic should be liberally injected into the soft tissue surrounding the planned

Ch 08 final.qxd

12/18/2006

15:27

Page 243

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

insertion site. A 2–3 cm incision is created at the midaxillary line in the sixth intercostal space, parallel to the axis of the ribs, to allow insertion of a finger and chest tube concurrently. A Kelly clamp is then used to create a tract over the sixth rib and advanced into the pleural space directly above the rib to avoid the neurovascular bundle lying on the inferior surface of each rib. The tract is dilated by spreading the Kelly clamp, and a finger is placed through the tract into the pleural space to confirm location and gently separate any adhesions to the lung. A 28-Fr chest tube is then advanced into the tract with the aid of the Kelly clamp to direct the tip posteriorly and towards the apex of the lung. Care should be taken to ensure that the last side hole of the chest tube is well within the pleural cavity. The tube is attached to a standard evacuation device (e.g. Pleur-evac), sutured in place, and an occlusive dressing applied. The chest tube can be removed in several days, provided there is no evidence of an air leak and re-accumulation of fluid in the right chest is not excessive (less than 200 ml of drainage per day). In experienced hands, diaphragm resection is associated with an acceptable morbidity and mortality rate comparable with other extensive procedures involved in the ovarian cancer cytoreduction. Feitoza et al. reported a 20% (8 of 41 patients) complication rate in ovarian cancer patients undergoing resection of the diaphragm at the Mayo Clinic. Morbidity and mortality included pneumothorax in two patients, symptomatic pleural effusion in four patients, subphrenic abscess in one patient, and gastropleural fistula in one patient. This last case was associated with the only death.4

Cytoreduction of liver disease As with the diaphragm, the liver surface and parenchyma are often involved by both primary and recurrent ovarian cancer. Sakai et al. evaluated 109 patients with stage I–IV primary ovarian cancer; 10% (11 of 109 patients) had liver or splenic capsule involvement.2 In their investigation, Bristow et al. reported that 44% (37 of 84 patients) of stage IV

ovarian cancer subjects had liver parenchymal involvement.1 Examining a similar population, Bonnefoi et al. reported that 26% (50 of 192 patients) of stage IV ovarian cancer subjects had involvement of liver parenchyma.17 Finally, Akahira et al. found 15% of 225 stage IV ovarian cancer patients to have liver metastases.9 Several studies have addressed the feasibility of liver resection as part of a maximal cytoreductive surgical effort for metastatic ovarian cancer. Chi et al. evaluated 12 patients who underwent liver resection for gynecologic malignancies; seven of these patients (58%) had ovarian cancer.6 The operations performed included a trisegmentectomy in four patients, a lobectomy in four patients, a segmentectomy in three patients and a wedge resection in one patient. Morbidity was acceptable, with one patient developing a low-output enterocutaneous fistula. In a followup report, Chi et al. described a major hepatic resection performed at the time of interval cytoreductive surgery following neoadjuvant chemotherapy for advanced ovarian cancer.18 Yoon et al. evaluated 24 women with recurrent ovarian or fallopian tube cancer including liver metastases, who underwent optimal debulking of their disease. The hepatic resections included two trisegmentectomies, two lobectomies, 17 segmentectomies and three wedge resections. These operations were performed with an acceptable complication rate.19 Investigators at the Mayo Clinic identified 26 patients requiring hepatic resection for recurrent ovarian carcinoma. Optimal debulking was achieved in 21 patients (81%). Morbidity included four patients who required transfusion of more than 4 units of packed red blood cells, one wound infection and 1 small bowel perforation.5 Multiple reports have observed that, despite the presence of liver involvement, optimal cytoreductive surgery for primary or recurrent ovarian cancer is associated with improved survival rates. Naik et al. evaluated 27 subjects with stage IV ovarian cancer on the basis of liver metastases.20 Multivariate analysis demonstrated optimal debulking to have a significant association with improved survival. Of 84 women

243

Ch 08 final.qxd

12/18/2006

15:27

Page 244

SURGERY FOR OVARIAN CANCER

with stage IV ovarian cancer, Bristow et al. noted that 37 patients had liver involvement.1 Optimal cytoreduction of both hepatic and extrahepatic disease was performed in 16% (six of 37 patients). These patients had improved outcome compared to patients who did not have optimal debulking of intrahepatic disease. In the investigation by Yoon et al., all 24 patients underwent optimal cytoreductive surgery for recurrent ovarian or fallopian tube cancer.19 Median survival was 62 months after hepatic resection. Merideth et al. reported significantly improved survival in patient with recurrent ovarian cancer involving the liver who were undergoing optimal cytoreductive surgery compared to those with suboptimal debulking (27.3 vs. 8.6 months).5

Superficial liver disease In the usual case of advanced-stage disease, ovarian cancer is disseminated throughout the abdominal cavity by spreading along peritoneal surfaces and, therefore, tends to involve only the liver surface. In these instances, cytoreduction of liver surface disease can be accomplished by simple resection using conventional techniques: excision with the CUSA, ablation with the argon beam coagulator, or a combination. For simple resection, electrocautery is used to demarcate a 1-cm margin of normal-appearing tissue around the surface implant by circumferentially incising Glisson’s capsule. The tumor implant resection margin is placed on counter traction with pickups or an Allis clamp, and electrocautery dissection continued within the substance of the superficial hepatic parenchyma (Figure 8.16a–c). The resection bed of the liver surface is then treated with the argon beam coagulator to secure hemostasis (Figure 8.16d). When excising surface implants along the reflection of the coronary ligament, it is important to ascertain the location of the vena cava and hepatic veins to avoid injury to the vessel wall. Resection of posterior surface disease in the region of Morison’s pouch is facilitated by dividing the right triangular ligament and completely mobilizing the right side of the liver.

244

Parenchymal liver disease Although uncommon, ovarian cancer may involve the liver parenchyma either through extensive centripetal growth from a surface metastasis or through hematogenous spread. In these instances, a more substantial resection of hepatic parenchyma may be indicated, depending on the extent and location of disease. Hepatic resections can be divided into three general categories: non-anatomic wedge resections; major hepatic resections; and lesser resections. Nonanatomic wedge resections are performed to remove small peripherally located lesions and are based on lesion size rather than vascular supply or biliary drainage. Major hepatic resections involve removal of multiple segments (see below) of the liver and may be associated with a significant impact on hepatic physiology. Lesser resections, also called anatomic resections or segmentectomies, are based on the anatomic descriptions of Couinaud and involve removal of individual hepatic segments, preserving more functional liver parenchyma than in the major hepatic resections. Intraoperative ultrasonography can be helpful in delineating the site of parenchymal lesions intraoperatively and identifying specific intrahepatic vascular and biliary structures, to enable a more precise hepatic resection to be performed. After partial or complete mobilization of the liver, saline is liberally applied over the liver to improve acoustic imaging prior to the placing of the ultrasound probe.

Non-anatomic wedge resection A non-anatomic wedge resection of the hepatic parenchyma is indicated for lesions measuring up to 4 cm in diameter that are peripherally located and easily accessible, or when there are multiple resectable lesions located in both sides of the liver. Isolation of specific segmental vascular and biliary channels is unnecessary. Excessive blood loss is unusual unless the lesion is located in close proximity to a major blood vessel, in which case a more anatomic resection should be considered.

Ch 08 final.qxd

12/18/2006

15:27

Page 245

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

a

b

c

d

Figure 8.16 Resection of superficial tumor implants on the liver surface. (a) The tumor implant is circumscribed with electrocautery and the resection margin raised. (b and c) Counter traction on the tumor implant provides for dissection within the superficial hepatic parenchyma. (d) Hemostasis of the resection bed is achieved with the argon beam coagulator

In performing a wedge resection, the liver should be inspected in its entirety to ensure that additional metastatic lesions are not missed. Normally, this should include division of the falciform, coronary and right triangular ligaments so that the posterior surface of the liver can be both visually and palpably inspected. The target lesion is then superficially circumscribed with electrocautery, leaving a 1–2 cm margin of normal-appearing tissue (Figure 8.17a). Parenchymal liver metastases are generally spherical but tend to be wider than they are deep. A series of full-thickness mattress stitches of large caliber (1 or 1-0) chromic suture on a gently curved long needle may be placed at this junction for additional hemostasis. These stitches are placed parallel to, but out-

side, the circumscribing cautery incision and overlap one another by 0.5–1.0 cm (Figure 8.17b). The sutures are tied so as to provide compression of the liver parenchyma without lacerating deeply into the tissue. The free edge of liver adjacent to the target lesion is grasped with an Allis clamp and placed on counter traction as the electrocautery, argon beam coagulator, or CUSA is used to divide the parenchyma, maintaining a three-dimensional margin of resection of at least 1–2 cm (Figure 8.17c). When vessels or biliary ducts of 2–3 mm are encountered, they should be secured with suture ligatures or vascular hemoclips (Figure 8.17d). The completed resection bed can be treated with the argon beam coagulator for additional hemostasis. If bile leakage is observed, the damaged

245

Ch 08 final.qxd

12/18/2006

15:27

Page 246

SURGERY FOR OVARIAN CANCER

Falciform ligament

Lesion

Cautery

a

b

Mattress sutures

d

c

e

Figure 8.17 Non-anatomic wedge resection. (a) The target lesion is circumscribed with elctrocautery, leaving a 1–2-cm margin of normal-appearing tissue. (b) Full-thickness mattress sutures are placed parallel to, but outside, the circumscribing incision, for added hemostasis. (c) The liver parenchyma is divided with electrocautery. (d) Hepatic vessels and ducts larger than 2–3 mm are individually ligated and divided. (e) Completed resection bed

duct should be sought and ligated. Routine closed suction drainage is unnecessary in the majority of cases, but should be considered if the surgical field is not completely dry.

Major and lesser hepatic resection: general principles A central tenet in performing major hepatic resections or the lesser (anatomic) resections is the

246

isolation of the vascular structures in the portal triad supplying the parenchyma to be removed. Typically, control of hepatic inflow is obtained initially, followed by control of hepatic outflow, and lastly the hepatic parenchyma is transected. The relevant portal triad pedicle can be isolated within the hilum of the liver and divided extrahepatically or, alternatively, it can be isolated intrahepatically by the performance of a series of limited hepatotomies to eventually reach the

Ch 08 final.qxd

12/18/2006

15:27

Page 247

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

pedicle. Ligation of the appropriate pedicle is associated with a change in color of the hepatic parenchyma and the line demarcating the specific anatomic segments to be removed, in addition to controlling vascular inflow. Before division of a pedicle, it is often helpful to apply an atraumatic vascular clamp to confirm that the pedicle does indeed supply the appropriate segment(s) of the liver inclusive of the planned resection. The individual components of each pedicle (portal vein, hepatic artery, biliary duct) may be secured and divided by traditional suture ligation or by one of the automated stapling devices utilizing a vascular staple load (2.5 mm). Adequate outflow control during a major hepatic or anatomical resection requires isolation and ligation of the appropriate hepatic vein(s) prior to transection of the liver parenchyma. In hepatectomy, the corresponding right or left hepatic vein is ligated, depending on which side of the liver is being removed. Additionally, the middle hepatic vein can be ligated in conjuction with removal of either the right or the left liver. Some hepatic surgery centers also use total vascular isolation of the liver in which total inflow occlusion as well as control of the infrahepatic and suprahepatic inferior vena cava is obtained. This technique, however, can be associated with significant hemodynamic instability and is unnecessary in the vast majority of cases. There are a variety of techniques and instrumentations available for division of hepatic parenchyma including blunt dissection, use of a Tissue Link device, the CUSA, the harmonic scalpel, standard electrocautery, the argon beam coagulator, or automated stapling devices. Blunt parenchymal dissection can be performed manually with finger fracture, with a Kelly clamp that crushes the tissues, or a small blunt suction device. Vascular or biliary structures that are encountered during blunt parenchymal dissection are ligated or clipped prior to division. The Tissue Link device combines an applicator with electrocautery and local perfusion of saline. As the hepatic parenchyma is transected, the collagen and surrounding blood and bile vessels contract to create hemosta-

sis and seal small bile ducts. As with the blunt techniques, CUSA dissection requires that vascular and biliary structures encountered during dissection be ligated or clipped prior to division. The harmonic scalpel relies on coagulation necrosis caused by ultrasound. Automated stapling devices are especially useful for division of hepatic parenchyma that also contains larger blood vessels and bile ducts. Newer techniques using pulsating water jets are also being developed. The superiority of any one of these techniques has not been established; consequently, the choice is based primarily on surgeon preference. Closed suction drainage is not routinely required unless the surgical field demonstrated residual venous or biliary leakage. Total inflow occlusion

Total inflow occlusion (Pringle maneuver) is used to reduce bleeding during parenchymal transection. Access to the porta hepatis is achieved by inserting the index finger into the foramen of Winslow and the thumb through a defect created in the gastrohepatic ligament (Figure 8.18). For total inflow occlusion, an

Figure 8.18 Total inflow occlusion (Pringle maneuver). The index finger is inserted through the foramen of Winslow and the thumb through a defect in the gastrohepatic ligament to access the porta hepatis. An atraumatic clamp is then placed for total inflow occlusion

247

Ch 08 final.qxd

12/18/2006

15:27

Page 248

SURGERY FOR OVARIAN CANCER

atraumatic vascular clamp or silastic loop can be placed and tightened. Temporary total inflow occlusion can reduce bleeding from the liver remnant that occurs even after selective inflow ligation of the pedicle supplying the liver to be removed. The noncirrhotic liver can tolerate total inflow occlusion beyond 60 min without irreversible damage. However, patients with liver dysfunction do not tolerate prolonged liver ischemia. In such cases, the Pringle maneuver should be used intermittently and for brief periods of time (e.g. 10 min separated by several minutes of perfusion) until parenchymal transection is completed.

a

b

c

d

Major hepatic resections Metastatic ovarian cancer presenting initially or at the time of recurrence as a sizable lesion or lesions involving multiple liver segments requires a major hepatic resection. The major hepatic resections consist of five procedures, the nomenclature for which is based on systems developed by Couinaud21 and Goldsmith and Woodburne.22 Using the terminology of Couinaud, a right hepatectomy is removal of segments 5, 6, 7 and 8. A left hepatectomy is resection of segments 2, 3 and 4. A right lobectomy, sometimes referred to as a right trisegmentectomy, is excision of segments 4, 5, 6, 7 and 8, and occasionally segment 1. A left lobectomy is resection of segments 2 and 3. Finally, an extended left hepatectomy is resection of segments 2, 3, 4, 5 and 8, and occasionally segment 1 (Figure 8.19). As a general rule, cholecystectomy is performed as part of a major hepatic resection, owing to the position of the gallbladder, lying adjacent to both the right and left sides of the liver within the gallbladder fossa. For the purposes of this discussion, only the right and left hepatectomy procedures will be described in detail. Right hepatectomy

A large lesion occupying the right side of the liver is appropriately managed by right hepatectomy with resection of segments 5, 6, 7 and 8 (Figure 8.20a). The liver is mobilized by dividing the falciform and

248

e

Figure 8.19 Common major hepatic resections. (a) Right hepatectomy; (b) left hepatectomy; (c) right lobectomy; (d) left lobectomy; (e) extended left hepatectomy

right coronary ligaments (Figure 8.20b). Since the gallbladder fossa is the dividing line between the right and left sides of the liver, the gallbladder should be removed as part of the resection (see below). Also, the right hepatic duct is easier to visualize after cholecystectomy has been performed. The right side of the liver is retracted cephalad and the right hepatic duct clamped, divided and suture ligated (Figure 8.20c). A transfixion ligature is placed for added security. The common hepatic duct and left hepatic duct are retracted medially and the right hepatic artery doubly ligated and divided (Figure 8.20d). Posteriorly, the portal vein, with its right and left branches, is clearly exposed. The right branch of the portal vein is doubly cross clamped with straight vascular clamps (e.g. Cooley) and divided; the ends of the vessels are oversewn with a continuous stitch of 4-0 polypropylene suture. The proximal end of the vein may be reinforced with a series of horizontal mattress stitches for additional security (Figure 8.20e).

Ch 08 final.qxd

12/18/2006

15:27

Page 249

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

Confluence of middle and left hepatic veins Left triangular ligament

Right hepatic vein

a

Branch of middle hepatic vein Right branch of portal vein Right triangular ligament Falciform ligament Ligated right hepatic duct

b

Ligated right hepatic artery d

Loop about common hepatic duct

Cystic artery c

Cystic duct Right branch of portal vein

e

Figure 8.20 Right hepatectomy. (a) Right hepatectomy is indicated for lesions occupying or in close proximity to segments 5, 6, 7 and 8. (b) The falciform ligament has been taken down and the right coronary ligament divided, exposing the underlying hepatic veins as they join the inferior vena cava. (c) Following cholecystectomy, the right hepatic duct is suture ligated and divided. (d) The right hepatic artery is similarly exposed, ligated and divided. (e) The right branch of the portal vein is clamped and divided; the vessel ends are oversewn; the proximal venous stump is further secured with an additional layer of horizontal mattress stitches. Continued

249

Ch 08 final.qxd

12/18/2006

15:27

Page 250

SURGERY FOR OVARIAN CANCER

Right hepatic vein

Ligature on caval branches

f Divided oversewn right hepatic vein

Vascular clamps on right hepatic vein g

Inferior vena cava

Ligated minor vessels and ducts

Line of resection vs. line of color demarcation

Residual portal structures i h

Figure 8.20 continued (f) The right side of the liver is exposed and the small hepatic veins emptying directly into the inferior vena cava are ligated and divided. (g) The right hepatic vein is clamped with vascular clamps and divided, and the ends are oversewn; the vessel stump on the caval side is further secured with a series of horizontal mattress stitches. (h) The line of color demarcation produced by inflow and outflow control identifies the plane of resection, which should be approximately 1 cm lateral to the color line. (i) The argon beam coagulator is applied to the resected surface for additional hemostasis

The right side of the liver is rotated medially and the right triangular ligament taken down to the level of division of the coronary ligament, completely exposing the bare area of the right side. Any remaining attachments to the diaphragm are divided with electrocautery, exposing the right hepatic vein and inferior vena cava. The small hepatic veins directly draining from the liver to the inferior vena cava are

250

carefully ligated and divided (Figure 8.20f). The right hepatic vein is carefully mobilized from the surrounding hepatic parenchyma to develop an adequate pedicle and secured with a vasa-loop. Enough mobility must be achieved to permit the application of two curved vascular clamps (e.g. Cooley, Satinsky) across the right hepatic vein. The right hepatic vein is divided and the ends of the vessels are oversewn in the

Ch 08 final.qxd

12/18/2006

15:27

Page 251

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

same fashion as for the portal vein, securing the vena caval side end with a series of horizontal mattress sutures (Figure 8.20g). Alternatively, an endoscopic automated GIA stapling device with a vascular load (2.5 mm) can be used to secure and divide the right hepatic vein. Following ligation of the blood supply, the line of color demarcation is used as a guide and the liver parenchyma divided using one of the techniques described above. The line of resection should be approximately 1 cm lateral to the line of color demarcation (Figure 8.20h). Minor bleeding or leakage of bile from the liver surface can be controlled with the argon beam coagulator (Figure 8.20i) Left hepatectomy

Significant lesions involving the left side of the liver may be excised with a left hepatectomy (resection of segments 2, 3 and 4) (Figure 8.21a). As the medial margin of segment IV extends to the gallbladder fossa, cholecystectomy is generally performed prior to proceeding with the hepatic resection, as for right hepatectomy. The liver is mobilized by division of the falciform, left coronary and left triangular ligaments (Figure 8.21b). With cephalad retraction, attention is first directed toward controlling inflow to the left side of the liver. The left hepatic duct is dissected proximal to the common hepatic duct and doubly suture ligated (Figure 8.21c). The left hepatic artery is isolated, divided and ligated with a transfixion ligature placed on the proximal pedicle (Figure 8.21d). The gastrohepatic ligament is evaluated for an accessory or replaced hepatic artery arising from the left gastric artery, which, if present, requires ligation. A preoperative hepatic angiogram may assist with identification of the variable blood supply to the left liver. Further dissection at the base of the umbilical fissure exposes the left branch of the portal vein, which is secured with vascular clamps, divided and oversewn (Figure 8.21e). Ligation at this location allows for identification and preservation of the branches to segment 1 (caudate lobe), which arise prior to entry of the left branch portal vein into the umbilical fissure. Once inflow has been divided, a line of demarcation devel-

ops along a plane extending from the gallbladder fossa to the left of the vena cava. In order to reduce blood loss from the procedure, outflow from the left side of the liver is controlled prior to division of the hepatic parenchyma. The left side of the liver is completely mobilized and drawn inferiorly, exposing the middle and left hepatic veins as they enter the suprahepatic inferior vena cava. The gastrohepatic ligament is divided and the ligamentum venosum is ligated at its entry to the left hepatic vein. A triangular tunnel formed by the left hepatic vein, the inferior vena cava and the upper surface of segment II is developed. The left and middle hepatic veins individually, or their common trunk, are doubly clamped with vascular clamps and divided, and the vessel ends are oversewn as previously described (Figure 8.21f). Alternatively, an endoscopic automated GIA stapling device with a vascular load (2.5 mm) can be used to secure and divide the middle and left hepatic veins. The parenchyma is then resected using the demarcation plane as a guide (Figure 8.21g). The resection plane should remain anterior to the ligamentum venosum to avoid injury to the caudate lobe. Finally, the resection bed is treated with the argon beam coagulator for added hemostatic security.

Anatomic resection (segmentectomy) An anatomic resection of the liver, or segmentectomy, results in the loss of less liver parenchyma than the major hepatic resections, but may be better suited than the non-anatomic wedge resections for removing bulky and/or centrally located metastatic lesions. Anatomic resection may include excision of a single segment (unisegmentectomy) or multiple adjacent segments (such as a bisegmentectomy or trisegmentectomy). The procedures are based on the anatomic descriptions of Couinaud (Figure 8.22). Resection of segments 5 and 8 (right anterior bisegmentectomy) and resection of segments 6 and 7 (right posterior bisegmentectomy) are also referred to as sectoral resections and are described in detail below. Resection of segment 1 (unisegmentectomy) and resection of

251

Ch 08 final.qxd

12/18/2006

15:27

Page 252

SURGERY FOR OVARIAN CANCER

Left coronary ligament

Lesion

Left triangular ligament

Falciform ligament a

Left lobectomy

Left lateral lobe

b

Left duct ligatures

Left hepatic artery ligatures

c

d

Curved vascular clamps on left hepatic vein

Ligature of left hepatic duct e

Falciform ligament f

Figure 8.21 Left hepatectomy. (a) Left hepatectomy is indicated for lesions occupying or in close proximity to segments 2, 3 and 4. (b) The falciform ligament has been taken down and the left coronary ligament divided. (c) Following cholecystectomy, the left hepatic duct is suture ligated and divided. (d) The left hepatic artery is similarly exposed, ligated and divided. (e) The left branch of the portal vein is clamped and divided; the vessel ends are oversewn; the proximal venous stump is further secured with an additional layer of horizontal mattress stitches. (f) The left hepatic vein is clamped with vascular clamps and divided, and the ends oversewn; the vessel stump on the caval side is further secured with a series of horizontal mattress stitches. Continued

252

Ch 08 final.qxd

12/18/2006

15:27

Page 253

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

Color demarcation Line of resection

Oversewn left hepatic vein Inferior vena cava

Ligated minor vessels and ducts g

h

Figure 8.21 continued (g) The line of color demarcation produced by inflow and outflow control identifies the plane of resection, which should be approximately 1 cm lateral to the color line. (h) The argon beam coagulator is applied to the resected surface for additional hemostasis

segments 4, 5 and 6 (trisegmentectomy) are other common examples of anatomic resections. Right anterior sectoral resection

A right anterior sectoral resection may be performed for lesions involving segments 5 and 8 located within the anterolateral substance of the right side of the liver. Prior to resection, the gallbladder should be removed and the liver completely mobilized. Division of the right anterior sectoral pedicle allows for inflow control. Confirmation of the isolation of the correct pedicle can be made by temporary occlusion and subsequent color change of segments 5 and 8. The liver is retracted upward, exposing the porta hepatis. The common hepatic duct is traced proximally and the right and left hepatic ducts are identified. The right hepatic duct is then followed distally until the bifurcation into anterior and posterior segmental ducts is located. The anterior segmental duct is carefully isolated, doubly suture ligated and divided, with care taken to preserve the posterior segmental duct (Figure 8.23a). The corresponding anterior segmental hepatic artery is similarly isolated, divided and ligated. The

right branch of the portal vein is then followed distally until the bifurcation into the right anterior and right posterior portal veins is located and gently mobilized. The right anterior portal vein is then isolated between curved vascular clamps and divided, and the ends of the vessels are oversewn (Figure 8.23b). Once the right anterior sectoral pedicle has been secured, the color demarcation line is used to guide resection of the liver parenchyma (Figure 8.23c). Both the right and middle hepatic veins are preserved at their respective junctions with the inferior vena cava; however, temporary placement of a vascular clamp on the right hepatic artery may reduce blood loss during parenchymal resection. Right posterior sectoral resection

For metastatic ovarian cancer limited to the posterolateral substance of the right side of the liver, a right posterior sectoral resection of segments 6 and 7 is indicated. Again, cholecystectomy should be performed and the liver mobilized prior to proceeding with resection. When dissecting along the posterior surface of segments 6 and 7, the retrohepatic veins

253

Ch 08 final.qxd

12/18/2006

15:27

Page 254

SURGERY FOR OVARIAN CANCER

hepatic artery is identified as it leaves the right hepatic artery, and it is also suture ligated and divided. The third component of the right posterior sectoral pedicle, the right posterior portal vein, is isolated between curved vascular clamps and divided, and the vessel ends are oversewn (Figure 8.24b). The resulting color line of demarcation defines the appropriate plane of hepatic parenchymal resection (Figure 8.24c). The right hepatic vein may be either temporarily clamped and preserved or ligated and divided at its insertion into the inferior vena cava. If the right hepatic vein is sacrificed, the remaining anterior sector will drain via the middle hepatic vein.

2 7

4A 8

3 4B

C

5

6

D

A

B

4

Unisegmentectomy

5 3 6

2 1

7 C

Figure 8.22 Lines

of

resection

for

various

anatomic

(segmental) resections. Right anterior sectorectomy – resection of segments 5 and 8 (follows lines A and B). Right posterior sectorectomy – resection of segments 6 and 7 (follows line B). Central hepatectomy – resection of segments 4, 5 and 8 (follows lines B and D). Trisegmentectomy – resection of segments 4, 5 and 6 (follows lines C, A and D)

entering the inferior vena cava directly should be individually ligated and divided. Inflow control is obtained through ligation of the right posterior sectoral pedicle. A deep groove, the incisura dextra of Gans, is present in the inferior surface of the liver near the bed of the gallbladder. The right posterior sectoral pedicle lies within the base of this groove in the majority of cases. Once again, confirmation of pedicle identity can be made by temporary occlusion and observing for color change in segments 6 and 7. The right hepatic duct is traced until it bifurcates, and the posterior segmental duct is isolated, doubly ligated and divided (Figure 8.24a). The posterior segmental

254

Segmental resection is indicated for ovarian cancer metastasis limited to a specific anatomical segment of liver parenchyma and is based on the underlying vascular, arterial and biliary ductal anatomy. Any individual segment can be removed as part of an anatomic resection; however, unisegmentectomy of segment 5 and segment 6 is most commonly performed for ovarian cancer (Figure 8.25). In general, removal of the gallbladder will improve exposure for anterior resections and should be performed initially. Inflow control is achieved by application of the Pringle maneuver. In contrast to the sectoral resections, pedicle isolation for unisegmentectomy is performed during the course of, rather than prior to, parenchymal dissection. In an anatomical resection of segment 5, the segmental margin is first outlined with electrocautery ventrally (Figure 8.25a) and dorsally (Figure 8.25b). Dissection proceeds by using a Kelly clamp to gently crush the liver parenchyma and identify small caliber vessels and ducts, which are individually clipped or ligated as they are encountered (Figure 8.25c). The anterior inferior hepatic duct and artery, tributaries to the right hepatic vein, and the right anterior inferior portal vein are clamped, divided and suture ligated as they are encountered during the course of parenchymal dissection (Figure 8.25d). If necessary, the argon beam coagulator can be used to secure additional hemostasis in the resection bed.

Ch 08 final.qxd

12/18/2006

15:27

Page 255

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

Anterior segmental duct Left hepatic duct Left hepatic artery Right hepatic artery Left hepatic duct Common hepatic duct Portal vein a

b

Figure 8.23 Right anterior sectoral resection. (a) The anterior segmental duct is isolated and divided, exposing the anterior segmental hepatic artery. (b) The right anterior portal vein is isolated with vascular clamps and divided, and the ends are oversewn. (c) Completed resection with preservation of the c

Radiofrequency ablation and cryoablation Radiofrequency ablation (RFA) has emerged as an important tool for treating primary and metastatic tumors of the liver including ovarian cancer.23,24 RFA may be particularly useful in patients who are not ideal candidates for liver resection because of compromised medical condition, advanced cirrhosis, or anatomic considerations that would make surgical excision unsafe. RFA uses alternating electric current in the range of radiofrequency waves through a multipronged needle electrode to produce focal thermal injury. RFA may be performed percutaneously under radiographic (computed tomography) guidance or at the time of open laparotomy, in which case ultrasonography is the imaging modality of choice. The

right and middle hepatic veins

procedure can be used in most areas of the liver, except in the region of the hilum, and is effective for lesions up to 5 cm in size.24 The RFA electrode is guided to the center of the tumor mass and the prongs (or tines) are deployed. The goal of RFA is to destroy the target tumor as well as a rim of adjacent normal hepatic parenchyma. A single ablation takes 10–30 min depending on the tissue characteristics and desired ablation size. As the ablation progresses, gas and debris will become apparent and are used to gauge the extent of tumor and surrounding normal tissue destruction. Significant complications from RFA are infrequent but may include thermal injury to nearby abdominal viscera, bile duct fistula, portal vein thrombosis and hemorrhage. The most commonly encountered minor complications from RFA are right

255

Ch 08 final.qxd

12/18/2006

15:27

Page 256

SURGERY FOR OVARIAN CANCER

Left hepatic duct Posterior segmental duct Left hepatic artery Right hepatic artery Right hepatic duct Common hepatic duct Portal vein a

b

Figure 8.24 Right posterior sectoral resection. (a) The posterior segmental duct is isolated and divided, exposing the posterior segmental hepatic artery. (b) The right posterior portal vein is isolated with vascular clamps and divided, and the ends are oversewn. (c) Completed resection with preservation of the right hepatic vein; the right hepatic vein may be c

upper quadrant pain, asymptomatic pleural effusion and transient elevation of liver function tests. Cryoablation is an alternative option to RFA and may be used alone or in combination with hepatic resection in patients with multiple metastases or those with underlying liver dysfunction. Intraoperative ultrasound is used to confirm the location of the target lesion (or lesions) and the relationship to surrounding vascular and biliary structures. Usually, needle localization of the tumor mass is performed first, followed by introduction of the cryoprobe. The size of the cryoprobe selected is based on the required iceball size necessary to generate a 1-cm margin around the tumor. The cryoprobe reaches a temperature below 160°C at the tip. A common cryoablation protocol consists of two 10-min freeze cycles with an interven-

256

sacrificed if necessary

ing 5-min thaw period. Goering et al. reported their experience with cryoablation of liver metastases in seven patients with ovarian cancer.25 Three of the seven patients died within 19 months of surgery, while all of the remaining four patients experienced hepatic recurrence of their ovarian cancer within 12 months of the procedure. The high local failure rate in this small series is concerning, and probably warrants consideration of RFA as the primary ablative option for ovarian cancer metastases that are not amenable to standard surgical resection.

Postoperative management and complications Continued advances in postoperative care have led to significant reductions in morbidity and mortality associated with hepatobiliary surgery. Several issues

Ch 08 final.qxd

12/18/2006

15:27

Page 257

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

a

b

c

d

Figure 8.25 Segmentectomy: resection of parenchymal liver metastasis of recurrent ovarian cancer to segment 5. (a) Electrocautery is used to circumscribe the boundaries of resection on the ventral surface. (b) The right lobe is retracted upward and the dorsal margin of resection is outlined with electrocautery. (c) The hepatic parenchyma is divided using a Kelly clamp to crush the tissue and identify individual hepatic vessels and ducts, which are individually clipped or ligated. (d) Completed resection of segment 5 with clamps on the anterior inferior portal vein (lower right) and tributaries to the right hepatic vein and anterior inferior hepatic duct and artery (upper left)

related specifically to resection of liver parenchyma must be considered, including the potential for altering liver physiology so as to affect elimination of anesthetics and other medications including sedatives, analgesics and neuromuscular blockers. The reduction in metabolism is unpredictable and needs to be monitored closely. Deficiencies in coagulation factors

may evolve rapidly and need to be anticipated. Many abnormalities may be corrected within 48 h with vitamin K. If more rapid reversal of the coagulopathy is required, fresh frozen plasma may be administered. Phosphorus replacement is typically necessary and should be added to replacement fluids. Respiratory function may be compromised by any upper abdomi-

257

Ch 08 final.qxd

12/18/2006

15:27

Page 258

SURGERY FOR OVARIAN CANCER

a

b

c

d

Figure 8.26 Cholecystectomy. (a) Metastatic ovarian cancer involving the gallbladder serosa. (b) An incision is created in the peritoneum over the dome of the gallbladder. (c) Dissection proceeds around either side of the gallbladder, releasing the areolar attachments to the gallbladder fossa with electrocautery. (d) The cystic duct is identified, ligated and divided

nal procedure, especially a hepatic resection, and prolonged intubation may be required. Jaundice may result from several factors, including hepatocellular injury or iatrogenic bile duct obstruction. Hepatocellular injury may be a consequence of low flow state, infection secondary to transfusion, or anesthetic toxicity. Both ovarian cancer and hepatobiliary surgery may predispose to ascites, and fluid shifting (‘thirdspacing’) may become an issue postoperatively. As with diaphragm resections, hepatobiliary surgery performed by technically proficient operators is associat-

258

ed with acceptable morbidity and mortality that are comparable to the complication rates seen with other major abdominal surgery.

Gallbladder and porta hepatis disease Cholecystectomy

On occasion, metastatic ovarian cancer will involve the serosa of the gallbladder such that cholecystectomy may be necessary for safe removal of the disease (Figure 8.26a). The decision to perform cholecystec-

Ch 08 final.qxd

12/18/2006

15:27

Page 259

CYTOREDUCTIVE SURGERY: RIGHT UPPER ABDOMEN

tomy for ovarian cancer cytoreduction should be considered within the context of the overall operation and likelihood that the procedure will contribute to an overall optimal volume of residual disease. The gallbladder can be removed from the fundus downward. The gallbladder is grasped with a Kelly clamp for counter traction and a 1-cm superficial incision is created along the peritoneum overlying the dome of the gallbladder using electrocautery (Figure 8.26b). A plane is then developed around both sides of the gallbladder and dissection continued downward towards the infundibulum (Figure 8.26c). The areolar attachments of the gallbladder to the adjacent gallbladder fossa are released with electrocautery. When the infundibular area of the gallbladder is reached, dissection is carefully carried out to isolate, ligate and transect the cystic artery. The infundibular cystic duct junction is identified and dissected (Figure 8.26d). There is no need to dissect the cystic duct and common bile duct junction, as this may contribute to a bile duct injury. The infundibular cystic duct junction is ligated and transected, and the remnant cystic duct stump secured with a transfixion stitch, with care being taken not to encroach upon the common bile duct. Hemostasis of the gallbladder fossa can be achieved by application of the electrocautery device or the argon beam coagulator. Porta hepatis disease

Ovarian cancer metastatic to the porta hepatis is one of the most challenging scenarios, at least in terms of cytoreductive surgery, that the surgeon operating on this group of patients will face. The porta hepatis is an often cited locale predicating a suboptimal surgical result. A clear understanding of the anatomy of the porta hepatis is necessary to assess whether such disease is safely resectable and whether doing so will contribute to the overall surgical result (i.e. optimal residual disease). The hepatoduodenal ligament runs from the porta hepatis to the duodenum and is covered by a peritoneal sheath and thin layer of fat. If the portal vein, hepatic artery and bile duct are identified and protected, disease can be on occasion dissected

off the structures within the hepatoduodenal ligament and porta hepatis; however, this is associated with a significant risk of injury and should only be undertaken once the safety of underlying structures has been assured. The CUSA may prove particularly useful for dissecting around the portal vein, hepatic artery and common bile duct, unless the tumor is extremely fibrotic. Occasionally, ovarian cancer may spread through the lymphatics of the intestinal tract to involve the portocaval lymph nodes. Nodal disease in this region is often more amenable to resection or enucleation than the typical peritoneal spread pattern. Resection of portocaval adenopathy should be attempted only once the vascular and biliary structures have been clearly identified and protected.

CONCLUSION Both primary and recurrent ovarian malignancies frequently involve the right upper quadrant of the abdomen. Optimal cytoreduction in a patient with right upper quadrant disease requires a comprehensive understanding of both the regional anatomy and the relevant surgical procedures. Mobilization of the liver and its consequent improvement in exposure is a prerequisite. Even in optimal conditions, disease in the porta hepatis may be most challenging and this may be the only area with significant residual disease. However, with an experienced ovarian cancer surgeon, many patients may achieve optimal cytoreduction with the addition of an aggressive approach to disease in the right upper quadrant.

REFERENCES 1.

Bristow RE, Montz FJ, Lagasse LD, et al. Survival impact of surgical cytoreduction in stage IV epithelial ovarian cancer. Gynecol Oncol 1999; 72: 278–87

2.

Sakai K, Kamura T, Hirakawa T, et al. Relationship between pelvic lymph node involvement and other disease sites in patients with ovarian cancer. Gynecol Oncol 1997; 65: 164–8

259

Ch 08 final.qxd

12/18/2006

15:27

Page 260

SURGERY FOR OVARIAN CANCER

3.

4.

5.

6.

7.

8.

9.

Griffiths CT, Finkler NJ. Surgery for carcinoma of ovary: extrapelvic cytoreduction. In Coppleson M, ed. Gynecologic Oncology, 2nd edn. London: Churchill Livingstone, 1992: 1325–33

14.

Rose PG. The cavitational ultrasonic surgical aspirator for cytoreduction in advanced ovarian cancer. Am J Obstet Gynecol 1992; 166: 843–6

15.

Cliby W, Dowdy S, Feitoza SS, et al. Diaphragm resection for ovarian cancer: technique and short-term complications. Gynecol Oncol 2004; 94: 655–60

Bristow RE, Montz FJ. Complete surgical cytoreduction of advanced ovarian carcinoma using the argon beam coagulator. Gynecol Oncol 2001; 83: 39–48

16.

Merideth MA, Cliby WA, Keeney GL, et al. Hepatic resection for metachronous metastases from ovarian carcinoma. Gynecol Oncol 2003; 89: 16–21

Brand E, Pearlman N. Electrosurgical debulking of ovarian cancer: a new technique using the argon beam coagulator. Gynecol Oncol 1990; 39: 115–18

17.

Chi DS, Fong Y, Venkatraman ES, Barakat RR. Hepatic resection for metastatic gynecologic carcinomas. Gynecol Oncol 1997; 66: 45–51

Bonnefoi H, A’Hern RP, Fisher C, et al. Natural history of stage IV epithelial ovarian cancer. J Clin Oncol 1999; 17: 767–75

18.

Montz FJ, Schlaerth JB, Berek JS. Resection of diaphragmatic peritoneum and muscle: role in cytoreductive surgery for ovarian cancer. Gynecol Oncol 1989; 35: 338–40

Chi DS, Temkin SM, Abu-Rustum NR, et al. Major hepatectomy at interval debulking for stage IV ovarian carcinoma: a case report. Gynecol Oncol 2002; 87: 138–42

19.

Kapnick SJ, Griffiths CT, Finkler NJ. Occult pleural involvement in stage III ovarian carcinoma: role of diaphragm resection. Gynecol Oncol 1990; 39: 135–8

Yoon SS, Jarnagin WR, Fong Y, et al. Resection of recurrent ovarian or fallopian tube carcinoma involving the liver. Gynecol Oncol 2003; 91: 383–8

20.

Naik R, Nordin A, Cross PA, et al. Optimal cytoreductive surgery is an independent prognostic indicator in stage IV epithelial ovarian cancer with hepatic metastases. Gynecol Oncol 2000; 78: 171–5

21.

Couinaud CL. Le foie: etudes anatomiques et chirurgicales. Paris: Mason 1957: 400–9

22.

Goldsmith NA, Woodburne RT. The surgical anatomy pertaining to liver resection. Surg Gynecol Obstet 1957; 105: 310–18

23.

Curley SA. New approaches to the treatment of hepatic malignancies: radiofrequency ablation of malignant liver tumors. Ann Surg Oncol 2003; 10: 338–47

24.

Bojalian MO, Machado GR, Swenson R, Reeves ME. Radiofrequency ablation of liver metastasis from ovarian adenocarcinoma: case report and literature review. Gynecol Oncol 2004; 93: 557–60

25.

Goering JD, Mahvi DM, Niederhuber JE, et al. Cryoablation and liver resection for noncolorectal liver metastases. Am J Surg 2002; 183: 384–9

Akahira JI, Yoshikawa H, Shimizu Y, et al. Prognostic factors of stage IV epithelial ovarian cancer: a multicenter retrospective study. Gynecol Oncol 2001; 81: 398–403

10.

Curtin JP, Malik R, Venkatraman ES, et al. Stage IV ovarian cancer: impact of surgical debulking. Gynecol Oncol 1997; 64: 9–12

11.

Hoskins WJ, McGuire WP, Brady MF, et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J Obstet Gynecol 1994; 170: 974–9, discussion 979–80

12.

Liu PC, Benjamin I, Morgan MA, et al. Effect of surgical debulking on survival in stage IV ovarian cancer. Gynecol Oncol 1997; 64: 4–8

13.

Adelson MD. Cytoreduction of diaphragmatic metastases using the Cavitron Ultrasonic Surgical Aspirator. Gynecol Oncol 1991; 41: 220–2

260

Ch 09 final.qxd

12/18/2006

15:29

Page 261

CHAPTER 9

Cytoreductive surgery: left upper abdomen Krishnansu S Tewari, Michael L Berman

INTRODUCTION Metastatic ovarian cancer may involve important anatomical structures in the left upper abdomen including the transverse colon, the spleen, the tail of the pancreas, the left hemi-diaphragm and the stomach. The proximate relationship of these structures to the greater omentum and gastrocolic ligament predisposes to tumor spread by direct extension from a large omental cake. The spleen and left hemi-diaphragm, and rarely the stomach, may also be involved by independent metastatic lesions that are discontinuous with an omental tumor. This chapter will outline the relevant regional anatomy and address the cytoreductive surgical procedures required for the operative management of ovarian cancer extending to the left upper abdomen.

REGIONAL ANATOMY Stomach, lesser sac, celiac axis and gastrics The stomach is a local expansion of the alimentary tube interposed between the esophagus and the duodenum. It is shaped like a cornucopia. The larger part of the stomach lies in the left hypochondrium, and the remainder in the epigastrium. It is supported by the sloping transverse mesocolon upon which it rests and is maintained in position via its continuity with the abdominal esophagus, which is fixed in turn solidly to the esophageal hiatus in the diaphragm. The stomach is also held firmly in position by its continuity with the

duodenum, which in turn is anchored securely over most of its extent to the posterior parietal peritoneum. The fundus is the expanded upper portion of the stomach, lying left of the cardia. The pylorus of the stomach makes up the right portion and is attenuated by the circular narrowing of the duodenopyloric constriction externally and the pyloric orifice internally. The body of the stomach lies obliquely between the fundus and the pylorus. The cardiac orifice occurs at the junction of the esophagus and the stomach, and represents the level where the curvatures begin. The convex greater curvature is on the left side of the stomach and is related to the arterial arch formed from the right and left gastroepiploic arteries lying within the gastrocolic ligament. The concave lesser curvature defines the right margin of the stomach and affords attachment for the gastrohepatic ligament or lesser omentum, which is related to the arterial circle arising from the right and left gastric (coronary) arteries (Figure 9.1). The upper portion of the fundus is in close proximity to the medial aspect of the spleen, separated by the gastrosplenic ligament which contains the splenic vessels. The splenic flexure of the colon, likewise, is in immediate proximity to the lateral aspect of the stomach and inferior portion of the spleen. The pancreas and the anterior surface of the left kidney are separated from the posterior aspect of the stomach by the peritoneum of the lesser sac. The anterior aspect of the body and tail of the pancreas is in close proximity to the posterior aspect of the stomach. Because of these relations, tumor involving the greater omentum may impinge upon adjacent structures such as the

261

Ch 09 final.qxd

12/18/2006

15:29

Page 262

SURGERY FOR OVARIAN CANCER

Celiac axis Left gastric Left inferior phrenic Superior polar gastric Left gastric Short gastrics anterior branch

Superior terminal gastric

Splenic Common hepatic

Inferior terminal gastric

Gastroduodenal Superior (dorsal) pancreatic Right gastric

Greater pancreatic Left gastroepiploic

Right gastroepiploic Posterior and anterior (superior and inferior) pancreaticoduodenals

Middle colic

Superior mesenteric

Inferior (transverse) pancreatic

Figure 9.1 Anatomy of the left upper abdomen with arterial supply to the stomach, spleen, pancreas and transverse colon

posterior wall of the stomach, pancreas, spleen, transverse colon and splenic flexure of the colon. An envelope of peritoneum gives the stomach a complete serous coat except at the omental reflections at the greater and lesser curvatures and a small portion of the cardiac area in direct contact with the diaphragm. The musculature consists of three layers of involuntary muscle, and at the pylorus, the circular middle layer thickens to form the pyloric sphincter. The mucosa is thick and highly vascular. The vasculature of the stomach arises from a rich arterial supply from the celiac axis that bathes the stomach. The celiac trunk arises from the aorta just above the neck of the pancreas and trifurcates immediately into the left gastric (coronary), splenic and hepatic arteries. These vessels and their branches

262

form two arterial arches adjacent and parallel to the greater and lesser curvatures (Figure 9.1). The arch of the lesser curvature consists of the right gastric branch of the hepatic artery running from right to left and the left gastric (coronary) artery running from left to right. The greater omentum attaches to the arch of the greater curvature, which is made up of the right gastroepiploic artery from the gastroduodenal and the left gastroepiploic artery from the splenic artery. The hepatic artery runs along the upper border of the pancreas and ascends to the liver in front of the epiploic foramen between the two layers of the lesser omentum, at which level it gives off the right gastric (pyloric) artery and subsequently the gastroduodenal artery when it reaches the upper border of the first

Ch 09 final.qxd

12/18/2006

15:29

Page 263

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

part of the duodenum. The gastroduodenal artery divides into the superior pancreaticoduodenal and right gastroepiploic arteries. The left gastric (coronary) artery divides into two parallel branches which anastomose with the right gastric (pyloric) artery. Finally, the splenic artery runs to the left under the peritoneal floor of the lesser sac along the upper margin of the body and tail of the pancreas, until it reaches the hilus of the spleen. Near its termination, the splenic artery gives rise to the short gastric and left gastroepiploic arteries which run forward in the gastrosplenic ligament to the greater curvature of the stomach, where the short gastric vessels are distributed to the fundus and the left gastroepiploic artery anastomoses with the right gastroepiploic artery (Figure 9.1). To sum up, the celiac axis supplies the stomach directly by the left gastric (coronary) artery and indirectly by the right gastric and right gastroepiploic arteries from the hepatic artery, and the short gastric and left gastroepiploic arteries from the splenic artery. Special mention of the posterior gastric artery should be made. Although it does not appear in many anatomic textbooks, it is important clinically. It is a branch from the middle third of the splenic artery which runs beneath the posterior peritoneum of the lesser sac to the fundus of the stomach, which it enters through the gastrohepatic ligament. Troublesome retroperitoneal hemorrhage may occur during partial gastrectomy if this artery, which can be paired, is not recognized and ligated. The gastric veins correspond to the arteries and empty into the large splenic and superior mesenteric trunks of the portal vein or directly into the portal vein itself. The left gastric (coronary) vein is a tributary to the portal system and anastomoses freely with the lower esophageal veins, which in turn anastomose with the upper esophageal veins. The esophageal veins are tributaries through the azygos veins with the caval venous system, forming an important anastomosis between the portal and systemic venous systems (Figure 9.2).

Pancreas The pancreas is an elongated, hammer-shaped, retroperitoneal gland immediately anterior to the upper abdominal aorta and vena cava. It is contained in the epigastrium and left hypochondrium and lies behind the serous floor of the lesser sac at the level of the first and second lumbar vertebrae. The pancreas is divided into a head, neck, body and tail. The head extends to the right of the midline and is surrounded by the first to third portions of the duodenum. The uncinate process of the head hooks behind the superior mesenteric vessels. The short neck supports the pyloric end of the stomach and, at its upper border, the common duct, portal vein and hepatic artery enter the lesser omentum. Behind the neck the superior mesenteric and splenic veins join to form the portal vein. The body of the pancreas is triangular and presents an anterior convexity in front of the vertebral column. The upper margin of the body has a close relation to the celiac trunk as evidenced by the grooving created by the tunneling of the splenic artery. The danger of attempting to reach the pancreas by a posterior lumbar approach for partial or complete pancreatectomy is great, because of the underlying splenic vein and left renal vessels. Finally, the body merges into the tail without demarcation. The tail lies within the peritoneal duplication of the lienorenal ligament and is related to the splenic flexure of the colon immediately below. The tail of the pancreas often extends to the splenic hilum, where it is surrounded by the splenic vessels. The surgeon should always consider the intimate relations of the pancreas to the spleen during the performance of splenectomy in order to avoid removing a portion of the tail of the pancreas. The main pancreatic duct of Wirsung begins in the tail and traverses the whole gland near its posterior surface, emerging from the right border of the head and, together with the common bile duct, opens into the ampulla of Vater in the duodenum. The branched accessory pancreatic duct of Santorini drains the upper part of the pancreatic head, opening both into

263

Ch 09 final.qxd

12/18/2006

15:29

Page 264

SURGERY FOR OVARIAN CANCER

Right gastric Portal Left gastric

Short gastric Splenic

Pancreatic Left gastroepiploic

Pancreaticoduodenal Communicating branch

Inferior mesenteric

Middle colic Right gastroepiploic Superior mesenteric

Figure 9.2 Anatomy of the left upper abdomen with venous drainage of the stomach, spleen, pancreas and transverse colon

the main pancreatic duct, and into the duodenum, cephalad to the opening of the main duct. Pancreatic arteries arise from the splenic artery itself or one of its divisions (e.g. the left gastroepiploic artery), enter the tail of the pancreas, and then form anastomoses with the inferior pancreatic artery and the pancreatica magna (a superior pancreatic branch of the splenic artery which enters the body of the pancreas). The veins of the pancreas join the splenic vein for the most part, but a large trunk issues from the dorsal aspect of the gland, running upward along the common bile duct to join the portal vein.

264

Spleen The largest of the ductless glands, the spleen is located in the left hypochondrium under cover of the 9th, 10th and 11th ribs. It is deeply concealed beneath the diaphragm and costal arch, and hidden by the stomach. The surface of the spleen is oval and confined entirely to the thoracic region, where it cannot be palpated under normal circumstances on physical examination. When it becomes involved with metastatic disease, the organ can extend from beneath the rib cage in an obliquely downward and medial direction,

Ch 09 final.qxd

12/18/2006

15:29

Page 265

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

as the phrenocolic ligament and splenic flexure hinder its enlargement directly downward. The layers of the gastrosplenic ligament separate at the hilus of the spleen. The outer layer invests the gland and is then reflected to the peritoneum over the anterior surface of the left kidney. The deep layer covers the splenic vessels and is continuous with the peritoneum of the floor of the lesser sac. Behind and medial to the spleen these two layers contain the splenic vessels and form the pancreaticosplenic or lienorenal ligament, which passes between the hilum of the spleen and the ventral aspect of the kidney. The inferior portion of the spleen lies upon the phrenocolic ligament which attaches the splenic flexure of the colon to the diaphragm. The spleen is maintained in position by pressure of the surrounding viscera, particularly the left kidney, the splenic flexure of the colon and the stomach (Figure 9.1). During resection of an omental cake and/or the take-down of the splenic flexure to facilitate a partial or total colectomy, downward traction can result in a splenic laceration and hemorrhage. If significant non-traumatic stretch to the ligaments occurs, a wandering spleen can result. The splenic artery runs a transverse course from the right to left along the superior margin of the pancreas. Near the hilum, and within the gastrosplenic ligament, the artery divides into numerous branches, including the superior polar, left gastroepiploic, superior terminal and inferior terminal arteries (Figure 9.1). The latter two merge into secondary branches which penetrate the splenic parenchyma at the hilum. The robust superior polar artery gives off the short gastric arteries to the stomach before entering the spleen. Meanwhile, the left gastroepiploic artery passes along the inferior pole of the spleen to supply part of the greater curvature of the stomach and the greater omentum. The splenic vein arises from several large branches leaving the hilum and pursues a much straighter course than the artery, running dorsal to the pancreas from left to right. Before joining with the superior mesenteric vein to form the portal vein, the splenic

vein receives the inferior mesenteric vein at a junction behind the head of the pancreas and ventral to the inferior vena cava (Figure 9.2).

Greater omentum, transverse colon and splenic flexure The transverse colon crosses the abdominal cavity from the hepatic to the splenic flexure. It has a downward curve and in the recumbent position the transverse colon reaches its lowest position in the midline, just below the umbilicus. The identity of the transverse colon can always be established by finding the greater omentum attached to its superoanterior surface. For the purposes of this discussion we will consider the distal transverse colon which is related closely to the greater curvature of the stomach and ascends slightly as it approaches the splenic flexure. The transverse colon is connected to the posterior abdominal wall by the transverse mesocolon. The transverse mesocolon forms a horizontal partition across the abdominal cavity, suspending the transverse colon from the posterior abdominal wall and separating the cavity of the omental bursa and the supramesocolic structures from the inframesocolic structures. The posterior parietal attachment of the transverse mesocolon is to the peritoneum overlying the anterior surface of the head, neck and body of the pancreas. The greater omentum develops from the primitive peritoneal duplication, the dorsal mesogastrium, which extends from the greater curvature of the stomach to the posterior abdominal wall, independent of the colon and transverse mesocolon. This fold bulges to the left and develops into a bag-like structure, the omental bursa. The epiploic foramen of this pouch is directed to the right, and the closed lower end hangs downward over the transverse colon, its mesocolon and much of the small intestine. The posterior layer of the greater omentum fuses with the serosal layer of the transverse colon and is in contact with its mesocolon. Generally, the cavity of the greater omentum is obliterated to within a short distance of the greater curvature of the stomach by adhesion between the

265

Ch 09 final.qxd

12/18/2006

15:29

Page 266

SURGERY FOR OVARIAN CANCER

opposed surfaces; obliteration of the space between the two leaves of the omentum on the left is not as complete as on the right, so that the unobliterated section on the left between the stomach and transverse colon usually offers ready access to the cavity of the lesser sac, even when extensive metastatic cancer is present. That part of the greater omentum connecting the greater curvature of the stomach with the transverse colon is the gastrocolic ligament. Because the left lobe of the liver is smaller than the right lobe, the left colic (splenic) flexure is placed higher than the right, and its angle is more acute than that of the hepatic flexure. The splenic flexure may overlie the left kidney, and is located deeply under cover of the costal margin. Examination and takedown of the splenic flexure can often be difficult. The upper and anterior aspects of the splenic flexure are also attached to the left margin of the greater omentum, and the posterior aspect of the splenic flexure is attached to the pancreas by that portion of the transverse mesocolon closest to this area. From the lateral aspect of the splenic flexure the peritoneum extends to the diaphragm as the left phrenocolic ligament, upon which the inferior pole of the spleen rests.

Superior and inferior mesenteric vasculature The superior mesenteric artery (SMA) arises from the aorta about 2 cm below the origin of the celiac trunk and behind the head of the pancreas (Figure 9.1). The arteries supplying the jejunum and ileum arise from the left aspect of the SMA. The middle colic artery arises from the SMA at the lower margin of the pancreas and runs in the transverse mesocolon, where it divides into right and left branches, which form anastomoses with the right and left colic arteries, respectively. The right branch supplies the right one-third of the transverse colon and the left supplies the left twothirds. The inferior mesenteric artery (IMA) arises from the aorta about 10 cm above its bifurcation. As it runs downward and slightly to the left, it gives off the left colic branch to the descending colon and the distal

266

portion of the transverse colon, and the sigmoid branches to the sigmoid colon. The superior mesenteric vein (SMV) returns blood from the small intestine and from the ascending and transverse colon (Figure 9.2). Behind the neck of the pancreas the SMV unites with the splenic vein to form the portal vein. Importantly, the SMV may be the site of a thrombosing phlebitis, resulting in intestinal venous engorgement, gangrene, intestinal obstruction and peritonitis. The inferior mesenteric vein (IMV) is formed by the superior hemorrhoidal, sigmoid and left colic veins; it joins the splenic vein and drains into the portal vein.

Left hemi-diaphragm The abdominal surface of the left hemi-diaphragm, except at the level of the pancreas and left kidney, is covered by peritoneum. It is related to the fundus of the stomach, the lateral surface of the spleen and the splenic flexure of the colon (Figure 9.3). Although there may be some variability in the origin of the phrenic artery, there is usually a single large inferior phrenic artery on each side of the body; these vessels usually arise from the aorta or from the celiac artery and during their ascent give off branches to the adrenal glands, the kidneys and renal capsule, and diaphragmatic crura. The venous drainage of the left hemi-diaphragm is accomplished by an unnamed vessel of considerable size, unaccompanied by an artery, which passes across the middle of the diaphragm to the inferior vena cava. The motor and sensory nerve supply of the diaphragm is derived primarily from the cervical plexus (C3–C5) through the two phrenic nerves, which descend in the thorax and finally terminate in branches that penetrate the diaphragm.

Left kidney and adrenal gland The left kidney and associated adrenal gland may be encountered during mobilization and resection of the left upper abdominal viscera. The adrenal gland is a small flattened gland located in the upper median

Ch 09 final.qxd

12/18/2006

15:29

Page 267

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

Left phrenic artery

Lateral branch of left phrenic artery

Left hemi-diaphragm

9th rib

Spleen

10th rib Splenocolic ligament

Stomach

Left colic (splenic) flexure

Pancreas Transverse colon

Descending colon

left renal artery and nerve enter and the renal vein, the principal lymphatics and the ureter emerge (Figure 9.4). This hilar aggregate forms the renal pedicle. The kidney is covered completely by a thin, resistant, fibroelastic membranous capsule (Glisson’s capsule). Large supernumerary renal vessels are common, being derived from the aorta as serially arranged stems; supernumerary renal arteries may also be derived from the internal spermatic and superior mesenteric arteries. The left renal vein, which is several centimeters longer than the right, regularly receives the following tributaries: suprarenal and inferior phrenic veins from above, and the ovarian and second or third lumbar veins from below. In contrast, on the right side, these veins drain into the inferior vena cava.

Figure 9.3 Anatomy of the left hemi-diaphragm and relationship to the stomach, spleen, tail of the pancreas and splenic flexure of the transverse colon

SURGICAL TECHNIQUES

portion of the kidney fossa between the upper pole of the kidney and the great vascular trunks of the abdomen. The arterial supply consists of a superior suprarenal artery from the inferior phrenic, a small direct branch from the aorta, and an inferior suprarenal branch from the renal artery. It is important to recognize that, although the suprarenal arterial supply is derived from the three sources described, this is at best an oversimplification as the arteries to the gland are numerous, often in excess of 50, and may be intimately associated with subdivisions of the main renal artery. The venous return is less complex, with the left suprarenal vein entering the left renal vein. The left kidney lies 2 cm higher than the right. About one-third of the left kidney lies in contact with the diaphragm, and about one-half lies above the pleural line, thus predisposing a patient to the risk of a pneumothorax with renal surgery. The left kidney is about 12 cm long and 7 cm broad, with its medial border hollowed out where the

For surgery on women with metastatic ovarian cancer, or other conditions which require excellent exposure of the left upper quadrant, the incision may be midline or paramedian but must extend well above the umbilicus. Operations which require extended exposure as part of an abdominopelvic debulking procedure, such as splenectomy, gastrectomy, diaphragmatic resection or total omentectomy, require that the incision extend to the level of the xyphoid process. Procedures which require access limited to the left upper quadrant of the abdomen may be midline, paramedian, or subcostal. When considering the left upper abdomen for surgical management of metastatic ovarian carcinoma, the organs and structures that may be involved with bulky disease include the greater omentum, lesser omentum, transverse colon, spleen, left hemidiaphragm, stomach and all associated peritoneal surfaces. The important anatomic details for these organs are described above. Upon entering the abdomen in the setting of metastatic disease, our preference typically is to begin

267

Ch 09 final.qxd

12/18/2006

15:29

Page 268

SURGERY FOR OVARIAN CANCER

Inferior phrenic arteries Left adrenal artery Left adrenal gland Left adrenal vein

Celiac axis Superior mesenteric artery

Left renal artery Left renal vein Left kidney Left ovarian artery

Right ovarian vein

Left ovarian vein

Right ovarian artery

Lumbar artery, posteriorly

Inferior mesenteric artery

Figure 9.4 Anatomy of the left kidney and adrenal gland and associated vasculature

the operation with the removal of the omentum as it is usually extensively involved with tumor, is an important source of serious gastrointestinal morbidity and ascites formation, and provides an ideal starting point for surgical extirpation of metastases in the left upper abdomen. From an anatomic perspective, contiguous extension to the transverse colon, spleen, pancreatic tail, diaphragm, or stomach may require an en bloc resection of one or a combination of these organs with the omental tumor. Even if the omental disease does not directly involve adjacent structures, its removal will improve exposure and facilitate mobilization of the surrounding anatomy so that subsequent resection of isolated metastases to the viscera of the left upper abdomen can be accomplished with maximum safety.

Omentectomy The omentum often appears hopelessly replaced by disease frequently involving the transverse colon, stomach and spleen. Having decided to begin the surgical attack on the greater omentum, the question often becomes where to start. If the disease has grown

268

into the anterior abdominal wall, this peritoneum must be resected first. Otherwise, we advocate delivering the omental ‘cake’ into the abdominal incision and, reflecting it cephalad, thus exposing the surgical plane between the posterior leaf and the serosa of the transverse colon. This avascular plane is incised with electrocautery just above the midpoint of the transverse colon and developed toward the hepatic and splenic flexures (Figure 9.5). When colonic involvement is too great to permit using electrocautery safely, scissor dissection is preferred. It may become necessary to take down either the hepatic flexure or the splenic flexure or both, in order to get around the disease. Care must be taken at the splenic flexure to avoid undue traction on the spleen which might cause troublesome bleeding from a splenic laceration or a capsule tear. The electrocautery is used to incise the anterior leaf of the greater omentum while employing downward traction on the adjacent colon to avoid injury. This step will separate the omentum fully from the transverse colon. Blood vessels in the anterior leaf must be ligated or hemoclipped separately. It is noteworthy, however, that even when the disease burden

Ch 09 final.qxd

12/18/2006

15:29

Page 269

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

Liver Stomach

Pancreas Greater omentum Approach to lesser sac a

the right and left gastroepiploic vessels and the intervening omental vessels arising from the gastric arcade along the greater curvature of the stomach (see Chapter 7: Figures 7.20–7.23). Postoperatively we prefer to decompress the stomach for 24 h via nasogastric suction to avoid overdistension and dislodging the suture ligatures along the greater curvature.

Omentectomy with en bloc resection of the transverse colon and splenic flexure Transverse colon

Greater omentum

Spleen

Transverse colon

b

Figure 9.5 Omentectomy. (a) The peritoneal reflection along the dorsal border of the transverse colon is incised. (b) The proper plane of dissection between the greater omentum and transverse colon mesentery permits access to the lesser sac

is extensive and the planes between the omentum and transverse colon obliterated, precise finger dissection and judicious use of the electrocautery will often allow the tumor to be separated from the colon. At this juncture, if an en bloc resection with contiguous structures is not required, the omentectomy is completed by identifying, clamping, dividing and ligating

When colonic involvement is such that the omentum cannot be separated from the transverse colon as it extends out to the splenic flexure, the omentum is removed in continuity with the involved portion of the transverse colon. Because the anastomotic vascular arcade between the middle colic and left colic arteries (marginal artery of Drummond) may be unpredictable in the region of the splenic flexure, it may be advisable to incorporate the proximal descending colon within the scope of resection. This will ensure a reliable blood supply to the planned anastomosis between the proximal transverse colon and the descending colon. The en bloc resection requires complete mobilization of the hepatic and splenic flexures by dividing the hepatocystocolic and phrenocolic ligaments, respectively. The lesser sac can be entered superiorly through the gastrocolic ligament, which should be completely divided from the greater curvature of the stomach to provide adequate access to the transverse colon and splenic flexure. The stomach is reflected cephalad, exposing the pancreas lying retroperitoneally and the mesentery of the transverse colon. The extent of bowel to be resected is delineated and windows are created in the subjacent mesentery. Standard GIA stapling devices are used to divide the colon proximally and distally (Figure 9.6). Care must be taken to avoid compromising the blood supply to the remaining bowel by confirming the integrity of the marginal artery in both proximal and distal segments. The mobility of the hepatic flexure is confirmed and the descending colon reflected medially by incising the peritoneal reflection of the paracolic gutter

269

Ch 09 final.qxd

12/18/2006

15:29

Page 270

SURGERY FOR OVARIAN CANCER

(white line of Toldt). Maximum mobilization of both proximal and distal bowel segments is necessary to ensure a tension-free anastomosis. We prefer performing a side-to-side stapled anastomosis when feasible; however, if this cannot be done free of tension, an end-to-end anastomosis employing the circular EEA stapling device might prove a better choice. These techniques are described in Chapter 7. Alternatively, a stapled end-to-end anastomosis can be accomplished by the triangulation technique, using three separate applications of the TA stapling device placed at 60° angles to one another (Figure 9.7). Stay sutures are placed at the antimesenteric border and approximately two-thirds of the way toward the mesenteric border on either side, dividing the planned anastomosis into three approximately equal sections. The first application of the TA stapling device produces an inverting staple line, while the latter two staple lines are placed such that the bowel edges are everted. Following creation of the anastomosis, the mesenteric defect should be repaired with interrupted delayed absorbable or silk sutures to prevent internal hernia

formation. Finally, if an anastomosis is not feasible because of undue tension, a permanent ascending colostomy can be developed. The preservation of as much length of the ascending colon as possible from a tumor debulking standpoint is paramount, so as to optimize fluid absorption and minimize the creation of a wet colostomy.

Splenectomy When metastatic ovarian cancer affects the spleen, it is typically due to direct extension of tumor from the greater omentum to the splenic hilum or capsule along the gastrosplenic or splenocolic ligaments. The precise incidence of splenic involvement at the time of diagnosis is unknown; however, splenectomy is performed as part of primary cytoreductive surgery for advanced ovarian cancer in 5–11% of cases.1–4 Splenic metastases are more common among patients with recurrent disease, with one large autopsy study reporting an incidence of 19.6%.5 In the setting of recurrent disease, the spleen may occasionally represent an isolated site of relapse.6–8 Clinically suspicious

Left gastric artery Common hepatic artery Splenic artery Pancreas

Middle colic artery Superior mesenteric artery

Phrenocolic ligament divided Peritoneal incision for mobilization of splenic flexure Line of incision of descending colon

Omentum and tumor

Figure 9.6 En bloc resection of omentum with transverse colon and take-down of the splenic flexure

270

Ch 09 final.qxd

12/18/2006

15:29

Page 271

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

Allis clamp Posterior traction suture Anterior traction suture

a

b

c

d

e

f

Figure 9.7 Triangulation technique of stapled end-to-end anastomosis. (a) The bowel edges are approximated with stay sutures placed two-thirds of the way between the mesenteric and antimesenteric borders and at the antimesenteric borders. (b) The posterior wall is closed with the TA stapler, inverting the bowel edges. (c) One of the antimesenteric stay sutures is placed through both bowel segments to approximate the edges. (d and e) The remaining bowel defect is closed with two additional applications of the TA stapler placed at a 60° angle to one another, everting the bowel edges. (f) The completed anastomosis

metastatic involvement of the spleen is not always confirmed histologically. In their series of 34 patients undergoing splenectomy for suspected ovarian cancer metastases, Ayhan et al. reported that 78% of specimens had pathologic evidence of splenic or perisplenic soft tissue tumor involvement.9 When splenectomy is indicated as part of an overall debulking procedure, the midline incision must be extended to the level of the xyphoid process. Rarely, splenectomy for ovarian cancer may be performed through a subcostal incision if the spleen is thought to be an isolated focus of recurrent disease (Figure 9.8). The surgical approach to splenectomy is individual-

ized and may proceed anteriorly through the gastrosplenic ligament or posteriorly through the lienorenal ligament, depending on the pattern and extent of tumor growth. In either case, a thorough understanding of the attachments and blood supply to the spleen is essential (Figure 9.9). Anterior approach to splenectomy

If the anterior surface of the spleen and splenic hilum are not obscured by tumor, the anterior approach is the most direct route to mobilize and control the splenic blood supply. Traction is exerted on the stomach medially, and the inferior margin of the

271

Ch 09 final.qxd

12/18/2006

15:29

Page 272

SURGERY FOR OVARIAN CANCER

B

A

Figure 9.8 Splenectomy may be approached through a vertical midline (A) or subcostal incision (B)

gastrosplenic ligament, containing the left gastroepiploic vessels, is divided between clamps and secured, if this was not previously performed as part of the omentectomy (Figure 9.10a). The body and tail of the pancreas are identified within the posterior wall of the lesser sac. The tortuous course of the splenic artery can be palpated along the upper margin of the pancreas and traced distally toward the splenic hilum. The peritoneum overlying the splenic artery is carefully incised and a right-angle clamp introduced beneath the artery, with care taken not to disturb the splenic vein lying immediately underneath. Two 2-0 silk ligatures are drawn beneath the splenic artery and carefully tied (Figure 9.10b). Preliminary ligation of the splenic artery permits blood to drain from the spleen, providing an autotransfusion, and allows the spleen to shrink in size, making its removal somewhat easier. After the splenic artery is secured, the two to five short gastric arteries running in the upper portion of the gastrosplenic ligament are serially ligated and

Left hemi-diaphragm Stomach

Gastrosplenic ligament

Lesser sac

Splenic vein

Splenic artery

Spleen Pancreas

Kidney

Lienorenal ligament

Figure 9.9 The spleen: ligamentous attachments and vasculature

272

Ch 09 final.qxd

12/18/2006

15:29

Page 273

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

Gastrosplenic ligament Stomach

a

Colon Spleen Splenocolic ligament

Splenic artery Stomach

b

Gastroepiploic vessels Pancreas

Spleen

divided; these sutures must not encroach upon the stomach wall, since this may lead to gastric necrosis and fistula formation. An attempt is made to deliver the spleen into the incision by passing a hand laterally and exerting anterior and medial traction (Figure 9.11a). The splenocolic ligament should be taken down if not done previously during mobilization of the colon. Dense adhesions may be present between the spleen and the anterior abdominal wall or diaphragm peritoneum, which are divided with electrocautery or excised with the splenic capsule by dissecting in the subperitoneal plane. As the spleen is mobilized, the lienorenal ligament is exposed on the posterior lateral surface of the spleen and carefully incised. A combination of sharp dissection and gentle blunt dissection are used to reflect the underlying kidney and adrenal gland posteriorly, allowing the spleen to be delivered entirely into the incision (Figure 9.11b–c). In the event that the tail of the pancreas abuts the splenic hilum, it should be gently mobilized away from the splenic vessels by blunt dissection with a laparotomy sponge (Figure 9.11d–e). The spleen is then drawn upward and laterally to expose the medial surface of the splenic vessels, which are carefully separated from adjacent tissues (Figure 9.12a). Three clamps are applied to the splenic artery and vein individually; the vessels are divided between the middle and distal clamps and doubly ligated (Figure 9.12b). Although the splenic artery has been previously ligated, many surgeons will place two additional ties distal to the initial ligature, with the final tie on the medial side secured in place with a transfixion stitch. Separate ligation of the splenic artery and splenic vein reduces the risk for the development of an arteriovenous fistula. Posterior approach to splenectomy

Figure 9.10 Splenectomy – anterior approach. (a) The gastrosplenic ligament, including the left gastroepiploic vessels, is divided between clamps. (b) The lower short gastric vessels are divided to further expose the anterior surface of the spleen, and the splenic artery is isolated and ligated

The posterior approach to splenectomy is advantageous when bulky omental disease with extensive tumor infiltration of the gastrocolic and gastrosplenic ligaments obscures the splenic hilum and allows only limited surgical access to the splenic vessels via the

273

Ch 09 final.qxd

12/18/2006

15:29

Page 274

SURGERY FOR OVARIAN CANCER

Point of incision Pancreas Spleen

a

Pancreas Lienorenal ligament

Lienorenal ligament c b

Spleen

Liver Diaphragm Perirenal fat Pancreas

Pancreas

e d

Figure 9.11 Splenectomy – anterior approach, continued. (a) Medial mobilization of the spleen exposing the lienorenal ligament. (b) The lienorenal ligament is sharply incised. (c) Continued medial traction allows the pancreas to fall away from the splenic vascular pedicle. (d) The tail of the pancreas is further mobilized from the splenic hilum by gentle blunt dissection with a laparotomy sponge. (e) The only remaining attachments are the splenic vessels

anterior approach (Figure 9.13a). In this technique, the stomach is placed on medial traction, but no attempt is made to divide the gastrosplenic ligament at this juncture; any attachments of the spleen to the anterior abdominal wall and diaphragm are taken down, and the splenocolic ligament is divided, if this was not performed previously. Placing the patient in the reverse Trendelenberg position permits the transverse colon and small bowel to fall free of the upper

274

abdomen and facilitates exposure of the spleen. The spleen is rotated medially and anteriorly, exposing the vessels in the pedicle from the lateral side (Figure 9.13b). Individual ligation of the splenic vessels is once again preferred. The lienorenal ligament is sharply incised and the splenic artery identified by palpation, and isolated through gentle sharp dissection. The tail of the pancreas should be adequately mobilized away from the splenic hilum. Three clamps

Ch 09 final.qxd

12/18/2006

15:29

Page 275

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

Original ligature on splenic artery

a Original ligature on splenic artery

the specimen has been removed. The spleen is gently compressed to ensure an autotransfusion before the splenic vein is secured between clamps and divided, and the proximal pedicle doubly ligated. The specimen is then further rotated into the incision, exposing the undersurface of the gastrosplenic ligaments and the associated short gastric vessels along the upper portion of the greater curvature of the stomach, which are individually skeletonized, ligated, divided and secured (Figure 9.13d). Care should be taken not to include a portion of the gastric wall within the ligatures, which may be especially challenging when the gastrosplenic ligament is extensively infiltrated with tumor. Following either approach to splenectomy, a drain should be placed in the splenic bed and, postoperatively, the serum and drain fluid amylase levels can be monitored to be certain that the pancreas has not been injured. Complications and literature review of splenectomy in ovarian cancer

Splenic artery

b

Splenic vein

Figure 9.12 Splenectomy – anterior approach, continued. (a) The spleen is drawn laterally, exposing the medial surface and the splenic vessels, which are carefully dissected from any remaining surrounding soft tissue. (b) The splenic artery and vein are clamped, divided and suture ligated

are placed across the splenic artery, and the artery is divided between the distal and middle clamps (Figure 9.13c). The proximal pedicle is doubly ligated, with the more superficial ligature secured with a transfixion stitch; it may be easier to defer this step until after

Perioperative complications associated with splenectomy include hemorrhage, splenic vein thrombosis, arteriovenous fistula between the splenic artery and vein, gastric fistula, pancreatitis, pancreatic pseudocyst formation, left-sided atelectasis, pneumonia and sepsis from encapsulated organisms. Sepsis following splenectomy is the most feared complication.10 The post-splenectomy diagnosis of sepsis is confounded by the fact that leukocytosis is a physiologic response to splenectomy, similar to the phenomenon of postsplenectomy thrombocytosis.11 Konstantinos et al. examined predictors of post-splenectomy sepsis in 118 trauma patients and found that a white blood cell count greater than 15 × 103/µl and a platelet count/white blood cell count ratio of less than 20 on postoperative day 5 were statistically significant and independent predictors of sepsis and should not be considered part of the physiologic response to splenectomy.12 A single one-time polyvalent pneumococcal vaccination (Pneumovax) and a vaccine against Haemophilus influenzae should be given 7–10 days preoperatively to patients for whom splenectomy

275

Ch 09 final.qxd

12/18/2006

15:29

Page 276

SURGERY FOR OVARIAN CANCER

Stomach

Lesser sac Gastrosplenic ligament Stomach Pancreas

a

b

Splenocolic ligament

Splenic artery Gastrosplenic ligament

Lesser sac d c

Pancreas

Spenic vein

Figure 9.13 Splenectomy – posterior approach. (a) Tumor obscures the gastrosplenic ligament and anterior access to the splenic hilum; the spleen is drawn medially to expose the lienorenal ligament. (b) The lienorenal ligament is incised and the splenic artery and vein identified within the areolar tissue. (c) The splenic artery and vein are clamped, divided and ligated. (d) After division of the vasculature, the lesser sac can be properly developed and any remaining tumor attachments to the gastrosplenic ligament taken down and the specimen removed

is planned, to reduce the risk of serious infection from encapsulated organisms. In the scenario of an unplanned splenectomy, the vaccines should be given immediately postoperatively. Once again, arteriovenous fistulas can be prevented by ligating the splenic artery and vein individually.

276

The existing literature does not contain many large series summarizing the outcomes of women undergoing splenectomy as a component of primary cytoreductive surgery for ovarian cancer. One of the earliest case series was reported in 1989 by Sonnendecker et al.13 Their group described six

Ch 09 final.qxd

12/18/2006

15:29

Page 277

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

women who underwent splenectomy in the setting of advanced disease; five of the procedures were performed for splenic metastases and one for a capsular avulsion injury and resultant bleeding. Late thrombotic complications occurred in four of the patients (67%), and resulted in the death of one patient from a massive pulmonary embolism 20 days after surgery. Three patients were disease free at 14 months, 25 months and 32 months after surgery, respectively. Morris et al. reviewed 45 patients undergoing splenectomy at the M D Anderson Cancer Center between 1970 and 1989, of which 27 operations (60%) were performed in women with advancedstage ovarian cancer.14 The procedure was planned preoperatively in only nine patients (20%). Thirteen patients (29%) underwent splenectomy because of injury to the spleen, most commonly due to traction during omentectomy, resulting in capsular laceration. In 32 patients (71%), the splenectomy was accomplished for tumor reduction, and residual tumor following cytoreduction was smaller than 2 cm in 62.5% of patients. Postoperative morbidity was observed in 13 women (29%), two of whom each had sepsis and gastrointestinal bleeding. No cases of late sepsis, severe thrombocytosis, subphrenic abscess or pancreatic pseudocyst were reported. In 1995, Nicklin et al. published a series of 18 patients who underwent splenectomy as a component of their surgery leading to optimal debulking of an ovarian carcinoma.15 The morbidity attributable to the splenectomy included atelectasis and/or effusion in eight women, pancreatic tail injury in four, and thrombocytosis of > 105/µl in three. One patient each suffered a pancreatic pseudocyst, a partial left adrenalectomy and a pulmonary embolism. There were no cases of serious post-splenectomy infection. Chen et al. reported 13 women who underwent splenectomy at the time of primary cytoreduction and 22 others for whom the procedure was performed during secondary debulking.7 Preoperative splenic involvement was diagnosed in 77.3% of recurrent cases versus 15.4% of primary cases. Cytoreduction to disease of less than 1 cm was achieved in 100% of

patients undergoing primary debulking and in 86% of patients operated on for recurrent disease. Major morbidity occurred in 23.1% of patients undergoing primary debulking and in 28.6% of patients operated on for recurrent disease. The investigators observed that combining splenectomy with other cytoreductive procedures may increase the morbidity of the splenectomy itself. The median progression-free interval was 24 months in primary patients and 14 months in secondary patients. Finally, among secondary patients, the median survival time after splenectomy and cytoreduction was 41 months with a median follow-up of 17 months. The technique of hand-assisted laparoscopic splenectomy has been reported in the setting of isolated recurrence of ovarian cancer.16,17 The splenectomy is carried out with the patient in the semilateral position with the right side down. In addition to the operative laparoscopic port sites, a transverse incision is made below the level of the umbilicus on the left side of the abdomen to admit the surgeon’s hand to help with the dissection. The spleen is grasped and rotated, and the endo-GIA stapler (Ethicon, Cincinnati, OH) guided across the splenic vessels and fired. The short gastric vessels are clipped and divided, and the freed spleen removed through the incision in the left lower abdomen. Splenorraphy may be an alternative to splenectomy for limited capsular involvement or to repair a traumatic laceration of the spleen sustained during removal of the omentum. In 1991, Patsner and Rose reported four patients who underwent cavitron ultrasonic surgical aspirator (CUSA) splenorraphy to resect capsular metastases in order to achieve optimal cytoreduction.18 Following total omentectomy, gentle traction was employed to mobilize the spleen. The CUSA straight handpiece attached to a CUSA System 200 Macro-Dissector (Valley Lab, Boulder, CO) employing an aspiration frequency of 23 000 Hz/s was used to fragment and suction ovarian tumor deposits. Two patients had tumor on the inferior and inferolateral portion of the spleen and two patients had tumor on the hilum. Deposits ranged from 2 to

277

Ch 09 final.qxd

12/18/2006

15:29

Page 278

SURGERY FOR OVARIAN CANCER

5 cm in size, and operating time for CUSA removal ranged from 10 to 15 min. Additional sharp dissection was not required in any patient, and all visible disease on the spleen surface was removed. No patient had immediate or delayed splenic capsule rupture or any postoperative infectious or embolic morbidity. The authors advocated CUSA splenorraphy as an alternative to splenectomy and its associated potential morbidities in patients with capsular metastases. The most common cause of splenic laceration is traction injury. With undue traction on the lienoomental band of Lord and Gourevitch during omentectomy or mobilization of the splenic flexure, avulsion of the splenic capsule may occur. When splenorraphy is attempted, the midline incision should be extended to the xyphoid process. Hemorrhage may be immediately controlled by manual compression of the splenic pedicle once the ligamentous attachments have been divided, and adequate mobilization of the spleen has brought it into the operative field. Small capsular lacerations may be controlled with pressure and the application of a hemostatic agent such as oxidized cellulose (e.g. Surgicel®, Johnson and Johnson, New Brunswick, NJ) or by using the argon beam coagulator to secure hemostasis. Larger surface injuries may be repaired by injecting powdered microfibrillar collagen (e.g. Avitene®, Davol Inc., Cranston, RI) into the laceration and closing the defect with through-and-through mattress sutures (Figure 9.14). Alternatively, a portion of healthy omentum can be sutured onto the bleeding site to tamponade it. Stapling devices may also be used to excise a portion of the spleen. If the injury is extensive or involves the hilar region, no attempt should be made to save the spleen and splenectomy is the appropriate surgical option. Other conditions for which splenectomy is indicated over splenorraphy include hemodynamic instability, coagulation defect, splenic parenchymal metastases and major hilar injuries. Taking these issues into consideration, Morris et al. suggest that the opportunities for splenorraphy in patients with ovarian cancer may be limited, while noting that complications associated with splenorraphy may include

278

increased blood loss and even re-operation for bleeding.14

Distal pancreatectomy There exists a limited role for distal pancreatectomy in women with metastatic ovarian cancer. Even when the omentum, stomach and spleen are densely involved with metastatic cancer, the pancreas can normally be dissected safely out of the field of resection. Nevertheless, ovarian cancer involving the splenic hilum may, on occasion, encroach upon or invade the tail of the pancreas, such that distal pancreatectomy is required for complete clearance of left upper abdomen disease (Figure 9.15). In addition, a radical resection of disease in this region may be associated with injury to the tail of the pancreas, mandating a partial resection or repair. Such an injury can result in troublesome bleeding or extravasation of pancreatic enzymes leading to pancreatitis, peritonitis and pancreatic pseudocyst formation. The surgeon

Powdered microfibrillar collagen

Figure 9.14 Splenorraphy. An avulsion injury to the splenic capsule is repaired by injecting microfibrillar collagen (Avitene) into the laceration and placing through-and-through mattress stitches of delayed absorbable suture to close the defect

Ch 09 final.qxd

12/18/2006

15:29

Page 279

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

should, therefore, be familiar with the principles of and surgical technique for this procedure. The surgical approach proceeds as for splenectomy described above, with wide mobilization of the structures in the left upper abdomen to provide adequate exposure to the splenic hilum. The gastrosplenic ligament should be taken down if possible, with individual ligation and division of the short gastric vessels, to allow adequate access to the posterior wall of the lesser sac. To permit mobilization of the pancreatic tail involved by tumor or to approach a surgical injury to the gland, the avascular peritoneum along the inferior border of the pancreas should be incised and the inferior mesenteric vein identified as it courses posterior to the pancreas, where it is joined by the splenic vein (Figure 9.16a). An umbilical tape should be used to encircle the distal pancreas anterior and to the left of the inferior mesenteric vein (Figure 9.16b). Both

the spleen and the tail of the pancreas are then drawn anteriorly into the operative field. The splenic artery is dissected with a right-angle clamp along the superior border of the pancreatic tail, ligated and divided as previously described (Figure 9.17). Gentle dissection of the fibrofatty retroperitoneal tissue will lead the

Stomach

Left gastric vessels Pancreas Inferior mesenteric vein Middle colic vessels

a

Line of incision

Liver

Umbilical tape Stomach

Splenic artery Gastrosplenic ligament Pancreas

Pancreas Inferior mesenteric vein

Spleen

b

Figure 9.16 Distal pancreatectomy. (a) The line of peritoneal incision is shown around the inferior and superior surfaces of the distal pancreas; the inferior mesenteric vein and left gastric Figure 9.15 Metastatic ovarian cancer involving the splenic

vessels are carefully identified. (b) The distal pancreas is

hilum with contiguous extension to involve the tail of the

isolated with an umbilical tape placed anterior to and to the left

pancreas

of the inferior mesenteric vein

279

Ch 09 final.qxd

12/18/2006

15:29

Page 280

SURGERY FOR OVARIAN CANCER

surgeon to the juncture of the splenic vein and the inferior mesenteric vein, permitting ligation of the splenic vein while preserving the inferior mesenteric vein. Rarely, it may be necessary to sacrifice the inferior mesenteric vein, which can be done safely but will produce a transient venous congestion of the descending and sigmoid colon. The remaining posterior attachments of the lienorenal ligament are taken down with electrocautery. Following ligation of both the splenic artery and vein, the en bloc resection of the spleen and distal pancreas is completed by dividing the tail of the pancreas with electrocautery along a surgical line that will clear the tumor (or the area of pancreatic injury) (Figure 9.18a). Caution must be employed to avoid injury to the underlying inferior mesenteric vein. The pancreatic duct of Wirsung should be diligently sought for and occluded with a small hemoclip or

Stomach

Spleen Tumor Umbilical tape

Pancreas

Figure 9.17 Distal pancreatectomy, continued: the spleen and distal pancreas are retracted anteriorly and the splenic artery and vein individually ligated and divided

280

suture. The divided pancreas is then closed with a running non-absorbable suture which is reinforced by a second layer of 2-0 silk placed in a mattress fashion (Figure 9.18b–c). Division of the pancreas with a stapler is not recommended, as this may increase the risk of pancreatic fistulae and pseudocyst formation. Liberal use of closed suction drainage is mandatory. Perioperative administration of somatostatin at a dose of 100–150 µg either subcutaneously or intravenously on call to the operating room or upon recognizing that a portion of the pancreas has been injured or will require resection will lead to a decrease in the incidence of postoperative pancreatic fistulae. The dose should be repeated 4 h later.

Diaphragmatic resection Because the route of spread of ovarian cancer into the upper abdomen follows a pathway along the paracolic gutters into the dense network of lymphatics that occupy the undersurface of the hemi-diaphragms, tumor involvement of the diaphragm is a common manifestation of epithelial ovarian cancer. While diaphragmatic disease is most often present on the right side, metastases to the left hemi-diaphragm may also occur, generally as bulky disease in conjunction with extensive omental and splenic involvement. Contiguous extension of disease to the left hemidiaphragm can be effectively extirpated by an en bloc resection of the spleen and/or omentum with the diaphragm peritoneum or inclusive of the underlying muscle, depending on the depth of invasion. Access to diaphragmatic disease is best achieved through a vertical midline incision that reaches the xyphoid process or a large subcostal incision. Exposure is facilitated by use of the Omni Trac retractor or Wilkinson Robot Arm to elevate the costal margin. Exposure of the left hemi-diaphragm is further enhanced if the spleen is removed initially; however, splenic capsular disease may extend to the diaphragm peritoneum or invade into the underlying muscle and penetrate the pleural cavity, requiring a simultaneous en bloc resection with the spleen.19 For disease extending along the left diaphragmatic leaf to

Ch 09 final.qxd

12/18/2006

15:29

Page 281

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

the midline, transection of the left and right triangular hepatic ligaments and the falciform ligament for full mobilization of the liver permits the greatest exposure and maneuverability for the surgeon. Ovarian cancer involving the splenic capsule and its ligamentous attachments can be inseparable from the left hemi-diaphragm peritoneum (Figure 9.19a). An initial attempt should be made to develop the sub-

peritoneal plane of dissection and resect the involved diaphragm peritoneum, leaving it attached to the splenic capsule, as this will allow sufficient mobilization of the spleen to safely identify and control the vascular pedicles during splenectomy. The spleen is retracted medially and the peritoneal covering of the diaphragm is grasped with long Allis clamps and placed on traction. The subperitoneal plane is opened

Splenic vessels divided

Tumor

a

Middle colic vessels

Line of resection Left adrenal gland

Mattress sutures Left gastric artery

b

c

Pancreas

Left renal vein

Figure 9.18 Distal pancreatectomy, continued. (a) The distal pancreas is resected en bloc with the spleen by dividing the tail with electrocautery distal to the ligated splenic vessels. (b) The distal end of the pancreas is closed with a running stitch of delayed absorbable suture followed by a reinforcing layer of 2-0 silk mattress stitches. (c) The completed resection bed after resection of the distal pancreas and spleen

281

Ch 09 final.qxd

12/18/2006

15:29

Page 282

SURGERY FOR OVARIAN CANCER

with electrocautery, and the line of incision carried along the perimeter of the disease (Figure 9.19b). The appropriate plane of dissection is often easiest to develop more caudally, in the region of the splenic flexure and phrenocolic ligament, elevating the peritoneum from the underlying muscle and proceeding

upward toward the area of gross tumor involvement. The Allis clamps are replaced onto the free edge of peritoneum on the specimen side for additional counter-traction, and the involved peritoneum entirely ‘stripped’ from the underlying diaphragm muscle (Figure 9.19c). Inadvertent entry into the

Left hemi-diaphragm Stomach

Gastrosplenic ligament

Lesser sac Splenic artery

Splenic vein

Spleen Pancreas

Kidney

Splenorenal ligament a Spleen

Line of incision Diaphragm (peritoneum)

b

c

Figure 9.19 Management of left hemi-diaphragm disease. (a) Cross-section of splenic metastasis extending to the diaphragm; the resection is initiated in the subperitoneal plane anterior to the tumor mass. (b) The diaphragm peritoneum is placed on tension with Allis clamps and incised with electrocautery. (c) Clamps are placed on the specimen side of the peritoneal margin to provide additional counter traction as it is elevated from the underlying muscle. For full-thickess resection, the diaphragm muscle is incised with electrocautery and the lesion circumscribed

282

Ch 09 final.qxd

12/18/2006

15:29

Page 283

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

chest cavity in this circumstance can be managed by suturing the defect with large-caliber (1-0) silk or permanent monofilament sutures in interrupted or figure-of-eight stitches and placing a chest tube to effect prompt expansion of the lung on the side of injury. Large bulky lesions of the spleen extending to and infiltrating into the diaphragm soft tissue can be managed by full-thickness resection of the muscle wall. If possible, it is preferable to secure and divide the splenic vessels and short gastric vessels initially and reflect the spleen laterally before addressing the diaphragm muscle resection. This will provide for maximal exposure to the posterior left upper abdomen to delineate the extent of resection required, ensuring that no more diaphragm muscle is excised than is necessary for disease clearance. Following assessment of the disease distribution, areas of tumor for resection are grasped with long Babcock or Allis clamps and, with the diaphragm partially everted, the tumor mass is circumscribed and resected with electrocautery. A 24 Fr chest tube is inserted through the chest wall under direct vision. The diaphragmatic defect is then closed with interrupted horizontal mattress stitches of 1-0 silk or permanent monofilament suture if the defect is small enough to permit primary closure. An airtight seal is confirmed by filling the upper abdomen with normal saline after placing the patient in steep Trendelenburg position. The lungs are hyperinflated by the anesthesiologist and the pool of saline in the left upper abdomen assessed for bubbles of air arising from the diaphragmatic surface. Defects in the repair are over-sewn and the test repeated. Large resections with resulting wide defects may be effectively closed using a Dexon mesh sutured circumferentially to the defect margin with interrupted stitches or permanent monofilament suture. Postoperatively, chest tubes are removed typically after 48–72 h, following the usual criteria for chest tube management. There are few case series examining the role of diaphragmatic muscle resection in the setting of metastatic ovarian cancer. In 1989, Montz et al. described the outcomes of 14 women who underwent

resection of the diaphragmatic peritoneum and/or muscle.20 Ninety-three per cent (13 of 14 patients) underwent a complete resection of diaphragmatic disease, rendering these patients optimally resected. The maximum size of specimens ranged from 12 to 17 cm. The mediastinum was entered in two patients and four had resection of diaphragmatic muscle. One patient who did not have muscle resection had a 30% pneumothorax that spontaneously resolved. No subdiaphragmatic hematomas or abscesses occurred. The mean patient survival was 23 months post-surgery with a mean follow-up of 27 months. Six patients were alive at the time of publication, 7–38 months after diaphragm resection. Kapnick et al. reported 11 patients who underwent diaphragmatic resection. Full-thickness infiltration of the diaphragm by tumor was present in all instances.19 Four patients required splenectomy to permit adequate exposure of the left hemi-diaphragm. Women who underwent diaphragmatic resection at initial surgery for ovarian cancer had a median survival of only 8 months, while those who underwent resection for persistent or recurrent disease seemed to do better, with four of five women alive at 16+ to 33+ months following surgery. Small-volume disease (e.g. studding on the undersurface of the diaphragm) that is not adherent to the splenic capsule may be ablated with the argon beam coagulator, excised with an electrocautery, or aspirated using the CUSA.21,22 Adelson reported his experience of 33 women who underwent cytoreduction of diaphragmatic metastases using the CUSA; 30 of these women had metastatic epithelial ovarian cancer, 13 of whom had nodules greater than 1.5 cm in diameter.23 At completion of the procedures, one patient had no gross diaphragmatic residual disease, and 30 had residual disease measuring 5 mm or less in diameter on the diaphragm. No postoperative complications were attributable to the use of the CUSA.

Partial gastrectomy Although an omental metastasis may encroach upon the greater curvature of the stomach, the need for a

283

Ch 09 final.qxd

12/18/2006

15:29

Page 284

SURGERY FOR OVARIAN CANCER

partial gastrectomy in the setting of advanced epithelial ovarian cancer is rare. Because most disease that involves the stomach affects only its serosa or seromuscular layer, such disease can usually be excised without a full-thickness resection of the stomach wall. Occasionally, however, disease involving the stomach wall cannot be dissected free and must be managed by partial gastrectomy. After the omentum is dissected off the transverse colon, the peritoneum overlying the transverse mesocolon is separated from the omentum and gastrocolic ligament, carefully avoiding injury to the middle colic vessels. Dissection is extended to the greater curvature of the stomach, the lesser sac is entered and developed, and the posterior wall of the stomach mobilized completely. If a splenectomy is to be a part of the surgical plan for cytoreduction, it should be performed at this time following the techniques outlined above. If splenectomy is not required, the spleen should be protected from excessive traction applied to the stomach by dividing the gastrosplenic ligament and ligating the short gastric vessels. If not already done, the left and right gastroepiploic vessels are ligated and divided. The omental tumor within the gastrocolic ligament should be completely mobilized except for the area of planned resection along the greater curvature of the stomach. The tumor mass is placed on downward traction, tenting the greater curvature of the stomach. An extended (90 mm) automated TA stapling device (TA90) is then placed parallel to the axis of the stomach above the area of tumor involvement and fired, laying down a double row of staples (Figure 9.20). The tumor-laden segment of the greater curvature is then sharply excised and the staple line oversewn with interrupted inverting (Lembert) stitches of 3-0 silk sutures to secure the closure and provide for additional hemostasis. Placement of an intraperitoneal closed suction drain and prolonged nasogastric decompression is advisable. Disease involving the lesser omentum requiring resection of the lesser curvature of the stomach is an extraordinarily rare occurrence. In the event that cytoreduction of disease in this location will

284

contribute significantly to an overall optimal surgical result, the resection should be approached via the classic Billroth II anterior gastrojejunostomy because of the proximity of the pylorus and duodenum to the lesser curvature of the stomach. Following removal of the greater omentum, a right-angle clamp is used to gently dissect the peritoneum to the base of the mesocolon until the inferior edge of the pancreas is identified. The peritoneum and fat overlying the pancreas are dissected and removed, moving in a cephalad direction. Exposure of the second portion of the duodenum and the head and body of the pancreas allows the dissection to be continued up to the first portion of the duodenum and superior edge of the pancreas. Since the duodenum does not need to be divided (as in cases of primary gastric carcinoma), the cholecystoduodenal ligament may be preserved. Gentle

Duodenum

Stapler

Antrum

Figure 9.20 Partial gastrectomy, greater curvature. The adherent omental tumor has been mobilized and the involved segment of the greater curvature of the stomach is resected with the TA stapling device; the staple line is reinforced with interrupted inverting stitches of 3-0 silk suture

Ch 09 final.qxd

12/18/2006

15:29

Page 285

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

traction of the distal stomach provides sufficient tension for blunt dissection of the hepatoduodenal ligament with a right-angle clamp, which is then transected, exposing the porta hepatis. The lesser omentum is taken down from its attachment to the liver (hepatogastric ligament) (Figure 9.21a). At this point, the left lobe of the liver is mobilized off the left diaphragm, retracted to the right side, and stabilized with a self-retaining retractor. While the surgeon retracts the stomach and lesser omentum to the left, the splenic artery is identified along the superior margin of the pancreas and followed back to its origin from the celiac axis, where the left gastric artery is located. The left gastric artery is skeletonized, ligated and divided (Figure 9.21b). Note that, for lesser curvature disease, the short gastric arteries are preserved, and for both lesser and greater curvature disease, the right gastric artery is preserved. The lesser omentum (hepatogastric ligament) and associated tumor are then folded over, and the distal stomach and lesser curvature resected between two applications of the TA90 automated stapling device, with the distal staple line placed just proximal to the pylorus (Figure 9.21c). A gastrojejunostomy is created to re-establish intestinal continuity by dividing a loop of proximal jejunum about 20 cm from the ligament of Treitz with the GIA stapler. The jejunal mesentery is then taken down so that the distal limb can reach the greater curvature of the stomach. In the posterior approach to gastrojejunostomy, the distal jejunal limb is brought up through a small hole that is created in the left mesocolon, placed near the base of the mesentery and to the left of the middle colic vessels and approximated against the posterior wall of the stomach with stay sutures approximately 3 cm proximal to the stomach staple line. Stab wounds are created in both the distal jejunal limb and the stomach wall, and one end of the GIA stapler is introduced into each lumen (Figure 9.22a). The GIA stapler is fired, and the common defect closed with the TA stapler (Figure 9.22b) and the patency of the lumen confirmed by palpation (Figure 9.22c). The posterior approach to gastrojejunostomy is usually easier to perform than

Esophagus Hepatogastric ligament Tumor

a Left gastric artery and vein

Stomach

b

Pancreas

Beginning of duodenum

Stapler

Antrum c

Figure 9.21 Partial gastrectomy, lesser curvature. (a) The lesser omentum and hepatogastric ligament are taken down between clamps medial to the area of tumor involvement. (b) The left gastric vessels are identified, ligated and divided. (c) The distal stomach and lesser curvature are resected between two applications of the TA stapling device

285

Ch 09 final.qxd

12/18/2006

15:29

Page 286

SURGERY FOR OVARIAN CANCER

Duodenum Pylorus Stomach

Ligament of Treitz Distal limb of jejunum brought up for gastrojejunostomy Proximal duodenum divided a

Stoma completed

b

Closure of stab wounds

Gastrojejunal stoma

the anterior approach, in which the distal jejunum is brought anterior to the transverse colon and the anastomosis created along the anterior surface of the greater curvature of the stomach. Finally, continuity between the proximal jejunal segment and the alimentary stream is re-established by completing a jejunojejunal anastomosis approximately 40 cm distal to the gastrojejunostomy.

Celiac axis Ovarian cancer involvement of the celiac trunk is fortunately uncommon. Bulky disease may spread retrograde, from intestinal metastasis, along the superior mesenteric vasculature to the celiac axis; this generally precludes optimal cytoreduction. On occasion, however, suprahilar para-aortic adenopathy can extend cephalad and manifest as enlarged celiac nodal disease that may be more amenable to resection or enucleation, provided exposure is adequate. With care taken not to injure the underlying splenic vessels, a vein retractor is used to apply gentle downward traction on the upper margin of the pancreas to expose the celiac axis. The stomach is then retracted cephalad and the peritoneum along the superior border of the pancreas is incised, entering the retroperitoneal space. The superior mesenteric ganglion may be seen as a shiny white bundle of nerves along the anterior surface of the aorta at the level of the superior mesenteric artery, and should not be sacrificed. The areolar sheath on the anterior surface of the aorta is incised and a plane of dissection is established between the vessel adventitia and the nodal mass. The dissection should proceed in a controlled fashion, moving from

Figure 9.22 Posterior gastrojejunostomy. (a) One prong of the GIA stapling device is introduced into either limb of the stomach and jejunum. (b) The common lumenal defect

GIA stapler to restore intestinal continuity

produced by the GIA stapler is closed with a single application of the TA stapler, incorporating the full-thickness of stomach and jejunal walls. (c) The completed anastomosis is checked for

c

286

patency by digital palpation; jejunojejunal anastomosis is completed 40 cm from gastrojejunostomy

Ch 09 final.qxd

12/18/2006

15:29

Page 287

CYTOREDUCTIVE SURGERY: LEFT UPPER ABDOMEN

Stomach Left gastric artery

Duodenum

Aorta Celiac trunk

Common hepatic artery

Splenic artery Superior pancreatic artery

Pancreas Superior mesenteric artery

Middle colic artery

Figure 9.23 Celiac axis disease. The superior margin of the pancreas is reflected downward, exposing the celiac vessels. The tumor mass is carefully resected from the surrounding vessels. Generally, only bulky adenopathy extending from suprahilar para-aortic nodes is amenable to resection in the region of the celiac axis

caudad to cephalad and lateral to medial (Figure 9.23). The nodal bundle is carefully excised from the juxtaposed common hepatic, left gastric, splenic and superior pancreatic arteries. Particular caution is necessary along the posterolateral surface of the aorta, as the middle suprarenal arteries, supplying the adrenal glands, may be encountered in this region and should be preserved.

vasculature is essential to the surgeon undertaking extirpative operations in this region. As with all cytoreductive operations for ovarian cancer, sound surgical judgment must be exercised to ensure that radical resection of disease involving spleen, stomach, transverse colon and surrounding structures is considered within the context of the overall operation and, therefore, will contribute significantly to achieving an end result of minimal residual disease.

CONCLUSION Ovarian cancer metastasic to the left upper abdomen is not uncommon among patients with advancedstage disease. A thorough knowledge of the anatomic relationships between the viscera and their associated

REFERENCES 1.

Gershenson DM. Primary cytoreduction for advanced epithelial ovarian cancer. Obstet Gynecol Clin North Am 1994; 21: 121–40

287

Ch 09 final.qxd

12/18/2006

15:29

Page 288

SURGERY FOR OVARIAN CANCER

2.

Bilgin T, Ozerkan K, Ozan H. Splenectomy in cytoreductive surgery for advanced ovarian cancer. Arch Gynecol Obstet 2005; 271: 329–31

3.

Malfetano JH. Splenectomy for optimal cytoreduction in ovarian cancer. Gynecol Oncol 1986; 24: 392–4

4.

Deppe G, Zbella EA, Skogerson K, et al. The rare indication for splenectomy as part of cytoreductive surgery in ovarian cancer. Gynecol Oncol 1983; 16: 282–7

5.

Rose PG, Piver MS, Tsukuda Y, Lau TS. Metastatic patterns in histologic variants of ovarian cancer. An autopsy study. Cancer 1989; 64: 1508–13

6.

Gemignani ML, Curtin CC, Barakat RR. Splenectomy in recurrent epithelial ovarian cancer. Gynecol Oncol 1999; 72: 407–10

7.

Chen L-M, Leuchter RS, Lagasse LD, et al. Splenectomy and surgical cytoreduction for ovarian cancer. Gynecol Oncol 2000; 77: 362–8

8.

9.

10. 11.

14.

Morris M, Gershenson DM, Burke TW, et al. Splenectomy in gynecologic oncology: indications, complications, and technique. Gynecol Oncol 1991; 43: 118–22

15.

Nicklin JL, Copeland LJ, O’Toole RV, et al. Splenectomy as part of cytoreductive surgery for ovarian carcinoma. Gynecol Oncol 1995; 58: 244–7

16.

Klingeler PJ, Smith SL, Abendstein BJ, et al. Handassisted laparoscopic splenectomy for isolated splenic metastasis from an ovarian carcinoma. Surg Laparosc Endosc 1998; 8: 49–54

17.

Yano H, Iwazawa T, Kinuta M, et al. Solitary splenic metastasis from ovarian cancer successfully treated by hand-assisted laparoscopic splenectomy: report of a case. Surg Today 2002; 32: 750–2

18.

Patsner B, Rose PG. CUSA splenorraphy for ovarian cytoreductive surgery. Gynecol Oncol 1991; 41: 28–9

19.

Kapnick SJ, Griffiths CT, Finkler NJ. Occult pleural involvement in stage III ovarian carcinoma: role of diaphragm resection. Gynecol Oncol 1990; 39: 135–8

20.

Pate JW, Peters TG, Andrews CR. Postsplenectomy complications. Am J Surg 1985; 51: 437–41

Montz FJ, Schlaerth JB, Berek JS. Resection of diaphragmatic peritoneum and muscle: role in cytoredutive surgery for ovarian cancer. Gynecol Oncol 1989; 35: 338–40

21.

Horowitz J, Leonard D, Smith J, Brotman S. Postsplenectomy leukocytosis: physiologic or an indicator of infection? Am Surg 1992; 58: 387–90

Deppe G, Malviya VK, Boike G, et al. Surgical approach to diaphragmatic metastases from ovarian cancer. Gynecol Oncol 1986; 24: 258–60

22.

Deppe G, Malviya VK, Boike G, et al. Use of cavitron surgical aspirator for debulking of diaphragmatic metastases in patients with advanced carcinoma of the ovaries. Surg Gynecol Obstet 1989; 168: 455–6

23.

Adelson MD. Cytoreduction of diaphragmatic metastases using the cavitron ultrasonic surgical aspirator. Gynecol Oncol 1991; 41: 220–2

Farias-Eisner R, Braly P, Berek JS. Solitary recurrent metastasis of epithelial ovarian cancer in the spleen. Gynecol Oncol, 1993; 48: 338–41 Ayhan A, Al RA, Baykal C, et al. The influence of splenic metastases on survival in FIGO stage IIIC epithelial ovarian cancer. Int J Gynecol Cancer 2004; 14; 51–6

12.

Konstantinos G, Toutouzas G, Velmahos GC. Leukocytosis after posttraumatic splenectomy: a physiologic event or sign of sepsis? Arch Surg 2002; 137: 924–9

13.

Sonnendecker EWW, Guidozzi F, Margolius KA, et al. Splenectomy during primary maximal cytoreductive

288

surgery for epithelial ovarian cancer. Gynecol Oncol 1989; 35: 301–6

Ch 10 final.qxd

12/18/2006

15:30

Page 289

CHAPTER 10

Second-look surgery Ilana Cass, Beth Y Karlan

INTRODUCTION The role of second-look surgery in ovarian cancer has evolved as a result of improved operative technology and treatment outcome. The original intent of the procedure was its therapeutic effect in the era when treatment of ovarian cancer was limited to surgery. Second-look surgery was subsequently used to determine treatment duration following the discovery that ovarian cancer could be successfully treated with chemotherapy. Currently, the debate over the secondlook procedure centers around the value of its prognostic information balanced against the lack of a demonstrated impact on survival. While a small subset of patients derive a therapeutic benefit from secondary cytoreduction at the time of the second-look procedure, second-look surgical assessment has not clearly altered the outcome of most patients. Retrospective reports of patient outcome after second-look surgery suffer from inconsistent study design that often fails to utilize the findings of second-look surgery to determine subsequent therapy. Recent advances in laparoscopic technique have fueled the debate because of the comparative ease and high patient acceptance of a laparoscopic second-look procedure.

HISTORY OF THE SECOND-LOOK PROCEDURE Originally, second-look evaluation was proposed by Wangensteen et al. to prolong survival by removing early recurrences in asymptomatic patients with intra-

abdominal malignancies who were at high risk of recurrence.1 The original second-look procedure prescribed that an asymptomatic patient with lymph node-positive visceral malignancies was taken for reoperation following prior curative surgery to detect recurrent disease. Gilbertsen and Wangensteen described the outcome of their second-look program in over 300 patients, including ovarian cancer patients, over a 13-year period.2 Patients were taken for second-look evaluation 6 months after curative surgery, and those with residual disease were again offered surgical re-evaluation at a later time to determine the salvage rate of the second-look procedure. Although half of the patients had residual disease, they reported a ‘negative final look’ in 10% of patients who had residual disease resected at their prior second-look procedure. They concluded that the second-look surgical program could salvage a small number of patients with no alternative therapies, and was therefore justifiable. Once chemotherapy was shown to have activity in ovarian cancer, second-look laparotomy was proposed as a technique to assess the pathologic response in those patients with a clinical response. Rutledge and Burns described 12 advanced-stage ovarian cancer patients who were taken for a second laparotomy after achieving a clinical response to the alkylating agent Lphenylalanine mustard. No visible residual disease was seen, although lymph nodes were not sampled.3 As more effective agents became available to treat ovarian cancer, second-look surgery gained popularity as the most accurate method to determine disease response. Given the significant toxicity associated

289

Ch 10 final.qxd

12/18/2006

15:30

Page 290

SURGERY FOR OVARIAN CANCER

with prolonged chemotherapy treatment, second-look surgery identified those patients who could safely discontinue therapy.4 Second-look surgery became part of the standard management of advanced-stage ovarian cancer patients because of its ability to assess disease response definitively. If disease was found at secondlook assessment, therapy was altered, using a variety of second-line agents. The second-look procedure then came under scrutiny when several studies failed to show any survival benefit from second-line therapies. Contemporary philosophy regarding the role of second-look surgery is that the procedure has important diagnostic utility, with potential therapeutic value in a small subset of patients. However, several factors have limited the use of second-look surgery as the standard of care. There is no consensus as to the extent of the procedure, and whether the laparoscope can substitute for laparotomy, especially in the setting of no macroscopic disease. There is also very little standardization regarding the timing of the secondlook assessment relative to the primary surgery and chemotherapy. The clinical significance of secondlook surgery is unclear, given that approximately half of patients with negative second-look surgery will have disease recurrence. Finally, the limited success of second-line therapies for patients with residual disease has curbed enthusiasm for the second-look procedure in the past. Considering the cost and associated morbidity of an extended second-look laparotomy, many have questioned the value of the procedure outside a clinical trial. In light of the inconsistent performance of non-invasive techniques in detecting residual or persistent disease, second-look surgery is still an integral part of some prospective randomized trials to establish pathologic response rates to new chemotherapy agents.5–7

DETERMINATION OF DISEASE STATUS The limited sensitivity and specificity of available non-invasive modalities to assess disease status under-

290

scores the value of the second-look procedure. Imaging techniques, such as computed tomography (CT), have a reported diagnostic accuracy of 20–75% for recurrent ovarian cancer. Brenner et al. found that a negative CT was associated with residual disease in 42% of cases.8 Serum tumor markers such as CA-125 are poor predictors of residual disease status. While an elevated CA-125 correlates with residual disease in > 90% of cases, residual tumor is found in 35–50% of patients with a normal CA-125.9 Newer techniques such as 2-[18F]fluor-2-deoxy-Dglucose positron emission tomography (FDG-PET) offer very promising imaging modalities to detect residual or recurrent ovarian cancer based upon the high glycolytic rate of tumors.10 It has been valuable in the detection of many cancers, including breast, lung, colorectal, lymphoma and ovarian.11 FDG-PET imaging exploits the differential metabolic uptake of viable cancer cells. FDG is a positron-emitting glucose analog that remains metabolically trapped in the cancer cell after it is taken up for glycolysis.12 The rate of uptake of FDG and the final concentration that remains in tissues help to differentiate between cancer and non-cancer cells. Sensitivity rates of 10–90% have been reported in recurrent ovarian cancer. Some factors that may contribute to the variable detection of recurrent disease include the small tumor volume, the baseline variable physiologic uptake of other organs, including the bladder and digestive tract, and the absence of identifiable anatomic structures.11 Emerging data from the new hybrid PET/CT imaging systems suggest that this modality may have superior sensitivity and specificity in the detection of residual disease to either modality alone. Ongoing studies are needed to confirm these early reports.13

PATIENT ELIGIBILITY Second-look surgery is considered for appropriately counseled patients with stage II–IV disease within 3–6 months of having completed primary chemotherapy. The patient should have achieved a clinical remission defined as a normal physical examination,

Ch 10 final.qxd

12/18/2006

15:30

Page 291

SECOND-LOOK SURGERY

serum CA 125, and in some series, imaging studies. Second-look surgery is not indicated in the completely stage I patient because of the low yield, and the small risk of recurrence. Multivariate analysis has consistently identified stage of disease as a predictor of residual disease at second-look assessment.14,15 Stage I patients have very low frequencies of positive second-look procedures. Podratz et al. found that only six (12%) of 52 patients with stage I disease had residual disease at second-look laparotomy.16 The Gynecologic Oncology Group (GOG) conducted one of the largest trials in early stage (I and II) ovarian cancer patients, and found that only five (5.2%) of 95 clinically asymptomatic patients had residual disease at second-look laparotomy.17

TECHNIQUE Laparotomy The current second-look procedure closely follows Wangensteen’s original description from 1951. An extensive evaluation of the prior operative area and sites of frequent metastases are performed. If no disease is seen, biopsies from prior cancer sites and those frequently involved with cancer are taken. If disease is encountered, an attempt is made to completely resect the tumor.1 Second-look laparotomy evaluation is performed via a midline incision. Lysis of adhesions is often required for adequate visualization of peritoneal surfaces and restoration of normal anatomy. Biopsy should include any sites of apparent disease, sites of prior disease, even if normal-appearing, and sites where tumor implantation from ovarian cancer is common, e.g. pelvic peritoneum, colonic gutters, falciform ligament and inferior aspects of the diaphragms. While there is no uniformly accepted definition of an adequate second-look assessment in the literature, most agree that visualization of the pelvic and abdominal surfaces as outlined above is necessary.14 Clough et al. employed a rigorous 18-site anatomic checklist to define an adequate second-look

assessment in their prospective study of patients who underwent both laparoscopy and laparotomy. Consequently, they reported the lowest rate of adequate laparoscopic visual assessment, and the highest rate of laparoscopic complications in the literature. Not surprisingly, even laparotomy inadequately assessed all 18 abdominopelvic sites in 5% of their patients.18 Routine lymphadenectomy at the time of secondlook evaluation remains controversial. The incidence of isolated retroperitoneal disease in an otherwise clinically negative second-look evaluation is small, unless the patient had lymph node disease at initial surgery.19 Childers et al. described isolated microscopic para-aortic lymph node disease in two patients with clinically negative second-look laparoscopy who had prior lymph node disease at primary surgery.20 Similarly, Goldberg et al. reported no positive lymph nodes at the time of second-look laparotomy in patients who were lymph node negative at initial surgery, but 33% persistent lymph node involvement in patients with prior lymph node metastases.21 Rubin et al. found no correlation between risk of recurrence and lymph node biopsies at second-look laparotomy.22,23 In the absence of grossly enlarged nodes, lymphadenectomy is generally reserved for those patients with documented lymph node metastases at primary surgery, or those who did not have complete surgical staging performed at the time of primary surgery.

Laparoscopic second-look assessment Laparoscopy has been described as an alternative to laparotomy for second-look assessment for over three decades. Despite rudimentary equipment, many authors reported that they were able to detect gross residual disease in 33–50% of patients, thereby sparing them laparotomy.4,24,25 The magnification and enhanced illumination of laparoscopy should enhance detection of residual disease in many areas such as the diaphragm and paracolic gutters.26,27 Advances in laparoscopic technology have led to wider acceptance of laparoscopic second-look assessment. The same

291

Ch 10 final.qxd

12/18/2006

15:30

Page 292

SURGERY FOR OVARIAN CANCER

systematic approach inspecting the pelvic and abdominal cavities used in reassessment laparotomy is followed in second-look laparoscopy.20,26 Preoperatively, patients should have a mechanical bowel preparation in order to facilitate mobilization of the bowel, and antibiotic prophylaxis in the event that bowel injury occurs or segmental resection is required. Patients should always be counseled preoperatively about the possible need to convert the procedure to a laparotomy in the event of: (1) bowel or large vessel injury; (2) inadequate visualization secondary to adhesive disease; or (3) requirement for secondary cytoreduction.

Patient positioning The patient may be placed in the dorsal lithotomy position, using padded leg stirrups, or the supine position. A sponge stick placed in the vagina assists with the assessment of the cul de sac. An orogastric or nasogastric tube is placed to aspirate stomach contents, which facilitates entry into the abdominal cavity if a left upper quadrant entry site is used. Decompression of the stomach also improves visualization of the left upper quadrant. The peritoneal cavity can be assessed either periumbilically or from the left upper quadrant. Periumbilical access can be obtained with an open or closed technique, although in general prior abdominal incisions should be avoided, owing to the risk of underlying adhesions between the bowel and abdominal wall. Great effort is made to avoid a prior abdominal incision with entry, in order to prevent adhesions. An open approach using a small periumbilical incision allows the surgeon to enter the peritoneal cavity sharply under direct visualization. A Hassan 8-mm or 10-mm trocar is inserted into the peritoneal cavity and secured to the fascia using anchoring sutures which are used to close the fascial defect. The left upper quadrant entry approach is often preferable to the periumbilical entry in patients with prior midline incisions, because it minimizes the risk of damage to adherent intraperitoneal structures. This approach can also provide good visualization for

292

adhesiolysis along the prior laparotomy site. A pneumoperitoneum should be established before the placement of the trocar. The Veress needle is placed perpendicular to the anterior abdominal wall in the left 10th intercostal space or subcostal margin between the midclavicular and anterior axillary lines to avoid injury to the superior epigastric vessels and the splenic flexure of the colon. This location takes advantage of the natural adherence of the peritoneum to the ribs and makes preperitoneal insufflation unlikely. Deep insertion of the Veress needle is unnecessary. Once an adequate pneumoperitoneum has been established, a 5-mm or 8-mm trocar is inserted perpendicularly just below the left subcostal margin, which capitalizes on the immobility of the rib cage to prevent downward displacement of the anterior abdominal wall.28 A potential limitation to this approach is left upper quadrant adhesions following a prior supracolic omentectomy or splenectomy. After thorough evaluation of the pelvic and abdominal cavity, accessory trocars are introduced under direct visualization lateral to the inferior epigastric vessels. An intraperitoneal washing from the pelvis and abdominal cavity should be obtained with at least 200–500 ml of normal saline and sent for cytologic evaluation. Adhesiolysis, as necessary, is performed in order to visualize peritoneal surfaces, the cul de sac, residual omentum and the hemidiaphragms. Biopsies are then obtained as outlined above in second-look laparotomy with frozen section analysis as indicated. The ideal number of biopsies that are necessary to optimize the laparoscopic detection of residual disease to be obtained is unclear, although various authors have reported a range from 7 to > 100 biopsies at second-look assessment.26,29–32 More biopsies are generally obtained at laparotomy than laparoscopy, although this does not seem to affect the sensitivity for detecting disease. In studies that have compared the two procedures, the range of median number of laparoscopic biopsies is 7–14, while the range is 16–24 at laparotomy.20,29,32,33 The distribution of biopsy sites is probably more important than a fixed number of

Ch 10 final.qxd

12/18/2006

15:30

Page 293

SECOND-LOOK SURGERY

tissue samples, especially for the patient with a clinically negative second-look laparoscopy. Most series of laparoscopic second-look procedures report concordance between cytologic washings and pathologic biopsy in the majority of patients. However, several studies have noted that cytology was the only positive laparoscopic finding in 4–46% of patients.25,26,29,32 In comparison, the rate of cytologypositive, biopsy-negative second-look laparotomies is 1–2%.14,32 This finding underscores the need for complete cytology sampling of both pelvic and abdominal cavities, especially in the patient with a clinically negative second-look laparoscopy.

Intraperitoneal port placement Patients with a clinically negative or small-volume (1–2 mm) positive second-look procedure may be candidates for the placement of an intraperitoneal port placement.34 The catheter can be inserted via a 5-mm re-usable port, grasped and pulled into the peritoneal cavity. The end of the catheter should be placed into the pelvis. The trocar sheath is then removed. The reservoir site is selected, ideally over an area that will provide stability and support to the reservoir base, such as the lower costal margin, which will facilitate port access. The reservoir pocket is created, and a long narrow clamp is advanced from the pocket to the catheter exit site, making a tunnel through the subcutaneous tissue. The catheter is grasped, and brought through the tunnel to the reservoir pocket. The catheter is attached to the reservoir, which is secured in place against the abdominal fascia by means of permanent sutures. Prior to removal of the laparoscope, the catheter is flushed with heparin solution (100 mg/ml) under direct visualization to verify its correct position.35

COMPLICATIONS Complications are reported in 0–15% of laparoscopic second-look procedures (Table 10.1). Complications associated with laparoscopy include superficial wound

infection, subcutaneous hematoma and emphysema, abdominal wall tumor implantation and vascular or bowel injury.20,32 The largest series to date of laparoscopic procedures in women with gynecologic malignancies has reported that subcutaneous tumor implantation occurs in 0.97% of patients.36 Most intraoperative complications relate to trocar injury of viscera secondary to adhesions. The major complications of bowel or vascular injury occur in 2–5% of cases in contemporary series, which is lower than that described in earlier reports.26,29,30 This reduction is probably secondary to the use of bowel preparation, open technique, avoiding prior incision sites and the use of the 5-mm scope, as necessary. Conversion to laparotomy occurs in 6–12% of second-look laparoscopy procedures, the majority of which have been performed to repair bowel perforations.20,26,30,37 Laparoscopic second-look surgery has documented decreased morbidity, cost and patient recovery time compared to laparotomy.29,30,32 Second-look

Table 10.1 Laparoscopic second-look surgery experience

Author

Year

n

Rate (%) of complete exploration with LSC

Major complications with LSC (%)

Smith4

1977

24

80

0

Quinn24

1980

62

76

5

Berek26

1981

57

73

10

Lele37

1986

51

80

6

Childers20

1995

44

88

14

Casey32

1996

126

89

8

Abu-Rustum29 1996

109

100

0

Nicoletto38

1997

54

96

0

Clough18

1999

20

41

15

Husain30

2001

150

98

3

Cass44

2002

108

91

6

LSC, laparoscopy

293

Ch 10 final.qxd

12/18/2006

15:30

Page 294

SURGERY FOR OVARIAN CANCER

laparotomy is associated with a major complication rate of 5–8%, and a minor complication rate of 11–40% of patients. Frequent complications associated with second-look laparotomy include an ileus in 2–15% of procedures, wound hematoma/infection in 6–10%, urinary tract infection in 5%, pulmonary infections in 3% and need for transfusion in 2–4% of patients.14,29,32 Mean hospital stay, blood loss, direct costs and operative time are significantly lower among patients who have laparoscopy. The average length of stay among laparoscopy second-look procedures was 1 day, compared to 7 days among those who had laparotomy. The average blood loss for laparoscopy was under 35 ml compared to 200 ml for laparotomy. The average cost of laparotomy was twice that of laparoscopy, and laparoscopic operative time was approximately two-thirds that for second-look laparotomy.29,32

EQUIVALENCY OF LAPAROSCOPY AND LAPAROTOMY Three factors must be considered to establish that laparoscopy is equivalent to laparotomy as a method of second-look assessment in ovarian cancer patients: first, the feasibility of the procedure, including associated cost and complications; second, the accuracy of laparoscopy to detect disease, and the validity of negative laparoscopic surgery; third, patient outcome should be comparable between second-look laparoscopy and laparotomy procedures. The utility of laparoscopic second-look assessment was limited in earlier studies by high rates of inadequate visualization due to adhesions from prior surgery (Table 10.1). Obscuring adhesions have been described in 3–60% of patients.18,20,24,26,30,32,38 Laparoscopic assessment in many earlier studies was often restricted to visualization and biopsy only, with limited technology for performing adhesiolysis.4,24,26 Patients found to have macronodular disease at laparoscopy that was amenable to resection were uniformly converted to laparotomy for secondary cyto-

294

reduction. Improved laparoscopic technology, including smaller and more advanced instruments, has allowed surgeons to perform more complex procedures including extensive adhesiolysis and tumor resection. Consequently, the rates of inadequate laparoscopic visualization have generally decreased over time to 2–12%. The detection of residual disease at second-look surgery is dependent upon several factors including the histopathologic features of the tumor, prior treatment administered and patient selection. In older reports of second-look surgery outcome, the length of treatment and interval between initial laparotomy and second-look assessment varied considerably. Many studies used prolonged treatment regimens, often with intervals of 12–45 months between initial laparotomy and second-look surgery. Some studies also varied with respect to clinical assessment of disease prior to second-look surgery and included patients with suspected clinical evidence of disease including elevated CA-125 (< 35 U/ml).7,18 Among those studies of patients with predominantly advanced-stage disease treated with combination chemotherapy, 40–70% of patients had residual disease at second-look assessment.3,14,24,39–40 Recent studies that have adhered to more uniform treatment and patient inclusion criteria have found comparable rates of residual disease at laparoscopy and laparotomy in 37–61% of patients.6,20,25,29–32,42,43 The most recent data from second-look surgery outcome in the era of taxane/platinum-based therapy suggests that optimized first-line therapy can further improve the pathologic response rates. A retrospective study of 108 optimally cytoreduced patients treated with paclitaxel and carboplatin reported residual disease in 28% of patients at second-look surgery. The frequency of residual disease was identical between those patients who had second-look laparoscopy versus those undergoing laparotomy.44 The most recent data from the GOG protocol 158 reported that 37% of optimally cytoreduced patients who underwent second-look laparotomy had residual disease after paclitaxel and carboplatin chemotherapy.6

Ch 10 final.qxd

12/18/2006

15:30

Page 295

SECOND-LOOK SURGERY

To determine the equivalency of laparoscopy and laparotomy in detecting residual disease at secondlook assessment requires more than just a comparison of the reported detection rates for each procedure. Laparoscopy should also be an accurate method of detecting residual disease. Laparotomy has historically been the gold standard for the detection of disease at second-look assessment. Laparoscopy has consistently high rates of sensitivity, with variable false-negative rates. False-negative rates of laparoscopy range from 10 to 77% (Table 10.2). Some of this variation relates to laparoscopic technology, and to the definition of negative laparoscopy. The false-negative rates are significantly lower in the more contemporary studies, presumably because of better operative techniques and equipment. Combining clinical and pathologic findings reduces the false-negative rates of laparoscopy to 13–24% among patients with adequate laparoscopic visualization. Several studies have directly compared the reliability of laparoscopy to laparotomy for the detection of residual disease (Table 10.2).4,18,24–26,38 However, these studies have differed in their indication and tim-

ing of conversion to laparotomy. In some studies, laparotomy was performed up to 20 months after laparoscopy, which may impact on the reported accuracy of each procedure.24,26 In most studies, an immediate laparotomy was performed on patients with an incomplete laparoscopic evaluation, or a negative laparoscopic visualization.4,18,25,29 Abu-Rustum et al. recently compared eight patients who had laparotomy immediately following a negative second-look laparoscopy, and found that laparotomy identified only one patient with residual disease not identified by laparoscopy.29 Two prospective studies required that patients had both procedures. Nicoletto et al. randomized 102 women in complete remission, as documented by normal CA-125, CT scan and negative laparoscopy, to second-look laparotomy or to no further intervention.38 They found that 11 patients (24%) had falsenegative laparoscopic assessment, although two patients did not have any cytology or biopsies performed because of extensive adhesions. Patients had comparable survival after a 60-month follow-up period, despite the use of second-line treatment in those

Table 10.2 Comparison between second-look laparoscopy (LSC) and exploratory laparotomy (ELAP) False-negative rate, False-negative LSC LSC visualization visualization/ (n) pathology evaluation (n)

Detection rate of residual disease LSC (n) with LSC (%)

ELAP (n)

Detection rate of residual disease with ELAP (%)

33

11

41

62

52

19

47

22

36

10

—

—

57

38

—

1995

—

44

56

—

1996

—

57

53

69

54

(1) 13%

31

55

8

61

(11) 24%

46

—

46

24

20

40

Author

Year

Smith4

1977

(5) 45%

—

24

Quinn24

1980

(7) 77%

33%*

Piver25

1980

(2) 20%

(1) 10%

Berek26

1981

Childers20 Casey32

Abu-Rustum29 1996

(4) 50%

Nicoletto38

1997

Clough18

1999

—

(2) 14%

20

30

Husain30

2001

—

—

150

54

—

* Two patients had incomplete visualization

295

Ch 10 final.qxd

12/18/2006

15:30

Page 296

SURGERY FOR OVARIAN CANCER

with residual disease detected at laparotomy. Multivariate analysis found that survival was not improved by the addition of laparotomy to laparoscopy. The authors concluded that second-look laparotomy was of no clinical utility in patients with a negative second-look laparoscopy.38 Clough et al. performed an immediate laparotomy following laparoscopy in 20 women, irrespective of disease status at laparoscopy.18 They mandated that second-look evaluation include 18 predetermined sites by both procedures. They found that some degree of adhesions limited their laparoscopic assessment in 60% of patients, but they were nonetheless clinically able to detect gross disease in five of eight patients who ultimately were found to have residual disease. Incorporating the histology results from laparoscopy, an additional patient was found to have residual disease at laparoscopy. In summary, they reported that laparoscopic second-look surgery had a sensitivity of 100% and a negative predictive value of 86%, with two false negatives among 14 patients without any residual disease at laparoscopy. The authors concluded that laparotomy was advised after a negative second-look laparoscopy for more accurate assessment of disease status.18 Evaluation of clinical outcome in other retrospective studies that compared second-look laparoscopy to laparotomy suggests that patient outcome is unaffected by the type of second-look assessment procedure. Casey et al. reported similar patient outcomes between 30 patients who had second-look laparoscopy and 37 who had second-look laparotomy at a median follow-up of 28 months.32 Fifty-three per cent of patients were without evidence of disease after a second-look laparoscopy versus 54% of those after second-look laparotomy.32 Abu-Rustum et al. reported comparable disease-free intervals between patients followed after negative second-look laparotomy versus laparoscopy, although there were a small number of events in each arm.29 Based on the literature, laparoscopy has excellent sensitivity, but has a false-negative rate of approximately 10–15%. Some authors therefore suggest that

296

laparoscopy is useful for identifying only those patients with clinically evident disease at second look. In those patients with macronodular, confined disease, secondary cytoreduction can be accomplished, either through the laparoscope or via laparotomy. Patients with only miliary disease do not benefit from additional surgical intervention (Figure 10.1). Management of the patient with a clinically negative laparoscopy remains controversial. Some argue that these patients mandate immediate laparotomy because of the reported false-negative rate associated with laparoscopy. Most authors, however, endorse laparoscopy as a reliable method of detection given its low morbidity and reasonable specificity when combining clinical and pathologic evaluation. The literature has established that laparoscopy is generally capable of achieving the goals of second-look surgical assessment while avoiding the morbidity, patient discomfort and extended hospitalization in the majority of patients.30

PREDICTORS OF RESIDUAL DISEASE The likelihood of residual disease at second-look surgery correlates directly with advanced stage of

Figure 10.1 Miliary disease seen at second-look laparoscopy

Ch 10 final.qxd

12/18/2006

15:30

Page 297

SECOND-LOOK SURGERY

disease and residual disease after primary cytoreduction (Table 10.3). Some authors have suggested that these are indirect markers of initial tumor burden and therefore reinforce the need for aggressive initial cytoreductive surgical effort.14,31,40,41 Advanced age and a slow normalization of CA-125 also correlate with the probability of positive second-look surgery.31,44 Neither tumor grade nor histology has been consistently related to findings at second look. Survival after second-look surgery is also directly related to histopathologic variables that reflect tumor burden. Advanced stage disease, large tumor metastases, and residual disease at both primary and

Table 10.3 Covariates affecting the outcome of second-look laparotomy. From reference 31, with permission Covariate Age (years) ≤ 62 > 62

Negative 32 11

Positive 16 19

second-look surgery are inversely correlated with patient survival.7,16,31,40,41,45 Several studies have found that secondary cytoreduction of macroscopic disease to microscopic disease at second-look surgery can significantly improve patient survival, thereby supporting the contention that second-look surgery can have therapeutic value.7,16,41,45 Multivariate analyses suggest that stage of disease and the presence of residual disease at second-look surgery are more predictive of patient survival than tumor grade, age, rate of normalization of CA-125 or elevated CA125.31,41,44 Five-year survival for patients with any macroscopic disease at second look range from 10 to 30%, while the median survival ranges from 10 to 20 months. In stark contrast, patients with negative second-look surgery enjoy 5- and 10-year survival rates of 50–80% (Figure 10.2, Table 10.4).

χ2 p = 0.01

Stage IIIC IV

38 5

30 5

p = 0.727

Grade 1 or 2 3

8 35

13 22

p = 0.66

Implants ≤ 65 > 65

30 13

13 22

p = 0.004

Ascites ≤ 1000 ml > 1000 ml

24 19

12 23

p = 0.058

Largest metastasis ≤ 10 cm > 10 cm

19 24

13 22

p = 0.529

Histology endometriod all others

5 38

1 34

p = 0.148

Taxol no yes

22 21

21 14

p = 0.435

PREDICTORS OF OUTCOME AFTER NEGATIVE SECOND-LOOK SURGERY Approximately 40% of patients with a surgically documented complete response to chemotherapy are alive 8–10 years after second-look surgery (Table 10.4). These long-term data confirm that a negative second-look evaluation is one of the most favorable prognostic markers for ovarian cancer patients.43,46 A negative second look, however, does not imply cure. Despite improved pathologic response rates following primary surgery and chemotherapy, 20–50% of patients will ultimately have disease recurrence. The median disease-free intervals reported for patients with a negative second-look procedure ranges from 14 to 45 months, although most reported disease-free intervals are < 20 months. Studies have established that tumor stage, grade and residual disease at primary surgery predict disease recurrence following a negative second look (Table 10.5, Figure 10.3) Tumor histology, number of biopsies at second look, and extended number of chemotherapy cycles do not significantly affect disease recurrence. Using these risk factors, patients can be

297

Ch 10 final.qxd

12/18/2006

15:30

Page 298

SURGERY FOR OVARIAN CANCER

1.25 Negative SLL

Cumulative survival probability

1.00

Positive SLL

0.75

0.50 n = 78 0.25 n = 30 p < 0.0001 0 0

10

20

30

40

50

60

70

80

90

100

Months from primary surgery

Figure 10.2 Overall survival by findings at second-look surgery (SLL). From reference 44

Table 10.4 Outcome of patients with negative second look Disease-free interval Median follow-up Median 5(months) (months) year

8year

Overall survival

10- Median 5year (months) year

Author

Year Stage

n

Recurrences

Cain40

1986 I–IV

104

17 (16%)

18.6

86%

Gershenson33

1985 III, IV

85

20 (24%)

18.5

85%

Friedman31

1997 III, IV

43

12 (28%)

28

15.3

70%

Podratz16

1988 III, IV

50

15 (30%)

41

58%*

Bolis42

50%

43% 39%

1996 III, IV

140

61(44%)

37

Chiara43

1995 II–IV

129†

63 (49%)

118

34

45%

Rubin46

1999 I–IV

91

54%

113

45

44%

Husain30

2001 II–IV

69

Cass44

2002 II–IV

78

* Survival estimates at 3.5 years; 31 months

298

28 (41%)

†includes

31

16

76

39

8year

10year

82%

47–85*

81 40%

66%

51%

54%

45%

61%

40%

> 31‡ 32%

17 patients with microscopic residual disease;

70% > 31‡ median

survival not yet reached at

Ch 10 final.qxd

12/18/2006

15:30

Page 299

SECOND-LOOK SURGERY

stratified into categories that may guide the clinician in the choice and timing of additional therapy.30,33,42,43,46 Patients at high risk for recurrence may benefit from a more aggressive approach, or lower clinical threshold to commence therapy, while lowrisk patients may be better candidates for observation or clinical trials. Some authors have proposed that one logical approach to prolong disease-free and overall survival is the specific use of consolidation therapy in patients who have demonstrated a prior complete, pathologic response. Second-look surgery can identify that subset of patients who are optimal candidates for consolidation therapy based upon prior tumor responsiveness.47 Although several studies did not find that the use of consolidation therapy prolonged survival, conclusions are limited by the varied consolidation regimens that were used in the past. Most patients received three additional cycles of current therapy, while a change in treatment modality was offered to the minority of patients in these studies.30,42,43

Table 10.5 Adjusted associations with disease recurrence in patients after negative second-look surgery. From reference 46, with permission Predictor Significant Stage* Grade† > 2-cm residual tumor after first surgery Non-significant‡ Age Lymph node sampling Undifferentiated or clear cell histology No. of biopsies No. of prior cycles of platinum-based chemotherapy

Relative risk

95% CI

p Value

2.02 2.00 3.19

1.2, 3.3 1.3, 3.2 1.2, 8.5

0.005 0.004 0.02

1.05 1.73 0.9

0.3, 3.4 0.9, 3.4 0.4, 1.7

0.95 0.11 0.65

1.31 1.24

0.7, 2.5 0.6, 2.5

0.42 0.55

*Odds ratio for each increase in stage or grade; †calculated for stage III disease; ‡power of data to detect difference approximately 25%

Recent single institution studies have reported their success in consolidation therapy using intraperitoneal chemotherapy following negative second-look procedures. Eighty-nine patients were treated with consolidation intraperitoneal cisplatin-based therapy, with a median overall survival of 8.7 years.34 Tournigand et al. reported similar results among 68 patients with pathologic complete response to cisplatin-based primary combination therapy subsequently treated with intraperitoneal cisplatin, etoposide and mitoxantrone. The median overall survival was 73 months, with 58% 5-year survival.48 Prospective clinical trials are clearly needed to confirm these encouraging results, and to gain better definition of the ideal consolidation regimen.

MANAGEMENT OF PATIENTS WITH POSITIVE SECOND LOOK Of patients found to have residual disease at secondlook surgery, the majority will have small-volume disease (Table 10.6). Only 20–45% of patients will have persistent macroscopic disease after primary therapy and be candidates for secondary cytoreduction at second-look surgery. While cytoreduction is feasible in most cases, only those patients who are cytoreduced to microscopic disease have demonstrated significantly improved survival in the available retrospective studies.7,16,41,45 Patients left with grossly visible residual disease after secondary cytoreduction have a modest improvement in survival, which is not statistically significant in most studies. Median survival for patients debulked to ≤ 5 mm at second-look surgery is approximately 18 months compared to 10 months for patients with residual disease of > 1 cm.7,14,41 Studies have consistently reported that patient survival is comparable for those patients found to have microscopic residual disease and those rendered microscopic after secondary cytoreduction at secondlook surgery. Secondary cytoreduction to microscopic residual disease at second-look surgery effectively

299

Ch 10 final.qxd

12/18/2006

15:30

Page 300

SURGERY FOR OVARIAN CANCER

1.00

Proportion surviving

0.75 Residual tumor < 2 cm (n = 19)

0.50

0.25 Residual tumor > 2 cm (n = 38)

0 0

60

120

180

Months after second-look laparotomy with negative findings

Figure 10.3 Recurrence-free survival in patients with stage III disease after negative second-look surgery by the amount of residual tumor present after primary surgery. From reference 46, with permission

Table 10.6 Findings among patients with positive second look (SL) Amount of residual disease > 1 cm

≤ 1 cm

116

39

51

III, IV

49

27*

22*

1989

I–IV

67

20*

28*

Childers20

1995

II–IV

24

5

Friedman31

1997

III, IV

35

20

10 (66%)

Williams7

1997

III, IV

153

69

29

36 (29%)

Husain30

2001

II–IV

81

15

Cass44

2002

II–IV

30

2

Author

Year

Stage

n

Podratz16

1988

I–IV

Lippman41

1988

Hoskins45

* Residual reported > 2 cm or ≤ 2 cm

300

55

Micro

No. of patients debulked at SL to micro < 5 mm

26

62 (69%)

17

16 (24%)

11 (73%) 16

12

5 (27%)

Ch 10 final.qxd

12/18/2006

15:30

Page 301

SECOND-LOOK SURGERY

doubles the median survival of patients compared to those left with macroscopic disease after second-look surgery, with a projected 5-year survival of 50% versus 10–19%, respectively7,16,45 (Figure 10.4). One frequently cited prospective study by Luesley et al. concluded that there was no survival benefit to second-look surgery.49 The authors randomized 169 patients to (a) second-look surgery followed by 12 courses of chlorambucil; or (b) second-look surgery followed by abdominal–pelvic radiation; or (c) no second look and 12 courses of chlorambucil. Unfortunately, the study was underpowered to examine differences in the mixed treatment regimens because of patient attrition. Moreover, there was a significant improvement in 2-year survival rates among those patients debulked to < 2 cm residual at secondlook surgery compared to those with inoperable residual disease of > 2 cm (p = 0.002). Although the authors concluded that there was no survival benefit to second-look surgery, their own data showed a survival benefit for that subset of patients who were successfully cytoreduced to microscopic residual disease at second look.47,49

The intuitive explanation for the improved survival of patients with microscopic as opposed to macroscopic disease is enhanced tumor kinetics. While this tenet has been confirmed in the setting of primary surgical cytoreduction, there are only retrospective data to support the same approach to cytoreduction at second-look surgery at the current time. The earlier identification of patients with residual disease, prior to the development of macroscopic, symptomatic disease, should improve the efficacy of chemotherapy. The lack of demonstrated survival benefit from second-look surgery to date might result from poor study design and minimally effective salvage regimens. It is noteworthy that several small studies that used ongoing reassessment surgery and mixed salvage regimens have documented a subsequent pathologic response and prolonged survival in 7–30% of patients with residual disease at second look.24,26,40 The amount of residual disease at second-look surgery must dictate the appropriate salvage therapy. Lessons learned from cumulative data on intraperitoneal chemotherapy as a second-line therapy

100

Per cent alive

80

60 Disease-free (77) 40

Microscopic (24) Macroscopic ≤ 5 mm (20)

20 Macroscopic > 5 mm (14) 0 0

1

2

3

4

5

Years after second look

Figure 10.4 Survival from second-look laparotomy to death from disease, according to residual tumor size at completion of reexploration. From reference 14, with permission

301

Ch 10 final.qxd

12/18/2006

15:30

Page 302

SURGERY FOR OVARIAN CANCER

illustrates this point. The ideal candidate for intraperitoneal therapy is the patient with smallvolume disease with prior chemotherapy-responsive disease. Many retrospective phase II trials of intraperitoneal cisplatin have shown surgical response rates of 40–50% and surgical complete response rates of 25–35% in patients with small-volume residual disease.34,50 Memorial Sloan-Kettering has recently reported the long-term follow-up of > 300 patients treated with intraperitoneal cisplatin-based therapy after positive second-look surgery. In a multivariate analysis, the only predictors of survival were amount of residual disease at the initiation of therapy, and tumor grade. The median survival for patients from the start of therapy by residual disease at second-look surgery was: microscopic, 4.8 years; < 1 cm, 3.3 years; and > 1 cm, 1.2 years.34

priate postoperative treatment based upon surgical findings to improve patient survival. The true value of second-look assessment may be realized as newer, more effective salvage and consolidation therapies are developed. The majority of data suggest that laparoscopy is the most appropriate surgical procedure for secondlook assessment. Modern laparoscopic technique permits the surgeon to lyse adhesions, assess disease status and cytoreduce gross residual disease in the majority of patients, resulting in significantly reduced morbidity, patient discomfort and cost. Second-look laparotomy should be reserved for those patients with inadequate laparoscopic second-look procedures.

REFERENCES 1.

Wangensteen OH, Lewis FJ, Tongen LA. The ‘second-look’ in cancer surgery: a patient with colic cancer and involved lymph nodes negative on the ‘sixth-look’. Lancet 1951; 71: 303–7

2.

Gilbertsen VA, Wangensteen OH. A summary of thirteen years’ experience with the second look program. Surg Gynecol Obstet 1962; 114: 438–42

3.

Rutledge F, Burns BC. Chemotherapy for advanced ovarian cancer. Am J Obstet Gynecol 1966; 96: 761–72

4.

Smith WG, Day TG Jr, Smith JP. The use of laparoscopy to determine the results of chemotherapy for ovarian cancer. J Reprod Med 1977; 18: 257–60

5.

McGuire WP, Hoskins WJ, Brady MF, et al. Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 1996; 334: 1–6

6.

Ozols RF, Bundy BN, Greer BE, et al. Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol 2003; 21: 3194–200

7.

Williams L, Brunetto VL, Yordan E, et al. Secondary cytoreductive surgery at second-look laparotomy in advanced ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1997; 66: 171–8

CONCLUSION To date, second-look surgery remains the best way to establish disease status after primary therapy of advanced-stage ovarian cancer. In addition to its prognostic value, there is clear benefit to some patients with secondary cytoreductive surgery at second-look, especially in the initially suboptimally cytoreduced or incompletely staged patient. The data clearly show that patients in clinical remission following primary therapy for ovarian cancer are an inherently heterogeneous group. Secondlook surgery is the most effective way to stratify these patients by their risk of recurrence, and to identify the optimal candidates for clinical trials of novel consolidation and salvage therapies. Proponents of secondlook assessment argue that the clinical significance of future clinical trials of consolidation and second-line agents will be diluted without some accurate selection process by status of residual disease to identify candidates for therapy. Arguably, any future clinical trial that involves second-look assessment following primary therapy in ovarian cancer should include appro-

302

Ch 10 final.qxd

12/18/2006

15:30

Page 303

SECOND-LOOK SURGERY

8.

9.

Brenner DE, Shaff MI, Jones HW, et al. Abdominopelvic computed tomography: evaluation in patients undergoing second-look laparotomy for ovarian carcinoma. Obstet Gynecol 1985; 65: 715–19 Berek JS, Knapp RC, Malkasian GD, et al. CA 125 serum levels correlated with second-look operations among ovarian cancer patients. Obstet Gynecol 1986; 67: 685–9

10.

Warburg O. On the origin of cancer cells. Science 1956; 123: 309–14

11.

Rose PG, Faulhaber P, Miraldi F, Abdul-Karim FW. Positive emission tomography for evaluating a complete clinical response in patients with ovarian or peritoneal carcinoma: correlation with second-look laparotomy. Gynecol Oncol 2001; 82: 17–21

12.

13.

14.

15.

Karlan BY, Hawkins R, Hoh C, et al. Whole-body positron emission tomography with 2-[18F]-fluoro-2deoxy-D-glucose can detect recurrent ovarian carcinoma. Gynecol Oncol 1993; 51: 175–81 Bristow RE, del Carmen MG, Pannu HK, et al. Clinically occult recurrent ovarian cancer: patient selection for secondary cytoreductive surgery using combined PET/CT. Gynecol Oncol 2003; 90: 519–28 Podratz KC, Malkasian GD Jr, Hilton JF, et al. Second-look laparotomy in ovarian cancer: evaluation of pathologic variables. Am J Obstet Gynecol 1985; 152: 230–8 Creasman WT. Second-look laparotomy in ovarian cancer. Gynecol Oncol 1994; 55: S122–7

16.

Podratz KC, Schray MF, Wieand HS, et al. Evaluation of treatment and survival after positive second-look laparotomy. Gynecol Oncol 1988; 31: 9–24

17.

Walton L, Ellenberg SS, Major F Jr, et al. Results of second-look laparotomy in patients with early-stage ovarian carcinoma. Obstet Gynecol 1987; 70: 770–3

18.

Clough KB, Ladonne JM, Nos C, et al. Second look for ovarian cancer: laparoscopy or laparotomy? A prospective comparative study. Gynecol Oncol 1999; 72: 411–17

19.

Moore JL Jr, Barter JF. Second-look laparotomy. Semin Surg Oncol 1994; 10: 276–82

20.

Childers JM, Lang J, Surwit EA, Hatch KD. Laparoscopic surgical staging of ovarian cancer. Gynecol Oncol 1995; 59: 25–33

21.

Goldberg GL, Scheiner J, Friedman A, et al. Lymph node sampling in patients with epithelial ovarian carcinoma. Gynecol Oncol 1992; 47: 143–5

22.

Rubin SC, Hoskins WJ, Hakes TB, et al. Recurrence after negative second-look laparotomy for ovarian cancer: analysis of risk factors. Am J Obstet Gynecol 1988; 159: 1094–8

23.

Rubin SC, Hoskins WJ, Saigo PE, et al. Prognostic factors for recurrence following negative second-look laparotomy in ovarian cancer patients treated with platinum-based chemotherapy. Gynecol Oncol 1991; 42: 137–41

24.

Quinn MA, Bishop GJ, Campbell JJ, et al. Laparoscopic follow-up of patients with ovarian carcinoma. Br J Obstet Gynaecol 1980; 87: 1132–9

25.

Piver MS, Lele SB, Barlow JJ, Gamarra M. Secondlook laparoscopy prior to proposed second-look laparotomy. Obstet Gynecol 1980; 55: 571–3

26.

Berek JS, Griffiths CT, Leventhal JM. Laparoscopy for second-look evaluation in ovarian cancer. Obstet Gynecol 1981; 58: 192–8

27.

Bagley CM Jr, Young RC, Schein PS, et al. Ovarian carcinoma metastatic to the diaphragm – frequently undiagnosed at laparotomy. A preliminary report. Am J Obstet Gynecol 1973; 116: 397–400

28.

Childers JM, Brzechffa PR, Surwit EA. Laparoscopy using the left upper quadrant as the primary trocar site. Gynecol Oncol 1993; 50: 221–5

29.

Abu-Rustum NR, Barakat RR, Siegel PL, et al. Second-look operation for epithelial ovarian cancer: laparoscopy or laparotomy? Obstet Gynecol 1996; 88: 549–53

30.

Husain A, Chi DS, Prasad M, et al. The role of laparoscopy in second-look evaluations for ovarian cancer. Gynecol Oncol 2001; 80: 44–7

31.

Friedman RL, Eisenkop SM, Wang HJ. Second-look laparotomy for ovarian cancer provides reliable prognostic information and improves survival. Gynecol Oncol 1997; 67: 88–94

32.

Casey AC, Farias-Eisner R, Pisani AL, et al. What is the role of reassessment laparoscopy in the management of gynecologic cancers in 1995? Gynecol Oncol 1996; 60: 454–61

33.

Gershenson DM, Copeland LJ, Wharton JT, et al. Prognosis of surgically determined complete respon-

303

Ch 10 final.qxd

12/18/2006

15:30

Page 304

SURGERY FOR OVARIAN CANCER

ders in advanced ovarian cancer. Cancer 1985; 55: 1129–35 34.

Barakat RR, Sabbatini P, Bhaskaran D, et al. Intraperitoneal chemotherapy for ovarian carcinoma: results of long-term follow-up. J Clin Oncol 2002; 20: 694–8

35.

Morrow CP, Curtin JP. Minimal access surgery. In Gynecologic Cancer Surgery. New York: Churchill Livingstone, 1996: 745–67

36.

37.

38.

Abu-Rustum NR, Rhee EH, Chi DS, et al. Subcutaneous tumor implantation after laparoscopic procedures in women with malignant disease. Obstet Gynecol 2004; 103: 480–7 Lele SB, Piver MS. Interval laparoscopy as predictor of response to chemotherapy in ovarian carcinoma. Obstet Gynecol 1986; 68: 345–7 Nicoletto MO, Tumolo S, Talamini R, et al. Surgical second look in ovarian cancer: a randomized study in patients with laparoscopic complete remission – a Northeastern Oncology Cooperative Group-Ovarian Cancer Cooperative Group study. J Clin Oncol 1997; 15: 994–9

39.

Smith EM, Sowers MF, Burns TL. Effects of smoking on the development of female reproductive cancers. J Natl Cancer Inst 1984; 73: 371–6

40.

Cain JM, Saigo PE, Pierce VK, et al. A review of second-look laparotomy for ovarian cancer. Gynecol Oncol 1986; 23: 14–25

41.

42.

304

Lippman SM, Alberts DS, Slymen DJ, et al. Secondlook laparotomy in epithelial ovarian carcinoma. Prognostic factors associated with survival duration. Cancer 1988; 61: 2571–7 Bolis G, Villa A, Ferraris C, et al. Survival of advanced ovarian cancer patients with microscopic

partial response after surgery and first-line chemotherapy. Eur J Cancer 1995; 31A: 1019–20 43.

Chiara S, Lionetto R, Campora E, et al. Long-term prognosis following macroscopic complete response at second-look laparotomy in advanced ovarian cancer patients treated with platinum-based chemotherapy. The Gruppo Oncologico Nord Ovest. Eur J Cancer 1995; 31A: 296–301

44.

Cass I, Madden A, Berek J, et al. Second look surgery as a prognostic tool for optimally cytoreduced ovarian carcinoma patients. Gynecol Oncol 2002; 84: 495

45.

Hoskins WJ, Rubin SC, Dulaney E, et al. Influence of secondary cytoreduction at the time of second-look laparotomy on the survival of patients with epithelial ovarian carcinoma. Gynecol Oncol 1989; 34: 365–71

46.

Rubin SC, Randall TC, Armstrong KA, et al. Tenyear follow-up of ovarian cancer patients after second-look laparotomy with negative findings. Obstet Gynecol 1999; 93: 21–4

47.

Podratz KC, Cliby WA. Second-look surgery in the management of epithelial ovarian carcinoma. Gynecol Oncol 1994; 55: S128–33

48.

Tournigand C, Louvet C, Molitor JL, et al. Long-term survival with consolidation intraperitoneal chemotherapy for patients with advanced ovarian cancer with pathological complete remission. Gynecol Oncol 2003; 91: 341–5

49.

Luesley D, Lawton F, Blackledge G, et al. Failure of second-look laparotomy to influence survival in epithelial ovarian cancer. Lancet 1988; 2: 599–603

50.

Markman M. Intraperitoneal therapy of ovarian cancer. Semin Oncol 1998; 25: 356–60

Ch 11 final.qxd

12/18/2006

15:31

Page 305

CHAPTER 11

Secondary cytoreductive surgery Robert L Coleman, Adnan Munkarah, Robert E Bristow

INTRODUCTION The concept of cytoreductive surgery for epithelial ovarian cancer has evolved since 1935 when Meigs first suggested that as much tumor as possible should be removed to enhance the effects of postoperative irradiation.1 Forty years after Meigs’ initial proposition, Griffiths published a landmark study that provided the first conclusive evidence of an inverse relationship between residual tumor size and patient survival.2 Perhaps the most convincing data for both the benefits and shortcomings of primary cytoreductive surgery is a detailed analysis of 637 patients with advanced ovarian cancer reported by Hoskins et al. for the Gynecologic Oncology Group (GOG) (protocols 52 and 97).3,4 In evaluating patients with both optimal and suboptimal residual disease status, these investigators demonstrated three distinct groups: microscopic residual disease (no gross residual), visible residual disease ≤ 2 cm, and residual disease of > 2 cm. The salient observations of these studies were that survival is inversely related to the volume of residual disease after primary surgery and that primary cytoreduction failed to have a meaningful effect on survival if the largest residual tumor dimension exceeded 2 cm, regardless of the extent of resection. Because the majority of patients with advancedstage ovarian cancer will experience a recurrence of their disease, the therapeutic value of repeating the initial surgical treatment plan (secondary tumor cytoreduction) has been widely debated. Theoretically, cytoreductive surgery can augment the response to subsequent chemotherapy by reducing cellular

kinetic and pharmacologic barriers to maximal tumor cytotoxicity. This rationale has been proposed as the basis for the efficacy of surgical cytoreduction at the time of initial diagnosis as well as for the subgroup of patients with chemotherapy-sensitive recurrent ovarian cancer. The ‘fractional cell kill hypothesis’ states that the efficiency with which chemotherapeutic agents are able to eliminate malignant cells is dependent on both the dose of the agent and the number of cells present.5 Each exposure to a particular dose of drug will kill a constant fraction of cells, such that multiple cycles of chemotherapy are necessary for complete eradication of a tumor. Cytoreductive surgery can achieve a dramatic reduction in tumor volume but does not completely eliminate all viable tumor cells. The Goldie–Coldman model of spontaneous tumor cell mutation suggests that the probability of developing a chemotherapy-resistant phenotype is related to the number of cell divisions that has occurred.6 The larger the number of cells present or the longer the delay in starting chemotherapy, the greater the chance that a resistant clone of cells has already emerged. Cytoreductive surgery can remove large tumor masses that potentially harbor drugresistant cells. The Norton–Simon model of Gompertzian tumor growth is a hypothetical model that holds that, as tumor volume decreases, the growth fraction (the proportion of tumor cells that doubles during a given increment of time) subsequently increases.7 By reducing tumor volume and recruiting previously dormant tumor cells into the active phase of the cell cycle, cytoreductive surgery may enhance chemotherapeutic cytotoxicity. Finally,

305

Ch 11 final.qxd

12/18/2006

15:31

Page 306

SURGERY FOR OVARIAN CANCER

by removing poorly perfused large tumor masses, debulking surgery is thought to enhance response rates to chemotherapy by leaving a population of cells with improved vascularization. Although similar in theoretical basis, the clinical benefits of secondary cytoreductive surgery for ovarian cancer have not been as clearly established as those for primary debulking procedures. Many studies include patients treated over long time intervals, during which concepts about adjuvant therapy were still evolving, with a resultant lack of uniformity in the therapeutic regimens employed prior to and/or after secondary surgery. The definition of the optimal volume of residual disease has undergone a similar evolution over time. Various early reports of secondary cytoreduction included heterogeneous patient populations with inconsistent patient inclusion criteria, making meaningful comparison of data between studies difficult. In addition, the majority of studies have been retrospective reviews, such that selection bias has been a common and potentially significant confounding variable. To justify the use of secondary surgery for recurrent ovarian cancer, several clinical questions must be addressed: (1) what is the potential survival benefit of successful secondary cytoreduction and what constitutes an optimal surgical resection? (2) what is the feasibility of a successful resection? (3) what is the associated risk of morbidity and mortality? and (4) are there clearly defined selection criteria by which to identify patients suitable for this approach? In addition, the clinician must have a working knowledge of the advantages and limitations of currently available surveillance techniques for detecting recurrent disease.

TERMINOLOGY Before engaging the above questions, however, it is worthwhile to review the common terminology used to categorize the clinical circumstances and subsequent outcomes reflected in the literature on sec-

306

ondary cytoreductive surgery for ovarian cancer. To separate patients on the basis of differences in biologic tumor behavior, secondary operations can be grouped into four distinct clinical scenarios: (1) Progressive disease. Patients who have evidence of clinical disease progression while receiving front-line chemotherapy. (2) Interval debulking. Patients with bulky, unresectable tumor discovered at initial surgery, who then undergo an abbreviated course of initial chemotherapy prior to a repeat (or interval) attempt at cytoreductive surgery. (3) Second-look surgery. Patients who are clinically and radiographically free of disease after primary surgery and front-line chemotherapy, who are found to have macroscopic disease at secondlook surgery. (4) Recurrent disease. Patients who enjoy a prolonged clinical disease-free interval (longer than 6–12 months) after completing primary therapy and then develop recurrent disease. This chapter will focus on the latter category, those patients with recurrent ovarian cancer manifesting after a sustained (≥ 6 months) disease-free interval following primary therapy. Patients receiving contemporary platinum-based initial therapy can also be described according to the platinum ‘sensitivity’ of their disease. Patients with clinical recurrence detected within 6 months of completing primary chemotherapy are typically categorized as having persistent, platinum-resistant disease, while patients who experience a disease-free interval of 6 months or longer are generally considered to have recurrent, platinum-sensitive disease.8 As with primary cytoreductive surgery, the surgical result of secondary operations for ovarian cancer has been variably classified according to the volume or size of residual disease. An optimal resection is typically taken to represent residual tumor with a maximal diameter no larger than 1.0–2.0 cm. More recently, patients undergoing optimal cytoreduction have

Ch 11 final.qxd

12/18/2006

15:31

Page 307

SECONDARY CYTOREDUCTIVE SURGERY

been separated further to distinguish between those in whom all visible disease is successfully extirpated (complete cytoreduction) and those left with optimal but visible disease (0.1 cm up to 1.0 cm or 2.0 cm). The corresponding clinical outcomes associated with these criteria are discussed later in this chapter.

SURVEILLANCE FOR RECURRENT DISEASE The most accurate and cost-effective surveillance program for patients with ovarian cancer in complete clinical or pathologic remission has yet to be determined. Clinical examination, serum biomarkers (e.g. CA-125), radiographic survey (ultrasound, magnetic resonance imaging (MRI) and computed tomography (CT)), and radiological measures of metabolic function (positron emission tomography (PET) and the related PET/CT fusion scan) are employed to some degree and have increased the precision with which suspected recurrent disease can be evaluated. Physical examination and assessment of serum biomarkers are generally performed at regular intervals (e.g. every 3 months), with adjunctive studies performed primarily for evaluation of abnormalities on examination or elevation of tumor markers. Alternatively, some clinicians recommend serial radiographic imaging at 6–12month intervals, particularly in those patients in whom serum biomarkers have been unreliable indicators of the presence of disease (e.g. low-level abnormal value at initial diagnosis or positive second-look surgery).

Serum biomarkers The serum CA-125 level is the most commonly used measure for ovarian cancer surveillance, although other biomarkers such as plasma YKL-40 and MUC-1 are being evaluated alongside CA-125 in an effort to increase diagnostic precision.9,10 In many cases where an individual’s CA-125 level was elevated at diagnosis and returned to normal following therapy, the biomarker correspondingly elevates at the time of recurrence. The clinical significance of a rising CA-125

level in this setting has been confirmed to reflect disease recurrence, with several authors reporting a sensitivity and specificity of 77% and 94%, respectively.11,12 When a rising CA-125 level is accompanied by the appearance of new disease-related symptoms or radiographic changes, the diagnosis of recurrent ovarian cancer is relatively straightforward. Not infrequently, however, an elevation of the serum CA125 level may be the only clinical manifestation suggestive of recurrent ovarian cancer. The precise definition of what constitutes a ‘significant’ change in CA125 level among asymptomatic patients has been debated, as well as the classification of patients with an apparent serologic recurrence in the absence of measurable disease. Rustin et al. retrospectively evaluated 131 patients entered on the North Thames Ovary Trial of five versus eight cycles of platinum chemotherapy for stages IC to IV ovarian cancer and proposed a model where two consecutive CA-125 values double the upper cut-off value of 30 U/ml (i.e. two serial samples of ≥ 60 U/ml) provided the most diagnostic precision in predicting recurrent disease.13 Using this criterion, the diagnostic model yielded a sensitivity of 84%, a specificity of 98% and a positive predictive value (PPV) of 99%. In addition, a marker change of two consecutive CA-125 values of ≥ 60 U/ml preceded conventional objective measures of recurrence by a median of 63 days (‘lead time’). Among ovarian cancer patients achieving a complete clinical response to initial therapy, the implications of a rising CA-125 level that remains within the accepted normal range (< 35 U/ml) have not been widely studied, although limited data indicate that this may herald disease recurrence in a significant number of patients. Wilder et al. recently reported a multi-institutional study of 11 patients with a progressive increase in at least three consecutive CA-125 values above the coefficient of variation (3%) for the assay that remained < 35 U/ml.14 Of the 11 patients identified, all developed recurrent ovarian cancer, six with documented progression on imaging studies and five by histologic confirmation. The average time interval between the third early rising serum CA-125

307

Ch 11 final.qxd

12/18/2006

15:31

Page 308

SURGERY FOR OVARIAN CANCER

level to clinical or radiographic confirmation of recurrence was 189 days (range 84–518 days). The above data suggest that a progressive increase in the CA-125 level, even if remaining within the normal range, is associated with a high risk of tumor recurrence, and that further investigation may be warranted. In theory, earlier detection of ovarian cancer recurrence would allow prompt treatment intervention at a time when tumor volume is relatively low, thereby increasing the duration of survival.

Radiographic imaging studies Radiographic imaging of patients with suspected recurrent ovarian cancer is critical for documentation of disease extent, patient counseling and estimating the likelihood of resectability among surgical candidates. The most frequently used imaging modality currently is the CT scan, which is often obtained in response to new symptomatology and/or an elevation in serum CA-125 trends. In the setting of primary disease, conventional CT imaging has been quite useful in establishing a cancer diagnosis, in correlating surgical findings and as a potential model to predict surgical outcome. In a recent publication, Bristow et al. developed a predictive index score from 41 stage III and IV ovarian cancer patients undergoing primary surgical cytoreduction.15 Findings from CT imaging were numerically weighted according to the strength of statistical association. For definition purposes, an optimal cytoreduction was considered less than 1 cm residual disease. Fourteen parameters (13 radiographic characteristics and performance status) were incorporated into a predictive model. Radiographic characteristics that were associated with a lower likelihood of optimal resection included involvement of the bowel mesentery, retroperitoneal adenopathy above the renal vessels and diffuse peritoneal thickening. Among the 20 (49%) patients achieving an optimal cytoreduction, 17 had a predictive index score of ≥ 4, yielding a specificity of 85%. The positive and negative predictive values (NPV) for an optimal resection were 88% and 100%, respectively, highlighting the potential clinical utility of such a model for surgical

308

planning. In the setting of recurrent disease, only limited data regarding the use of CT imaging to predict resectability are available. Several studies have used CT imaging with diagnostic resolution of 1–2 cm to correlate radiographic findings with disease at secondlook operations.16–18 Reported sensitivities and specificities range from 59 to 83% and 83 to 88%, respectively. The major limitations of CT imaging for detecting recurrent ovarian cancer are similar to those observed at primary diagnosis, namely, difficulty in detecting small implants on the peritoneal surfaces and disease in the bowel mesentery. In addition, changes associated with previous surgery (such as adhesions, lymphocysts, abscess, hematoma) further complicate the detection of recurrent ovarian cancer. Whether newer and higher resolution imagery from spiral CT equipment will improve these important limitations is still to be determined. MRI offers enhanced resolution of soft tissues and has been suggested as an alternative to CT for identification of recurrent ovarian cancer. Ricke et al. correlated MRI findings in 39 patients with suspected recurrent disease who subsequently underwent surgical evaluation.19 Six specific anatomic sites (upper abdomen, bowel, lower pelvis, abdominal wall, lymph nodes and diffuse peritoneal carcinomatosis) were evaluated radiographically and compared against the surgical findings. MRI demonstrated an overall sensitivity of 67–83%, specificity of 60–89%, PPV of 65–93% and NPV of 47–83% for recurrent disease. The diagnostic accuracy for the individual imaging sites is shown in Table 11.1. While the PPV was acceptable for most disease sites, the NPV was < 50% for tumor involving the bowel surface and retroperitoneal lymph nodes, a limitation similar to that of CT imaging. Several clinical trials have prospectively compared MRI to CT imaging in patients with both primary and recurrent ovarian cancer and demonstrated similar diagnostic accuracy (MRI, 88%; CT, 91%) as well as a high degree of interobserver correlation.16,20 Buist et al. concluded that little additional information was gained by the addition of the alternative study.20

Ch 11 final.qxd

12/18/2006

15:31

Page 309

SECONDARY CYTOREDUCTIVE SURGERY

Table 11.1 Diagnostic accuracy of magnetic resonance imaging for recurrent ovarian cancer according to anatomic imaging sites (from reference 19) Site

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Upper abdomen

67

89

88

70

Bowel

72

70

88

47

Lower pelvis

73

83

73

83

Abdominal wall

83

60

77

69

Lymph nodes

67

86

93

46

Diffuse peritoneal carcinomatosis

69

74

65

77

PPV, positive predictive value; NPV, negative predictive value

Radiographic measures of metabolic function PET imaging displays the differential metabolic activity within various tissues to distinguish between physiologic and malignant processes. Its utility has been extensively investigated in functional studies of the brain and heart. In the clinical practice of oncology, the use of the PET scan with the radiopharmaceutical, 2-[18F]fluoro-2-deoxy-D-glucose (FDG) has been a valuable asset in treatment planning. Limited experience has been reported among ovarian cancer patients – primarily in the evaluation of clinically occult disease. Rose et al. reported the results of FDG–PET in 22 advanced ovarian and peritoneal cancer patients who had a normal physical examination and CA-125, and conventional imaging, and who consented to a second-look operation following primary chemotherapy.21 Persistent disease was identified in 13 (59%) patients. Only one of nine sites with macroscopic and none of four sites with microscopic disease were accurately predicted. The sensitivity was 10% and the specificity 42%, causing the authors to conclude that FDG–PET has limited utility as a substitute for second-look surgery. However, all falsenegative studies had implants between 3 and 6 mm in size and it has been suggested that, in clinical treatment planning, particularly for patients being considered for secondary cytoreduction, this level of resolution may, in fact, be advantageous.22 With an accurate

depiction of metabolically active implants greater than 1 cm, it is anticipated that this diagnostic modality will improve the precision of selecting the appropriate patients for salvage surgery. Comparing CT, MRI and FDG–PET, Kubik-Huch et al. examined 11 patients for recurrence, which was confirmed histologically.23 Because of the small sample size, no significant difference between techniques could be detected in the accuracy of identifying disease, despite a range in sensitivity from 43% for CT scan to 90% for FDG–PET. However, only the FDG–PET study identified a solitary transverse colon implant in one patient, again highlighting its potential utility in identifying surgical candidates. Several larger studies have confirmed the potential utility of FDG–PET for detecting recurrent ovarian cancer. Zimny et al. reported results of 106 FDG–PET scans in 54 patients undergoing surveillance for ovarian cancer following completion of primary therapy.24 Fifty-eight scans were performed in patients with suspected recurrence based on physical examination findings, an abnormal anatomic imaging study, or a rising CA-125 level. In these cases, the sensitivity of FDG–PET for ovarian cancer recurrence was 94% and the PPV was 98%. The sensitivity was 96% when the suspicion of recurrence was based solely on an abnormal CA-125 level. Similarly, Chang et al. studied the utility of FDG–PET imaging in 28

309

Ch 11 final.qxd

12/18/2006

15:31

Page 310

SURGERY FOR OVARIAN CANCER

patients with asymptomatic serum CA-125 elevation (> 35 U/ml) following a complete clinical response to primary therapy.25 All patients had negative or equivocal conventional imaging studies. In this setting, the overall accuracy of FDG–PET was 93%. Nineteen of 20 patients with confirmed recurrence were correctly identified by PET imaging, yielding a sensitivity of 95%. PET/CT fusion imaging is an emerging technology that combines the important physiologic and metabolic information of FDG–PET with the anatomic precision of CT imaging.26 These novel scanners acquire PET and CT images that are cotemporaneous and co-registered to localize elevated FDG uptake with improved anatomic specificity (Figure 11.1). The PET images are corrected for attenuation using coefficients obtained by scaling the CT numbers from the CT images to the PET energy (511 keV). The helical CT scan is reconstructed into images with a slice thickness of 3.4 mm to match the PET scan. Makhija et al. reported a limited retrospective experience using this technology in eight ovarian and fallopian tube cancer patients undergoing secondary surgical evaluation (all with disease).27 There were five positive PET/CT scans, all in patients with negative CT findings and in two with negative CT and CA-125 findings. There were three patients with false-negative studies: one who had a suspicious CT scan, one with an elevated CA-125 level (normal CT) and one with all negative clinical parameters. More recently, Bristow et al. studied the utility of combined PET/CT imaging in 22 patients with suspected recurrent ovarian cancer based on a rising serum CA-125 level and negative or equivocal conventional CT findings.28 All patients underwent surgical exploration. In this study, 18 of 22 patients with clinically occult disease by conventional imaging were found to have recurrent ovarian cancer measuring ≥ 1 cm in maximal diameter, of which 72.2% underwent complete secondary cytoreduction to no gross residual disease. The overall patient-based accuracy of combined PET/CT in detecting recurrent disease of ≥ 1 cm was 81.8%, with a sensitivity of 83.3% and PPV of 93.8%. The authors

310

concluded that combined PET/CT may be a useful imaging technique for patients with biochemical evidence of recurrence and negative conventional imaging that can facilitate early surgical intervention, thereby improving the likelihood of a successful secondary surgical resection. Prospective trials are underway to determine the precise role of PET/CT fusion scanning in clinical decision-making.

SURVIVAL OUTCOME AFTER SECONDARY CYTOREDUCTIVE SURGERY An early report in 1983 from Berek et al. forms the basis for many current concepts regarding cytoreductive surgery for recurrent ovarian cancer, despite including a heterogeneous patient population.29 These investigators retrospectively identified a group of 32 patients with a median interval between primary and secondary surgery of 12 months (range 6–48 months). An optimal resection, defined as a maximum diameter of residual tumor of < 1.5 cm, was accomplished in 38% of cases. The median survival time for this group (20 months) was significantly longer than that of patients left with larger disease residuum (5 months, p < 0.01). Survival time was also found to be associated with the interval between initial and secondary operations, with patients having an interval of 12 months or more surviving significantly longer (median 16 months) than patients with an interval of less than 12 months (median 9 months, p < 0.05). In 1989, Morris et al. described their experience at the MD Anderson Cancer Center with 30 patients undergoing secondary cytoreduction for recurrent ovarian cancer after a disease-free interval of at least 6 months.30 However, in this study, the median survival time for patients with residual disease measuring < 2 cm in maximal diameter was 18 months, which was not statistically different from the 13.3 months’ median survival for those left with larger residual tumor. The duration of the disease-free interval between primary therapy and secondary surgery was also found to correlate with survival time,

Ch 11 final.qxd

12/18/2006

15:31

Page 311

SECONDARY CYTOREDUCTIVE SURGERY

b

a

c

Figure 11.1 Positron emission tomography PET/computed tomography (CT) fusion scan for detecting recurrent ovarian cancer. (a) Whole body PET/CT image (coronal section) of a 68-year-old woman with stage IV ovarian cancer. A prior CT scan had been interpreted as demonstrating a solitary liver lesion consistent with a cyst or malignant process. (b) Axial image of the same patient through the level of the kidneys, demonstrating metabolic activity of a left paraortic node. (c) Intraoperative view of the identified paraortic node near the porta hepatis

but this association again did not reach statistical significance. Notably, the response to second-line therapy was only 11%, which mirrors the poor overall outcome of this study population and illustrates that the potential benefit of an optimal secondary resection is not sustained without access to effective second-line chemotherapy.

During the past 15 years, additional studies evaluating the survival impact of secondary cytoreductive surgery for patients with recurrent ovarian cancer have included more homogeneous populations. Given that the majority of contemporary reports describe patients who were initially treated during the era of universal platinum-based combination chemotherapy

311

Ch 11 final.qxd

12/18/2006

15:31

Page 312

SURGERY FOR OVARIAN CANCER

for advanced ovarian cancer, these studies, for the most part, reflect the experience with patients considered to have platinum-sensitive disease. With the exception of two large series, the contemporary body of literature consists of retrospectively designed studies (Table 11.2).29–45 In 1992, a report by Jänicke et al. described 30 patients undergoing secondary cytoreduction after a median disease-free interval of 16.5 months (range 4–110 months).31 Optimal debulking (< 2 cm) was achieved in 87% of patients, and in 47% of cases all macroscopic disease was removed. These authors observed that the most significant survival advantage was obtained with complete secondary cytoreduction. The median survival time was 29 months for those patients cytoreduced to only microscopic residual disease, whereas patients left with even small-volume gross residual disease (< 2 cm) had a median survival time of only 9 months (p = 0.004). While the length of the preceding clinical remission and type of postoperative second-line therapy were also associated

with survival outcome, multivariate analysis revealed that residual tumor after secondary cytoreduction was the most important independent variable. Segna et al., in one of the largest series, reported results of secondary cytoreduction in 100 patients with recurrent or progressive ovarian cancer following primary resection and platinum-based chemotherapy.32 In this study, 73% of patients had a disease-free interval in excess of 12 months. Optimal secondary cytoreduction (< 2 cm) was accomplished in 61% of patients and predicted a significantly longer median survival time (27.1 months) compared to patients left with suboptimal residual disease (9 months, p = 0.0001). After adjusting for other clinical characteristics by multivariate analysis, successful secondary surgery remained a significant predictor of survival. In 1995, Vaccarello et al. from the Memorial Sloan-Kettering Cancer Center reported on 57 patients with ovarian cancer relapse, which occurred at a median of 33 months after documented complete surgical response (negative second-look surgery) to

Table 11.2 Compiled studies of secondary cytoreductive surgery for recurrent ovarian cancer, 1983–2004

n Author

Year

(total)

Berek29

1983

32

Morris30

Janicke31

Segna32

1989

1992

1993

Eisenkop33* 1995

30

30

100

36

Median

Median

disease-free

overall

interval

survival Morbidity resection

(months) (months) 12

42.5**

16

NA

22

Bowel

10

16.3

18

16.6

43

Residual

Median survival

Hazard

(%)

(%)

criteria

n

%

(months)

ratio

34.4

53.1

< 1.5 cm

12

37.5

20

> 1.5 cm

20

62.5

5

< 2 cm

17

57

18

> 2 cm

13

43

13.3

no gross

14

46.7

29

< 2 cm

12

40

9

> 2 cm

4

13.3

3

< 2 cm

61

61

27.1

> 2 cm

39

39

9

no gross

30

83.0

43

gross

6

17.0

5

36.7

23.3

13.0

30.1

43.3

59.4

38.0

47.2

disease

Significance

p < 0.01

p < 0.2

p = 0.004

p = 0.0001

p = 0.03 continued

312

Ch 11 final.qxd

12/18/2006

15:31

Page 313

SECONDARY CYTOREDUCTIVE SURGERY

Table 11.2 Continued

n Author

Year

Vaccarello34 1995

(total) 57

Median

Median

disease-free

overall

interval

survival Morbidity resection

(months) (months) 20

Bowel

19 (mean)

Residual

Median survival

Hazard

(%)

(%)

criteria

disease n

%

(months)

ratio

23.7

26.3

< 0.5 cm

14

36.8

41†

> 0.5 cm

24

63.2

23

no surgery

9

p < 0.0001

29

NA

Landoni35 1998

38

22

29

NA

21.0

no gross

38

Cormio36

21

25

29

42.9

38.1

no gross

15

71.4

32

gross

6

28.6

9

no gross

17

(56.7)

37

13

43.3

19

< 1 cm

23

38.3

19

> 1 cm

37

61.7

8

no gross

14

63.6

NA

5

22.7

NA

> 1 cm

3

13.6

NA

no gross

87

82.1

44.4

gross

19

17.9

19.3

1999

Gadducci37 2000

Zang38

Chen39

2000

2000

Eisenkop40* 2000

Scarabelli41*2001

Munkarah42 2001

Tay43

2002

Meredith44 2003

Zang45

2004

30

60

22

106

149

25

46

26

117

17.5

13

26

16.8

NA

21

NA

41

34.4

NA

37.6**

26

29.4

15.4

20.0

5.0

28.6

32.1

26.2

44.0

22.5

26.3

22

8.7

23.1

7.7

36.7

26.7

50.0

46.2

33.6

32.0

35.0

42.3

88.0

100

Significance

p = 0.02

p = 0.04

p = 0.0000

NA

p = 0.0007

no gross

53

35.6

1

0.1–1.0 cm

51

34.2

2.65

1.43–4.92††

> 1 cm

45

30.2

5.79

2.99–11.21††

< 2 cm

18

72.0

56.9

> 2 cm

7

28.0

25.1

no gross

19

41.3

38

0.1–1.0 cm

14

30.4

14.5

> 1.0 cm

13

28.3

11

< 1 cm

21

80.8

27.3

> 1 cm

5

19.2

8.6

no gross

11

9.4

84‡

0.1–1.0 cm

61

52.1

26

> 1 cm

45

38.5

14.5

p = 0.08

p = 0.002

p = 0.031

p < 0.0001

NA, not available; * Prospective study; †75% probability; ‡median survival not yet reached; ††95% confidence interval; **diagnosis to recurrence interval

313

Ch 11 final.qxd

12/18/2006

15:31

Page 314

SURGERY FOR OVARIAN CANCER

surgery, the authors concluded that secondary cytoreduction can benefit those patients having had a good response to primary surgery and chemotherapy (pathologic complete response followed by a prolonged disease-free interval), because these qualities would predict a favorable response to second-line therapy. The ideal methodology for assessing the value of secondary cytoreductive surgery for patients with recurrent ovarian cancer would be a prospective trial that randomizes patients to surgical or non-surgical arms followed by equivalent salvage chemotherapy. While such a trial has yet to be successfully conducted, prospectively collected data are available in the form of two sequential reports from Eisenkop et al.33,40 In 1995, the initial report detailed their experience with 36 patients with large-volume recurrent ovarian cancer relapsing at a minimum of 6 months after completing primary platinum-based chemotherapy.33 In

primary platinum-based chemotherapy.34 Thirty-eight patients (67%) underwent laparotomy at the time of recurrence, of which 36 had bulky disease (> 0.5 cm) prior to resection. Of the 23 patients in whom cytoreductive surgery was attempted, 14 (61%) completed surgery with optimal (< 0.5 cm) residual disease and had a median survival time of 41 months, which was significantly longer than both patients undergoing suboptimal resection (median survival 23 months) and those who were not explored (median survival 9 months). Tumor cytoreduction to residual disease of < 0.5 cm was the only statistically significant predictor of survival. Further analysis revealed a threetiered stratification according to whether surgical cytoreduction was undertaken at all, and if so, the volume of residual disease remaining afterwards (Figure 11.2). Although this study was limited by the relatively small number of patients undergoing

1.0

Proportion surviving

0.8

0.6

0.4

0.2

0.0 0

14

28

42

55

70

84

Time (months)

Figure 11.2 Survival from recurrence for no laparotomy group ( ( ; 24 patients, seven censored) and < 0.5 cm (

; 19 patients, 0 censored), and laparotomy patients with > 0.5 cm

; 14 patients, 11 censored) residual disease (median: 9 months, 23 months, and not

reached, respectively; p < 0.0001). For residual disease of < 0.5 cm, the probability of surviving 41 months is 75%. From reference 34, with permission

314

Ch 11 final.qxd

12/18/2006

15:32

Page 315

SECONDARY CYTOREDUCTIVE SURGERY

this study, optimal secondary cytoreduction was defined as no grossly evident residual disease (microscopic residual) and was associated with a dramatic survival advantage (median survival 43 months) compared to patients left with any visible disease (median survival 5 months). In 2000, these authors reported an update of their initial series, which included 106 patients with platinum-sensitive recurrent ovarian cancer submitted to an attempt at secondary cytoreduction.40 The criteria for study inclusion are shown in Table 11.3. A complete secondary resection of all macroscopic disease was accomplished in 82.1% of patients and often required extensive surgical procedures. Residual disease measuring < 0.5 cm was left in 2.8% of patients, while 15.1% of patients had bulky (> 5 cm) residual disease. For all patients, the median survival time from secondary cytoreduction was 34.4 months. Patients who were rendered visibly disease free at secondary cytoreduction had a median survival time of 44.4 months, which was statistically significantly longer than the median survival time of 19.3 months for patients left with any visible disease (p = 0.0007) (Figure 11.3). On multivariate analysis,

Table 11.3 Criteria

for

consideration

of

secondary

cytoreduction of recurrent ovarian cancer (from reference 40)

1. Completion of primary surgery and chemotherapy with a clinical, radiographic and serologic disease-free interval of at least 6 months after primary adjuvant therapy 2. A rising CA-125 level and/or radiographic or physical findings suggestive of recurrence 3. Absence of unresectable extra-abdominal or hepatic metastases 4. Patient willingness to be treated with chemotherapy or radiation therapy after recovery from surgery 5. Absence of medical contraindications to an extensive surgical procedure 6. Gynecologic Oncology Group performance status < 4 (completely disabled; no self-care)

complete secondary cytoreduction retained significance as an independent predictor of survival. Scarabelli et al. reported the only other prospectively collected series of secondary cytoreductive surgery for recurrent ovarian cancer, in 2001.41 These authors studied 149 patients undergoing secondary debulking operations for a minimum of 6 months after completing primary therapy, with most patients (85.9%) having a disease-free interval of < 24 months. Fifty-three patients (35.6%) were rendered visibly disease free (complete cytoreduction), 51 patients (34.2%) had residual disease of ≤ 1 cm and 45 patients (30.2%) had residual intra-abdominal disease larger than 1 cm. After controlling for confounding variables, residual disease after secondary cytoreduction was the clinical factor most predictive of subsequent survival outcome, with patients left with no visible residual disease surviving significantly longer than both patients with macroscopic residual disease of ≤ 1 cm (hazard ratio 2.65, 95% confidence interval (CI) 1.43–4.92) and those with > 1 cm residuum (hazard ratio 5.79, 95% CI 2.99–11.21). Recently, Tay et al. from the Royal Hospital for Women published their experience with 46 patients undergoing secondary cytoreduction for recurrent ovarian cancer after a median disease-free interval of 26 months.43 This was a highly selected study group, with 89% of patients having a disease-free interval longer than 12 months and 54% of patients having localized disease. The overall median post-recurrence survival time for all patients was 22.5 months. Complete cytoreduction was achieved in 41% of patients and was associated with a statistically significant survival advantage compared to patients left with any visible residual disease (median survival 38 vs. 11 months, p = 0.002). The survival effects of important clinical factors were evaluated in a multivariate analysis, which revealed that patients with no residual disease experienced a statistically significant and independent 70% relative risk reduction (95% CI 0.13–0.71) compared to patients with residual disease measuring > 1 cm.

315

Ch 11 final.qxd

12/18/2006

15:32

Page 316

SURGERY FOR OVARIAN CANCER

1.0 0.9

Cumulative survival

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

12

24

36

48

60

72

84

96

Survival time (months)

Figure 11.3 Survival determined by cytoreductive outcome. Solid line, visibly disease-free; dashed line, not visibly disease-free; p = 0.007 (multivariate analysis). From reference 40, with permission

Collectively, the 17 studies depicted in Table 11.2 represent a total of 889 patients undergoing secondary cytoreductive surgery for recurrent ovarian cancer. (Since the two reports from Eisenkop et al. represent a continuing series of patients, only the most recent publication from this group was used to compile summary data.) Of the 15 studies that included a statistical analysis of survival according to cytoreductive surgical outcome or volume of residual disease, 13 (86.7%) reported a significant association between optimal secondary cytoreduction and median post-recurrence survival time. The median postrecurrence survival time for patients undergoing complete secondary cytoreduction, ranges from 29 to 44.4 months.31,33,35–37,40,41,43 In one study reported by Zang et al., the median survival time for patients completely cytoreduced had not yet been reached at 84 months.45 In the two studies that did not find a statistically significant survival advantage, patients left with minimal residual disease had a median survival time that was 1.4-fold and 2.3-fold longer than that of patients undergoing suboptimal resection.30,42 From the summary data in Table 11.2 it is also apparent that

316

the majority of patients left with bulky residual disease (larger than 1–2 cm) experienced a median survival time ranging from 3 to 14.5 months.29–32,38,43,45 Two studies by Vaccarello et al. and Munkarah et al., however, reported median survival times of 23 months and 25.1 months, respectively, for suboptimally debulked patients.34,42 Despite the absence of randomized prospective data, the consistency of the positive relationship between residual disease and subsequent survival outcome suggests that, in appropriately selected patients, successful secondary cytoreduction is associated with a clinically meaningful survival advantage, with the most significant benefit noted in those patients in whom all visible disease can be removed.

FEASIBILITY AND MORBIDITY OF SECONDARY OPERATIONS The feasibility of achieving a successful secondary resection for recurrent ovarian cancer is dependent upon multiple factors, not the least of which is the

Ch 11 final.qxd

12/18/2006

15:32

Page 317

SECONDARY CYTOREDUCTIVE SURGERY

surgeon’s criteria for defining an optimal volume of residual disease. The standard for an optimal result must take into account not only the technical achievability of surgical cytoreduction (i.e. how much disease can safely be extirpated) but also the surgical intent (i.e. the point at which the surgeon is satisfied that the procedure has achieved its desired purpose and can be concluded). Overall results for the 889 patients compiled in Table 11.2 reveal that 577 (64.9%) underwent an optimal resection by the authors’ criteria, which ranged from < 2 cm to no gross residual disease. Ten studies, including 559 patients, reported the proportion of patients undergoing complete cytoreduction, which was achieved in 298 cases (53.3%).31,33,35–37,39–41,43,45 The reported rates of surgical success reflect both the use of discriminating patient selection criteria for surgery (detailed below) as well as the degree of surgical radicality that is often required in secondary operations. From the collected literature, resection of a portion of the intestinal tract was necessary in 44.1% of cases (392 of 889 patients), with the rate in individual series ranging from 21 to 88%. Small-bowel resection is performed in 16–48% of cases, while resection of the colon was performed in 9.4–58% of secondary operations for recurrent ovarian cancer.32,40,43,46 As with primary cytoreductive surgery, secondary operations for recurrent ovarian cancer are associated with a significant but acceptable risk of morbidity. In the collected series, the overall incidence of postoperative complications was 21.5% (183 of 851 patients) (Table 11.2). The most frequently reported complications were intestinal ileus and minor infectious morbidity such as wound cellulites, urinary tract infection and pneumonia.29–31,34,36,40 Significant life-threatening morbidity is uncommon. Enterocutaneous fistula has been reported in 3.3–6.2% of patients following secondary cytoreductive surgery, and anastomotic leak may complicate as many as 16% of intestinal resections performed for recurrent disease.31,40,46 The mortality risk associated with secondary operations is comparable to that of initial debulking surgery but

varies somewhat according to the size of the study population. Most authors have reported a risk of perioperative death ranging from 0 to 3.3%.29,30,32,34,37,40,43

SELECTION CRITERIA The selection criteria utilized to identify appropriate candidates for secondary cytoreductive surgery for recurrent ovarian cancer are based on two principal elements: (1) clinical characteristics other than residual disease that correlate with subsequent survival outcome (i.e. prognostic factors); and (2) factors associated with, or predictive of, surgical outcome.

Prognostic factors In clinical practice, the decision of whether to proceed with an attempt at secondary cytoreduction is usually influenced by the presence or absence of prognostic characteristics that have been consistently associated with a favorable survival outcome independent of residual tumor volume (Table 11.4). While this practice necessarily introduces a component of selection bias into interpreting the degree of clinical benefit, it also serves to distinguish the subset of patients most likely to derive a survival advantage from surgical exploration and cytoreduction. Of the clinical characteristics that have been studied, the disease-free interval following completion of primary therapy has been most consistently associated with post-recurrence survival time. The prognostic impact of time to recurrence on survival is not peculiar to patients who undergo secondary cytoreduction, however. A retrospective review of data from multiinstitutional Canadian chemotherapy trials found that significant factors for post-relapse survival were: time from diagnosis to first relapse or progression; tumor grade at diagnosis; and performance status at diagnosis.47 Although the ideal discriminating time criterion to select surgical candidates has been difficult to define, in general the longer the disease-free interval the better the prognosis. More specifically, patients with a disease-free interval of at least 12

317

Ch 11 final.qxd

12/18/2006

15:32

Page 318

SURGERY FOR OVARIAN CANCER

Table 11.4 Prognostic factors associated with prolonged survival following secondary cytoreductive surgery other than residual disease Age < 55 years32 Disease-free interval > 12 months31,32,39,44 Disease-free interval > 17.5 months37 Disease-free interval > 24 months43 Disease-free interval > 36 months33,40 Optimal (< 2 cm) primary cytoreductive surgery32 Complete clinical response to platinum primary chemotherapy32 GOG performance status 3;33 ECOG performance status 245 Ascites38 Size of largest tumor < 10 cm40 Number of recurrence sites > 145 Administration of any postoperative treatment31 Six or more cycles of salvage chemotherapy45 Secondary cytoreductive surgery prior to salvage chemotherapy33,40 GOG, Gynecologic Oncology Group; ECOG, Eastern Cooperative Oncology Group

months have been independently associated with an improved survival outcome compared to patients who have earlier disease recurrence.31,32,38,44 Consequently, the 12-month mark is generally taken as an indication of chemotherapy or platinum-sensitive disease and used as a minimum requirement before considering secondary operative intervention. Other authors have reported that longer disease-free intervals, ranging from 17.5 to 36 months, are more predictive of improved post-recurrence survival.33,37,40,43 Eisenkop et al. found that post-recurrence survival time was directly proportional to an increasing duration of the disease-free interval, independent of surgical outcome.40 Patients diagnosed with recurrent disease of > 36 months after completing initial therapy survived significantly longer (median survival time

318

56.8 months) than those with a disease-free interval of 13–36 months (median survival time 44.4 months) and 6–12 months (median survival time 25.0 months, p = 0.005) (Figure 11.4). Tay et al. found that a disease-free interval after completing primary therapy of ≥ 24 months had a more favorable prognosis (median survival 39 months) compared to an interval of 12–24 months (median survival 11 months) or < 12 months (median survival 6 months, p = 0.001).43 In their multivariate analysis, a disease-free interval of > 24 months was the only factor other than residual disease that was a significant and independent predictor of survival (Figure 11.5). Specifically, a disease-free interval of ≥ 24 months carried a 75% relative risk reduction (95% CI 0.08–0.75) compared to a diseasefree interval of less than 12 months. Additional patient and disease-related characteristics have also been variably associated with survival outcome following secondary cytoreductive surgery. Gynecologic Oncology Group (GOG) and Eastern Cooperative Oncology Group performance status scores have both been correlated with survival outcome, but it is unclear whether this is a reflection of tumor burden at the time of recurrence, the capacity to tolerate an aggressive multi-modality treatment approach, or both.40,45 Clinical indicators of the extent or pattern of recurrence, such as the presence of ascites, the number of recurrent disease sites and the size of largest disease recurrence, have also been reported to correlate with survival, although the strength of these associations has been inconsistent.38,40,45 The sequence of post-recurrence treatment (i.e. surgery prior to chemotherapy or chemotherapy prior to secondary surgery) has also been investigated as a potentially important prognostic factor. Of the 106 patients with recurrent ovarian cancer reported by Eisenkop et al., 60.4% of patients underwent secondary cytoreduction before receiving salvage chemotherapy, while the remaining patients received some salvage chemotherapy before undergoing secondary surgery.40 To ensure equivalent starting points for survival comparisons and to adjust for the variable

Ch 11 final.qxd

12/18/2006

15:32

Page 319

SECONDARY CYTOREDUCTIVE SURGERY

1.0 0.9

Cumulative survival

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

12

24

36

48

60

72

84

96

Survival time (months)

Figure 11.4 Survival determined by disease-free interval. Solid line, 6–12 months; dashed line, 13–36 months; dotted line, > 36 months; p = 0.005 (multivariate analysis). From reference 40, with permission

sequence of adjuvant therapy administration, survival data were calculated from the date of recurrence to the date of last contact or death. In this study, the median post-recurrence survival time was significantly longer for patients undergoing secondary cytoreduction prior to receiving salvage chemotherapy (48.4 months) compared to patients who received chemotherapy before undergoing secondary surgery (median survival 24.9 months, p = 0.0002) (Figure 11.6). After controlling for other variables, including surgical outcome, by multivariate analysis, this association retained significance as an independent predictor of survival. However, other investigators have not found the sequence of post-recurrence treatment to be significantly associated with post-recurrence survival time.42,45

Predictors of surgical outcome Because surgical outcome, or rather residual disease, is one of the most consistent factors associated with survival, identification of characteristics that predict a favorable operative result are of value in targeting the subset of patients most likely to derive a survival ben-

efit from secondary cytoreduction (Table 11.5). Patient-related characteristics predictive of optimal secondary surgery include younger age and a more functional performance status, and underscore the requisite capacity to tolerate an extensive procedure and aggressive multi-modality therapy.32,33,40 A history of optimal primary cytoreduction and complete clinical response to initial platinum-based chemotherapy followed by a prolonged disease-free interval have also been associated with a higher likelihood of successful surgical resection at the time of disease recurrence.31,32,37 As with primary treatment, the availability of an effective adjuvant therapy (e.g. platinumsensitive disease) is closely linked to the degree of potential benefit from secondary surgical cytoreduction. Eisenkop et al. reported that the administration of salvage chemotherapy prior to secondary surgical intervention had a statistically significant and independent adverse effect on the probability of achieving complete cytoreduction (64.3%) compared to when surgery preceded chemotherapy (93.8%).40 These authors suggested that treating relapsed ovarian cancer with chemotherapy prior to considering secondary

319

Ch 11 final.qxd

12/18/2006

15:32

Page 320

SURGERY FOR OVARIAN CANCER

p Value No prior platinum

0.58

Residual disease 0.01

Nil < 1 cm

0.30

Disease-free interval 0.43 12–24 months > 24 months

0.01 0

0.25

0.5

0.75

1

1.25

Relative risk

Survival benefit

1.5

1.75

2

Survival detriment

Figure 11.5 Effect of prognostic factors on survival after secondary cytoreduction. Risk ratios ( ) and 95% confidence intervals. From reference 43, with permission

1.0 0.9

Cumulative survival

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

12

24

36

48

60

72

84

96

Survival time (months)

Figure 11.6 Survival determined by use of preoperative salvage chemotherapy. Solid line, no preoperative salvage chemotherapy; dashed line, preoperative salvage chemotherapy given; p = 0.001 (multivariate analysis). From reference 40, with permission

320

Ch 11 final.qxd

12/18/2006

15:32

Page 321

SECONDARY CYTOREDUCTIVE SURGERY

Table 11.5 Characteristics associated with a higher likelihood of optimal secondary cytoreductive surgery Age < 55 years32 Disease-free interval > 12 months31,32 Disease-free interval > 17.5 months37 Optimal (< 2 cm) primary cytoreductive surgery32,37 Complete clinical response to platinum primary chemotherapy32 GOG performance status 333,40 Size of largest tumor < 10 cm33,40 Number of recurrence sites 1 versus > 137 Secondary cytoreductive surgery prior to salvage chemotherapy40 GOG, Gynecologic Oncology Group

surgery may select for patients whose disease is resistant to chemotherapy and permit additional time for tumor growth, thereby reducing the likelihood of surgical success. Characteristics of individual tumor biology are also considered to have a significant impact on the ‘resectability’ of recurrent ovarian cancer. Intuitively, it would seem that patients with anatomically localized recurrent disease would have a higher likelihood of undergoing a successful or complete resection; however, the precise definition of localized disease has been somewhat elusive. Tay et al. noted that no patient with disseminated disease underwent a complete secondary cytoreductive operation, whereas 76% of patients with ‘localized’ disease underwent a complete resection.43 Zang et al. also found that the pattern and extent of relapse influenced the likelihood of an optimal (≤ 1 cm) secondary resection, which was achieved in 87.9% of patients with a solitary site and 51.2% of patients with multifocal recurrent disease (p = 0.0002).45 Of the 25 patients reported by Munkarah et al., all were suspected of having a solitary recurrence preoperatively.42 At surgery, 60% of patients had one site of recurrence, 24% had two sites, and 16% had three or more sites. However,

these authors found that the number of tumor sites detected at the time of surgical exploration was not significantly associated with the surgical outcome. Bulky disease involving the bowel mesentery and underlying structures, the porta hepatis and lesser sac disease have been cited as most often precluding a meaningful cytoreductive effort.43 In those circumstances in which the tumor location or distribution is uncertain, consideration may also be given to performing an initial laparoscopy to evaluate the abdominopelvic cavity and directly assess the prospects for a successful resection. Taken together, the above data suggest that secondary cytoreductive surgery is most likely to result in an optimal or complete resection and be associated with prolonged post-recurrence survival in those patients who: (1) have achieved and maintained a complete clinical response to primary platinum-based chemotherapy for a period of 12 months or longer; (2) have localized (≤ 3 sites) disease that is technically resectable; (3) are of satisfactory health and performance status to tolerate a major operation; (4) are willing to receive subsequent salvage chemotherapy; and (5) undergo surgical exploration prior to the administration of salvage chemotherapy.

CLINICAL APPLICATIONS The sites for recurrence of ovarian cancer mirror the distribution of disease at initial diagnosis and include the pelvic cavity, the abdominal peritoneal surfaces, the small and large bowel, the retroperitoneal lymphatic chains, and less commonly the spleen and liver.43 Intraoperatively, the initial steps should include a detailed assessment of the extent of recurrent disease and the feasibility of completely resecting the tumor(s). One added factor to consider is related to the adhesion formation that occurs after a prior surgery. Particularly in ovarian cancer patients who have undergone successful pharmacologic therapy for disseminated abdominal carcinomatosis, the peritoneal and serosal implants are often replaced by

321

Ch 11 final.qxd

12/18/2006

15:32

Page 322

SURGERY FOR OVARIAN CANCER

fibrotic adhesions that form as the tumor cells die in response to chemotherapy and may obliterate the anatomic planes, making the surgical dissection more difficult. The use of sharp dissection is very important to avoid any tear injuries that usually occur as a result of blunt separation of densely adherent tissues. All of the surgical principles and techniques that are used in primary cytoreduction, described earlier in this text, are applicable in the setting of secondary operations for recurrent disease. The following sections review the clinical application of these techniques to some of the more common (gastrocolic ligament and central pelvis) and more challenging (spleen and liver) disease sites encountered during secondary operations for ovarian cancer.

Resection of the gastrocolic ligament A frequent location of recurrent ovarian cancer, particularly in a patient who had a limited or subtotal evaluation of the upper abdomen, is within the residual supracolic omentum or gastrocolic ligament. The surgical approach should begin by developing the lesser sac and delineating the extent of disease in relation to the transverse colon and its mesentery, the greater curvature of the stomach, and the spleen. Previous dissection in the lesser sac may obliterate this plane, increasing potential injury to the colonic mesentery and vasculature. Careful dissection and resection in this region is required, as ischemic injury to the stomach may follow extensive extirpation. In most circumstances, however, the gastroepiploic vascular arcade along the greater curvature of the stomach can be safely sacrificed, as long as the left gastric artery (with its transmural blood supply) is left intact. In many cases, remnants of the omental vessels from the right and left gastroepiploic artery and vein are present and resection of the residual tissue requires division of these vessels. Given the potential for postoperative gastric distension, transfixion of vascular pedicles to the muscularis along the greater curvature of the stomach is suggested along with routine nasogastric decompression. Particular caution must also be observed in the area of the gastrosplenic ligament, as

322

division of the short gastric vessels is often necessary to gain access to the omental tissue between the spleen and fundus of the stomach. Overzealous vascular ligation may lead to ischemia of the cardia and fundus of the stomach, while excessive traction on the stomach increases the potential for a splenic laceration.

Resection of central pelvic recurrence Whereas in patients with newly diagnosed ovarian cancer, direct retroperitoneal invasion of the tumor into the rectovaginal space is rather infrequent, the retroperitoneal space between the bladder and rectum is not an uncommon site for recurrence in women who have previously undergone a hysterectomy (Figure 11.7). Complete resection of tumor in this specific situation can be complicated but is usually achievable. Resection of the rectosigmoid colon, or a modified posterior exenteration, has reportedly been necessary in as many as 30.2% of patients with recurrent disease, in order to achieve an acceptably high rate of optimal residual disease.40 In this approach, the pelvic sidewalls are opened lateral to the iliac vessels, and the major iliac vessels and ureters identified and freed from the peritoneal reflection. The pararectal and paravesical spaces are opened bilaterally, thus isolating the parametrial and paravaginal tissues. The branches of the hypogastric vessels supplying the tumor including the uterine and vaginal arteries and veins can be ligated and transected, thus significantly decreasing the blood supply to the tumor. The parametrial and paravaginal attachments of the tumor are also severed. The cul-de-sac peritoneum, covering the rectovaginal mass, is then incised. Using sharp dissection, the tumor is separated from the base of the bladder anteriorly. In patients with large rectovaginal tumors, causing significant distortion of the anatomy, placing a spongestick or an end-to-end anastomosissizing instrument in the vagina can help identify the boundaries between the vaginal cuff and the bladder. If the tumor is invading into the bladder muscularis layer, a partial cystectomy might be indicated. One should be careful to avoid injury to the bladder

Ch 11 final.qxd

12/18/2006

15:32

Page 323

SECONDARY CYTOREDUCTIVE SURGERY

Figure 11.7 Photograph of an obliterated cul-de-sac during a secondary cytoreduction procedure. Dissection is complicated by distorsion of anatomic planes and complex and dense adhesions involving the bowel and bladder

trigone. Occasionally, resection of the distal part of the ureter followed by ureteroneocystostomy is necessary to obtain tumor-free margins. On the posterior aspect, the recurrent tumor frequently invades the rectal wall. Because of the absence of a serosal layer covering the rectum below the cul-de-sac reflection, dissecting the tumor from the colonic muscularis is often impossible without significantly damaging the rectal wall. Therefore, a partial resection of the distal sigmoid and proximal rectum becomes necessary. The rectum is mobilized by developing the presacral space and elevating the rectum out of the sacral hollow and pelvic floor. The extent of tumor involvement is determined and the rectum is transected distal to the lowest part of tumor invasion using the techniques described in Chapter 5. Similarly, the sigmoid is transected proximal to the area of tumor invasion. At this point, the rectovaginal tumor is freed laterally from its parametrial and paravaginal attachments, anteriorly from the bladder, and is removed along with a segment of the colon. Reanastomosis of the sigmoid to

the rectum can then be performed as described in Chapter 5.

Splenectomy The spleen is not an uncommon site of isolated recurrence of ovarian cancer, particularly in those patients undergoing a maximal cytoreductive effort at initial surgery. Consequently, splenectomy may be required to achieve an optimal secondary resection in as many as 13.2% of secondary operations for recurrent disease and can be accomplished by either an anterior or a posterior approach, as described in Chapter 9.32,35,40,41,46 Chen et al. from Cedars Sinai Medical Center reported on 22 patients with recurrent epithelial ovarian cancer undergoing splenectomy as part of a secondary cytoreductive surgical procedure.39 The median time to recurrence was 26 months and the average size of splenic metastasis was 5.2 cm. In eight of 22 patients (36.4%), recurrent ovarian cancer was localized to the left upper abdomen (i.e. no other sites

323

Ch 11 final.qxd

12/18/2006

15:32

Page 324

SURGERY FOR OVARIAN CANCER

of recurrent disease). Despite the fact that multiple additional surgical procedures were required, including bowel resection in 50% of cases, the morbidity was acceptable, with 28.6% of patients experiencing significant complications. In this series, 19 of 22 patients (86.4%) were left with an optimal volume of residual disease (< 1 cm), with 14 patients (63.6%) undergoing complete cytoreduction to no gross residual. Although the survival analysis was not specified according to residual disease status, the entire cohort of patients undergoing secondary cytoreduction including splenectomy experienced a medial survival time from the date of recurrence of 41 months. The authors recommended vigilant surveillance for isolated ovarian cancer recurrence in the interests of identifying the specific subgroup of patients with a good performance status and platinum-sensitive disease that might benefit from an aggressive surgical approach including splenectomy. Scarabelli et al. reported on 14 patients undergoing splenectomy as part of a secondary cytoreductive surgical procedure and noted a 2-year survival rate of 78% for patients left with no visible residual disease compared to 24% for patients with any macroscopic residual disease.48 The median survival times for these two groups were 27 months and 16 months, respectively. Finally, Gemignani et al. reported the experience at Memorial Sloan-Kettering with a very select group of six patients who underwent splenectomy for recurrent epithelial ovarian cancer confined to the splenic parenchyma.49 In this report, secondary cytoreduction was performed at a median of 57 months following initial diagnosis and all patients were without evidence of disease at the median follow-up time of 25.5 months.

Resection of hepatic disease In the setting of primary surgery for advanced-stage ovarian cancer, parenchymal liver disease is a frequently cited exclusion criterion for optimal cytoreduction such that the surgical effort may be abbrevi-

324

ated or abandoned altogether. Early studies of secondary cytoreductive surgery also refer to hepatic recurrence as precluding an attempt at optimal secondary debulking.29.30 More contemporary reports, however, have described successful resection of hepatic disease in 4–9.4% of patients undergoing secondary operations for recurrent ovarian cancer using the techniques described in Chapter 8.32,40,41,43,46 Recently, investigators from the Mayo Clinic reported the only series of patients undergoing hepatic resection exclusively for recurrent ovarian cancer.44 The authors retrospectively identified 26 patients who had complete segmentectomies (69.2%), right or left hepatectomy (19.2%), or trisegmentectomy (11.5%). Twenty patients required cytoreduction of extrahepatic tumor, with 42.3% undergoing an intestinal resection. Overall, 21 patients (80.8%) were left with an optimal volume of residual disease (< 1 cm), with 18 patients (69.2%) having no macroscopic residual disease. Six patients (23.1%) experienced significant postoperative morbidity. For the entire patient cohort, the median survival time following hepatic resection for recurrent disease was 26.3 months, with patients undergoing optimal cytoreduction surviving significantly longer (median survival time 27.3 months) than those left with suboptimal residual disease (median survival time 8.6 months). Further analysis revealed that neither the number of liver lesions nor whether the tumor had a unilobar or multilobar distribution were significantly associated with survival outcome. The authors concluded that patients with recurrent ovarian cancer involving the liver experience comparable survival rates to patients undergoing successful secondary cytoreduction without a hepatic component, as long as an optimal volume of residual disease can be achieved. Therefore, the decision to incorporate hepatic resection into the treatment of recurrent ovarian cancer with liver metastasis should be based primarily on the likelihood of achieving an optimal cytoreductive surgical result.

Ch 11 final.qxd

12/18/2006

15:32

Page 325

SECONDARY CYTOREDUCTIVE SURGERY

SELECTION OF POSTOPERATIVE ADJUVANT THERAPY As outlined above, two principal characteristics of patients best suited for secondary cytoreductive surgery are: (1) a long progression-free interval following primary therapy; and (2) disease that appears more or less isolated and of limited volume, and that seems to be completely (visibly) resectable. These characteristics also portend a favorable subsequent response to chemotherapy, and such patients are classically defined, if by no other characteristic than treatment-free interval, as ‘chemosensitive’. In addition, since this cohort may be left with small-volume disease or visibly no gross residual disease, alternative routes of therapy may be considered, such as intraperitoneal administration of radiocolloids and chemotherapy. These patients arguably have the widest range of treatment options as well as the highest expectations for response among women with recurrent ovarian cancer. Consequently, in the absence of randomized prospective data, it is difficult to quantify the ‘added-value’ of cytoreductive surgery in this cohort of patients. An exhaustive review of individual chemotherapy agents, combinations and strategies in the setting of recurrent ovarian cancer is beyond the scope of this text (comprehensive reviews can be found elsewhere50,51). Because cisplatin and carboplatin are generally regarded as the most active agents in advanced ovarian cancer, platinum-based therapy is often considered first when ‘chemosensitive’ disease recurs. Patients who have a prolonged duration of response to first-line platinum-based therapy have an increased likelihood of responding to re-treatment with a platinum agent. Combined data from the landmark Gore and Markman studies show higher response rates with increasing treatment-free intervals up to 24 months and beyond.52,53 Specifically, the response rate to a platinum agent after a treatmentfree intervals of 5–12 months is 22% compared to 31% and 59% after intervals of 12–24 months and > 24 months, respectively.

Similar effects related to treatment-free intervals are also seen with many of the novel non-platinum agents. For instance, in a large randomized trial of recurrent ovarian cancer patients treated with a single agent (liposomal doxorubicin or topotecan), the clinical response rates of these two agents were 12% and 7%, respectively, in platinum-resistant patients and 28% and 29%, respectively, among platinumsensitive patients.54 Additional data suggest that progression-free survival may be the most important factor in determining subsequent tumor response and overall survival. In a review of eight GINECO phase II and III chemotherapy clinical trials involving 583 patients with recurrent ovarian cancer, PujadeLauraine et al. demonstrated that response and survival were closely related to the initial progressionfree interval.55 Patients recurring within 3–12 months of their initial treatment had a response rate to salvage therapy of 35% and a median overall survival time of 393 days. In contrast, patients with a progression-free survival of at least 18 months had a response rate of 62% and a median overall survival time of 957 days. Importantly, these effects were independent of the individual chemotherapeutic agents administered in the salvage setting. Thus, it would be reasonable to expect that patients considered for surgery would enjoy, by virtue of selection criteria, longer survival in the presence of subsequent chemotherapy than those with shorter treatment-free intervals either with or without surgical intervention. Non-randomized data, as presented previously, would support this conclusion. A popular treatment strategy in patients with long treatment-free intervals is the use of combination chemotherapy. Some clinicians maintain that a combination regimen may have greater efficacy than single-agent carboplatin or cisplatin, since a number of phase II trials suggest somewhat higher response rates and disease-free survivals with combination treatment. Dizon et al. reported a 70% overall response rate and a 42% complete clinical response rate to a combination of paclitaxel and carboplatin

325

Ch 11 final.qxd

12/18/2006

15:32

Page 326

SURGERY FOR OVARIAN CANCER

among 66 drug-sensitive patients who had a median treatment-free interval of 22 months.56 The median progression-free interval following salvage therapy was 13 months, and the 3-year overall survival rate was 72%. These favorable responses must be carefully interpreted within the context of their study cohort, which was significantly heterogeneous. Prospective data are very limited in this setting. Bolis et al. reported on 190 platinum-sensitive (recurrence > 6 months after primary therapy) patients who were randomized to receive either single-agent carboplatin (300 mg/m2) or combination carboplatin (300 mg/m2) and epidoxorubicin (120 mg/m2).57 The median platinum-free interval was 17 months and a minimum of five cycles of therapy were planned. The overall response rate was 55% for single-agent therapy compared to 58% for the combination. Median progression-free and overall survival rates favored the combination regimen but were not statistically significant; however, a limitation in sample size may have contributed to the lack of effect. Recently, the ICON collaborators conducted a large randomized trial of paclitaxel and platinum versus conventional single-agent platinum therapy in first-relapse platinum-sensitive patients.58 This study of 802 patients represents the largest such trial completed to date. The overall response rates were 66% for the combination arm and 54% for the singleagent arm (p = 0.06). However, at a median follow-up time of 42 months, significant improvements in both progression-free survival (hazard ratio (HR) = 0.76, p < 0.001) and overall survival (HR = 0.82, p = 0.023) were observed for patients receiving the combination of paclitaxel and platinum. These provocative data, if confirmed, will probably justify combination paclitaxel and platinum chemotherapy in the cohort of patients also strongly considered for secondary surgery, and may be the most promising strategy following a surgical intervention. Another modality that may offer some therapeutic advantage in the setting of the small volume of residual disease following a successful surgical debulking is intraperitoneal therapy. The vehicle product in this

326

case has traditionally been single or combination chemotherapeutics; however, other modalities such as cytokines, interferon (α and γ), radiotherapy (32P) and radioimmunoconjugates (yttrium – 90Y, lutetium – 177Lu) have been administered with mixed results.59–64 In most cases, the amount of residual disease prior to treatment greatly impacts the observed efficacy. For instance, the GOG reported the results of intraperitoneal paclitaxel (60 mg/m2 per week) in patients with small-volume (less than 5 mm) persistent or recurrent disease.65 In this phase II trial, the presence of any macroscopic disease was associated in patients with a significantly lower likelihood of a negative third-look laparotomy (3%) compared to patients surgically debulked to microscopic residual disease at second-look surgery (54%) and those found to have only microscopically persistent disease at second look (67%). It is unknown whether a strategy that prospectively employs a secondary cytoreduction maneuver will enable a cohort to enjoy the favorable findings of these salvage treatment modalities. Based on the available data, it would seem that only patients with completely resected recurrent disease confined to the peritoneal cavity should be considered for salvage therapy administered via the intraperitoneal route.

CONCLUSION Renewed interest in surgical intervention for patients with recurrent ovarian cancer following a prolonged disease-free interval stems from accumulating data suggesting that survival may be improved in this setting. Bolstered with refinements in surgical techniques and operative goals, as well as longer progression-free (and therefore treatment-free) intervals seen after standard front-line chemotherapy combinations, more clinicians are considering this strategy for their patients. A clear understanding of prognostic factors and expectations of therapy are necessary to safely identify appropriate surgical candidates. Ultimately,

Ch 11 final.qxd

12/18/2006

15:32

Page 327

SECONDARY CYTOREDUCTIVE SURGERY

however, randomized trials will be necessary to ferret out the true ‘value-added’ benefit of secondary surgery in this otherwise chemosensitive cohort of patients.

11.

Meier W, Baumgartner L, Stieber P, et al. CA125 based diagnosis and therapy in recurrent ovarian cancer. Anticancer Res 1997; 17: 3019–20

12.

Meier W, Stieber P, Hasholzner U, et al. Prognostic significance of CA125 in patients with ovarian cancer and secondary debulking surgery. Anticancer Res 1997; 17: 2945–7

13.

Rustin GJ, Nelstrop AE, Tuxen MK, Lambert HE. Defining progression of ovarian carcinoma during follow-up according to CA 125: a North Thames Ovary Group Study. Ann Oncol 1996; 7: 361–4

14.

Wilder JL, Pavlik E, Straughn JM, et al. Clinical implications of a rising serum CA-125 within the normal range in patients with epithelial ovarian cancer: a preliminary investigation. Gynecol Oncol 2003; 89: 233–5

15.

Bristow RE, Duska LR, Lambrou NC, et al. A model for predicting surgical outcome in patients with advanced ovarian carcinoma using computed tomography. Cancer 2000; 89: 1532–40

16.

Skipper HE. Adjuvant chemotherapy. Cancer 1978; 41: 936–40

Prayer L, Kainz C, Kramer J, et al. CT and MR accuracy in the detection of tumor recurrence in patients treated for ovarian cancer. J Comput Assist Tomogr 1993; 17: 626–32

17.

6.

Goldie JH, Coldman JA. A mathematic model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat Rep 1979; 63: 1727–33

Reuter KL, Griffin T, Hunter RE. Comparison of abdominopelvic computed tomography results and findings at second-look laparotomy in ovarian carcinoma patients. Cancer 1989; 63: 1123–8

18.

7.

Norton L, Simon R. Tumor size, sensitivity to therapy, and design of treatment schedules. Cancer Treat Rep 1977; 61: 1307–17

Megibow AJ, Bosniak MA, Ho AG, et al. Accuracy of CT in detection of persistent or recurrent ovarian carcinoma: correlation with second-look laparotomy. Radiology 1988; 166: 341–5

8.

Thigpen JT, Herrin VE. Recent developments in the treatment of recurrent ovarian carcinoma. In Rubin SC, Sutton GP, eds. Ovarian Cancer, 2nd edn. Philadelphia: Lippincott Williams & Wilkins, 2001: 301–14

19.

Ricke J, Sehouli J, Hach C, et al. Prospective evaluation of contrast-enhanced MRI in the depiction of peritoneal spread in primary or recurrent ovarian cancer. Eur Radiol 2003; 13: 943–9

20.

Buist MR, Golding RP, Burger CW, et al. Comparative evaluation of diagnostic methods in ovarian carcinoma with emphasis on CT and MRI. Gynecol Oncol 1994; 52: 191–8

21.

Rose PG, Faulhaber P, Miraldi F, Abdul-Karim FW. Positive emission tomography for evaluating a complete clinical response in patients with ovarian or peritoneal carcinoma: correlation with second-look laparotomy. Gynecol Oncol 2001; 82: 17–21

REFERENCES 1.

Meigs JV. Tumors of the Female Pelvic Organs. New York: McMillan, 1935

2.

Griffiths CT. Surgical resection of tumor bulk in the primary treatment of ovarian carcinoma. Monograph Natl Cancer Inst 1975; 42: 101–4

3.

4.

5.

Hoskins WJ, Bundy BN, Thigpen JT, Omura GA. The influence of cytoreductive surgery on recurrence-free interval and survival in small-volume stage III epithelial ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1992; 47: 159–66 Hoskins WJ, McGuire WP, Brady MF, et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J Obstet Gynecol 1994; 170: 974–80

9.

Obermair A, Schmid BC, Packer LM, et al. Expression of MUC1 splice variants in benign and malignant ovarian tumours. Int J Cancer 2002; 100: 166–71

10.

Dehn H, Hogdall EV, Johansen JS, et al. Plasma YKL40, as a prognostic tumor marker in recurrent ovarian cancer. Acta Obstet Gynecol Scand 2003; 82: 287–93

327

Ch 11 final.qxd

12/18/2006

15:32

Page 328

SURGERY FOR OVARIAN CANCER

22.

Bristow RE, Simpkins F, Pannu HK, et al. Positron emission tomography for detecting clinically occult surgically resectable metastatic ovarian cancer. Gynecol Oncol 2002; 85: 196–200

34.

Vaccarello L, Rubin SC, Vlamis V, et al. Cytoreductive surgery in ovarian carcinoma patients with a documented previously complete surgical response. Gynecol Oncol 1995; 57: 61–5

23.

Kubik-Huch RA, Dorffler W, von Schulthess GK, et al. Value of (18F)-FDG positron emission tomography, computed tomography, and magnetic resonance imaging in diagnosing primary and recurrent ovarian carcinoma. Eur Radiol 2000; 10: 761–7

35.

Landoni F, Pellegrino A, Cormio G, et al. Platin-based chemotherapy and salvage surgery in recurrent ovarian cancer following negative second-look laparotomy. Acta Obstet Gynecol Scand 1998; 77: 233–7

36.

24.

Zimny M, Siggelkow W, Schröeder W, et al. 2[Fluorine-18]-fluoro-2-deoxy-D-glucose positron emission tomography in the diagnosis of recurrent ovarian cancer. Gynecol Oncol 2001; 83: 310–15

Cormio G, di Vagno G, Cazzolla A, et al. Surgical treatment of recurrent ovarian cancer: report of 21 cases and a review of the literature. Eur J Obstet Gynecol Rep Biol 1999; 86: 185–8

37.

25.

Chang WC, Hung YC, Kao CH, et al. Usefulness of whole body positron emission tomography (PET) with 18F-fluoro-s-deoxyglucose (FDG) to detect recurrent ovarian cancer based on asymptomatically elevated serum levels of tumor marker. Neoplasma 2002; 49: 329–33

Gadducci A, Iacconi P, Cosio S, et al. Complete salvage surgical cytoreduction improves further survival of patients with late recurrent ovarian cancer. Gynecol Oncol 2000; 79: 344–9

38.

Zang RY, Zhang ZY, Li ZT, et al. Impact of secondary cytoreductive surgery on survival of patients with advanced epithelial ovarian cancer. Eur J Surg Oncol 2000; 26: 798–804

39.

Chen LM, Leuchter RS, Lagasse LD, Karlan BY. Splenectomy and surgical cytoreduction for ovarian cancer. Gynecol Oncol 2000; 77: 362–8

40.

Eisenkop SM, Friedman RL, Spirtos NM. The role of secondary cytoreductive surgery in the treatment of patients with recurrent epithelial ovarian carcinoma. Cancer 2000; 88: 144–53

41.

Scarabelli C, Gallo A, Carbone A. Secondary cytoreductive surgery for patients with recurrent epithelial ovarian carcinoma. Gynecol Oncol 2001; 83: 504–12

42.

Munkarah A, Levenback C, Wolf JK, et al. Secondary cytoreductive surgery for localized intra-abdominal recurrences in epithelial ovarian cancer. Gynecol Oncol 2001; 81: 237–41

26.

27.

28.

Beyer T, Townsend DW, Brun T, et al. A combined PET/CT scanner for clinical oncology. J Nucl Med 2000; 41: 1369–79 Makhija S, Howden N, Edwards R, et al. Positron emission tomography/computed tomography imaging for the detection of recurrent ovarian and fallopian tube carcinoma: a retrospective review. Gynecol Oncol 2002; 85: 53–8 Bristow RE, del Carmen MG, Pannu HK, et al. Clinically occult recurrent ovarian cancer: patient selection for secondary cytoreductive surgery using combined PET/CT. Gynecol Oncol 2003; 90: 519–28

29.

Berek JS, Hacker NF, Lagasse LD, et al. Survival of patients following secondary cytoreductive surgery in ovarian cancer. Obstet Gynecol 1983; 61: 189–93

30.

Morris M, Gershenson DM, Wharton JT, et al. Secondary cytoreductive surgery for recurrent epithelial ovarian cancer. Gynecol Oncol 1989; 34: 334–8

43.

Jänicke F, Holscher M, Kuhn W, et al. Radical surgical procedure improves survival time in patients with recurrent ovarian cancer. Cancer 1992; 70: 2129–36

Tay EH, Grant PT, Gebski V, Hacker NF. Secondary cytoreductive surgery for recurrent epithelial ovarian cancer. Obstet Gynecol 2002; 99: 1008–13

44.

Segna RA, Dottino PR, Mandeli JP, et al. Secondary cytoreduction for ovarian cancer following cisplatin therapy. J Clin Oncol 1993; 11: 434–9

Meredith MA, Cliby WA, Keeney GL, et al. Hepatic resection for metachronous metastases from ovarian carcinoma. Gynecol Oncol 2003; 89: 16–21

45.

Eisenkop SM, Friedman RL, Wang HJ. Secondary cytoreductive surgery for recurrent ovarian cancer. A prospective study. Cancer 1995; 76: 1606–14

Zang RY, Li ZT, Tang J, et al. Secondary cytoreductive surgery for patients with relapsed epithelial ovarian carcinoma: who benefits? Cancer 2004;100: 1152–61

46.

Lichtenegger W, Sehouli J, Buchmann E, et al. Operative results after primary and secondary debulk-

31.

32.

33.

328

Ch 11 final.qxd

12/18/2006

15:32

Page 329

SECONDARY CYTOREDUCTIVE SURGERY

Study Group. Carboplatin alone vs carboplatin plus epidoxorubicin as second-line therapy for cisplatin- or carboplatin-sensitive ovarian cancer. Gynecol Oncol 2001; 81: 3–9

ing-operations in advanced ovarian cancer (AOC). J Obstet Gynaecol Res 1998; 6: 447–51 47.

48.

Hoskins P, Tu D, James K, et al. Factors predictive of survival after first relapse or progression in advanced epithelial ovarian carcinoma: a prediction tree analysis-derived model with test and validation groups. Gynecol Oncol 1998; 70: 224–30 Scarabelli C, Gallo A, Campagnutta E, Carbone A. Splenectomy during primary and secondary cytoreductive surgery for epithelial ovarian carcinoma. Int J Gynaecol Cancer 1998; 8: 215–21

58.

Parmar MK, Ledermann JA, Colombo N, et al; ICON and AGO Collaborators. Paclitaxel plus platinumbased chemotherapy versus conventional platinumbased chemotherapy in women with relapsed ovarian cancer: the ICON4/AGO-OVAR-2.2 trial. Lancet 2003; 361: 2099–106

59.

Edwards RP, Gooding W, Lembersky BC, et al. Comparison of toxicity and survival following intraperitoneal recombinant interleukin-2 for persistent ovarian cancer after platinum: twenty-four-hour versus 7-day infusion. J Clin Oncol 1997; 15: 3399–407

60.

Berek JS, Markman M, Stonebraker B, et al. Intraperitoneal interferon-alpha in residual ovarian carcinoma: a phase II gynecologic oncology group study. Gynecol Oncol 1999; 75: 10–14

61.

Berek JS, Markman M, Blessing JA, et al. Intraperitoneal alpha-interferon alternating with cisplatin in residual ovarian carcinoma: a phase II Gynecologic Oncology Group study. Gynecol Oncol 1999; 74: 48–52

49.

Gemignani ML, Chi DS, Gurin CC, et al. Splenectomy in recurrent epithelial ovarian cancer. Gynecol Oncol 1999; 72: 407–10

50.

Fields AL, Runowicz CD. Current therapies in ovarian cancer. Cancer Invest 2003; 21: 148–56

51.

Salom E, Almeida Z, Mirhashemi R. Management of recurrent ovarian cancer: evidence-based decisions. Curr Opin Oncol 2002; 14: 519–27

52.

Gore ME, Fryatt I, Wiltshaw E, et al. Cisplatin/carboplatin cross-resistance in ovarian cancer. Br J Cancer 1989; 60: 767–9

53.

Markman M, Rothman R, Hakes T, et al. Second-line platinum therapy in patients with ovarian cancer previously treated with cisplatin. J Clin Oncol 1991; 9: 389–93

62.

Gordon AN, Fleagle JT, Guthrie D, et al. Recurrent epithelial ovarian carcinoma: a randomized phase III study of pegylated liposomal doxorubicin versus topotecan. J Clin Oncol 2001; 19: 3312–22

Alvarez RD, Huh WK, Khazaeli MB, et al. A Phase I study of combined modality (90)Yttrium-CC49 intraperitoneal radioimmunotherapy for ovarian cancer. Clin Cancer Res 2002; 8: 2806–11

63.

Pujade-Lauraine E, Paraiso D, Cure H, et al. Predicting the effectiveness of chemotherapy (Cx) in patients with recurrent ovarian cancer (ROC): a GINECO study. Proc Am Soc Clin Oncol 2002: Abstr 829

Favalli G, Odicino F, Torri V, Pecorelli S. Early stage ovarian cancer: the Italian contribution to clinical research. An update. Int J Gynecol Cancer 2001; 11 (Suppl 1): 12–19

64.

Varia MA, Stehman FB, Bundy BN, et al; Gynecologic Oncology Group. Intraperitoneal radioactive phosphorus (32P) versus observation after negative second-look laparotomy for stage III ovarian carcinoma: a randomized trial of the Gynecologic Oncology Group. J Clin Oncol 2003; 21: 2849–55

65.

Markman M, Brady MF, Spirtos NM, et al. Phase II trial of intraperitoneal paclitaxel in carcinoma of the ovary, tube, and peritoneum: a Gynecologic Oncology Group study. J Clin Oncol 1998; 16: 2620–4

54.

55.

56.

57.

Dizon DS, Hensley ML, Poynor EA, et al. Retrospective analysis of carboplatin and paclitaxel as initial second-line therapy for recurrent epithelial ovarian carcinoma: application toward a dynamic disease state model of ovarian cancer. J Clin Oncol 2002; 20: 1238–47 Bolis G, Scarfone G, Giardina G, et al, Associazione per la Ricerca in Ginecologia Oncologia (ARGO 96)

329

Ch 12 final.qxd

12/18/2006

15:33

Page 331

CHAPTER 12

Laparoscopic surgery David E Cohn, Inbar Ben-Shachar, Jeffrey M Fowler

INTRODUCTION Laparoscopy has been utilized in the management of gynecologic conditions for decades. In performing laparoscopic sterilization, gynecologists recognized that this surgical approach accomplished the technical goal of contraception with a reduction in surgical morbidity relative to standard open sterilization. Since that time, laparoscopy has been recognized as a means to accomplish a surgical goal, rather than a separate end itself. It is the concept of laparoscopy as a surgical tool that has led to the expansion of the procedures in which this technique has been employed. One of the most recent applications of laparoscopy has been that of its utility in cancer surgery. Traditionally, the surgical management of cancers was based on the radical excision of the primary tumor and potential routes of spread. However, over the past few decades, cancer surgeons have found themselves balancing surgical radicality with outcome. Interestingly, many diseases previously managed with radical excision are currently being treated using less radical techniques without sacrificing the potential for cure. Concomitant with this change in surgical philosophy was the acceptance of laparoscopy as a viable technique to accomplish this modern surgical goal. Consistently, laparoscopy has been shown to lead to a decreased length of hospital stay and a relatively shorter recovery, often with decreased postoperative pain. However, the technique is often complicated, resulting in longer operating and anesthesia time for completion of a procedure. Also, the lack of tactile response during laparoscopy theoretically limits a sur-

geon’s ability to use palpation as a mechanism to evaluate the surgical site. Potentially the greatest risk with a new procedure in patients with cancer, however, is the potentially negative impact on disease recurrence or survival. In this chapter, the applications of laparoscopy in the management of women with ovarian cancer will be explored. Evaluation of a pelvic mass, primary laparoscopic management of apparently early ovarian cancer, management of advanced ovarian cancer and reassessment laparoscopy will be described in detail. Likewise, the indications, outcomes, limitations and risks specific to laparoscopy in the diagnosis and management of ovarian cancer will be evaluated.

LAPAROSCOPIC EVALUATION OF A PELVIC MASS Despite the widespread use (and often overuse) of transvaginal ultrasonography and CA-125 testing in women with a pelvic mass, these tests are neither 100% specific nor sensitive in confirming or excluding the diagnosis of ovarian cancer. Thus, the definitive evaluation of such a mass is surgery. Although the indications for the evaluation of an adnexal mass by laparotomy or laparoscopy are identical, the decision regarding which method to use for surgical evaluation of a pelvic mass is quite complex. Factors such as the size and characteristics of the mass, the age and weight of the patient, and the skills of the surgeon will certainly be factors in the decision to proceed with either laparoscopy or laparotomy. Adnexal masses occurring

331

Ch 12 final.qxd

12/18/2006

15:33

Page 332

SURGERY FOR OVARIAN CANCER

in women with a history of breast cancer or nongynecologic malignancies are at high risk for both primary and secondary ovarian cancers and should be managed by a gynecologic oncologist.1 When a plan is made for the laparoscopic evaluation of a pelvic mass, the surgical team must determine preoperatively how they will manage a malignancy discovered at the time of laparoscopy. Likewise, it is imperative that patients be counseled regarding the potential need for the surgical management of cancer. If a competent gynecologic laparoscopist were incidentally to discover an ovarian neoplasm clinically suspicious for cancer at the time of laparoscopy, management should include careful inspection of all peritoneal surfaces, including the pelvis, pouch of Douglas, diaphragm, paracolic gutters, omentum and bowel surfaces. It is recommended that peritoneal washings be obtained and that the ovarian mass be removed without spilling cyst contents into the peritoneal cavity. Uncontrolled puncture, biopsy and partial resection of suspicious masses are not appropriate management strategies. Morcellation of ovarian masses should not be performed. Frozen section should be performed on all suspicious ovarian masses when surgery is performed in a location were facilities exist for the appropriate management of ovarian cancer. In general, if a surgeon stratifies the risk of an adnexal mass being malignant into high, medium and low, then the exclusion of high-risk masses from laparoscopy would lead to the diagnosis of cancer in approximately 1–3% of adnexal masses evaluated laparoscopically.2–13 Since no randomized trials of laparoscopy versus laparotomy for the management of ovarian neoplasms exist, the data reported herein regarding the laparoscopic management of ovarian tumors is primarily derived from retrospective series and case reports. Canis et al.2 laparoscopically evaluated 757 patients with 819 masses. After inspection of the masses, these were punctured and their gross characteristics were examined. In this investigation, 6% of the masses were suspicious for malignancy, and 41% of these suspicious masses were found to be malignant or of low

332

malignant potential (LMP). No malignant masses were missed, but seven of the 15 malignant tumors were ruptured (not in a controlled fashion) during laparoscopy. In a follow-up study,3 230 masses suspicious for malignancy by ultrasound were evaluated by laparoscopy. At surgery, 62 of the 230 masses, including all 25 cases of cancer or tumors of LMP, were determined to be suspicious for malignancy. Eight of the 25 cancers were ruptured before histologic confirmation. Twelve of the 15 invasive cancers were immediately managed with laparotomy and staging, whereas the remaining three cases were determined to be without malignancy by frozen section, and therefore definitive management was delayed until the final histologic diagnosis was rendered. Dottino et al.5 reported on a selected series of 160 patients with a pelvic mass who were referred to a gynecologic oncology practice. Laparoscopy was performed for masses that did not extend above the umbilicus, where there was no gross evidence of extraovarian disease, and in patients without other apparent pathology that would make them better suited for laparotomy. These authors encountered nine ovarian cancers (6%), eight tumors of LMP (5%) and four non-gynecologic cancers. Laparoscopic management was successful in 141 (88%) patients. Of the nine patients with invasive ovarian cancers, five underwent immediate laparotomy, three were staged laparoscopically, and one patient who was initially diagnosed with a borderline tumor on frozen section whose tumor was determined to be an invasive ovarian cancer on final histology had no further management at the time of laparoscopy. Overall, 3% of frozen section diagnoses were inconsistent with the final report. Importantly, in the cases in which a final diagnosis of cancer was rendered and not recognized at the time of laparoscopy, no significant delay in the definitive treatment was attributable to the change in diagnosis. From these data, most authors conclude that suspicious adnexal masses can be managed laparoscopically. Ovarian neoplasms should be carefully inspected for clinical evidence of malignancy to direct

Ch 12 final.qxd

12/18/2006

15:33

Page 333

LAPAROSCOPIC SURGERY

surgical therapy. The pelvis and abdomen (including para-colic gutters, diaphragmatic domes and omentum) should be inspected. To minimize the risk of intraperitoneal tumor dissemination and delay in diagnosis, frozen sections of any lesion or suspicious masses are essential. Morcellation of a solid or suspicious tumor should be avoided. Importantly, arrangements for the appropriate management of an ovarian cancer diagnosed at laparoscopy (i.e. laparoscopic staging, laparotomy for staging or cytoreduction) must be established prior to surgery to minimize the need for re-operation.

SURGICAL OUTCOMES OF OPERATIVE LAPAROSCOPY Limited studies exist that describe the outcomes of women treated laparoscopically for ovarian cancer. However, extrapolating from the few studies comparing laparoscopy and laparotomy in the management of benign adnexal masses,12,13 it seems that laparoscopy is associated with less blood loss, decreased pain, decreased postoperative analgesic requirements and shorter length of stay without an apparent increase in complications. Of the few series of surgical outcomes in women with ovarian cancer treated with laparoscopy, the immediate surgical results described in women with benign ovarian neoplasms seems similar to those reported for women with ovarian cancer. Childers et al.4 reported on 138 patients with an adnexal mass suspicious for malignancy by ultrasound and/or elevated CA-125, 19 of whom (14%) were diagnosed with ovarian cancer. In this series, three major complications were reported: sigmoid enterotomy; injury to the inferior vena cava (both repaired laparoscopically); and port site bowel hernia requiring laparotomy and bowel resection. In the series by Dottino et al.,5 of 160 patients with ovarian masses, only four out of the nine patients with invasive ovarian cancer were staged laparoscopically. All eight tumors of LMP were treated with an attempt at laparoscopic management; however, two required

conversion to laparotomy for vascular injuries caused by trocar insertion. In total, there were five operative complications: three vascular injuries; one smallintestinal injury; and one patient who required laparotomy for persistent bleeding. Overall, it appears that immediate surgical outcomes of women managed laparoscopically for ovarian cancer are similar to those of other reported series of operative laparoscopy for benign gynecologic conditions, including a recent retrospective series of 6451 women treated for gynecologic conditions. In this single-institution series, a major complication rate of 0.8% (39/4865) was reported for operative laparoscopy; there was one death, one stomach injury, three major vascular injuries, five ureteral injuries, ten intestinal injuries and 23 bladder injuries.14

PORT SITE METASTASIS A risk specific to the laparoscopic management of ovarian cancer is the potential for seeding of a laparoscopic port site leading to a novel site of disease metastases or recurrence. The first description of such a port site metastasis (PSM) was reported by Dobronte et al. in 1978,15 2 weeks following endoscopy for ovarian cancer. Since that time, approximately 50 cases of ovarian cancer (both invasive and borderline) managed laparoscopically have been described in which PSMs have resulted (Table 12.1).15–31 Although this condition is likely to be underreported in the medical literature, the relative rarity of its occurrence makes it difficult to draw conclusions regarding incidence, risks and outcomes following PSM. Retrospective series have reported an incidence of PSM in 1–16% of all laparoscopic procedures for ovarian cancer.21,25,26` From the reported cases, it appears PSM after laparoscopic diagnosis and staging of ovarian cancer usually occurs in women with advanced disease and ascites and not in those with early-stage disease. In general, all but one case of PSM have been in women with primary stage II, III, or IV disease, or at the time of second-look surgery.

333

Ch 12 final.qxd

12/18/2006

15:33

Page 334

SURGERY FOR OVARIAN CANCER

Table 12.1 Reported cases of port site metastasis in ovarian cancer

Ascites

Time to diagnosis (days)

Author

Cases

Stage

Procedure

Dobronte15

1

III

Laparoscopic biopsy

Yes

14

Stockdale16

1

IV

Laparoscopic biopsy

Yes

8

Hsiu17

2

LMP III

Laparoscopic biopsy

NR

21

Miralles18

1

I

Laparoscopy after open adnexectomy

NR

365

Gleeson19

3

LMP IC III, III

Oophorectomy and morcellation Laparoscopic biopsy

NR Yes

14 < 14

Shepherd20

1

LMP IA

Laparoscopic cyst aspiration and excision

No

42

Childers21

1

IIA

Second-look laparoscopy (positive for persistent disease)

No

56

Kindermann22 14

IC–III

Laparoscopic biopsies, cystectomy, morcellation

NR

8–60

Gungor23

1

IIIB

Third-look laparoscopy

Yes

240

Chu24

1

IC (teratoma) Laparoscopic salpingo-oophorectomy

Yes

14

Kruitwagen25 7

IIIC–IV

Diagnostic laparoscopy

Yes

9–35

van Dam26

9

IIIC–IV

Laparoscopy for tissue diagnosis and assessment of operability

Yes

4–90

Hopkins27

3

LMP apparent Laparoscopic drainage and morcellation, partial excision, early stage oophorectomy (tumor rupture)

NR

14–28

Morice28

5

III–IV

Laparoscopic biopsy

Yes

NA

Haughney29

1

Early stage

Laparoscopic oophorectomy

NR

3650

Carlson30

1

IIIC

Laparotomy for cancer Recurrence in laparoscopic port from surgery for benign disease

Yes

820

Huang31

8

I–IIIC

Laparoscopic biopsy, cystectomy, oophorectomy (tumor rupture)

NR

11–390

Total

60

4–3650

LMP, low malignant potential; NR, not reported

However, at least five cases of apparent stage I tumors of LMP have developed PSM. In each of these cases, it is unclear whether comprehensive staging was undertaken, and in four of the cases, laparoscopic morcellation,27 partial ovarian excision,19 or accidental rupture of the bag containing the ovarian tumor27 may have accounted for the PSM. The mechanism by which PSM or intraperitoneal metastasis occurs after laparoscopy has recently been investigated, and a number of different theories to

334

explain this phenomenon have been proposed.32 Intuitively, the removal of cancerous tissue through the abdominal wall may lead to direct contamination from tumor or instruments. Likewise, tumor cells may be more likely to exist in an aerosolized state during laparoscopy, and may contaminate the surgical sites, especially during decompression of the pneumoperitoneum. Possible factors specific to laparoscopy that may lead to an increased rate of PSM are the effect of increased intra-abdominal pressure with the pneu-

Ch 12 final.qxd

12/18/2006

15:33

Page 335

LAPAROSCOPIC SURGERY

moperitoneum, metabolic and immunologic effects of CO2, alterations in peritoneal humidity, abdominal wall stretching, electrostatic interactions with trocars, and pressure-flow effects related to the insufflation with gas. Investigators have demonstrated that insufflation with CO2 leads to increased peritoneal blood flow,33 with a potential for vascular uptake of cancer cells and PSM or intraperitoneal dissemination. Despite these data regarding increased blood flow during laparoscopy, there are no data that suggest that PSM occurs as a result of hematogenous spread of disease. Investigation of the role of the CO2 pneumoperitoneum in the incidence of PSM was evaluated by randomizing groups of rats to no surgery, laparotomy, or the equivalent of laparoscopy with or without the establishment of pneumoperitoneum following injection of mammary adenocarcinoma cells.34 These authors demonstrated an increased risk of PSM with pneumoperitoneum compared with laparoscopy without pneumoperitoneum, and an increased rate of PSM with laparoscopy compared with laparotomy. Other investigators, however, reported that there was greater tumor growth and wound metastasis in rats undergoing laparotomy, while there was increased tumor dissemination with laparoscopy.35 Research into the molecular mechanisms of PSM has also been undertaken. Hyaluronic acid, which causes adhesion between cancer cells and mesothelial cells, has been implicated in the increased risk for PSM in an animal model after CO2 pneumoperitoneum.36 These researchers also demonstrated that PSM was enhanced in a murine model when mice were treated with hyaluronic acid along with CO2 pneumoperitoneum compared with those undergoing pneumoperitoneum alone.37 Interestingly, little has been reported regarding the prognostic implications of PSM. Two retrospective reports have analyzed the survival related to ovarian cancer in women who had PSM from the laparoscopic management26 and laparoscopic surgery or paracentesis25 for stage III and stage IV disease. Both studies failed to demonstrate a statistically significant difference in survival between the limited numbers of

women with port or paracentesis site metastasis compared with those women without PSM. Owing to the very small sample sizes reported in these studies, it is difficult to draw conclusions regarding the prognostic implications of PSM.

RISK OF RUPTURE AND TUMOR SPILLAGE OF MALIGNANT OVARIAN CYSTS Although not unique to the laparoscopic management of a suspicious ovarian tumor, rupture of an ovarian malignancy may occur more commonly during laparoscopy than with laparotomy due to increased adnexal manipulation and lack of tactile feedback during minimally invasive surgery. Limited retrospective data suggest that ovarian cancer rupture before and, importantly, during surgery may be a negative prognostic factor for survival from disease due to dissemination of malignant cells into the peritoneal cavity. Dembo et al.38 reviewed the records of 519 patients with stage I epithelial ovarian cancer managed with laparotomy. In a multivariate analysis, these authors found that intraoperative cancer rupture did not affect prognosis; however, the majority of these patients did not undergo comprehensive staging. A multivariate analysis of 394 patients with early-stage ovarian carcinoma by Sjovall et al.39 demonstrated that there was no difference in survival between patients whose tumors had intact capsules and patients whose tumors ruptured during laparotomy. On the other hand, a significant reduction in survival was found between patients in whom rupture occurred before surgery compared to a group with intraoperative rupture. Similarly, Sevelda et al.40 found a 76% 5-year survival in groups of 30 patients with stage I ovarian cancer with and without capsular rupture. Again, limited data regarding the degree of comprehensive staging were reported. In contrast, Webb et al.41 demonstrated a statistically significant survival advantage in women with stage I ovarian cancer who did not have capsular rupture compared with those who had intraoperative

335

Ch 12 final.qxd

12/18/2006

15:33

Page 336

SURGERY FOR OVARIAN CANCER

rupture (5-year survival of 78% vs. 56%, respectively). Vergote et al.42 retrospectively reviewed 1545 patients with stage I ovarian cancer in whom ovarian cancer rupture before and during surgery had a significant negative prognostic impact on disease-free survival (hazard ratio 2.65 and 1.64, respectively). Data regarding the risk of laparoscopic rupture of malignant ovarian tumors are less complete. Theoretically, one could argue that laparoscopic rupture could portend a worse prognosis relative to rupture at the time of laparotomy due to seeding of port sites or more diffuse spread of malignancy from the increased intra-abdominal pressure from the pneumoperitoneum. With the limited data describing laparoscopic rupture of ovarian malignancies, Dottino et al.5 described one case of intraperitoneal spillage during a laparoscopic removal of an apparent stage I stromal tumor of the ovary that had a pelvic recurrence 2 years after diagnosis. Childers et al.4 reported a patient with an apparent stage I epithelial ovarian cancer in whom capsular rupture occurred during laparoscopy. This patient developed disease recurrence following chemotherapy and a pathologically negative second-look laparoscopy. Canis et al.3 reported tumor dissemination from one of seven cancers managed laparoscopically. A patient undergoing laparoscopic evaluation for a presumed mature teratoma was treated by unilateral oophorectomy with morcellation. Final histology revealed a grade 1 immature teratoma, and 3 weeks later, diffuse spread of immature implants was discovered. An interesting case was reported by Mayer et al.43 in which peritoneal implants of squamous cell carcinoma were discovered at staging laparotomy following the laparoscopic rupture of a mature teratoma that was found to contain malignant squamous elements on final histology. From these cases, no definitive conclusions can be drawn regarding the rates of cancer rupture at laparoscopy or regarding the prognostic implications of such an outcome. Regardless of the type of procedure that led to the cyst rupture, however, surgeons must be aware of the potential risk of tumor spread following cancer rupture, and the potential implications on prognosis

336

and therefore the possible need for postoperative adjuvant therapy. Various suggestions have been made to minimize the risk of rupture of an ovarian cancer during laparoscopy. Laparotomy should always be performed for suspicious masses of > 10 cm, masses < 10 cm but adherent to the pelvic sidewall, and masses in patients in whom ovarian preservation is desired. In all cases managed laparoscopically, masses should be removed in a commercially available laparoscopic bag through the largest laparoscopy port or through a colpotomy incision to minimize the risk of peritoneal contamination.44 If a cystic ovarian mass is too large to be removed through these routes, the cyst fluid can be drained in the bag prior to its removal. In this situation, a Cook needle can be placed against the cyst wall and connected to suction to create a seal against the cyst wall. A needle is then inserted through the aspirator and the capsule pierced. The aspirator can be placed through a laparoscopic looped ligature prior to drainage so the puncture site can be easily sealed following aspiration.45 If cyst rupture does occur in the peritoneal cavity, the abdomen and pelvis should be thoroughly irrigated.

MANAGEMENT OF OVARIAN CANCER DIAGNOSED DURING LAPAROSCOPY Despite the use of extensive preoperative studies, such as vaginal ultrasonography and tumor markers, it is not always possible to distinguish between benign and malignant ovarian masses before surgical exploration. Even with the most diligent preoperative triage of patients, the gynecologist may encounter unexpected malignancies when undertaking laparoscopic removal of an ovarian mass. Once malignancy is diagnosed during laparoscopy, the surgeon is faced with a decision either to proceed with the appropriate surgery for the ovarian cancer or to abort the procedure in favor of rescheduling the definitive surgery at another time. This decision is best made at the preoperative consultation; if this is not the case,

Ch 12 final.qxd

12/18/2006

15:33

Page 337

LAPAROSCOPIC SURGERY

factors such as the amount of disease, the anticipated likelihood of successful surgical outcome, and the training and experience of the surgeon must influence the surgical plan. If the decision is made to abort the procedure with a plan to reschedule the patient at a later date, the risk of delaying the definitive procedure and potential need for adjuvant therapy must be considered. Although the quantification of this risk is difficult, some series have evaluated the effect of delaying surgery on survival after laparoscopic diagnosis of ovarian cancer. Lehner et al.46 reviewed 48 cases of delayed laparotomy after laparoscopic diagnosis of ovarian cancer. In half of these cases, laparotomy was delayed for more than 17 days and, in the other half, it was performed within 17 days from laparoscopy. Although the sample size is small, the results demonstrate that delay in laparotomy led to a higher chance of more advanced disease at laparotomy. Interestingly, in the 46 cases of ovarian cancer thought to be confined to the ovary at laparoscopy, 27 were found to have more advanced disease at staging laparotomy. Others even suggest that a delay of 7 days or more from laparoscopic diagnosis to definitive surgery could allow for disease progression.22 Intuitively, a plan for therapy of either repeat surgery or adjuvant treatment should be established in a short interval from the time of diagnosis of ovarian cancer to minimize the risk of spread of disease.

LAPAROSCOPY IN THE STAGING OF APPARENT EARLY OVARIAN CANCER Patients with apparent early ovarian cancer must be comprehensively staged, because of the risk of extrapelvic spread in up to 30% of cases thought to be confined to the pelvis. Regardless of whether a patient’s ovarian cancer is managed by laparoscopy or laparotomy, comprehensive staging includes obtaining pelvic cytology, resection of the adnexa(e), infracolic omentectomy, pelvic and para-aortic lymphadenectomy and multiple peritoneal biopsies.47 Likewise, the

serosal and mesenteric surfaces of the large and small intestines must be evaluated. Unfortunately, this portion of a staging surgery may be compromised during laparoscopy, owing to the limited ability to manipulate the intestines fully. The first laparoscopic attempt to stage an ovarian cancer patient was reported by Reich et al.48 in 1990. In this case, the patient did not undergo a para-aortic lymph node dissection. In 1993, Querleu et al.49 described a comprehensive laparoscopic staging procedure. Since that report, multiple series of laparoscopic staging of ovarian cancers have been reported, and are presented in Table 12.2.48–57 Most of these patients had previously undergone incomplete staging of their ovarian cancer and were managed with comprehensive laparoscopic staging. Since none of these 98 cases of laparoscopic staging were compared with staging laparotomy, no conclusions can be drawn regarding the relative rates of complications and outcomes between patients treated with these different techniques. Commonly, laparoscopy has been used in the re-staging of patients who were incidentally discovered to have a malignancy at initial surgery in whom comprehensive staging was not performed. Because repeat laparotomy may increase the total recovery time relative to laparoscopy and potentially delay initiation of adjuvant therapy if required, laparoscopy is an attractive alternative. The Gynecologic Oncology Group (GOG) is currently reviewing those data from a prospective trial of laparoscopic re-staging of ovarian cancer (GOG 9302). Until these data mature, conclusions regarding laparoscopic re-staging of ovarian cancer can only be drawn from limited retrospective series and case reports. Childers et al. reported on 14 patients undergoing laparoscopic staging in whom all procedures were completed successfully without major complications.52 The same year (1995), Pomel et al. reported on an additional ten patients who completed laparoscopic re-staging.53 One hemorrhage was reported requiring laparotomy. Querleu et al.57 recently reported a series of 30 women who underwent staging laparoscopy of an unstaged ovarian tumor of LMP. These authors successfully managed all women

337

Ch 12 final.qxd

12/18/2006

15:33

Page 338

SURGERY FOR OVARIAN CANCER

Table 12.2 Laparoscopic staging of ovarian cancer patients

Author

Cases

OR Time (mean)

PLN (mean)

PALN (mean)

EBL (mean)

Complications IOC POC

LOS (days)

Reich48

1

300

11

NR

200

None

NR

2

Querleu49

2

200

12

9

NR

None

Hematoma

2

Querleu50

8

227

NR

8.6

< 300

None

Hematoma

2.8

Spirtos51

4

193

20.8

7.9

< 100

VI–2

DVT-2 SBO-2

2.7

Childers52

19

180

NR

NR

NR

VI–1

None

1.6

Pomel53

10

313

6

8

NR

None

Bleeding-1 PE-1

4.75

Possover54

13

187

13.4

8.6

< 200

VI–2 GI-1

None

5.6

Dottino55

3

130

11.9

3.7

83.4

VI–1

None

3.6

Scribner56

8

240

18.1

11.9

321.1

NR

NR

2.8

Querleu57

30

165

NR

NR

NR

VI–1

Hematoma

2.7

Total

98

235

10.6

8.6

< 233

8.5%

12.8%

2.85

OR time, median minutes of operation; PLN, median number of pelvic nodes removed; PALN, median number of para-aortic nodes removed; LOS, mean length of hospital stay; IOC, intraoperative complication; POC, postoperative complication; EBL, estimated blood loss; DVT, deep vein thrombosis; GI, gastrointestinal injury; PE, pulmonary embolism; SBO, small-bowel obstruction; VI, vascular injury; NR, not reported

laparoscopically, with a delay between primary and staging surgery of approximately 10 weeks, and a reported incidence of two major complications. From these small series, it can be concluded that re-staging laparoscopy is feasible and safe in selected patients with unstaged ovarian malignancies.

SECOND-LOOK LAPAROSCOPY The purpose of second-look surgery is to assess a patient’s pathologic response to chemotherapy after a complete clinical response (i.e. no disease on examination, on radiography, or by CA-125), and is generally performed only in the setting of a clinical trial. Traditionally, this procedure was performed through a vertical midline incision, with extensive exploration

338

of the abdomen, including peritoneal washings, multiple peritoneal biopsies and often retroperitoneal lymphadenectomy. In current practice, patients may undergo a second surgical procedure after completing adjuvant chemotherapy that would not technically be considered second-look surgery (i.e. for a concurrent surgical condition such as cholelithiasis). In this setting, as well as for classic second-look surgery, laparoscopy may be an alternative to laparotomy. A potential advantage of laparoscopy over laparotomy includes the magnification afforded by the endoscopic equipment. However, the lack of ability to palpate the abdominal and peritoneal surfaces may limit a surgeon’s ability to detect small-volume disease that may be detected at the time of laparotomy. Although the initial literature regarding second-look laparoscopy reported poor visualization, frequent false-negative pathologic results and frequent complications,58–61

Ch 12 final.qxd

12/18/2006

15:33

Page 339

LAPAROSCOPIC SURGERY

more recent literature has reported success. AbuRustum et al.62 and Casey et al.63 found that, compared with laparotomy, laparoscopy was associated with less blood loss, shorter operative time and hospital length of stay, and with fewer intra- and postoperative complications. Importantly, the rates of recurrence following second-look laparotomy and laparoscopy were not different, suggesting reliable pathologic results from the laparoscopic procedure. Husain et al.64 reported 150 cases of second-look laparoscopy in which the rate for conversion to laparotomy was 12%, the majority of them (72%) secondary to the need for cytoreduction that could not be performed laparoscopically. An important study evaluating the accuracy and efficacy of second-look laparoscopy with laparotomy was performed by Clough et al.,65 in which laparotomy was performed immediately following second-look laparoscopy in 20 patients with advanced ovarian cancer. The positive predictive value of laparoscopy for the diagnosis of residual disease was 100%, while the negative predictive value was 86% (2/14 false negatives). Owing to postoperative adhesions, only 41% of the patients had complete evaluation of their peritoneal cavity during laparoscopy. These authors concluded that secondlook surgery via laparoscopy was less reliable than that performed by laparotomy. Evidence regarding comparable survival between second-look surgery performed by laparoscopy and laparotomy comes from Abu-Rustum et al.,66 who compared the outcome of 131 women undergoing second-look laparoscopy with 139 women who underwent the procedure by laparotomy. No significant difference in the overall survival was noted between these groups, suggesting that CO2 pneumoperitoneum does not adversely affect survival in women undergoing second-look surgery. Canis et al.67 proposed the following management for patients with ovarian cancer who need secondlook evaluation. The second look should begin by laparoscopy. If peritoneal carcinomatosis or metastasis is found, the procedure should be stopped and secondary cytoreduction or initiation of treatment for

persistent disease should be started. A negative laparoscopic evaluation can be considered reliable only if (1) the entire peritoneal cavity was inspected; (2) complete adhesiolysis, if required, had been achieved; and (3) multiple random biopsies were taken. If any of the above could not be accomplished, laparotomy should be considered. Patients eligible for second-look surgery usually have had previous extensive cytoreduction followed by chemotherapy that places them at risk for significant adhesions. Therefore, second-look laparoscopy can often be an extremely challenging endoscopic procedure.

LAPAROSCOPY TO DETERMINE THE FEASIBILITY OF CYTOREDUCTION IN PRIMARY OVARIAN CANCER Because of the significant survival advantage afforded by primary optimal cytoreduction for advanced ovarian cancer, advances in surgical techniques to achieve this goal have been developed. In gynecologic oncology centers, it is now possible to achieve a complete tumor resection in 40–90% of the cases. Despite the development of algorithms to predict the ability to achieve optimal cytoreduction (which include disease location, CA-125, volume of ascites and disease volume), it is often difficult to determine prospectively which patients will be left with small-volume residual disease and which will have bulky disease remaining after primary cytoreduction. For this reason, laparoscopy has been suggested to evaluate the potential for optimal cytoreduction prior to laparotomy. Vergote et al.68 described 77 patients with clinical and radiological findings predictive of unresectable advanced ovarian carcinoma by imaging criteria who underwent open laparoscopy to evaluate the ability for cytoreduction. Laparoscopy was demonstrated to be safe, and primary optimal cytoreduction was accomplished in 79% of the patients who were subjected to laparotomy after their laparoscopic assessment. Thus, laparoscopy serves as a reasonable mechanism to assess resectability in a patient with

339

Ch 12 final.qxd

12/18/2006

15:33

Page 340

SURGERY FOR OVARIAN CANCER

advanced ovarian cancer prior to laparotomy. In patients in whom optimal cytoreduction is not thought to be feasible, a biopsy can be performed to confirm the clinical impression of ovarian cancer, and neoadjuvant chemotherapy can be initiated quickly after surgery. Recently, hand-assisted laparoscopic surgery (HALS) has been described for the management of malignancies of the colon, kidney, liver and spleen. This technique utilizes an airtight port the size of a glove, into which the surgeon inserts the nondominant hand to combine the minimally invasive advantages of laparoscopy with the tactile advantage of laparotomy. HALS is a developing technique that has been described to lead to a decreased rate of conversion to laparotomy due to inability to complete a procedure in patients with colon and kidney cancer. It is anticipated that this technique may expand the utility of laparoscopy for primary or secondary cytoreduction in advanced or recurrent ovarian cancer.

position in Allen stirrups (minimizing the flexion at the hips) to maximize the surgical technique. Both arms are tucked at the patient’s side, and a naso- or orogastric tube is placed on suction to minimize gastric distension common with pre-oxygenation prior to endotracheal intubation (Figure 12.1). Four trocar sites are typically used for this procedure. We prefer placing 12-mm adaptable trocars in the midline and 5-mm ports laterally. The largerdiameter port being in the suprapubic midline allows for larger specimens (e.g. omental biopsy and lymph nodes) to be easily removed from the peritoneal cavity. We begin by initially placing a 12-mm port in the periumbilical region (either supra- or infraumbilical depending on patient factors); this port is generally used as the camera port. This port is always placed using an open technique to minimize the risk of vascular injury. Another 12-mm port is then placed approximately 2–3 cm above the symphysis pubis in

LAPAROSCOPIC SURGICAL TECHNIQUES

Monitor

A successful operation must begin prior to the initiation of the procedure. Careful patient selection, familiarity and accessibility of appropriate instrumentation, education of assistants and the experience of the surgeon are all factors that can be modified to improve surgical outcomes. This section briefly describes the surgical procedures related to laparoscopic management of ovarian cancer.

Knees flat

Patient positioning, incisions and port placement The patient undergoing operative laparoscopy is prepared identically to a patient undergoing laparotomy for ovarian cancer. Our preparation includes clear liquid diet for 24 h before surgery, a full mechanical bowel preparation, oral antibiotics, enemas the morning of surgery and prophylactic intravenous antibiotics immediately prior to entry into the operating suite. The patient is placed in the modified ski

340

Allen stirrups Ulnar padding 30° Trendelenburg

Figure 12.1 Appropriate patient positioning in the operating suite for laparoscopy. Minimization of the risk of neurologic injury is accomplished through positioning and padding of the upper and lower extremities

Ch 12 final.qxd

12/18/2006

15:33

Page 341

LAPAROSCOPIC SURGERY

the midline, and the 5-mm ports are placed equidistant from the midline ports just lateral to the inferior epigastric blood vessels (which can be identified laparoscopically as they travel through the lateral umbilical fold of the abdominal wall). All ports are then secured to the abdominal wall skin with suture to prevent them from being inadvertently pulled from the peritoneal cavity (Figure 12.2). Following port placement, the peritoneal cavity is thoroughly inspected through the umbilical port to look for the presence of extrapelvic disease, and pelvic washings are obtained. The patient is then placed in steep Trendelenburg position with the operating table at the

lowest height. To maximize exposure, the intestines and omentum are placed in the upper abdomen by inspecting the peritoneum through the suprapubic port while using graspers to manipulate these organs through the other ports.

Laparoscopic oophorectomy Usually, the oophorectomy or salpingo-oophorectomy will be the initial procedure after obtaining the pelvic washings, so that a frozen section diagnosis can be made prior to further staging. This can generally be accomplished by initially dividing the round ligament with the electrosurgical unit or harmonic scalpel and developing the pararectal space, as described in the section on laparoscopic pelvic lymphadenectomy. When the lateral peritoneum is incised, the medial leaf of the broad ligament is tented up to maximize the visualization of the ureter. When the ureter is identified, the ovarian blood vessels in the infundibulopelvic ligament can be divided using a stapler, harmonic scalpel, or electrosurgical unit. This ligament is then skeletonized above the ureter to the uteroovarian ligament, which is then transected similarly to the infundibulopelvic ligament. The ovary and fallopian tube can be placed in a bag and removed through the 12-mm port.

Transperitoneal laparoscopic aortic lymphadenectomy

Figure 12.2 Trocar sites for laparoscopic staging, including pelvic and aortic lymphadenectomy. A supraumbilical (demonstrated) or infraumbilical (not shown) incision can be made for the camera port depending on the patient’s anatomy

In general, we begin the laparoscopic lymphadenectomy with the para-aortic lymph nodes. To perform the right para-aortic lymphadenectomy, the operating surgeon stands on the patient’s left side at the level of the hips and uses endoscopic scissors and graspers in the suprapubic and left lateral port. Monitors are generally placed at the level of the patient’s shoulder. We prefer the camera in the umbilical port rather than in the suprapubic port (Figure 12.3). The peritoneum is opened cephalad to the right ureter along the right common iliac artery. The peritoneal dissection continues cephalad to the base of the aorta, and continues along this plane to the duodenum. To perform

341

Ch 12 final.qxd

12/18/2006

15:33

Page 342

SURGERY FOR OVARIAN CANCER

Right ovarian vein

Figure 12.3 Surgeon

positioning

for

the

right

aortic

From right lateral port

Ureter

lymphadenectomy. The primary surgeon is on the patient’s left,

From left lateral port

with the monitor adjacent to the patient’s right shoulder.

From suprapubic port

Laparoscopic graspers are used through the suprapubic trocar to facilitate removal of the lymph nodes through the 12-mm port

Figure 12.4 Laparoscopic view of the right common iliac and aortic dissection. The surgical assistant, standing on the

the right aortic lymphadenectomy, the assistant surgeon retracts the right ureter laterally out of the field of dissection. The operating surgeon then dissects the lymph nodes from the base of the inferior vena cava at the right common iliac artery to the level of the renal vein. Dissection continues from medial to right lateral. Care must be taken to isolate and divide with monopolar cautery small perforating blood vessels that come off the vena cava to prevent hemorrhage that may require laparotomy for repair (Figure 12.4). The lymph nodes are always removed through the 12mm suprapubic port to minimize nodal fracture. The left para-aortic lymphadenectomy is performed by retracting the mesentery of the sigmoid colon anteriorly and dissecting the areolar tissue between the left common iliac artery and mesentery

342

patient’s right, reflects the right ureter laterally out of the operative field

just below the inferior mesenteric artery (IMA) until the left psoas muscle is identified. In this tissue, the left ureter should be identified to ensure that it is out of the operative field. Once this dissection is completed, the assistant surgeon maintains lateral retraction on this tissue. The surgeon then grasps the nodal bundle along either the aorta or the left common iliac artery and lifts while depressing the vascular structures with the scissors (Figure 12.5). The cephalad and caudad extent of the left aortic lymph nodes are then isolated, transected with the electrosurgical unit and removed through the 12-mm port.

Ch 12 final.qxd

12/18/2006

15:33

Page 343

LAPAROSCOPIC SURGERY

Inferior mesenteric artery

Left ovarian vein Inferior mesenteric artery Left para-aortic lymph nodes

Right ovarian vein

Interaortocaval lymphatic and fatty tissue

Reflected mesentery of sigmoid colon

Right ovarian vein

From left lateral port

From right lateral port From suprapubic port

Left proximal common iliac lymph nodes

From right lateral port

Position of umbilicus

From left lateral port

Figure 12.5 Laparoscopic view of the left common iliac and

Figure 12.6 Resection of the high aortic lymph nodes. The

inframesenteric aortic dissection. Again, the surgical assistant is

duodenum has been reflected cephalad out of the surgical field.

on the patient’s ipsilateral side and retracts the ureter laterally

This dissection may be facilitated by the surgeon standing

out of the operative field

between the patient’s legs

When necessary, the aortic lymph nodes cephalad to the IMA can be removed, although exposure is often limited. The peritoneum between the IMA and duodenum is incised, and the renal veins and ovarian blood vessels are exposed. The assistant maintains the ureters laterally out of the field of dissection. Often, the ovarian vessels will need to be clipped or cauterized for adequate performance of the lymphadenectomy (Figure 12.6). Depending on the relative risks and benefits, the IMA may need to be ligated at its base to enhance exposure of the high left aortic lymph nodes.

Transperitoneal laparoscopic pelvic lymphadenectomy The pelvic lymphadenectomy is also performed with the operating surgeon on the contralateral side of the table from that of the pelvic dissection. The monitors are generally placed at the level of the patient’s knees, simulating the operating position during an open pelvic lymphadenectomy (Figure 12.7). The camera, again through the umbilical port, can be angled in such a fashion that the external iliac blood vessels are visualized horizontally, again similar to the view during an open pelvic lymphadenectomy. The

343

Ch 12 final.qxd

12/18/2006

15:33

Page 344

SURGERY FOR OVARIAN CANCER

Assistant

Surgeon

Monitor 1 Monitor 2

Figure 12.7 Surgeon positioning for the right pelvic lymphadenectomy, with the primary surgeon on the patient’s left. The monitors are adjacent to the patient’s knees. Again, the graspers are placed through the suprapubic trocar to facilitate removal of the lymph nodes through the 12-mm port

peritoneum lateral to the iliac vessels is then opened along the psoas muscle and parallel to the ovarian blood vessels. Often, the infundibulopelvic ligament will have already been divided during the initial salpingo-oophorectomy. Any adhesions of the cecum and sigmoid colon are divided, and these structures are retracted further cephalad. The pararectal and paravesical spaces are then opened. The ureter is identified and retracted medially by the assistant surgeon. The lymph nodes are removed by dissecting the lateral nodal tissue away from the psoas muscle, using caution to avoid injury to the genitofemoral nerve. As the dissection continues caudally, the assistant retracts in the paravesical space until the circumflex iliac vein is identified (Figure 12.8). The electrosurgical unit is used to transect the specimen at this level. The obturator space is then developed beneath the

344

external iliac vein until the obturator nerve is identified. The assistant retracts under the vein to maximize the exposure of small perforating vessels that should be divided only after the nerve is identified (Figure 12.9). If the round ligament had not been transected as part of a salpingo-oophorectomy, it is divided at this time.

Extraperitoneal laparoscopic aortic lymphadenectomy An attempt at further refinement of laparoscopic lymphadenectomy has led to the evaluation of an extraperitoneal approach. Following the initial description of the technique for extraperitoneal laparoscopic lymphadenectomy by Vasilev and McGonigle,69 other investigators described and

Ch 12 final.qxd

12/18/2006

15:33

Page 345

LAPAROSCOPIC SURGERY

Medical retraction of specimen

Accessory obturator vein

External iliac vein

Circumflex iliac vein

Obturator nerve

External iliac artery Psoas muscle External iliac vein

Figure 12.9 Laparoscopic view of the right obturator lymphadenectomy. In this view, the assistant surgeon reflects Figure 12.8 Laparoscopic view of right pelvic lymphadenec-

the external iliac vein laterally to expose the obturator nerve

tomy. The assistant reflects the external iliac artery laterally to facilitate resection of the external iliac lymph nodes. As the dissection continues towards the common iliac artery, the assistant will reflect the ureter cephalad out of the operative field

refined the technique.70,71 Occelli et al. have demonstrated that extraperitoneal aortic lymphadenectomy resulted in significantly fewer adhesions than the transperitoneal laparoscopy in a porcine model.72 No randomized studies have been performed to compare the different approaches of para-aortic lymph node dissection, and it is therefore difficult to compare the results of the different studies. The mean number of nodes retrieved by the three techniques (transperitoneal, bilateral and left extraperitoneal) is similar,69 and is comparable to or even higher than that in laparotomy.73,74 Importantly, the incidence of metastatic para-aortic lymph nodes was similar in the different studies: 18–22%.71,73,75

Laparoscopic staging The remainder of the laparoscopic staging procedure can be performed after lymphadenectomy. Peritoneal biopsies are taken of the anterior and posterior culde-sac, the pelvic sidewalls and the pelvic gutters. These biopsies are easily obtained with endoscopic scissors and graspers. To biopsy the diaphragms, often the camera will need to be placed in the suprapubic port and long instruments placed through the umbilical port to reach the peritoneum in this location. An omental sampling is performed similar to that done during laparotomy; avascular windows are created sharply in the infracolic omentum. Vascular pedicles are then divided with the electrosurgical unit, clips, or a harmonic scalpel. The omentum can be placed in a bag and removed through a 12-mm port. In cases in which an appendectomy is necessary, the cecum is grasped atraumatically and elevated towards the liver. The appendix is then grasped, and a window between

345

Ch 12 final.qxd

12/18/2006

15:33

Page 346

SURGERY FOR OVARIAN CANCER

it and the mesoappendix is created close to the cecum. The appendix is transected with a gastrointestinal anastomotic (GIA) stapler, followed by transaction of the mesoappendix in similar fashion.

Complications of advanced operative laparoscopy In counseling patients regarding the route of operation to complete surgery for ovarian cancer, one consideration must be the expected rate of complications. Although there are few large series reporting the complications of specific advanced laparoscopic procedures in patients with ovarian cancer, data from other series of operative laparoscopy allow for some conclusions to be drawn. In 1995, the American Association of Gynecologic Laparoscopists (AAGL) reported on over 45 000 cases of operative laparoscopy performed in 1993.76 The rate of fatal complications was reported to be 6.7/100 000 procedures, essentially stable from the series reported of cases in 1991. However, the rate of significant complications (injury to the intestinal or urinary system, unintended laparotomy, hemorrhage, or transfusion) increased during this interval, probably due to the increased complexity of cases being performed laparoscopically. Specific complications during operative laparoscopy can be divided into those that are general, vascular, intestinal, urinary, neurologic and incisional. Injury to the epigastric blood vessels occurs commonly, with some series estimating more than 2% of cases. Major bleeding can be controlled with direct suturing through the abdominal wall under laparoscopic visualization, or by placing a large urethral catheter through the abdominal wall and overinflating the balloon to tamponade the hemorrhage. Fortunately, major vascular injury is rare, occurring in less than 0.1% of cases. Knowledge of pelvic anatomy is imperative in preventing vascular injury. The use of open laparoscopy as described above will further minimize the risk of major vascular injury resulting from blind insertion of a laparoscopic trocar or Veress needle. Injury to the intestines has been reported in 0.05–3% of cases of major operative laparoscopy, and would be expected to be even high-

346

er in patients undergoing repeat laparoscopic surgery for ovarian cancer. Open laparoscopy does not decrease the risk of injury to the intestinal tract, but placement of the initial trocar in a location that has not previously been operated upon (such as the left upper quadrant) may minimize this risk. Electrical injury from monopolar current may occur from direct application of current to tissue, or as a result of lateral spread or inadvertent conduction. Because the amount of tissue damage from the electrosurgical unit can exceed that which is clinically apparent, intestinal resection is usually recommended for intestinal injury from cautery. Laparoscopic bladder injuries are relatively rare, and occur in less than 1% of cases. Complete bladder drainage and insertion of the suprapubic port under direct visualization will minimize risk. Small injuries can be managed by continuous bladder drainage, although a large cystotomy should be repaired. Fortunately, injury to the ureter is less common, and occurs in less than 0.1% of cases.77 It is commonly injured in similar locations to those that occur during laparotomy, mainly at the pelvic brim and at the level of the endocervix. During lymphadenectomy, the ureter should be maintained out of the surgical field to minimize direct or indirect injury. Besides nerve palsies of the upper or lower extremities resulting from inappropriate positioning or pressure during laparoscopy, neurologic injury is rare during laparoscopy. Transection of the genitofemoral or obturator nerves is the most common major injury; prevention of such injuries is accomplished through the identification and retraction of these nerves out of the surgical field. Incisional hernias are extremely rare when trocar diameters are 5 mm or less; therefore, closing the fascia through a defect from such a trocar is not necessary. Intestinal herniation through a 10- or 12-mm port occurs rarely, and has been reported in between 0.2 and 2% of cases, and occurs most commonly in the region of the umbilical port.78 Care should be taken in closing the fascia beneath these larger ports to prevent incisional complications.

Ch 12 final.qxd

12/18/2006

15:33

Page 347

LAPAROSCOPIC SURGERY

CONCLUSION The clinical applications of laparoscopy in the management of women with ovarian cancer continue to evolve. In general gynecologic surgery, laparoscopy has been shown to lead to a decreased length of hospital stay and a relatively shorter recovery, often with decreased postoperative pain. Evaluation of an adnexal mass, re-staging following inadequate initial surgery, and reassessment surgery following primary therapy are procedures that appear to be well suited to the laparoscopic approach. As more data regarding the safety and efficacy of minimally invasive surgery for ovarian cancer become available, the frequency with which it is incorporated into clinical practice will probably increase. The surgeon operating on women with known or suspected ovarian cancer should have a thorough understanding of the potential risks, benefits and specific techniques of laparoscopic surgery to ensure appropriate case selection and optimization of patient outcome.

REFERENCES 1.

2.

3.

4.

5.

Chi DS, Curtin JP, Barakat RR. Laparoscopic management of adnexal masses in women with a history of nongynecologic malignancy. Obstet Gynecol 1995; 86: 964–8 Canis M, Mage G, Pouly JL, et al. Laparoscopic diagnosis of adnexal cystic masses: a 12-year experience with long-term follow-up. Obstet Gynecol 1994; 83: 707–12 Canis M, Pouly JL, Wattiez A, et al. Laparoscopic management of adnexal masses suspicious at ultrasound. Obstet Gynecol 1997; 89: 679–83 Childers JM, Nasseri A, Surwit EA. Laparoscopic management of suspicious adnexal masses. Am J Obstet Gynecol 1996; 175: 1451–7 Dottino PR, Levine DA, Ripley DL, Cohen CJ. Laparoscopic management of adnexal masses in premenopausal and postmenopausal women. Obstet Gynecol 1999; 93: 223–8

6.

Hulka JF, Parker WH, Surrey MW, Phillips JM. Management of ovarian masses. AAGL 1990 survey. J Reprod Med 1992; 37: 599–602

7.

Mage G, Canis M, Manhes H, et al. Laparoscopic management of adnexal cystic masses. J Gynecol Surg 1990; 6: 71–9

8.

Mecke H, Lehmann-Willenbrock E, Ibrahim M, Semm K. Pelviscopic treatment of ovarian cysts in premenopausal women. Gynecol Obstet Invest 1992; 34: 36–42

9.

Nezhat F, Nezhat C, Welander CE, Benigno B. Four ovarian cancers diagnosed during laparoscopic management of 1011 women with adnexal masses. Am J Obstet Gynecol 1992; 167: 790–6

10.

Parker WH, Berek JS. Management of selected cystic adnexal masses in postmenopausal women by operative laparoscopy: a pilot study. Am J Obstet Gynecol 1990; 163: 1574–7

11.

Wenzl R, Lehner R, Husslein P, Sevelda P. Laparoscopic surgery in cases of ovarian malignancies: an Austria-wide survey. Gynecol Oncol 1996; 63: 57–61

12.

Hidlebaugh DA, Vulgaropulos S, Orr RK. Treating adnexal masses. Operative laparoscopy vs. laparotomy. J Reprod Med 1997; 42: 551–8

13.

Mais V, Ajossa S, Piras B, et al. Treatment of nonendometriotic benign adnexal cysts: a randomized comparison of laparoscopy and laparotomy. Obstet Gynecol 1995; 86: 770–4

14.

Wang PH, Lee WL, Yuan CC, et al. Major complications of operative and diagnostic laparoscopy for gynecologic disease. J Am Assoc Gynecol Laparosc 2001; 8: 68–73

15.

Dobronte Z, Wittmann T, Karacsony G. Rapid development of malignant metastases in the abdominal wall after laparoscopy. Endoscopy 1978; 10: 127–30

16.

Stockdale AD, Pocock TJ. Abdominal wall metastasis following laparoscopy: a case report. Eur J Surg Oncol 1985; 11: 373–5

17.

Hsiu JG, Given FT Jr, Kemp GM. Tumor implantation after diagnostic laparoscopic biopsy of serous ovarian tumors of low malignant potential. Obstet Gynecol 1986; 68: 90–3

18.

Miralles RM, Petit J, Gine L, Balaguero L. Metastatic cancer spread at the laparoscopic puncture site.

347

Ch 12 final.qxd

12/18/2006

15:33

Page 348

SURGERY FOR OVARIAN CANCER

Report of a case in a patient with carcinoma of the ovary. Case report. Eur J Gynaecol Oncol 1989; 10: 442–4

laparoscopic surgery for benign disease. Gynecol Oncol 2002; 85: 529–31 31.

Huang KG, Wang CJ, Chang TC, et al. Management of port-site metastasis after laparoscopic surgery for ovarian cancer. Am J Obstet Gynecol 2003; 189: 16–21

32.

Neuhaus SJ, Watson DI, Ellis T, et al. Wound metastasis after laparoscopy with different insufflation gases. Surgery 1998; 123: 579–83

33.

Brundell SM, Tsopelas C, Chatterton B, et al. Experimental study of peritoneal blood flow and insufflation pressure during laparoscopy. Br J Surg 2002; 89: 617–22

Kindermann G, Maassen V, Kuhn W. Laparoscopic preliminary surgery of ovarian malignancies. Experiences from 127 German gynecologic clinics. Geburtshilfe Frauenheilkd 1995; 55: 687–94

34.

Hopkins MP, Dulai RM, Occhino A, Holda S. The effects of carbon dioxide pneumoperitoneum on seeding of tumor in port sites in a rat model. Am J Obstet Gynecol 1999; 181: 1329–33

23.

Gungor M, Cengiz B, Turan YH, Ortac F. Implantation metastasis of ovarian cancer after thirdlook laparoscopy. J Pak Med Assoc 1996; 46: 111–12

35.

24.

Chu HS, Jung NW, Kim JH, et al. Tumor implantation along abdominal trocar site after pelviscopic removal of malignant ovarian tumor – a case report. J Korean Med Sci 1996; 11: 440–3

Canis M, Botchorishvili R, Wattiez A, et al. Tumor growth and dissemination after laparotomy and CO2 pneumoperitoneum: a rat ovarian cancer model. Obstet Gynecol 1998; 92: 104–8

36.

Yamaguchi K, Hirabayashi Y, Suematsu T, et al. Hyaluronic acid secretion during carbon dioxide pneumoperitoneum and its association with port-site metastasis in a murine model. Surg Endosc 2001; 15: 59–62

37.

Yamaguchi K, Hirabayashi Y, Shiromizu A, et al. Enhancement of port site metastasis by hyaluronic acid under CO2 pneumoperitoneum in a murine model. Surg Endosc 2001; 15: 504–7

38.

Dembo AJ, Davy M, Stenwig AE, et al. Prognostic factors in patients with stage I epithelial ovarian cancer. Obstet Gynecol 1990; 75: 263–73

39.

Sjovall K, Nilsson B, Einhorn N. Different types of rupture of the tumor capsule and the impact on survival in early ovarian carcinoma. Int J Gynecol Cancer 1994; 4: 333–6

40.

Sevelda P, Dittrich C, Salzer H. Prognostic value of the rupture of the capsule in stage I epithelial ovarian carcinoma. Gynecol Oncol 1989; 35: 321–2

41.

Webb MJ, Decker DG, Mussey E, Williams TJ. Factor influencing survival in Stage I ovarian cancer. Am J Obstet Gynecol 1973; 116: 222–8

42.

Vergote I, De Brabanter J, Fyles A, et al. Prognostic importance of degree of differentiation and cyst rup-

19.

Gleeson NC, Nicosia SV, Mark JE, et al. Abdominal wall metastases from ovarian cancer after laparoscopy. Am J Obstet Gynecol 1993; 169: 522–3

20.

Shepherd JH, Carter PG, Lowe DG. Wound recurrence by implantation of a borderline ovarian tumour following laparoscopic removal. Br J Obstet Gynaecol 1994; 101: 265–6

21.

Childers JM, Aqua KA, Surwit EA, et al. Abdominalwall tumor implantation after laparoscopy for malignant conditions. Obstet Gynecol 1994; 84: 765–9

22.

25.

26.

Kruitwagen RF, Swinkels BM, Keyser KG, et al. Incidence and effect on survival of abdominal wall metastases at trocar or puncture sites following laparoscopy or paracentesis in women with ovarian cancer. Gynecol Oncol 1996; 60: 233–7 van Dam PA, DeCloedt J, Tjalma WA, et al. Trocar implantation metastasis after laparoscopy in patients with advanced ovarian cancer: can the risk be reduced? Am J Obstet Gynecol 1999; 181: 536–41

27.

Hopkins MP, von Gruenigen V, Gaich S. Laparoscopic port site implantation with ovarian cancer. Am J Obstet Gynecol 2000; 182: 735–6

28.

Morice P, Viala J, Pautier P, et al. Port-site metastasis after laparoscopic surgery for gynecologic cancer. A report of six cases. J Reprod Med 2000; 45: 837–40

29.

30.

348

Haughney RV, Slade RJ, Brain AN. An isolated abdominal wall metastasis of ovarian carcinoma ten years after primary surgery. Eur J Gynaecol Oncol 2001; 22: 102–3 Carlson NL, Krivak TC, Winter WE 3rd, Macri CI. Port site metastasis of ovarian carcinoma remote from

Ch 12 final.qxd

12/18/2006

15:33

Page 349

LAPAROSCOPIC SURGERY

ture in stage I invasive epithelial ovarian carcinoma. Lancet 2001; 357: 176–82 43.

44.

55.

Dottino PR, Tobias DH, Beddoe A, et al. Laparoscopic lymphadenectomy for gynecologic malignancies. Gynecol Oncol 1999; 73: 383–8

56.

Scribner DR Jr, Walker JL, Johnson GA, et al. Laparoscopic pelvic and paraaortic lymph node dissection: analysis of the first 100 cases. Gynecol Oncol 2001; 82: 498–503

57.

Querleu D, Papageorgiou T, Lambaudie E, et al. Laparoscopic restaging of borderline ovarian tumours: results of 30 cases initially presumed as stage IA borderline ovarian tumours. Br J Obstet Gynecol 2003; 110: 201–4

58.

Ozols RF, Fisher RI, Anderson T, et al. Peritoneoscopy in the management of ovarian cancer. Am J Obstet Gynecol 1981; 140: 611–19

59.

Staging announcement: FIGO cancer committee. Gynecol Oncol 1986; 25: 383

Quinn MA, Bishop GJ, Campbell JJ, et al. Laparoscopic follow-up of patients with ovarian carcinoma. Br J Obstet Gynaecol 1980; 87: 1132–9

60.

Reich H, McGlynn F, Wilkie W. Laparoscopic management of stage I ovarian cancer. A case report. J Reprod Med 1990; 35: 601–4

Berek JS, Griffiths CT, Leventhal JM. Laparoscopy for second-look evaluation in ovarian cancer. Obstet Gynecol 1981; 58: 192–8

61.

Gadducci A, Sartori E, Maggino T, et al. Analysis of failures after negative second-look in patients with advanced ovarian cancer: an Italian multicenter study. Gynecol Oncol 1998; 68: 150–5

62.

Abu-Rustum NR, Barakat RR, Siegel PL, et al. Second-look operation for epithelial ovarian cancer: laparoscopy or laparotomy? Obstet Gynecol 1996; 88: 549–53

63.

Casey AC, Farias-Eisner R, Pisani AL, et al. What is the role of reassessment laparoscopy in the management of gynecologic cancers in 1995? Gynecol Oncol 1996; 60: 454–61

64.

Husain A, Chi DS, Prasad M, et al. The role of laparoscopy in second-look evaluations for ovarian cancer. Gynecol Oncol 2001; 80: 44–7

65.

Clough KB, Ladonne JM, Nos C, et al. Second look for ovarian cancer: laparoscopy or laparotomy? A prospective comparative study. Gynecol Oncol 1999; 72: 411–17

66.

Abu-Rustum NR, Sonoda Y, Chi DS, et al. The effects of CO2 pneumoperitoneum on the survival of women with persistent metastatic ovarian cancer. Gynecol Oncol 2003; 90: 431–4

Mayer C, Miller DM, Ehlen TG. Peritoneal implantation of squamous cell carcinoma following rupture of a dermoid cyst during laparoscopic removal. Gynecol Oncol 2002; 84: 180–3 McCartney AJ, Johnson N. Using a vaginal tube to exteriorize lymph nodes during a laparoscopic pelvic lymphadenectomy. Gynecol Oncol 1995; 57: 304–6

45.

Boike GM, Graham JE. Ovarian cancer surgery by operative laparoscopy. Op Tech Gynecol Surg 1997; 2: 171–9

46.

Lehner R, Wenzl R, Heinzl H, et al. Influence of delayed staging laparotomy after laparoscopic removal of ovarian masses later found malignant. Obstet Gynecol 1998; 92: 967–71

47. 48.

49.

Querleu D. Laparoscopic paraaortic node sampling in gynecologic oncology: a preliminary experience. Gynecol Oncol 1993; 49: 24–9

50.

Querleu D, LeBlanc E. Laparoscopic infrarenal paraaortic lymph node dissection for restaging of carcinoma of the ovary or fallopian tube. Cancer 1994; 73: 1467–71

51.

Spirtos NM, Schlaerth JB, Spirtos TW, et al. Laparoscopic bilateral pelvic and paraaortic lymph node sampling: an evolving technique. Am J Obstet Gynecol 1995; 173: 105–11

52.

Childers JM, Lang J, Surwit EA, Hatch KD. Laparoscopic surgical staging of ovarian cancer. Gynecol Oncol 1995; 59: 25–33

53.

Pomel C, Provencher D, Dauplat J, et al. Laparoscopic staging of early ovarian cancer. Gynecol Oncol 1995; 58: 301–6

54.

Possover M, Krause N, Plaul K, et al. Laparoscopic para-aortic and pelvic lymphadenectomy: experience with 150 patients and review of the literature. Gynecol Oncol 1998; 71: 19–28

349

Ch 12 final.qxd

12/18/2006

15:33

Page 350

SURGERY FOR OVARIAN CANCER

67.

Canis M, Botchorishvili R, Manhes H, et al. Management of adnexal masses: role and risk of laparoscopy. Semin Surg Oncol 2000; 19: 28–35

73.

Buchsbaum HJ. Extrapelvic lymph node metastases in cervical carcinoma. Am J Obstet Gynecol 1979; 133: 814–24

68.

Vergote I, De Wever I, Tjalma W, et al. Neoadjuvant chemotherapy or primary debulking surgery in advanced ovarian carcinoma: a retrospective analysis of 285 patients. Gynecol Oncol 1998; 71: 431–6

74.

Finan MA, DeCesare S, Fiorica JV, et al. Radical hysterectomy for stage IB1 vs IB2 carcinoma of the cervix: does the new staging system predict morbidity and survival? Gynecol Oncol 1996; 62: 139–47

69.

Vasilev SA, McGonigle KF. Extraperitoneal laparoscopic para-aortic lymph node dissection. Gynecol Oncol 1996; 61: 315–20

75.

70.

Dargent D, Ansquer Y, Mathevet P. Technical development and results of left extraperitoneal laparoscopic paraaortic lymphadenectomy for cervical cancer. Gynecol Oncol 2000; 77: 87–92

Vergote I, Amant F, Berteloot P, Van Gramberen M. Laparoscopic lower para-aortic staging lymphadenectomy in stage IB2, II, and III cervical cancer. Int J Gynecol Cancer 2002; 12: 22–6

76.

Querleu D, Dargent D, Ansquer Y, et al. Extraperitoneal endosurgical aortic and common iliac dissection in the staging of bulky or advanced cervical carcinomas. Cancer 2000; 88: 1883–91

Hulka J, Peterson HB, Phillips JM, Surrey MW. Operative laparoscopy: American Association of Gynecologic Laparoscopists’ 1993 membership survey. J Am Assoc Gynecol Laparosc 1995; 2: 133–6

77.

Grainger DA, Soderstrom RM, Schiff SF, et al. Ureteral injuries at laparoscopy: insights into diagnosis, management, and prevention. Obstet Gynecol 1990; 75: 839–43

78.

Montz FJ, Holschneider CH, Munro MG. Incisional hernia following laparoscopy: a survey of the American Association of Gynecologic Laparoscopists. Obstet Gynecol 1994; 84: 881–4

71.

72.

350

Occelli B, Narducci F, Lanvin D, et al. De novo adhesions with extraperitoneal endosurgical para-aortic lymphadenectomy versus transperitoneal laparoscopic para-aortic lymphadenectomy: a randomized experimental study. Am J Obstet Gynecol 2000; 183: 529–33

Ch 13 final.qxd

12/18/2006

15:34

Page 351

CHAPTER 13

Palliative surgery for ovarian cancer Laura J Havrilesky, Daniel L Clarke-Pearson

INTRODUCTION The majority of patients with advanced ovarian cancer will undergo recurrence and ultimately succumb to their disease. Therapies such as secondary cytoreduction and salvage chemotherapeutics may prolong life but are almost never curative. Recurrent ovarian cancer usually presents as intraperitoneal disease, which can result in symptomatic ascites, carcinomatous ileus, or bowel obstruction. Thus, many patients with recurrent disease will experience distressing symptoms of pain, abdominal distension, nausea, vomiting and dyspnea. The surgeon caring for ovarian cancer patients is likewise faced with difficult decisions regarding potentially palliative procedures that may improve quality of life but which also carry the potential to cause significant morbidity and mortality. This chapter will outline the diagnostic evaluation and surgical management of patients with symptomatic advanced ovarian cancer.

SMALL-BOWEL OBSTRUCTION Intestinal obstruction is a common problem among women with progressive ovarian cancer, occurring clinically in up to one-third of patients1 and found at autopsy in 51% of cases2. Many patients presenting with bowel obstruction have previously been treated with multiple chemotherapy regimens or radiotherapy and have limited therapeutic options. Survival of patients with malignant bowel obstruction in the

absence of any intervention is usually less than 3 months.1,3 The differential diagnosis for patients with advanced ovarian cancer who present with persistent nausea and vomiting includes gastroenteritis, carcinomatous ileus and mechanical bowel obstruction. Gastroenteritis may be of infectious etiology or secondary to prior chemotherapy or radiation treatments. Carcinomatous ileus is a non-obstructive condition characterized by the lack of efficient peristalsis due to carcinomatosis coating the bowel serosa and mesentery. Mechanical bowel obstruction in patients with progressive ovarian cancer is usually due to malignancy, but a ‘benign’ etiology, such as adhesions or radiation stricture, is the cause in 9–23% of cases.1,4,5 Risk factors for bowel obstruction include multiple prior operations and prior radiotherapy. The decision for surgical intervention must take into account the likely findings at surgery and probability of successful intervention. Multiple retrospective series are available to aid in the determination of likely outcomes. At laparotomy, the site of obstruction is the small bowel in 44–61% of cases, large bowel in 18–33% and a combination of small and large bowel in 9–22%.1,3,4,6,7 Procedures most often performed to alleviate obstruction are bypass procedures in 22–63%, resection of bowel with anastomosis in 12–25%, colostomy in 8–33%, ileostomy in 4–15% and lysis of adhesions in 15%.1,3–5,7–9 Inoperable disease preventing a definitive operative procedure is found in up to 20% of patients undergoing exploratory laparotomy for ovarian cancer-associated bowel obstruction.3–5,7,8

351

Ch 13 final.qxd

12/18/2006

15:34

Page 352

SURGERY FOR OVARIAN CANCER

Criteria for the success of the operative procedure are not well-defined. The most commonly used criterion, proposed by Castaldo et al., is a postoperative survival longer than 8 weeks. By this criterion, these authors reported a 78% success rate of palliative surgery for intestinal obstruction in recurrent ovarian cancer.10 Subsequent retrospective studies have likewise reported success rates of 62–73% using the same criterion.4,5,7 Rubin et al. proposed that success be defined by the ability of the patient to eat a solid diet on discharge from the hospital and reported a 62% success rate using this criterion.3 Median survival among patients undergoing palliative exploratory laparotomy for bowel obstruction ranges from 2.5 to 5.8 months.3–5,7–9 Major complications of surgery occur in 31–49% of cases and include fistulae, recurrent bowel obstruction, sepsis, pulmonary embolus and wound dehiscence4,5,8,10. Operative mortality, defined as death within 30 days of surgery, is consistently between 13 and 26%. 1,3–5,7–10 Survival longer than 1 year is achieved in 14–26% of patients, including those found to have a non-malignant source of obstruction at laparotomy.3,4,9,10 It is clear that the overall prognosis for most patients presenting with malignant bowel obstruction is grim. Several authors have therefore attempted to develop clinical criteria for operative intervention based on risk factors. Krebs and Goplerud identified age > 65, severe malnourishment, extent of tumor burden, ascites, previous chemotherapy and previous radiotherapy as predictive of survival less than 8 weeks postoperatively and suggested a scoring system using these criteria.7 In subsequent studies the scoring system was applied to other retrospective series with mixed results. Larson et al., in a separate retrospective analysis of 33 patients, found significant correlation of Krebs score with survival.9 However, in separate analyses of their series, Lund et al.11 and Rubin et al.3 did not find the Krebs scoring system predictive. Clarke-Pearson et al. subsequently performed a multivariate logistic regression model and determined that the most important preoperative risk factors for survival of less than 2 months were clinical tumor status

352

and serum albumin concentration. An equation to estimate the probability of survival for longer than 60 days postoperatively was created.12 Although this equation has not been validated using a separate data set, it seems reasonable to use tumor burden and nutritional status as two determinants of a patient’s risk for surgical complications when deciding between surgical intervention and conservative management.

Diagnostic evaluation The staple of diagnostic evaluation of suspected small-bowel obstruction is the plain film of the abdomen with flat and upright views. Findings consistent with small-bowel obstruction include air–fluid levels, dilated loops of small bowel and a paucity of colonic gas. Patients with carcinomatous ileus are more likely to have gas throughout the small and large bowel. The accuracy of plain films in the diagnosis of small-bowel obstruction is 30–70%.13,14 Limitations include the inability to diagnose the cause of obstruction or to diagnose multiple points of obstruction. In addition, in some cases of malignant bowel obstruction, military tumor growth can encase the bowel and prevent dilatation, which limits diagnostic capability. Small-bowel contrast studies are generally more accurate (70–100%) than plain film studies in diagnosing small-bowel obstruction.15,16 A small-bowel follow-through requires administration of barium or water-soluble contrast either orally or via a nasogastric tube. Consecutive plain films at progressive time intervals are obtained to identify the site of bowel obstruction. Failure of contrast to reach the colon within 24 h is generally consistent with high-grade or complete obstruction. An enteroclysis study is a modification of the small-bowel follow-through in which the duodenum is first intubated under fluoroscopy and contrast injected under pressure. This study is highly reliable for both cases of low- and high-grade obstruction, because it bypasses the stomach where washout of contrast can occur.17 It is important to remember that imaging of both small and large bowel is prudent in the evaluation of patients with probable malignant bowel obstruction.

Ch 13 final.qxd

12/18/2006

15:34

Page 353

PALLIATIVE SURGERY FOR OVARIAN CANCER

Concurrent small- and large-bowel obstructions have been noted at surgery in up to 22% of cases.1,3,4,6,7 Therefore, patients undergoing small-bowel contrast studies should also be evaluated using a contrast enema. Multiple sites of obstruction are suggestive of carcinomatosis and may aid the surgeon both in the decision on whether to perform surgery and in selection of the procedure. Computed tomography (CT) imaging is emerging as a reliable method of diagnosing malignant bowel obstruction and has been endorsed by the American College of Radiology as highly appropriate in this endeavor.18 Studies have shown CT to be 78–100% sensitive and more than 90% specific in diagnosing small-bowel obstruction.19,20 One advantage of CT over small-bowel contrast studies is the ability to identify the source of obstruction. In cases of malignant obstruction, the radiographic assessment of tumor burden, specific sites of tumor involvement and volume of ascites may aid the surgeon in the evaluation of risk factors for surgical intervention. In addition, CT can sometimes identify bowel ischemiarelated injury which might hasten surgical intervention.

Management The most important goal of management of malignant small-bowel obstruction is the relief of symptoms with improvement in quality of life. Potential prolongation of life through debulking or by allowing the patient to recover enough function to receive further chemotherapy should be seen as a secondary goal that is less likely to be achieved. The immediate objective is to allow the patient to eat a reasonable diet without excessive nausea or vomiting. The initial step in the management of suspected small-bowel obstruction is placement of a nasogastric tube and administration of intravenous fluids with correction of electrolyte abnormalities. A complete nutritional assessment should be initiated, including serum levels of albumin and transferrin, and a total lymphocyte count. Patients in whom bowel perforation or ischemia are suspected may present with peritoneal signs, elevated white blood cell count and/or

free air on radiographic imaging, and may require immediate laparotomy. However, malignant smallbowel obstruction is rarely an urgent or emergency situation, and prolonged nasogastric suction is an acceptable form of management even in cases of suspected high-grade obstruction.21 Patients with known carcinomatosis and without suspicion for an acute process should be given a substantial trial of nasogastric suction, sometimes up to 2 weeks. If patients do not respond to nasogastric suction within several days, further imaging studies should be considered, to delineate the cause and location of the obstruction. If there is no clinical response to conservative measures by 10–14 days, a decision should be made regarding operative versus medical management. When radiographic imaging is complete and the patient has failed conservative measures, the surgeon must assess the risks and benefits of surgical intervention. Risk factors such as the patient’s nutritional status, suspected tumor burden, amount of ascites and life expectancy should be taken into account. Generally, patients with a life expectancy of less than 2 months are not considered operative candidates. If the physician feels that operative intervention may be of benefit, an informed discussion should be held with the patient regarding options including medical management of symptoms with no further active treatment and surgical intervention. Patients should be made aware that the goals of treatment are palliative rather than curative. Expected morbidity and mortality rates as well as the expected effects of surgery on quality of life should be discussed. When contemplating laparotomy in patients with severe nutritional depletion, a perioperative course of total parenteral nutrition (TPN) should be considered, as this has been shown to lower postoperative complications.22 Patients for whom major surgical intervention is not felt to be an option may benefit from medical management of their symptoms. Opioid analgesics such as morphine and hydromorphone can ease the pain of intestinal obstruction and slow intestinal motility, which may alleviate symptoms.23 An aggressive antiemetic regimen should be instituted and

353

Ch 13 final.qxd

12/18/2006

15:34

Page 354

SURGERY FOR OVARIAN CANCER

should include agents that can be given rectally (prochlorperazine, promethazine, hydroxyzine), sublingually (ondansetron), or subcutaneously (haloperidol). Patients who require nasogastric decompression for alleviation of symptoms may benefit from placement of a percutaneous endoscopic gastrostomy (PEG) tube or percutaneous radiologic gastrostomy tube. These procedures are less invasive than open gastrostomy tube placement and may significantly improve the quality of life.24,25 Patients with upper abdominal tumor and ascites are at a higher risk for gastrostomy-related complications, but the procedure is usually successful among patients with advanced malignancy.26 Consideration of total parenteral nutrition in patients with malignant bowel obstruction who are not surgical candidates is controversial. Abu-Rustum et al. retrospectively reviewed 21 patients with ovarian cancer and small-bowel obstruction who were managed using gastrostomy tube drainage and chemotherapy either with or without TPN. Median survival among patients who received TPN and chemotherapy was 89 days, compared to 71 days for patients receiving chemotherapy only. Although this survival difference was statistically significant, survival was so short that the difference was not felt to justify TPN in this group of patients.27 The risks of line sepsis and metabolic alterations as well as the monetary and human costs of TPN are not likely to be outweighed by the minimal survival benefit. However, as in all decisions surrounding end-of-life care, patients should be evaluated on an individual basis.

Surgical techniques Prior to laparotomy for suspected malignant smallbowel obstruction, a patient’s medical condition should be optimized. Electrolyte abnormalities and clotting factor deficiencies should be corrected completely. For patients with severe nutritional depletion, a perioperative course of TPN should be considered. Antibiotic prophylaxis with coverage for enteric organisms is recommended. Nasogastric decompres-

354

sion should be maintained until surgery to decompress dilated bowel. The major goal at laparotomy for malignant bowel obstruction is location and correction of the bowel obstruction. Debulking intraperitoneal tumor is a secondary goal which is dependent on the individual surgeon’s judgment. Usually, dilated loops of small bowel can be traced to the transition point with adhesiolysis. The procedure required to relieve the obstruction ranges from simple adhesiolysis to resection of tumorinvolved bowel with anastomosis or ileostomy. In cases of malignant small-bowel obstruction, a decision must often be made whether to resect or bypass obstructing tumor-involved bowel. This decision should be based on the feasibility of the resection, the overall tumor burden and the patient’s co-morbidities. In cases of obstructed radiation-damaged distal ileum, it is preferable to resect the damaged segment of bowel along with part of the ascending colon if healthy bowel can be identified for anastomosis. If damaged bowel cannot safely be resected, bypass with ileo-ascending or ileo-transverse enterocolostomy is acceptable, with creation of a mucous fistula from the bypassed loop to prevent ‘blind loop’ syndrome (see below). If intraperitoneal disease is such that the obstruction cannot be relieved, consideration should be given to placement of a gastrostomy tube. Gastrojejunostomy

Gastrojejunostomy is indicated for cases of gastric outlet or duodenal obstruction, usually caused by bulky retroperitoneal adenopathy. This situation is most commonly encountered in advanced or recurrent cervical cancer and is rarely encountered with progressive ovarian cancer. The goal of gastrojejunostomy is creation of an anastomosis between stomach and jejunum to allow diversion of the gastric contents to the functional portion of small bowel. Gastrojejunostomy may be performed using either the anterior or the posterior wall of the stomach. In the former approach, the jejunal loop is brought through a window in the infracolic omentum and over the transverse colon. In the latter approach, the jejunal

Ch 13 final.qxd

12/18/2006

15:34

Page 355

PALLIATIVE SURGERY FOR OVARIAN CANCER

loop is brought through a window in the transverse mesocolon. The anterior is the more common approach and usually preferred unless the patient is obese or has a redundant transverse colon. Anterior gastrojejunostomy is begun by bringing a proximal jejunal loop through a defect created in the infracolic omentum. The loops of bowel and stomach are aligned so that the jejunum is not twisted, and stay sutures are placed to approximate the anterior stomach to the jejunum at proximal and distal ends of the proposed anastomosis (Figure 13.1a). A posterior outer layer of interrupted seromuscular stitches of 2-0 silk is placed with knots inverted. The stomach and jejunum are then opened sharply in parallel over a

a

c

length of 3–5 cm (Figure 13.1b). Interrupted or a running full-thickness stitch of 2-0 or 3-0 polyglactic acid suture is placed to close first the inner posterior layer and then the inner anterior layer of the anastomosis. An outer anterior seromuscular layer of interrupted 20 silk stitches is placed. The anastomosis is palpated for patency and checked for leakage. Alternatively, a stapled anastomosis may be performed. Following placement of the stay sutures, matching 5-mm gastrostomy and enterotomy incisions are created at the proximal end (Figure 13.2a). A side-to-side anastomosis is created using the GIA stapling device with one arm of the device in each lumen (Figure 13.2b–c). The staple line is inspected for hemostasis

b

d

e

Figure 13.1 Anterior gastrojejunostomy, hand-sewn. (a) A proximal loop of jejunum is brought through a defect created in the infracolic omentum. The jejunum is approximated to the stomach using two stay sutures. (b) A posterior interrupted seromuscular imbricating stitch is placed with knots on the serosal side. Parallel 5-cm enterotomies are made in jejunum and stomach. Closure of the (c) posterior, then (d) anterior mucosal layer is performed using an interrupted stitch with knots in the bowel lumen. (e) The anterior seromuscular imbricating closure is then performed

355

Ch 13 final.qxd

12/18/2006

15:34

Page 356

SURGERY FOR OVARIAN CANCER

a c

b

d

Figure 13.2 Anterior gastrojejunostomy, stapled. (a) A proximal loop of jejunum is brought through a defect created in the infracolic omentum. The jejunum is approximated to the stomach using stay sutures. Matching 5-mm gastrotomy and enterotomy incisions are made. (b) Anastomosis is created using a GIA stapling device. (c) The intralumenal staple line is inspected for bleeding. (d) The remaining enterotomy is closed using a TA55 stapling device

and the remaining defect closed using a TA55 stapler (Figure 13.2d). The anastomosis is palpated for patency and checked for leakage. Posterior gastrojejunostomy is accomplished first by entering the lesser sac sharply through the gastrocolic ligament. A window is made sharply in the avascular portion of the left transverse mesocolon. The site for anastomosis on the posterior wall of the stomach is pulled gently through the window and approximated to the cut edge of the mesocolon to prevent herniation. The gastrojejunostomy is then performed as described above. Finally the anterior edge of the transverse mesoenterotomy is approximated to the stomach wall.

356

Small-bowel bypass

In cases of malignant small-bowel obstruction with diffuse carcinomatosis the treatment of choice may be to bypass the obstructed bowel. The purpose of the bypass procedure is to re-establish communication of the intestine proximal and distal to the obstruction. Side-to-side enteroenterostomy is associated with a lower risk for anastomotic leak, breakdown and stenosis than resection and anastomosis, and therefore may be desirable in the patient with progressive disease. The technique of side-to-side anastomosis is usually performed to bypass a segment of obstructed bowel that cannot be resected. The most distal portion of healthy small bowel proximal to the obstruc-

Ch 13 final.qxd

12/18/2006

15:34

Page 357

PALLIATIVE SURGERY FOR OVARIAN CANCER

tion should be connected with the most proximal healthy small or large bowel distal to the obstruction. Side-to-side anastomosis can be accomplished either by a hand-sewn technique or by using the GIA stapler. Side-to-side bypass procedures may reanastomose small bowel to small bowel or to ascending or transverse colon. In the setting of a fistula or perforation, the segment of small bowel that is diseased may be more completely isolated from the fecal stream by being transected, and both proximal and distal lumena brought to the skin as mucous fistulae to prevent ‘blind loop’ syndrome (Figure 13.3). After the obstructed segment of bowel has been identified, both loops of bowel are occluded proximally and distally atraumatically and moist laparotomy pads placed into the field to collect any spill of

enteral contents. The two loops are approximated in a side-to-side fashion using stay sutures 6 cm apart on the antimesenteric surface of the bowel. Matching 5mm enterotomy incisions are created in each loop (Figure 13.4a). The anastomosis is then created using a GIA stapler with one arm of the instrument in each limb and the bowel mesenteries rotated away from each other (Figure 13.4b–c). The intralumenal staple lines are inspected for hemostasis and the edges of the remaining defect approximated using Allis clamps. The defect is closed just beneath the Allis clamps using a TA55 stapler (Figure 13.4d). The anastomosis is palpated for patency and checked for integrity. A hand-sewn anastomosis is also possible. After the two loops of small bowel have been approximated, an outer posterior layer of 3-0 silk interrupted sutures

Mucous fistula stoma

a Diseased segment remains in abdomen

b

Figure 13.3 Small-bowel resection with isolation of the diseased segment and creation of a mucous fistula. (a) The segment of diseased bowel is isolated while anastomosis of proximal and distal bowel is accomplished. (b) Proximal and distal segments of diseased bowel are brought to the skin surface, and mucous fistulae are created to allow drainage from the isolated segment. The mesenteric defect is closed

357

Ch 13 final.qxd

12/18/2006

15:34

Page 358

SURGERY FOR OVARIAN CANCER

a

c

b

d

Figure 13.4 Small-bowel bypass, stapled side-to-side anastomosis. (a) Antimesenteric surfaces of small bowel are approximated and matching 5-mm enterotomies are made to introduce the GIA stapling device. (b) The stapling device is fired, creating the anastomosis. (c) The remaining enterotomy is elevated and the intralumenal suture line inspected for hemostasis. (d) The enterotomy is closed using the TA55 stapling device

is placed in the seromuscular layer with knots inverted. Matching longitudinal 5-cm enterotomies are made in the bowel loops 4 mm away from the posterior layer of stitches. A posterior running or interrupted full-thickness stitch of 2-0 polyglactic acid suture is placed posteriorly (Figure 13.5a). An anterior running or interrupted full-thickness stitch of 2-0 polyglactic acid suture is placed. If an interrupted stitch is used, the knots should be placed within the bowel lumen. An anterior outer layer of interrupted seromuscular stitches is placed using 3-0 silk. The anastomosis is palpated for patency and checked for leakage (Figure 13.5b). Alternatively, an anastomosis of proximal

358

small bowel to colon may be performed in a similar fashion. Ileostomy

Ileostomy is indicated in cases of malignant bowel obstruction for diversion of bowel contents when no functional distal bowel is available for anastomosis, when bowel is too edematous for anastomosis, or to protect a distal anastomosis when the risk of anastomotic leak is high. Loop ileostomy is generally preferred over an end ileostomy because of the decreased incidence of complications such as retraction, necrosis and stenosis.

Ch 13 final.qxd

12/18/2006

15:34

Page 359

PALLIATIVE SURGERY FOR OVARIAN CANCER

a

b

Figure 13.5 Small-bowel bypass, hand-sewn side-to-side anastomosis. (a) Matching longitudinal 5-cm enterotomies are created in the antimesenteric surface of both segments of bowel. The ‘back wall’ of the anastomosis is approximated first using an interrupted imbricating seromuscular layer with knots on the serosal side. The lumen is then closed with a mucosal layer beginning with the back wall and continuing on to the anterior wall until the entire lumen is closed. (b) The anastomosis is completed by closing the anterior seromuscular layer with imbricating sutures

The preferred location for an ileostomy is in the right lower quadrant; the ileostomy should tunnel through the rectus muscle to prevent prolapse (Figure 13.6a). The site should be at least 5 cm from the umbilicus, iliac spine and incision to facilitate placement of the ostomy bag. A circle of skin 2 cm in diameter is excised sharply. Blunt dissection is performed to expose the rectus fascia, which is incised in a cruciate fashion. The rectus muscles are bluntly separated vertically and a cruciate incision made in the posterior rectus sheath and peritoneum. The tract should

accommodate two fingers. A loop of ileum is selected and a mesenteric defect is created adjacent to the bowel. A red rubber 16-Fr catheter is placed through this mesenteric defect and the loop of bowel with the catheter pulled through the abdominal wall tunnel so that approximately 6 cm of ileum is above the skin surface (Figure 13.6b). The loop of ileum should be under no tension. After the abdominal incision has been closed, the stoma is matured. A transverse incision is made in the bowel closest to the superior limb and about 1 cm from the skin edge (Figure 13.6c). The edges of bowel are then everted using Allis clamps, and the full-thickness edge of the bowel is sutured to the subcuticular layer of the skin using circumferential, interrupted stitches of 2-0 polyglactic acid suture (Figure 13.6d). The ostomy bag is placed in the operating room. The red rubber catheter is removed 7 days postoperatively. Small-bowel resection

Small-bowel resection with anastomosis is often necessary to correct an obstruction if a segment of bowel has been injured in the course of dissection, when removal of bowel involved with tumor is desired, or when a segment of bowel has been chronically narrowed by adhesions or radiation injury. Resection and re-anastomosis may be accomplished using either hand-sewn methods or surgical staplers. Either method is equally acceptable as long as three basic principles are followed. (1) The anastomosis should have adequate blood supply from both sides. (2) The anastomosis should not be under tension. (3) The anastomosis should have an adequate lumen. The techniques of small-bowel resection with end-toend and functional end-to-end anastomoses, using hand-sewn and stapled closure methods, are described in Chapter 7.

359

Ch 13 final.qxd

12/18/2006

15:34

Page 360

SURGERY FOR OVARIAN CANCER

c

a

b

d

Figure 13.6 Loop ileostomy. (a) The abdominal wall defect is created in the right lower quadrant in the middle of the rectus muscle. (b) A defect is created in the mesentery of the ileum and a red rubber catheter brought through the defect. The loop of ileum is brought through the abdominal wall defect. (c) The ostomy is matured by incising the bowel transversely. (d) Edges of bowel are everted and sutured circumferentially to the subcuticular skin layer. The red rubber catheter is removed 7 days postoperatively

Gastrostomy decompression

Open placement of a decompressing gastrostomy tube is indicated when, at laparotomy for malignant bowel obstruction, operative correction of the obstruction is not technically possible or re-obstruction is felt to be highly likely. Patients with malignant bowel obstruction who are poor candidates for operative intervention should be considered for radiologically or endoscopically guided gastrostomy tube placement rather than open gastrostomy. Palliative tube gastrostomy obviates the need for long-term nasogastric decompression, which is uncomfortable and may predispose to aspiration. Open gastrostomy tube placement is performed under general anesthesia. In most cases the patient has already been explored through a midline incision. Alternatively, a 5–7-cm vertical left subcostal incision may be made. A site on the lower body of the stomach that is accessible to the anterior abdominal wall is selected. The site should not be adjacent to major

360

blood vessels. A tube with multiple fenestrations such as an 18–20-Fr Malekot is brought through a puncture wound in the anterior abdominal wall at a site where the stomach can approximate the abdominal wall (or through the small vertical incision). The stomach wall is grasped using Babcock clamps (Figure 13.7a). Two concentric purse-string stitches of 2-0 polyglactic acid suture are placed in the stomach wall and a small gastrostomy incision is made (Figure 13.7b). The Malekot drain is placed through the gastrostomy using a tonsil clamp to elongate the end of the drain, and the purse-string stitch is tied down. The gastrostomy site is approximated to the anterior abdominal wall and secured there using four interrupted stitches of 2-0 polyglactic acid suture. The tube is secured to the skin using 2-0 silk (Figure 13.7c). Postoperatively the gastrostomy tube may be placed to dependent drainage or low intermittent suction. The tube should be flushed with saline several

Ch 13 final.qxd

12/18/2006

15:34

Page 361

PALLIATIVE SURGERY FOR OVARIAN CANCER

a

b

c

Figure 13.7 Gastrostomy tube, open. (a) The stomach wall is grasped using a Babcock clamp to assess mobility. Two purse-string stitches are placed. (b) Gastrotomy incision is made. A 20-Fr Malekot drain is brought through a separate stab incision in the left upper quadrant and placed into the stomach. (c) Additional interrupted stitches are placed to approximate the stomach to the abdominal wall parietal peritoneum. The Malekot drain is secured to the skin

times a day to prevent obstruction. The skin stitch may be removed 10–14 days postoperatively. Leakage around the tube may occur and can cause skin irritation, which can be treated by direct application of over-the-counter antacids. In cases of copious leakage an ostomy bag may be needed. If the patient experiences discomfort, nausea, or vomiting, a radiologic tube study may be necessary to ensure correct positioning.

and end-to-end anastomosis. If surgical intervention is not an option, medical management consists of broad spectrum antibiotics, a low-fat diet and administration of specific nutrients such as vitamin B12. Blind loop syndrome may be prevented at the time of bypass procedure by creation of a mucous fistula from the proximal end of the bypassed bowel (Figure 13.3a–b).

Complications

Short bowel syndrome results from the inadequate absorption of nutrients and electrolytes by the bowel, and is caused by either malfunctioning bowel or its surgical absence. In gynecologic cancer patients, the most common causes of short bowel syndrome are radiation enteritis and surgical resection of bowel. Radiation-induced short bowel syndrome is most common among patients who have received whole abdominal radiation but has rarely occurred after whole pelvic radiation. The surgical removal of the small bowel is not usually associated with malnutrition, but varying degrees of malnutrition are seen with extensive resections. Up to 50% of the bowel can be removed without affecting nutritional status. The average length of the small bowel in women is 600 cm (20 ft). Serious nutritional

Blind loop syndrome

Blind loop syndrome rarely develops as a consequence of bowel bypass procedures for palliation of advanced ovarian cancer (such as enteroenterostomy and enterocolostomy). The syndrome is due to stasis of intestinal contents in the bypassed loop of bowel which leads to bacterial overgrowth, bowel dilatation and inflammation. The syndrome can lead to fistula formation or even perforation. Symptoms of blind loop syndrome include crampy abdominal pain, distension, nausea, vomiting and diarrhea. Patients may present with weight loss, megaloblastic anemia, or other signs of malnutrition. The treatment for blind loop syndrome is resection of the diseased portion of bowel

Short bowel syndrome

361

Ch 13 final.qxd

12/18/2006

15:34

Page 362

SURGERY FOR OVARIAN CANCER

deficiencies can occur with 200 cm (6 ft) of bowel while survival without TPN requires at least 70–100 cm (3 ft) of functional small bowel. Patients are able to survive after total colectomy without supplementation if at least 150–200 cm of small bowel remains.28 The small bowel is responsible for absorption of most major nutrients. The simple sugars and fatty acids are absorbed throughout the small bowel, while amino acids are absorbed in the jejunum. The ileum absorbs water, bile salts, fat-soluble vitamins and vitamin B12. The colon absorbs water, electrolytes and short-chain fatty acids. Loss of portions of the small bowel can lead to steatorrhea, diarrhea, excessive water loss with dehydration, mixed anemia, electrolyte abnormalities and vitamin deficiencies. The ileum normally absorbs bile acids, which can stimulate colonic fluid and electrolyte secretion, causing diarrhea after ileal resection. Patients with short bowel syndrome are at increased risk for cholesterol gallstone formation due to a decrease in the bile acid pool size. Renal stone formation is also more likely, owing to an increase in oxalate absorption by the large intestine. The initial management of patients with known short bowel syndrome is TPN. Patients with varying degrees of nutritional depletion will benefit from additional measures to optimize remaining bowel function. H2 blockers control gastric secretions, and antimotility agents (such as loperamide and diphenoxylate plus atropine) are administered to control diarrhea or high ostomy output. Cholestyramine may be given to absorb bile salts and improve diarrhea. If more than 100 cm of terminal ileum has been resected, cholestyramine is not likely to be effective, because excessive amounts of bile acids are being excreted fecally and the liver is unable to synthesize enough bile acids to compensate for the depleted pool. This prevents adequate jejunal fat absorption and causes worsening steatorrhea. Enteral feedings via a feeding jejunostomy tube are often initiated and can stimulate hypertrophy of the remaining functional bowel. Intestinal adaptation to

362

slow enteral feeding may lead to the ability to wean the patient from chronic TPN. Adaptation of the small intestine to extensive resection may take up to 12 months. Oral feedings should consist of a highcarbohydrate, low-fat diet. Patients with colonic continuity should receive a low-oxalate diet to prevent urinary stones. Patients with progressive metastatic ovarian cancer are not generally candidates for home TPN and should be counseled for palliative care.

LARGE-BOWEL OBSTRUCTION Large-bowel obstruction accounts for up to one-third of all cases of ovarian cancer-related bowel obstruction.1,3,4,6,7 Like patients with malignant small-bowel obstruction, these patients usually present in the advanced stages of disease. Presenting symptoms of large-bowel obstruction include progressive constipation, hematochezia, abdominal distension and abdominal pain with or without nausea and vomiting. Physical findings include a distended abdomen, right lower quadrant tenderness and high-pitched bowel sounds. The differential diagnosis for patients with suspected large-bowel obstruction and a history of ovarian cancer includes fecal impaction, diverticular disease, primary gastrointestinal malignancy and colonic pseudo-obstruction (Ogilvie’s syndrome). Malignant large-bowel obstruction is usually partial in nature and may therefore initially be managed conservatively. If left untreated, progressive large-bowel obstruction can lead to cecal perforation, sepsis and death. The most commonly performed operative procedures for alleviation of malignant large-bowel obstruction are colostomy (with or without partial colectomy) and ileostomy. The prognosis of patients undergoing palliative surgery for malignant largebowel obstruction is poor. Median survival ranges from 3.6 to 9.2 months and is not significantly longer than among patients undergoing palliative surgery for small-bowel obstruction4–6. However, in one study,

Ch 13 final.qxd

12/18/2006

15:34

Page 363

PALLIATIVE SURGERY FOR OVARIAN CANCER

postoperative complications were significantly less frequent among patients with large-bowel obstruction than among those with small-bowel obstruction.4

options prior to proceeding with exploratory laparotomy.

Management Diagnostic evaluation The diagnostic evaluation of patients with suspected large-bowel obstruction begins with plain films of the abdomen. Findings characteristic of large-bowel obstruction include air–fluid levels in both large and small bowel with dilated large bowel proximal to the site of obstruction and a paucity of distal colonic gas. A dilated right colon with gas extending to the rectum is consistent with Ogilvie’s syndrome, or colonic pseudo-obstruction. These patients are at risk for cecal perforation and need to be followed closely. Plain films revealing a non-obstructive gas pattern with stool throughout the descending and rectosigmoid colon are diagnostic of fecal impaction. Other useful imaging studies include CT and contrast enema. CT has the advantage of potentially diagnosing the cause of obstruction, such as mass or diverticular abscess, as well as giving important information about tumor burden and ascites. However, colonic ileus has sometimes been misinterpreted as a large-bowel obstruction on CT.20 Contrast enema is a highly reliable study to delineate a large-bowel obstruction and can also reveal the presence of metastatic implants on the colon.15,29 The barium enema is contraindicated in patients in whom perforation is suspected and in those with extreme cecal dilatation of 10 cm or greater; these patients are at high risk for barium peritonitis which can be lifethreatening.15 When perforation cannot be excluded and a contrast enema is desired, water-soluble contrast should be used. Flexible sigmoidoscopy and colonoscopy are valuable techniques for imaging the colon and may be useful in diagnosing the source of a partial obstruction. Endoscopically guided biopsies will often determine whether an obstructing malignancy is a recurrent ovarian neoplasm or a new primary gastrointestinal tumor. This may be helpful in discussing treatment

Patients with suspected malignant large-bowel obstruction who do not have extreme cecal dilatation may initially be managed with nasogastric decompression, intravenous hydration, electrolyte correction and intravenous antibiotics. If the patient remains stable without peritoneal signs or extreme cecal dilatation, appropriate contrast studies may be obtained. The decision on whether to proceed with exploratory laparotomy may then be made with all imaging information available. This decision must be based on the known risk factors influencing survival, including poor nutritional status and extreme tumor burden. Other relative contraindications to surgery for malignant colonic obstruction include largevolume ascites, distant metastases with impairment of other vital organs, multiple sites of intestinal obstruction and rapidly progressive disease. An informed discussion should be held with the patient regarding the significant morbidity and mortality of the procedure and its palliative rather than curative intent. In general, however, we prefer to correct/relieve colonic obstruction rather than allow a patient to die of sepsis from colonic perforation. Patients with extreme cecal dilatation require emergency decompression to avoid spontaneous perforation. This is accomplished at exploratory laparotomy and usually requires colostomy or cecostomy. Likewise, patients with peritoneal signs or elevated white blood cell count may not be candidates for a trial of conservative management. These patients should not receive mechanical bowel preparation but broad-spectrum intravenous antibiotics are necessary. Patients who present with evidence of colonic perforation are managed by emergency laparotomy and loop ileostomy. For non-acute patients presenting with partial colonic obstruction, a mechanical and antibiotic bowel preparation should be performed over several days. This will reduce infectious complications and the risk of anastomotic breakdown.

363

Ch 13 final.qxd

12/18/2006

15:34

Page 364

SURGERY FOR OVARIAN CANCER

For patients who are not felt to be operative candidates, new colonic stenting devices are becoming available. Several authors have reported success with the use of expandable stents placed using either endoscopic or fluoroscopic guidance for palliative management of large-bowel obstruction. Placement of such devices was initially successful in 64–100% of cases of colonic obstruction due to malignancy and has been successful in ovarian cancer patients.30,31 Potential complications include bowel perforation and stent migration. It remains to be seen whether these techniques will become standard in the palliative management of malignant large-bowel obstruction.

Surgical techniques The primary surgical objective in cases of malignant large-bowel obstruction is palliation of symptoms and prevention of sepsis. This can often be accomplished by performing a diverting colostomy or ileostomy. In cases in which tumor recurrence is isolated as an obstructing mass, the surgeon may choose to resect bowel along with obstructing tumor. When diffuse peritoneal carcinomatosis and excessive tumor burden are found at laparotomy, no benefit is likely to be realized by performing a colonic resection, and diverting colostomy is the procedure of choice. The diversion should be performed in the most distal portion of the large bowel which is proximal to the obstruction. Diverting colostomy

Diverting colostomy is often the simplest procedure to palliate a large-bowel obstruction. The decision about which type of colostomy should be performed should be made by an experienced surgeon. In general, end colostomy is a planned procedure performed when a colostomy is expected to be permanent. Loop colostomy is often performed as an urgent or emergency procedure for obstruction of the left colon, sigmoid colon, or rectum. A transverse loop colostomy is the preferred procedure in such cases. The loop colostomy may remain as a permanent colostomy, although an end colostomy is preferable when permanency is desired, because there is a lower chance of prolapse.

364

Transverse loop colostomy A transverse 5–6-cm skin

incision is made between the umbilicus and the left costal margin (Figure 13.8a). The anterior fascia is incised transversely, rectus muscles are separated longitudinally, and the posterior sheath and peritoneum incised transversely. The transverse colon is usually easily found, but may be located by elevating visible omentum as well. The omentum is dissected free from the distal surface of the colon (Figure 13.8b). The mesentery of the transverse colon is opened just beneath the serosa of the transverse colon (Figure 13.8c). A red rubber 16-Fr catheter is passed beneath the colon and elevated above the level of the skin to prevent retraction of the loop of colon into the abdominal cavity (Figure 13.8d). Omentum, distal and proximal colon are placed back into the abdominal cavity. A fascial bridge stitch of 0 PDS suture is placed to bring fascia together beneath the loop of colon. A similar skin bridge stitch of 2-0 polyglactic acid suture is placed. The colon is opened along the taeniae on the antimesenteric border (Figure 13.8e). The free edges of the open colon are sutured to the subcuticular layer of the skin using 2-0 polyglactic acid suture (Figure 13.8f). A clear collection bag is placed over the colostomy in the operating room. The red rubber catheter is removed 5–7 days postoperatively. End colostomy End colostomy is usually a planned

procedure performed for permanent diversion when the distal bowel is being removed. If distal diseased or obstructed bowel remains, a mucous fistula or loop colostomy is necessary. The most common type of end colostomy is in the sigmoid colon. In most cases the patient has been explored through a midline laparotomy incision. The sigmoid colon has been divided using a GIA stapler proximal to the obstruction and the distal colon resected. The ostomy is ideally placed halfway between the umbilicus and iliac spine and should tunnel through the rectus muscle to prevent hernia formation (Figure 13.9a). A circle of skin, 2.5–3 cm in diameter, is excised at the proposed ostomy site. Blunt dissection through subcutaneous fat will reveal the anterior rectus fascia, which is incised

Ch 13 final.qxd

12/18/2006

15:34

Page 365

PALLIATIVE SURGERY FOR OVARIAN CANCER

Anterior leaf of omentum

a

b

Transverse colon freed from attachment to posterior leaf of omentum c Small defect created in transverse mesocolon

Red rubber catheter

Colon oipened along taenia

d

e

f

Figure 13.8 Transverse loop colostomy. (a) The site for colostomy is in the left or right upper quadrant and the defect is created through the rectus muscle. (b) A segment of transverse colon is selected and the omentum dissected free from the distal surface of the colon. (c) The transverse mesocolon mesentery is opened just beneath the serosa of the transverse colon. (d) A red rubber 16-Fr catheter is passed beneath the colon. (e) The omentum and distal proximal colon are placed back into the abdominal cavity with the red rubber catheter beneath the loop colostomy at the skin level. The colon is opened along its taeniae on the antimesenteric border. (f) Sutures are placed in the skin and colon, thus creating a ‘double-barrel’ colostomy. The red rubber catheter is removed 7 days postoperatively

in a cruciate fashion. The rectus muscle is split longitudinally and the posterior sheath and peritoneum incised. The tunnel should easily accommodate two fingers. A Babcock clamp is placed through the ostomy tunnel and the end of the sigmoid colon grasped

and brought through, with care not to twist the bowel. The colon should be adequately mobilized so that it sits easily above the skin. Appendices epiploicae may need to be trimmed. The wall of the colon is sutured circumferentially to the subcuticular layer of skin

365

Ch 13 final.qxd

12/18/2006

15:34

Page 366

SURGERY FOR OVARIAN CANCER

using interrupted stitches of 2-0 polyglactic acid suture (Figure 13.9b). Before the abdominal incision has been closed, the colon is inspected for viability and adequate blood supply. After the abdominal incision is closed, the end of the colon is opened by sharply excising the staple line (Figure 13.9c). The full-thickness bowel edge is secured circumferentially to the subcuticular layer of the skin using interrupted 2-0 polyglactic acid suture (Figure 13.9d). A drainage bag is placed. Cecostomy

When the cecum is persistently dilated more than 10–12 cm and endoscopic decompression is not possible, owing to obstruction, or not successful, tube cecostomy is indicated to prevent spontaneous perfo-

a

ration. A small transverse right lower quadrant incision is made. The anterior fascia is incised transversely, rectus muscles are separated longitudinally, and the posterior sheath and peritoneum incised transversely. The cecum is identified and may be decompressed using a spinal needle. Two concentric purse-string stitches of 2-0 polyglactic acid suture are placed in the wall of the cecum. A 20-Fr Malekot drain is brought through a separate stab incision in the abdominal wall. The cecum is opened sharply within the purse-string sutures and the Malekot drain placed (Figure 13.10a). The purse-string sutures are tied down (Figure 13.10b). Additional interrupted stitches of 2-0 polyglactic acid suture are placed to approximate the cecal wall to the abdominal wall (Figure 13.10c). The Malekot drain is sutured to the

b

d

c

e

Figure 13.9 Descending end colostomy. (a) The site for colostomy is usually in the left lower quadrant and the defect is created through the rectus muscle. (b) The previously divided distal segment of colon is brought through the abdominal wall defect using a Babcock clamp. The bowel wall is sutured circumferentially to the subcuticular skin layer. (c) The staple line is sharply excised. (d) Bowel edges are everted and the full-thickness bowel edge is approximated circumferentially to subcuticular skin

366

Ch 13 final.qxd

12/18/2006

15:34

Page 367

PALLIATIVE SURGERY FOR OVARIAN CANCER

skin using 2-0 silk suture. The abdominal incision is closed and a drainage bag is connected to the Malekot.

Complications Common complications of colostomy include necrosis, retraction, para-stomal hernia formation, stenosis and prolapse. Necrosis and retraction both result from excessive tension on the bowel and can lead to stomal stenosis. The treatment for stenosis is surgical revision versus colostomy takedown if no longer necessary. Para-stomal hernia is more common following end than loop colostomy and may require repair using permanent mesh. Prolapse occurs preferentially in the distal limb of a loop colostomy and may be corrected surgically by suturing the distal limb to the anterior abdominal wall using permanent suture or by taking the colostomy down.

a

URINARY TRACT OBSTRUCTION b

c

Because its presentation is often silent, clinically evident ureteral obstruction is uncommon in ovarian cancer patients. At autopsy, 28% of patients with advanced ovarian cancer have some degree of ureteral obstruction, which is usually unilateral.2 Patients who are symptomatic present with pyelonephritis or renal colic. Malignant ureteral obstruction of any type is commonly an end-stage event with a median survival of 7 months.32 Causes of death from ureteral obstruction include sepsis and renal failure in rare cases of bilateral obstruction. However, most patients with advanced ovarian cancer with urinary obstruction die of another cause. Urinary diversion, which is associated with morbidity and discomfort, is not always necessary in these patients.

Figure 13.10 Tube cecostomy. (a) Two purse-string stitches are placed in the wall of the cecum. A 20-Fr Malekot drain is brought through a separate stab incision in the abdominal wall. (b) The cecum is incised, Malekot drain placed and the pursestring sutures tied down. (c) Additional interrupted stitches are placed to approximate the cecum to the abdominal wall

Management Ureteral obstruction is often unilateral and asymptomatic in patients with advanced ovarian cancer. It is usually identified at the time of abdominal imaging

367

Ch 13 final.qxd

12/18/2006

15:34

Page 368

SURGERY FOR OVARIAN CANCER

performed to evaluate progression of disease. Occasionally an acute elevation in the serum creatinine level leads to suspicion of ureteral obstruction. However, increased serum creatinine levels are difficult to interpret in patients who have experienced fluid loss or who have previously been treated with renally toxic drugs such as cisplatin. Furthermore, the serum creatinine level often remains normal as long as one kidney has good function. Symptoms of renal colic or a clinical presentation of pyelonephritis can be suggestive of ureteral obstruction, but these symptoms are non-specific and can occur in the absence of urinary tract compromise. The evaluation of a patient with suspected ureteral obstruction may be accomplished using renal ultrasound, intravenous pyelogram, radionuclide renography, or CT.33 CT is the preferred method of imaging in patients with normal renal function, since information can be gained about both the degree of obstruction and the location and volume of the obstructive tumor. Both intravenous pyelogram and CT utilize intravenous contrast dye, which is nephrotoxic. Patients with significant renal compromise who cannot tolerate an intravenous dye load may be evaluated using renal ultrasound, radionuclide renography, or retrograde dye studies. Renal ultrasound is highly sensitive and specific for the detection of ureteral obstruction.34 Radionuclide renography has the advantage of giving information about each kidney’s relative function, which may be valuable when making the decision of whether to divert an obstruction. Once ureteral obstruction has been diagnosed in a patient with advanced ovarian cancer, the oncologist must decide whether to intervene. Intervention to salvage an obstructed kidney may be warranted to relieve symptoms of renal colic or recurrent pyelonephritis. Alternatively, urinary diversion may preserve renal function enough to allow a patient to continue to receive chemotherapeutic treatment. Unfortunately, most patients with recurrent, progressive disease are not expected to respond to third- and fourth-line salvage treatments and therefore any

368

diverting intervention is likely to be permanent. Invasive interventions have the potential to cause significant discomfort and lead to multiple hospital admissions without proven survival benefit and therefore are not always warranted. The surgeon must consider both the expected benefits of available treatments as well as their effects on the patient’s quality of life in making recommendations regarding urinary diversion. Several options are available to relieve ureteral obstruction. Percutaneous nephrostomy can usually be performed by an interventional radiologist under ultrasound or CT guidance and provides immediate relief of obstruction. The procedure is relatively safe, with the most common immediate complications being hematoma (4%) and perinephric abscess (4%).35 Normalization of renal function may be expected in 70–88% of cases.35,36 Significant longterm complications of percutaneous nephrostomy include positive urine cultures or pyelonephritis (62–70%) and leakage from or blockage of the catheter, requiring reinsertion (65%).35,36 In addition, these catheters require exchange every 3 months. Internal ureteral stents can be placed in a retrograde fashion cystoscopically, or antegrade placement can be performed during initial percutaneous nephrostomy placement. Most internal stents are constructed of flexible synthetic materials and are in a ‘double J’ configuration with one end coiled in the renal pelvis and the other coiled in the bladder to prevent stent migration. Flexible guidewires are often used to negotiate narrowed or distorted ureters. The success of internal stenting for malignant ureteral obstruction is variable. While initial placement of the stent is often possible, there is a high failure rate. Such complications as urinary infections and blocked stents are common, and frequent hospitalizations are often required.32 More recently, some success has been reported with internal metallic, self-expanding stents with a lower rate of re-obstruction and infection.32 The morbidity associated with ureteral decompression techniques in the face of malignancy is

Ch 13 final.qxd

12/18/2006

15:34

Page 369

PALLIATIVE SURGERY FOR OVARIAN CANCER

significant. In one series of 78 patients undergoing ureteral decompression for malignancy using either percutaneous nephrostomy or cystoscopic decompression, there was a 50% complication rate, with twothirds of complications requiring hospitalization for treatment.37 A second series of 103 patients with advanced malignancy undergoing urinary diversion with either percutaneous nephrostomy or internal stent reported a 63% complication rate with a median survival of 5 months after decompression. Of the patients’ remaining life, 50% was spent in the hospital.38 Given the high rate of morbidity associated with decompression and the fact that ureteral obstruction is not likely to be symptomatic or be the primary cause of death in ovarian cancer patients, it may be prudent in most asymptomatic cases to allow one kidney to fail. In terminal patients with bilateral obstruction in whom all surgical and chemotherapeutic options have been exhausted, consideration should be given to conservative management, hospice care and a peaceful death from uremia.

ASCITES Ascites is a common presenting problem among women with advanced-stage ovarian cancer. Primary surgical debulking and platin-based chemotherapy are usually successful in resolving symptomatic ascites. Patients with recurrent cancer often develop recurrent ascites as well. Small-volume ascites is usually asymptomatic and can be ignored, but a large volume of ascites can cause abdominal distension, nausea, vomiting, anorexia and respiratory difficulties which significantly impact on the patient’s quality of life. Therefore, palliative attempts at ascites management are important in these patients. The two primary factors involved in the pathogenesis of ascites formation are increased production of ascitic fluid and impaired drainage. Neoplastic production of vascular endothelial growth factor and other substances increases the number and available area of microvessels lining the peritoneal cavity while

also increasing capillary permeability. This leads to intraperitoneal accumulation of proteins which raises peritoneal fluid oncotic pressure and promotes fluid filtration into the peritoneal cavity. Meanwhile, accumulation of tumor in subdiaphragmatic lymphatics inhibits fluid evacuation from the peritoneal cavity.39 The net result is ascites accumulation.

Management Malignant ascites usually regresses when salvage chemotherapy is effective. In the setting of chemotherapy-resistant progressive disease, ascites is traditionally managed using serial therapeutic paracentesis. Unfortunately, repetitive paracentesis carries the risk of injury to the bowel and peritoneal infection. Repeated removal of large volumes of ascites can lead to electrolyte imbalances and protein depletion. In addition, loculated ascites may develop which is more difficult to drain by blind paracentesis. In these cases, CT or ultrasound imaging may be required to direct needle or catheter placement (Figure 13.11). Alternatives to serial paracentesis include permanent drainage catheter placement and peritoneovenous shunting. Permanent tunneled peritoneal drainage catheters can be placed under conscious sedation with radiographic guidance, resulting in excellent success rates and low infectious morbidity.40 The procedure is usually performed in an out-patient setting and therefore may be preferred over more invasive procedures. Alternatively, several groups have reported peritoneovenous shunting procedures whereby ascites is siphoned directly into the femoral or internal jugular vein. Successful palliation of symptoms was achieved in 60–88% of cases. Complications included shunt occlusion (5–36%), infection (2.4–18%) and pulmonary edema (0–15%).41 Various intraperitoneal therapies including interferon, cisplatin, chromic phosphate (32P) and Corynebacterium parvum have been tried to control recurrent ascites.42–44 It has been our observation that these treatments are of no significant value in the setting of ascites and recurrent ovarian cancer.

369

Ch 13 final.qxd

12/18/2006

15:34

Page 370

SURGERY FOR OVARIAN CANCER

a

b

Figure 13.11 Radiographically guided semipermanent catheter placement for malignant ascites. (a) Loculated upper abdominal ascites in a patient with progressive ovarian cancer. The patient had abdominal discomfort, nausea and vomiting. (b) A semipermanent catheter has been placed into the ascites under computed tomography guidance; 2000 ml of ascites was drained and the patient had significant relief

PLEURAL EFFUSION Malignant pleural effusions are fairly common among patients with advanced ovarian cancer and can cause unpleasant symptoms including dyspnea, chest discomfort and exercise intolerance. Malignant pleural effusions often resolve spontaneously in the course of initial platin-based chemotherapy for primary ovarian cancer. Intervention is indicated for effusions that fail to resolve with chemotherapy treatments and which cause significant symptoms. The pathophysiology of malignant pleural effusions is best described with an understanding that pleural fluid homeostasis depends upon a balance between capillary and interstitial oncotic and hydrostatic pressures. Pleural effusion formation is thought to begin with metastases of tumor cells to the visceral pleura with subsequent seeding of the parietal pleura.45 The pleural tumor cells may contribute to pleural effusion formation in several ways. First, tumor may directly block lymphatics, leading to fluid accumulation in the pleural space. Second, direct injury of the pleural capillary bed by tumor may cause protein, then fluid, to leak into the space. Third, tumor cells pro-

370

duce vascular endothelial growth factor, which promotes endothelial permeability and is thought to play a role in effusion formation.46,47 Patients with advanced disease who are malnourished develop hypoalbuminemia and hypoproteinemia, which lower plasma oncotic pressure and contribute further to effusion formation.

Management Dyspnea is the most common presenting symptom in patients with pleural effusions. A plain chest radiograph is usually sufficient to identify a clinically important pleural effusion. A CT scan often identifies smaller effusions which may not be clinically recognizable. CT is also useful in the identification of mediastinal or parenchymal disease, although these are less frequently identified in ovarian cancer patients. Diagnostic thoracentesis may be useful in identifying malignancy as the cause of an effusion, although among patients with known recurrent or metastatic disease these tests are usually not necessary. Positive cytology is diagnostic of malignant effusion, but sensitivity is variable and ranges from 62 to 90%.47

Ch 13 final.qxd

12/18/2006

15:34

Page 371

PALLIATIVE SURGERY FOR OVARIAN CANCER

Malignant effusions are usually exudative, with protein content greater than 3 g/dl, pleural protein/serum protein ratio greater than 0.5, pleural lactate dehydrogenase (LDH)/serum LDH ratio greater than 0.6, and specific gravity greater than 1.015. Malignant effusions are more likely to be bloody and have a higher white cell count.48 Thoracentesis often serves as the primary diagnostic and therapeutic modality in the patient with newly diagnosed ovarian cancer. These patients can be expected to receive some relief from the procedure, and cytology may be obtained to confirm the stage of disease. The malignant effusion will usually respond to administration of primary platin-based chemotherapy. Therapeutic thoracentesis should be limited to a maximum of 1–1.5 l because of the risk of mediastinal shift and re-expansion pulmonary edema.47 Most patients with known malignant effusion and recurrent or progressive ovarian cancer, as well as some newly diagnosed stage IV patients, require chest tube placement with subsequent pleurodesis for resolution of the effusion. Open placement of large-bore chest tubes is becoming less common, as several studies have shown similar success rates with placement of small (8-14 Fr)-bore catheters, often using radiographic guidance49–51 (Figure 13.12). Catheters are

a

b

typically left in place until pleural drainage of less than 150 ml/day is obtained and adequate lung reexpansion is seen on repeat radiographs. Intrapleural sclerosis, or pleurodesis, is then performed to prevent re-accumulation of the effusion. Agents commonly used for sclerosis include doxycycline (500 mg), bleomycin (15–240 units) and talc (2.5–10 g). The sclerosing agent is typically infused in a solution of 50–100 ml sterile saline. The chest tube is then clamped for 1–2 h and the patient may be rotated several times to facilitate distribution of the sclerosing agent. The chest tube is then opened to 20 cmH20 suction until the 24-h output is less than 150 ml. Sideeffects from pleurodesis include chest discomfort (7–40%) and fever (16–31%). Success rates are between 72 and 93%.47 Patients who develop recurrent pleural effusions despite pleurodesis usually have loculated effusions that are identified on decubitus plain films or chest CT. Loculated effusions often prevent further palliation by thoracentesis or simple chest tube placement. Depending on the patient’s condition and life expectancy, video-assisted thorascopic surgery may be considered. The procedure allows direct visualization of the pleural space. Adhesions and loculations are sharply taken down and adequate lung re-expansion

c

Figure 13.12 Radiographically guided chest tube placement and pleurodesis. (a) Large right pleural effusion in a patient with progressive ovarian cancer. (b) A pigtail catheter has been placed into the right pleural space under computed tomography guidance. (c) Following talc pleurodesis, the right lung is clear

371

Ch 13 final.qxd

12/18/2006

15:34

Page 372

SURGERY FOR OVARIAN CANCER

assured, followed by thoracoscopic pleurodesis. Thoracoscopic management of malignant effusions has a success rate around 90%.52

6.

Solomon HJ, Atkinson KH, Coppleson JV, et al. Bowel complications in the management of ovarian cancer. Aust NZ J Obstet Gynaecol 1983; 23: 65–8

7.

Krebs HB, Goplerud DR. Surgical management of bowel obstruction in advanced ovarian carcinoma. Obstet Gynecol 1983; 61: 327–30

8.

Piver MS, Barlow JJ, Lele SB, et al. Survival after ovarian cancer induced intestinal obstruction. Gynecol Oncol 1982; 13: 44–9

9.

Larson JE, Podczaski ES, Manetta A, et al. Bowel obstruction in patients with ovarian carcinoma: analysis of prognostic factors. Gynecol Oncol 1989; 35: 61–5

10.

Castaldo TW, Petrilli ES, Ballon SC, et al. Intestinal operations in patients with ovarian carcinoma. Am J Obstet Gynecol 1981; 139: 80–4

11.

Lund B, Hansen M, Lundvall F, et al. Intestinal obstruction in patients with advanced carcinoma of the ovaries treated with combination chemotherapy. Surg Gynecol Obstet 1989; 169: 213–18

12.

Clarke-Pearson DL, DeLong ER, Chin N, et al. Intestinal obstruction in patients with ovarian cancer. Variables associated with surgical complications and survival. Arch Surg 1988; 123: 42–5

13.

Maglinte DD, Kelvin FM, O’Connor K, et al. Current status of small bowel radiography. Abdom Imaging 1996; 21: 247–57

14.

Daneshmand S, Hedley CG, Stain SC. The utility and reliability of computed tomography scan in the diagnosis of small bowel obstruction. Am Surg 1999; 65: 922–6

15.

Ericksen AS, Krasna MJ, Mast BA, et al. Use of gastrointestinal contrast studies in obstruction of the small and large bowel. Dis Colon Rectum 1990; 33: 56–64

16.

Anderson CA, Humphrey WT. Contrast radiography in small bowel obstruction: a prospective, randomized trial. Mil Med 1997; 162: 749–52

17.

Shrake PD, Rex DK, Lappas JC, et al. Radiographic evaluation of suspected small bowel obstruction. Am J Gastroenterol 1991; 86: 175–8

18.

DiSantis DJ, Ralls PW, Balfe DM, et al. The patient with suspected small bowel obstruction: imaging strategies. American College of Radiology. ACR

CONCLUSION The natural history of advanced ovarian cancer dictates that many patients will ultimately experience symptomatic progression of disease, which may manifest as visceral obstruction or accumulation of malignant effusion. The surgeon caring for patients with ovarian cancer must have an intimate knowledge of the pathophysiology of these processes, the surgical options available to palliate symptoms, as well as their associated complications. The application of palliative surgery for symptomatic ovarian cancer must be individualized, considering each patient’s specific condition and goals combined with a careful assessment of the potential benefits and risks of surgical intervention.

REFERENCES 1.

2.

3.

4.

5.

372

Tunca JC, Buchler DA, Mack EA, et al. The management of ovarian-cancer-caused bowel obstruction. Gynecol Oncol 1981; 12: 186–92 Dvoretsky PM, Richards KA, Angel C, et al. Survival time, causes of death, and tumor/treatment-related morbidity in 100 women with ovarian cancer. Hum Pathol 1988; 19: 1273–9 Rubin SC, Hoskins WJ, Benjamin I, et al. Palliative surgery for intestinal obstruction in advanced ovarian cancer. Gynecol Oncol 1989; 34: 16–19 Clarke-Pearson DL, Chin NO, DeLong ER, et al. Surgical management of intestinal obstruction in ovarian cancer. I. Clinical features, postoperative complications, and survival. Gynecol Oncol 1987; 26: 11–18 Redman CW, Shafi MI, Ambrose S, et al. Survival following intestinal obstruction in ovarian cancer. Eur J Surg Oncol 1988; 14: 383–6

Ch 13 final.qxd

12/18/2006

15:34

Page 373

PALLIATIVE SURGERY FOR OVARIAN CANCER

Appropriateness Criteria. Radiology 2000; 215 (Suppl): 121–4 19.

30.

Harris GJ, Senagore AJ, Lavery IC, et al. The management of neoplastic colorectal obstruction with colonic endolumenal stenting devices. Am J Surg 2001; 181: 499–506

31.

Carter J, Valmadre S, Dalrymple C, et al. Management of large bowel obstruction in advanced ovarian cancer with intraluminal stents. Gynecol Oncol 2002; 84: 176–9

32.

Russo P. Urologic emergencies in the cancer patient. Semin Oncol 2000; 27: 284–98

33.

Perioperative total parenteral nutrition in surgical patients. The Veterans Affairs Total Parenteral Nutrition Cooperative Study Group. N Engl J Med 1991; 325: 525–32

Cronan JJ. Contemporary concepts in imaging urinary tract obstruction. Radiol Clin North Am 1991; 29: 527–42

34.

Isbister WH, Elder P, Symons L. Non-operative management of malignant intestinal obstruction. J R Coll Surg Edinb 1990; 35: 369–72

Frohlich EP, Bex P, Nissenbaum MM, et al. Comparison between renal ultrasonography and excretory urography in cervical cancer. Int J Gynaecol Obstet 1991; 34: 49–54

35.

Soper JT, Blaszczyk TM, Oke E, et al. Percutaneous nephrostomy in gynecologic oncology patients. Am J Obstet Gynecol 1988; 158: 1126–31

36.

Dudley BS, Gershenson DM, Kavanagh JJ, et al. Percutaneous nephrostomy catheter use in gynecologic malignancy: M.D. Anderson Hospital experience. Gynecol Oncol 1986; 24: 273–8

37.

Donat SM, Russo P. Ureteral decompression in advanced nonurologic malignancies. Ann Surg Oncol 1996; 3: 393–9

38.

Shekarriz B, Shekarriz H, Upadhyay J, et al. Outcome of palliative urinary diversion in the treatment of advanced malignancies. Cancer 1999; 85: 998–1003

39.

Tamsma JT, Keizer HJ, Meinders AE. Pathogenesis of malignant ascites: Starling’s law of capillary hemodynamics revisited. Ann Oncol 2001; 12: 1353–7

40.

Barnett TD, Rubins J. Placement of a permanent tunneled peritoneal drainage catheter for palliation of malignant ascites: a simplified percutaneous approach. J Vasc Interv Radiol 2002; 13: 379–83

41.

Zanon C, Grosso M, Apra F, et al. Palliative treatment of malignant refractory ascites by positioning of Denver peritoneovenous shunt. Tumori 2002; 88: 123–7

42.

Jackson GL, Blosser NM. Intracavitary chromic phosphate (32P) colloidal suspension therapy. Cancer 1981; 48: 2596–8

Furukawa A, Yamasaki M, Furuichi K, et al. Helical CT in the diagnosis of small bowel obstruction. Radiographics 2001; 21: 341–55

20.

Megibow AJ, Balthazar EJ, Cho KC, et al. Bowel obstruction: evaluation with CT. Radiology 1991; 180: 313–8

21.

Krouse RS, McCahill LE, Easson AM, et al. When the sun can set on an unoperated bowel obstruction: management of malignant bowel obstruction. J Am Coll Surg 2002; 195: 117–28

22.

23.

24.

25.

26.

27.

Scheidbach H, Horbach T, Groitl H, et al. Percutaneous endoscopic gastrostomy/jejunostomy (PEG/PEJ) for decompression in the upper gastrointestinal tract. Initial experience with palliative treatment of gastrointestinal obstruction in terminally ill patients with advanced carcinomas. Surg Endosc 1999; 13: 1103–5 Gemlo B, Rayner AA, Lewis B, et al. Home support of patients with end-stage malignant bowel obstruction using hydration and venting gastrostomy. Am J Surg 1986; 152: 100–4 Campagnutta E, Cannizzaro R. Percutaneous endoscopic gastrostomy (PEG) in palliative treatment of non-operable intestinal obstruction due to gynecologic cancer: a review. Eur J Gynaecol Oncol 2000; 21: 397–402 Abu-Rustum NR, Barakat RR, Venkatraman E, et al. Chemotherapy and total parenteral nutrition for advanced ovarian cancer with bowel obstruction. Gynecol Oncol 1997; 64: 493–5

28.

Scolapio JS. Short bowel syndrome. J Parenter Enteral Nutr 2002; 26 (Suppl): S11–16

29.

Bundrick T, Cho SR, Ammann A, et al. Intraperitoneal carcinomatosis: incidence of its radiographic findings and description of a new sign. Br J Radiol 1983; 56: 13–16

373

Ch 13 final.qxd

12/18/2006

15:34

Page 374

SURGERY FOR OVARIAN CANCER

43.

Currie JL, Gall S, Weed JC Jr, et al. Intracavitary Corynebacterium parvum for treatment of malignant effusions. Gynecol Oncol 1983; 16: 6–14

44.

Bezwoda WR, Seymour L, Dansey R. Intraperitoneal recombinant interferon-alpha 2b for recurrent malignant ascites due to ovarian cancer. Cancer 1989; 64: 1029–33

45.

Rodriguez-Panadero F, Borderas Naranjo F, Lopez Mejias J. Pleural metastatic tumours and effusions. Frequency and pathogenic mechanisms in a postmortem series. Eur Respir J 1989; 2: 366–9

46.

Antunes G, Neville E. Management of malignant pleural effusions. Thorax 2000; 55: 981–3

47.

Antony VB, Loddenkemper R, Astoul P, et al. Management of malignant pleural effusions. Eur Respir J 2001; 18: 402–19

48.

Martensson G, Pettersson K, Thiringer G. Differentiation between malignant and non-malig-

374

nant pleural effusion. Eur J Respir Dis 1985; 67: 326–34 49.

Parker LA, Charnock GC, Delany DJ. Small bore catheter drainage and sclerotherapy for malignant pleural effusions. Cancer 1989; 64: 1218–21

50.

Morrison MC, Mueller PR, Lee MJ, et al. Sclerotherapy of malignant pleural effusion through sonographically placed small-bore catheters. AJR Am J Roentgenol 1992; 158: 41–3

51.

Clementsen P, Evald T, Grode G, et al. Treatment of malignant pleural effusion: pleurodesis using a small percutaneous catheter. A prospective randomized study. Respir Med 1998; 92: 593–6

52.

Viallat JR, Rey F, Astoul P, et al. Thoracoscopic talc poudrage pleurodesis for malignant effusions. A review of 360 cases. Chest 1996; 110: 1387–93

Ch 14 final.qxd

12/18/2006

15:35

Page 375

CHAPTER 14

Postoperative management Lee-may Chen

INTRODUCTION The postoperative care of the low-risk ovarian cancer patient may differ little from the postoperative care of women undergoing abdominopelvic surgery for nonmalignant conditions. In general, early ambulation and pulmonary toilet are encouraged, with careful attention to diet and fluid management. On the other hand, patients with advanced-stage disease are often sicker than the typical gynecologic oncology patient. These women tend to be older, have more medical comorbidities, be nutritionally depleted and require more extensive surgical procedures for optimal resection of their disease, and thus are at a higher risk of infection and other postoperative complications. In patients with advanced-stage ovarian cancer undergoing tumor cytoreduction surgery, proper preoperative planning and intraoperative management can optimize hemodynamic conditions for the anticipated postoperative fluid shifts. Particularly in the presence of ascites, both mechanical bowel preparation and prolonged fasting prior to surgery contribute to an already depleted intravascular state. During surgery, several liters of ascites may be removed, and peritoneal surfaces may be stripped of tumor, exposing bowel and peritoneal surfaces for prolonged lengths of time. Adequate intravenous access and monitoring are essential to replace fluid deficits appropriately and maintain intravascular volume. In addition to thirdspace losses, intraoperative blood loss may be significant, requiring transfusion for correction of anemia and/or coagulopathy. Central hemodynamic monitoring and intensive care unit (ICU) admission are useful

in selected patients. Perioperative neuraxial blockade with epidural analgesia has been suggested to decrease intraoperative anesthetic requirements and promote early return of bowel function. Selected patients may benefit from perioperative β-blockade to decrease cardiac co-morbidity. Optimal pain and symptom management are also essential to promoting a timely and comfortable recovery.

INTENSIVE CARE The ICU is an essential resource for the management of the most critically ill patients following ovarian cancer surgery. Depending on the acuity of patients and aggressiveness of the surgeon, the incidence of ICU utilization for patients following ovarian cancer surgery ranges from 20 to 30%. In a population-based analysis, patient age and ICU stay were the most prominent predictors of length of hospital stay as well as total charges for ovarian cancer surgery.1 A multivariate analysis of ovarian cancer patients admitted to the ICU for short (less than 24 h) versus longer stays found that the patient’s preoperative medical condition was less important than perioperative factors in utilization of ICU resources. Patients who had required bowel resection and placement of a pulmonary artery catheter, and remaining intubated postoperatively, were most likely to require ICU care.2 Severity of illness by the Acute Physiology and Chronic Health Evaluation (APACHE II) classification system has also been correlated with survival of critically ill gynecologic oncology patients.3

375

Ch 14 final.qxd

12/18/2006

15:35

Page 376

SURGERY FOR OVARIAN CANCER

Identification of patients who may be at greatest risk for needing ICU care allows for appropriate anesthesia and perioperative planning to optimize care and outcomes for the patients. The ICU must be a collaborative setting, involving interactions between the surgeon, critical caretrained physicians, subspecialty consultants, nurses, respiratory therapists, nutritionists, physical therapists and other support staff, all working together for the common goals of the patient. Particular ICUs at individual hospitals may differ in organization; whether they are run by surgeons, anesthesiologists, or pulmonologists, all ICU environments are resource intensive and emotionally intensive, particularly in this age of continued technologic advancements in critical care medicine. While the surgeon may know the acute issues of the patient and her family best, the intensivist may be able to add perspective to goals of care, particularly if there are conflicts in the level of care being provided to critically ill patients who may be facing the end of life.4 Primary care providers are also important figures in this dynamic, particularly if they have longstanding relationships with the patient and her family. Regardless of whether this is an ‘open’ model, where the surgical team continues to provide the primary service, or a ‘closed’ model, where the ICU team provides the primary service while in the ICU, the involvement of an ICU physician can help co-ordinate patient care and improve patient outcome.5

CARDIAC COMPLICATIONS The stresses and hemodynamic shifts related to ovarian cancer surgery may have a significant impact on the cardiovascular system. Cardiac disease is a common co-morbidity, and may lead to postoperative events including ischemia and arrhythmia. Recognizing patients at risk perioperatively allows for prompt evaluation and intervention when problems arise.

376

Clinical evaluation of patients undergoing noncardiac surgery includes a review of systems to evaluate whether patients are at significant risk for coronary artery disease. The Goldman Cardiac Risk Index based on nine risk factors, as well as the subsequent Revised Cardiac Risk Index based on six independent predictors of cardiac complications, are both only estimates of risk.6,7 Exercise tolerance also helps predict perioperative risk and the level of invasive monitoring needed. The use of neuraxial blockade or epidural anesthesia may reduce the stress response of surgery and further reduce the incidence of ischemic cardiac events.8 Cardiac events are estimated to account for more than 50% of perioperative deaths. In particular, tachycardia is probably the single most common hemodynamic abnormality associated with ischemia. An increase in heart rate results in both increased myocardial oxygen consumption and decreased oxygen delivery by reducing time for coronary artery perfusion during diastole. An imbalance between myocardial oxygen supply and demand results in ischemia. The incidence of myocardial infarction after non-cardiac surgery in patients with ischemic heart disease is as high as 5–6%, with the greatest risk occurring within 2 days of surgery.9 Chest pain is not a typical symptom, and the electrocardiogram (ECG) may not necessarily show Q-wave changes. The initiation of perioperative β-adrenergic receptor blockade (atenolol or metoprolol) for 1 week up to 1 month in patients with coronary artery disease or at risk for coronary artery disease has been associated with decreased incidence of perioperative myocardial infarction, a reduction in associated mortality, and other long-term benefits.10,11 Patients undergoing intra-abdominal surgery with cardiac risk criteria of ischemic heart disease, cerebrovascular disease, diabetes mellitus requiring insulin, or renal insufficiency (serum creatinine level ≥ 2.0 mg/dl) should be considered for the use of perioperative β-blockers. Patients with minor clinical criteria should also be considered for perioperative therapy (Table 14.1). The medication dose may be titrated to lower the heart rate to

Ch 14 final.qxd

12/18/2006

15:35

Page 377

POSTOPERATIVE MANAGEMENT

65 beats/min or less, carefully monitoring the blood pressure. Aggressive therapy in the first 48–72 h after surgery is likely to provide maximal benefit, although intravenous medications should be judiciously administered if oral intake is limited. Patients who are taking anti-hypertensive medications preoperatively should continue on these drugs as able, with careful follow-up of their blood pressure as affected by perioperative pain and fluid management. Hypotension and tachycardia may also be related to a relative hypovolemic state. Despite aggressive intraoperative fluid resuscitation and tumor cytoreduction, ovarian cancer patients may re-accumulate ascites and ‘third-space’ fluid for 2–3 days after surgery. A low serum albumin level and oncotic pressure may permit continued intravascular depletion. In critically ill patients, vasopressors may be initiated, but the underlying hypovolemia must also be corrected before the patient will improve. Spontaneous mobilization of extravascular fluid normally begins by the third or fourth postoperative day; however, some patients may require initial diuretic therapy to facilitate this process when renal function is suboptimal or clinically significant fluid overload is present. Central venous pressure monitoring may be helpful in assessing the patient’s volume status, although pulmonary

Table 14.1 Criteria for use of perioperative β-blockers. Adapted from reference 10 Major surgery (intraperitoneal, intrathoracic, supra-inguinal vascular) Plus one of the following Ischemic heart disease (angina, prior myocardial infarction) Cerebrovascular disease Diabetes mellitus requiring insulin Renal insufficiency (serum creatinine level ≥ 2.0 mg/dl) Or two minor clinical criteria Age ≥ 65 years Hypertension Smoking Serum cholesterol level ≥ 240 mg/dl Diabetes mellitus not on insulin

artery catheters are rarely needed in the absence of significant cardiac disease or pulmonary hypertension. Increasingly, echocardiography is becoming a less invasive modality to evaluate left ventricular function and vascular pathology. Shock is a state of inadequate tissue perfusion, which may be categorized according to the predisposing clinical condition: hypovolemic, cardiogenic, neurogenic, septic, or anaphylactic.12 Symptoms include hypotension, tachycardia, hypoxia, decreased cardiac index, low urine output and peripheral vasoconstriction. Hemodynamic monitoring can help differentiate the different etiologies of shock13,14 (Table 14.2). After surgery, the most common hemodynamic condition is hypovolemia. Treatment to improve oxygen delivery and decrease oxygen consumption usually centers around volume replacement, keeping the patient warm, correcting coagulopathy and controlling pain. Replacement of fluids can be with either crystalloid or colloid solutions.15,16 Despite the hypoalbuminemia in many advanced ovarian cancer patients, there is no evidence to support the use of albumin in fluid replacement.17 Bolus of fluids rather than continuous infusion is preferable to allow for evaluation of response. The response of a hemodynamic parameter such as central venous pressure (CVP) to a fluid bolus may provide information about the volume status as well – if the CVP remains unchanged after a fluid bolus, the patient is still likely to be intravascularly depleted. Vasopressor agents may act through increasing myocardial contractility (inotrope or chronotrope), or by increasing systemic vascular resistance. Caution should be used with dobutamine if the hypotension is not due to cardiogenic shock. Intermediate-dose dopamine and epinephrine (adrenaline) have the effects of both an inotrope/chronotrope and a peripheral vasoconstrictor. Phenylephrine and norepinephrine (noradrenaline) are vasoconstrictors that increase systemic vascular resistance and thus improve blood pressure. Both systemic and individual organ perfusion must be monitored in the pharmacologic management of shock. Hemorrhagic shock is fully reversible with

377

Ch 14 final.qxd

12/18/2006

15:35

Page 378

SURGERY FOR OVARIAN CANCER

Table 14.2 Hemodynamic parameters of shock. Adapted from reference 14 Pulmonary capillary wedge pressure

Cardiac output

Systemic vascular resistance

Hypovolemic

Low

Low

High

Cardiogenic

High

Low

Low

Sepsis (early)

Low

High

Low

Sepsis (late)

High

Low

High

Normal hemodynamic parameters: systemic arterial pressure, 100–140/60–90 mmHg; mean systemic arterial pressure (MAP) = [systolic pressure + (2 × diastolic pressure)/3], 70–100 mmHg; right atrial pressure (CVP), 0–6 mmHg; pulmonary artery pressure, 15–30/5–13 mmHg; mean pulmonary artery pressure, 10–18 mmHg; mean pulmonary capillary wedge pressure, 2–12 mmHg; cardiac output (CO) = stroke volume × heart rate, 3–7 l/min; cardiac index = cardiac output/body surface area, 2.5–4.5 l/min per m2; systemic vascular resistance = (MAP – mean CVP) × 80/CO, 800–1200 dynes/s per cm–5

prompt replacement of circulating volume and oxygen-carrying capacity. Prolonged shock may trigger a cascade of local and systemic cytokines, resulting in a systemic inflammatory response syndrome, which may eventually lead to multiple organ failure. Myocardial ischemia may manifest as anginal pain or be largely asymptomatic with only ECG changes. Tachycardia and elevated catecholamines are a common response to surgery, therefore control of pain and anxiety may help decrease the incidence of postoperative cardiac events. Prompt recognition and treatment of ischemia allows for the prevention of myocardial infarction and its sequelae, including dysrhythmias, congestive heart failure and death. Immediate treatment of suspected cardiac ischemia includes the rapid administration of supplemental oxygen, nitrates to decrease myocardial demand by venodilation and β-blockers to decrease heart rate and contractility. Serial evaluation of ECG changes and measurement of serum troponin I can reflect the evolution or resolution of myocardial ischemia. In surgical patients, use of anticoagulation such as heparin must be considered with respect to the risk of inducing postoperative bleeding. Aspirin can usually be started in the postoperative period, as the benefits outweigh the potential risks. Calcium channel blockers may be useful in the setting of reperfusion after ischemia. Antiplatelet drugs are also to be considered at the discretion of the managing surgeon and cardiologist. An echo-

378

cardiography study may also evaluate myocardial function to determine the role of additional cardiology interventions. Sinus tachycardia is common in the postoperative ovarian cancer patient who may be volume depleted or experiencing pain. Treatment involves addressing the underlying physiologic condition. Supraventricular tachyarrhythmias include atrial fibrillation, atrial flutter and other sinus node re-entrant tachycardias.18 Electrocardiographic distinction is often subtle for narrow complex tachycardia, and management is often to treat the hemodynamic impact of the rhythm disturbance by controlling the rapid ventricular rate, rule out ischemia and look for underlying causes. Paroxysmal supraventricular tachycardia can be managed first with vagal maneuvers or adenosine for diagnosis. Atrial fibrillation or atrial flutter should be managed first by controlling the ventricular response rate. In a hemodynamically stable patient, calcium channel blockers, β-blockers, digoxin, or amiodarone can each be used for rate control. Low doses of electrical energy through synchronized cardioversion may also override and reset the abnormal pulse generator. Cardioversion would be the first-line treatment for the hemodynamically unstable patient. Atrial fibrillation in the postoperative patient may represent underlying cardiac disease, but is often paroxysmal with spontaneous conversion to sinus rhythm within 24 h. Echocardiography studies may be

Ch 14 final.qxd

12/18/2006

15:35

Page 379

POSTOPERATIVE MANAGEMENT

used to investigate thrombus as well as other structural etiologies, to determine the role of additional cardiology interventions.

PULMONARY COMPLICATIONS Any surgical procedure requiring intubation for general anesthesia increases the risk of pulmonary complications. The presence of an acute respiratory condition creates significant concerns in the perioperative patient. Acute infections should be treated before surgery in most non-emergency situations. Patients with high-risk conditions including asthma, bronchitis, emphysema, or smoking, should be optimized for their medical condition prior to surgery, if possible.19 Even in the healthiest of patients, the preoperative evaluation for ovarian cancer surgery will typically include a chest X-ray for staging. Common complications after abdominal surgery include atelectesis, pneumonia, aspiration and pulmonary edema. Large malignant pleural effusions may be drained to optimize respiratory function. Similar to the cardiac preoperative risk indices, pulmonary multifactorial risk indices have been developed and validated to identify patients at increased risk for postoperative pneumonia, so that appropriate respiratory interventions can be instituted.20 Age, poor functional status, upper abdominal surgery, general anesthesia, chronic obstructive pulmonary disease, transfusion, steroid use and smoking all contribute to perioperative pulmonary risks.21 Upper abdominal surgery is associated with diaphragm dysfunction and is more likely to result in pulmonary complications.22 The use of neuraxial blockade or epidural analgesia may be advantageous in managing pain while minimizing opiates to facilitate early ambulation and pulmonary toilet. At the end of surgery, most patients are awakened from anesthesia and extubated. In cases of largevolume fluid resuscitation or prolonged operative time, extubation may be delayed until the patient is in the recovery room or ICU. Laryngeal edema increases the likelihood of airway obstruction. In addition,

bowel edema may increase the intra-abdominal pressure, which in turn limits respiratory excursion and functional residual capacity. Mechanical ventilation supports respiration by delivering positive pressure through either volume control or pressure control. Intermittent mandatory ventilation delivers a set rate and volume, with unassisted spontaneous breaths allowed in between. In pressure-controlled ventilation, the volume delivered is determined by a pre-set level of pressure; the patient’s inspiratory effort triggers the ventilator to deliver the breath. Delivering positive end-expiratory pressure (PEEP) or continuous positive airway pressure (CPAP) provides support to keep open previously collapsed alveoli and reduces the overall work of breathing. For patients who remain intubated following surgery because of airway concerns, the management of mechanical ventilation involves supportive care until a trial of weaning can be performed (Table 14.3). The fraction of inspired oxygen (FiO2) should be weaned to minimize injury from oxygen radicals, and the peak inspiratory pressure should be kept low to minimize barotrauma. As long as oxygen and carbon dioxide exchange is adequate, the patient may be extubated after airway edema resolves, often the day following surgery. Pulmonary toilet is emphasized in the postoperative period by encouraging incentive spirometry and early ambulation as able. Atelectasis results in retention of bronchial secretions, may be a source of fever and increases the likelihood of pneumonia. Aspiration pneumonitis without infection is treated by supportive measures. Pneumonia in the postoperative period is a common cause of respiratory compromise, a leading cause of nosocomial infection and ICU deaths.23–25 Thoracic or upper abdominal surgery is a significant clinical risk factor, as is a history of respiratory disease, as well as a bedridden status resulting in atelectasis.26,27 In the face of impaired host defenses, bacterial entry into the lower respiratory tract may be from aspiration of oropharyngeal secretions, inhalation of aerosolized bacteria, or bacteremia from other sites. Patients requiring mechanical ventilation for longer than 24 h have a higher risk of nosocomial

379

Ch 14 final.qxd

12/18/2006

15:35

Page 380

SURGERY FOR OVARIAN CANCER

Table 14.3 A typical ventilator management and weaning protocol Mode: intermittent mandatory ventilation Tidal volume: 8–10 ml/kg Rate: 10–14 breaths/min, to keep minute ventilation of 70 ml/kg per min Fraction of inspired oxygen (FiO2): 100%, titrating rapidly downward to keep PaO2 > 80 mmHg, goal of 30–40% Positive end expiratory pressure (PEEP): 5 cmH2O Pressure support: 5 cmH2O Increase minute ventilation for hypercapnea Add FiO2 or PEEP for hypoxemia Weaning parameters are measured when the patient is on minimal ventilatory support, as when the patient is on a continuous positive airway pressure (CPAP) or T-piece trial. These include tidal volume greater than 5 ml/kg, rate less than 25/min, FiO2 less than 40%, negative inspiratory force less than –20 cmH2O. After 20–30 min, consider extubation if the respiratory rate is less than 25 and the pulse is less than 120

pneumonia.28 The microbiology of hospital-acquired pneumonia is significantly different from that of community-acquired pneumonia. Severely ill patients have a higher risk of being colonized with Gramnegative bacilli, and the bacteria are often polymicrobial.29,30 The usual presentation is that of fever, sputum production, pulmonary consolidation on physical examination and a localized infiltrate on X-ray. Particularly in critically ill patients, the clinical picture may vary. Fever or leukocytosis may be from one of several sources. Diagnosis of an infiltrate on chest X-ray may be confounded by the presence of atelectasis, or a malignant pleural effusion. Gram-stain is more valid as an index of pulmonary infection, as most cultures will reveal airway flora. Therapy should be directed at the underlying cause, keeping in mind the antibiotic used for surgical prophylaxis as one to which the organism is presumably resistant. Thirdgeneration cephalosporins in combination with an aminoglycoside, or fourth-generation cephalosporins are effective regimens for hospital/ventilator-acquired pneumonia, as are β-lactamase inhibitors or fluoroquinolones. Chronic obstructive pulmonary disease (COPD) is usually smoking related, and may range from mild to severe. Clinical symptoms may not develop for 10–20 years, after irreversible lung damage has already occurred. As in asthma, chronic inflammation plays a

380

significant role in COPD, but with different inflammatory mediators, proteases and oxidative stresses. Bronchodilators remain the mainstay of therapy, with a role for inhaled steroids and antibiotics in acute exacerbations.31 The ventilator management of COPD should be modified to mitigate the risk of dynamic hyperinflation. Because of reduced elastic recoil, alveoli are prone to overdistension with early airway collapse, causing air trapping and inadvertent positive end-expiratory pressure (‘autoPEEP’). Without a longer exhalation time, the positive airway pressure can accumulate, resulting in hyperinflation of the lungs, barotrauma and hemodynamic compromise. Extrinsic PEEP may be helpful to decrease circuit pressure, and setting a low ventilator sensitivity allows for easier triggering of breaths. Acute respiratory distress syndrome (ARDS) is unusual after ovarian cancer surgery, although it may occur under particularly stressful circumstances. Predisposing causes include direct lung injury from aspiration or pneumonia, as well as indirect phenomena of massive fluid shifts, coagulopathy, transfusion and sepsis. In 1994, a consensus conference recognized the definition of ARDS as a condition with acute onset, bilateral infiltrates on chest X-ray, pulmonary-artery wedge pressure of < 18 mmHg (if available) in the absence of clinical evidence of left atrial hypertension, and a gradient of partial pressure

Ch 14 final.qxd

12/18/2006

15:35

Page 381

POSTOPERATIVE MANAGEMENT

of arterial oxygen (PaO2) to FiO2 of less than 200.32 A PaO2 to FiO2 ratio of less than 300 is considered acute lung injury. Treatment consists of supportive care, searching for an underlying cause, treating underlying nosocomial infection and preventing the development of multi-system organ failure. Anatomically, ARDS can be a heterogeneous process, with some areas of the lung remaining normal while other areas become poorly compliant. A positive pressure breath with go preferentially to normal lung; overdistension of more compliant normal lung zones can then result in greater stretch injury. Ventilator strategies include using lower tidal volumes and higher PEEP, avoiding overdistension of normal alveoli. A phase III study by the Acute Respiratory Distress Syndrome Network demonstrated a 22% decrease in mortality with this strategy compared to the use of traditional tidal volumes (6 ml/kg vs. 12 ml/kg).33 Restricting fluids may decrease pulmonary edema, but studies have been not shown a mortality benefit. The use of neither glucocorticoids nor surfactant has shown any outcome benefit.34

HEMATOLOGIC COMPLICATIONS Coagulopathy and cancer may be associated with thromboembolic or hemorrhagic phenomena. In a series of patients undergoing primary ovarian cytoreduction surgery at Cedars-Sinai Medical Center, 6.7% of patients developed a perioperative coagulopathy.35 Abnormal coagulation studies were identified before significant dilution or transfusion had occurred. Ascites volume, blood loss, hypoalbuminemia and a large burden of metastatic disease were associated with an increased risk for perioperative coagulopathy. Laboratory studies after ovarian cancer surgery with significant blood loss should include prothrombin and activated thromboplastin time. Additional screening tests for disseminated intravascular coagulation may include quantitative fibrinogen levels, D-dimers and platelet counts.

Packed red blood cells are transfused for anemia and to increase the oxygen-carrying capacity of the patient36,37 (Table 14.4). Fresh frozen plasma (FFP) is separated from whole blood and frozen within 6 h of collection to minimize loss of coagulation factors V and VIII. FFP is indicated for coagulopathy from transfusion as well as correction or warfarin effect. Platelets are usually pooled from six units of whole blood and transfused to treat thrombocytopenia or deficits of platelet function. Typically, spontaneous bleeding is uncommon with platelet levels about 20 000/dl, although patients undergoing major operative procedures should generally have platelet counts about 70 000/dl for effective hemostasis. Cryoprecipitated antihemophilic globulin is also pooled from whole blood and serves as a therapeutic source of fibrinogen. Although cryoprecipitate is rich in factor VIII, the treatment of choice for hemophiliacs currently is factor VIII concentrate. Current blood bank practice involves screening for disease transmission of syphilis, hepatitis B, hepatitis C and human immunodeficiency virus. More recently, activated recombinant factor VIIA has been used in incidental cases of intra-abdominal hemorrhage that fails to respond to conventional therapy with dramatic response by promoting clotting through the extrinsic pathway.38 The major drawback to date is its short half-life of about 2 h and its high cost. Adverse reactions to transfusion may include febrile reactions that may be caused by recipient antibodies to leukocyte antigens reacting to leukocyte fragments in the transfused blood.36,37 A temperature rise of 1.0°C from baseline is considered a fever in this setting. These reactions are most commonly seen in patients with a history of multiple blood transfusions or pregnancies, which can stimulate the development of leukocyte antibodies. Approximately one in eight patients with such reactions will have a similar reaction with subsequent transfusions. Allergic urticarial reactions are seen in approximately one in 100 transfusion recipients. This reaction is probably caused by foreign plasma proteins. Premedication with acetominophen and diphenhydramine may help

381

Ch 14 final.qxd

12/18/2006

15:35

Page 382

SURGERY FOR OVARIAN CANCER

Table 14.4 Contents of blood products Component

Volume (ml)

Content

Response

Packed red blood cells

250–350

50–80% hematocrit Small amount plasma Variable leukocyte content

Increases hematocrit 3%

Whole blood

450

35–45% hematocrit Few platelets Decreased factors V, VIII

Increases hematocrit 3%

Platelets

40–60

5.5 × 1010 platelets Small amount of plasma Usually pooled as 6-packs

Increases platelets 5000–10 000/µl

Single-donor platelets

200–400

3 × 1011 platelets

Increases platelets 7.5 × 109/l

Fresh frozen plasma

200–275

Fibrinogen Factors V, VII, IX, XI Protein C and S Antithrombin III

Increases coagulation factors 2%

Cryoprecipitate

5–15

Fibrinogen Factor VIII von Willebrand’s Factor

Increases factor VIII activity 30–50%

minimize febrile and allergic reactions. Fever is also the most frequent manifestation of acute hemolytic transfusion reactions. These reactions are most likely to occur when a group O patient is mistakenly transfused with group A, B, or AB blood. Symptoms may include fevers, chills, chest tightness, tachycardia, hypotension and hemoglobinemia with subsequent hemoglobinuria and hyperbilirubinemia. If a transfusion reaction is suspected, the transfusion must be stopped immediately and supportive measures undertaken. Fluid and diuretic therapy may be indicated, and blood specimens from the patient and the transfused blood product should be collected to confirm hemolysis. Delayed hemolytic reactions may occur in patients who have developed antibodies from prior transfusion and may not manifest until 4–8 days after transfusion. Such delayed reactions may not be detected, as the red cell destruction occurs slowly, and are diagnosed only with a falling hematocrit and positive direct antiglobulin (Coombs) test. Transfusionrelated acute lung injury (TRALI) is a rare complication of transfusion manifested by abrupt non-

382

cardiogenic pulmonary edema. This reaction is probably associated with the presence of donor antibodies reactive to recipient leukocyte antigens, or with inflammatory mediators in stored blood components. Severe cases may require ventilator support, but usually resolve within 72 h. Because the reaction is most often donor specific, donors who have been implicated in a case of TRALI are usually excused from future blood donation. The popularity of autologous and designated donor blood programs has increased, while the estimated rates of transfusion-related infections has decreased39,40 (Table 14.5). Autologous and designated donor programs vary by institution, but have been proposed for procedures where the mean transfusion requirement exceeds one unit, where more than 10% of patients undergoing the procedure require transfusion, and the anticipated blood loss is greater than 20% of the patient’s estimated blood volume. In most cases, it is probably not cost effective, as surgical blood loss is difficult to predict.

Ch 14 final.qxd

12/18/2006

15:35

Page 383

POSTOPERATIVE MANAGEMENT

Physical examination of the perioperative coagulopathic patient should include an assessment of bleeding from an anatomic site versus continued consumptive coagulopathy. Particularly in extensive ovarian cancer debulking surgery, there are large surfaces of tumor beds where oozing may continue until the coagulation disorder is corrected. Inadequate surgical hemostasis may be considered if the patient’s condition appropriately responds to transfusion but then deteriorates again. If the physical examination findings suggest a hematoma in a confined space, the bleeding should be self-limited; transfusion of blood products is continued as necessary to support hemodynamic parameters. If the physical examination is consistent with continued hemorrhage, such as from a bleeding pedicle (increasing tense abdomen, gross blood in an abdominal drain), re-exploration may be necessary to control hemostasis.

THROMBOEMBOLIC COMPLICATIONS Deep vein thrombosis and pulmonary embolism continue to be a leading cause of postoperative morbidity and mortality in surgery. Patients undergoing gynecologic oncology surgery, in particular, often have multiple risk factors for a thromboembolic event including the presence of malignancy, older age, longer surgical

procedures, limitation with mobility after abdominal surgery, and frequent dissection around pelvic vessels. In a prospective study of gynecologic cancer patients undergoing major radical surgery, the incidence of deep vein thrombosis detected by 125I-fibrinogen scanning was 38%, the majority of which were evident within 24 h of surgery.41 Only 10% of deep vein thromboses were clinically evident, and bilateral disease was present in 20% of cases. The clinical diagnosis of pulmonary embolism is equally inaccurate, with 70% of patients with fatal pulmonary embolism being diagnosed at autopsy.42 A high clinical index of suspicion for thromboembolic events is critical for early detection and timely therapeutic intervention. Some patients will have a pre-existing diagnosis of thromboembolism, or have a mechanical heart valve in place, and be on anticoagulation therapy prior to surgery. As there is a risk of hypercoagulability with discontinuation of warfarin, patients at significantly high risk for thrombosis should be transitioned to heparin before and after surgery.43 The International Normalized Ratio (INR) should fall to less than 1.3–1.5 before surgery. Intravenous heparin should be stopped 6 h prior to incision. Elective surgery should be avoided in the first month after acute venous thromboembolism. Beyond that, a vena caval filter should be considered in situations of acute venous thromboembolism, or if the risk of bleeding on intra-

Table 14.5 Risks of blood transfusion. Adapted from reference 39 Risk factor

Estimated frequency per unit

Deaths per million units

Hepatitis A

1/1 000 000

0

Hepatitis B

1/30 000–1/250 000

0–0.14

Hepatitis C

1/30 000–1/150 000

0.5–17

HIV

1/200 000–1/2 000 000

0.5–5

Bacterial contamination

1/12 000–1/500 000

0.1–21

Acute hemolytic reaction

1/250 000–1/1 000 000

0.67

Delayed hemolytic reaction

1/1000

0.4

Transfusion-related acute lung injury

1/5000

0.4

383

Ch 14 final.qxd

12/18/2006

15:35

Page 384

SURGERY FOR OVARIAN CANCER

Table 14.6 Prevention of deep vein thrombosis (DVT) after gynecologic surgery. Adapted from reference 44 Regimen Untreated control patients

No. trials

No. patients

DVT incidence (%)

95% CI

12

945

16

14–19

Oral anticoagulants

5

183

13

8–18

Pneumatic compression

3

253

9

6–13

11

1092

7

6–9

1

104

0

0–3

Low-dose unfractionated heparin Elastic stockings

venous heparin is high. Intravenous heparin should be restarted without bolus no sooner than 12 h after surgery, and potentially longer if there is continued concern for surgical bleeding. Patients with acute venous thromboembolism after the first month may not require intravenous heparin before surgery, but should be started on intravenous heparin as able, before restarting warfarin. Prophylactic interventions can decrease the incidence of deep vein thrombosis by 50% (16% to 8%), and fatal pulmonary embolism by 75% (0.4% to 0.1%)44 (Table 14.6). Consensus panel guidelines from the American College of Chest Physicians recommend routine prophylaxis of gynecologic surgery patients with three daily doses of low-dose unfractionated heparin. Alternatively, low-dose unfractionated heparin may be combined with pneumatic compression devices, or prophylactic doses of lowmolecular-weight heparin may be given (Table 14.7). Several prospective clinical trials at Duke University Medical Center have demonstrated that low-dose heparin or pneumatic calf compression are both effective means of reducing the incidence of deep vein thrombosis in patients following gynecologic surgery.45–48 A randomized trial of 208 patients comparing unfractionated heparin to external pneumatic compression did not demonstrate any clinically significant benefit to heparin prophylaxis.46 Five thousand units of unfractionated heparin administered subcutaneously every 8 h, from three doses prior to surgery until postoperative day 7, was associated with a deep vein thrombosis incidence of 6.5%, com-

384

Table 14.7 Guidelines from the American College of Chest Physicians for thromboprophylaxis in gynecologic oncology patients. Adapted from reference 44 Low-dose unfractionated heparin 5000 units every 8 h Low-dose unfractionated heparin 5000 units every 12 h with elastic stockings or pneumatic compression Low-molecular-weight heparin daily or every 12 h The duration of thromboprophylaxis remains unresolved

pared to the pneumatic calf compression initiated with the induction of anesthesia and continued for 7 postoperative days, where the incidence of deep vein thrombosis was 4%. Patients receiving heparin did have a higher transfusion rate postoperatively, a significant consideration for patients with ovarian cancer surgery who may be more coagulopathic from their extensive surgery. Compared to unfractionated heparin, lowmolecular-weight heparins have a better anticoagulation profile, based on better bioavailability, longer half-life, dose-independent clearance and decreased binding to plasma proteins and endothelial cells. They have less anti-thrombin activity, more anti-factor Xa activity, and less effect on partial thromboplastin time. They also have less platelet inhibition and do not increase microvascular permeability, and therefore fewer bleeding complications are seen with their use. Patients with renal impairment may have an increase in exposure to the low-molecular-weight

Ch 14 final.qxd

12/18/2006

15:35

Page 385

POSTOPERATIVE MANAGEMENT

heparin enoxaparin, and its dosage should be adjusted downward when the creatinine clearance is less than 30 ml/min. A subsequent randomized trial of 211 gynecologic oncology patients at Duke University Medical Center comparing the low-molecular-weight heparin dalteparin to external pneumatic compression sleeves also showed both strategies to be similarly effective in preventing venous thromboembolic events.47 Patients were randomized to receive either 2500 units of dalteparin subcutaneously every 12 h, from one dose prior to surgery until postoperative day 5, or external pneumatic compression initiated with the induction of anesthesia and continued through postoperative day 5 or discharge. There were no cases of symptomatic pulmonary embolism or deep vein thrombosis in either group; using Doppler ultrasound, the incidence of venous thrombosis in the low-molecular-weight heparin group was 1.9%, while the incidence in the pneumatic compression group was 0.9%. The frequency of bleeding complications and transfusion was similar between the two groups. A study of the role of patient preference has concluded that there is a high level of satisfaction with both methods of prophylaxis, with comparable levels of compliance.48 Despite the efficacy of intermittent pneumatic calf compression, some patients still develop deep vein thrombosis. Through a retrospective review of gynecologic surgery patients, Clarke-Pearson et al. identified risk factors that were most strongly associated with clinically significant postoperative venous thromboembolic events to include: cancer diagnosis, history of deep vein thrombosis and age greater than 60 years.45 Patients with two or three of these risk factors had a 3.2% incidence of developing thromboemboli, as compared with a 0.6% incidence in patients with zero or one risk factor(s). These patients may represent a group where a second prophylactic measure (i.e. heparin) may be added. Combination therapy with unfractionated heparin and pneumatic compression has been used in high-risk patients undergoing general surgery, with lower rates of postoperative deep vein thrombosis. While the combina-

tion of low-molecular-weight heparin and pneumatic compression are often prescribed, the efficacy of this regimen has not yet been proved in a well-designed clinical trial.49 The development of deep vein thromboses and pulmonary emboli share a common pathogenesis. A prothrombotic state in the cancer patient may result from the host inflammatory response to the tumor, and surgical patients will inevitably have endothelial injury triggering coagulation. Consequently, a high index of suspicion must be maintained for venous thromboembolism, even in a patient with other aberrations of the coagulation cascade from hemorrhage or consumption. Venous thromboembolic events may not be clinically apparent or symptoms may be mistakenly attributed to other perioperative processes. Classic clinical findings of deep vein thrombosis include leg swelling, pain, increased warmth and erythema. Doppler ultrasonography is the most commonly used diagnostic procedure to diagnose deep vein thrombosis50,51 (Figure 14.1). Particularly after pelvic surgery, Doppler ultrasound is limited to evaluation of veins below the inguinal ligament. In the situation of a negative or indeterminate study with high clinical suspicion, contrast venography still remains the standard. Magnetic resonance venography may be useful in patients who require evaluation of the pelvis, or are unable to receive iodinated contrast. Negative quantitative enzyme-linked immunosorbent assay (ELISA) D-dimer assays, combined with negative non-invasive imaging, have been found to have a high negative predictive value in out-patients, but these results are not yet confirmed to have validity in cancer patients.52,53 A negative test in a high-risk patient does not exclude pulmonary embolism, and a positive D-dimer assay in a postoperative coagulopathic patient may not be specific either.54 Pulmonary embolism is the most serious potential consequence of deep vein thrombosis. Classic clinical findings include hypoxia, chest pain, hemoptysis, shortness of breath and tachycardia. In a critically ill postoperative ovarian cancer patient, the differential diagnosis is broad and may include fluid overload,

385

Ch 14 final.qxd

12/18/2006

15:35

Page 386

SURGERY FOR OVARIAN CANCER

Suspected DVT

Doppler ultrasound

Positive study

Indeterminate study or high clinical suspicion

Negative study

Treat acute DVT

Venography

No DVT

Positive study

Negative study

Treat acute DVT

No DVT

Figure 14.1 Diagnostic algorithm for evaluation of suspected deep vein thrombosis (DVT)

pneumonia, effusion, atelectasis, or a number of other cardiac events. The most common chest radiography findings include atelectasis, infiltrate, or a small effusion, all of which are non-specific in a perioperative setting.55 Findings on electrocardiography may include ST-segment or T-wave changes, with few patients showing electrical changes suggestive of right heart strain. An arterial blood gas assessment may help determine the degree of hypoxia and hypercapnea, although a quarter of patients with acute pulmonary embolism will have a normal PaO2. A computed tomography (CT) angiogram, or spiral CT, can reliably diagnose most clinically significant pulmonary emboli, but is less sensitive in evaluating the subsegmental vessels (Figure 14.2). Overall sensitivity and specificity rates vary by institution and available technologies. A ventilation/perfusion (V/Q) scan has the advantage of not using iodinated contrast, but has the disadvantage of determining only indirect evidence of

386

pulmonary embolism. In the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study, only 14% of patients had a high probability scan, while 77% of patients had an indeterminate, nondiagnostic study.56 Pulmonary angiography remains the standard of diagnosis for pulmonary embolism, but is considered rather invasive by requiring catheterization of the main pulmonary vessels, and a considerable amount of contrast dye, which can be particularly significant in a patient population whose renal function may be compromised from intravascular depletion or acute tubular necrosis. For a critically ill patient in the ICU, transesophageal echocardiography may also be performed at the bedside to evaluate pulmonary embolism.57 While transesophageal echocardiography provides poor visualization of the left pulmonary and lobar arteries, it is able to identify right ventricular volume and pressure overload associated with pulmonary embolism.

Ch 14 final.qxd

12/18/2006

15:35

Page 387

POSTOPERATIVE MANAGEMENT

Suspected PE

Spiral CT scan

Positive for PE

Indeterminate

Treat acute PE

Doppler ultrasound

Positive study

Normal

Non-PE pathology

If explains all symptoms, no PE, otherwise follow indeterminate pathway

Negative Doppler ultrasound

Treat DVT

Pulmonary angiogram

Positive study

Negative Indeterminate or high suspicion

Treat DVT

No PE diagnosed Pulmonary angiogram

Figure 14.2 Diagnostic algorithm for evaluation of suspected pulmonary embolism (PE). CT, computed tomography; DVT, deep vein thrombosis

Clinically, the most practical examination may be spiral CT pulmonary angiography.58,59 Other lung pathology may also be concurrently diagnosed. Combining spiral CT of the chest with indirect CT venography may also obviate the need for lower extremity Doppler examinations if either a pulmonary embolism or a deep vein thrombosis is identified. Although a contrast dye load is still required, it is not particularly invasive, and allows for evaluation of pelvic and lower extremity thrombi with over 90% sensitivity and specificity.60,61 Heparin remains the primary therapy for postoperative patients with venous thromboembolic events. Failure to rapidly anticoagulate a patient with an

acute event has been associated with late recurrence of thrombosis. On the other hand, the exuberance to anticoagulate a patient must be tempered by the patient’s postoperative condition and likelihood of persistent surgical bleeding or coagulopathy. If an epidural catheter is in place, it should be removed before the initiation of anticoagulation if the patient is stable enough to wait 1 h before starting heparin.62 The short half-life of intravenous unfractionated heparin allows for rapid reversal of anticoagulation in a patient at high risk of bleeding, or requiring an invasive procedure (Figure 14.3). Once the patient is hemodynamically stable and therapeutically anticoagulated with heparin, warfarin can be started for a

387

Ch 14 final.qxd

12/18/2006

15:35

Page 388

SURGERY FOR OVARIAN CANCER

Laboratory investigations prior to initiation of heparin: activated partial thromboplastin time (aPTT), prothrombin time/International Normalized Ratio (PT/INR), platelets, urinalysis, stool guiac Loading dose: 80 units/kg (minimum 5000 units, maximum 10 000 units) Continuous infusion: 18 units/kg per h Obtain first aPTT 4 h after initiation of heparin therapy. If < 55 s, follow table below. If ≥ 55 s, re-check aPTT in 4 h. Adjust rate using table below. Re-check aPTT every 6 h until two consecutive aPTT values are within therapeutic range, then obtain daily complete blood cell count (CBC) with platelets, aPTT

aPTT (s)

Re-bolus (units)

< 40 40–47 48–54

70 U/kg 0 0

Therapeutic range 55–80 81–90 91–100 > 100

0 0 0 0

Hold drip (min)

Change drip (units/h)

Repeat aPTT (h)

0 0 0

+200 +200 +100

6 6 6

0 0 30 60

0 –100 –200 –300

6 until stable, then every morning 6 6 6

Warfarin may be initiated on day 1 of heparin therapy. A minimum of 5 days is recommended. The therapeutic range for warfarin should be an INR of 2.0–3.0. A PT/INR should be ordered daily until a steady-state dose is reached Day

INR

Warfarin dosage (mg)

1

< 1.1

5

2

< 1.5 1.5–1.9 2–2.5 > 2.5

5 2.5 1–2.5 0

3

< 1.5 1.5–1.9 2–3 >3

5–10 1.5–5 0–2.5 0

4

< 1.5 1.5–1.9 2–3 >3

10 5–7.5 0–5 0

5

< 1.5 1.5–1.9 2–3 >3

10 7.5–10 0–5 0

Figure 14.3 Algorithm for anticoagulation of acute deep vein thrombosis (DVT) or pulmonary embolism (PE) with high bleeding risk. Note that these guidelines are recommendations, not intended to replace an individual clinician’s judgment

course of long-term anticoagulation. A patient with active bleeding may benefit from an inferior vena cava filter, placed to prevent new or recurrent pulmonary

388

emboli. While the placement of a filter reduces the acute risk of pulmonary embolism, the filter may still become occluded. Once stable, patients typically

Ch 14 final.qxd

12/18/2006

15:35

Page 389

POSTOPERATIVE MANAGEMENT

require lifelong anticoagulation to minimize their later risk of recurrent deep vein thrombosis.63 Currently, new filters are being developed that may be able to be removed within 2 weeks of placement. These may be useful in patients with acute preoperative deep vein thrombosis and pulmonary emboli. Several meta-analyses of clinical trials comparing low-molecular-weight heparin to unfractionated heparin for the treatment of patients with deep vein thrombosis have demonstrated low-molecular-weight heparin to be advantageous with respect to reduction of thrombus size, as well as to a decrease in recurrent venous thromboembolism.64–66 Low-molecular-weight heparin also appears to be as safe and effective as unfractionated heparin in the management of pulmonary embolism.67 In the early postoperative period, however, intravenous unfractionated heparin may be more appropriate for management of venous thromboembolism, as low-molecular-weight heparin is not reversible if problems with hemostasis occur. Later, the patient can be transitioned to low-molecularweight heparin, especially if they tolerate intravenous heparin without bleeding. Recommended doses of low-molecular-weight heparin depend on whether the indication is prevention or treatment. A prophylactic dose of dalteparin is 5000 IU 10–12 h before surgery, and once daily after surgery.68 A prophylactic dose of enoxaparin is 40 mg 2 h before surgery, and once daily after surgery, while a therapeutic dose for anticoagulation is 1 mg/kg every 12 h. Monitoring is not required for prophylactic doses, nor in most patients for treatment, but measuring an anti-Xa level can be considered for patients with renal insufficiency, morbid obesity, or refractoriness to therapy.69 Heparin-associated thrombocytopenia (HAT) is a non-immune-mediated process that is relatively common (5–30% incidence) within a period of 4 days of a patient starting heparin.70 The etiology is probably direct platelet activation by heparin, resulting in a mild and reversible platelet abnormality that is selflimited without major complications by itself. Platelet counts typically fall to 100 000–150 000 and recover in a matter of days. In contrast, heparin-induced

thrombocytopenia (HIT) is triggered by antibodies directed against complexes of heparin and platelet factor 4, and can result in both venous and arterial thrombosis in 1–3% of patients treated with heparin for 4–14 days.70 Platelet counts typically fall to 20 000–150 000. The diagnosis is predominantly clinical, based on the clinical presentation of thrombocytopenia with or without thrombosis while on heparin. Without treatment, mortality of HIT is as high as 20–30%. Management should include discontinuation of heparin, and instituting an alternative anticoagulant. Platelet transfusion is relatively contraindicated, as this may result in new thromboembolic events. Warfarin should not be used alone acutely, and treatment focuses around direct thrombininhibiting agents such as lepirudin or argatroban. Low-molecular-weight heparin is less likely to result in HIT; however, there is still a high degree of antibody cross-reactivity and its use is probably best avoided in a patient with a history of its use. Arterial thrombotic events are rare after gynecologic cancer surgery but nevertheless require prompt recognition and should be treated as a vascular surgical emergency.71 Hypercoagulability, tissue trauma and patient immobilization are etiologic risk factors that may exacerbate underlying vascular disease. Classic clinical findings in the affected extremity include pain, pulselessness, paresthesia, pallor and paralysis. Limb ischemia may lead to muscle necrosis and compartment syndrome within hours after onset. Prompt diagnosis and rapid revascularization allows the best opportunity for limb salvage and avoidance of amputation.

FLUIDS The daily fluid requirement for an average adult is 2–3 l/day.72 Several formulas are available to estimate maintenance fluid requirements: a simple one includes 4 ml/kg per h for the first 10 kg of body weight, 2 ml/kg per h for the second 10 kg, and 1 ml/kg per h for each subsequent kg in weight.

389

Ch 14 final.qxd

12/18/2006

15:35

Page 390

SURGERY FOR OVARIAN CANCER

Gastrointestinal and urinary losses are adequately replaced with lactated Ringer’s solution or normal saline crystalloid solutions (Tables 14.8 and 14.9). In the postoperative patient, fluid management considerations must include accounting for insensible losses related to the surgical procedure in addition to the usual insensible losses associated with skin, lung and fecal material. Weighing the patient daily is the best means of assessing total body fluid status. Supplemental intravenous fluids may be weaned off as the patient’s oral intake increases. In surgical patients, the degree of fluid shifting into the extracellular ‘third space’ reflects the magnitude of the operative procedure.73 Postoperative sodium retention is a response to a decrease in extracellular volume.74 For a minimally traumatic procedure, fluid replacement during surgery should include 4 ml/kg per h of lactated Ringer’s solution or normal

saline; for a moderately traumatic procedure, 6 ml/kg per h; and for an extremely traumatic procedure such as a large ovarian debulking of a patient with ascites, 8 ml/kg per hour. With normal cardiac and renal function, the retained fluid begins to mobilize back into the intravascular space 3–4 days after surgery. Inadequate clearance of this fluid, by impaired renal or cardiac function, may result in pulmonary edema or congestive heart failure. A patient slow to diurese spontaneously may only need a single dose of furosemide to facilitate the process. Clinical evaluation of hypovolemia includes looking for signs of oliguria, supine hypotension and orthostatic hypotension.75 Urine output of 0.5 mg/kg per h typically suggests adequate renal perfusion. Laboratory parameters may reflect hemoconcentration, azotemia, or low urinary sodium. The hematocrit is usually unchanged by acute hemorrhage until fluids

Table 14.8 Commonly used parenteral solutions pH

mOsm/l

kcal/l

Na (mEq/l)

D5 W

4.3

253

170

—

D5 1/2 NS

4.4

405

170

D5 NS

4.4

560

NS

5.6

D5 LR LR

K (mEq/l)

Ca (mEq/l)

Cl (mEq/l)

Lactate (mEq/l)

—

—

—

—

77

—

—

77

—

170

154

—

—

154

—

308

—

154

—

—

154

—

5.0

530

170

130

4

3

109

28

6.3

275

—

130

4

3

109

28

DS, 5% dextrose in water; W, water; NS, normal saline, LR, lactated Ringer’s

Table 14.9 Gastrointestinal fluid content Na (mEq/l)

K (mEq/l)

Cl (mEq/l)

20

10

120

0

1000–2500

Pancreatic

140

5

75

80

500–1000

Bile

148

5

100

35

300–1000

Small bowel

110

5

105

30

1000–3000

Diarrheal stools

120

25

90

45

500–17 000

Gastric

390

HCO3 (mEq/l)

Daily volume (ml)

Ch 14 final.qxd

12/18/2006

15:35

Page 391

POSTOPERATIVE MANAGEMENT

are administered or hemodilution occurs, with fluid shifting from the interstitial to the intravascular space. A blood urea nitrogen/serum creatinine ratio of greater than 20 may suggest dehydration. In prerenal oliguria, urine sodium is low, due to increased sodium and water resorption. A fractional excretion of sodium (FENa = [(UNa × PCr) / (PNa × UCr)] × 100) of less than 1% is suggestive of a prerenal state that is best managed by volume expansion. This calculation may be less diagnostic in patients who are elderly, have pre-existing renal disease, or who have received diuretics. Even if the urine output is low while volume resuscitative efforts may be ongoing, continued urine production and a stable creatinine level indicate adequate renal perfusion. Prevention of acute oliguria in the postoperative ovarian cancer patient is best accomplished by recognition of prerenal events including blood loss, surgical trauma and re-accumulation of ascites.76 The therapeutic goal of treating hypovolemia is to restore volume with fluid similar to the fluid that was lost.75 Patients with significant blood loss are managed by transfusion and administration of colloids intraoperatively. While blood products are required to treat anemia and coagulopathy, lactated Ringer’s solution or normal saline without added dextrose are the crystalloid solutions of choice for continued hypotension or shock. Large volumes of lactated Ringer’s solution may result in hyperkalemia, while large volumes of normal saline may result in hyperchloremic acidosis. Intravenous albumin (250 ml of 5% concentration) may not be any more effective than crystalloid solutions in replacing drained ascites, eventually leaking from the intravascular space into the peritoneal cavity, yet albumin may reach equilibrium slowly enough to allow for adequate resuscitation of the patient.

SODIUM Disorders of sodium concentration, hyponatremia and hypernatremia, reflect the relative excesses or deficits

of extracellular fluid. Pseudohyponatremia may occur with hyperproteinemia or hyperlipidemia, where protein or lipids displace water from plasma, producing a low plasma concentration of sodium. True hyponatremia (sodium < 136 mmol/l) may develop rapidly or chronically; acute hyponatremia may produce neurologic changes from cerebral edema, while chronic hyponatremia may trigger compensatory mechanisms which then require slow correction.77 Serum sodium level is frequently low in the postoperative ovarian cancer patient. With the administration of crystalloid fluids during surgery, hyponatremia may be associated with increased total body sodium concentration. In the immediate postoperative patient, there may still be a state of relative hypovolemia, which then results in secretion of antidiuretic hormone (ADH), which in turn preserves intravascular volume. Most patients with a serum sodium level > 125 mmol/l have few symptoms. A rapid decline of the sodium level to < 130 mEq/l may result in mental status changes, and even seizures. These patients with symptoms require rapid but controlled correction with hypertonic saline. An assessment of total body sodium concentration may be made by measuring urine sodium level and osmolarity. While third spacing is ongoing during the immediate postoperative period, the urine sodium level is low (< 10–15 mEq/l), and urine osmolarity is high (> 400 mOsm/kg). Treatment of hyponatremia with high serum osmolarity is directed towards the restriction of both sodium and water. Hypovolemic hyponatremic patients are treated with 0.9% saline. Hypervolemic hyponatremic patients are managed with fluid restriction, occasionally increasing free water excretion with the administration of diuretics. Hypernatremia is usually associated with a total fluid deficit, and may also result in neurologic changes. It can result from pure water loss (e.g. diabetes insipidus) or hypotonic sodium loss (e.g. nasogastric drainage) but may also be iatrogenic from hypertonic sodium loading.78 Correction of hypovolemia must be performed slowly with hypotonic solutions, to avoid precipitating cerebral edema or seizures.

391

Ch 14 final.qxd

12/18/2006

15:35

Page 392

SURGERY FOR OVARIAN CANCER

POTASSIUM Disorders of potassium homeostasis may be affected by the balance of intake versus excretion.79 Specifically, potassium may enter the body through oral or parenteral means, and leaves through renal excretion, which may be affected by acid/base status. Potassium is the major intracellular cation, such that plasma potassium may be a poor reflection of total body potassium. At the time of surgery, adrenergic stress may acutely mobilize potassium from the intravascular to the intracellular space. Additional gastrointestinal fluid losses will produce hypokalemia, unless the potassium deficit is adequately replaced. Hypokalemia interferes with muscle contractility; thus, ileus and generalized weakness are common manifestations of hypokalemia (potassium level < 3.0 mmol/l). Profound hypokalemia is associated with an increased risk of cardiac arrhythmias, and also predisposes to digitalis toxicity. Treatment of hypokalemia involves potassium replacement, and this must be performed cautiously (10–20 mEq/h) to avoid overcorrection and hyperkalemic complications. Hyperkalemia may occur through excess ingestion or intravenous administration, but more frequently through renal impairment.80 The most lethal manifestations of hyperkalemia include cardiac conduction abnormalities, including prolongation of the PR interval, decrease in P wave amplitude, and widening of the QRS complex, resulting in ventricular fibrillation or asystole. Cardiac effects are negligible when the potassium level is below 6 mEq/l. Indications for treatment include the presence of ECG changes, or when the serum concentration of potassium is greater than 7 mEq/l. Treatment of hyperkalemia involves elimination of exogenous potassium, reversing membrane hyperexcitability, and removing potassium from the body. Acute therapy includes administration of calcium, insulin and glucose, sodium bicarbonate, diuretics, or cation exchange resins (via the gastrointestinal tract). Dialysis should be reserved for patients with renal fail-

392

ure, or life-threatening hyperkalemia unresponsive to conventional treatment.

CALCIUM Approximately 50% of serum calcium is unbound, while the remainder is complexed, primarily to albumin. Hypoalbuminemia, as seen in many ovarian cancer surgery patients, alters total serum calcium concentration; clinical decisions should therefore be based on ionized calcium levels. If ionized calcium cannot be measured, the total serum calcium level can be corrected by adding 0.8 mg/dl for each 1.0 g/dl the serum albumin is below 4.0 g/dl. In the perioperative patient receiving multiple transfusions, hypocalcemia may be attributable to chelation by citrate. Hyperphosphatemia may precipitate calcium or decrease intestinal absorption of calcium. Hypomagnesemia may suppress the production of parathyroid hormone. Clinical manifestations of hypocalcemia (ionized calcium level < 0.7 mmol/l) include neuronal membrane irritability and tetany. Acute management includes replacement of calcium intravenously, and correction of other electrolyte abnormalities.81 Hypercalcemia occurs most commonly with bone resorption, when calcium enters the extracellular volume more rapidly than can be excreted by the kidneys. Bone metastases are uncommon in ovarian cancer, and hypercalcemia (total serum calcium level > 13 mg/dl or ionized calcium level > 1.3 mmol/l) is also less common in the perioperative gynecologic oncology patient.

MAGNESIUM Magnesium also plays an important role in neuronal conduction. Although hypomagnesemia is common in perioperative gynecologic oncology patients, symptoms are uncommon unless the serum magnesium level is less than 1.0 mg/dl, at which time patients may have symptoms of weakness, lethargy, muscle spasms,

Ch 14 final.qxd

12/18/2006

15:35

Page 393

POSTOPERATIVE MANAGEMENT

paresthesias and depression. Causes of hypomagnesemia may include excessive losses through the gastrointestinal tract, or inability of the kidneys to conserve magnesium, such as after platinum-based chemotherapy. The sodium–potassium pump is magnesium dependent; attempts to correct potassium deficits may not be successful unless the magnesium deficit is simultaneously corrected. Treatment is by replacing magnesium intravenously, with reduced doses given in patients with renal insufficiency.82 Most cases of hypermagnesemia are iatrogenic and will be corrected with urinary excretion.

PHOSPHATE Phosphate provides the primary energy bond in ATP, is an essential element of second messenger systems, and is a major component of cellular membranes and nucleic acids; significant phosphate depletion also results in cellular energy depletion.81 Severe hypophosphatemia (< 1 mg/dl) usually indicates total body phosphate depletion, and may manifest with paresthesias, muscle weakness, malaise, encephalopathy, seizures and coma. Moderate hypophosphatemia (1–2.5 mg/dl) is usually attributable to renal losses, and a decrease in gastrointestinal absorption. Hypophosphatemic patients are often hypokalemic and hypomagnesemic. Intravenous administration of phosphorus should also be given cautiously to patients with renal dysfunction, or hypocalcemia. Hyperphosphatemia in the postoperative patient is usually due to decreased renal excretion.

ACID–BASE DISORDERS Prompt recognition and treatment of acid–base disturbances and electrolyte imbalances are important to the homeostasis of postoperative patients. A pH of 7.4 reflects the normal hydrogen concentration, which is maintained in balance with PaCO2 and HCO 3–. Acidemia can result from either an increased PaCO2

or a decreased HCO –3 concentration. Alkalemia can result from either decreased PaCO2 or increased HCO –3. Most of the time, the clinical process is mixed.81 Metabolic acidosis (pH < 7.35) results from a decrease in HCO –3 (< 21 mEq/l) due to either loss of bicarbonate or accumulation of acid.81 Loss of HCO –3 may be through diarrhea, biliary drainage, urinary diversion, or renal tubule losses – hyperchloremic metabolic acidosis is associated with a normal anion gap (calculated as [Na+] – ([Cl–] + [HCO –3])). A high-anion-gap acidosis may be due to excess production of acids (lactic acidosis or ketoacidosis), increased retention of waste products (sulfate or phosphate), or ingestion of toxins (salicylic acid, ethylene glycol, or methanol). A compensatory response is seen through hyperventilation and a fall in PaCO2. In the postoperative ovarian cancer patient, metabolic acidosis is most commonly due to gastrointestinal fluid losses and renal failure. Treatment of metabolic acidosis should focus on correcting the underlying metabolic condition. In a mechanically ventilated patient, a compensatory hyperventilation should be included in ventilator settings. Administration of sodium bicarbonate or other alkalinizing agents should be reserved for severe acidemia.83 Metabolic alkalosis (pH > 7.45) results from either a loss of H+ or a gain in HCO –3 (> 27 mEq/l).81 Loss of H+ may be through nasogastric suction or diuretic administration. The re-absorption of HCO –3 in the distal renal tubules results in hypokalemia and hypovolemia, resulting in a so-called ‘contraction alkalosis.’ Treatment of metabolic alkalosis therefore includes the replacement of volume and electrolytes.84 Fluid resuscitation with lactated Ringer’s solution may be better than with normal saline, as HCO –3 can be generated from lactate. Respiratory acidosis (pH < 7.35) is characterized by hypercarbia (PaCO2 ≥ 45 mmHg), which occurs when ventilation is insufficient to eliminate CO2.81 Over time, the kidneys compensate by excreting H+ and retaining HCO –3. Postoperative patients are particularly at risk of respiratory depression with upper

393

Ch 14 final.qxd

12/18/2006

15:35

Page 394

SURGERY FOR OVARIAN CANCER

abdominal incisions, or when receiving opiates for perioperative pain management. Supplemental oxygen minimizes the incidence of hypoxia, even when decreased ventilation increases the risk of hypercarbia. Severe respiratory acidosis is an indication for intubation and ventilatory support. Respiratory alkalosis (pH > 7.45) occurs with increased ventilation, causing hypocarbia (PaCO2 ≤ 35 mmHg).81 In a mechanically ventilated patient, overbreathing can also result in respiratory alkalosis. In the patient on the postsurgical unit, pain, anxiety, central nervous system disease, or sepsis may all be causes of hyperventilation. An active approach may include interventions of sedation and reassurance if anxiety is the cause of hyperventilation.

RENAL DISORDERS The perioperative setting remains one of the most common risk factors for acute renal failure among hospitalized patients.85 The overall incidence is highest in patients undergoing cardiac or vascular surgery, although patients with pre-existing renal disease, hypertension, cardiovascular disease, diabetes and advanced age are all considered to be at higher risk. The elderly have a lower glomerular filtration rate and are more susceptible to volume depletion and other nephrotoxic insults. Despite the advancements in intensive care and renal replacement therapy, mortality from acute renal failure remains high, perhaps because acute renal failure often occurs in the setting of multi-organ failure. Traditionally, the evaluation of acute renal failure includes consideration of prerenal, intrarenal and postrenal causes85,86 (Table 14.10). A hypotensive insult, resulting in an ischemic–reperfusion injury, is the most common cause of prerenal acute renal failure and is manifest as acute tubular necrosis. Full renal recovery can occur after several days, barring any further injury. Renal injury from nephrotoxic drugs is unusual in a healthy well-hydrated patient, but an older, volume-depleted woman with chronic

394

renal insufficiency undergoing ovarian cancer surgery is more vulnerable to renal injury. Non-steroidal antiinflammatory drugs inhibit renal vasodilatation. Aminoglycoside antibiotics may injure the renal tubules through direct toxicity or immunologically mediated interstitial nephritis. Angiotensin converting enzyme (ACE) inhibitors and angiotensin II receptor antagonists disrupt the intrinsic renal autoregulation, also increasing the risk of renal hypoperfusion injury. Radiographic contrast is also nephrotoxic, although this risk can be reduced with the administration of N-acetylcysteine before and after administering the contrast dye.87 Postrenal causes of renal failure include urinary retention and ureteral injury. Radical pelvic dissection in ovarian tumor debulking can often result in bladder dysfunction in the immediate postoperative period. While a thoracic level epidural should not impede bladder function, high doses of opiates for pain control may delay spontaneous voiding. A Foley catheter can easily drain the bladder. The only indication of ureteral obstruction may be a transient rise in the serum creatinine level. A renal ultrasound examination may confirm hydroureteronephrosis, but a non-dilated collecting system does not exclude the possibility of obstruction. For patients with a high index of clinical suspicion, a CT urogram may evaluate the entire urinary system, and also identify any associated postsurgical findings. A period of hemodynamic compromise is the most common cause of postoperative renal insufficiency. Isolated renal failure may recover, but renal failure may be only part of a syndrome of multiple organ failure, prompting the need for dialysis or hemofiltration. Prerenal acute renal failure can be reversible if renal perfusion is maintained, and additional renal insults are avoided. In administering the patient’s medications, dosing of renally excreted drugs may require dose adjustment for creatinine clearance. Fluid, electrolyte and nutrition replacement must take into account the degree of renal insufficiency, both acute and chronic. Urinalysis, urine microscopy and urine electrolytes may be helpful in the

Ch 14 final.qxd

12/18/2006

15:35

Page 395

POSTOPERATIVE MANAGEMENT

Table 14.10 Common causes of acute renal failure in the surgical setting. Adapted from reference 85 Prerenal

Intrarenal

Postrenal

Hypotension

Drugs: non-steroidal anti-inflammatory drugs, aminoglycosides, amphoteracin B, radiographic contrast

Ureteral obstruction

Hypovolemia Arterial occlusion or stenosis Cardiac failure

Toxins: endotoxins

Bladder dysfunction Urethral obstruction

Pigment: myoglobin

Sepsis

distinction between prerenal and intrarenal processes. In prerenal renal failure, the specific gravity will be high (> 1.020), urine sodium level low (< 20 mmol/l), and the fractional excretion of sodium less than 1%.76,86 The treatment of acute renal failure is mainly supportive, treating the underlying cause of renal failure, and correcting fluid and electrolyte imbalances. Low-dose dopamine has not been shown to be effective in protecting or improving renal function.88,89 Loop diuretics are commonly used for converting oliguric to non-oliguric renal insufficiency, although response may be merely a demonstration of less severe kidney damage. In severe renal failure, large doses of diuretics may be required, and recent studies have suggested that a continuous furosemide infusion may be more effective than bolus therapy.90 When planning renal replacement therapy, management of volume and solute toxicity are parallel but separate goals. Intermittent hemodialysis is the most common approach for renal replacement therapy. Indications for dialysis in the acute setting include severe fluid overload, hyperkalemia, metabolic acidosis and uremia. In a stable patient with acute renal failure, hemodialysis allows for the rapid removal of fluids and toxic metabolites. In a critically ill ovarian cancer patient after surgery, hypotension from sepsis or multi-organ failure often precludes the use of intermittent hemolysis. Continuous hemofiltration or hemodiafiltration allows for a slower rate of fluid removal, and theoretically results in fewer hemodynamic shifts in the unstable ICU patient.76,91–93 In continuous venovenous hemofiltration, vascular

access is achieved through a central vein, and anticoagulation with heparin is given to maintain the extracorporeal circuit. Despite these new technologies, however, renal replacement is only a means of support, and ultimate mortality reduction still requires recovery of the kidneys and other affected organs.

NUTRITIONAL SUPPORT Malnutrition in ovarian cancer patients may result from the physical compression of the bowel from tumor masses, omental caking and ascites. In addition, cancer cachexia may cause metabolic changes, including increasing resting energy expenditure, increased anaerobic glycolysis, and a high turnover of glycerol and free fatty acids.94 Treatment of malnutrition may include surgery to relieve obstruction, nutritional support and other pharmacologic approaches. Those patients who are nutritionally depleted are at greater risk of surgical morbidity, including increased risk of infection, prolonged hospital stay, and mortality.95 Patients with a low serum albumin level (less than 3.4 g/dl) on admission to an acute care hospital have been observed to have over a three-fold increased risk for mortality when compared to patients with a normal serum albumin level.96 Although many studies have demonstrated that improving nutritional support improved the nutritional parameters of patients, strong data are lacking to demonstrate improvement in clinically significant endpoints. Perioperative nutritional assessment may

395

Ch 14 final.qxd

12/18/2006

15:35

Page 396

SURGERY FOR OVARIAN CANCER

help identify patients who are most likely to benefit from nutritional support. Prior to surgery, patients suspected of having ovarian cancer should have a nutritional assessment, including evaluation of weight loss, serum albumin and absolute lymphocyte count. Triceps skinfold thickness, levels of serum protein, transferrin and prealbumin, as well as antigenic skin testing may also be appropriate nutritional assessment techniques. A period of starvation has been common practice after major abdominal surgery in general surgical and gynecologic oncology patients. There is a growing body of evidence, however, that strongly supports the role of early postoperative feeding to improve return of bowel function and speed time to discharge. Beyond its digestive capacities, the gastrointestinal tract is recognized as being an immunologic barrier against infection.11 Tolerating a regular diet also allows for earlier administration of oral analgesic medications. Several prospective clinical trials at the State University of New York at Stony Brook have demonstrated that early feeding is safe and advantageous in gynecologic oncology patients. Placement of a nasogastric tube intraoperatively does not necessarily mean that the patient continues with nasogastric intubation in the postoperative period. Patients frequently complain of discomfort from the tube, and while more traditional surgeons argue that nasogastric decompression decreases gastric and intestinal distension, these benefits have not been confirmed in clinical trials. Rather, in a prospective study of 110 gynecologic oncology patients undergoing intra-abdominal surgery randomized to postoperative nasogastric tube or intraoperative orogastric tube, Pearl et al. demonstrated that there was no difference in bowel complications, time to tolerating a regular diet or length of hospital stay.97 A meta-analysis of 26 clinical trials including 3964 patients evaluating nasogastric decompression also found that patients managed without nasogastric tubes had significantly less febrile morbidity, atelectesis and pneumonia.98 There was greater abdominal distension and vomiting, but this

396

was not associated with any increase in complications (wound dehiscence, infection, anastamotic leak) or length of stay. Early oral feeding was first suggested to be safe in gastrointestinal and colorectal surgery.99 To study the safety and efficacy of early postoperative feeding after gynecologic cancer surgery, Pearl et al. randomized 200 patients at Stony Brook to clear liquids on postoperative day 1 versus nil by mouth until passage of flatus.100 Time to development of bowel sounds (1.8 vs. 2.3 days, p = 0.07), tolerance of clear liquids (1.2 vs. 3.5 days, p < 0.0001) and regular diet (2.3 vs. 4.2 days, p < 0.0001) and hospital stay (4.6 vs. 5.8 days, p = 0.001) were significantly shorter in the early feeding group. These findings were confirmed by a separate prospective randomized study at Indiana University comparing clear liquids on postoperative day 1 with nil by mouth until bowel sounds, flatus or bowel movement, or subjective hunger.101 The study group had a higher incidence of emesis, but tolerated a regular diet 1 day earlier than the control group. A meta-analysis of randomized controlled trials in elective gastrointestinal surgery comparing enteral feeding to nil by mouth suggested that early feeding also reduced infection risk (0.72 relative risk, 95% CI 0.54–0.98), as well as length of hospital stay (mean reduction 0.84 days, p = 0.001).102 While there was a higher incidence of emesis, the risk of anastomotic dehiscence, wound infection, pneumonia and mortality were also improved, but did not reach statistical significance (Figure 14.4). A regular diet may also be safe and efficacious, providing higher caloric intake, and potentially decreasing the risk of aspiration. Total parenteral nutrition may be considered in patients who are expected to remain without enteral feeding for a prolonged period of time, but the timing of nutrition support is also influenced by the hemodynamic stability of the patient in the postoperative period. The altered metabolic state of increased catecholamines, glucocorticoids and glucagon favors the process of gluconeogenesis, glycogenolysis and fatty acid oxidation. In the immediate postoperative period, attention should be given to hemodynamic

Ch 14 final.qxd

12/18/2006

15:35

Page 397

POSTOPERATIVE MANAGEMENT

resuscitation and deferring the nutritional assessment until after the patient is stabilized. Nutritional support has become a standard of care for these patients, yet the acceptable duration of starvation without clinically significant sequelae remains an unanswered question.103 Total parenteral nutrition is not recommended for patients with an intact gastrointestinal tract, but many ovarian cancer surgery patients requiring critical care have also had bowel resection as part of their cytoreductive surgery and can be expected to have a delayed return of bowel function. A patient admitted to the ICU who is expected to be unable to eat for 7 days is more likely to benefit from total parenteral nutrition, compared to a previously healthy well-nourished patient expected to resume feedings within 7 days. A meta-analysis of 26 prospective randomized trials including 2211 patients comparing total parenteral nutrition with (standard) oral diet and intravenous dextrose did not show any improve-

ment in overall mortality of surgical or critically ill patients (risk ratio 1.03, 95% CI 0.81–1.31).104 The rate of major complications was lower among malnourished patients receiving total parenteral nutrition (risk ratio 0.52, 95% CI 0.31–0.91), although not among the parenterally fed patient population as a whole. Compared to parenteral nutrition, enteral nutrition promotes decreased gastrointestinal mucosal permeability and better wound healing. The enteral route is also associated with fewer metabolic disturbances, and is less expensive than parenteral nutrition. Enteral nutrition is contraindicated in patients with a postoperative ileus and in those at risk for aspiration. Patients with short gut syndrome also may require parenteral supplementation until the remaining bowel adapts.105 Feeding into the small bowel may be preferable over gastric feeding; this avoids the risks of regurgitation and aspiration associated with delayed gastric

2

95% CI

1.5

1.0

0.99 0.85

0.84

0.74 0.5

0.53

0.52

Death

Vomiting

Abdominal abscess

Pneumonia

Wound infection

Anastamotic dehiscence

0

Figure 14.4 Results from meta-analyses of randomized trials comparing early enteral feeding with nil by mouth. Relative risks and 95% CI adapted from reference 102

397

Ch 14 final.qxd

12/18/2006

15:35

Page 398

SURGERY FOR OVARIAN CANCER

emptying, particularly in an ovarian cancer surgery patient, in whom gastric distension may compromise pedicles on the short gastric vessels from an infragastric omentectomy. However, bypassing the stomach does not necessarily increase the tolerance to feeds, and either approach to enteral feeding may be acceptable. Randomized trials of prokinetic agents have not shown them to be advantageous in improving postoperative ileus.106 In the immediate postoperative period, a critically ill, nutritionally depleted ovarian cancer patient will be more likely to be considered for total parenteral nutrition, but enteral feeding should be attempted after recovery of bowel function. Caloric requirements are assessed according to the patient’s stored reserves and body catabolism.107 The Harris–Benedict equation is based on a basal energy expenditure calculated using the age, gender, height and weight of the individual, then multiplied by a ratio for stress and activity (Figure 14.5). A reasonable estimate of energy in a critically ill adult patient is 25–30 kcal/kg. Carbohydrates should usually account for 60–70% of non-protein calories. Fats should constitute 25–30% of calories, and include essential fatty acids. Protein requirements range from 1.2 to 2.0 g/kg per day. Fluids, electrolytes, vitamins and trace elements complete the daily formulation of total parenteral nutrition (Table 14.11). Histamine2 receptor antagonists are compatible as additives in parenteral nutrition solutions and should be regularly

prescribed. Regular insulin is also compatible as an additive and should be prescribed as indicated in parenteral nutrition solutions. Central administration of parenteral nutrition is recommended. Owing to hyperosmolarity and vein irritation, the final dextrose concentration in peripheral parenteral nutrition is only 10%, and the amino acid concentration just 2.5%, limiting the amount of calories that can be delivered. Concurrent administration of fat emulsion may reduce the amount of vein irritation. Re-feeding syndrome may occur among severely malnourished patients, with intracellular incorporation of phosphate resulting in severe hypophosphatemia and possibly respiratory failure. Potassium and magnesium shift intracellularly as well, resulting in hypokalemia and hypomagnesemia. Regular monitoring of laboratory profiles including electrolytes, glucose, liver function and lipid panels can prevent metabolic complications of parenteral nutrition. Excess glucose can result in hyperglycemia and other hyperosmolar states. Excess lipids can result in hyperlipidemia, and excess protein can worsen azotemia or encephalopathy. A 24-h urine urea nitrogen excretion can be collected after stable protein intake for 3–5 days to calculate the patient’s nitrogen balance.108 Nitrogen intake = protein intake (g)/6.25. Nitrogen output = urine urea nitrogen + insensible losses (constant of 3). A nitrogen balance of +4 to 5 g is required for anabolism; nutrition specialists should be involved to

For women, the Harris–Benedict equation: Basal energy expenditure = 655 + 9.6 × weight (kg) + 1.7 × height (cm) – 4.7 × age (years) Multiply by activity factor: Confined to bed Out of bed

1.2 × 1.3 ×

Multiply by stress factor: Minor surgery Soft tissue trauma Peritonitis Major sepsis

1.0–1.2 × 1.1–1.4 × 1.2–1.5 × 1.4–1.8 ×

Figure 14.5 Total estimated calorie requirement is based on the basal energy expenditure, and stress and injury factors

398

Ch 14 final.qxd

12/18/2006

15:35

Page 399

POSTOPERATIVE MANAGEMENT

help optimize parenteral nutrition orders. The transition from parenteral nutrition to oral diet should ensure that the patient maintains adequate caloric intake throughout. Calorie counts may help quantitate the patient’s intake, and parenteral nutrition should be continued until the patient is taking at least 50% of calories through the enteric route.

ENDOCRINE DISORDERS The stress of surgery stimulates the hypothalamic– pituitary–adrenal axis to increase the secretion of cortisol. Patients on chronic exogenous steroids have a suppressed axis and are unable to mount an appropriate response when compared to normal individuals. Traditionally, these steroid-dependent patients have been prescribed supraphysiologic ‘stress doses’ perioperatively to prevent hypotensive crisis, although the risks of steroids regarding impaired wound healing, immunosuppression and other adverse effects are well recognized. The recovery of adrenal function after discontinuation of steroid therapy is also variable. Preoperative adrenocorticotropic hormone (ACTH) Table 14.11 Standard parenteral nutrition formulations per liter Central formulation Dextrose 25%, 250 g Amino acids 4.25%, 42.5 g Sodium 45 mEq Potassium 40 mEq Calcium 4.5 mEq Magnesium 5 mEq Chloride 43 mEq Phosphorus 15 mmol/l Acetate 41 mEq Trace elements Vitamin K (weekly) Osmolarity Calories Nitrogen content

1825 mOsmol/l 1020 kcal 6.7 g

Peripheral formulation 10%, 100 g 2.5%, 25 g 41 mEq 18 mEq 4 mEq 5 mEq 35 mEq 9 mmol/l 28 mEq

880 mOsmol/l 440 kcal 4.0 g

stimulation testing has been suggested to identify patients with appropriate adrenal gland response, who probably do not require perioperative glucocorticoid coverage.109 Few studies have monitored the primary outcome of hypotensive crisis, and there is no evidence that perioperative ‘stress dose’ steroids are necessary to prevent hemodynamic instability secondary to adrenal insufficiency.110 However, case series have suggested that there is a low incidence of Addisonian crisis (1–2%), and given the risk of death associated with adrenal insufficiency, perioperative glucocorticoids are still recommended until prospective randomized controlled data become available (Table 14.12). In the diabetic patient, glycemic control in the perioperative period depends on the patient’s preoperative regimen, and when she resumes oral intake. While awaiting return of bowel function and advancement of diet after surgery, glucose management depends upon the level of metabolic stress, concentration of dextrose infusion and insulin administration.111 Subcutaneous sliding scale regular insulin is most frequently given every 4–6 h, with more refractory cases treated by dextrose infusion and an insulin drip. Insulin resistance and hyperglycemia are common in critically ill patients, even if they do not have a previous diagnosis of diabetes mellitus. For patients with hyperglycemia on ICU admission (glucose level > 215 mg/dl), maintaining blood sugars between 80 and 110 mg/dl has been shown in a randomized controlled study to decrease morbidity and mortality in a surgical ICU setting.112 Mortality benefit was greatest in patients remaining in the ICU for longer than 5 days, with a significant reduction in death from multi-organ failure due to sepsis.

PAIN MANAGEMENT Postoperative pain management of ovarian cancer surgery patients may include intravenous or epidural administration of analgesic medications. An intravenous patient-controlled analgesia machine can

399

Ch 14 final.qxd

12/18/2006

15:35

Page 400

SURGERY FOR OVARIAN CANCER

effectively deliver systemic opioids such as hydromorphone or morphine to treat postoperative pain. Early oral analgesia in gynecologic oncology patients is also safe and efficacious, but this depends on the philosophy of early postoperative feeding.113 Epidural analgesia with fentanyl and ropivicaine (or bupivicaine) is also an effective therapy for the management of pain after major abdominal surgery. Meta-analyses of randomized trials of epidural analgesia after laparotomy have shown that using local anesthetics with reduced doses of epidural opioids speeds the return of bowel function, allowing for earlier hospital discharge, but is associated with increased pain.8 Epidural analgesia has also been associated with decreased intraoperative blood loss, fewer thromboembolic events and early ambulation, but this has not been adequately studied in gynecologic cancer surgery.114 Several small series of gynecologic oncology patients have suggested that thoracic epidural analgesia can provide better pain control with earlier return to bowel function and discharge. However, they have not included high-risk gynecologic oncology patients with transfusion requirements, bowel resection procedures, infectious complications and postoperative chemotherapy.115

INFECTIOUS MORBIDITY Postoperative fever is a common occurrence with potentially serious implications. While infection must

Table 14.12 Standard perioperative stress dose steroids. Adapted from reference 109 Surgical stress

Daily steroid regimen (IV)

Minor 25 mg hydrocortisone (e.g. inguinal dissection)

Duration (days) 1

Moderate (e.g. hysterectomy)

50–75 mg hydrocortisone

1–2

Major (e.g. tumor debulking)

100–150 mg hydrocortisone

2–3

400

be considered, fever may also be a result of tissue inflammation. An accepted definition of fever in the postoperative period is a temperature elevation of more than 38–38.5°C in a 24-h period. Reported incidences of postoperative fever range from 15 to 47%, with a source of infection identified in only 5–36% of patients, and bacteremia identified in less than 3% of patients. A traditional ‘fever work up’ of complete blood cell count, urine culture, blood culture and chest X-ray may be an inefficient use of resources.116 Within the first 3 days after surgery, postoperative fevers are often attributable to atelectasis. Clinical history is important to help assess the patient’s overall risk for postoperative infection. Length of surgery, blood loss, surgical contamination, pre-existing infection, nutritional status, an immunocompromised state and the presence of malignancy all contribute to patient risk. Targeting cultures and laboratory tests to the population at highest risk helps to minimize superfluous testing of low yield. The presence of fever, tachycardia, leukocytosis and localizing symptoms may more reliably suggest an underlying process. Foreign bodies can be a nidus for infection. Drains, including Foley catheters, ureteral stents and intraperitoneal pelvic drains should be removed as soon as appropriate.117 Use of antibiotic-coated or antiseptic impregnated central venous catheters appears to decrease the incidence of catheter colonization and catheter-related blood infections.118 Intra-abdominal and pelvic infections can be insidious in onset. The presence of free air on X-ray may persist for several days after abdominopelvic surgery. An abscess may take time to consolidate before CT drainage can be undertaken. A variety of antibiotics can be considered to cover a broad spectrum of aerobic and anaerobic organisms. A combination such as ampicillin with clavulanic acid or piperacillin with tazobactam is a commonly employed regimen. An alternative may be a triple regimen of ampicillin, gentamicin and clindamycin or metronidazole; or a combination of a fluoroquinolone and metronidazole. Duration of treatment is dictated for the most part by

Ch 14 final.qxd

12/18/2006

15:35

Page 401

POSTOPERATIVE MANAGEMENT

severity of infection, but will generally be given for a 7–14-day course. Suspicion for intra-abdominal infection or systemic sepsis should prompt the initiation of broadspectrum antibiotics. One-quarter of patients with clinical suspicion of sepsis will have no identifiable source of infection on microbiology studies; these patients have similar predisposing risk factors to those with positive cultures, and also a similar risk for death.119 Antibiotics are essential for the management of septic shock, but unfortunately are not sufficient for optimal treatment. Early recognition of systemic inflammatory response and organ dysfunction may identify patients who may benefit from recombinant activated protein C (drotrecogin-α). In a large multi-center randomized study, treatment with drotrecogin-α for 96 h, within 24 h of meeting inclusion criteria, significantly reduced mortality in patients with severe sepsis, but was associated with a higher risk of significant bleeding.120 Serious bleeding occurred primarily in those predisposed to bleeding – including patients with coagulopathy, severe thrombocytopenia (less than 30 000), gastrointestinal bleeding, or trauma. Selection of the antimicrobial agent is usually empiric and should provide broad-spectrum coverage. Wound infections are frequently due to Staphylococcus or Streptococcus organisms, which are normally sensitive to penicillins or first-generation cephalosporins. Clindamycin may be a favorable alternative to cover Gram-positive organisms in those patients with a penicillin allergy. Intra-abdominal organisms such as Escherichia coli or Bacterioides fragilis may be covered with piperacillin with tazobactam, with possible addition of an aminoglycoside. More virulent Gram-negative rods may respond better to imipenem with cilastin or fourth-generation cephalosporins. Pathologic fungal infections should be treated with amphotericin B or fluconazole. Newer antifungals such as liposomal amphoteracin B or caspofungin may have better toxicity profiles. Antimicrobial resistance is a serious concern, with increasing resistance having been observed to several classes of antibiotics. Limiting the use of drugs such as

fluoroquinolones, or increasing the treatment dose to reduce the risk of mutant selection, are strategies that may prove valuable in maintaining the antibiotic armamentarium.

WOUND CARE Wound infection remains a major cause of postoperative morbidity, and is a major cause of subsequent incisional hernias. The wound should be covered by a sterile occlusive dressing for the first 24–48 h after surgery, the time it takes for the skin to reepithelialize.121 Normal wound healing after a surgical incision consists of the inflammatory phase, the proliferative or fibroblastic phase, and the remodeling or maturation phase. The bacteriocidal activity of neutrophils is mediated by oxidative killing, which can be improved by the simple administration of supplemental oxygen in the perioperative period. A prospective randomized study of supplemental perioperative oxygen – FiO2 0.80 versus 0.30, keeping the O2 saturation at 99–100% – in colorectal resection surgery patients decreased the incidence of wound infection by 50% (11–5%).122 Promoting pain relief, keeping the patient warm and improving perfusion by optimizing the patient’s volume status can also enhance wound healing and reduce wound infection. Careful repletion of intravascular fluids allows for better tissue oxygenation and subsequent collagen deposition.123 Maintaining normothermia during and after colorectal surgery can decrease the incidence of wound infection by as much as two-thirds (19–6%).124 Fibroblast proliferation reaches its peak about a week after surgery. During this phase, hypoxia, ischemia, infection and malnutrition are most likely to have an adverse effect on wound healing. Incisional seromas and hematomas may be drained without opening the incision as long as there is no sign of infection. Probing between skin staples, or reapproximating the wound with suture or tape, may help maintain skin approximation. An infected or dirty wound requires opening and dressing changes. If

401

Ch 14 final.qxd

12/18/2006

15:35

Page 402

SURGERY FOR OVARIAN CANCER

the wound is associated with cellulitis, antibiotic therapy with a penicillin, a cephalosporin, or clindamycin may also be necessary. Wet to dry dressings are useful for debridement of necrotic tissue, but removal of dry dressings from granulation tissue may damage the new tissue and should not be performed. Moist wound care is preferred, using agents such as Sorbsan or calcium alginate on exudative wounds, and hydrogel or aquaphor on dry wounds. Wound irrigation or hydrotherapy may also be useful in removing loose necrotic tissue. A healthy granulating wound may be surgically closed using permanent monofilament suture at the bedside or in the office after discharge.125 Re-infection rates are low, and time to complete healing is significantly shorter than if the wound is allowed to heal by secondary intention. A deep draining wound also heals slowly. Application of constant negative pressure, such as with a vacuumassisted closure device, may enhance the wound environment by removing excess drainage and speed time to closure.126 Fascial dehiscence and evisceration is a major complication with significant morbidity and mortality if not treated promptly. In the surgical literature, risk factors for wound dehiscence included malnutrition, hypoalbuminemia, anemia, emergency procedures, ileus, vomiting, coughing and chronic lung disease.127 In the gynecologic patient population, the majority of abdominal wound dehiscences occur following vertical incisions, with the mean time from surgery of 7–8 days, a time period where many ovarian cancer patients are being prepared for discharge.128 Removal of skin staples may yield drainage that is most likely to indicate a wound seroma. Probing the fascia may identify a large weakness or defect. Management of fascial dehiscence should prepare the patient for reoperation and wound closure. Oral intake should stop, and intravenous antibiotics should be started. If evisceration has occurred, the bowel should be covered with warm saline soaks. In the operating room, all necrotic or infected tissue within the wound should be debrided. In most cases, the suture is found to be cutting through tissue, with a minority of cases

402

due to suture breakage or knot slippage. Those patients with multiple risk factors should be considered for full-thickness abdominal wall retention sutures on re-closure. The skin should be left open, but may be closed by delayed primary closure at a later time.

MANAGEMENT OF EFFUSIONS AND ASCITES Malignant pleural effusions may accumulate, owing to neoplastic involvement of the pleura, or by direct metastatic dissemination. The pathogenesis may involve lymphatic channel disruption by malignant cells, or increased endothelial permeability. The patient may present with dyspnea, or the effusion may be found on clinical examination or chest radiography. Thoracentesis is the first step for diagnosis and treatment, as reactive effusions are sometimes seen in ovarian cancer, particularly in patients with diaphragm disease or splenectomy, or in the postoperative period. The diagnosis of malignant pleural effusion indicates advanced-stage disease, although survival of stage IV ovarian cancer patients with pleural effusion as their only manifestation of stage IV disease may be somewhat better than that of patients with parenchymal liver lesions or other disseminated disease. The management of malignant pleural effusions from metastatic ovarian cancer depends on the patient’s overall medical condition.129,130A large effusion may interfere with perioperative ventilation, and may need to be drained prior to surgery. Similarly, effusions in the critically ill patient on the ventilator may need to be drained to facilitate extubation. Most malignant effusions also improve with the initiation of chemotherapy, but a symptomatic patient may require thoracentesis. Controlled drainage of no more than 1000–1500 ml at any given time is recommended to prevent re-expansion pneumonitis or pulmonary edema. Repeated thoracentesis predisposes to adhesions and loculation of the effusions. Tube

Ch 14 final.qxd

12/18/2006

15:35

Page 403

POSTOPERATIVE MANAGEMENT

thoracostomy may be useful for larger effusions, and facilitates the instillation of a sclerosing agent. Recently, small-bore catheters (8–14 Fr) have been reported as having similar success to large-bore chest tubes (24–32 Fr) in drainage and pleuridesis.131 These pigtail catheters can be placed by interventional radiologists with less discomfort to the patient than the traditional chest tube. Lung re-expansion is a prerequisite to successful pleuridesis; the output should be less than 100–150 ml/day before sclerotherapy is attempted. A number of sclerosing agents have been used for pleuridesis, although the ideal agent remains to be found132 (Table 14.13). Tetracycline was the most popular agent until its production was discontinued in the 1990s. Doxycycline has been substituted with modest efficacy. Bleomycin is the most widely used antineoplastic agent, but since the mechanism of action is chemical sclerosis, its success rate still does not justify its cost. Sterile talc is well tolerated and inexpensive, with 80–90% efficacy for pleuridesis. After the lung is re-expanded, a slurry is made with 5 g of talc in 50 ml of normal saline and instilled into the pleural space. Lidocaine may also be diluted in the slurry for analgesia. The tube remains clamped for 2 h before the fluid is allowed to drain. Rotation of the patient is probably unnecessary. Provided the lung remains expanded and the fluid drainage low, the tube may be removed in 24–48 h. Chest pain and fever are common with all pleuridesis agents. A potential serious adverse effect of talc is the rare association with acute respiratory distress syndrome. Effusions refractory for pleuridesis may also be approached with thorascopic surgery and pleurectomy. Malignant ascites is best treated by tumor resection and chemotherapy. While there may be some reaccumulation of ascites in the postoperative period, this is usually self-limited and does not require additional paracentesis. In fact, given the fluid shifts surrounding ovarian cancer surgery, paracentesis is usually discouraged, as this would tend to further deplete the patient’s intravascular volume as the ascites reaccumulates.

Abdominal compartment syndrome is an uncommon, but critical, condition where intra-abdominal pressure affects the perfusion of abdominal viscera.133 The syndrome was first described in trauma patients with complicated intraoperative and postoperative courses with large amounts of fluid resuscitation. The pressure may originate from re-accumulation of ascites, intra-abdominal hemorrhage, or bowel edema. Without prompt recognition, increased systemic vascular resistance, decreased venous return and organ failure may ensue.134 Typically, the diagnosis is made when increased ventilatory pressure is associated with increased central venous pressure, decreased urine output and increased abdominal distension. Early recognition allows for early intervention.135 Intraabdominal pressure may be measured most easily by transducing the bladder with 50–100 ml of fluid instilled. After clamping the catheter, a central venous monitoring line is attached to a 16-gauge needle inserted into the aspiration port of the Foley catheter. A bladder pressure of more than 15–25 mmHg in the setting of cardiopulmonary and renal impairment may be indicative of abdominal compartment syndrome. Decompression of the abdomen may be accomplished by removal of ascites, but successful therapy may require decompression by re-opening the abdomen, at which point there is usually an immediate improvement of the patient’s hemodynamic status. Delayed closure of the incision may be performed in a staged fashion after appropriate Table 14.13 Efficicacy of methods of palliating malignant pleural effusions. Adapted from reference 130 Mean 30-day control rate (%) Thoracentesis alone

2

Tube drainage

18

Tetracycline

60–70

Doxycycline

75

Bleomycin

84

Talc

95

403

Ch 14 final.qxd

12/18/2006

15:35

Page 404

SURGERY FOR OVARIAN CANCER

resuscitation. The mortality rate for abdominal compartment syndrome is 63–72%. In a patient with extensive bowel surgery and fluid resuscitation for ovarian cancer cytoreduction, it is unknown whether placement of drains may help to reduce the risk of abdominal compartment syndrome, but may serve as an outlet valve for decompression of large-volume ascites re-accumulation, avoiding the need for repeated paracentesis.

CHEMOTHERAPY IN THE POSTOPERATIVE PERIOD There is a lack of evidence regarding the optimal time interval between surgical staging/debulking and the initiation of adjuvant chemotherapy for patients with ovarian cancer. Surgical removal of tumor may stimulate growth of the remaining tumor cells. In a mouse model, cyclophosphamide was most effective against metastatic growth when given preoperatively or immediately after tumor removal.136 A delay in the initiation of systemic therapy has been hypothesized to increase the probability of drug resistance.137 In breast cancer, early start of adjuvant chemotherapy may improve outcome. A late analysis of the International (Ludwig) Breast Cancer Study Group (IBCSG) Trial V suggested that premenopausal node-positive patients with negative estrogen receptors had better survival when cyclophosphamide, methotrexate and fluorouracil chemotherapy was initiated within 20 days of breast surgery, including axillary lymph node dissection.138 With a median follow-up of 7.7 years, 10-year diseasefree survival was 60% in the early initiation group, and 34% in the later initiation group (p = 0.0003). This survival advantage was not seen in patients with estrogen receptor-expressing tumors. In ovarian cancer, neoadjuvant chemotherapy has been effective in managing symptoms of ascites in patients who are not candidates for primary cytoreductive surgery. Similarly, the ascites and effusions that re-accumulate in the postoperative period

404

usually respond well to prompt initiation of cytotoxic chemotherapy. The incidence of neutropenia and infectious complications of current paclitaxel and carboplatin chemotherapy is low enough for the potential benefits to outweigh the risks. In addition, if the patient is to be followed with a medical oncologist for subsequent chemotherapy cycles, initiating the chemotherapy prior to discharge avoids any delay in treatment during the period of transition from the gynecologic oncologist to the medical oncologist. Under normal circumstances, the patient should be medically stable and ready for discharge before chemotherapy is initiated – both to minimize the potential confounding effects of chemotherapyinduced nausea on the return of bowel function, and to minimize the risk of chemotherapy nadirs on potential postoperative complications. Patients are often re-assured to receive their first cycle of chemotherapy in the in-patient setting, where nurses and clinical pharmacists can also devote more time to chemotherapy teaching. If a central line was placed intraoperatively, this access can be used for chemotherapy before removal. The patient’s intravenous access should be re-assessed prior to discharge for evaluation of whether an indwelling intravenous port or peripherally inserted central catheter may be indicated for subsequent chemotherapy infusion. Because of the potential fluid shifts and fluctuations in laboratory values in the perioperative period, chemotherapy doses should be calculated using the patient’s current weight. For patients weighing more than 80 kg, an adjusted body weight should be used for calculating creatinine clearance. Twenty-four-hour urine collection is notoriously poor at measuring creatinine clearance. As long as the serum creatinine level is stable, the Cockcroft–Gault equation ([(140 – age) × kg] / [72 × serum creatinine] × 0.85) is quite adequate in measuring renal function. For patients who have had a particularly unstable postoperative course, there may be benefit in allowing the patient a longer recovery before initiating postoperative chemotherapy.

Ch 14 final.qxd

12/18/2006

15:35

Page 405

POSTOPERATIVE MANAGEMENT

HORMONE REPLACEMENT THERAPY The role of hormone replacement therapy for all women remains in evolution. A small retrospective study of women receiving hormone therapy after ovarian cancer treatment did not demonstrate any difference in disease-specific or overall mortality.139 Nevertheless, the attributable risk of ovarian cancer incidence and mortality from estrogen remains in question. In the American Cancer Society Prevention Project II, over 211 000 postmenopausal women with no history of cancer or ovarian surgery, and who were using estrogen therapy at baseline, had higher death rates than those who had never used estrogen (1.51 relative risk, 95% CI 1.16–1.96).140 Duration of use was associated with increased risk and the cessation of use was associated with a persistence of risk. Followup of over 44 000 postmenopausal women in the Breast Cancer Detection Demonstration Project suggested that estrogen use was associated with an increased incidence of ovarian cancer (1.6 relative risk, 95% CI 1.2–2.0), with a particularly increased risk for women who had used estrogen-only replacement for 10 or more years.141 Combination hormone therapy including estrogen and progestins was not associated with a significantly increased risk. For perspective, the incidence of ovarian cancer was only 0.74% in this cohort, thus an increased relative risk still did not confer a large absolute risk from hormone therapy. In the immediate postoperative period, estrogen therapy may be indicated for those women who are particularly symptomatic from surgical menopause. These patients may be the strongest candidates for estrogen therapy; however, the relief of these symptoms needs to be weighed against the small increased risk of estrogen and venous thromboembolic events.

ily for postoperative expectations. Psychosocial resources from the primary team, social workers and support groups may also ease the stress of facing a new cancer diagnosis. Criteria for discharge may include the following: afebrile without evidence of uncontrolled infection, tolerating a normal diet without nausea or vomiting, evidence of bowel and bladder function, and evidence of appropriate wound healing. With the economic pressure of minimizing the length of hospital stay, home nursing services are often valuable for establishing close follow-up until the patient returns to the out-patient office.

REFERENCES 1.

Brooks SE, Ahn J, Mullins CD, et al. Resources and use of the intensive care unit in patients who undergo surgery for ovarian carcinoma. Cancer 2002; 95: 1457–62

2.

Amir M, Shabot MM, Karlan BY. Surgical intensive care unit care after ovarian cancer surgery: an analysis of indications. Am J Obstet Gynecol 1997; 176: 1389–93

3.

Van Le L, Fakhry S, Walton LA, et al. Use of the APACHE II scoring system to determine mortality of gynecologic oncology patients in the intensive care unit. Obstet Gynecol 1995; 85: 53–6

4.

Eachempati SR, Miller FE, Fins JJ. The surgical intensivist as mediator of end-of-life issues in the care of critically ill patients. J Am Coll Surg 2003; 195: 847–53

5.

Pollack MM, Katz RW, Ruttiman EU, et al. Improving the outcome and efficiency of intensive care: the impact of an intensivist. Crit Care Med 1988; 16: 11

6.

Goldman L, Caldera DL, Nussbaum SR, et al. Multifactorial index of cardiac risk in noncardiac surgical procedures. N Engl J Med 1977; 297: 845–50

7.

Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 1999; 100: 1043–9

8.

Rodgers A, Walker N, Schug S, et al. Reduction of postoperative mortality and morbidity with epidural

DISCHARGE PLANNING Discharge planning should be initiated during the preoperative visit, to prepare the patient and her fam-

405

Ch 14 final.qxd

12/18/2006

15:35

Page 406

SURGERY FOR OVARIAN CANCER

or spinal anaesthesia: results from overview of randomised trials. Br Med J 2000; 321: 1493 9.

20.

Arozullah AM, Daley J, Henderson WG, et al. Multifactorial risk index for predicting postoperative respiratory failure in men after major noncardiac surgery. Ann Surg 2000; 232: 242–53

21.

Arozullah AM, Knuri SF, WG Henderson, et al. Development and validation of multifactorial risk index for predicting postoperative pneunomia after major noncardiac surgery. Ann Intern Med 2001; 135: 847–57

22.

Smetena G. Preoperative pulmonary evaluation. N Engl J Med 1999; 340: 937–44

23.

Gross PA, Neu HC, Aswapokee P, et al. Deaths from nosocomial infections: experience in a university and community hospital. Am J Med 1980; 68: 219

24.

Podnos SD, Toews GB, Pierce AK. Nosocomial pneumonia in patients in intensive care units. West J Med 1985; 143: 622–7

25.

Craven DE, Steeger KA. Epidemiology of nosocomial pneumonia. New perspectives on an old disease. Chest 1995; 108: 1S–16S

26.

Martin LF, Asher EF, Casey JM, et al. Postoperative pneumonia. Arch Surg 1984; 119: 379–83

27.

White R, Dilworth P. Pneumonia in the hospital. Br J Dis Chest 1988; 82: 121–6

28.

Schierhout G, Roberts I. Fluid resuscitation with colloid or crystalloid solutions in critically ill patients: a systematic review of randomized trials. Br Med J 1998; 316: 961–4

Cross AS, Roup B. Role of respiratory assistance devices in endemic nosocomial pneumonia. Am J Med 1981; 70: 681–5

29.

Choi PTL, Yip G, Qinonez LG, et al. Cystalloids vs. colloids in fluid resuscitation: a systematic review. Crit Care Med 1999; 27: 200–10

Johanson WG, Pierce AK, Sanford JP. Changing pharygeal bacterial flora of hospitalized patients: emergence of gram negative bacilli. N Engl J Med 1969; 281: 1137–40

30.

Rouby JJ, Martin De Lassale E, Poete P, et al. Nosocomial bronchopneumonia in the critically ill. Histologic and bacteriologic aspects. Am Rev Respir Dis 1992; 146: 1059–66

31.

Barnes P. Chronic obstructive pulmonary disease. N Engl J Med 2000; 343: 269–80

Badner NH, Knill RL, Brown JE, et al. Myocardial infarction after noncardiac surgery. Anesthesiology 1998; 88: 572–8

10.

Mangano DT, Layug EL, Wallace AW, et al. Effect of atenolol on mortality and cardiovascular morbidity after noncardiac surgery. N Engl J Med 1996; 335: 1713–20

11.

Deitch EA, Winterton J, Li M, et al. The gut as a portal of entry for bacteremia. Role of protein malnutrition. Ann Surg 1987; 205: 681–92

12.

13.

14.

15.

16.

17.

Rodriguez RM, Rosenthal MH. Etiology and pathophysiology of shock. In Murray MJDC, Pearl RG, Prough DS, eds. Critical Care Medicine: Perioperative Medicine. Philadelphia: Lippincott Williams & Wilkins, 2002: 192–205 Kollef M. Critical care. In Carey CFLH, Woeltje KF, eds. The Washington Manual of Medical Therapeutics, 29th edn. Philadelphia: Lippincott Williams & Wilkins, 1998: 170–89 Kollef M. Critical care parameters and formulas. In Carey CFLH, Woeltje KF, eds. The Washington Manual of Medical Therapeutics, 29th edn. Philadelphia: Lippincott Williams & Wilkins, 1998: 562–4

Wilkes MM, Navickis RJ. Patient survival after human albumin administration: a meta-analysis of randomized, controlled trials. Ann Intern Med 2001; 135: 149–64

18.

White R. Acute therapy in patients with cardiac arrhythmias. In Murray MJDC, Pearl RG, Prough DS, eds. Critical Care Medicine: Perioperative Medicine. Philadelphia: Lippincott Williams & Wilkins, 2002: 320–32

32.

Bernard GR, Artigas A, Brigham KL, et al. The American–European Consensus Conference on ARDS: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Med 1994; 149: 818–24

19.

Behr J. Optimizing preoperative lung function. Curr Opin Anaestheiol 2001; 14: 65–9

33.

The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared

406

Ch 14 final.qxd

12/18/2006

15:35

Page 407

POSTOPERATIVE MANAGEMENT

with traditional tidal volumes for acute lung injury and the actue respiratory distress syndrome. N Engl J Med 2000; 342: 1301–8 34.

Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000; 342: 1334–49

35.

Brown JV, Karlan BY, Greenspoon JS, et al. Perioperative coagulopathy in patients undergoing primary cytoreduction. Cancer 1993; 71: 2557–61

36.

UCLA Medical Center Division of Transfusion Medicine. Transfusion Therapy Manual. Edina, MN: Burgess Publishing, 1996

37.

The Blood Bank & Transfusion Service Staff The University of Michigan Hospitals. Blood Transfusion Policies and Standard Practices. Ann Arbor, MI, 1995

38.

39.

Danilos J, Goral A, Paluszkiewicz P, et al. Successful treatment with recombinant factor VIIa for intractable bleeding at pelvic surgery. Obstet Gynecol 2003; 101: 1172–3 Goodnough LT, Brecher ME, Kanter, et al. Transfusion medicine: blood transfusion. N Engl J Med 1999; 340: 438–47

40.

Goodnough LT, Brecher ME, Kanter, et al. Transfusion medicine: blood conservation. N Engl J Med 1999; 340: 525–33

41.

Crandon AJ, Koutts J. Incidence of post-operative deep vein thrombosis in gynaecological oncology. Aust NZ J Obstet Gynaecol 1983; 23: 216–19

pneumatic calf compression for the prevention of deep venous thrombosis after gynecologic oncology surgery. Am J Obstet Gynecol 1993; 168: 1146–53, discussion 1153–4 47.

Maxwell GL, Synan I, Dodge R, et al. Pneumatic compression versus low molecular weight heparin in gynecologic oncology surgery: a randomized trial. Obstet Gynecol 2001; 98: 989–95

48.

Maxwell GL, Synan I, Hayes RP, et al. Preference and compliance in postoperative thromboembolism prophylaxis among gynecologic oncology patients. Obstet Gynecol 2002; 100: 451–5

49.

Wille-Jorgensen P, Hauch O, Dimo B, et al. Prophylaxis of deep venous thrombosis after acute abdominal operation. Surg Gynecol Obstet 1991; 172: 44–8

50.

Lensing AW, Prandoni P, Brandjes D, et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med 1989; 320: 342–5

51.

Cronan JJ. Venous thromboembolic disease: the role of US. Radiology 1993; 186: 619–30

52.

Ginsberg JS, Kearon C, Douketis J, et al. The use of D-dimer testing and impedance plethysmographic examination in patients with clinical indications of deep vein thrombosis. Arch Intern Med 1997; 157: 1077–81

53.

Lee AY, Julian JA, Levine MN, et al. Clinical utility of a rapid whole-blood D-dimer assay in patients with cancer who present with suspected acute deep venous thrombosis. Ann Intern Med 1999; 131: 417–23

42.

Rubenstein MD, Wall RT, Wood GS, et al. Complications of therapeutic apheresis, including a fatal case with pulmonary vascular occlusion. Am J Med 1983; 75: 171–4

43.

Kearon CA, Hirsh J. Management of anticoagulation before and after elective surgery. N Engl J Med 1997; 336: 1506–11

54.

Kline J, Johns K, Colucciello S, et al. New diagnostic tests for pulmonary embolism. Ann Emerg Med 2000; 35: 168–80

44.

Geerts WH, Heit JA, Clagett GP, et al. Prevention of venous thromboembolism: Sixth ACCP Consensus Conference on Antithrombotic Therapy. Chest 2001; 119: 132S–75S

55.

45.

Clarke-Pearson DL, Dodge RK, Synan I, et al. Venous thromboembolism prophylaxis: patients at high risk to fail intermittent pneumatic compression. Obstet Gynecol 2003; 101: 157–63

Stein PD, Terrin ML, Hales CA, et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest 1991; 100: 598–603

56.

Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). The PIOPED Investigators. JAMA 1990; 263: 2753–9

46.

Clarke-Pearson DL, Synan IS, Dodge R, et al. A randomized trial of low-dose heparin and intermittent

407

Ch 14 final.qxd

12/18/2006

15:35

Page 408

SURGERY FOR OVARIAN CANCER

57.

Viellard-Baron A, Antakly Y, et al. Transesophageal echocardiology as a diagnostic and therapeutic aid in a critical care setting. Chest 1995; 107: 774–9

58.

Remy-Jardin M, Remy J. Spiral CT angiography of the pulmonary circulation. Radiology 1999; 212: 615–36

59.

60.

61.

62.

63.

64.

65.

66.

67.

408

tionated heparin for acute pulmonary embolism. The THESEE Study Group. Tinzaparine ou Heparine Standard: Evaluations dans l’Embolie Pulmonaire. N Engl J Med 1997; 337: 663–9 68.

Weitz JI. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. The Columbus Investigators. N Engl J Med 1997; 337: 657–62

69.

Weitz JI. Low-molecular-weight heparins. N Engl J Med 1997; 337: 688–98

70.

Pravinkumar E, Webster NR. HIT/HITT and alternative anticoagulation: current concepts. Br J Anaesth 2003; 90: 676–85

71.

Hamilton CA, Robinson WR. Femoral artery occlusion following pelvic cancer surgery. Gynecol Oncol 1996; 63: 151–3

72.

Pestana C. Maintainance. In Pestana C, ed. Fluids and Electrolytes in the Surgical Patient. Philadelphia: Lippincott Williams & Wilkins, 2000: 1–23

73.

Pestana C. Replacement. In Pestana C, ed. Fluids and Electrolytes in the Surgical Patient. Philadelphia: Lippincott Williams & Wilkins, 2000: 24–37

74.

Roberts JP, Roberts JD, Skinner C, et al. Extracellular fluid deficit following operation and its correction with Ringer’s lactate. A reassessment. Ann Surg 1985; 202: 1–8

75.

Pestana C. Volume changes. In Pestana C, ed. Fluids and Electrolytes in the Surgical Patient. Philadelphia: Lippincott Williams & Wilkins, 2000: 38–55

76.

Klahr S, Miller SB. Acute oliguria. N Engl J Med 1998; 338: 671–5

77.

Adrogue HJ, Madias NE. Hyponatremia. N Engl J Med 2000; 342: 1581–9

Siragusa S, Cosmi B, Piovella F, et al. Low-molecularweight heparins and unfractionated heparin in the treatment of patients with acute venous thromboembolism: results of a meta-analysis. Am J Med 1996; 100: 269–77

78.

Adrogue HJ, Madias NE. Hypernatremia. N Engl J Med 2000; 342: 1493–9

79.

Gennari FJ. Hypokalemia. N Engl J Med 1998; 339: 451–8

Lensing AW, Prins MH, Davidson BL, et al. Treatment of deep venous thrombosis with low-molecular-weight heparins. A meta-analysis. Arch Intern Med 1995; 155: 601–7

80.

Kim HJ, Han SW. Therapeutic approach to hyperkalemia. Nephron 2002; 92 (Suppl 1): 33–40

81.

Prough DS, Mathru M, Lang JD. Diagnosis and management of acid–base and electrolyte abnormalities. In Murray MJDC, Pearl RG, Prough DS, eds. Critical Care Medicine: Perioperative Medicine.

Goodman LR, Curtin JJ, Mewissen MW, et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. Am J Roentgenol 1995; 164: 1369–74 Garg K, Kemp JL, Wojcik D, et al. Thromboembolic disease: comparison of combined CT pulmonary angiography and venography with bilateral leg sonography in 70 patients. Am J Roentgenol 2000; 175: 997–1001 Loud PA, Katz DS, Bruce DA, et al. Deep venous thrombosis with suspected pulmonary embolism: detection with combined CT venography and pulmonary angiography. Radiology 2001; 219: 498–502 Horlocker TT, Wedel DJ, Benzon H, et al. Regional anesthesia in the anticoagulated patient: defining the risks (the second ASRA Consensus Conference on Neuraxial Anesthesia and Anticoagulation). Reg Anesth Pain Med 2003; 28: 172–97 Decousus H, Leizorovicz A, Parent F, et al. A clinical trial of vena caval filters in the prevention of pulmonary embolism in patients with proximal deepvein thrombosis. Prevention du Risque d’Embolie Pulmonaire par Interruption Cave Study Group. N Engl J Med 1998; 338: 409–15 Leizorovicz A, Simonneau G, Decousus H, et al. Comparison of efficacy and safety of low molecular weight heparins and unfractionated heparin in initial treatment of deep venous thrombosis: a metaanalysis. Br Med J 1994; 309: 299–304

Simonneau G, Sors H, Charbonnier B, et al. A comparison of low-molecular-weight heparin with unfrac-

Ch 14 final.qxd

12/18/2006

15:35

Page 409

POSTOPERATIVE MANAGEMENT

Philadelphia: Lippincott Williams & Wilkins, 2002: 206–24 82.

Pestana C. Specific electrolytes. In Pestana C, ed. Fluids and Electrolytes in the Surgical Patient. Philadelphia: Lippincott Williams & Wilkins, 2000: 81–99

83.

Adrogue HJ, Madias NE. Management of life-threatening acid–base disorders – first of two parts. N Engl J Med 1998; 338: 26–34

84.

Adrogue HJ, Madias NE. Management of life-threatening acid–base disorders – second of two parts. N Engl J Med 1998; 338: 107–11

85.

Carmichael P, Carmichael AR. Acute renal failure in the surgical setting. Aust NZ J Surg 2003; 73: 144–53

86.

Thadhani R, Pascual M, Bonventre JV. Acute renal failure. N Engl J Med 1996; 334: 1448–60

87.

Birck R, Krzossok S, Markowetz F, et al. Acetylcysteine for prevention of contrast nephropathy: meta-analysis. Lancet 2003; 362: 598–603

ology and management. Anticancer Res 2001; 21: 2941–7 95.

Heyland DK. Nutritional support in the critically ill patients. A critical review of the evidence. Crit Care Clin 1998; 14: 423–40

96.

Herrmann FR, Safran C, Levkoff SE, et al. Serum albumin level on admission as a predictor of death, length of stay, and readmission. Arch Intern Med 1992; 152: 125–30

97.

Pearl ML, Valea FA, Fischer M, et al. A randomized controlled trial of postoperative nasogastric tube decompression in gynecologic oncology patients undergoing intra-abdominal surgery. Obstet Gynecol 1996; 88: 399–402

98.

Cheatham ML, Chapman WC, Key SP, et al. A metaanalysis of selective versus routine nasogastric decompression after elective laparotomy. Ann Surg 1995; 221: 469–76, discussion 476–8

99.

Reissman P, Teoh TA, Cohen SM, et al. Is early oral feeding safe after elective colorectal surgery? A prospective randomized trial. Ann Surg 1995; 222: 73–7

88.

Burger A. Low-dose dopamine in patients with early renal dysfunction: a placebo-controlled randomised trial. Lancet 2000; 356: 2139–43

89.

Kellum JA, Decker J. Use of dopamine in acute renal failure: a meta-analysis. Crit Care Med 2001; 29: 1526–31

100. Pearl ML, Valea FA, Fischer M, et al. A randomized controlled trial of early postoperative feeding in gynecologic oncology patients undergoing intraabdominal surgery. Obstet Gynecol 1998; 92: 94–7

90.

Dormans M, Tom PJ, van Meyel M, et al. Diuretic efficacy of high dose furosemide in severe heart failure: bolus injection versus continuous infusion. J Am Coll Cardiol 1996; 28: 376–82

101. Schilder JM, Hurteau JA, Look KY, et al. A prospective controlled trial of early postoperative oral intake following major abdominal gynecologic surgery. Gynecol Oncol 1997; 67: 235–40

91.

Misset B, Timsit JF, Chevret S, et al. A randomized cross-over comparison of the hemodynamic response to intermittent hemodialysis and continuous hemofiltration in ICU patients with acute renal failure. Intensive Care Med 1996; 22: 742–6

102. Lewis SJ, Egger M, Sylvester PA, et al. Early enteral feeding versus ‘nil by mouth’ after gastrointestinal surgery: systematic review and meta-analysis of controlled trials. Br Med J 2001; 323: 773–6

92.

93.

94.

Forni LG, Hilton PJ. Continuous hemofiltration in the treatment of acute renal failure. N Engl J Med 1997; 336: 1303–9 Bellomo R, Farmer M, Boyce N. A prospective study of continuous venovenous hemodiafiltration in critically ill patients with acute renal failure. J Intensive Care Med 1995; 10: 187–92 Gadducci A, Cosio S, Fanucchi A, et al. Malnutrition and cachexia in ovarian cancer patients: pathophysi-

103. Wong PW, Enriquez A, Barrera R. Nutritional support in critically ill patients with cancer. Crit Care Clin 2001; 17: 743–67 104. Heyland DK, MacDonald S, Keefe L, et al. Total parenteral nutrition in the critically ill patient: a metaanalysis. JAMA 1998; 280: 2013–19 105. Johnson M. Management of short bowel syndrome – a review. Support Line 2000; 22: 11–25 106. Smith AJ, Nissan A, Lanouette NM, et al. Prokinetic effect of erythromycin after colorectal surgery: ran-

409

Ch 14 final.qxd

12/18/2006

15:35

Page 410

SURGERY FOR OVARIAN CANCER

domized, placebo-controlled, double-blind study. Dis Colon Rectum 2000; 43: 333–7 107. The University of Michigan Medical Center Parenteral and Enteral Nutrition Manual. Ann Arbor, MI, 1994

119. Brun-Buisson C, Doyon F, Carlet J, et al. Incidence, risk factors, and outcome of severe sepsis and septic shock in adults. A multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis. JAMA 1995; 274: 968–74

108. Long CL, Schaffel N, Geiger JW, et al. Metabolic response to injury and illness: estimation of energy and protein needs from indirect calorimetry and nitrogen balance. J Parenter Enteral Nutr 1979; 3: 452–6

120. Bernard GR, Vincent J-L, Laterre P-F, et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001; 344: 699–709

109. Salem MTR, Bromberg J. Perioperative glucocorticoid coverage: a reassessment 42 years after emergence of a problem. Ann Surg 1994; 219: 416–25

122. Greif R, Akca O, Horn EP, et al. Supplemental perioperative oxygen to reduce the incidence of surgicalwound infection. Outcomes Research Group. N Engl J Med 2000; 342: 161–7

110. Buie BC. Perioperative stress dose steroids: do they make a difference? J Am Coll Surg 2001; 193: 678–86 111. Jacober SJ. An update on perioperative managment of diabetes. Ann Intern Med 1999; 159: 2405–11 112. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001; 345: 1359–67 113. Pearl ML, McCauley DL, Thompson J, et al. A randomized controlled trial of early oral analgesia in gynecologic oncology patients undergoing intraabdominal surgery. Obstet Gynecol 2002; 99: 704–8 114. Rapp SE, Ready LB, Greer BE. Postoperative pain management in gynecology oncology patients utilizing epidural opiate analgesia and patient-controlled analgesia. Gynecol Oncol 1989; 35: 341–4 115. de Leon-Casasola OA, Karabella D, Lema MJ. Bowel function recovery after radical hysterectomies: thoracic epidural bupivacaine–morphine versus intravenous patient-controlled analgesia with morphine: a pilot study. J Clin Anesth 1996; 8: 87–92 116. Badillo AT, Sarani B, Evans SR. Optimizing the use of blood cultures in the febrile postoperative patient. J Am Coll Surg 2002; 194: 477–87, quiz 554–6 117. Roy S. Perioperative infections: prevention and therapeutic options. In Vasilev S, eds. Perioperative and Supportive Care in Gynecologic Oncology: Evidencebased Management. New York: Wiley-Liss, 2000: 411–35 118. Veenstra DL, Saint S, Saha S, et al. Efficacy of antiseptic-impregnated central venous catheters in preventing catheter-related bloodstream infection. JAMA 1999; 281: 261–7

410

121. Chrintz H, Vibits H, Cordtz TO, et al. Need for surgical wound dressing. Br J Surg 1989; 76: 204–5

123. Hartmann M, Johnson K, Zederfeldt B. Effect of tissue perfusion and oxygenation on accumulation of collagen in healing wounds. Randomized study in patients after major abdominal operations. Eur J Surg 1992; 158: 521–6 124. Kurz A, Sessler D, Lenhardt R, et al. Perioperative normothermia to reduce the incidence of surgicalwound infection and shorten hospitalization. N Engl J Med 1996; 334: 1209–16 125. Dodson MK, Magann EF, Meeks GR. A randomized comparison of secondary closure and secondary intention in patients with superficial wound dehiscence. Obstet Gynecol 1992; 80: 321–4 126. Schimp VL, Worley C, Brunello S, et al. Vacuumassisted closure in the treatment of gynecologic oncology wound failures. Gynecol Oncol 2004; 92: 586–91 127. Makela JT, Kiviniemi H, Juvonen T, et al. Factors influencing wound dehiscence after midline laparotomy. Am J Surg 1995; 170: 387–90 128. Helmkamp BF. Abdominal wound dehiscence. Am J Obstet Gynecol 1977; 128: 803–7 129. Antunes G, Neville E. Management of malignant pleural effusions. Thorax 2000; 55: 981–3 130. Hausheer FH, Yarbro JW. Diagnosis and treatment of malignant pleural effusion. Semin Oncol 1985; 12: 54–75 131. Parulekar W, Di Primio G, Matzinger F, et al. Use of small-bore vs. large-bore chest tubes for treatment of malignant pleural effusions. Chest 2001; 120: 19–25

Ch 14 final.qxd

12/18/2006

15:35

Page 411

POSTOPERATIVE MANAGEMENT

132. Sahn SA. Talc should be used for pleurodesis. Am J Respir Crit Care Med 2000; 162: 2023–4, discussion 2026 133. von Gruenigen VE, Coleman RL, King MR, et al. Abdominal compartment syndrome in gynecologic surgery. Obstet Gynecol 1999; 94: 830–2 134. Schein M, Wittmann DH, Aprahamian CC, et al. The abdominal compartment syndrome: the physiological and clinical consequences of elevated intraabdominal pressure. J Am Coll Surg 1995; 180: 745–53 135. Tiwari A, Haq AI, Myint F, et al. Acute compartment syndromes. Br J Surg 2002; 89: 397–412 136. Fisher B, Gunduz N, Saffer EA. Influence of the interval between primary tumor removal and chemotherapy on kinetics and growth of metastases. Cancer Res 1983; 43: 1488–92 137. Goldie JH, Coldman AJ. A mathematic model for relating the drug sensitivity of tumors to their sponta-

neous mutation rate. Cancer Treat Rep 1979; 63: 1727–33 138. Colleoni M, Bonetti M, Coates AS, et al. Early start of adjuvant chemotherapy may improve treatment outcome for premenopausal breast cancer patients with tumors not expressing estrogen receptors. The International Breast Cancer Study Group. J Clin Oncol 2000; 18: 584–90 139. Eeles RA, Tan S, Wiltshaw E, et al. Hormone replacement therapy and survival after surgery for ovarian cancer. Br Med J 1991; 302: 259–62 140. Rodriguez C, Patel AV, Calle EE, et al. Estrogen replacement therapy and ovarian cancer mortality in a large prospective study of US women. JAMA 2001; 285: 1460–5 141. Lacey JV Jr, Mink PJ, Lubin JH, et al. Menopausal hormone replacement therapy and risk of ovarian cancer. JAMA 2002; 288: 334–41

411

Index.qxd

12/18/2006

15:38

Page 413

Index

abdominal compartment syndrome 403–4 abdominal wall anatomy 59–61 ablation liver disease 255–6 mesenteric disease 221 pelvic peritoneal implants 161–2 access to specialist care 103–7 acid–base disorders 393–4 acute renal failure 394–5 acute respiratory distress syndrome (ARDS) 380–1 adjuvant chemotherapy 70–2 postoperative period 404 secondary cytoreductive surgery 325–6 see also chemotherapy adrenal arteries 173 adrenal glands 176 left 266–7 right 236–7 adrenal veins 173 age distribution 1 ampulla of Vater 235 analgesia, postoperative 399–400 anastomosis see intestinal anastomosis anastomotic dehiscence 158–9 androblastoma 25 anemia 40, 381 antibiotics postoperative infection management 400–1 prophylaxis 44–5 bowel preparation 45–6 aorta 130 abdominal 172–3 appendectomy 70

appendix 197 argon beam coagulator (ABC) 48 peritoneal implant ablation 161–2 ascites 8–9 anastomotic dehiscence and 158 palliative surgery 369 postoperative 159 following liver disease cytoreduction 258 management 403–4 atrial fibrillation 378 automated stapling devices 49–52 bilaterality 12 biliary ducts 232, 235 biopsies 65 second-look surgery 292–3 bladder 127 partial cystectomy 153–4 blind loop syndrome 361 blood supply 130–1 abdominal wall 61 diaphragm 230 large intestine 199–201 liver 232 omentum 202 ovary 7 pancreas 264 retroperitoneum 172–5 small intestine 196–7 stomach 262–3 blood transfusion reactions 381–2 Boari bladder flap 156 Bookwalter self-retaining retractor 47

413

Index.qxd

12/18/2006

15:38

Page 414

SURGERY FOR OVARIAN CANCER

borderline malignancies 16–18, 72–4 management 72–4 staging 72, 177–8 bowel involvement 97–8, 195 radical oophorectomy, type II modification 140–53 see also intestinal anastomosis; intestinal tract bowel preparation 45–6 bowel resection see intestinal anastomosis; intestinal tract BRCA1/BRCA2 mutations 4–5 breast cancer familial syndrome 3–5 screening 6 Brenner tumors 15–16 broad ligament 129 bubble test 146 CA19-9 42 CA27.29 42 CA-125 41, 290 preoperative assessment 41–2 recurrent disease surveillance 307–8, 309–10 screening 6 calcium levels, postoperative management 392 Call–Exner bodies 23, 26 caloric requirements 398–9 carcinoembryonic antigen (CEA) 42 carcinoid tumors 20 carcinosarcoma 15 cardiac complications 376–9 preoperative risk evaluation 376 cardinal ligaments 129 Cautlie’s line 232 cavitron ultrasonic surgical aspirator (CUSA) 48–9 mesenteric disease excision 221 peritoneal implant aspiration 162 cecostomy 366–7 cecum 197 celiac axis involvement 286–7 cervix 127 Cheatle incision 206 chemotherapy 70–2, 108

414

germ cell tumors 78 postoperative period 404 prior tumor mass reduction 89 regimen 115–16 response following optimal cytoreduction 94 interval cytoreductive surgery and 109–10 second-look surgery and 289–90 secondary cytoreductive surgery and 325–6 sex cord–stromal tumors 80 see also neoadjuvant chemotherapy Cherney incision 62 chest radiograph, preoperative 41 chest tube insertion 242–3, 371 cholecystectomy 258–9 choriocarcinoma 22 chronic obstructive pulmonary disease (COPD) 380 circular end-to-end anastomosis (CEEA) 144–8 stapler 51, 52, 150–1 classification 11–28 epithelial ovarian cancer 11–18 germ cell tumors 19–22, 75 gonadoblastoma 27 metastatic tumors 27–8 primary peritoneal carcinoma 18–19 radical oophorectomy 136 sex cord–stromal tumors 22–7, 78 clear cell carcinoma 15 coagulation abnormalities 381 studies 41, 381 colectomy subtotal transverse, with omental disease 217–18, 269–70 total 219 see also hemicolectomy; intestinal anastomosis colic arteries 199–200 colon 197–9 ascending 198 descending 199 transverse 198, 265 see also bowel involvement; colectomy; intestinal anastomosis; intestinal tract; sigmoid colon

Index.qxd

12/18/2006

15:38

Page 415

INDEX

colonic fistulae 223 colonoscopy 363 preoperative 41 colorectal anastomosis see intestinal anastomosis colostomy, diverting 364–6 complications 367 end colostomy 364–6 transverse loop colostomy 364 complete blood count (CBC) 40 compression, thromboembolic prophylaxis 46, 384, 385 computed tomography (CT) 42–3, 290 bowel obstruction investigation 353, 363 pulmonary embolism diagnosis 386, 387 recurrent disease surveillance 308, 310 continuous hemofiltration 395 continuous positive airway pressure (CPAP) ventilation 379 coronary artery disease see cardiac complications coronary ligaments 230–2 cryoablation, liver disease 256 Cushing’s syndrome 27 cystectomy, partial 153–4 cystic artery 234 cytoreductive surgery 88–9, 133–5, 305–6 access to specialist care 103–7 bulky retroperitoneal nodal mass 185–6 complications 99–101 cardiac complications 376–9 hematologic complications 381–3 infectious morbidity 400–1 pulmonary complications 379–81 thromboembolic complications 383–9 degree of radicality 98–9 delayed, following neoadjuvant chemotherapy 111–15 rationale 111–12 selection criteria 112–15 diaphragm 237–43, 280–3 gastrectomy 283–6 historical perspective 87 intestinal tract 209–11 complications 196, 221–3

exposure 209 ileocecal resection 213–14 indications for 195–6 large intestine 217–19 small intestine 211–13 surgical approach 209–11 liver disease 243–58 ablation 255–6 hepatic resection 246–54 non-anatomic wedge resection 244–6 parenchymal disease 244 postoperative management 256–8 superficial disease 244 omentum 211, 214–17, 268–70 exposure 209 surgical approach 209–11 optimal cytoreduction 91–2 chemotherapy response and 94 feasibility/achievability 101 pancreatectomy 278–80 pelvic adenopathy 162–4 surgical techniques 163–4 rationale 89–90 growth fraction increase 89 immunological competence enhancement 89–90 maximal cytoreduction 90–1 tumor mass reduction prior to chemotherapy 89 tumor perfusion improvement 89 secondary see secondary cytoreductive surgery splenectomy 270–8 stage IV disease 101–3 surgery versus biology 92–3 surgical approach 96–8 survival 94–5 see also interval cytoreductive surgery; oophorectomy; pelvic peritoneal implants debulking surgery see cytoreductive surgery deep vein thrombosis (DVT) 46–7, 383–5, 387–9 dermoid cyst 19–20 diabetes, postoperative management 399

415

Index.qxd

12/18/2006

15:38

Page 416

SURGERY FOR OVARIAN CANCER

diagnosis 88 diagnostic lymphadenectomy 176–8 laparoscopic evaluation 331–3 large bowel obstruction 363 pleural effusion 370–1 small bowel obstruction 352–3 diaphragm 229–30, 266 apertures 229–30 innervation 230 metastatic disease 237 muscular components 229 surgical procedures 237–43 peritonectomy 239–41 postoperative management 242–3 resection 237, 241–2, 280–3 vascular supply 230 dietary factors 2–3 discharge planning 405 diverting colostomy see colostomy, diverting duodenum 235–6 dysgerminoma 20–1, 75 early-stage cancer management fertility preservation 74–5 germ cell tumors 75–8 incision selection 61–3 low malignant potential (LMP) tumors 72–4 sex cord–stromal tumors 78–80 staging 63–70 adjuvant therapy and 70–2 rationale for 57–9 see also staging electrocardiography, preoperative 41 electrolytes see serum electrolytes electrosurgery 47–52 argon beam coagulator (ABC) 48 automated stapling devices 49–52 cavitron ultrasonic surgical aspirator (CUSA) 48–9, 221 electrosurgical unit (ESU) 47–8, 160, 220 embryonic carcinoma 21–2 end colostomy 364–6 end-to-end anastomosis 206–7

416

hand-sewn techniques 148–9, 206 stapled techniques 144–8, 206–7 end-to-side anastomosis 207–8 hand-sewn technique 151–2, 208 stapled techniques 149–51, 207 endocrine disorders, postoperative management 399 endodermal sinus tumors 21 endometrioid carcinoma 14–15 enteral nutrition 397–8 enterocutaneous fistulae 222–3 epidemiology 1–6 familial syndromes 3–5 racial trends 2 risk factors 1–3 screening 5–6 epigastric arteries 61 epiploic foramen of Winslow 202 epithelial ovarian cancer 11–19 clear cell 15 endometrioid 14–15 fertility preservation 74–5 mixed epithelial types 16 mucinous 13–14 primary peritoneal carcinoma (PPC) 18–19 serous 11–13 squamous 16 transitional cell 15–16 tumors of low malignant potential 16–18 undifferentiated carcinoma 16 see also early-stage cancer management etiology 1–2 falciform ligament 230 fallopian tubes 127 familial syndromes 3–5 fascial dehiscence 402 femoral nerve 132–3 fertility preservation 74–5 germ cell tumors 77, 78 low malignant potential (LMP) tumors 73 sex cord–stromal tumors 79 fever, postoperative 400 fibroma 24–5

Index.qxd

12/18/2006

15:38

Page 417

INDEX

fibrosarcoma 24–5 fibrothecoma 24 FIGO staging 9–11, 57–9, 88 fistulae 222–3 fluids, postoperative management 389–91 frozen pelvis 140 functional end-to-end anastomosis (FEEA) 152–3, 209 gallbladder 234–5 cholecystectomy 258–9 Gambee one-layer closure 205 gastrectomy, partial 283–6 gastric artery 262–3 gastric veins 263 gastrocolic ligament resection 322 gastroduodenal artery 262–3 gastroenteritis 351 gastroepiploic artery 202, 262, 263, 265 gastrointestinal anastomosis (GIA) stapler 50–1, 52, 212 gastrojejunostomy 354–6 gastrostomy decompression 360–1 genitofemoral nerve 132 germ cell tumors 19–22, 75 choriocarcinoma 22 dysgerminoma 20–1 early-stage tumor management 75–8 postoperative therapy 78 embryonic carcinoma 21–2 endodermal sinus tumors 21 mixed germ cell tumors 22 polyembryoma 22 staging 77–8 teratomas 19–20 Glisson’s capsule 230 gonadoblastoma 20, 27 granulosa cell tumors 22–4, 79, 80 adult 23 juvenile 23–4, 79 greater omentum 201–2, 265–6 see also omental disease; omentectomy gynandroblastoma 26

hand-assisted laparoscopic surgery (HALS) 340 hematologic complications 381–3 hemicolectomy, with omental disease left 218–19 right 218 hemodialysis 395 hemorrhage management 164–6, 383 during lymphadenectomy 186–8 heparin, thromboembolic prophylaxis 46, 383–5, 387–9 heparin-associated thrombocytopenia (HAT) 389 heparin-induced thrombocytopenia (HIT) 389 hepatic artery 232, 262, 263 hepatic ducts 232, 234 hepatic flexure 198 hepatic mobilization 238–9 hepatic resection 246–54 left hepatectomy 251 right hepatectomy 248–51 secondary cytoreductive surgery 324 segmentectomy 251–4 right anterior sectoral resection 253 right posterior sectoral resection 253–4 unisegmentectomy 254 total inflow occlusion 247–8 hepatic veins 173, 232–4 hereditary breast–ovarian cancer syndrome 3–5 hereditary non-polyposis colon cancer (HNPCC) 5 hormone replacement therapy 405 hypercalcemia 392 hyperkalemia 40–1, 392 hypermagnesemia 392, 393 hypernatremia 391 hyperphosphatemia 392, 393 hypoalbuminemia 392 hypocalcemia 392 hypogastric artery 130 hypogastric nerves 133 hypogastric plexuses 133 hypogastric vein 130 injury to 165 hypokalemia 40–1, 392 hypomagnesemia 392, 393

417

Index.qxd

12/18/2006

15:38

Page 418

SURGERY FOR OVARIAN CANCER

hyponatremia 40–1, 391 hypophosphatemia 393 hypovolemia 377, 390–1 hysterectomy 64 ileocecal resection 213–14 ileostomy 358–9 ileum 196 ileus 221–2 carcinomatous 351 iliac arteries 130, 175 iliac veins 175 injury 165 iliohypogastric nerve 61 ilioinguinal nerve 61 imaging techniques 42–3, 290 recurrent disease surveillance 308–9 immune suppression 89–90 incessant ovulation theory 1 incision 136 closure 62–3 selection early-stage cancer 61–3 intestinal tract surgery 209 right upper abdominal surgery 237 infectious morbidity 400–1 antibiotic prophylaxis 44–5 wound care 401–2 inferior mesenteric artery (IMA) 130, 173, 199–200, 266 inferior mesenteric vein (IMV) 200, 266 inferior vena cava (IVC) 173–5 inguinal adenopathy 189–90 intensive care unit (ICU) 375–6 interval cytoreductive surgery 107–17, 306 chemotherapy regimen 115–16 following suboptimal primary cytoreduction 109–11 rationale 109 selection criteria 109–11 timing of interval surgery 116–17 intestinal anastomosis 202–9 end-to-end anastomosis 206–7

418

hand-sewn techniques 148–9, 206 stapled techniques 144–8, 206–7 end-to-side anastomosis 207–8 hand-sewn technique 151–2, 208 stapled techniques 149–51, 207 side-to-side functional end-to-end anastomosis 152–3, 209 techniques 202–6 hand-sewn anastomoses 148–9, 151–2, 202–5 stapled anastomoses 144–8, 149–51, 205–6 intestinal tract cytoreductive surgery 209–11 complications 196, 221–3 exposure 209 indications for 195–6 large intestine 217–19 omentum 211, 215–17 small intestine 211–14 surgical approach 209–11 see also intestinal anastomosis large intestine 197–201, 217–19 vasculature 199–201 mesenteric disease 219–21 ablative techniques 221 excisional techniques 220–1 obstruction see large bowel obstruction; small bowel obstruction small intestine 196–7 bypass 356–8 ileocecal resection 213–14 resection 211–12, 359 vasculature 196–7 see also bowel involvement intraperitoneal port placement 293 J-incision 62 Jackson’s membrane 198 jejunum 196 kidney 175 left 266–7 right 236 see also renal disorders

Index.qxd

12/18/2006

15:38

Page 419

INDEX

Krukenberg tumors 27 management 223–4 laparoscopy 291–2, 331–47 complications 293–4, 346 determination of cytoreduction feasibility 339–40 equivalence to laparotomy 294–6 hand-assisted laparoscopic surgery (HALS) 340 management of cancer diagnosed during laparoscopy 336–7 pelvic mass evaluation 331–3 risks of port site metastasis 333–5 rupture and tumor spillage 335–6 second-look laparoscopy 291–2, 338–9 staging 337–8, 345–6 surgical outcomes 333 surgical techniques 340–6 extraperitoneal aortic lymphadenectomy 344–5 oophorectomy 341 patient positioning 340, 341 port placement 340–1 transperitoneal aortic lymphadenectomy 341–3 transperitoneal pelvic lymphadenectomy 343–4 large bowel obstruction 362–7 complications of surgery 367 diagnostic evaluation 363 management 363–4 postoperative 222 surgical techniques 364–7 cecostomy 366–7 diverting colostomy 364–6 large intestine 197–201 ascending colon 198 cytoreductive surgery 217–19 descending colon 199 transverse colon 198, 265 vasculature 199–201 see also bowel involvement; intestinal anastomosis; intestinal tract; large bowel obstruction

left upper abdomen anatomy 261–7 adrenal gland 266–7 greater omentum 265–6 hemi-diaphragm 266 kidney 266–7 pancreas 263–4 spleen 264–5 stomach 261–3 transverse colon 265 surgical techniques 267–87 celiac axis involvement 286–7 distal pancreatectomy 278–80 omentectomy 268–70 partial gastrectomy 283–6 splenectomy 270–8 lesser omentum 201–2 see also omental disease; omentectomy Leydig cell tumors 26 Sertoli–Leydig cell tumor 25–6, 79 liver 230–4 biliary ducts 232 cytoreduction 243–58 ablation 255–6 hepatic resection 246–54 non-anatomic wedge resection 244–6 parenchymal liver disease 244 postoperative management 256–8 secondary cytoreductive surgery 324 superficial liver disease 244 ligamentous attachments 230–2 metastatic disease 243 mobilization 238–9 segmental anatomy 232–4 surfaces 230 vascular system 232 loop electrosurgical excision procedure (LEEP) 221 low malignant potential (LMP) tumors 16–18, 72–4 management 72–4 staging 72, 177–8 lumbar arteries 173 lumbar plexus 132 lumbar veins 173–5

419

Index.qxd

12/18/2006

15:38

Page 420

SURGERY FOR OVARIAN CANCER

lymph node metastases 65–8, 98, 134, 162, 171–2 diagnostic lymphadenectomy 176–8 extra-abdominal 189 inguinal 189–90 supraclavicular 190 low malignant potential (LMP) tumors 74, 177–8 see also lymphadenectomy lymphadenectomy 65–8, 98, 162–4 diagnostic 176–8 during second-look surgery 291 extra-abdominal 189 inguinal 189–90 supraclavicular 190 laparoscopic 341–5 extraperitoneal aortic 344–5 transperitoneal aortic 341–3 transperitoneal pelvic 343–4 morbidity 186 vascular complications 186–8 para-aortic nodes 68–70 infrarenal 180–2 suprarenal 182–4 pelvic nodes 68, 178–80 retroaortic nodes 184–5 systematic 178–84 technique 68–70, 163–4, 178–85 intact capsular resection 163 progressive subtotal enucleation 163 lymphatic drainage 7–8, 9, 65, 131–2, 176 large intestine 201 small intestine 197 Lynch II syndrome 5 Maffucci syndrome 23 magnesium levels, postoperative management 392–3 magnetic resonance imaging (MRI) 42–3 recurrent disease surveillance 308–9 malignant ascites see ascites malignant Brenner tumors 15–16 malignant mixed mesodermal (Müllerian) tumors (MMMTs) 14–15 malignant pleural effusion see pleural effusion malnutrition 43–4

420

postoperative nutritional support 395–9 marginal artery of Drummond 200 maximum surgical effort 87 Maylard incision 62 mechanical ventilation 379–80 medically compromised patients 114–15 Meigs’ syndrome 24 mesenteric arteries 130, 173, 196–7, 199–200, 266 mesenteric disease 219–21 ablative techniques 221 excisional techniques 220–1 mesenterotomy 212 metabolic acidosis 393 metabolic alkalosis 393 metastases, extraovarian 59 appendix 70 diaphragm 237 intestinal see intestinal tract laparoscopy risks 333–5 liver 243 lymph nodes 65–8, 134, 162, 171–2 diagnostic lymphadenectomy 176–8 extra-abdominal 189 inguinal 189–90 low malignant potential (LMP) tumors 74, 177–8 supraclavicular 190 spleen 270–1 metastatic tumors to ovaries 27–8 midline incision 62, 63 mismatch repair (MMR) gene mutations 4, 5 mortality 1 mucinous ovarian cancer 13–14 mumps, as risk factor 2 myocardial ischemia 378 nasogastric tube 396 neoadjuvant chemotherapy 108 followed by delayed primary cytoreduction 111–15 rationale 111–12 selection criteria 112–15 regimen 115–16

Index.qxd

12/18/2006

15:38

Page 421

INDEX

see also chemotherapy nutritional support postoperative 395–9 preoperative 43–4 oblique muscle external 61 internal 61 obturator nerve 132 obturator vein injury 165–6 Ollier disease 23 omental cake 97, 211, 214, 215–17, 268 omental disease left hemicolectomy 218–19 right hemicolectomy 218 subtotal transverse colectomy 217–18, 269–70 see also omentectomy omentectomy 65, 209–11, 214–17, 268–70 exposure 209 indications for 195 surgical approach 209–11 see also omental disease omentum 201–2, 265–6 vasculature 202 see also omental disease; omentectomy Omni retractor 47 oophorectomy historical perspective 87 laparoscopic 341 prophylactic 7 radical 135–59 after prior hysterectomy 158 classification 136 clinical outcomes 159 criteria for 135–6 morbidity 158–9 surgical technique 136–9 type I modification 139–40 type II modification 140–53 type III modification 153–9 see also salpingo-oophorectomy oral contraceptive, risk reduction 6–7 outcome

following cytoreductive surgery 94–5 secondary cytoreductive surgery 310–16, 319–21 following operative laparoscopy 333 predictors of 319–21 following negative second-look surgery 297–9 ovarian arteries 7, 130, 173 ovarian cancer see specific forms of cancer ovarian veins 7, 130, 173 ovaries 7, 127 blood supply 7 lymphatic drainage 7–8, 9, 131–2 pain management, postoperative 399–400 palliative surgery 88, 351–72 ascites 369 large bowel obstruction 362–7 complications 367 diagnostic evaluation 363 management 363–4 surgical techniques 364–7 pleural effusion 370–2 small bowel obstruction 351–62 complications 361–2 diagnostic evaluation 352–3 management 353–4 surgical techniques 354–61 urinary tract obstruction 367–9 pancreas 263–4 pancreatectomy, distal 278–80 pancreatic arteries 264 pancreatic ducts 263–4 para-aortic lymph node dissection 68–70 paracentesis 369 pararectal spaces 130 paravesicle spaces 130 parenteral nutrition bowel obstruction management 353, 354 preoperative 44 pelvic adenopathy 162–4 surgical techniques 163–4 pelvic anatomy 7–8, 127–33 ligaments 128–9

421

Index.qxd

12/18/2006

15:38

Page 422

SURGERY FOR OVARIAN CANCER

lymphatic drainage 7–8, 131–2 neural anatomy 32–133 potential spaces 129–30 vascular anatomy 7, 130–1 visceral anatomy 127–8 pelvic lymph node dissection 68 pelvic peritoneal implants 159–62 ablation 161–2 aspiration 162 local excision 160 peritonectomy 160–1 percutaneous nephrostomy 368 peritoneal stripping see peritonectomy peritonectomy 160–1 diaphragm 239–41 mesenteric 220 Peutz–Jeghers syndrome 26 Pfannenstiel incision 62 phosphate levels, postoperative management 393 phrenic arteries 173, 230 phrenic nerve 230 pleural effusion 370–2, 379 management 370, 371, 402–3 pleurodesis 371, 403 pneumonia 379–80 pneumothorax 242 polyembryoma 22 porta hepatis 235 disease 259 portal vein 200–1, 232 positive end-expiratory pressure (PEEP) ventilation 379, 380, 381 positron emission tomography (PET) 43, 290 recurrent disease surveillance 309–10 postoperative management 375–405 ascites 403–4 cardiac complications 376–9 chemotherapy 404 diaphragm surgery 242–3 discharge planning 405 endocrine disorders 399 fluids 389–91 germ cell tumors 78

422

hematologic complications 381–3 hormone replacement therapy 405 infectious morbidity 400–1 intensive care 375–6 liver surgery 256–8 nutritional support 395–9 pain management 399–400 pleural effusion 402–3 pulmonary complications 379–81 renal disorders 394–5 secondary cytoreductive surgery 325–6 serum electrolytes 391–4 acid–base disorders 393–4 calcium 392 magnesium 392–3 phosphate 393 potassium 392 sodium 391 sex cord–stromal tumors 80 thromboembolic complications 383–9 wound care 401–2 potassium levels postoperative management 392 preoperative assessment 40–1 Potter syndrome 23 preoperative evaluation 39–43 imaging techniques 42–3 laboratory and routine testing 40–1 risk assessment 39–40 tumor markers 41–2 preoperative preparation 43–7 antibiotic prophylaxis 44–5 bowel preparation 45–6 nutritional supplementation 43–4 second-look surgery 292 thromboembolic prophylaxis 46–7 presacral space 130 prevalence see epidemiology prevention strategies 6–7 oral contraceptive 6–7 prophylactic oophorectomy 7 screening 6 primary peritoneal carcinoma (PPC) 7, 18–19

Index.qxd

12/18/2006

15:38

Page 423

INDEX

primary surgery, reasons for 87–9 cytoreduction 88–9 diagnosis 88 palliation 88 staging 88 Pringle maneuver 247–8 progressive disease 306 psammocarcinoma 13 psammoma bodies 11, 13 pudendal nerve 133 pulmonary complications 242–3, 370–2, 379–81 pulmonary embolism 383, 385–9 pyramidalis muscle 60 racial trends 2 radiation exposure, as risk factor 2 radiofrequency ablation (RFA), liver disease 255–6 rectosigmoid colon resection 322–3 rectovaginal space 130 rectus abdominis muscle 59–60 rectus sheath 60–1 recurrent disease 306, 321 central pelvic recurrence 322–3 gastrocolic ligament 322 liver 324 spleen 323–4 surveillance for 307–10 imaging studies 308–9 radiographic measures of metabolic function 309–10 serum biomarkers 307–8 see also secondary cytoreductive surgery Reinke crystals 26 renal arteries 173, 267 renal disorders postoperative management 394–5 renal insufficiency 40–1 renal replacement therapy 395 renal veins 173, 267 residual disease 91–2 following secondary cytoreductive surgery 310–16 predictors of 296–7 respiratory acidosis 393–4

respiratory alkalosis 394 retractors 47 retroperitoneal anatomy 172–6 adrenal glands 176 kidney 175 lymphatics 176 ureter 175–6 vasculature 172–5 retroperitoneal lymph node dissection 65–8, 162–4 morbidity 186 vascular complications 186–8 para-aortic nodes 68–70 infrarenal 180–2 suprarenal 182–4 pelvic nodes 68, 178–80 retroaortic nodes 184–5 technique 68–70, 163–4, 178–85 see also lymph node metastases retroperitoneal lymphatic involvement 65, 98, 134, 162, 171–2 retropubic space of Retzius 130 right upper abdomen anatomy 227–37 adrenal gland 236–7 diaphragm 229–30 duodenum 235–6 gallbladder 234–5 kidney 236 liver 230–4 surgical procedures 237–59 diaphragm disease 237–43 gallbladder disease 258–9 incision 237 liver disease 243–58 porta hepatis disease 259 surgical approach 237 risk assessment, preoperative 39–40 risk factors 1–3 Roticulator stapling device 49 round ligaments 129 routes of spread 8–9 rupture of malignant tumors, intraoperative 64–5

423

Index.qxd

12/18/2006

15:38

Page 424

SURGERY FOR OVARIAN CANCER

sacral artery 173 sacral nerves 133 sacral plexus 133 salpingo-oophorectomy 64, 77 unilateral 74–5, 77, 79 Schiller–Duval bodies 21 sciatic nerve 133 sclerosing stromal tumors 25 sclerosis, intrapleural 371 screening 5–6 breast cancer 6 second-look surgery 289–302, 306 complications 293–4 history of 289–90 intraperitoneal port placement 293 laparoscopy 291–2, 338–9 equivalence to laparotomy 294–6 laparotomy 291 outcome predictors after negative second-look surgery 297–9 patient eligibility 290–1 patient positioning 292–3 positive second look patient management 299–302 residual disease predictors 296–7 secondary cytoreductive surgery 305–27 clinical applications 321–4 gastrocolic ligament resection 322 hepatic disease resection 324 resection of central pelvic recurrence 322–3 splenectomy 323–4 feasibility 316–17 morbidity 317 postoperative adjuvant therapy 325–6 selection criteria 317–21 predictors of surgical outcome 319–21 prognostic factors 317–19 surveillance for secondary disease 307–10 survival outcome 310–16 terminology 306–7 self-retaining retractors 47 serous adenocarcinomas 11–13 Sertoli cell tumor 25

424

Sertoli–Leydig cell tumor 25–6, 79 serum electrolytes postoperative management 391–4 acid–base disorders 393–4 calcium 392 magnesium 392–3 phosphate 393 potassium 392 sodium 391 preoperative assessment 40–1 serum tumor markers 41–2 recurrent disease surveillance 307–8 sex cord–stromal tumors 22–7, 78 early-stage tumor management 78–80 fibroma 24–5 fibrosarcoma 24–5 granulosa cell tumors 22–4, 79, 80 gynandroblastoma 26 sclerosing stromal tumors 25 Sertoli cell tumor 25 Sertoli–Leydig cell tumor 25–6, 79 sex cord tumors with annular tubules 26 steroid-cell tumors 26–7 Leydig cell tumors 26 not otherwise specified 27 stromal luteoma 26 thecoma 24 shock 377–8 short bowel syndrome 361–2 side-to-side functional end-to-end anastomosis (FEEA) 152–3, 209 sigmoid arteries 200 sigmoid colon 128 radical oophorectomy, type II modification 140–53 see also intestinal anastomosis sigmoidoscopy 363 preoperative 41 sinus tachycardia 378 small bowel obstruction 351–62 complications of surgery 352, 361–2 blind loop syndrome 361 short bowel syndrome 361–2

Index.qxd

12/18/2006

15:38

Page 425

INDEX

diagnostic evaluation 352–3 management 353–4 postoperative 222 surgical techniques 354–61 gastrojejunostomy 354–6 gastrostomy decompression 360–1 ileostomy 358–9 small-bowel bypass 356–8 small-bowel resection 359 small intestine 196–7 ileocecal resection 213–14 resection 211–12 vasculature 196–7 see also bowel involvement; intestinal tract; small bowel obstruction sodium levels postoperative management 391 preoperative assessment 40–1 space of Treves 196 specialist care access to 103–7 ovarian cancer care team 106–7 spleen 264–5 splenectomy 270–8 anterior approach 271–3 complications 275–8 literature review 276–8 posterior approach 273–5 secondary cytoreductive surgery 323–4 splenic artery 263, 265 splenic flexure 198, 266 splenic vein 265 splenorraphy 277–8 spread, routes of 8–9 squamous carcinoma 16 staging 9–11, 58, 88 adjuvant therapy and 70–2 germ cell tumors 77–8 laparoscopic 337–8, 345–6 low malignant potential (LMP) tumors 72, 177–8 rationale for 57–9 surgical technique 63–70 appendectomy 70

biopsies 65 exploration 64 omentectomy 65 primary tumor management 64–5 retroperitoneal lymph node dissection 65–70 stapling devices, automated 49–52 see also intestinal anastomosis steroid-cell tumors 26–7 Leydig cell tumors 26 not otherwise specified 27 stromal luteoma 26 stomach 261–3 partial gastrectomy 283–6 vasculature 262–3 stromal luteoma 26 stromal tumors see sex cord–stromal tumors struma ovarii 20 superior mesenteric artery (SMA) 173, 196, 199, 266 superior mesenteric vein (SMV) 197, 200, 266 supraclavicular adenopathy 190 survival following cytoreductive surgery 94–5 secondary cytoreductive surgery 310–16, 319–21 outcome predictors 319–21 after negative second-look surgery 297–9 sutures 63 symptoms, early-stage cancer 57 T-incision 62 tachycardia 376–7, 378 talc, as risk factor 1–2 teratomas 19–20 immature 20 mature cystic 19–20 monodermal 20 thecoma 24 thoracentesis 370, 371, 402–3 thoracoabdominal nerves 61 thoracoabdominal (TA) stapler 49, 52 thrombocytopenia 40 thrombocytosis 40

425

Index.qxd

12/18/2006

15:38

Page 426

SURGERY FOR OVARIAN CANCER

thromboembolic complications 46–7, 383–9 prophylaxis 46–7 total inflow occlusion, liver 247–8 total parenteral nutrition (TNP) bowel obstruction management 353, 354 postoperative 396–7 transabdominal ultrasonography 42 transfusion reactions 381–2 transfusion-related acute lung injury (TRALI) 382 transitional cell carcinoma 15–16 transureteroureterostomy (TUU) 157 transvaginal ultrasonography (TVUS) 42 screening 6 transversalis fascia 61 transverse loop colostomy 364 transversus abdominis muscle 60–1 triangular ligaments 230–2 tumor markers 41–2 undifferentiated carcinoma 16 upper abdomen see left upper abdomen; right upper abdomen ureter 128, 175–6 complications of decompression 368–9 internal stents 368 obstruction 367–8 ureterectomy, partial 154–8

426

uretero–ilio–neocystotomy (UINC) 156–7 ureteroneocystotomy (UNC) 155 ureteroureterostomy 155 urinary tract obstruction 367–9 complications of surgery 368–9 resection 153 uterosacral ligaments 129 uterus 127 vasa recti 196 vascular anatomy 7, 61, 130–1 vena cava inferior (IVC) 173–5 injury to 188 ventilation support 379–80 vesicovaginal space 130 vessel sealers 48 viral infection, as risk factor 2 water test 146 white line of Toldt 198 windows of Deaver 196 wound care 401–2 yolk sac tumors 21