Imaging of Vertebral Trauma (Cambridge Medicine)

Citation preview

This page intentionally left blank

Imaging of Vertebral Trauma Third Edition

Imaging of Vertebral Trauma Third Edition Richard H. Daffner, MD, FACR Professor of Radiologic Sciences, Drexel University College of Medicine and Department of Diagnostic Radiology, Allegheny General Hospital, Pittsburgh, USA

CAMBRID GE UNIVERSIT Y PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo, Mexico City Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521897013 First Edition © Aspen Publishers 1988 Second Edition © Lippincott–Raven Publishers 1996 Third Edition © Cambridge University Press 2011 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First Edition published by Aspen Publishers 1988 Second Edition published by Lippincott–Raven Publishers 1996 Third Edition published by Cambridge University Press 2011 Printed in the United Kingdom at the University Press, Cambridge A catalog record for this publication is available from the British Library Library of Congress Cataloging in Publication data

ISBN 978-0-521-89701-3 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate. Every effort has been made in preparing this book to provide accurate and up-to-date information which is in accord with accepted standards and practice at the time of publication. Although case histories are drawn from actual cases, every effort has been made to disguise the identities of the individuals involved. Nevertheless, the authors, editors, and publishers can make no warranties that the information contained herein is totally free from error, not least because clinical standards are constantly changing through research and regulation. The authors, editors, and publishers therefore disclaim all liability for direct or consequential damages resulting from the use of material contained in this book. Readers are strongly advised to pay careful attention to information provided by the manufacturer of any drugs or equipment that they plan to use.

In remembrance of Morris M. Daffner, William F. Barry, Jr., MD, and George J. Baylin, MD – teachers, scholars, and friends – and Earl L. Weaver III, whose example inspired all

Contents List of contributors ix Preface to the Third Edition xi Preface to the Second Edition xiii Preface to the First Edition xv Acknowledgments xvii 1

Overview of vertebral injuries 1 Richard H. Daffner

7

Mechanisms of injury and their “fingerprints” 88 Richard H. Daffner

2

Anatomic considerations 12 Richard H. Daffner

8

3

Biomechanical considerations 36 Richard H. Daffner

Radiologic “footprints” of vertebral injury: the ABCS 126 Richard H. Daffner

9

Imaging of vertebral trauma I: indications and controversies 45 Richard H. Daffner

Vertebral injuries in children 165 Geetika Khanna and Georges Y. El-Khoury

10 Vertebral stability and instability 181 Richard H. Daffner

4

5

Imaging of vertebral trauma II: radiography, computed tomography, and myelography 53 Richard H. Daffner

6

Imaging of vertebral trauma III: magnetic resonance imaging 72 Bryan S. Smith and Richard H. Daffner

11 Normal variants and pseudofractures 192 Richard H. Daffner

Index

221

vii

Contributors

Richard H. Daffner, MD, FACR Professor of Radiologic Sciences, Drexel University College of Medicine, Department of Diagnostic Radiology, Allegheny General Hospital, Pittsburgh, PA, USA

Geetika Khanna, MD Assistant Professor of Radiology, Mallinckrodt Institute of Radiology, Barnes-Jewish Hospital, St. Louis, MO, USA

Georges Y. El-Khoury, MD Professor of Radiology and Orthopedic Surgery, Department of Radiology, University of Iowa Hospitals and Clinics, Iowa City, IA, USA

Bryan S. Smith, MD Musculoskeletal Imaging Fellow, Department of Diagnostic Radiology, Allegheny General Hospital, Pittsburgh, PA, USA

ix

Preface to the Third Edition

The imaging methods used to evaluate patients with suspected vertebral injuries have undergone radical changes since the publication of the second edition of Imaging of Vertebral Trauma in 1996. The most significant of these changes has been the ascendency of computed tomography (CT) to become the primary tool for studying these patients. Radiography now assumes a secondary role, serving mainly for follow-up of known injuries or as a tool to solve problems with CT studies, such as motion or metallic artifacts. Furthermore, there has been an ongoing dialog in the radiologic and trauma literature regarding the indications for imaging in trauma patients, as well as the methods of choice. Of most recent note are the issues of high-radiation dose associated with CT studies as well as the continuing debates on health care reform and cost containment. The first edition dealt mainly with radiography supplemented with polydirectional or computed tomography and magnetic resonance (MR) imaging. The second edition expanded the discussion of the roles of MR in vertebral injuries. This new edition presents an in-depth discussion on the indications and methods of imaging the spine based on the

evidence available in the current literature. Each chapter has been revised with those precepts in mind and the majority of the illustrations have been changed to represent state-ofthe-art imaging. There are still a large number of radiographs since they present teaching points on principles that transfer directly to CT studies. Furthermore, this book is used in parts of the world where high-speed multislice CT scanners may not be available, as they are in the United States. The section on imaging has been divided into three chapters: an introduction, radiography and CT, and MR imaging. A new chapter on pediatric injuries has been added by Drs. George El-Khoury and Geetika Khanna. Dr. Bryan Smith has revised and updated the chapter on MR imaging. I hope that the third edition of Imaging of Vertebral Trauma will continue to fill the gaps that were present in the first two editions and that it will provoke a thoughtful reassessment of the imaging of patients with suspected vertebral or spinal cord injury. Richard H. Daffner, MD, FACR

xi

Preface to the Second Edition

Since the publication of the first edition of Imaging of Vertebral Trauma in 1988, major developments have been made in the evaluation of patients with suspected vertebral injury. Most of these have been in the realm of magnetic resonance imaging, but new reports have also given us a better understanding of some important anatomic relationships. There is a greater awareness of the subtle signs of injury, and there has been a reassessment of exactly how “significant” many of them may be. The current emphasis on health care reform and cost containment has prompted a reassessment of indications for radiography and computerized imaging of the vertebral column. The first edition dealt mainly with plain film radiography supplemented with polydirectional or computed tomography and magnetic resonance imaging. This edition continues that focus by addressing some of the new issues that have surfaced since 1988. In addition, two contributing authors have written chapters. Dr. Andrew L. Goldberg, a neuroradiologic colleague of mine at Allegheny General Hospital, has written an indepth review of the use of magnetic resonance imaging in the

diagnosis of vertebral and spinal cord injuries. Dr. Stanley P. Bohrer, a musculoskeletal radiologist at Bowman Gray School of Medicine, has written a chapter on the use of flexion and extension radiographs in patients with suspected ligamentous injuries in the cervical region. A third new chapter deals with the biomechanics of the vertebral column and biomechanical considerations in vertebral injury. The topic of vertebral stability and instability is now described in a chapter of its own. Finally, each chapter has been carefully reviewed and revised to reflect the state of the art in vertebral imaging, and the index has been expanded and made more user friendly. As the centennial of the discovery of the roentgen ray is celebrated, we should be cognizant of how far we have come in so short a time. I hope that the second edition of Imaging of Vertebral Trauma will fill the gaps that were present in the first edition and that it will provoke a thoughtful reassessment of the imaging of patients with suspected vertebral or spinal cord injury. Richard H. Daffner, MD

xiii

Preface to the First Edition

Vertebral trauma is a major cause of permanent disability. Although there has been an increasing number of vertebral injuries due to motor vehicle accidents, improved medical technology has salvaged the lives of individuals who suffer what were once considered uniformly fatal injuries. The key to the administration of prompt therapy and rehabilitation is the ability to properly diagnose the full extent of these injuries. The discovery of the roentgen ray was the first major technological breakthrough in diagnosing vertebral trauma, and this method remained the chief method for diagnosis until the development of computed tomography and magnetic resonance imaging. With these methods it is now possible to define the full extent of injury and, in the latter method, to determine the extent of spinal cord involvement. I became interested in the subject of vertebral injury through my long and close association with Dr. John A. Gehweiler, Jr., who described many signs of subtle injury to the cervical vertebrae. The advent of multiplanar imaging confirmed the validity of the signs described by Dr. Gehweiler and other individuals

interested in vertebral trauma. This book grew out of a series of lectures that I have given over the past decade and represents a systematic and practical approach to the radiography of vertebral trauma. This book is not encyclopedic in scope and does not describe every variation of every type of vertebral injury. It does, however, provide a working basis for the practicing radiologist in the community hospital as well as in the large medical center, who is often the first person called on to interpret radiographs of a patient with vertebral injury. The book relies on the premise that all injuries (vertebral and nonvertebral) occur in a predictable and reproducible fashion that is solely dependent on the mechanism of injury. As such, each type of injury produces indelible signs that I have termed “fingerprints.” By following this logical approach and by applying the principles outlined in the text, the reader will gain confidence in his or her diagnostic skills and ability to diagnose even the most subtle injury. Richard H. Daffner, MD

xv

Acknowledgments

No book of this scope could be produced without the technical assistance of many individuals. I am extremely grateful to Maggie Cauley for her efforts in manuscript preparation, editing, and collation. I am indebted to Donna Spillane, of the Creative Services Department of Allegheny General Hospital, for production of the original illustrations. I thank Randy McKenzie, medical illustrator, for the new drawings as well as Scott Williams for the superb original artwork. Many thanks to Peter Brondar, of the Computer Laboratory at Carnegie Mellon University, for rescuing electronic images from the previous editions and restoring them so they could be used.

I also acknowledge the artwork of Debbie Whitman and Maurice Williams, as well as the photography of Gary Stark and Douglas Whitman for the illustrations that were reused from the first and second editions of Imaging of Vertebral Trauma. For their assistance in clinical correlation of the case material in the book, I wish to thank Aurelio Rodreguez, MD, Head of the Trauma Center of Allegheny General Hospital, and his colleagues. Finally I thank my dear wife, Alva, for her encouragement, support, and patience, as well as her editorial skills in proofreading and editing the manuscript.

xvii

Chapter

1

Overview of vertebral injuries Richard H. Daffner

In the course of human history, no injuries have evoked greater fear than vertebral fracture and dislocation. They are among the most devastating of insults and result in a gamut of abnormalities ranging from mild pain and discomfort to severe paralysis and even death. Despite improved technology for the diagnosis and treatment of vertebral fracture and dislocation, the physician who is confronted with a spine-injured patient often feels incapable of interpreting the imaging studies that would delineate the full extent of injury. This book presents the systematic approach to the diagnosis of vertebral trauma that my colleagues and I have used for the interpretation of images (radiographs, computed tomography [CT] scans, and magnetic resonance [MR] images) of patients suspected of having vertebral injury. Furthermore, it amplifies several concepts that I have developed – namely, that vertebral injuries occur in a predictable pattern, that the imaging findings produced by a generic injury are similar, and that findings for injuries caused by the same mechanism are identical no matter where they are encountered within the vertebral column [1]. This chapter defines the descriptive terms pertaining to fractures and dislocations, reviews the terminology used for reporting these abnormalities, and discusses basic mechanisms of injury. Succeeding chapters discuss anatomy, biomechanics, imaging methods available for diagnosing vertebral injuries, and the basic diagnostic principles that make possible a logical and systematic approach to diagnosing vertebral injuries and to determining whether or not vertebral stability has been maintained. The final chapter discusses pseudofractures and normal variants. In addition, we will discuss the current controversies in imaging patients suspected of having vertebral injuries.

Fractures Most medical dictionaries define a fracture as a disruption, either complete or incomplete, in the continuity of a bone, physis, or cartilaginous joint surface. I prefer a definition that has a more practical significance: a fracture is a soft tissue injury in which a bone is broken. This definition is of greatest importance in injuries to the skull and to the vertebral column, where the bony disruption itself may be the least important component of the injury, and damage to

the meninges, brain, spinal cord, blood vessels, or peripheral nerves is more serious. A number of descriptive terms are used in regard to fractures. Most of these are applicable to the peripheral skeleton. A complete fracture is one in which both cortices of a bone have been broken; an incomplete fracture involves only one cortex. In closed (or simple) fractures, there is no communication of the fracture site with the exterior of the body; in open (or compound) fractures, there is communication between the fracture site and the external environment. Most fractures of the vertebral column are closed. Open fractures generally result from missile injuries. Operative intervention converts a closed fracture to an open one. Fractures can be the result of either direct or indirect injury. In a direct injury, force is applied directly to the bone, and fracture occurs at the site of impact. In the vertebral column, this is most likely to occur in a spinous process (Fig. 1.1) [2]. Most vertebral injuries result from indirect trauma in which force is applied at a distance from the involved vertebra (Fig. 1.2). In the case of a cervical injury, a loading force applied to the head or trunk is transmitted directly to the vertebral column, producing a deformity as a result of exceeding the normal physiologic range of motion (as explained in Chapter 3). Sudden acceleration or deceleration of the head relative to the trunk, or vice versa (as often occurs in motor vehicle crashes and falls), can also produce indirect injury, particularly in the cervical region [1,3–15].

Joint injuries Joint injuries result from the same types of force that produce fractures. The mildest form of joint injury is a ligamentous sprain caused by stretching of the ligament fibers beyond their normal range of elasticity. This produces small tears and hemorrhages. Rupture of a ligament may occur with more severe injury. The only difference between a sprain and a rupture is the degree of injury. Sprain or rupture of a ligament or a combination of ligaments can result in three types of joint instability: occult instability, subluxation, and dislocation. Occult instability is recognizable radiographically only when a joint is stressed in flexion or extension (Fig. 1.3). Subluxation is a more severe joint injury in which there is a partial loss of contact between

1

1 Overview of vertebral injuries

A

B

Fig. 1.1 Spinous process (“clay-shoveler”) fracture. (A) Sagittal reconstructed CT image shows the fracture in the spinous process of C7 (large arrow). Note the teardrop extension fracture of the body of C2 (small arrow). The small ossific density along the inferior aspect of C3 is another avulsion fracture. (B) Axial image shows the fracture (arrow).

Fig. 1.2 Flexion teardrop fracture of C5. Patient dove into shallow water. Note the retrolisthesis of the body of C5 (arrowhead) and widening of the facet joints (arrows).

A

C

2

B

Fig. 1.3 Flexion sprain C4–C5. (A) Lateral radiograph shows reversal of lordosis and widening of the interlaminar space between C4 and C5 (*). (B) Frontal radiograph shows widening of the interspinous space (double arrow). (C) T1-weighted MR sagittal image shows rupture of the posterior longitudinal ligament (arrow).

1 Overview of vertebral injuries

A

B

Fig. 1.4 Flexion sprain C6–C7. (A) Lateral radiograph shows widening of the interlaminar space (*) and wide facet joints. (B) Sagittal reconstructed CT image shows the subluxation of the facet joint (arrow).

A

B

Fig. 1.5 Atlanto-axial dislocation. Axial (A) and sagittal (B) reconstructed CT images show widening of the predental space (*).

3

1 Overview of vertebral injuries

A

B

E

C

D

F

Fig. 1.6 Unilateral facet lock C3–C4. (A,B) Lateral radiograph (A) and sagittal reconstructed CT image (B) show anterolisthesis of C3 on C4 (arrows). Note the pillar duplication producing a “bowtie sign” (* in A). (C) Sagittal reconstructed CT image shows the locked facet (arrow) with multiple fracture fragments. (D) Sagittal CT image further medial shows a lamina fracture (arrow) as well as the anterolisthesis and a fracture off the inferior body of C3. (E) Axial CT image shows pillar and pedicle fractures extending into the transverse foramen on the left (arrows) as well as the body and laminar fractures of C3. (F) Axial CT image shows facet fragmentation as well as an unpaired facet on the left (arrow). Compare with the normal “hamburger bun” appearance of the facet joint on the right.

apposing joint surfaces (Fig. 1.4). Dislocation (luxation) is the complete loss of contact between the apposing articular surfaces (Fig. 1.5). The term locking refers to an abnormal relationship between articular surfaces that results from dislocation (Fig. 1.6).

Descriptive terminology Fractures and dislocations in the axial skeleton are described, with one important exception, by the same terms as those in the peripheral skeleton. By convention, an injury should be defined at the level or levels at which it has occurred. When an injury occurs at a disc level, it is defined by the vertebra above it. Thus, an injury to the C4–C5 disc space is said to have occurred at the C4 disc space.

4

Descriptive terms such as avulsion, impaction, distraction, rotation, compression, and burst should all be used. The plane of fracture (horizontal, transverse, coronal, or sagittal) and displacement of major fragments should also be identified and described. In addition, if a fracture appears to have a pathologic etiology, this should be stated. Figures 1.7 through 1.11 show examples of various fractures and the descriptive terminology used for these injuries. Subluxations and dislocations are described by relating the direction taken by the upper vertebra with regard to the one below. This is in contradistinction to the descriptive terminology used for peripheral fractures, in which the position and angulation of the distal fragments are described in relation to the proximal fragments. Figures 1.12 through 1.14 show variations of joint injuries and their descriptions.

1 Overview of vertebral injuries

A

B

C

Fig. 1.7 Simple compression fracture of L4. (A) Lateral radiograph shows the compression fracture of the anterior superior margin (arrowhead). The posterior vertebral body line (arrows) is intact. (B) Axial CT image shows fracture of the anterior margin of the vertebra with an intact posterior vertebral body line. (C) Sagittal reconstructed CT image shows the fracture to involve the anterior superior margin of the vertebra only (arrow).

A number of terms are used throughout this book in regard to the mechanisms of injury [1,2,4,7–9,13,14]. Although these terms are defined in further detail in Chapters 3 and 7, they require a brief description at this time. Flexion injuries result from a forward bending motion of the vertebral column at any level. Such injuries are the result of either posterior impact of a force on the vertebral column or anterior impact of the torso on a solid object [7,14,15]. Extension injuries are caused by a posterior bending of the vertebral column in response to either an anterior force or sudden deceleration against a solid object posteriorly [1,12,13,15]. Shearing injuries are the result of horizontal or oblique linear forces being transmitted to the vertebral column from any

direction. Limited motion in flexion, extension, and rotation are permitted within the vertebral column. However, horizontal (translational) or oblique linear motion is never normal [1,15]. Rotational injuries result from abnormal torque applied to the vertebral column. The normal vertebral column is permitted limited motion in flexion and extension and even less motion in rotation [1,15]. Thoracolumbar rotary injuries usually result in severe neurologic compromise because they are extremely disruptive [1]. All of these mechanisms may occur in combination. In addition, they take into account the effect of axial loading.

5

1 Overview of vertebral injuries

A

B

C

D

Fig. 1.8 Burst fracture of L3. (A) Lateral radiograph shows compression of the superior portion of the vertebral body as well as retropulsion of bone fragments from the posterior body (arrow). (B) Frontal radiograph shows widening of the interpedicle distance (double arrow). (C) Axial CT image shows the retropulsed bone fragment (*) narrowing the vertebral canal by 50%. (D) Sagittal reconstructed CT image shows the retropulsed fragment in the canal (arrowhead).

6

1 Overview of vertebral injuries

A

B

Fig. 1.9 Unilateral Jefferson fracture of C1on the left. (A) Open-mouth radiograph shows offset of the lateral mass of C1 on the left (arrow). (B) The CT image shows fractures of the anterior and posterior arches of C1 on the left (arrows).

A

B

Fig. 1.10 Chance-type fracture of L2. (A) Frontal radiograph shows horizontal fractures through the body, pedicles (arrowheads), and transverse process processes. (B) Lateral radiograph shows the posterior extension of the fracture through the pedicles (arrow).

7

1 Overview of vertebral injuries

A

8

B

A

B

C

D

Fig. 1.11 Pathologic fractures. (A) Lateral radiograph shows complete collapse of the body of C4 with resulting kyphosis due to metastatic disease. There is destruction of C3 and C5. (B) Disc space infection has resulted in collapse of the bodies of T8 and T9.

Fig. 1.12 Extension injuries. (A,B) Lateral radiograph (A) and sagittal reconstructed CT image (B) show an extension teardrop fracture of the inferior body of C2 (arrows). Note the prevertebral soft tissue swelling in A (*). (C) Sagittal CT image shows a hyperextension sprain at C6–C7 in another patient. Note the wide disc space (arrow). (D) Sagittal short-tau inversion recovery (STIR) MR image shows the torn anterior longitudinal ligament (arrow) as well as an occult fracture of the body of C6 (arrowheads).

1 Overview of vertebral injuries

A

B

Fig. 1.13 Unilateral facet lock. (A) Sagittal CT image shows a fracture of the facet and locking (arrows). (B) Axial CT image shows severely comminuted fractures of the facet and lamina on the left as well as locking (arrow).

Etiology of vertebral injuries Most vertebral injuries result from motor vehicle crashes [3,16,17], which account for 85% of the patients seen at the Trauma Center of Allegheny General Hospital. In almost all of these cases, three elements coincide: speed, generally greater than 15 miles per hour over the posted limit; alcohol intoxication, greater than 0.08 mg/dL (the legal limit in most states); and lack of the use of seat belts, which might have prevented most injuries. Interestingly, a higher incidence of soft tissue injury (sprains) occurs in belted vehicle occupants [18]. Although air bags have been available in the USA on all domestic and most foreign cars manufactured after 1993, not enough data have been gathered by trauma centers to determine their effectiveness in preventing vertebral injuries, particularly when seat belts are not used in conjunction with the air bag.

Approximately 14% of vertebral injuries encountered at my institution resulted from falls, primarily in patients over 65 years of age. Miscellaneous causes, such as diving accidents and missile (gunshot) injuries, account for the remaining 1% of injuries [1]. The Trauma Center of Allegheny General Hospital admits some 400 patients with vertebral injury each year. As a level I trauma center, we primarily treat victims of high-speed vehicular trauma, falls, and industrial accidents. Spinal cord injury is a frequent occurrence in patients with vertebral trauma; at my institution, it is found in 40% of spine-injured patients. Of patients with head injuries, 10–15% also have vertebral injury with spinal cord involvement. Not surprisingly, 75% of patients with spinal cord injury have associated injuries, many of which are life threatening. Furthermore, the conditions of up to 10% of patients with spinal cord injury were worsened by prehospital care, despite efforts to reduce the incidence through education of paramedical personnel. Surprisingly, these percentages have not changed in the quarter century I have been working at the Allegheny Trauma Center. Similar numbers are encountered by colleagues at other trauma centers in the USA.

9

1 Overview of vertebral injuries

A

B

Fig. 1.14 Thoracic dislocation with facet locking in an abused child. (A) Lateral radiograph shows dislocation of T11 on T12 (arrow). (B) Sagittal CT image shows the facet locking (arrow).

10

1 Overview of vertebral injuries

References 1.

2.

3.

4.

5.

6.

Daffner RH, Deeb ZL, Rothfus WE. “Fingerprints” of vertebral trauma: a unifying concept based on mechanisms. Skeletal Radiol 1986;15:518–525. Cancelmo JJ Jr. Clay shoveler’s fracture: a helpful diagnostic sign. AJR Am J Roentgenol 1972;115:540–543. Alker GJ Jr., Oh YS, Leslie EV, et al. Postmortem radiology of head and neck injuries in fatal accidents. Radiology 1975;114:611–617. Atlas SW, Regenbogen V, Rogers LF, et al. The radiographic characterization of burst fractures of the spine. AJR Am J Roentgenol 1986;147:575–582. Bohlman HH. Acute fractures and dislocations of the cervical spine: an analysis of three hundred hospitalized patients and review of the literature. J Bone Joint Surg 1979;61A:1119–1142. Braakman R, Penning L. Injuries of the Cervical Spine. London: Excerpta Medica, 1971.

7.

8.

9.

10.

11.

12.

13.

Chance GQ. Note on a type of flexion fracture of the spine. Br J Radiol 1948;21:542–543. Daffner RH, Deeb ZL, Rothfus WE. Thoracic fractures and dislocations in motorcyclists. Skeletal Radiol 1987;16:280–284. Dehner JR. Seat belt injuries of the spine and abdomen. AJR Am J Roentgenol 1971;111:833–843. Harris JH Jr. Radiographic evaluation of spine trauma. Orthop Clin North Am 1986;17:75–86. Harris JH Jr., Mirvis SE. The Radiology of Acute Cervical Spine Trauma, 3rd edn. Baltimore, MD: Williams & Wilkins,  1996. Holdsworth FW. Fractures, dislocations, and fracture–dislocations of the spine. J Bone Joint Surg 1970;52A:1534–1551. Roaf R. A study of the mechanics of spinal injuries. J Bone Joint Surg 1960; 42B;810–823.

14. Smith WS, Kaufer H. Patterns and mechanisms of lumbar injuries associated with lap seatbelts. J Bone Joint Surg 1969;51A:239–254. 15. White AA, Panjabi MM. Clinical Biomechanics of the Spine, 2nd edn. Philadelphia, PA: JB Lippincott, 1990. 16. Jonsson H, Jr., Bring G, Rauschning W, Sahlsedt B. Hidden cervical spine injuries in traffic accident victims with skull fractures. J Spinal Disord 1991; 4:251–263. 17. Stabler A, Eck J, Penning R, et al. Cervical spine: postmortem assessment of accident injuries: comparison of radiographic MR imaging, anatomic, and pathologic findings. Radiology 2001; 221:340–346. 18. Bourbeau R, Desjardins D, Maag U, et al. Neck injuries among belted and unbelted occupants of the front seat of cars. J Trauma 1993;35:794–799.

11

Chapter

2

Anatomic considerations Richard H. Daffner

The vertebral column is composed of 33 irregular bones that extend from the base of the skull through the entire length of the neck and trunk. With the attachment of muscles, ligaments, and supporting intervertebral discs, the column forms a strong, flexible support for the body while protecting the spinal cord and its surrounding meninges. The column can be divided into the upper 24 presacral vertebrae, which remain separate throughout life, and the fixed vertebrae, which constitute the five sacral and four coccygeal segments. Although a detailed explanation of the anatomy is beyond the scope of this text, an understanding of the basic anatomic features of the vertebral column is necessary to appreciate the abnormalities that may be encountered in spine-injured patients. Readers who desire a detailed treatment of the anatomy are referred specifically to The Radiology of Vertebral Trauma by Gehweiler and colleagues [1] or to a basic textbook on anatomy [2,3].

Abbreviations used in the figures For the two-part figures in this chapter, part A is a photograph and part B is a radiograph unless otherwise indicated in the legend. The abbreviations used within the figures are explained below. Aa anterior arch of atlas Al arcuate line Ap articular pillar B body C central tubercule of atlas D dens F transverse foramen Ia inferior articular facet L lamina Lm lateral mass M mammillary process P pedicle Pa posterior arch of atlas Pi pars interarticularis R rib facet S spinous process Sa superior articular facet Sc sacral canal

12

Sf Si Sl SS T U

sacral foramen sacroiliac joint spinolaminar line sacral spine transverse process uncinate process

Normal vertebral development Each vertebra develops from several ossification centers. For the purposes of this discussion, three patterns will be considered: the atlas (C1), the axis (C2), and the remainder of the vertebral column. Bailey [4], in a classic article, described the normal developmental anatomy in the cervical region in infants and children. He was one of the first to correlate this anatomy with the normal radiographic appearance in the pediatric age group. The atlas develops from three separate ossification centers (centra). The first is that of the anterior arch or “body” (centrum), and it is not ossified at birth. Ossification usually begins in the center during the first year after birth. Occasionally, two centers are present. If the anterior arch fails to develop, forward extension of the neural arches may take its place, with resultant anterior arch cleft formation. There are bilateral neural arch ossification centers, which form the lateral masses and the posterior arch. The neural arch ossification centers appear bilaterally about the seventh fetal week. Bailey [4] reported that the anterior-most portion of the superior articular facet of the lateral mass is usually formed by the ossification center of the anterior arch. There is a synchondrosis of the posterior arch, which generally unites by the third year. Failure of union results in absence of the spinolaminar line (the junction of the laminae to form the spinous processes), a common finding. The neurocentral synchondroses join the ossification centers of the neural arches and the anterior arch. These generally fuse by the seventh year. Finally, there is an occasional ossification in later life of the ligaments surrounding the superior vertebral notch, through which the vertebral artery passes (Kimmerle anomaly). Figure  2.1 shows the atlas ossification centers and synchondroses [1,4]. The axis develops from four primary ossification centers. The first is the body (centrum), which begins to ossify by the

2 Anatomic considerations

Fig. 2.1 Atlas ossification centers and synchondroses.

Fig. 2.2 Axis ossification centers and synchondroses.

fifth fetal month. The dens actually arises from two separate ossification centers that fuse with each other by the seventh fetal month to form a single bone by birth. The dens is separated from the body by a synchondrosis that fuses between three and six years of age. There are two neural arch centers, which will form articular pillars, pedicles, laminae, and eventually a spinous process. These generally appear by the fifth fetal month and fuse with each other by the seventh fetal month. The synchondrosis between the dens and the neural arch is present at birth and also fuses by age three to six years [1,4]. Occasionally, the presence of this synchondrosis on a lateral radiograph may be misinterpreted as evidence of a fracture [5]. Figure 2.2 shows the ossification centers and synchondroses of the axis. The remaining cervical vertebrae, as well as those in the thoracic and lumbar regions, develop from three primary ossification centers. The ossification center of the body appears by the fifth fetal month. Bilateral centers of the neural arches that form the pedicles, laminae, and eventually the spinous process appear by the seventh to ninth fetal week. Like the atlas, the synchondrosis between the body and the neural arch fuses between three and six years of age. The synchondrosis between the spinous processes generally unites by two to three years of age. An unfused spinous synchondrosis, when seen through the vertebral body, is sometimes misinterpreted as a vertical fracture of the body. Each vertebra also has a superior and inferior apophyseal ring along the disc margin. These ring apophyses appear at puberty and unite with the body at approxi-

mately 25 years of age. Occasionally, they fail to unite but still have a typical rounded appearance [4,5]. Figure 2.3 shows the developmental aspects of the typical cervical vertebra. The thoracic and lumbar vertebrae have similar development. Vertebral development provides many opportunities for a large number of normal variants and anomalies to develop [6]. These will be discussed in detail in Chapter 11.

Bones All of the movable presacral vertebrae except the atlas (C1) and the axis (C2) have certain common characteristics that define the “typical” vertebra. The basic parts of a vertebra are (1) the body, which is weight-bearing and located anteriorly; and (2) the vertebral arch, which acts as a protective shell for the spinal cord and its meninges and blood vessels and which is located posteriorly. The vertebral arch comprises two pedicles and two laminae. The pedicles attach the arch to the vertebral body. The laminae join the pedicles and form the posterior wall of the vertebral foramen, which encloses the spinal cord and its coverings and vessels. The vertebral foramina, when normally aligned, form the vertebral canal. The vertebral arch supports seven projections or processes: four articular processes, two transverse processes, and one spinous process (Fig. 2.4). The transverse processes and spinous process serve as levers on which muscles pull. The orientation of the articular processes determines the direction and degree of motion of the vertebral

13

2 Anatomic considerations

column [1–5,7]. This will be discussed in greater detail in Chapter 3.

Cervical vertebrae The cervical vertebral column can be divided into typical and atypical vertebrae. Vertebrae C3 through C6 constitute the typical vertebrae; C1, C2, and C7 are the atypical vertebrae. All cervical vertebrae have a common distinguishing feature – a foramen in each of the transverse processes [1,3].

Typical cervical vertebrae In the typical cervical vertebra (Figs. 2.5 to 2.10), the vertebral body is elliptical, being wider in its transverse diameter than in its sagittal diameter. The upper surface of a typical vertebra has a slightly convex appearance from front to back and a concave appearance transversely because of the presence of the uncinate processes. The superoanterior surface is beveled to receive the protruding rim of the anteroinferior surface of the body above. Conversely, the posteroinferior surface of the body is concave in its sagittal direction and convex transversely to accommodate the uncinate processes of the vertebra below [1–3]. On a lateral radiograph, the posterior margin of the vertebral body appears as a sclerotic vertical line that is uninterrupted. This is the posterior vertebral body line (posterior cortical line), an important marker of integrity of the vertebral canal (Fig. 2.7B). Any displacement, duplication, rotation, angulation, or absence of this line is abnormal [8]. Fig. 2.3 Ossification centers of a typical cervical vertebra.

14

A

B

A

B

Fig. 2.4 Parts of a “generic” vertebra (L2) view from below.

Fig. 2.5 Typical cervical vertebra (C5), anterior view.

2 Anatomic considerations

A

B

Fig. 2.6 Typical cervical vertebra (C5), posterior view.

A

B

Fig. 2.7 Typical cervical vertebra (C5), lateral view.

A

B

Fig. 2.8 Typical cervical vertebra (C5), view from above.

A

B

Fig. 2.9 Typical cervical vertebra (C5), view from below.

Uncinate processes are not present at birth but develop during adolescence, reaching full height in the adult. Anthropomorphically, they are believed to prevent lateral displacement during cervical motion. They project cranially from the upper lateral margins of the posterior aspect of the

vertebral bodies of C3 through C6 and are found along the posterolateral upper margin of C7 and T1 [1–5]. Small notches develop along the undersurface of the adjacent vertebrae at the same time. The cervical pedicles are short and stout and arise from the posterolateral aspect of the vertebral body. They are

15

2 Anatomic considerations

A

B

Fig. 2.10 Typical cervical vertebra (C5), oblique view.

directed posteriorly and laterally. Typically, they are notched equally on both superior and inferior surfaces [1]. The laminae are narrow and thin. The laminae join posteriorly to form the spinous process. On a lateral radiograph, this junction can be recognized as a sclerotic line termed the spinolaminar line, an important normal landmark. The ring formed by the laminae, pedicles, and vertebral body is called the vertebral foramen; it has a triangular shape in the cervical region [1–3]. The superior and inferior articular processes sit on either side of a rhomboid articular pillar, often misnamed the lateral mass. The pillars project laterally from the junction point of the lamina and the pedicle. Each articular process contains a facet that articulates with its neighbor. When viewed on a lateral radiograph, the inferior borders of the pillars overlap the superior aspect of their neighbors in an orderly manner like shingles on a roof. This appearance has been termed imbrication [1–3]. The spinous processes are short and directed posteriorly and inferiorly. Typically, they are bifid in Caucasians and single in those of African descent. Similarly, the transverse processes are short and thin and point inferiorly. This makes it possible to easily distinguish a cervical transverse process from a thoracic transverse process, which points cephalad [1–3]. All cervical vertebrae have a round transverse foramen within each transverse process. In the upper six, the vertebral artery and vein and a sympathetic nerve plexus are contained within the transverse foramen. The vertebral artery does not pass through the transverse foramen of C7. The transverse processes have posterior and anterior roots. The posterior root arises from the junction of the lamina and pedicle. Its tip is bulbous and is referred to as the posterior tubercle. The anterior root of the transverse process also ends in a tubercle, called the anterior tubercle. This anterior root is also referred to as the costal process [1–5,7].

Atypical cervical vertebrae The atlas (C1), the axis (C2), and the seventh cervical vertebra are considered atypical. The atlas differs from all other cervical vertebrae because it lacks both a body and a spinous process. There are essentially five parts of the ring-shaped

16

atlas. The anterior arch constitutes the anterior one-fifth; the posterior arch makes up two-fifths; and the lateral masses are the remaining two-fifths (Figs. 2.11 to 2.15). The anterior and posterior arches contain central tubercles on their outer surfaces to which the anterior longitudinal ligament, the posterior longitudinal ligament, the nuchal ligament (ligamentum nuchae), and several muscles attach. The posterior surface of the anterior arch is slightly concave and contains in its midportion a smooth, rounded depression that articulates with the dens of C2. A small bursa between the bones makes this a true synovial joint and it is a frequent target in rheumatoid arthritis. The cranial surface of the posterior arches is grooved to accommodate the vertebral artery as it courses through the transverse foramen to enter the skull [1–3,7,9,10]. The lateral masses of C1 are relatively large. They represent the “body” of the atlas. The superior articular facets articulate with the occipital condyles. The inferior articular facets articulate with the superior articular surface of C2 and permit rotation of the head at the atlanto-axial joint. Anteromedially, a small tubercle projects medially from each lateral mass; this is the site for anchoring of the transverse ligament of the atlas. This ligament holds the dens in position against the anterior arch of the atlas [1–3,7–9]. The transverse processes of the atlas are the longest in the cervical region. They serve as anchoring points for muscles that assist in rotation of the head. The second cervical vertebra, or axis, is easily recognized by a toothlike projection (the dens [preferred term] or odontoid process) that extends from the upper end of the body (Figs. 2.16 to 2.20). There is a distinct narrowing of the dens to form a neck just above the junction of this structure within the body of the axis. The dens contains a rounded facet anteriorly for articulation with the posterior margin of the anterior arch of the atlas. Posteriorly, it is grooved to accommodate the transverse ligament of the atlas and a synovial sac [1–3,7,9]. The posterior margin of the dens is a continuation of the posterior vertebral body line of C2. Under normal circumstances, this line has no interruption. The pedicles of the axis are large and strong. Similarly, the laminae are thick. The superior and inferior articular processes

2 Anatomic considerations

A

B

Fig. 2.11 Atlas (C1), anterior view.

A

B

Fig. 2.12 Atlas (C1), posterior view.

A

B

Fig. 2.13 Atlas (C1), lateral view.

17

2 Anatomic considerations

A

B

Fig. 2.14 Atlas (C1), view from above.

A

B

Fig. 2.15 Atlas (C1), view from below.

A

Fig. 2.16 Axis (C2), anterior view.

18

B

2 Anatomic considerations

A

B

Fig. 2.17 Axis (C2), posterior view.

A

B

Fig. 2.18 Axis (C2), lateral view showing Harris’ ring. 1, superior articular facet; 2, posterior vertebral body line; 3, inferior margin of transverse foramen; 4, portion of anterior vertebral body.

A

B

Fig. 2.19 Axis (C2), view from above.

19

2 Anatomic considerations

A

B

Fig. 2.20 Axis (C2), view from below.

actually extend above and below these areas, and a distinct pars interarticularis may be discerned. The transverse processes are short; the spinous process is also thick. On a lateral radiograph, overlapping of the images of four distinct structures in the axis results in the appearance of a circular structure, which is referred to as Harris’ ring (Fig. 2.18B) [10]. The upper arc represents the upper margin of the superior articular facet; the posterior arc represents the posterior vertebral body line of the axis; the inferior arc represents the inferior margin of the transverse foramen; and the anterior arc represents the pedicle and anterior portion of the body of the axis. The seventh cervical vertebra is distinguished by its long, thin, nonbifid spinous process. This structure can be easily palpated, accounting for the other name of C7 – the vertebra prominens. This vertebra is further distinguished by large transverse processes that can extend as far laterally as the first thoracic transverse process [1–3,7]. Occasionally, a cervical rib develops from the anterior root of the transverse process (Fig. 2.21).

Os odontoideum is variously described as either a congenital failure of fusion of the dens to the body of C2 or the result of previous trauma. There is evidence to support both theories of etiology. In addition to the absence of fusion of the dens to the body of C2, the most salient radiographic feature of this abnormality is thickening and increased density of the anterior arch of the atlas (Fig. 2.22). This entity is described in more detail in Chapter 11. The anomaly encountered most commonly in the cervical region is failure of fusion of the posterior arch of the atlas (Fig. 2.23). It usually occurs as a single anomaly that is easily recognized by the absence of the spinolaminar line. As a rule, the anterior arch of the atlas is thickened and of increased density. Other anomalies include failure of fusion of the anterior arch of the atlas and partial or complete absence of a portion of the posterior ring of C1. Failure of fusion may occur at other levels as well, but less commonly than in the atlas (dysraphism, spina bifida).

Cervical anomalies Anomalies of segmentation and assimilation frequently occur in the cervical region. The most common of these is congenital failure of segmentation, generically and erroneously referred to as the Klippel–Feil anomaly. This disorder can be recognized as conjoint vertebrae with fusion of varying degrees anteriorly, posteriorly, or universally. Conjoint vertebrae are usually “taller” than normal. The overall height of the conjoint segment is that of two normal vertebrae and a normal intervertebral disc space. This is to be distinguished from surgical fusion, which generally consists of narrowing of the disc space and open facet joints posteriorly. Occipito-atlanto-axial fusions are complex anomalies that are usually readily recognized by the lack of normal segmentation. Imaging of these anomalies by conventional radiography is difficult. These patients are best evaluated by CT with multiplanar sagittal and coronal reconstruction.

20

Fig. 2.21 Cervical ribs (*). Note that the cervical transverse process (C) points caudad; the thoracic (T) process points cephalad.

2 Anatomic considerations

A

B

Fig. 2.22 Os odontoideum. (A) Lateral radiograph shows the os (arrow) just behind a hypertrophied anterior arch of the atlas (*). (B) Sagittal CT reconstructed image shows the os (O) to advantage.

Thoracic vertebrae The 12 thoracic vertebrae are recognized by the presence of costal facets superiorly and inferiorly on either side of the body and along the transverse processes. These facets articulate with the heads and articular tubercles of the ribs. The last two thoracic vertebrae lack facets on their transverse processes. As in the cervical region, there are typical and atypical vertebrae. The T2 to T8 vertebrae are considered typical; T1 and T9 through T12 are atypical [1–3].

Typical thoracic vertebrae The typical thoracic vertebra has a reniform shape with a small dorsal waist. Although the transverse and sagittal diameters are approximately equal, thoracic vertebrae are slightly taller posteriorly than anteriorly, resulting in the normal thoracic kyphosis. This anatomic difference becomes important when images are analyzed for minimal compression fractures, since the typical vertical height is up to 2 mm less anteriorly than the

A

vertical height posteriorly at the same level [1]. The posterior vertebral body lines of the upper thoracic vertebrae are solid. At the lower levels, the lines are interrupted centrally by a nutrient foramen, similar to the lumbar vertebrae. The lateral aspect of the vertebral bodies contains demifacets for articulation with the rib on either side of the disc space. The transverse process, which is long and club shaped, also contains a facet for articulation with the rib. Similarly, all the typical ribs contain demifacets superiorly and inferiorly and a third facet along their tubercles for articulation with the transverse processes. The articular facet of the neck of any typical rib always articulates with the transverse process of its own numbered vertebra [1–3]. It is important to recognize this relationship when evaluating patients who may have suffered rotary, shearing, or lateral flexion injuries, in which disruption of the costovertebral joints occurs. All thoracic transverse processes point cephalad, as opposed to cervical transverse processes, which point caudad. The spinous process of a typical thoracic vertebra is long and slender and slopes inferiorly, overlapping the spinous process of the vertebra below. There is variation in the degree of the slope of the spinous processes: T1 and T2 are almost always horizontal; T5 through T8 are nearly vertical; and T11 and T12 are horizontal. Figs. 2.24 to 2.28 show typical thoracic vertebrae. Use of CT allows us to see more of the vertebral anatomy than radiographs ever did. Furthermore, by adjusting the window and level settings on the PACS (picture archiving and communication system) monitor, it is possible to “enhance” the appearance of certain structures. All vertebrae have small vascular channels traversing the bodies. These are most prominent in the thoracic and lumbar regions. These channels typically radiate from the center of the vertebral body and have sclerotic margins (Fig. 2.29). This last feature serves to differentiate them from fractures, which have no sclerotic borders. If there is ever a question of whether these lucencies represent a fracture or not, MR imaging is useful in making the differentiation.

B

Fig. 2.23 Unfused posterior arch of the atlas. (A) Lateral radiograph shows absence of the spinolaminar line of the atlas (?). Note the line in the axis (arrow). The anterior arch of the atlas is hypertrophied. (B) The CT image shows hypoplasia of the posterior arch (arrow). Note a cleft in the anterior arch.

21

2 Anatomic considerations

A

B

Fig. 2.24 Typical thoracic vertebra (T7), anterior view.

Fig. 2.25 Typical thoracic vertebra (T7), posterior view.

A

22

B

Fig. 2.26 Typical thoracic vertebra (T7), lateral view.

2 Anatomic considerations

A

B

Fig. 2.27 Typical thoracic vertebra (T7), view from above.

Fig. 2.28 Typical thoracic vertebra (T7), view from below.

Fig. 2.29 Vascular channels (arrows). Note the sclerotic borders.

Atypical thoracic vertebrae

They are easily distinguished from their mates elsewhere in the column by their size and lack of costal facets. The spinous processes are large and rectangular; the transverse processes are thin. The typical lumbar vertebral body is large and reniform with a shallow, dorsal concavity abutting a triangular vertebral foramen. Like the thoracic vertebrae, the first two lumbar vertebral bodies are taller posteriorly than anteriorly. The reverse is true for L4 and L5; this results in a lumbar lordosis [1–5]. On a lateral radiograph or sagittal reconstructed CT image, the central portion of the posterior vertebral body line is interrupted by a nutrient foramen (Fig. 2.32B) [8]. As with thoracic vertebrae, on axial CT images vascular channels can be seen traversing the body, often in a stellate pattern. Again, these may be recognized by their sclerotic margins for differentiation from fractures (Fig. 2.29). The lumbar pedicles are short and arise from the upper lateral margin of the vertebral bodies. The inferior vertebral notches are deeper than the superior ones. Similarly, the laminae of the lumbar vertebrae are large and thick. The superior

The first thoracic vertebra and the ninth through twelfth thoracic vertebrae are considered atypical [1]. The first thoracic vertebra resembles a cervical vertebra. It is the only thoracic vertebra that contains an uncinate process [1–3]. The entire head of the first rib articulates in a full facet along the lateral superior aspect of this vertebra. Vertebrae T9 and T10 vary in the arrangement of their costal facets: T9 has a demifacet superiorly and no facet inferiorly, whereas T10 has a full facet superiorly and no facet inferiorly. Vertebra T11 and T12 resemble lumbar vertebrae; their short transverse processes lack facets for rib articulations, and their bodies are quite large [1–5]. Indeed, the absence of ribs from T12, a common anomaly, can result in mistaken identification of this vertebra as L1.

Lumbar vertebrae The lumbar vertebrae are the largest and heaviest segments of the presacral part of the vertebral column (Figs. 2.30 to 2.34).

23

2 Anatomic considerations

A

B

Fig. 2.30 Typical lumbar vertebra (L3), anterior view.

A

B

Fig. 2.31 Typical lumbar vertebra (L3), posterior view.

A

B

Fig. 2.32 Typical lumbar vertebra (L3), lateral view. Note the interruption of the posterior vertebral body line by a nutrient vessel B in (arrow).

24

2 Anatomic considerations

A

B

Fig. 2.33 Typical lumbar vertebra (L3), view from above.

A

B

Fig. 2.34 Typical lumbar vertebra (L3), view from below.

and inferior articular processes project above and below the laminae, respectively, just behind the pedicles. The portion of the lamina between these two processes is known as the pars interarticularis [1–3]. It is through this site that spondylolysis occurs (Fig. 2.35). A small, knobby protuberance, the mammillary process, extends from the posterolateral tip of each superior articular facet. This structure is the site for the attachment of posterior vertebral muscles. The transverse processes are thin, flattened, and elongated. At the base of the transverse process is found a small, rough tubercle known as the accessory process, which is the site of muscle attachment. If the accessory process is more than 5 mm long, it is termed the styloid process (Fig. 2.36) [1–5]. The transverse processes of L1–L3 point laterally; those of L4 and L5 point slightly cephalad. The transverse processes of L3 are typically the longest.

A

B

C

D

Thoracic and lumbar anomalies A number of anomalies that occur in the thoracolumbar and lumbosacral areas can cause confusion about the numbering of lumbar vertebrae. These common anomalies include absence of the twelfth rib, presence of a first lumbar rib, sacralization of L5 (Fig. 2.37), and lumbarization of the first sacral segment (Fig. 2.38) [1,5,7]. These anomalies present diagnostic difficulties when it is absolutely necessary to be able to identify a particular vertebral level for the site of injury, site of myelographic abnormality, or site for surgical intervention.

Fig. 2.35 Bilateral pars interarticularis defects of L4 with spondylolisthesis. (A) Lateral radiograph shows defects in the pars (arrow). (B) Axial CT image shows the bilateral defects (arrows). (C,D) Sagittal reconstructed CT images show the defects in the pars (arrows).

25

2 Anatomic considerations

Fig. 2.37 Sacralization of L5. A horizontal line drawn across the iliac crests passes through or close to the L4–L5 junction.

Fig. 2.36 Lumbar styloid processes of L1 (arrows). Fig. 2.38 Lumbarization of S1.

When such an anomaly is encountered, three methods can be used to determine the correct lumbar levels. The first method requires the availability of a chest radiograph, a thoracic vertebral radiograph, or a full thoracic and lumbar CT. If such a study is available, it is a simple task to count the thoracic vertebrae. If no other studies are available, the second method should be used. This method is based on the fact that a line drawn across the iliac crests passes through or near the L4 intervertebral disc space. The third method relies on the fact that the transverse processes of L3 are the most horizontal and are usually the longest [1]. Furthermore, as mentioned above, the transverse processes of L4 and L5 often angle cephalad. Occasionally, it is impossible to identify a lumbar level with confidence by any method. In these unusual circumstances, it is best to identify the level of abnormality by counting from the last rib-bearing vertebra. For example, if a radiologist tells a surgeon

26

that a burst fracture involves the second non-rib-bearing vertebra from above, the clinician has a definite point of reference. Uncommon occurrences include congenital absence of a pedicle, lamina, articular process, or the entire posterior element complex on one side. Such an anomaly can be differentiated from a destructive process by the presence of sclerosis of the pedicle on the opposite side (Fig. 2.39). The sclerosis occurs as a response to increased stress placed on the normal pedicle through weight-bearing. Use of CT scans or MRI may be necessary to solve the dilemma. As mentioned, the posterior portion of the vertebral body is delineated on a lateral radiograph by a single uninterrupted sclerotic line in the cervical and upper thoracic regions. In the lower thoracic and lumbar regions, this line is interrupted centrally by a nutrient vessel. At C2, the posterior vertebral body line continues uninterrupted along the back of the dens (Fig. 2.18B). Any disruption, displacement, angulation, rotation, duplication, or absence of this line is abnormal [8]. In the trauma setting, burst, shearing, and rotary fractures are the most likely causes of these abnormalities. However, neoplasms and infections can also destroy the posterior vertebral body line.

Sacrum and coccyx The sacrum comprises five sacral vertebrae fused in adults to form a wedge-shaped bone (Figs. 2.40 to 2.43). The sacrum articulates with the iliac bones laterally, and its base articulates with the last lumbar vertebra. The coccyx attaches inferiorly. The pelvic surface of the sacrum is concave. Along the pelvic surface, there are four transverse ridges, which form the pelvic sacral foramina. The superior aspects of these foramina are easily recognizable on frontal radiographs as thin, archlike densities, referred to as the sacral arcuate lines (Fig. 2.40). They are important in diagnosing occult sacral fractures (Fig. 2.44), which commonly occur in conjunction with pelvic fractures.

2 Anatomic considerations

A

B

C

Fig. 2.39 Congenital absence of posterior elements. (A) Frontal radiograph shows absence of the left pedicle of L4 and hypertrophy of the right pedicle (arrow). (B) Frontal radiograph shows the lamina of L5 on the left to be missing (*). Note the increased sclerosis of the right pedicle (arrow). (C) CT image shows the defect in the lamina (arrow) and the compensatory hypertrophy on the right (*).

A

B

Fig. 2.40 Sacrum, anterior view.

A

B

Fig. 2.41 Sacrum, posterior view.

27

2 Anatomic considerations

Fig. 2.42 Sacrum, lateral view. A

B

A

B

Fig. 2.43 Sacrum, view from above.

The coccyx is formed by four rudimentary vertebrae. Injuries to the coccyx generally present no diagnostic difficulties from an imaging standpoint.

Joints and ligaments

Fig. 2.44 Disrupted sacral arcuate lines on the right (arrows) in a patient with bilateral pubic bone fractures.

28

The vertebral column is articulated through a series of joints and supporting ligaments (Fig. 2.45). Two series of joints unite the individual vertebrae; the only exceptions are the joints between the occiput and the atlas and between the atlas and the axis, owing to their special anatomy [9]. There are essentially two types of joint – slightly movable (amphiarthrodial) symphyseal joints and freely movable (diarthrodial) synovial joints. The intervertebral discs are typical amphiarthrodial joints. The apophyseal, or facet, joints are diarthrodial joints that are enclosed in a fibrous capsule lined by a synovial membrane. Motion in these joints is of a gliding nature since the surfaces of these

2 Anatomic considerations

Fig. 2.45 Articulated vertebral column. (A) Anterior view. (B) Posterior view. (C) Lateral view.

A

B

joints are relatively flat [11]. Hence, movement is permitted by laxity in the articular capsule and is limited by the ligaments and osseous structures surrounding the joint. Motion about the disc spaces is markedly limited and depends mainly on disc thickness. The greatest degree of motion is present in the cervical and lumbar regions, where the discs are the thickest. Anatomically, intervertebral discs are composed of a laminated outer portion, the annulus fibrosus, and an inner portion, the nucleus pulposus (Fig. 2.46) [1–5,7]. Both of these structures derive embryologically from notochordal remnants. The nucleus pulposus is eccentrically located when viewed in the sagittal plane. The shorter distance to the vertebral canal accounts for the fact that herniation of disc material occurs more commonly posteriorly (into the canal) than anteriorly. Central herniations produce Schmorl nodes. Anterior herniations often displace a small fragment of bone from the anterosuperior or anteroinferior margin of the adjacent vertebrae, which is referred to as a vertebral edge separation or limbus fragment. Rarely, the same process can produce a posterior limbus. In the cervical region, synovial joints develop between the uncinate processes and the intervertebral discs. These are referred to as Luschka or uncovertebral joints. There is some

C

Fig. 2.46 Schematic depiction of abnormalities of the intervertebral disc space. The intervertebral disc is composed of the annulus fibrosus and the nucleus pulposus (A). Discovertebral disruption may result in posterior herniation of nuclear material (B), anterior herniation (C), intra-osseous herniation to produce a Schmorl node (D), or intra-osseous herniation with a corner fracture of the vertebral body (E) to produce a vertebral edge separation (limbus deformity).

29

2 Anatomic considerations

controversy about whether they are true joints. Most authorities now believe that the Luschka joints are actually fissures without true synovial linings [1,5,7]. Because the uncinate processes are not present at birth, these joints develop as the person grows. As with true synovial joints, however, osteophytes develop in response to stress in these regions, and they encroach on the nerves leaving the intervertebral foramina. The vertebral bodies are linked by two strong ligamentous bands (Fig. 2.47). The anterior longitudinal ligament is located over the anterolateral surfaces of the vertebral bodies. It is thin-

30

A

B

D

E

nest at its attachment to the base of the skull on the occiput. It is also thicker in the thoracic region than in the cervical and lumbar regions and is thickest over the central concavity of each vertebral body, where it actually blends with the periosteum (Fig. 2.47B,C) [1,3]. The posterior longitudinal ligament is on the posterior surface of the vertebral bodies within the vertebral canal (Fig. 2.47D,E). In the cervical region, it attaches to the body of the axis and becomes continuous with the tectorial membrane. The posterior longitudinal ligament is firmly bound to

C

Fig. 2.47 Vertebral ligaments. (A) Schematic overview in sagittal section. (B,C) Anterior longitudinal ligament. (D,E) Posterior longitudinal ligament (sectioned to show vertebral veins). ALL, anterior longitudinal ligament; LF, ligamentum flavum; IS, interspinous ligament; P, pedicle (sectioned); PLL, posterior longitudinal ligament; SS, supraspinous ligament; V, vertebral veins.

2 Anatomic considerations

Fig. 2.48 Nuchal ligament ossification. Lateral radiograph shows extensive degenerative change as well as nuchal ligament ossification (arrow).

Fig. 2.49 Degenerative changes between lumbar spinous processes (Baastrup disease) (arrows).

the intervertebral discs, where it spreads laterally. This is in contradistinction to the anterior longitudinal ligament, which is not intimately bound to the same extent. It is separated from the vertebral bodies by the venous plexuses, however [1,3]. Posterior to the vertebral bodies are the apophyseal or facet joints, which are the important articulations [1,3,7]. These are true synovial joints that are surrounded by a thin fibrous capsule attached to the outer surfaces of the articular processes. Unlike the fibrocartilage of the intervertebral disc space, the articular processes are covered by thin hyaline cartilage. Posterior support to the vertebral column is given by the ligamenta flava and by the supraspinous ligament. The ligamenta flava are actually paired ligaments connecting the laminae. They arise from the anterior surface of the lower lamina and attach to the upper portion of the posterior surface of the next succeeding lamina. They are separated in the midline by venous structures. The supraspinous ligament is composed of thin layers of fibrous tissue coursing over the tips of the spinous processes. There is variation in the attachments of these ligamentous fibers. Shorter fibers connect adjacent spinous processes, and longer ones connect several vertebrae. In the lumbar region, deep fibers merge with those of the interspinous ligament laterally. In the cervical region, the supraspinous ligament becomes part of the nuchal ligament [9]. Ossification, a normal variant, may occur within the nuchal ligament (Fig. 2.48). The interspinous ligament is a thin structure extending between adjacent spinous processes. Degeneration can occur with aging, and osteoarthrosis (Baastrup disease) may occur, particularly in the lumbar region (Fig. 2.49). The inherent stability of the vertebral column depends on the integrity of these ligamentous structures and adjacent bones. Denis [12] proposed a three-column approach to determining vertebral stability. The anterior column lies between the anterior longitudinal ligament and a vertical line through the junction of the middle and posterior third of the intervertebral

Fig. 2.50 The threecolumn delineation used in determining vertebral stability. Disruption of any single column will not result in instability; disruption of two contiguous columns, however, will. A, anterior; M, middle; P, posterior columns.

disc and body. The middle column extends from this line to the posterior longitudinal ligament. The posterior column extends through the posterior arch of the vertebra to the supraspinous ligament (Fig. 2.50). According to Denis, disruption of any single column will not result in instability. Disruption of two contiguous columns, however, produces instability. This is discussed in greater depth in Chapter 10. The atlanto-axial articulation is complex and consists of three joints – a middle atlanto-axial joint and two paired lateral joints (Figs. 2.51 to 2.53). The middle atlanto-axial joint is a pivotal type of joint with two small synovial sacs on either side of the dens; the posterior synovial sac is the larger of the two. Except for their size, the lateral atlanto-axial joints are similar to the facet joints found elsewhere in the vertebral column [9]. The tectorial membrane is a broad band of fiber that extends from the posterior longitudinal ligament along the lower aspect of the body of C2 and stretches cranially to attach to the inner aspect of the base of the occiput (Fig. 2.54). It covers the dens and the other ligaments. Just anterior to the tectorial membrane is the cruciform ligament. It has two components: a transverse portion (the transverse ligament) and a vertical

31

2 Anatomic considerations

Fig. 2.51 Atlanto-axial articulation, frontal view. Note the relation of the lateral masses of C1 and their articulations to the body of C2. Normally, there should never be more than 2 mm of unilateral or bilateral atlanto-axial overlap at their lateral margins. Similarly, there should never be more than 2 mm difference in the distance between the lateral margins of the dens (D) and the medial margins of the lateral masses (Lm).

Fig. 2.53 Atlanto-axial joint, from above. AA, anterior arch of atlas; PA, posterior arch of atlas; SA, superior articular facet of atlas; SP, synovial pads; TL, transverse ligament of dens. Compare with Fig. 2.54B.

portion. The transverse ligament attaches to the small tubercles on the medial side of the lateral masses of C1. This ligament holds the dens against the anterior arch of the atlas. A synovial sac is located between the transverse ligament and the dens [1,9]. The small atlanto-axial ligaments are collectively referred to as “check” ligaments, since they check, or stop, excessive motion between C1 and C2. There is a normal relationship between the posterior arch of C1 and the spinous process of C2. Lovelock and Schuster [13] have shown that the ratio between the height of the spinolaminar line of C1 and the flexion interval distance of the interspinous space between C1 and C2 is 2.0 or less. The ratio is extremely useful in determining the presence of a flexion injury at this level. These authors also found that the maximal interspinous distance between the atlas and axis should not exceed 18 mm in flexion (18 mm is the diameter of a dime,

32

Fig. 2.52 Atlanto-axial articulation, lateral view. The predental space (*) should not exceed 3 mm in an adult or 5 mm in a child. The spinolaminar line (straight line) should be smooth and uninterrupted. The posterior vertebral body line of the axis continues uninterrupted into the dens.

which can be conveniently used for measurement on a radiograph if a ruler is not handy). These measurements can readily be performed on CT scans The occipito-atlantal joint is also complex and is formed by the convex occipital condyles and the concave superior articular surfaces of the atlas (Fig. 2.54). These joints are enclosed by a synovial-lined articular capsule. The fibrous anterior occipito-atlantal membrane and posterior occipitoatlantal membrane are quite broad; the anterior membrane is the denser of the two. The posterior membrane is analogous to the ligamentum flavum in its relation to the vertebral canal. It adheres to the posterior margin of the foramen magnum and also to the posterior arch of the atlas [1,9]. The radiographic anatomy of the occipito-atlantal area has received considerable attention in the radiologic literature [14,15]. The normal relationships of this area were often difficult to assess on radiographs until recently. Occipito-atlantal dislocations and subluxations were once considered uniformly fatal injuries. Major trauma centers, however, are seeing more patients who survive this severe injury, often with little or no neurologic deficits. Several methods have been proposed to describe the normal relationships of the occiput to the atlas [14–17]. The Powers ratio [16] and Lee method [17] depend on accurately locating the opisthion, the posterior margin of the foramen magnum. Although this structure is often not clearly demonstrable on lateral cervical radiographs, it can be clearly identified on sagittal reconstructed CT images. The Harris method [14] is the easiest to perform, since it requires locating only the basion, the anterior margin of the foramen magnum and the posterior vertebral body line of the dens. A line drawn along the posterior vertebral body margin of the dens and extended cephalad should be no less than 6 mm and no more than 12 mm posterior to the basion (Fig. 2.55). This method is accurate for both anterior and posterior occipito-atlantal dislocations, whereas the Powers and Lee methods are inac-

2 Anatomic considerations

A

B

C

D

Fig. 2.54 Craniovertebral junction. (A). Anterior view. (B) Sagittal section. (C) Posterior view, showing superficial structures. (D) Posterior view, showing deep structures. A-A, lateral atlanto-axial joint; AAL, accessory atlanto-axial ligament; ALD, apical ligament of dens; AL, alar ligament; ALL, anterior longitudinal ligament; AO-A, anterior occipito-atlantal membrane; At, lateral mass of atlas; Ax, body of axis; C, cruciform ligament; CL, cruciform ligament over dens; D, dens; LF, ligamentum flavum; O, occipital bone; O-A or OA, occipito-atlantal joint; PLL, posterior longitudinal ligament; TL, transverse ligament of atlas; TM, tectorial membrane (superficial and deep layers); VA, vertebral artery.

curate in posterior dislocations. These methods are discussed in Chapter 8. Today, the Powers and Lee methods are considered obsolete and are of interest only from a historic perspective. Sutherland and associates [18] performed an anatomic and biomechanical study to determine whether the interspace between the dens and the lateral masses of the atlas may normally be asymmetric in the presence of intact ligaments. They dissected 10 human cadaveric atlanto-axial specimens in which the ligaments were intact. They found measurable asymmetry between the dens and lateral masses of C1 in the neutral

position where the ligament complex was intact. This finding results from minor degrees of head rotation. The investigators also found that the interspace is increased on the side toward which the head is rotated. They, therefore, concluded that asymmetry of this space is not an indicator of cervical instability, particularly in a person with no symptoms [18]. Finally, the sacroiliac joints are composed of true synovial joints anteriorly and fibrous joints posteriorly. Accessory ligaments bolster the strength of the sacroiliac joint and provide sacroiliac stability (Figs. 2.56 and 2.57).

33

2 Anatomic considerations

A B

Fig. 2.55 Normal craniovertebral relationships (Harris method) in a diagram (A) and an anatomic specimen (B). A line drawn along the posterior vertebral body margin and dens should be no less than 6 mm and no more than 12 mm posterior to the basion (*).

34

Fig. 2.56 Sacroiliac joint, anterior view. ALL, anterior longitudinal ligament; ASL, anterior sacroiliac ligament; GS, greater sciatic foramen; IL, iliolumbar ligament; LS, lumbosacral ligament; LSF, lesser sciatic foramen; PL, pectineal ligament; SS, sacrospinous ligament; ST, sacrotuberous ligament.

Fig. 2.57 Sacroiliac joint, posterior view. DSC, dorsal sacrococcygeal ligament; DSI, long and short dorsal sacroiliac ligaments; GSF, greater sciatic foramen; LSF, lesser sciatic foramen; SS, sacrospinous ligament; ST, sacrotuberous ligament.

2 Anatomic considerations

References 1.

2.

3.

4.

5.

6.

7.

8.

Gehweiler JA Jr., Osborne RL, Jr., Becker RF. The Radiology of Vertebral Trauma. Philadelphia, PA: WB Saunders, 1980, pp. 3–88. Agur AMR, Dalley AF. Grant’s Atlas of Anatomy, 11th edn. Baltimore, MD: Lippincott, Williams & Wilkins, 2005. Standring S. Gray’s Anatomy: The Anatomical Basis of Clinical Practice, 39th edn. Philadelphia, PA: ChurchillLivingstone, 2005, pp. 727–798. Bailey DK. The normal cervical spine in infants and children. Radiology 1952;59:712–719. Swischuk LE. Imaging of the Cervical Spine in Children, 2nd edn. New York: Springer, 2004, pp. 13–38. Keats TE, Anderson MW. Atlas of Normal Roentgen Variants that May Simulate Disease, 8th edn. Philadelphia, PA: Mosby, 2007, pp. 155–372. Schmorl G, Junghanns H. The Human Spine in Health and Disease, 5th edn. New York: Grune & Stratton, 1971. Daffner RH, Deeb ZL, Rothfus WE. The posterior vertebral body line:

9.

10.

11.

12.

13.

14.

importance in the detection of burst fractures. AJR Am J Roentgenol 1987; 148:93–96. von Torklus D, Gehle W. The Upper Cervical Spine: Regional Anatomy, Pathology, and Traumatology – A Systematic Radiological Atlas and Textbook. New York: Grune & Stratton, 1972. Harris JH Jr., Burke JT, Ray RD, et al. Low (type III) odontoid fracture: a new radiographic sign. Radiology 1984; 153:353–356. Penning L. Normal movements of the cervical spine. AJR Am J Roentgenol 1978;130:317–326. Denis F. The three-column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 1983;8:817–831. Lovelock JE, Schuster JA. The normal posterior atlantoaxial relationship. Skeletal Radiol 1991;20:121–123. Harris JH Jr., Carson GC, Wagner LK. Radiologic diagnosis of traumatic occipitovertebral dissociation. 1. Normal occipitovertebral

15.

16.

17.

18.

relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 1994;162:881–886. Harris JH Jr., Carson GC, Wagner LK, et al. Radiologic diagnosis of traumatic occipitovertebral dissociation. 2. Comparison of three methods of detecting occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 1994;162:887–892. Powers B, Miller MD, Kramer RS, et al. Traumatic anterior atlanto-occipital dislocation. Neurosurgery 1979; 4:12–17. Lee C, Woodring JH, Goldstein SJ, et al. Evaluation of traumatic atlantooccipital dislocations. AJNR Am J Neuroradiol 1987;8:19–26. Sutherland JP Jr., Yaszemski MJ, White AA, III. Radiographic appearance of the odontoid lateral mass interspace in the occipitoatlantoaxial complex. Spine 1995;20:2221–2225.

35

Chapter

3

Biochemical considerations Richard H. Daffner

The previous chapter dealt with the pertinent anatomy of the vertebral column. This chapter discusses the salient biomechanical principles of vertebral motion. An understanding of basic vertebral biomechanics is necessary in order to fully comprehend the principles needed to diagnose vertebral injuries by imaging. A detailed discussion of vertebral biomechanics is beyond the scope of this book. Much of the material contained in this chapter was gleaned from the excellent text by White and Panjabi, Clinical Biomechanics of the Spine [1], to which the reader is referred for a more in-depth discussion.

Definitions This discussion uses a number of biomechanical terms, some of which may be unfamiliar to the reader. For this reason, I include the following glossary: Dynamics: the branch of mechanics that studies the loads and motions of interacting bodies. Kinematics: the branch of mechanics that studies the motion of bodies without taking into account the forces that produce that motion. It is the study of motion without regard to forces. Kinetics: the branch of mechanics that studies the relationships between forces acting on a body and the changes that those forces produce in body motion. More simply, it is the study of forces as well as motion. Translation: the movement of a body in the same direction relative to a fixed point. Rotation: any spinning motion or angular displacement of a body about an axis. Shear: the application of any force parallel to the surface on which it acts. Coupling: the phenomenon in which any motion involving translation or rotation of a body about an axis consistently produces simultaneous translation or rotation about another axis. Degrees of freedom: the motion of a rigid body in translation back and forth about a straight axis or rotation back and forth about any axis. Vertebrae have six degrees of freedom – translation along and rotation about either direction of three orthogonal axes (X, Y, and Z). The sagittal plane is the Y–Z plane; the coronal is the X–Y plane; and the horizontal plane is the X–Z plane (Fig. 3.1) [2].

36

Compression: the force that tends to push components of a body together. The unit of measurement is the newton (N). Distraction – the force that tends to pull components of a body apart. (Again, measured in newtons.) Bending: the deformity that occurs in a structure when load is applied to an area of that structure that is not directly supported. Pattern of motion: the configuration of the path made by the geometric center of a body moving through its range of motion. Functional spinal unit: a term used for two adjacent vertebrae and their associated soft tissues. This is often referred to as a motion segment. Neutral zone: the distance between the neutral position and the onset of intrinsic resistance to physiologic motion. This is expressed in degrees in the vertebral column. Elastic zone: the distance from the end of the neutral zone to the end of the physiologic range of motion. Once the elastic zone is exceeded, structural damage occurs. Plastic zone: the distance from the end of the elastic zone to the point of structural failure. Microtrauma begins in this zone, eventually leading to failure. Major injuring vector: the direction of the principal force that resulted in vertebral (or other skeletal) injury. For purposes of this discussion and subsequent chapters, four such vectors are recognized: flexion, extension, shearing, and rotation.

Structural considerations Chapter 2 described the various skeletal and soft tissue elements of the vertebral column. Let us now reexamine these structures, not as individuals but as components of the functional spinal unit (FSU). The vertebral bodies transmit the bulk of the weight imposed on the vertebral column. As a rule, the compressive strength of the vertebrae increases from C1 to L5. Unfortunately, osteoporosis is a naturally occurring aspect of aging, and as a result bone strength decreases significantly in women, but also in men, after around 40 years of age. The rate of decrease becomes more gradual after age 60. Studies on gender and age differences using densitometry of vertebral bodies suggest that the higher incidence of osteoporotic compression fractures in elderly women compared with men does

3 Biochemical considerations

A

B

Fig. 3.1 (A) Central coordinate system illustrating the three orthogonal axes (X, Y, and Z). The sagittal plane is the Y–Z plane; the coronal plane is the X–Y plane; and the horizontal plane is the X–Z plane. (B) The six vertebral degrees of freedom – translation along and rotation about either direction of the three orthogonal axes. (A modified from Panjabi and White [2].)

37

3 Biochemical considerations

not result from a greater degree of osteoporosis in women but rather from the overall increased size of the male vertebrae [3]. The intrinsic trabecular compressive strength in a vertebral body is greatest in the center and weakest on the outside of the posterior region [1,4]. This is somewhat surprising in view of the fracture pattern, which affects primarily the anterior cortical surface. However, the posterior elements also contribute to the overall structural strength, as described below. The vertebral endplate is the junction between the vertebra and the intervertebral disc. It is generally the first structure to fail with compressive loads. The pattern of failure depends on whether degeneration has occurred within the intervertebral disc. Compression of a nondegenerated disc increases pressure on the nucleus pulposus. This, in turn, increases the compressive load at the middle of the vertebral endplate and places some tension on the periphery. The net result is that high bending stresses occur in the center, which may produce Schmorl nodes (Fig. 3.2A) [1,4]. If the disc is degenerated, the compressive load is distributed through the annulus, with the result that endplate loading is more at the periphery. When failure occurs, it is generally in the vertebral body itself (Fig. 3.2B). The facet joints are important stabilizing structures within the vertebral column. They change in their orientation from the cervical to the lumbar region (Fig.  3.3) [1,5]. Depending on the body’s posture, facet joints carry up to 30% of the compressive load [1,4,6]. They also contribute up to 45% of the torsional strength of an FSU. Torsional stiffness is determined by the design and orientation of the facet joints. It increases from the T7–T8 FSU to the L3–L4 FSU. The highest torsional stiffness is found at the thoracolumbar junction (T12–L1 FSU) [1,5]. It is not surprising, therefore, that there is a high incidence of injury at this level. Figure 3.4 illustrates the factors that contribute to torsional stiffness.

A

Ligaments Seven sets of ligaments tie the individual vertebrae to one another (Fig. 3.5). Each of the ligaments is permitted a small degree of motion. In flexion, all the ligaments are stretched except the anterior longitudinal ligament. In extension, all are stretched except the posterior longitudinal ligament. Lateral bending stretches the ligamenta flava and transverse ligaments. In axial rotation, one of the capsular ligaments on the opposite side of rotation is stretched, as is the supraspinous ligament. Because ligaments transfer the tensile loads from bone to bone, failure may occur either within the ligament or at its attachment point. Failure of the ligament–bone complex depends on the rate of loading. If the rate is slow, failure usually occurs through the bone. If it is rapid or high, the ligament itself fails [1,7]. No studies of vertebral ligaments have established the exact failure point. Studies of the cruciate ligaments of the knee, however, suggest that the site depends on both the rate of application of the loads and the status of the bone [8].

Muscles Three types of muscle provide motion to the vertebral column: flexors, extensors, and rotators [1,5,7,9]. The flexors are all of the anterior muscles, including the abdominal musculature. The extensors are the posterior muscles. The rotators are obliquely oriented anterior or posterior muscles that contract independently of their mates on the opposite side of the body. They are responsible for lateral flexion. Muscle activity in concert with ligamentous ties and torsional stiffness produces coupling. A good example of coupling can be seen with cervical rotation and lateral bending. When the head is tilted to the

B

Fig. 3.2 Deformity of the vertebral endplate as the result of axial loading. (A) Compression of the nondegenerated disc increases pressure on the nucleus pulposus (P). The compressive load is greater at the center (middle arrows). This results in Schmorl node formation, with a central deformity of the vertebral body. (B) In the degenerated disc, the compressive load is distributed through the annulus with greater peripheral compression (arrows). The net result is deformity of the vertebral body with anterior compression. (Modified from White and Panjabi [1].)

38

3 Biochemical considerations

A

B

Fig. 3.3 Typical facet orientation in the cervical, thoracic, and lumbar regions. The differences in the spatial alignment of the facet joints produce differences in kinematics in each region (dashed lines). (Modified from White and Panjabi [1].)

C

A

B

Fig. 3.4 Torsional stiffness as related to the orientation of the facet joints. (A) At T5–T6, facet orientation allows rotation of the vertebra. (B) At T12–L1, facet orientation does not permit any significant rotation. Consequently, severe rotational forces result in fracture, dislocation, or both. (Modified from White and Panjabi [1].)

39

3 Biochemical considerations

Fig. 3.5 The normal ligaments of the vertebral column. (Modified from White and Panjabi [1].)

right, the spinous processes move to the left, and vice versa. In the cervical region, the spinous processes move toward the convexity of the curve on lateral bending [1,7,9]. In the lumbar and lumbosacral regions, they move toward the concavity. The reason for these differences lies in the orientation of the muscle groups and the facet joints.

the average allowable degrees of motion. Ranges are given when one region entails more than one FSU. In the absence of disease, a gradual decrease in the range of motion, primarily within the lumbar region [1,10], occurs intrinsically up to age 35. After age 35, there is little loss of motion. Diseases that affect the intervertebral discs, vertebral bodies, or facet joints generally cause a further decrease in mobility, particularly with osteophyte or syndesmophyte formation. Some diseases, such as rheumatoid arthritis, can actually result in increased mobility, particularly in the atlanto-axial region, as the result of ligamentous laxity or disruption of the ligament–bone interface by synovial proliferation or pannus.

Allowable vertebral motion Experimental data have determined the range of motion for each FSU in combined flexion and extension, unilateral lateral bending, and unilateral axial rotation [1,7,9,10]. Table 3.1 lists

Table 3.1 Average allowable degrees of vertebral motion

Level

Flexion/extension X axis

Lateral bending Z axis

C0–1

25

5

5

C1–2

20

5

40

C2–7

10–20a

7–11b

3–7a

C7–T1

9

4

2

T2–9

4–6

5–6

6–8b

T9–10

6

6

4

T10–11

9

7

2

T11–L1

12

7–9

2

a

L1–5

12–16

6–8

2

L5–S1

17

3

1

a b

40

Axial rotation Y axis

Increases with lower level. Decreases with lower level.

3 Biochemical considerations

Biomechanical basis for vertebral injury

on age and presence and extent of disease. Furthermore, age plays an important role in the location of vertebral injuries. In the cervical region, we have found that there is a higher incidence of fractures in the C1–C2 region [11,12]. The reason for this is that as patients age and the spine degenerates, the atlanto-axial region becomes the most mobile.

Bones, ligaments, and muscles are basically specialized forms of architectural material. When they are placed under stress, they naturally deform. The degree of deformity follows a wellestablished curve known as Wolff ’s law. The typical curve for Wolff ’s law contains four zones: the neutral zone, the elastic zone, the plastic zone, and the failure zone (Fig. 3.6). Within the neutral zone, little effort is required to deform the ligaments. The deformity is not permanent, and there is no structural damage. Within the elastic zone, the deformation requires much more force. Here, a release of the force results in reversal of the deformity and a return to the resting configuration. The neutral zone and elastic zone form the physiologic range. An increase of force above that of the physiologic range is found in the traumatic range. Further increases in force cause microfractures in bones and microtears in ligaments. A cessation of force does not result in a return of the structure to its normal resting state, and a permanent deformity occurs. This is referred to as the plastic zone. A classic example of this is encountered in children with plastic bowing injuries of the extremities. Once sufficient microtrauma has occurred to the structural system, catastrophic failure (manifested as gross fracture, ligamentous rupture, or both) occurs as the failure zone is reached. Aging and disease produce changes in ligamentous elasticity and bony rigidity [1,10]. As a result, the length and slope of the Wolff ’s law curve differ from patient to patient depending

Any time a solid body moves along a plane, there is a point at every instant that does not move within the body or at some hypothetical extension of that body. Lines drawn perpendicular to reference points along the plane of motion pass through a point that is called the instantaneous axis (or center) of rotation (IAR) for the motion at that instant (Fig. 3.7). For the occiput (C0) on C1, the IAR is in the clivus (Fig. 3.8A, left and center) [13]. For C1 on C2, the IAR is centered in the dens (Fig. 3.8A, right). For the lower cervical vertebrae, the IAR is centered in the anterior vertebral body of the subjacent vertebra of an FSU (Fig. 3.8B). For the thoracic vertebrae, the IAR is in the body anteriorly for lateral bending and for flexion and extension, and in the posterior body for axial rotation (Fig. 3.8C). For the lumbar vertebrae, the IAR is in the anterior body in flexion, in the posterior body in extension, on the right and left sides of the body in lateral bending, and in the center of the body in axial rotation [1] (Fig. 3.8D). The concept of IAR may seem difficult. However, knowledge of IARs is essential for

Fig. 3.6 Relationship of deformation or strain in response to load or stress (Wolff ’s law). Within the neutral zone, little effort is required to deform the ligaments and bones. Any deformity is not permanent, and there is no structural damage. Within the elastic zone, more force is required to produce deformation. Release of the force results in reversal of the deformity and a return to the resting configuration. Within the plastic zone, further increases in force cause permanent deformity. A release of force does not result in return to the resting configuration. Further stress produces failure, which will be manifest as either rupture of the ligament or fracture as the failure zone is entered. The neutral and elastic zones constitute the physiologic range of stress response; the plastic and failure zones constitute the traumatic range.

Fig. 3.7 Instantaneous axis of rotation. Lines drawn perpendicular to reference points along the plane of motion pass through a fixed point in space called the instantaneous axis or center of rotation for the motion at that instant. The instantaneous axis does not move while the remainder of the vertebra does. (Modified from White and Panjabi [1].)

Instantaneous axis of rotation

41

3 Biochemical considerations

Fig. 3.8 Instantaneous axes of rotation in lateral bending, flexion, extension, and axial rotation at various levels. Circled areas indicate location of instantaneous axes of rotation. (A modified from White and Panjabi [1]; B–D modified from White and Panjabi [13].)

42

3 Biochemical considerations

understanding the mechanisms of injury, since they dictate the patterns of deformation that occur in the FSU. As an example, a vertical force applied anterior to the IAR produces flexion; if the force is applied posterior to the IAR extension results. There is a load spectrum for each mechanism of injury. Although it is convenient to classify injuries on the basis of one particular mechanism (flexion, extension, rotary, or shear), actually a combination of factors contributes to the major (or primary) injuring vector (Fig.  3.9). This vector typically may be determined from imaging studies when they are analyzed together with the available history of the injury and the findings on physical examination. This will be elaborated upon in Chapter  7. Some generalizations for interpretation of imaging studies of vertebral injuries can be made based on biomechanical principles [1,4]. Bone fails first along lines of tensile strength. This is generally through shear or compression. Triangular anterosuperior or anteroinferior fractures may occur as the result of flexion combined with shearing or extension (Fig.  3.10). The earlier presumption was that these fractures resulted from avulsion of the peripheral annulus fibrosus fibers. “Teardrop” fractures, either from flexion or extension, may have associated comminuted fractures of the vertebral body (Fig. 3.11). Compression of the FSU produces endplate fractures first. However, disc injury may occur in the absence of fracture. Wedging is caused by compression from eccentric forces [1,4,5]. Under normal circumstances with most loading vectors, bone fails before ligaments. Wide separation between anterior and posterior elements indicates ligamentous rupture, most likely from axial rotation about the Y axis. Narrowing of a disc space above a fractured vertebra suggests failure of the annulus at its attachments, usually as the result of flexion. Conversely, widening of the disc space also indicates annulus failure, but in an extension injury [1].

Fig. 3.10 Production of teardrop fractures. Triangular teardrop fractures are the result of either combined compression and shearing in flexion mechanisms or tension and avulsion in extension mechanisms. These fragments can be either anterosuperior or anteroinferior. (Modified from White and Panjabi [1].)

Fig. 3.9 Load spectrum for flexion and extension mechanisms of injury. The gray arrows within the circles indicate the major injuring vector. The size of the arrow indicates the degree of effect that vector has on the overall injury pattern. Primary vectors along the Y axis represent axial loading (compression); forward or backward vectors along the Z axis represent flexion or extension forces, respectively. The bottom illustration represents pure axial loading. As one progresses to the left of neutral, the typical changes of flexion injuries occur. At the far left, a pure flexion mechanism results in significant posterior distraction. As one progresses to the right of neutral, the typical changes of extension injuries occur. At the far right, a pure extension mechanism results in significant anterior distraction and a wide disc space. (Modified from White and Panjabi [1].)

Fig. 3.11 Common fractures associated with teardrop fractures. These include burst of the vertebral body (arrows) and fracture of the lamina. (Modified from White and Panjabi [1].)

43

3 Biochemical considerations

Summary The biomechanical implications of vertebral injuries must be understood for proper interpretation of imaging studies of patients affected. The bones, joints, intervertebral discs, ligaments, and muscles work in concert to produce normal

References 1.

2.

3.

4.

5.

44

White AA, Panjabi MM. Clinical Biomechanics of the Spine, 2nd edn. Philadelphia, PA: JB Lippincott, 1990. Panjabi MM, White AA, Brand RA. A note on defining body parts configurations. J Biomech 1974;7:385. Gilsanz V, Boechat MI, Gilsanz R, et al. Gender differences in vertebral sizes in adults: biomechanical implications. Radiology 1994;190:678–672. Roaf R. A study of the mechanics of spinal injuries. J Bone Joint Surg 1960; 42B:810–823. Tanguy A. Biomechanics of the normal thoracolumbar spine and their application to fractures. In Floman Y, Farcy J-PC, Argenson C, eds. Thoracolumbar Spine Fractures. New York: Raven Press, 1993, pp. 45–57.

6.

physiologic motion. These motion parameters have limitations, however, and when these limitations are exceeded, injury to the bone, soft tissues, or both results. Mechanisms of injury are far more complex than previously assumed and generally are the result of multiple factors acting in concert at one or more FSUs.

Markolf KL. Deformation of the thoracolumbar intervertebral joints in response to external loads: a biomechanical study using autopsy material. J Bone Joint Surg 1972;54A: 511–533. 7. Maiman DJ, Yoganandan N. Biomechanics of cervical spine trauma. Clin Neurosurg 1991;37:543–570. 8. Noyes FR, DeLucas JL, Torvik PJ. Biomechanics of anterior cruciate ligament failure: an analysis of strainrate sensitivity and mechanisms of failure in primates. J Bone Joint Surg 1974;56A:236–253. 9. Penning L. Normal movements of the cervical spine. AJR Am J Roentgenol 1978;130:317–326. 10. Tanz SS. Motion of the lumbar spine: a roentgenologic study. AJR Am J Roentgenol 1953;69:399–412.

11. Daffner RH, Goldberg AL, Evans TC, Hannon DP, Levy DB. Cervical vertebral injuries in the elderly: a 10 year study. Emerg Radiol 1998;5:38–42. 12. Ong AW, Rodriguez A, Kelly R, et al. Detection of cervical spine injuries in alert, asymptomatic geriatric blunt trauma patients; who benefits from radiologic imaging? Amer Surgeon 2006;72:773–777. 13. White AA, III, Panjabi MM. Spinal kinematics: the research status of spinal manipulative therapy. NINCDS Monogr 1975;15:93.3

Chapter

4

Imaging of vertebral trauma I: indications and controversies Richard H. Daffner

The referring physician and the radiologist have many imaging techniques available for the diagnosis of the extent of vertebral injury. These include radiography, CT, MR imaging, and myelography. These techniques are used alone or in combination to arrive at the correct diagnosis. However, the imaging of patients with suspected vertebral trauma has also been one of the most controversial topics across many specialty lines since the mid-1990s [1,2]. This controversy has engendered a number of questions. Which patients need imaging? Are there certain clinical and historical factors that will identify those trauma patients who are at high or low risk for vertebral injury? If imaging is indicated, which modality should be used? Is there a role for radiography? Should CT be the method of choice? To answer these questions, a number of factors (ease of performance, efficacy of making a diagnosis, time required for the study, cost, and radiation exposure) that influence the selection of the appropriate imaging study must be examined. This chapter will explore these issues, review the current concepts regarding selection of appropriate imaging studies, and will address the topic of “clearing” the spine in comatose patients. Detailed descriptions of the use of each of the established imaging formats will be described and illustrated in Chapters 5 and 6.

Indications We live in an era in which the cost of health care is in the public spotlight daily. To cut costs, the public and third-party payers (including the government) are demanding that the indications for performing diagnostic imaging be reevaluated. Not surprisingly, one of the “hot button topics” in trauma care relates to the assessment of possible vertebral injuries. Trauma patients are of great concern not only to the physicians who must treat them but also to the hospital administrators seeking ways to augment hospital income, as well as to third-party health care providers, who are seeking cost containment. The emergency physician is often caught between the “rock” of protocol-driven requirements to obtain (cervical) radiographs (coupled with the fear of malpractice litigation of missing an injury), and the “hard place” that results from the efforts of medical cost containment.

Certainly, as mentioned above, no subject has generated more controversy in trauma care than that of imaging for vertebral injuries. Most trauma centers follow a series of protocols that are aimed at efficiently identifying all the abnormalities in this group of critically injured patients. Sometimes, these protocols are followed with religious zeal, without any careful thought as to the individual patient, the mechanism, and the risk of particular injuries. These protocols have been developed by trauma surgeons and are, for the most part, effective. A significant amount of cervical imaging is performed solely because the patient arrives at the hospital wearing a cervical collar. Paramedical personnel apply these collars as standard procedure regardless of the history of the injury. I had an opportunity to witness this practice while at an ice skating rink. A man fell and sustained a laceration that would require suturing over the bridge of his nose. He walked without assistance off the ice, where rink personnel administered first aid and then called for an ambulance. On arrival, one of the first things the paramedical crew did was apply a cervical collar to the otherwise ambulatory and symptom-free patient. I subsequently learned that cervical radiographs were performed at the local hospital, despite his protestations that his neck was fine. There are conservative estimates in the literature that indicate that more than a million patients with blunt trauma, who have the potential for cervical spine injuries, are seen in emergency departments in the USA annually. With numbers such as these, it is important to have a reliable method of properly screening patients to insure that those who need imaging have it, and to exclude those who do not. In the late 1980’s there were calls for cost containment by the federal government and health insurers. This led to several studies investigating ways to decrease the number of imaging studies being performed, particularly on patients seen in the emergency setting. Among the earliest responses were two studies that produced the so-called Ottawa Rules for reducing the number of ankle and knee radiographs following injury [3,4]. Radiologists also began seeking ways to reduce the number of imaging examinations in trauma patients. One of the first studies was by Mirvis and colleagues in 1989 [5], which revealed that protocol-driven imaging was not only time consuming and expensive, but also resulted in the unnecessary expenditure of hundreds of thousands of dollars (in 1989). His group found

45

4 Imaging I: indications and controversies

that 34% of the patients imaged were mentally alert and without symptoms referable to the cervical region. They urged radiologists to work closely with trauma surgeons to develop more rational methods for determining risk of injury in an effort to improve efficacy and reduce costs [5], and thus began a series of investigations designed to assess risk factors. The first report on risk-based indications for cervical spine imaging was published by Vandemark in 1990 [6]. He proposed 10 criteria that would identify patients at high risk for having a cervical injury: high-velocity blunt trauma, generally from a motor vehicle crash; presence of multiple fractures, particularly from large bones; presence of pain, spasm, or deformity of the cervical spine; altered mental status from alcohol and/ or drugs or the injury itself (Glasgow Coma Score [GCS]   60  mph [100 km/h]), motor vehicle crash with rollover or ejection of the victims, crash of any form of motorized recreational vehicle (motorcycle, snowmobile, water craft), or any bicycle collision. The CCR go further in defining low-risk factors. These include the history of a “simple” or rear end motor vehicle crash. This is a crash in which the victim’s vehicle was not hit by a bus or large truck; was not struck by a high-speed vehicle, was not pushed into oncoming traffic, did not involve a rollover, and in which the impact was so slight that the air bags did not deploy. The remaining low-risk factors include history of the victim being ambulatory at any time, sitting in the emergency department, delayed onset of neck pain, and absence of midline cervical tenderness [9]. Stiell and his Canadian colleagues went on to state that imaging is indicated if high-risk factors are present; if they are absent, the clinician should assess for a range of motion. If low-risk factors are also absent they recommend no imaging; if present, they assess for motion (flexion, extension, 45° rotation to each side). If motion is normal, they perform no imaging; if abnormal they recommend imaging [9]. How effective is the CCR? Stiell and colleagues claimed that when their criteria were applied to the 8924 patients in their study, only four fractures would have been missed. Each of these injuries were considered to be of a “minor” nature, meaning they produced no neurologic deficits and were mechanically stable [9]. Furthermore, when they compared the CCR with the NEXUS low-risk criteria in alert and stable patients, they found that the CCR was superior to the low-risk criteria in both sensitivity and specificity for ruling out cervical vertebral injury. In addition, they felt that by using the CCR they could eliminate up to 25% of the imaging that was performed for suspected cervical injury [10]. Finally, the American College of Radiology (ACR) began developing Appropriateness Criteria in 1993 as a guide for clinicians and radiologists to provide “the right examination, for the right reasons, performed the right way.” Each panel producing the criteria consists of a group of radiologists, considered expert in their particular discipline. In addition, the panel contains several non-radiologists. The musculoskeletal panel includes an orthopedic surgeon and an emergency medicine physician. Each panel conducts a literature review on the subject of study, and after this review formulates a document that is based primarily on the evidence in the peer-review literature as well as on their own personal experience. The Expert Panel on Musculoskeletal Imaging in conjunction with their counterparts in neuroradiology formulated the latest document on appropriate imaging for patients with suspected vertebral trauma. The panel reviewed literature covering 55 000 patients, including the NEXUS and CCR studies. Their findings were published on the ACR web site (www.acr.org) as well as in the Journal of the American College

4 Imaging I: indications and controversies

of Radiology [2]. They concurred that adult patients who satisfy any of the low-risk criteria (as outlined above) need no imaging [2]. Their specific recommendations on the type of imaging that should be performed on patients who do not fall into the lowrisk category will be listed in the sections to follow. The indications for thoracic and lumbar imaging parallel those for the cervical spine. In our experience, as well as that of others who work at large trauma centers, approximately 25% of patients with one vertebral injury have another at a noncontiguous site (cervical–thoracic, cervical–lumbar, thoracic–lumbar, multilevel same segment). Calenoff and associates [11] were the first to describe common combinations of multiple noncontiguous vertebral injuries. Gehweiler and coworkers [12] reported on the incidence of contiguous cervical injuries. Other researchers [13,14] reported the incidence of multilevel injuries to be as high as 10%. Gupta and Masri [14], in a series of 935 patients, observed that 50% of those patients with multilevel injuries had neurologic lesions that were incomplete. They, therefore, recommended a complete examination of the vertebral column if one injury was found. Similar results by Powell and associates [14] led them to recommend individualizing the examination using the same guidelines as for isolated injury. There is a strong correlation between thoracic cage injury and thoracic spine injuries. Jones and coworkers [15] observed an association between sternal fractures and unstable thoracic vertebral fractures. The sternum combines with the ribs to provide a stabilizing force in the thoracic vertebral column. Fractures of these structures allow excessive motion within the thoracic column, which often results in an unstable injury. Woodring and colleagues [16] reviewed 100 patients with chest trauma and found that nine had associated vertebral injury. Finally, in addition to the same criteria indicating that cervical imaging is necessary, we must add the presence of rigid spine disease, defined as ankylosing spondylitis or DISH [17–19]. How does this impact our trauma practice? The type of trauma seen at any particular institution may determine protocols for evaluation of patients with suspected cervical injury. My institution, Allegheny General Hospital in Pittsburgh, Pennsylvania, is a Level I trauma center that admits about 1900 major trauma cases a year. Of this group, some 400 patients with vertebral injury are encountered annually: 85% present after motor vehicle crashes and 14% after falls. A large number of these patients are over age 65. Most of the patients involved in motor vehicle trauma have a compromised sensorium as the result of head injury, alcohol, or other drug ingestion. Consequently, we have a large, selected group of trauma patients in the high-risk category for whom cervical imaging is performed in virtually every case. We see few patients with minor cervical injuries. This is in contrast with many of the smaller suburban hospitals in the Pittsburgh area, which receive a preselected patient population from accidents that are considered minor fender benders. This selection process is the result of the training of the local paramedical personnel, who automatically refer victims of severe trauma to one of the three large Level I trauma centers in Pittsburgh. In hospitals in which

the usual degree of injury is less severe, emergency physicians may have the luxury of being able to adequately assess their patients clinically to determine whether imaging is needed. Our experience parallels that of other large Level I trauma centers that deal primarily in high-speed blunt trauma. Table 4.1 summarizes the high- and low-risk criteria for vertebral injury. Table 4.1 High- and low-risk criteria for spine injury

Criteria High-risk criteria [2,7,9]

Altered mental status (Glasgow Coma Scale  65 years “Dangerous mechanism” Fall of > 1 m Axial load to head High-speed motor vehicle crash Motor vehicle crash with large vehicle Motor vehicle crash with rollover, ejection Pedestrian struck by vehicle Crash from motorized recreational vehicle Paresthesias in extremities Rigid spinal disease Ankylosing spondylitis Diffuse idiopathic skeletal hyperostosis

Canadian C-Spine Rule: no imaging [9]

Absence of high-risk factors Low-risk factors that allow safe assessment of active range of motion (flexion/extension, 45° right and left rotation) “Simple” rear-end motor vehicle collision Sitting position in emergency department Ambulatory at any time Delayed onset of neck pain Absence of midline cervical tenderness Able to actively rotate neck 45º left and right and flex and extend

NEXUS criteria (low risk) [8]

No midline cervical tenderness No focal neurologic deficits No intoxication or indication of brain injury No painful distracting injuries Normal alertness

Indications for thoracic and lumbar CT

Known cervical injury Rigid spine disease (ankylosing spondylitis, diffuse idiopathic skeletal hyperostosis)

Radiography versus computed tomography Until recent years, radiography was the standard initial “screening” examination performed to evaluate patients with suspected (cervical) vertebral trauma [12,20]. Most large trauma centers in the USA are now performing CT scanning for that purpose. However, many places in the world do not have access to CT scanners and radiography remains the mainstay for evaluation of trauma patients. One of the most dramatic changes in trauma management occurred in the past decade, when helical CT scanning without or with multidetector technology was shown to be much more efficient for screening than was

47

4 Imaging I: indications and controversies

radiography [21–28]. CT has superseded radiography because it is easy to perform, is faster, and most importantly, it is more efficient at detecting fractures. This transition has generated several questions regarding the future of vertebral radiography: Should radiography be performed at all? If so, how many views are needed? How long does the typical examination take? Is it cost-effective? What about radiation dose of CT compared with radiography? The answer to the first question, in the opinion of the author, is that there still is a role, albeit limited, for radiography [1]. In our level I trauma center, the majority of our trauma patients receive portable radiographs of their chest and pelvis in the resuscitation bay. From there, they go directly to the CT scanner (located in the emergency department). There scans of the brain, neck, thorax, abdomen, and pelvis are performed in rapid succession, using a multidetector unit. Thoracic and lumbar CT images are reconstructed from data obtained during the thorax–abdomen–pelvis studies. Radiographs of extremities are obtained after the CT images have been reviewed by either a trauma radiologist attending or the radiology resident. We have found cervical radiographs useful in two situations: motion artifacts and horizontal fractures. In the first instance some motion artifacts can look like fractures (Fig. 4.1). In the second, horizontal fractures in the plane of the scan may not always be detected (Fig. 4.2). Lateral cervical radiographs are generally confirmatory. In addition, we find radiography useful in patients with severe cervical spondylosis, where additional correlation is needed. Finally, we also perform radiography instead of CT on children under the age of 16 years. It has been our experience that children do not suffer the same subtle types of injury that are found in adults. Imaging studies in pediatric age patients are either normal or grossly abnormal (Fig. 1.14). In the majority of instances, frontal and lateral radiographs will suffice. Chapter 9 will address the issues of pediatric vertebral injury.

A

We have found CT to be far superior to radiography for identifying fractures throughout the vertebral column. This is particularly true for fractures of the pedicles, articular pillars and facets, laminae, and transverse and spinous processes. However, CT is not infallible. In a recent study at our institution, we found that of 297 cervical fractures in 5121 patients seen over a two year period, radiography missed 138; CT missed two fractures, both at C2, one of which was horizontal and the other one was obscured by dental artifacts [29]. For this reason we recommended a single lateral view to supplement the CT study if that examination was obtained on a multislice machine of 16 slices or fewer. If cervical radiography is to be performed, how many views are needed? Prior to the 1970’s it was standard to obtain a lateral radiograph only. However, as pointed out by Gehweiler and others the single lateral view was not sufficient to identify all fractures [12]. For the next decade a three-view cervical series (anterior–posterior, lateral, open mouth) became the norm. However, even these were felt to be inadequate, and in the 1980’s a six-view cervical series (adding bilateral supine oblique and swimmer views) was performed [12]. The supine oblique views were useful to look at the articular pillars and pedicles, as well as for evaluating the cervicothoracic junction. Unfortunately, obtaining such an extensive radiographic series is time consuming, and time is the enemy of proper trauma care. In a 2000 study, I found that the average time for obtaining these six views was 22 minutes. Moreover, 79% of the patients required one or more views to be repeated [30].

Efficacy, costs, and radiation dose How much more efficient is CT compared with radiography? In 2000, we began obtaining cervical CT examinations while the patient was undergoing cranial imaging. In 2001, the same methodology that we used for evaluating the time to perform cervical radiography was applied to cervical CT. Using

B

Fig. 4.1 Motion artifact. (A) Sagittal reconstructed CT image suggests a fracture of the body of C2 (arrow). (B) Axial CT image shows a motion artifact. (C) Lateral radiograph shows no fracture. The soft tissues are normal.

48

C

4 Imaging I: indications and controversies

A

C

B

Fig. 4.2 Horizontal fractures of C1 and C2. (A,B) Axial images through C1 do not depict the fracture. (C) Lateral radiograph of the upper spine show the horizontal fractures (arrows).

a helical scanner without multidetector capability, we found that the time for obtaining a satisfactory cervical study was 12 minutes [31]. With our new 64-slice multidetector scanner, we have found that we have diagnostic images in only a few minutes. Furthermore, sagittal and coronal images are reconstructed “instantly.” Is CT more cost-effective than radiography for trauma screening? Cost-effectiveness should be based on the actual fixed costs of the examination, including the scanner, time required, supplies used, and personnel. It should not be based on billing. The true cost-effectiveness, however, is measured by how well a particular examination establishes the diagnosis in terms of time and accuracy. When these parameters are taken into consideration, CT has been shown to be more cost-effective than radiography in the vertebral column [32–34]. An additional “cost” to be considered in comparing radiography with CT is radiation exposure. Multidetector CT examinations carry a higher radiation dose than radiography [35]. Efforts are now under way to decrease the exposure through lowering the milli-amperage. When one considers the number of repeat radiographs needed to adequately evaluate patients with suspected vertebral injuries, as well as the efficacy

of diagnosis, the higher radiation from CT may not be as significant. Frequently, solving one problem creates additional problems. While CT has now provided us with a very efficient tool for finding fractures, we need to consider if every fracture needs to be treated. How significant are some of these injuries? A look into the past may shed some light on these questions. In 1971, Martin Abel proposed an 11-view cervical radiographic series to find “occult” fractures [36]. His special angled views enabled radiologists to identify more fractures than were found on the standard three or six-view studies. However, on closer analysis, it became apparent that many of the fractures were of little or no clinical consequence and the patients required only symptomatic and supportive treatment. Now, over 30 years later, we have a diagnostic tool that is even more efficient at identifying fractures. To address this dilemma, my colleagues and I reviewed the imaging findings of 30 distinct cervical injuries or injury complexes. This allowed us to propose a new classification of cervical spine injuries using two categories: major and minor [37]. Major injuries are defined as those that produce neurologic symptoms or vertebral instability – or have a propensity

49

4 Imaging I: indications and controversies

to do so. These injuries usually require operative intervention and/or stabilization. Minor injuries are defined as those that produce no neurologic symptoms or vertebral instability and have no propensity to do so. They require symptomatic and supportive treatment only. Major injuries may be identified if any of the following findings are present: displacement greater than 2  mm, widening of a vertebral body, wide interlaminar space, wide facet joint, disrupted posterior vertebral body line, and widening of a disc space. These will be elaborated upon in more detail in Chapter 10. In addition, three specific cervical injuries are also considered major: burst fractures, locked facets (unilateral or bilateral), and type III occipital condyle fracture, where a bone fragment is displaced into the foramen magnum. All other injuries, such as isolated pillar or facet fracture, transverse process or spinous fracture, may be considered minor [37]. These principles apply to the thoracic and lumbar regions as well as the cervical spine. No discussion of radiography versus CT would be complete without addressing the issue of flexion and extension radiographs. In my institution we never use flexion–extension radiography to determine stability. The presence of muscle spasm that follows an acute neck injury results in a limited examination. Flexion–extension radiography is used primarily for those patients with minor degrees of antero- or retrolisthesis to determine whether the deformity is fixed. When these findings are accompanied by disc space or facet joint narrowing, a diagnosis of spondylosis is supported. We never use flexion–extension radiographs in unconscious patients (see below). Other authors concur with this practice [2,38–43].

“Clearing” the comatose patient The final controversy centers on the problem of “clearing” the spine on comatose patients. These patients pose a number of challenges to those involved with their trauma care. Firstly, comatose patients are unable to tell their care givers about any discomfort referable to the spine. Secondly, the severe neurologic compromise from the cerebral injury often obscures changes from a spinal cord injury. Finally, there are nursing concerns that need to be addressed. These include the possibility of skin breakdown under a rigid cervical collar and the need to turn or move the patient. It is, therefore, prudent to evaluate the integrity of the patient’s vertebral column as soon as possible to assure that there is no underlying skeletal or ligamentous injury. Is there an ideal method for achieving that goal? And, more importantly, what is the end point? A number of methods have been suggested for “clearing” the spine in comatose patients: dynamic flexion–extension fluoroscopy, lateral traction radiographs, CT, and MRI. Davis and colleagues were early advocates of dynamic fluoroscopy [38,41]. He and other investigators felt that when properly performed, fluoroscopic flexion–extension was a safe procedure [38,39,43]. However, following an incident in which a patient became quadriplegic, Davis reversed his position [41]. In addition, Anglen and colleagues [42] concluded that although

50

dynamic fluoroscopy was safe, it was not cost-effective. In their study of 837 patients over a five year period, they found that one third of the studies were inadequate. Furthermore, they found only four patients with abnormal motion on fluoroscopy. None required surgery or suffered neurologic injury [42]. Our position on the subject is that dynamic flexion– extension fluoroscopy is a procedure that is fraught with the possibility of disaster in inducing quadriplegia; we would never perform it on an unconscious patient or one with sensorial compromise. Another innovative method was proposed by Kaplan at the University of Virginia. Their protocol involved placing horizontal head traction on patients in whom radiographs and CT were normal. A head halter was placed on the patient and two lateral radiographs were obtained: first without traction and then after 15 pounds (7  kg) of horizontal traction was applied. Disc space or facet joint widening greater than 2 mm was interpreted as evidence of ligament injury (P. Kaplan, personal communication 2000). Both of the methods mentioned above have been abandoned in favor of CT. Numerous investigators have concluded that a CT examination with a modern (64-slice) multidetector scanner produces images adequate for ruling out injuries that would cause vertebral instability [42,44–47]. Should MRI be used? CT can certainly show fractures, displacements of bone fragments, and disc herniations. But what about demonstrating ligament injuries? The use of MRI has been advocated for studying patients with suspected vertebral ligamentous injury since 1989 [48]. In addition to showing the integrity of the ligaments it can also show whether there are other areas of soft tissue damage [49,50]. But is MRI necessary for all patients? Hogan and colleagues at the University of Maryland Shock Trauma unit do not think so [45], and they rely on CT as their primary screen for instability. In a study of 1400 patients they found that CT had 99% and 100% negative predictive values for ligament injury and vertebral instability, respectively [45]. At Allegheny General Hospital, we use a slightly different approach for our obtunded patients. Cervical CT is performed at the same time that the cranial scan is obtained. If this is normal, we leave the cervical collar on for the first 24 hours. After that time, if the patient is still comatose, we perform a limited MRI examination consisting of fast spin-echo T1 and T2-weighted and short-tau inversion recovery (STIR) sagittal images of the cervical region. We specifically search for evidence of ligament damage, disc herniation, and soft tissue edema, which indicates an underlying occult injury. If the study is abnormal, we obtain a complete MR examination (as will be outlined in Chapter 6). If the MR study is normal, we advise the surgeons that it is safe to remove the cervical collar. Spinal precautions are maintained in moving the patient until they regain consciousness. We generally do not do MRI of the thoracic and lumbar regions unless there is an abnormality on the CT of those areas [47]. Figure 4.3 shows our protocol for cervical spine “clearance.”

4 Imaging I: indications and controversies

•

•

Fig. 4.3 Allegheny General Hospital protocol for “clearing” the spine.

•

Conclusions and recommendations At Allegheny General Hospital we follow the recommendations of the ACR Appropriateness Criteria [2]. We also endorse the Canadian Rules [9]. • Alert adult patients who satisfy the low-risk criteria (no loss of consciousness, no alcohol and/or drugs, no cervical

Daffner RH. Controversies in cervical spine imaging in trauma patients. Semin Musculoskeletal Radiol 2005; 9:105–115.

2.

Daffner RH, Hackney DB. ACR Appropriateness Criteria on suspected spine trauma. J Am Coll Radiol 2007; 4:762–775.

3.

Stiell IG, Greenberg GH, McKnight RD, et al. Decision rules for the use of radiography in acute ankle injuries. Refinement and prospective validation. JAMA 1993;269:1127–1132.

4.

5.

6.

7.

Stiell IG, Greenberg GH, Wells GA, et al. Derivation of a decision rule for the use of radiography in acute knee injuries. Ann Emerg Med 1995; 26:405–413. Mirvis SE, Diaconis JN, Chirico PA, et al. Protocol-driven radiologic evaluation of suspected cervical spine injury: efficacy study. Radiology 1989; 170:831–834.

•

identify high-risk patients for helical CT screening. AJR Am J Roentgenol 2000;174:713–717.

References 1.

•

tenderness, no distracting injury, and no neurologic findings) should not have imaging. Patients who do not fall into this category should undergo thin-section CT examination of the entire vertebral column, which includes sagittal and coronal multiplanar reconstructed images. The thoracic and lumbar images may be derived from the thorax–abdomen–pelvis study. Those patients who cannot be examined by CT should have, as a minimum, a threeview radiographic examination to provide preliminary assessments of the likelihood of vertebral injury until CT can be performed. Radiography is recommended for children under age 14; above that age they should be examined the same as adults. Magnetic resonance imaging for evaluating possible spinal cord injury or compression, as well as ligamentous injuries in acute cervical trauma. Flexion–extension radiography is best reserved for the follow-up of symptomatic patients beyond the initial hospitalization.

8.

Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. N Engl J Med 2000; 343:94–99.

9.

Stiell IG, Wells GA, Vandemheen, et al. The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA 2001;286:1841–1848.

10. Stiell IG, Clement CM, McKnight RD, et al. The Canadian C-spine rule versus the NEXUS low-risk criteria in patients with trauma. N Engl J Med 2003; 349:2510–2518. 11. Calenoff L, Chessare JW, Rogers LF, et al. Multiple level spinal injuries: importance of early recognition. AJR Am J Roentgenol 1978;130:665–669. 12. Gehweiler JA Jr., Osborne RL Jr., Becker RF. The Radiology of Vertebral Trauma. Philadelphia, PA: WB Saunders, 1980.

Vandemark R. Radiology of the cervical spine in trauma patients: practice pitfalls and recommendations for improving efficiency and communication. AJR Am J Roentgenol 1990;155:465–472.

13. Gupta A, El Masri WS. Multilevel spinal injuries: incidence, distribution, and neurological patterns. J Bone Joint Surg 1989;71B:692–695.

Hanson JA, Blackmore CC, Mann FA,Wilson AJ. Cervical spine injury: a clinical decision rule to

14. Powell JN, Waddell JP, Tucker WS, et al. Multiple-level noncontiguous spinal injury. J Trauma 1989;29:1146–1151.

15. Jones HK, McBride GG, Mumby RC. Sternal fractures associated with spinal injury. J Trauma 1989;29:360–364. 16. Woodring JH, Lee C, Jenkins K. Spinal fractures in blunt chest trauma. J Trauma 1988;28:789–793. 17. Burkus JK, Denis F. Hyperextension injuries of the thoracic spine in diffuse idiopathic skeletal hyperostosis. J Bone Joint Surg 1994;76A:237–243. 18. Graham B, Van Peteghem K. Fractures of the spine in ankylosing spondylitis: diagnosis, treatment, and complications. Spine 1989;14:803–807. 19. Hendrix RW, Melany M, Miller F, et al. Fracture of the spine in patients with ankylosis due to diffuse skeletal hyperostosis: clinical and imaging findings. AJR Am J Roentgenol 1994; 162:899–904. 20. Harris JH Jr., Mirvis SE. The Radiology of Acute Cervical Spine Trauma, 3rd edn. Baltimore, MD: Williams & Wilkins, 1996. 21. Nuñez DB Jr., Ahmad AA, Coin CC, et al. Clearing the cervical spine in multiple trauma victims: a timeeffective protocol using helical CT. Emerg Radiol 1994;1:273–278. 22. Nuñez DB Jr., Zuluaga A, FuentesBernardo DA, et al. Cervical spine trauma: how much more do we learn by routinely using helical CT? Radiographics 1996;16:1307–1318.

51

4 Imaging I: indications and controversies

23. Lawrason JN, Novelline RA, Rhea JT, et al. Can CT eliminate the initial portable lateral cervical spine radiograph in the multiple trauma patient? A review of 200 cases. Emerg Radiol 2001;8:272–275. 24. Ptak T, Kihiczak D, Lawrason JN. Screening for cervical spine trauma with helical CT: experience with 676 cases. Emerg Radiol 2001;8:315–319. 25. Li AE, Fishman EK. Cervical spine trauma: evaluation by multidetector CT and three-dimensional volume rendering. Emerg Radiol 2003; 10:34–39. 26. Brohi K, Healy M, Fotheringham T, et al. Helical computed tomographic scanning for the evaluation of the cervical spine in the unconscious, intubated trauma patient. J Trauma 2005;58:897–901. 27. Brown CV, Antevil JL, Sise MJ, Sack DI. Spiral computed tomography for the diagnosis of cervical, thoracic, and lumbar fractures: its time has come. J Trauma 2005;58:890–895. 28. Holmes JF, Akkinepalli R. Computed tomography versus plain radiography to screen for cervical spine injury: a meta-analysis. J Trauma 2005;58:902–905. 29. Daffner RH, Sciulli RL, Rodriguez A, Protetch J. Imaging for evaluation of suspected cervical trauma: a 2-year analysis. Injury 2006;37:652–658. 30. Daffner RH. Cervical radiography for trauma patients: a time-effective technique? AJR Am J Roentgenol 2000; 175:1309–1311. 31. Daffner RH. Cervical helical CT for trauma patients: a time analysis. AJR Am J Roentgenol 2001;177:677–679. 32. Blackmore CC, Ramsey ST, Mann FA, Deyo RA. Cervical spine screening with CT in trauma patients: a costeffectiveness analysis. Radiology 1999; 212:117–125.

52

33. Saini S, Seltzer SE, Bramson RT. Technical cost of radiologic examinations: analysis across imaging modalities. Radiology 2000; 216:269–272. 34. Saini S, Sharma R, Levine LA, et al. Technical cost of CT examinations. Radiology 2001;218:25–26. 35. Rybicki F, Nawfel RD, Judy PF, et al. Skin and thyroid dosimetry in cervical spine screening: two methods for evaluation and a comparison between a helical CT and radiographic trauma series. AJR Am J Roentgenol 2002; 179:933–937. 36. Abel MS. Occult Traumatic Lesions of the Cervical Vertebrae. St. Louis, MO: Warren Green, 1971. 37. Daffner RH, Brown RR, Goldberg AL. A new classification for cervical vertebral injuries: influence of CT. Skeletal Radiol 2000;29:125–132. 38. Davis JW, Parks SN, Detlefs CL, et al. Clearing the cervical spine in obtunded patients: the use of dynamic fluoroscopy. J Trauma 1995; 39:435–438. 39. Brady WJ, Moghtader J, Cutcher D, et al. ED use of flexion–extension cervical spine radiography in the evaluation of blunt trauma. Am J Emerg Med 1999;17:504–508. 40. Dwek JR, Chung CB. Radiography of cervical spine injury in children: are flexion–extension radiographs useful for acute trauma? AJR Am J Roentgenol 2000;174:1617–1619. 41. Davis JW, Kaups KL, Cunningham MA. Routine evaluation of the cervical spine in head-injured patients with dynamic fluoroscopy: a reappraisal. J Trauma 2001;50:1044–1047. 42. Anglen J, Metzler M, Bunn P, Griffiths H. Flexion and extension views are not cost-effective in a cervical spine clearance protocol for obtunded patients. J Trauma 2002;52:54–59.

43. Spiteri V, Kotnis R, Singh P, et al. Cervical dynamic screening in spinal clearance: now redundant. J Trauma 2006;61:1171–1177. 44. Padayachee L, Cooper D J, Irons S, et al. Cervical spine clearance in unconscious traumatic brain injury patients: dynamic flexion–extension fluoroscopy versus computed tomography with three-dimensional reconstruction. J Trauma 2006; 60:341–345. 45. Hogan GJ, Mirvis SE, Shanmuganathan K, Scalea TM. Exclusion of unstable cervical spine injury in obtunded patients with blunt trauma: is MR imaging needed when multi-detector row CT findings are normal? Radiology 2005;237:106–113. 46. Diaz JJ Jr., Aulino JM, Collier B, et al. The early work-up for isolated ligamentous injury of the cervical spine: does computed tomography scan have a role? J Trauma 2005; 59:897–903. 47. Sekula RF Jr., Daffner RH, Quigley MR, et al. Exclusion of cervical spine instability in patients with blunt trauma with normal multidetector CT (MDCT) and radiography. Br J Neurosurg 2008;22:669–674. 48. Emery SE, Pathria MN, Wilber RG, et al. Magnetic resonance imaging of posttraumatic spinal ligament injury. J Spinal Disord 1989;2:229–233. 49. Benzel EC, Harr BL, Ball PA, et al. Magnetic resonance imaging for the evaluation of patients with occult cervical spine injury. J Neurosurg 1996; 85:824–829. 50. Saifuddin A. MRI of acute spinal trauma. Skeletal Radiol 2001; 30:237–246.

Chapter

5

Imaging of vertebral trauma II: radiography, computed tomography, and myelography Richard H. Daffner

The previous chapter discussed the indications for imaging patients with suspected vertebral injuries. This chapter discusses the three imaging modalities that use ionizing radiation: radiography, CT, and myelography. These techniques are frequently used in combination to arrive at the correct diagnosis. This chapter reviews each of these imaging formats and illustrates their uses in the diagnosis of vertebral injury. Chapter 6 will discuss MR imaging. Chapters 7 and 8 describe the integrated use of multiple imaging techniques.

Radiography Radiography was the foundation on which the diagnosis of vertebral injury was made [1,2]. As Chapter 4 illustrated, in an era when specialized imaging techniques such as CT and MR are commonplace, radiography has taken a “back seat.” In the past, radiography was used extensively and relied upon to screen the patient and to make an initial diagnosis. Special imaging techniques were then used to confirm the initial impression and to outline the extent of damage. However, despite the advantages of CT, in many instances it is still necessary to refer to radiographs for guidance, particularly for operative planning. Furthermore, there are still many places in the world where CT is not readily available on an emergency basis and trauma physicians still must utilize radiography. For this reason we include radiography in this discussion.

Techniques What is the “routine” series of radiographs for examining an adult with an acute vertebral injury? This question is frequently asked of radiologists by their surgical colleagues. As mentioned, there are differing opinions of which views should be routine.

Cervical region Once it is determined that a patient needs cervical radiography, what is the least number of views required to ensure that a significant injury has been excluded? Several studies have attempted to address this question [1–5]. Most investigators agree that the absolute minimal radiographic views are the supine lateral, the anterior–posterior (AP), and, where possible, the atlanto-axial (odontoid). Freemyer and coworkers

[1] and MacDonald and colleagues [2] believed the three-view study was sufficient. To a great extent, the American College of Radiology Appropriateness Criteria agrees with this premise [6]. However, it has been our experience at the Allegheny General Hospital, as well as that of colleagues at other large trauma centers, that the average trauma patient is quite large (in excess of 100 kg [220 lb]). In these patients, it is extremely difficult to completely evaluate the cervicothoracic junction on lateral radiographs. In most instances, however, the supine (“trauma”) oblique views adequately demonstrate this region. I agree with advocates of the three-view cervical series that the supine oblique views generally do not provide significant additional information about injuries. However, the fact that the supine oblique views can adequately demonstrate the anatomy of the cervicothoracic junction in almost every instance would seem to justify its use. The experience at Allegheny’s trauma center indicates that the combination of a normal AP and normal bilateral supine oblique radiographs is sufficient to adequately evaluate the cervicothoracic junction. Of course, if the patient is able to undergo cervical CT, this issue is moot. In the cervical region, the most important projection is the lateral. Gehweiler and colleagues [7] pointed out that at least two thirds of significant pathology can be detected on this view (Table 5.1). It is mandatory, therefore, that the surgeon and the radiologist not rely solely on the lateral view to clear the cervical region in a trauma patient [6–8]. The hazards of this practice are illustrated in Fig. 5.1. From a practical standpoint, however, the presence of life-threatening injury outside the vertebral column may dictate that the patient be taken immediately to surgery before a complete cervical series can be obtained. In dealing with patients such as this, active consultation and cooperation among trauma surgeons, anesthesiologists, radiologists, neurosurgeons, and orthopedic surgeons is necessary. The treatment of life-threatening injuries always precludes obtaining a complete series of radiographs. At Allegheny General Hospital, all radiographs of the spine are obtained with the patient in a supine position. We do not turn the patient for lateral views or oblique views. A portable X-ray unit usually is adequate for filming. Ambulatory patients may be studied in the upright position. All images are processed and displayed on our digital imaging system.

53

5 Imaging II: radiography, CT, and myelography

Table 5.1 Efficacy of lateral radiographs for cervical trauma Injury

Demonstrable

Occipito-atlantal dislocation

+

Atlanto-axial dislocation

+

Atlas fractures  Posterior arch

+

 Anterior arch

+

 Burst (Jefferson)

+/−

 Lateral mass

−

 Transverse process

−

Axis fractures  Dens

+

 Vertebral body

+

 Vertebral arch (“hanged-man”)

+

Lower cervical vertebral bodies  Simple flexion

+

 Burst

+

 Uncinate process

−

Lower cervical vertebral arch  Spinous process

+

 Locked facets

+

 Articular pillar

+/−

 Lamina

+/−

 Pedicle

−

 Transverse process

−

Flexion sprain

A

+

Extension sprain

+

Flexion fracture–dislocation

+

Extension fracture–dislocation

+

The lateral view is obtained by means of a horizontal beam with a grid cassette and using 40 inch focal film distance. The cassette is placed adjacent to the patient’s head and as close to the shoulders as possible. Gentle traction is placed on the shoulders to facilitate imaging of C7 (Fig. 5.2). Under no circumstances should traction be applied to the head. In individuals with upper limb fractures, it may be impossible to place traction on the upper limbs. Additional views with the “swimmer’s” technique may be necessary for complete imaging of the lower cervical region. Despite all these efforts, however, it still may be impossible to see C6 and C7 in muscular or obese patients with heavy shoulders. In these patients, CT with sagittal reconstruction will be necessary to clear this area. Nevertheless, in most instances, the supine oblique views are sufficient to demonstrate the region. After an adequate lateral radiograph has been obtained, the X-ray tube is placed in an upright position with 20° of cranial angulation of the central beam (Fig. 5.3). The point of entry is at the cricoid cartilage (C6). The cassette is placed beneath the patient’s neck. This can be accomplished easily by placing the film under the backboard on which the patient is lying. Again a 40 inch focal film distance is used [7]. The next view obtained is that of the atlanto-axial region with the patient’s mouth open when possible (Fig. 5.4). This is the most frequently repeated view [9]. This view can be delayed until the patient is able to fully cooperate. For this view, it may be necessary to remove the anterior portion of the cervical collar in which the patient has arrived. To prevent motion, sandbags should be placed at either side of the patient’s head and secured with a generous amount of tape; alternatively, an assistant can hold the head. Angled views for demonstrating the arches of C1 may also be necessary [10]. One of the more interesting and valuable views is the supine, or “trauma,” oblique projection. This view was developed

B

C

D

Fig. 5.1 Hazards of relying on a lateral view only. (A) Lateral radiograph shows a fracture of the anterior margin of the body of C4 (arrows). (B) Frontal view shows fractures of C5 and C6 (arrows). (C,D) The CT images of C5 (C) and C6 (D) show sagittal body fractures and laminar fractures (arrows) of these burst injuries.

54

5 Imaging II: radiography, CT, and myelography

Fig. 5.2 Normal lateral cervical radiograph. The anterior and posterior vertebral body lines align. The spinolaminar line (arrows) is smooth and uninterrupted. The facet joints overlap in an orderly fashion (imbrication). The interlaminar (interspinous) distances are uniform. The posterior vertebral body line is solid.

Fig. 5.3 Normal anterior–posterior cervical radiograph. Alignment at the lateral margins is normal. The pedicles are normally aligned, and the distances between them (double arrow) do not deviate more than 2 mm from level to level. Note the cervical transverse processes (C7) point downward and the thoracic (T1) point up.

independently at approximately the same time by Gehweiler and colleagues [7] and Abel [11]. For this projection, the cassette is placed adjacent to the head and neck with the patient supine on the table. The X-ray tube is angled 30 to 40° off the horizontal. We modified this view at Allegheny by adding a 15° cranial tilt of the tube in addition to the off-horizontal tilt, because the lower cervical region is not always adequately shown because of the patient’s shoulders. The result of this additional angulation is that the cervicothoracic junction is demonstrable in most patients, even those with heavy shoulders. The resulting images from either of these techniques show distortion because of the angulation. Nevertheless, the vertebral bodies, pedicles, articular pillars, and laminae are adequately demonstrated [7,11]. In addition, the posterior arch of the atlas is clearly seen. Less well recognized is the fact that a pair of these radiographs essentially represents two views of the same region at approximately 90º to each other. Therefore, the cardinal principle of radiographic diagnosis – to examine an injured part with two views at 90º – is preserved. As mentioned above, a diagnosis in the lower cervical region can be made with confidence by means of a combination of the AP view and both supine oblique views. At Allegheny General Hospital, active flexion and extension views are used on a limited basis. Other hospitals use these routinely. Although Bohrer and colleagues [12] conducted a study showing the value of routine flexion and extension views, other specialists, myself included, believe that they are not necessary

Fig. 5.4 Normal atlanto-axial (open-mouth) view. There is normal alignment between the lateral masses of C1 and the lateral margin of C2 (arrows). The spaces between the dens and the lateral masses are uniform.

in every case. As mentioned above, we find these views to be of limited value in the evaluation of patients with acute trauma. Flexion and extension views should be reserved for patients who have minor degrees of anterolisthesis or retrolisthesis. In most cases, the cause of the listhesis is degenerative disc disease at the same level (Fig. 5.5). Under no circumstances should the patient’s head be passively moved for this study. It is best to leave a cervical collar in place until the patient is able to cooperate fully. Flexion and extension radiography is a hands-off examination for the radiologist and the technologist. A mentally alert patient is instructed to flex and extend to the point of discomfort only. A physician should be present to supervise. The value of such supervision is to reiterate to the patient to stop moving if they experience pain. In 25 years of experience in emergency department radiology at a level I trauma center, I know of no alert patient who has injured himself or herself performing these movements, nor do my colleagues in similar practices. Moreover, there is no report in the literature of the development of significant injury when these studies are performed as described above. No discussion of flexion–extension radiography would be complete without mentioning “whiplash,” a common term used with regard to cervical trauma. Whiplash, in fact, is a descriptive term that attempts to define a mechanism of injury to the cervical column (hyperextension followed by flexion)

55

5 Imaging II: radiography, CT, and myelography

A

B

rather than an injury itself [12–14]. The term is not unlike “twisted ankle.” When the specific pathologic injury is not obvious clinically or radiologically, these terms or equally nonspecific ones are often used to define the injury itself. This may be a satisfactory lay term, but it should be clear that it is not a specific pathologic diagnosis. Some studies try to define the radiologic findings of whiplash injury [12]. The actual injuries incurred from this mechanism fall in a spectrum from minor soft tissue injury to fracture or dislocation [14]. Consequently, attempting to define a single pathologic lesion for whiplash or twisted ankle clinically or radiologically is futile. A more useful approach is to define stable and unstable injuries resulting from the whiplash mechanism and to try to distinguish serious injuries (those likely to have long-term symptoms) from minor injuries (those likely to improve rapidly). Soft tissue extension or flexion injuries range from isolated muscular, ligamentous, and capsular stretching and tearing, with or without hematomas, to complete unstable hyperflexion subluxation or hyperextension dislocation injuries. The most common and possibly the most practical use of the term whiplash injury is to reserve it for soft tissue injuries that occur after extension or flexion trauma when there is no apparent fracture or dislocation: injuries frequently called “neck sprains.” Although Bohrer and colleagues [12] recommended flexion and extension radiographs, MR imaging is the procedure of choice to identify the specific soft tissue pathology. The use of MR imaging for this entity is discussed in Chapter 6. In the cervical region, however, there are certain pitfalls and limitations to radiography, not the least of which are the cumbersome nature of the procedure and the time needed to obtain a satisfactory examination. Woodring and Lee [15] reviewed the limitations of radiographs and indications for CT

56

Fig. 5.5 Malalignment secondary to spondylosis. (A) Retrolisthesis of C4 on C5. Note the malalignment of the spinolaminar line at that level (arrows) and the narrow disc space. (B) Anterolisthesis of L4 on L5. There is narrowing of the L4 disc space and the facet joint of L4–L5 (lower arrow). Compare with L3–L4 (upper arrow).

in evaluating patients suspected of having cervical injury. They found that cervical radiographs could not always be relied on solely to make a diagnosis. They recommended that CT be used whenever radiographs found an abnormality. If radiographs were normal but the patient was at high risk for cervical injury, they recommended CT. As mentioned above, CT has replaced radiographs in our institution because of the ability of the CT examination to find more fractures in a fraction of the time a radiographic study requires [9,16]. Adequate radiographic visualization of the cervicothoracic junction is frequently difficult. Muscular or obese patients present special diagnostic problems. Failing to adequately demonstrate the cervicothoracic junction presents the hazard of missing an occult fracture or dislocation (Fig. 5.6). Every effort should be made to obtain adequate demonstration of this area, making use of all of the imaging resources to accomplish this goal. What about children? Children under 16 years of age do not need CT; radiography is adequate. Those over 16 years should be studied the same way as adults, primarily with CT [6]. Furthermore, radiographic examination of the cervical vertebrae in children need not be as extensive as in adults. Children do not suffer the same types of injury as adults do, primarily because of the increased suppleness of the pediatric cervical region. Cervical radiographs in children with suspected vertebral injury tend to fall into two categories: normal or grossly abnormal. The subtle radiographic findings found in adults (discussed in Chapter 8) are rarely present in children. Injuries commonly found in children include occipito-atlantal disruptions, atlanto-axial rotary subluxation or fixation, and occasional physeal injuries. Therefore, my institution limits the pediatric cervical radiographic examination to lateral, AP, and open-mouth views. We do not obtain flexion or extension

5 Imaging II: radiography, CT, and myelography

A

D

B

C

Fig. 5.6 Failure to adequately demonstrate the cervicothoracic junction. (A) Lateral radiograph shows six complete vertebrae and a portion of C7. No abnormalities are detected. (B) Same patient with shoulders pulled down shows anterior dislocation of C7 on T1 with perching of the facets (arrow). (C) Lateral radiograph in another patient shows six complete vertebrae and only the top of C7. (D) The anterior–posterior radiograph shows dislocation of C7 on T1 to the right. Note the malalignment of the spinous processes (arrows).

views on these patients. Chapter 9 discusses pediatric vertebral injuries in detail. Another area of concern is patients in whom there is straightening or reversal of the normal cervical lordosis. This is encountered with both radiography as well as CT. How can these patients, in whom the radiographic abnormality is caused purely by position or muscle spasm, be differentiated from patients with true ligamentous injury? There are several helpful clues on the lateral radiograph that can provide the answer (Fig. 5.7): • if the abnormality is purely a result of position, the angle of the mandible is close to the cervical column (“military” posture) • there is no disruption of the spinolaminar line; this indicates that no posterior ligamentous damage has occurred • there is no evidence of soft tissue abnormality in the prevertebral region

•

•

in an older individual, in whom there are minor degrees of anterolisthesis or retrolisthesis (Fig. 5.5A), there is evidence of degenerative disease at the level of the abnormality it may be necessary to obtain flexion and extension views to determine whether or not the deformity is fixed.

Thoracic and lumbar regions The same techniques are used for adults and children in the thoracic and lumbar regions. The radiographic examination of the thoracic vertebral column can be accomplished with the patient supine. An AP view (Fig. 5.8A) is obtained immediately after chest radiography and then a horizontal beam lateral radiograph (Fig. 5.9A) is obtained. Because of the shoulders and arms, the upper thoracic region usually is poorly demonstrated, and a swimmer’s view may be necessary as well. In the lumbar region, supine AP and cross-table lateral radiographs should be adequate to diagnose most injuries

57

5 Imaging II: radiography, CT, and myelography

A

B

C

Fig. 5.7 Loss of lordosis. (A) Normal lateral cervical radiograph in a patient with spasm. (B) Reversal of lordosis and kyphotic angulation at C3–C4 in a patient with a hyperflexion sprain. Note the wide interlaminar distances at C3 and C4 (*). (C) Sagittal CT reconstructed image shows same findings.

A

58

B

A

B

Fig. 5.8 Normal anterior–posterior radiographs of thoracic (A) and lumbar (B) vertebrae. The vertebral margins, pedicles, and spinous processes align. The interspinous spaces and interpedicle distances (double arrow) are normal. The paravertebral soft tissues are normal.

Fig. 5.9 Normal lateral radiographs of thoracic (A) and lumbar (B) vertebrae. In the thoracic region, there is a gentle kyphosis and there is a lordosis in the lumbar region. The posterior vertebral body lines are interrupted in the middle by a nutrient foramen (arrows).

(Figs. 5.8B and 5.9B). Sacral injuries are usually the result of pelvic fractures. An AP view of the pelvis is included as part of our routine trauma screening series. It is not necessary to obtain oblique views of the lumbar column; frontal and lateral radiographs are usually adequate.

The only radiographic pitfall in diagnosing an injury in the thoracic region is improper imaging of the upper thoracic column in the lateral position, and CT with sagittal reconstruction should be performed to evaluate all areas of suspected (upper) thoracic abnormality (Fig. 5.10). There are no

5 Imaging II: radiography, CT, and myelography

A

B

C

Fig. 5.10 High thoracic fracture. (A) Chest radiograph shows widening of the paraspinal lines (arrows). (B,C) Magnetic resonance imaging with T2-weighted (B) and inversion recovery sagittal (C) images show a severe fracture of T2 with retropulsion of bone into the vertebral canal (arrow in B). Note the cord hemorrhage (*) in C as well as the bright signal posterior to the cord injury indicating severe posterior ligamentous damage.

radiographic pitfalls, other than overlying bowel gas and content, in diagnosing injuries in the lumbar region.

Computed tomography The development of CT in the early 1970s revolutionized medicine and the practice of diagnostic radiology. For the first time, it was possible to obtain cross-sectional images of areas hitherto unseen by noninvasive diagnostic methods. It soon became apparent that one of the prime diagnostic uses for CT would be for the evaluation of patients with vertebral trauma [17–23]. Today, multidetector CT, with its rapid scan time and ability for excellent multiplanar and three-dimensional reconstruction, allows improved diagnoses of vertebral injuries [6,24–28]. Vertebral CT is easy to perform. We obtain the cervical scan at the same time as the patient undergoes a cranial scan. The patient need only lie in a supine position and not move. Occasionally, it may be necessary to induce immobility pharmacologically to obtain an adequate study in an acutely injured patient. Young children, if they are scanned, frequently require sedation to obtain an adequate study. There is no predetermined number of images. Our standard procedure for a cervical scan on our 64-slice multidetector CT unit is to obtain contiguous 2 mm slices from the skull base to the bottom of T1 or T2. Then 1  mm axial images are reconstructed from the data using both bone and soft tissue windows (Fig. 5.11). In addition, sagittal and coronal multiplanar images are also reconstructed from the 1  mm data set. Intravenous contrast enhancement is not required unless a CT angiogram is ordered. Thoracic and lumbar scans are obtained either as freestanding studies or, more commonly, from the data gathered during the thorax–abdomen–pelvis body scan. For scans

obtained by either method, the data are collected from 5 mm slice thickness, reconstructed to axial images at 2 mm. Sagittal and coronal multiplanar images are also reconstructed to complete the study. A stand-alone thoracic scan is obtained from C6 through L1 and a lumbar study from T11 through the lower sacrum. Scout views are obtained to determine the level of the scan. They are performed in the lateral position in the cervical region, in the AP position in the thoracic region, and either in the AP or lateral position for free-standing studies in the lower thoracic and lumbar regions. Those derived from the thorax– abdomen–pelvis use an AP scout view. Enlarged scout images are displayed with and without level annotations. It is important that the thoracolumbar scout views include the pelvis for reference points to properly determine levels. A significant amount of information about the extent of injury is also provided by CT [18,25–30]. It is the best method for determining the presence and degree of canal encroachment (Fig. 5.12) or intervertebral foramen encroachment (Fig.  5.13). It is also useful for demonstrating fractures of the laminae, pedicles (Figs. 5.14 and 5.15), and articular pillars, particularly those associated with perched or locked facets where the images of both facets of the joint are present (Fig.  5.16 and Fig.  5.17) [30]. The ability to perform sagittal, coronal, or three-dimensional reconstruction with CT is an additional benefit (Fig. 5.18) [29,31]. Most imaging in the USA is performed with a digital imaging system utilizing picture archiving and computer storage (PACS). As computers have become more sophisticated, it is now possible to manipulate data to improve images. These manipulations include the ability to darken or lighten an image or to shift data to improve demonstration of certain

59

5 Imaging II: radiography, CT, and myelography

A

B

C

D

A

B

Fig. 5.11 Value of bone and soft tissue windows. (A) Sagittal reconstructed CT image at bone window shows a fracture of the articular pillar of C5 (arrow). (B) Same image at soft tissue window shows a clot (arrow) in the adjacent vertebral artery. The clot shows in A but not as well as in B. Note the fracture is not as well defined in B. (C) Axial CT image at bone window shows locking of the facet on the right (arrow). (D) Soft tissue image at same level shows a disc herniation (*).

C

Fig. 5.12 Lumbar burst fracture. (A) Lateral radiograph shows compression of the vertebra with retropulsion of bone fragments from the posterior vertebral line (arrow). (B) Sagittal CT reconstructed image shows this displaced fragment to advantage (arrow). (C) Axial CT image shows the comminuted fracture of the vertebra with significant canal compromise (*).

60

5 Imaging II: radiography, CT, and myelography

A

B

Fig. 5.14 Lamina/spinous process fractures (arrows) in cervical (A) and lumbar (B) vertebrae.

Fig. 5.13 Foramen encroachment by an articular pillar fracture. The left-sided pedicle fracture enters the transverse foramen. The contrast-filled vertebral artery is displaced medially by a hematoma (arrow).

A

Fig. 5.15 Pedicle fracture (arrow) in a lower thoracic vertebra. Note the paraspinal hematoma.

B

Fig. 5.16 Unilateral facet fractures with locking. The arrow shows the point of lock. Note the spinous process is rotated toward the side of locking.

Fig. 5.17 Unilateral facet fractures with locking (same patient as in Fig. 5.16). Arrow shows the facet lock.

areas. This has become particularly important when reviewing images on patients who are not lying perfectly straight in the CT gantry (Fig. 5.18). A useful adjunct to cervical CT is the use of CT angiography [32–34]. It is used primarily for patients who have fractures that involve the transverse foramina and who are suspected of having injury or occlusion to the vertebral artery. It is also used in patients who have sustained a penetrating injury to the neck in order to determine the integrity of the carotid arteries. The typical scan is performed from the level of the orbits to the aortic arch (as determined on an AP scout view). A nonionic contrast (100 ml) is injected intravenously at a rate of 3.5–4.0 ml/s. Scanning is begun 12–18 seconds after the injection begins and is performed at 2 mm intervals, with axial reconstruction at 1 mm. In addition, sagittal and coronal three-dimensional volumetric reconstruction is performed for interpretation (Fig. 5.19). In many institutions, CT is combined with myelography using water-soluble contrast media to evaluate traumatic

61

5 Imaging II: radiography, CT, and myelography

A

B

D

C

Fig. 5.18 Image manipulation with a digital imaging system (PACS). (A) Sagittal reconstructed CT image shows orientation of the original scan (arrows). (B) Axial image through C1 and the tip of the dens (D) at the level shown in A. Note the posterior arch of C1 is not demonstrated. (C) Sagittal image shows orientation line (arrows) following electronic manipulation to demonstrate data through the anterior and posterior arches of C1. (D) Reconstructed axial image shows the entirety of C1.

A

62

B

Fig. 5.19 CT angiography of the same patient as in Fig. 5.11. (A) Coronal reconstructed image shows a clot in the vertebral artery on the left (arrow). (B) Subtracted image shows occlusion of the left subclavian artery (arrow). Note that the right vertebral artery (R) is patent.

5 Imaging II: radiography, CT, and myelography

encroachment of the subarachnoid space and spinal cord by bone fragments or herniated intervertebral disc fragments (Figs. 5.20 and 5.21). CT myelography is also useful for studying cervical nerve root avulsions (Fig. 5.22) [35] and posttraumatic cystic myelopathy [36]. In many institutions, CT myelography is performed because MR imaging is unavailable, the patient is too unstable, or the patient has a contraindication for the study [6].

The craniovertebral and cervicothoracic junctions Use of CT allows better evaluation of the two regions that posed significantly difficult diagnostic problems in the past  – the craniovertebral and the cervicothoracic junctions. Fractures of the craniocervical junction, specifically of the occipital condyles, were once considered rare. Fractures of the cervicothoracic junction are also difficult to see on radiographs. However, CT demonstrates these areas not only on axial views but also on the sagittal and coronal multiplanar reconstructed images.

A

A

What is the true incidence of “occult” fractures in the craniocervical junction? Furthermore, how significant are many of these fractures? The publications by Blacksin and Lee [37] and Link and associates [38] first called these injuries to our attention. Occipital condyle fractures had been considered rare [39,40], but this rarity may simply have been the result of a failure of recognition. Most of these injuries are either impacted fractures of the condyle as a result of axial loading (type I) or fractures that result from the extension of a basilar skull fracture into the condyle (type II) (Fig. 5.23) [37,40]. Most occipital condyle injuries, if unaccompanied by more serious soft tissue injury of either the brain or spinal cord, are probably associated with nothing more than pain in the craniocervical region or possible headache. The most rare, but most significant, of these fractures are caused by avulsion of all or a portion of the condyle by the alar ligaments (type III) [37,40]. These are likely to produce craniocervical instability, neurologic impairment, or both. In most instances, there is clear-cut soft tissue swelling on lateral radiographs of the cervical region.

Fig. 5.20 Computed tomography myelogram. (A) Sagittal reconstructed image shows a displaced fragment of bone in the vertebral canal (*). (B) Axial image shows effacement of the column of contrast in the subarachnoid space (arrows) by the fragment.

B

B

C

Fig. 5.21 Computed tomography myelogram showing cervical disc herniation. (A) Lateral radiograph shows indentation (arrow) on the anterior aspect of the contrast-filled subarachnoid space. (B,C) The CT images show effacement of the subarachnoid space (arrow) by the herniated disc.

63

5 Imaging II: radiography, CT, and myelography

A

C

B

Fig. 5.22 Computed tomography myelogram showing nerve root avulsion. (A) Axial image shows filling of the nerve sheath on the left (large arrow) and extravasation of contrast into the epidural space (small arrow). (B) Image slightly higher shows filling of the nerve sheath (arrow). (C) Coronal reconstructed image shows the normal roots (black arrows) and the contrast filling the sheath of the avulsed root on the left (*).

Pitfalls and limitations The use of CT has pitfalls and limitations [41,42], both patient related and technical. Patient-related problems result primarily from motion, patient size, and artifacts from dental fillings and metallic implants. Motion during the study results in blurred images and the possibility of a missed diagnosis. Motion can also disrupt multiplanar reconstructions. In many instances, the motion artifact may resemble a fracture (Fig. 5.24). Motion artifacts are easily identified on the axial images. On sagittal reconstructed images, the region of the artifact frequently has a shift in data in the surrounding tissues. When in doubt, a repeat scan, referral to the scout view, or radiographs may be used to solve the dilemma. The patient’s weight is a serious consideration. Most modern CT machines have a patient weight limit of 400 to 450  lb (180–205 kg) because the table must project into the gantry for the examination. The presence of dental fillings [43] (Fig. 5.25) or other metallic implants (e.g., rods, hooks, screws, plates, vascular clips, bullet fragments) can cause artifacts that severely compromise the radiographic image. Many of the new scanners have metal-suppression software that can reduce these artifacts.

64

Three technical pitfalls can result in incomplete information being obtained from the CT examination: the partial volume averaging effect, poor level calibration, and fractures in the plane of the scan. The partial volume averaging effect is a well-known CT pitfall. Normally, CT gives an image that represents an average of the radiographic densities of all structures contained within that section of tissue. Any normal structure or abnormality that is not completely located within that plane may be distorted or totally discarded from the final image (Fig. 5.26). In most instances, a fracture will be seen on more than one crosssection. It is possible, however, particularly in dealing with thin structures such as the laminae or posterior arch of C1, that a fracture is demonstrated on one view only (Fig. 5.27). Furthermore, where there is normal overlap of structures from adjacent vertebrae, lines of interruption will be apparent in the bony shadow (Fig. 5.28); these should not be misinterpreted as fractures. On some CT machines, a discrepancy in annotation on the scout film can result in erroneous information being obtained about the level of injury. This generally does not present a significant problem when knowledge of anatomy (dens, ribs, sacrum) is used to determine the levels of fracture.

5 Imaging II: radiography, CT, and myelography

A

B

C

D

A

D

B

Fig. 5.23 Occipital condyle fracture (arrows). (A,B) Axial CT images. (C,D) Coronal reconstructed CT images. These fractures probably would not have been seen on radiographs.

C

Fig. 5.24 Motion artifacts. (A) Axial image clearly shows motion. (B) Sagittal reconstructed CT image in another patient suggests a fracture at the base of the dens (arrow). (C) Axial image of the same patient shows fuzziness along the margins of the bone (arrows), indicating motion. (D) Lateral radiograph shows no evidence of fracture. The soft tissues are normal, as well, making a fracture unlikely.

65

5 Imaging II: radiography, CT, and myelography

A

C

B

Fig. 5.25 Artifact from dental fillings. (A) Sagittal reconstructed CT image shows an apparent break in the posterior body of C2 (long arrow). Note the streak artifacts from dental fillings in the horizontal plane (short arrows). (B) Axial image through same region shows no fracture. Note the streak artifacts. (C) Lateral radiograph shows no fracture. Note the normal soft tissues.

Fractures that are oriented in the horizontal plane may not always be demonstrated by CT [43]. This occurs most commonly with fractures of the dens or body of C2 (Fig. 5.29). To overcome this pitfall, it may be necessary to tilt the gantry to bring the fracture out of the plane of the scan. In some cases, it is necessary to resort to radiography to identify the fracture.

Myelography Myelography was used extensively in the past in the evaluation of acute spinal injuries to determine blockage of flow of cerebrospinal fluid caused by bone fragments in the vertebral canal or disc herniation (Figs. 5.30 and 5.31). Use of MR imaging has generally superseded myelography for this purpose. However,

66

CT myelography is useful for showing extradural lesions such as herniated intervertebral discs associated with acute skeletal injury (Figs. 5.20 and 5.21). The diagnosis of acute traumatic dural tears (Fig. 5.32) and the assessment of nerve root avulsions (Fig. 5.33; also see Fig. 5.22) [35,44,45] are other indications. In addition, CT myelography is useful for evaluating posttraumatic cystic myelopathy and the development of syringomyelia whenever MR cannot be performed [36]. Water-soluble contrast can be introduced into the subarachnoid space either from the lumbar region or from the atlantoaxial region. Nonionic water-soluble contrast is the preferred medium, particularly iohexol or iopamidol because of their limited side effects. The introduction of water-soluble contrast

5 Imaging II: radiography, CT, and myelography

A

B

Fig. 5.26 Partial volume averaging effect. (A) Axial image shows a lucent line (arrow) through the lateral aspect of the body of C2. This really represents the junction between C1, which is lying posterior, and C2. (B) Scout view shows the orientation of the scan. Slice portrayed in A is through the region shown by the two arrows and line.

A

B

Fig. 5.27 Jefferson fracture of C1 shown on one CT image only. (A) Lateral radiograph shows fractures in the anterior arch (short arrow) and posterior arch (long arrow). (B) The CT image shows only the anterior arch fracture (arrow). Additional images failed to show the posterior arch fracture.

A

B

Fig. 5.28 Pseudofracture due to partial volume averaging effect. (A) Scout radiograph from a thoracolumbar CT shows severe thoracic and lumbar scoliosis. The horizontal line shows the location of the image shown in B. (B) An axial CT image that shows partial volume averaging displaying portions of L1 and L2 simultaneously. The border between the two vertebrae could be misinterpreted as representing a fracture.

67

5 Imaging II: radiography, CT, and myelography

A

68

B

C

D

E

A

B

Fig. 5.29 Horizontal fractures in plane of scans. (A) Lateral radiograph shows a horizontal fracture of the dens (arrow). (B) The CT image fails to show the fracture. (C) Lateral radiograph shows pars defects of L3 (arrow). (D) The CT image fails to show the defects. (E) Scout view shows the pars defects (arrow) in the plane of the scan (solid white line).

Fig. 5.30 Myelogram showing blockage of flow of contrast. (A) Frontal radiograph shows a shearing fracture–dislocation of T12 on L1. (B) Myelogram image shows complete obstruction (*) to flow of contrast just below the injury.

5 Imaging II: radiography, CT, and myelography

A

A

D

Fig. 5.31 Myelogram showing blockage of flow of contrast. The patient suffered a fracture–dislocation involving L3 and L4. The contrast column ends at the L2 disc space (arrows) because of hematoma and debris in the vertebral canal. (A) Frontal view. (B) Lateral view.

B

B

C

Fig. 5.32 Traumatic dural tear. (A) Coronal reconstructed CT image shows a shearing fracture dislocation at L2–L3. (B) Coronal reconstructed image of the CT myelogram shows contrast extravasated through the fracture (arrow). (C,D) Axial images show contrast extravasation on the right (arrows).

69

5 Imaging II: radiography, CT, and myelography

A

is generally followed by CT examination (Figs. 5.20 and 5.21). Air has also been used as a contrast medium [45]. Pay and associates found that air myelography was useful for evaluating cervical trauma without bony deformity as well as for delin-

References 1.

2.

3.

4.

5.

6.

7.

8.

70

Fig. 5.33 Myelogram showing cervical nerve root avulsion. Radiographs show the extravasated contrast along the C7 nerve root sheath (arrows).

B

Freemyer B, Knopp R, Piche J, et al. Comparison of five-view and threeview cervical spine series in the evaluation of patients with cervical trauma. Ann Emerg Med 1989; 18:818–821. MacDonald RL, Schwartz ML, Mirich D, et al. Diagnosis of cervical spine injury in motor vehicle crash victims: how many X-rays are enough? J Trauma 1990; 30:392–397. Murphey MD. Trauma oblique cervical spine radiographs. Ann Emerg Med 1993;22:728–730. Turetsky DB, Vines FS, Clayman DA, et al. Technique and use of supine oblique views in acute cervical spine trauma. Ann Emerg Med 1993; 22:685–689. Daffner RH. Cervical radiography in the emergency department: who, when, how extensive? Emerg Radiol 1995; 2:261–263. Daffner RH, Hackney DB. ACR Appropriateness Criteria on suspected spine trauma. J Am Coll Radiol 2007; 4:762–775. Gehweiler JA Jr., Osborne RL Jr., Becker RF. The Radiology of Vertebral Trauma. Philadelphia, PA: WB Saunders, 1980. Shaffer MA, Doris PE. Limitation of the cross-table lateral view in detecting cervical spine injuries: a retrospective

9.

10.

11.

12.

13.

14.

15.

16.

17.

eating the thoracic spinal cord in the lateral projection and demonstrating cord atrophy in the postacute state. The protocol that they developed can be used with either air or water-soluble contrast media if MR imaging cannot be performed [45].

analysis. Ann Emerg Med 1981; 10:508–513. Daffner RH. Cervical radiography for trauma patients: a time-effective technique? AJR Am J Roentgenol 2000; 175:1309–1311. England AC, Shippel AH, Ray MJ. A simple view for demonstration of fractures of the anterior arch of C-1. AJR Am J Roentgenol 1985;144: 763–764. Abel MS. The exaggerated supine oblique view of the cervical spine. Skeletal Radiol 1982;8:213–219. Bohrer SP, Chen IM, Sayers EG. Cervical spine flexion patterns. Skeletal Radiol 1990;19:521–525. Evans RW. Some observations on whiplash injuries. Neurol Clin 1992; 10:975–997. Spitzer WO, Skovron ML, Salmi LR, et al. Scientific monograph of the Quebec Task Force on WhiplashAssociated Disorders: redefining “whiplash” and its management. Spine 1995;20(8 Suppl):1S–73S. Woodring JH, Lee C. Limitations of cervical radiography in the evaluation of acute cervical trauma. J Trauma 1993;34:32–39. Daffner RH. Cervical helical CT for trauma patients: a time analysis. AJR Am J Roentgenol 2001;177:677–679. Fielding JW, Stillwell WT, Chynn KY, et al. Use of computed tomography for

18.

19.

20.

21.

22.

23.

24.

the diagnosis of atlanto-axial rotatory fixation. J Bone Joint Surg 1978; 60A:1102–1104. Brant-Zawadzki M, Miller EM, Federle MP. CT in the evaluation of spine trauma. AJR Am J Roentgenol 1981; 136:369–375. Handel SF, Lee YY. Computed tomography of spinal fractures. Radiol Clin North Am 1981;19:69–89. Steppé R, Bellemans M, Boven F, et al. The value of computed tomography scanning in elusive fractures of the cervical spine. Skeletal Radiol 1981;6:175–178. Post MJD, Green BA. The use of computed tomography in spinal trauma. Radiol Clin North Am 1983; 21:327–375. Gellad FE, Levine AM, Joslyn JN, et al. Pure thoracolumbar facet dislocation: clinical features and CT appearance. Radiology 1986;161:505–508. Acheson MB, Livingston RR, Richardson ML, et al. High-resolution CT scanning in the evaluation of cervical spine fractures: comparison with plain film examinations. AJR Am J Roentgenol 1987;148:1179–1185. Nuñez DB, Ahmad AA, Coin CG, et al. Clearing of the cervical spine in multiple trauma victims: a time-effective protocol using helical computed tomography. Emerg Radiol 1994;1:273–278.

5 Imaging II: radiography, CT, and myelography

25. El-Khoury GY, Kathol SJ, Daniel WW. Imaging of acute injuries of the cervical spine: value of plain radiography, CT, and MR imaging. AJR Am J Roentgenol 1995;164:43–50. 26. Berne JD, Velmahos GC, El-Tawil Q, et al. Value of complete cervical helical computed tomographic scanning in identifying cervical spine injury in the unevaluable blunt trauma patient with multiple injuries: a prospective study. J Trauma 1999;47:896–903. 27. Lawrason JN, Novelline RA, Rhea JT, et al. Can CT eliminate the initial portable lateral cervical spine radiograph in the multiple trauma patient? A review of 200 cases. Emerg Radiol 2001;8:272–275. 28. Ptak T, Kihiczak D, Lawrason JN. Screening for cervical spine trauma with helical CT: experience with 676 cases. Emerg Radiol 2001;8:315–319. 29. Li AE, Fishman EK. Cervical spine trauma: evaluation by multidetector CT and three-dimensional volume rendering. Emerg Radiol 2003;10:34–39. 30. Yetkin Z, Osborn AG, Giles DS, et al. Uncovertebral and facet joint dislocations in cervical articular pillar fractures: CT evaluation. AJR Am J Roentgenol 1985;6:633–637. 31. Wojcik WG, Edeiken-Monroe BS, Harris JH Jr. Three-dimensional computed tomography in acute cervical

32.

33.

34.

35.

36.

37.

38.

spine trauma: a preliminary report. Skeletal Radiol 1987;16:261–269. Cothren CC, Moore EE, Biffl WL, et al. Cervical spine fracture patterns predictive of blunt vertebral artery injury. J Trauma 2003;55:811–813. Miller PR, Fabian TC, Croce MA, et al. Prospective screening for blunt cerebrovascular injuries: analysis of diagnostic modalities and outcomes. Ann Surg 2002;236:386–393. Biffl WL, Egglin T, Benedetto B, Gibbs F, Cioffi WG. Sixteen-slice computed tomographic angiography is a reliable noninvasive screening test for clinically significant blunt cerebrovascular injuries. J Trauma 2006;60:745–751. Petras AF, Sobel DF, Mani JR, et al. CT myelography in cervical nerve root avulsion. J Comput Assist Tomogr 1985;9:275–279. Seibert CE, Dreisbach JN, Swanson WB, et al. Progressive post-traumatic cystic myelopathy: neuroradiologic evaluation. AJR Am J Roentgenol 1981;136:1161–1165. Blacksin MF, Lee HJ. Frequency and significance of fractures of the upper cervical spine detected by CT in patients with severe neck trauma. AJR Am J Roentgenol 1995;165:1201–1204. Link TM, Schuierer G, Hufendiek A, et al. Substantial head trauma:

39.

40.

41.

42.

43.

44.

45.

value of routine CT examination of the cervicocranium. Radiology 1995;196:741–745. Kirshenbaum KJ, Nadimpalli SR, Fantus R, et al. Unsuspected cervical spine fractures associated with significant head trauma: role of CT. J Emerg Med 1990;8:183–198. Clayman DA, Sykes CH, Vines FS. Occipital condyle fractures: clinical presentation and radiologic detection. AJNR Am J Neuroradiol 1994;15: 1309–1315. Kowalski HM, Cohen WA, Cooper P, et al. Pitfalls in the CT diagnosis of atlantoaxial rotary subluxation. AJNR Am J Neuroradiol 1987;8:697–702. Woodring JH, Lee C. The role and limitations of computed tomographic scanning in the evaluation of cervical trauma. J Trauma 1992;33:698–708. Daffner RH, Sciulli RL, Rodriguez A, Protetch J. Imaging for evaluation of suspected cervical spine trauma: a 2-year analysis. Injury 2006;37: 652–658. Morris RE, Hasso AN, Thompson JR, et al. Traumatic dural tears: CT diagnosis using metrizamide. Radiology 1984;152:443–446. Pay NT, George AE, Benjamin MV, et al. Positive and negative contrast myelography in spinal trauma. Radiology 1977;123:103–111.

71

Chapter

6

Imaging of vertebral trauma III: magnetic resonance imaging Bryan S. Smith Richard H. Daffner

Since the mid 1970s, MR imaging has reached the stage of a mature technology. It is indispensable for the diagnosis of a broad spectrum of vertebral and spinal cord pathology as well as the sequelae of trauma. While radiography and CT can reveal important detail about fractures and abnormal alignment, it became clear from the outset that MR imaging was unique in its depiction of intrinsic spinal cord injury [1–5]. As technical advances occurred, MR imaging became recognized for its value in assessing vertebral fractures and in demonstrating ligamentous disruption [6–8]. Spinal cord compression by bone fragments, disc herniation, and epidural or subdural hematomas could also be diagnosed [9]. Hemorrhagic contusion within the cord could be observed with serial examinations, which could reveal the onset of posttraumatic progressive myelopathy [10]. Furthermore, refinement of MR angiography (MRA) has provided adjunctive information about vascular structures (e.g., vertebral artery dissection or occlusion) [11,12]. The information gleaned from MR imaging, especially when supplemented by CT, has succeeded in significantly reducing the need for myelography [13]. The risks associated with myelography are increased in the setting of acute trauma, and it is now reserved for situations in which MR imaging is contraindicated, where the technical challenges of the MR examination result in suboptimal image quality, or for parts of the world where MR imaging is unavailable.

Technical considerations A discussion of technique requires both an understanding of the logistics involved in the scanning of acutely injured patients in a safe manner and knowledge of the appropriate pulse sequences to achieve diagnostic efficacy. Optimal patient positioning differs slightly depending on whether the suspected injury is in the cervical vertebral column or the thoracolumbar region. A patient with a suspected cervical injury arrives on a wooden backboard for stabilization, and patient and board are placed on the scanner table. The patient is carefully positioned in a Helmholtz surface coil, which provides sufficient immobilization to ensure a safe procedure. At our institution, we do not maintain traction, other than the cervical collar, during the examination.

72

In the thoracolumbar region, the patient most often remains on the backboard, which results in a diminished signal-to-noise ratio and a consequent loss of imaging detail. Close consultation with the referring physician helps to permit removal of the backboard whenever possible. Assurance of the patient’s safety during MR imaging is of fundamental importance. The MR imaging suite should have the capacity for video observation of the patient. Verbal interaction is desirable for patients able to speak. For those who cannot communicate because of anesthesia, sedation, or the nature of their injury, physiologic monitoring is imperative, because hemodynamic stability must be maintained. The following recommendations have been made by the Safety Committee of the International Society of Magnetic Resonance [14,15] as well as by the American College of Radiology and the American Society of Neuroradiology [16–21]. Monitoring of blood oxygenation by a pulse oximeter is strongly recommended for sedated patients [15]. Electrocardiographic and blood pressure monitoring should also be performed, and the readings should be displayed within the control room. It is also imperative to prevent thermal injury on monitored patients [16–18,21]. Care must be taken to avoid coiled or looped wires in which electrical current can be induced. This is particularly important in patients with sensory deficits, who may not feel the heat. Patients with a known history of metalworking, surgical prosthesis implantation, or shrapnel/bullet wound should be carefully screened with appropriate radiographs before undergoing an MR examination. Any potential retained or implanted metallic object must be either proved MRI compatible utilizing available resources (such as the guidelines from the Institute for Magnetic Resonance Safety, Education, and Research [21]) or visualized, utilizing conventional radiography, remote from vital structures. Ferromagnetic metallic objects in close association to sensitive structures can have disastrous complications when exposed to the strong magnetic fields of the MRI magnet due to local motion or thermal effects.

Imaging parameters Important advances in the development of MR imaging pulse sequences have been made that have been particularly valuable

6 Imaging III: magnetic resonance imaging

Table 6.1 Specific parameters used for MR imaging in vertebral trauma at the Allegheny General Hospital

Parameter

Sagittal spin echo

Sagittal turbo spin echo

Axial spin echo

Axial gradient echo

Sagittal STIR

Repetition time (ms)

550

3500

550

527

4000

Time to echo (ms)

15

21 103

15

15

89

Field of view (cm)

250 (cervical); 280 (thoracolumbar)

250 (cervical); 280 (thoracolumbar)

230

230

230 (cervical); 340 (thoracolumbar)

Slice thickness (mm)

4

3

5

4

3 (cervical); 4 (thoracolumbar)

Acquisitions

1

3

2

3

2

Matrix

192 × 256

192 × 256

192 × 256

Flip angle (°)

192 × 256

75 × 256

15

150

STIR, short-tau inversion-recovery.

in enhancing the speed with which a given examination can be completed [20]. However, spin echo sequences with a short repetition time (TR) have remained essential components of the examination, since alignment of the vertebral axis in the midsagittal plane can be effectively assessed and the external morphology of the spinal cord is readily depicted. These sequences are also useful in demonstrating alteration of vertebral body marrow signal caused by compression fractures and can reveal signal abnormality indicative of hemorrhage, particularly in the subacute time frame. Acute hemorrhage, however, is often best demonstrated on either T2*-weighted gradient echo sequences or fast (or turbo) spin echo T2weighted sequences. In these sequences, there is selective acquisition of high-contrast raw data in K-space [22]. In general, a 50–70% decrease in time expenditure results, with a consequent alleviation of motion artifact. Fast spin echo sequences are slightly less sensitive to the magnetic susceptibility effects of acute hemorrhage but at the same time help to minimize the artifact that can occur if metallic fixation devices are present. To improve sensitivity to hemorrhage, obtaining a gradient echo sequence in a complementary (usually axial) plane is a practical option. Short tau inversion recovery (STIR) is a fat suppression technique where the signal of fat is zero. In comparison with a conventional spin echo, fat signal is darkened [6]. Because body fluids have both a long T1 and a long T2, STIR offers extremely sensitive detection of edema. The specific parameters used at our institution in the MR imaging of patients who have suffered vertebral trauma are given in Table 6.1. In patients who are also evaluated for possible vascular injury using MRA, the parameters given in Table 6.2 are used to generate a three-dimensional time-of-flight examination. The assessment is generally performed in the head coil with dual overlapping slabs to achieve coverage of both the cervical and intracranial vessels. A presaturation band placed over the superior sagittal sinus effectively eliminates signal from major venous structures.

Table 6.2 Parameters used to evaulate possible vascular injury using MR angiography Parameter

MR angiography

Repetition time (ms)

38

Time to echo (ms)

7 (minimum)

Field of view (cm)

200

Slice thickness (mm)

1 (64 three-dimensional partitions)

Acquisitions

1

Matrix

192 × 256

Flip angle (°)

20

Pathologic aspects Acute spinal cord injury The internal structure of the spinal cord is depicted by MR imaging to an extent not previously possible. With regard to cord injury following trauma, a spectrum of abnormalities has been described, including swelling, edema, hemorrhagic contusion, and transection [1,23–25]. Cord swelling implies enlargement of the contour of the spinal cord without necessarily alteration of the internal signal intensity, and this finding is generally best appreciated on T1-weighted images. By contrast, cord edema results in prolongation of both T1 and T2, and is most reliably identified as an area of hyperintensity on long TR spin echo or fast spin echo sequences; as such, it resembles edema anywhere in the body. Figure 6.1 shows edema in the cord as a result of a dens fracture. Many serious cord injuries are complicated by the occurrence of intramedullary hemorrhage. Hemorrhage is associated with signal abnormalities on both T1- and T2-weighted images, but for prompt recognition in the acute period, T2weighted or T2*-weighted (gradient recalled echo) images are essential. Hypointensity on these images indicates the magnetic susceptibility effect of deoxyhemoglobin, a constituent of

73

6 Imaging III: magnetic resonance imaging

A

B

C

Fig. 6.1 Cord edema in a patient with a dens fracture and central cord syndrome. (A) Sagittal reconstructed CT image shows a dens fracture (arrow) with retrolisthesis of the dens fragment. (B) Sagittal T2-weighted MR image shows faint increase in signal within the spinal cord immediately behind the fracture (*). (C) Sagittal STIR image shows the edema (*) to advantage. Note the soft tissue changes anterior and posterior to the vertebral column. These findings would be identical in the thoracic and lumbar regions.

A

B

an acute hematoma. Hyperintensity represents accompanying edema. Frequently, the findings coexist in a given patient. In Fig. 6.2, another dens fracture, both intramedullary hemorrhage and surrounding edema are present on the accompanying T2-weighted image. Once the T1-shortening effect of methemoglobin formation has occurred (after 24 to 36 hours), T1-weighted images are also useful in the depiction of hemorrhagic contusion in the spinal cord. This often is the period in which the patient, having been medically or mechanically stabilized, first presents for imaging. Sequences obtained using short TR spin echo parameters show

74

Fig. 6.2 Acute cord hemorrhage in a patient with a dens fracture. (A) T2-weighted and (B) STIR sagittal MR images show swelling of the spinal cord with a central zone of decreased signal (*) with a surrounding zone of increased signal representing edema.

hyperintensity in the injured cord parenchyma. Such a case is illustrated in Fig. 6.3, in which anterior fixation of the traumatized cervical column using surgical plates and screws has been performed. The use of titanium instrumentation has resulted in relatively minimal artifact compared with that seen with other metallic devices. Severe injuries may result in spinal cord transection. Complete discontinuity of the cord may occur at any level as a result of fracture–dislocation from a variety of mechanisms (Fig. 6.4). On rare occasions the cord will retract leaving an “empty” space between the severed ends (Fig. 6.4C,D).

6 Imaging III: magnetic resonance imaging

A

A

D

Fig. 6.3 Subacute cord hemorrhage. The MR imaging was not obtained immediately after the injury, only after surgical stabilization with plates and screws. (A) Sagittal T1-weighted image shows mottled areas of increased signal intensity (arrows). (B) Axial T1-weighted image shows similar findings (*).

B

B

E

C

Fig. 6.4 Cord transection. (A) Extension fracture–dislocation at C4–C5. Sagittal gradient echo MR image shows the transection (arrow). (B) Sagittal STIR MR image shows transection (arrow) at T10–T11 caused by a rotary injury. (C) Cord transection with hematoma in a victim of child abuse. Sagittal T2-weighted fat saturated image shows the severed cord ends (arrows) and a large intervening hematoma (*). (D) Sagittal STIR MR image shows cord transection at T10–T11 with retraction of the torn ends leaving a gap (*) (“empty cord” sign). (E) Axial STIR MR image shows no cord (*) in the vertebral canal. Note the extensive surrounding edema.

75

6 Imaging III: magnetic resonance imaging

Complete neurologic deficit occurs not only in these transactions but in almost all instances of intramedullary hemorrhage documented by MR imaging [1,2,9,26–28]. In patients with incomplete spinal cord syndromes, MR imaging can reveal cord injuries that may be unaccompanied by any significant CT or radiographic evidence of fracture or

A

B

A

B

dislocation. This is the concept of spinal cord injury without radiologic abnormality (SCIWORA) (Fig. 6.5) [29,30]. Use of CT imaging has significantly reduced the percentage of cases that fall into this category. I prefer to refer to those as spinal cord injury with minimal radiographic abnormalities (SCIMRA) (Fig. 6.6).

Fig. 6.5 Spinal cord injury without radiographic abnormalities (SCIWORA) in a child quadriparetic following a motor vehicle crash. (A) Lateral cervical radiograph is normal. (B) Sagittal T2-weighted MR image shows central cord hemorrhage (arrow) and an epidural hematoma (*) anterior to the cord. (Courtesy of Leonard Swischuk, MD, Galveston TX, USA.)

C

Fig. 6.6 Spinal cord injury with minimal radiographic abnormalities (SCIMRA). (A) Sagittal reconstructed CT image shows a spinous process fracture of C4 (arrow). There are mild degenerative changes. (B) The T1-weighted sagittal MR image shows mottling of the signal in the spinal cord (arrows) representing acute hemorrhage. (C) The T2-weighted sagittal MR image shows a zone of low-signal hemorrhage (large arrow) surrounded by highsignal edema. Note rupture of the anterior longitudinal ligament at C3–C4 (small arrow) as a manifestation of a hyperextension injury not apparent on the CT scan.

76

6 Imaging III: magnetic resonance imaging

Chronic spinal cord injury Sequential MR imaging examinations reveal important information about the evolution of chronic cord injury. Specifically, MR imaging can evaluate for posttraumatic cyst formation (cystic myelopathy, syringomyelia), which may be a source of progressive neurologic deficit. Indeed, MR imaging has an important role in differentiating myelomalacia from cystic myelopathy [31]. The latter complication may be amenable to surgical decompression, and intraoperative ultrasound may be a valuable adjunct to preoperative MR imaging in further defining internal septations within the posttraumatic cyst. The appearance of such lesions varies from small cysts with surrounding myelomalacia (Fig. 6.7) to more extensive cysts (Fig. 6.8). Myelomalacia has a similar appearance to cord edema, consisting of decreased signal intensity on T1-weighted images and increased signal intensity on T2-weighted and inversion recovery sequences. However, in cord edema, the signal changes are ill defined, whereas in myelomalacia the areas of abnormal signal are well defined.

Extradural compressive lesions While MR imaging provides unique information about cord hemorrhage and other intrinsic injury, it is also essential to evaluate the possibility of extradural cord or nerve root compression. The diagnosis of such compression, which was previously achieved by myelography, offers the best chance for neurologic improvement if expeditious surgical intervention ensues. Important causes of cord compression include herniated intervertebral disc, retropulsed bone fragments, and epidural

A

B

hematoma. It is sometimes difficult to distinguish these entities on the basis of signal intensity characteristics alone. Analysis of the morphology and location of the compressive lesion is particularly helpful in establishing the diagnosis. As always, the MR images should be interpreted in light of the CT findings. Disc herniation is usually contiguous with the vertebral interspace and eccentric anteriorly within the vertebral canal (Fig. 6.9). If migration of the fragment occurs, it is almost always in a cephalocaudal direction and not lateral. The signal intensity of an extruded fragment generally approximates that of the parent disc, but the herniation may be of higher intensity, particularly on T1-weighted images. Disc herniations are typically associated with partial or complete tears of the posterior longitudinal ligament. A bone fragment can generally be identified by the sharp linear hypointensity associated with its cortical margin and by indentation of the ventral aspect of the subarachnoid space (Fig. 6.10). Again, reference to the accompanying CT study will show the true nature of the injury. It is not unusual to have an associated epidural hematoma. In contrast to bone fragments, a hematoma tends to be circumferential rather than only ventral on axial sequences and tends to have a greater longitudinal extent on sagittal images. In some cases, these entitities coexist and are likely to be associated with marked deformity of the adjacent vertebral body (Fig. 6.10B).

Bony and ligamentous injury Although MR imaging can be useful in the identification of retropulsed bone fragments, it cannot replace CT and

C

Fig. 6.7 Chronic posttraumatic cord changes. (A) Compression deformity of C5 with associated cervical kyphosis as the result of previous trauma. Abnormal hyperintensity in the spinal cord extends from C1 through C5 (arrows) on this sagittal T2-weighted image. (B) Sagittal T1-weighted image distinguishes a focal cord cyst at the C4–C5 level (arrows) from the more extensive zone of myelomalacia seen in A. (C) Unlike congenital hydromyelia, this posttraumatic cyst is located eccentrically to the right within the spinal cord (arrow) on this axial image.

77

6 Imaging III: magnetic resonance imaging

A

D

A

B

C

Fig. 6.8 Posttraumatic syringomyelia in a patient with a dens fracture. (A) Sagittal reconstructed CT image shows the dens fracture (arrow). (B,C) Sagittal T1-weighted (B) and STIR (C) images obtained three months later show the cyst in the central cord (*) extending from C2 through C7. (D) Axial gradient echo image shows the cyst to be well defined and centrally located (*). Compare with Fig. 6.7C.

B

Fig. 6.9. Disc herniation (arrows) secondary to trauma. (A) Sagittal STIR image. Note slight anterolisthesis of C6 on C7 and prevertebral hemorrhage. (B) Axial STIR image.

78

6 Imaging III: magnetic resonance imaging

A

A

Fig. 6.10 Burst fracture of L1. (A) Sagittal reconstructed CT image shows compression of L1 as well as a retropulsed fragment of bone (arrow) in the vertebral canal. (B) The T2-weighted sagittal MR image shows similar findings (large arrow) as well as an epidural hematoma (small arrow).

B

B

radiographs in the overall evaluation of fractures. Because of its superior depiction of cortical bone, CT correlation is particularly important in evaluating neural arch fractures. The clarity of CT relative to MR imaging in such a case is illustrated in Fig. 6.11. In the vertebral body, MR imaging is sensitive to the alteration of signal that occurs in cancellous bone following acute injury. Specifically, on T1-weighted images, the normal hyperintensity of fatty marrow is replaced by the lower signal resulting from intra-osseous edema and hemorrhage. An example of abnormally hypointense vertebral signal is illustrated in Fig. 6.12, a burst fracture of T11. The T2-weighted images have more variable sensitivity in the diagnosis of marrow injury. The STIR sequences represent an improvement in this realm. Specifically, fractures are seen as areas of abnormal hyperintensity on these images [6].

Fig. 6.11 Burst fracture of L1 (same patient as in Fig. 6.10). (A) Axial CT image shows bony detail of the fracture as well as of the retropulsed fragments in the vertebral canal. (B) Axial STIR MR image shows the bone fragments in the canal (arrows), but to a degree considerably less than the CT images. The details of the fracture are not well defined.

These signal changes within the marrow allow us to identify fractures that are not apparent on CT or radiographic examinations (Fig. 6.13). In general in the clinical setting of acute trauma, there is often confusion about the possibility of underlying malignant disease when an abnormal vertebral body signal is encountered. Such uncertainty can occur when an abnormal vertebral body signal is encountered in older patients [32]. Again, CT correlation is helpful; signs supporting a benign etiology (trauma or osteoporosis) include identification of multiple well-defined interconnecting fracture lines (“jigsaw puzzle” sign), the presence of a retropulsed fragment, and an intravertebral vacuum phenomenon [32,33]. A thin paravertebral soft tissue mass with tapering ends also tends to exclude a malignant process such as metastasis or myeloma, for which a prominent focal mass is more common. On MR images, fractures of benign

79

6 Imaging III: magnetic resonance imaging

A

A

D

80

Fig. 6.12 Signal changes in bone resulting from trauma. (A) Sagittal reconstructed CT image shows compression of T3 (*). There is minimal loss of height of T4 and T5. (B) Sagittal STIR MR image shows the compression deformity of T3. There is increased signal intensity in T2, T4, and T5, indicating injuries to those levels as well. Note the posterior epidural hematoma (arrows) as well as hemorrhage in the interspinous region (*) between T3 and T4. Imaging with MR is excellent for identifying multiple levels of injury that would not be apparent on radiographs or CT.

B

B

C

Fig. 6.13 Chance-type fractures in two patients. (A) Sagittal reconstructed CT image shows compression of the anterior portion of L1 and an increase in height posteriorly. (B,C) The T1-weighted (B) and STIR sagittal (C) MR images show signal changes of a fracture in L1 (arrow). However, there are similar changes in T12 (*), which appears normal on the CT. (D) A T2-weighted sagittal MR image in another patient shows a Chance-type fracture of L4 and abnormal signal in L5 (arrow). The CT (not shown) demonstrated abnormalities in L4 only.

6 Imaging III: magnetic resonance imaging

A

B

Fig. 6.14 Ligament ruptures. (A) The anterior longitudinal ligament is stripped away from C6 and is torn (arrow) in a patient with unilateral facet lock. Note the prevertebral edema and cord swelling. (B) Anterior and posterior longitudinal ligament tears at C5–C6 (white arrows) secondary to an extension injury. Note the wide disc space and the significant edema in the spinal cord (black arrow).

origin usually have linear signal abnormalities; fractures resulting from tumor or infection have globular signal changes. In diagnosing ligament injuries, MR imaging in the sagittal plane is analogous to radiographs in showing the characteristic malalignments of flexion and extension injuries. Such findings aid the assessment of mechanical stability as defined in Chapter 10. Important constituents of the stable vertebral column include the anterior and posterior longitudinal ligaments [34], the ligamenta flava, and the interspinous ligament. Ligamentous disruption is identifiable on MR imaging by discontinuity or nonvisualization of the normal band of low signal intensity that represents the ligament (Fig. 6.14) [34,35]. This discontinuity may be accompanied by hyperintensity on T2-weighted images in the intervertebral or paravertebral soft tissues. In one series, 30% of thoracolumbar burst fractures were associated with posterior ligamentous disruption [10]. In most instances there will be significant edema in the deep or superficial soft tissues. By virtue of its capacity to examine long vertebral segments, MR imaging can show coexistent ligamentous injuries at noncontiguous levels that were unsuspected from either CT or radiographs [13,34,35]. Remember that as many as 25% of patients with a vertebral fracture will have multiple noncontiguous injuries. Figure 6.15 shows disruption of the ossified anterior longitudinal ligament at T10–T11 in a patient with

Fig. 6.15 Multiple noncontiguous injuries. Sagittal T2-weighted MR image shows disruption of the ossified anterior longitudinal ligament at T10–T11 (arrow). Note increased signal intensity in the bodies of T1, T2, T8, and T9.

81

6 Imaging III: magnetic resonance imaging

A

C

B

D

idiopathic skeletal hyperostosis (DISH). However, there were also abnormal signal changes at T1, T2, T8, and T9 from injuries that were not shown on the CT examination. The facet joint capsules are also important contributors to stability, and sagittal MR images are particularly advantageous in demonstrating facet locking or perching (Fig. 6.16) [36]. As with CT, the confusion that is often created by overlapping structures on radiographs is generally eliminated by MR imaging. Facet locking frequently coexists with disc herniation [36], and recognition of this abnormality is imperative for appropriate surgical management. Facet fracture or dislocation can also result in a vertebral artery injury (Fig. 6.17). Absence of the normal flow void in the vertebral

82

Fig. 6.16 Facet lock: unilateral (A,B) and bilateral (C,D). (A) Sagittal reconstructed CT image shows locking of the facet of C6 on the fractured pillar of C7 (arrow). (B) Sagittal STIR MR image shows the facet lock (arrow) as well as considerable paraspinal edema. (C) Sagittal reconstructed CT image shows anterolisthesis of C6 on C7. Note the canal narrowing. The patient has had a previous anterior fusion between C5 and C6. (D) Sagittal STIR MR image shows cord hemorrhage (arrow) and surrounding edema in the cord. There is also significant prevertebral soft tissue edema (*).

artery on standard spin echo images can be confirmed by MR angiography. Patients with degenerative spondylosis, DISH, or ankylosing spondylitis suffer neurologic injuries from even mild hyperextension without fracture [25], and MR imaging is essential in identifying these injuries [7,35]. In ankylosing spondylitis or DISH, hyperextension disrupts the ossified anterior longitudinal ligament and results in pseudoarthrosis at the discovertebral junction (“broken DISH”). Use of MR imaging shows ligamentous disruption and irregularity of the vertebral endplates (Fig. 6.18). The presence of adjacent edema supports the diagnosis of an acute injury. These findings may be difficult to detect on radiographs because of osteoporosis,

6 Imaging III: magnetic resonance imaging

B

A

C

Fig. 6.17 Vertebral artery injury in a 15 year old who suffered unilateral locked facet at C3–C4. (A) Lateral cervical radiograph shows C3 and the vertebrae above rotated. There is slight anterolisthesis of C3 on C4. (B) Sagittal STIR MR image shows central cord hemorrhage at C3 (arrow). Note the surrounding cord edema. (C,D) The T1-weighted axial image with fat saturation (C) and axial STIR image (D) show absence of the normal flow void in the right vertebral artery (arrows).

D

osteophytes, and extensive bony fusion [37]. However, sagittal reconstructed CT images are much more effective in demonstrating these abnormalities. Occipito-atlantal dislocation is an especially severe form of ligamentous disruption. The consequences are often fatal, but survival is possible when a pure distraction mechanism predominates (most commonly in children) [38]. The surviving patients usually have subtle radiographic findings (see Chapter 5). In Fig. 6.19, MR imaging demonstrates disarticulation of the occipito-atlanto-axial complex with intramedullary hemorrhage in the high cervical cord and massive soft tissue swelling. The ability of MR imaging to demonstrate the craniovertebral ligaments has led to hope that patients suffering from the so-called “whiplash syndrome” could have this condition more easily diagnosed. The presence of surrounding edema, as well as frank ligament disruption, is ample evidence of such an injury. However, for patients with the history of whiplash who

develop chronic neck pain, the results have been disappointing at best, and controversial [39–43]. Finally, what is the role of MRI in “clearing” the spine of trauma victims? There is controversy as to whether CT can be used instead of MR imaging for that purpose. Certainly MR is useful for showing whether or not there is ligament damage that would result in instability (see Chapter 10). However, a number of excellent studies have found that CT examinations with reconstructed sagittal and coronal images was just as effective as MRI for ruling out an unstable injury [44–50]. The authors acknowledged that MRI would identify microtrabecular fractures, intraspinous ligament injuries, cord signal abnormalities, and epidural hematomas. However, these investigators found that in patients who are neurologically intact none of these abnormalities resulted in a change in management [48]. In our institution, we still perform an MR examination on comatose patients, with normal CT examinations of their spine after 48 hours.

83

6 Imaging III: magnetic resonance imaging

A

C

84

B

D

Fig. 6.18 Extension injuries in two patients with diffuse idiopathic skeletal hyperostosis (“broken DISH”). (A,B) Cord contusion in a patient with DISH central cord syndrome. (A) Sagittal reconstructed CT image shows ossification of the posterior longitudinal ligament (black arrow) narrowing the vertebral canal. There is disruption of the ossified anterior longitudinal ligament at C3–C4 and C6 (white arrows). (B) Sagittal STIR MR image shows increased signal within the pinched spinal cord (arrow) as well as precervical hemorrhage (*). (C) Sagittal reconstruction in the same patient as in Fig. 6.15, with injury at T10–T11 shows the disruption of the ossified anterior longitudinal ligament (arrow). (D) Sagittal STIR MR image shows increased signal in the T10–T11 disc space (arrow).

6 Imaging III: magnetic resonance imaging

A

B

Fig. 6.19 Occipito-atlantal disruption. (A) Coronal reconstructed CT image shows widening of the occipito-atlantal joints (*). (B) Sagittal STIR MR image shows rupture of the apical ligaments (black arrow), cord hemorrhage (white arrow), and massive prevertebral hematoma (*). (C) Axial CT image through C1–C2 shows an epidural hematoma anteriorly (*).

C

References 1.

2.

3.

Kulkarni MV, McArdle CB, Kopanicky D, et al. Acute spinal cord injury: MR imaging at 1.5 T. Radiology 1987; 164:837–843. Kulkarni MV, Bondurant FJ, Rose SL, et al. 1.5 Tesla magnetic resonance imaging of acute spinal trauma. Radiographics 1988;8:1059–1082. Benzel EC, Hart BL, Ball PA, et al. Magnetic resonance imaging for the evaluation of patients with occult cervical spine injury. J Neurosurg 1996; 85:824–829.

4.

5.

6.

Vaccaro AR, Kreidl KO, Pan W, Cotler JM, Schweitzer ME. Usefulness of MRI in isolated upper cervical spine fractures in adults. J Spinal Disord 1998; 11:289–293. Saifuddin A. MRI of acute spinal trauma. Skeletal Radiol 2001; 30:237–246. Meyers SP, Wiener SN. Magnetic resonance imaging features of fractures using the short tau inversion recovery sequence: correlation with radiographic findings. Skeletal Radiol 1991;20: 499–507.

7.

8.

9.

Goldberg AL, Rothfus WE, Deeb ZL, et al. Hyperextension injuries of the cervical spine: magnetic resonance findings. Skeletal Radiol 1989;18: 283–288. Emery SE, Pathria MN, Wilber RG, Masaryk T, Bohlman HH. Magnetic resonance imaging of posttraumatic spinal ligament injury. J Spinal Disord 1989;2:229–233. Flanders AE, Tartaglino LM, Friedman DP, et al. Magnetic resonance imaging in acute spinal injury. Semin Roentgenol 1992;27:271–298.

85

6 Imaging III: magnetic resonance imaging

10. Petersilge CA, Pathria MN, Emery SE, et al. Thoracolumbar burst fractures: evaluation with MR imaging. Radiology 1995;194:49–54. 11. Willis BK, Griener F, Orrison WW, et al. The incidence of vertebral artery injury after midcervical spine fracture or dislocation. Neurosurgery 1994; 34:435–442. 12. Cothren CC, Moore EE, Biffl WL, et al. Cervical spine fracture patterns predictive of blunt vertebral arterial injury. J Trauma 2003;55:811–813. 13. Kalfas I, Wilberger JE, Goldberg AL, et al. Magnetic resonance imaging in acute spine cord trauma. Neurosurgery 1988;23:295–299. 14. Kanal E, Shellock FG. Policies, guidelines, and recommendations for MR imaging safety and patient management. J Magn Reson Imaging 1992;2:247–248. 15. Kanal E, Shellock FG. Patient monitoring during clinical MR imaging. Radiology 1992;185:623–629. 16. Shellock FG. Magnetic Resonance Procedures: Health Effects and Safety. Boca Raton, FL: CRC Press, 2001. 17. Shellock FG. Reference Manual for Magnetic Resonance Safety, Implants, and Devices, 2005 edn. Los Angeles, CA: Biomedical Research Publishing, 2005. 18. Shellock FG, Crues JV. MR procedures: biologic effects, safety, and patient care. Radiology 2004;232:635–652. 19. Kanal E, Barkovich A, Bell C, et al. ACR Guidance Document for Safe MR Practices. AJR Am J Roentgenol 2007; 188:1447–1474. 20. American College of Radiology. ACR-ASNR Practice Guideline for the Performance of Magnetic Resonance Imaging (MRI) of the Adult Spine. Reston, VA: American College of Radiology, 2006. 21. Institute for Magnetic Resonance Safety, Education, and Research. Guidelines to Prevent Excessive Heating and Burns Associated with Magnetic Resonance Procedures. www.imrser.org, 2010. 22. Sze G, Meriam M, Oshio K, et al. Fast spin-echo imaging in the evaluation of intradural disease of the spine. AJNR Am J Neuroradiol 1992;13:1383–1392. 23. Hackney DB, Asato R, Joseph PM, et al. Hemorrhage and edema in acute spinal cord compression: demonstration by MR imaging. Radiology 1986;161:387–390.

86

24. Quencer RM, Sheldon JJ, Post MJD, et al. Magnetic resonance imaging of the chronically injured cervical spinal cord. AJNR Am J Neuroradiol 1986; 7:457–464. 25. Regenbogen VS, Rogers LF, Atlas SW, et al. Cervical spinal cord injuries in patients with cervical spondylosis. AJR Am J Roentgenol 1986;146:277–284. 26. Davis PC, Reisner A, Hudgins PA, et al. Spinal injuries in children: role of MR imaging. AJNR Am J Neuroradiol 1993; 14:607–617. 27. Flanders AE, Schaefer DM, Doan HT, et al. Acute cervical spine trauma: correlation of MR imaging findings with degree of neurologic deficit. Radiology 1990;177:25–33. 28. Silberstein M, Tress BM, Hennessy O. Prediction of neurologic outcome in acute spinal cord injury: the role of CT and MR imaging. AJNR Am J Neuroradiol 1992;13:1597–1608. 29. Grabb PA, Pang D. Magnetic resonance imaging in the evaluation of spinal cord injury without radiographic abnormality in children. Neurosurgery 1994;35:406–414. 30. Mirvis SE, Geisler FH, Jelinek JJ, et al. Acute cervical spine trauma: evaluation with 1.5 T MR imaging. Radiology 1988;166:807–816. 31. Falcone S, Quencer RM, Green BA, et al. Progressive posttraumatic myelomalacic myelopathy: imaging and clinical features. AJNR Am J Neuroradiol 1994;15:747–754. 32. Baker LL, Goodman SB, Perkash I, et al. Benign versus pathologic compression fractures of vertebral bodies: assessment with conventional spin-echo, chemical shift, and STIR MR imaging. Radiology 1990;174:595–602. 33. Laredo J-D, Lakhdari K, Bellaïche L, et al. Acute vertebral collapse: CT findings in benign and malignant nontraumatic cases. Radiology 1995;194:41–48. 34. Brightman RP, Miller CA, Rea GL, et al. Magnetic resonance imaging of trauma to the thoracic and lumbar spine: the importance of the posterior longitudinal ligament. Spine 1992;17:541–550. 35. Davis SJ, Teresi LM, Bradley WG Jr., et al. Cervical spine hyperextension injuries: MR imaging findings. Radiology 1991:180:245–251. 36. Doran SE, Papadopoulos SM, Ducker TB, et al. Magnetic resonance imaging documentation of coexistent traumatic

locked facets of the cervical spine and disc herniation. J Neurosurg 1993; 79:341–345. 37. Goldberg AL, Keaton NL, Rothfus WE, et al. Ankylosing spondylitis complicated by trauma: MR findings correlated with plain radiographs and CT. Skeletal Radiol 1993;22:333–336. 38. Kaufman RA, Dunbar JS, Botsford JA, et al. Traumatic longitudinal atlantooccipital distraction injuries in children. AJNR Am J Neuroradiol 1982;3: 415–419. 39. Kongsted A, Sorensen JS, Anderson H, et al. Are early MRI findings correlated with long-lasting symptoms following whiplash injury? A prospective trial with 1-year follow-up. Eur Spine J 2008; 17:996–1005. 40. Ichihara D, Okada E, Kazuhiro C, et al. Longitudinal magnetic resonance imaging study on whiplash injury patients: minimum 10-year follow-up. J Orthop Sci 2009;14:602–610. 41. Vetti N, Kråkenes J, Eide GE, et al. MRI of the alar and transverse ligaments in whiplash-associated disorders (WAD) grades 1–2: high-signal changes by age, gender, event and time since trauma. Neuroradiology 2009;51:227–235. 42. Krakenes J, Kaale BR. Magnetic resonance imaging assessment of craniovertebral ligaments and membranes after whiplash trauma. Spine 2006;31:2820–2826. 43. Myran R, Kvistad KA, Nygaard OP, et al. Magnetic resonance imaging of the alar ligaments in whiplash injuries: A case-control study. Spine 2008; 33:2012–2016. 44. Hogan GJ, Mirvis SE, Shanmuganathan K, Scalea TM. Exclusion of unstable cervical spine injury in obtunded patients with blunt trauma: is MR imaging needed when multi-detector row CT findings are normal? Radiology 2005;237:106–113. 45. Como JJ, Thompson MA, Anderson JS, et al. Is magnetic resonance imaging essential in clearing the cervical spine in obtunded patients with blunt trauma? J Trauma 2007;63:544–549. 46. Stelfox HT, Velmahos GC, Gettings E, Bigatello LM, Schmidt U. Computed tomography for early and safe discontinuation of cervical spine immobilization in obtunded multiple injured patients. J Trauma 2007;63;630–636.

6 Imaging III: magnetic resonance imaging

47. Tomycz ND, Chew BG, Chang YF, et al. MRI is unnecessary to clear the cervical spine in obtunded/comatose trauma patients: the four year experience of a level I trauma center. J Trauma 2008; 64:1258–1263. 48. Muchow RD, Resnick DK, Abdel MP, Munoz A, Anderson PA. Magnetic

resonance imaging (MRI) in the clearance of the cervical spine in blunt trauma: a meta-analysis. J Trauma 2008; 64:179–189. 49. Stassen NA, Williams VA, Gestring ML, Cheng JD, Bankey PE. Magnetic resonance imaging in combination with helical computed tomography provides

a safe and efficient method of cervical clearance in the obtunded trauma patient. J Trauma 2006;60:171–177. 50. American College of Radiology. ACR Appropriateness Criteria. Suspected Spine Trauma. Reston, VA: American College of Radiology, 2009.

87

Chapter

7

Mechanisms of injury and their “fingerprints”  Richard H. Daffner

Many classifications are used to define vertebral injuries. Some surgeons believe that using a classification system based on mechanisms is beneficial in planning surgical reduction and stabilization. Two of the earliest classifications were by Whitley and Forsyth [1] and Holdsworth [2], who emphasized mechanisms of injury. Roaf [3] made a plea to classify vertebral injuries according to the principles of classic dynamics. Unfortunately, this scholarly approach is not useful to radiologists. Gehweiler and coworkers [4] addressed the needs of radiologists by stressing the radiographic features of the Holdsworth classification. In 1982, Allen and associates [5] reviewed a series of their own cases and observed a spectrum of injuries in the cervical vertebral column that they called phylogenies. They expanded the existing classifications along mechanistic lines and defined six common groups: • compressive flexion • vertical compression (pure axial loading) • distractive flexion • lateral flexion • compressive extension • distractive extension. These investigators also proposed that the probability of an associated neurologic lesion was directly related to the type and severity of the lesion [5]. Ferguson and Allen [6] applied a similar mechanistic classification to thoracolumbar fractures and established the following categories: • compressive flexion, comprising three subcategories: · anterior wedge fracture · anterior wedge with posterior distraction, with facet fracture or dislocation, or with both · burst with middle element failure and retropulsion of bone fragments (classic burst fracture) • distraction flexion injuries, in which the abnormalities are primarily posterior (Chance-type fractures) • lateral flexion • torsional flexion (rotary “grinding”) • translational (shearing) • vertical compression (pure axial loading).

88

Of all these classifications, the only one that addresses the concerns and needs of the radiologist is the one by Gehweiler and colleagues [4]. With so many variants and subtypes of injury, however, the classification process can become cumbersome, particularly if the average radiologist does not see large numbers of vertebral injuries. Vertebral fractures, like fractures in the peripheral skeleton, occur in predictable and reproducible patterns that are related to the kind of force applied to the affected bone. The same force applied to the cervical, thoracic, or lumbar column results in injuries that have a remarkably similar appearance [7]. A review of 4000 injuries to the vertebral column, which I observed over a 25-year period, suggests that there are essentially four mechanisms of injury: • flexion • extension • rotary or torque • shearing. These injuries can occur as isolated events or in combination with one another. The severity and extent of the damage produced by any one mechanism depend on the incident force, the position of the victim at the time of injury, and the victim’s velocity. This results in a pattern of recognizable radiographic signs that form a spectrum extending from mild soft tissue damage to severe skeletal and ligamentous disruption. I call these patterns the fingerprints of the injury [7]. This chapter reviews these four basic types of vertebral injury on the basis of their mechanism and the radiographic fingerprints that result from each. By learning the generic fingerprints of each type of injury, the reader should have no difficulty in recognizing the nature of the traumatic process no matter where it is located. Many of these findings have already been described elsewhere [4,7,8]. There are, of course, differences in occurrence of injury based on the relative flexibility and mobility of certain portions of the vertebral column, as mentioned in Chapter 3. For example, extension injuries, common in the cervical region, are uncommon in the less mobile thoracic and lumbar areas. Lateral flexion injuries in the cervical region tend to produce compression fractures of articular pillars; in the thoracic and lumbar regions, the same forces produce lateral burst injuries of vertebral bodies.

7 Mechanisms of injury and their “fingerprints” 

Levels of injury also depend on the age of the patient. Kiwerski [9] reviewed 1687 injuries and found that compression fractures and flexion injuries were most common at C5 and lower. Dislocations also tended to be more common at those levels. By comparison, extension injuries tended to occur in the upper levels, usually at C2. Compression, flexion, and burst injuries occurred more commonly in younger patients than older patients, the latter being more likely to suffer extension injuries and dislocations [9]. Miller and colleagues [10], in reviewing 400 cases at Duke, found that most injuries clustered around C5–C7. Multiple-level injuries in the cervical area occurred in two thirds of the patients and were found higher in the column than injuries of only a single level. As mentioned above, approximately 25% of patients have multiple noncontiguous injuries [11–13]. We have observed a high preponderance of cervical injury at C2 in patients over age 65 [14,15]. The incidence is just over twice that in younger patients. The reasons for this most likely relate to the loss of flexibility in the cervical vertebral column as one ages. As a result, the most mobile portion of the cervical area is at C2. In younger persons, most injuries are clustered around C5 and C6. Interestingly, the syndrome of spinal cord injury without radiographic abnormalities (SCIWORA) is twice as common in the elderly as in the young.

Background My colleagues and I have observed that the radiographic changes produced by vertebral injuries have a similar appearance regardless of their location. Two premises were considered: (1) vertebral injuries occur in a predictable pattern that depends on the mechanism, and (2) vertebral injuries caused by a particular mechanism produce the same radiographic changes regardless of location. Just as a criminal leaves fingerprints that link him or her to the crime, the patterns of injury represent the radiographic “fingerprints” that define the full extent of injury [7]. These observations were based on retrospective and prospective studies of 4000 vertebral injuries seen between 1983 and 2008 at the Trauma Center of Allegheny General Hospital in Pittsburgh. Of these injuries, 2123 (53%) were cervical, 886 (22%) were thoracic, and 991 (25%) were lumbar; 1042 (26%) involved multiple levels. Of interest, since we began using CT for trauma screening in 2000, we have diagnosed more isolated fractures of the transverse processes, spinous processes, facets, and articular pillars. Of the 2123 cervical injuries, 1444 (68%) were caused by flexion mechanisms (with or without axial load); 594 (28%) were caused by extension; 64 (3%) were the result of rotation (rotary subluxation or fixation of C1 on C2); and 21 (1%) were the result of shearing. Of the 886 thoracic injuries, 744 (84%) were caused by flexion mechanisms; 26 (3%) were caused by extension; 71 (8%) were the result of rotary forces; and 45 (5%) were the result of shearing. Of the 991 lumbar injuries, 872 (88%) were caused by flexion; 10 (1%) were caused by extension; 90 (9%) were the result of rotation; and 19 (2%) were the result of shearing

mechanisms. Most of the rotary and shearing injuries occurred in the thoracolumbar region (T11–L2). A number of injuries occurred under special circumstances. As mentioned, 1042 injuries involved multiple levels. Eightyseven patients had either ankylosing spondylitis or diffuse ankylosing skeletal hyperostosis (DISH). Of these, there were 57 (66%) cervical, 23 (26%) thoracic, and 7 (8%) lumbar injuries. All were extension injuries. Gunshot wounds accounted for 26 injuries. Finally, 213 patients sustained cervical unilateral facet lock, which accounted for 10% of all the cervical injuries. The cause of the injury was motor vehicle crashes in 85%, falls in 14%, and the remaining injuries had multiple causes, the most common of which were diving accidents. The injuries that resulted from motor vehicle crashes were associated almost universally with a deadly triad of alcohol use, high speed, and, in almost all cases, lack of seat-belt use. Surprisingly, automotive air bags have not been shown to change the incidence of vertebral injury when not used in conjunction with seat belts. All patients between 1983 and 1999 were evaluated by various imaging techniques, including radiography, CT, polydirectional tomography, and MR imaging. Beginning in 2000, CT became the prime screening method. The following anatomic regions were defined to take advantage of the natural clustering of injuries at certain levels: craniocervical (C0), atlanto-axial (C1–C2), lower cervical (C3–C7), upper thoracic (T1–T6), lower thoracic (T7–T10), thoracolumbar (T11–L2), and lower lumbar (L3–L5). Injuries were then categorized on the basis of mechanism: flexion, extension, rotation or torque, shearing, combined [1–6,16]. All categories included injuries in which axial loading was a factor.

Flexion injuries Flexion injuries are the most frequent type encountered in patients with vertebral trauma. They are the result of varying degrees of forward bending with the posterior third of the intervertebral disc space as the fulcrum (Fig. 7.1). With initial flexion, the upper and lower anterior vertebral endplates are compressed. When their structural compression limits are exceeded, cracks begin along the anterosuperior or anteroinferior margins. As the force continues, the target area becomes the vertebral body, particularly when combined with axial loading. This results in the literal explosion of the vertebral body in various configurations (burst fracture). At the same time, distractive forces are applied on the posterior vertebral structures. With sufficient distractive force, the posterior ligaments tear, beginning at the supraspinous ligament and proceeding anteriorly in anatomic order to eventually involve the posterior longitudinal ligament and the posterior portion of the intervertebral disc. Distractive forces result in widening of the distances between the posterior vertebral structures. Flexion injuries occur as isolated events or, more commonly, in combination with axial loading. Not surprisingly,

89

7 Mechanisms of injury and their “fingerprints” 

motor vehicle crashes account for most flexion injuries. In the typical scenario, an unrestrained occupant of a motor vehicle strikes the vertex of the head on a solid object. In the case of the driver or front-seat passenger, this object is the windshield (Fig. 7.2). If the victim is a rear-seat passenger, the object struck is usually the roof. Secondary impacts may

Fig. 7.1 Mechanism of flexion injury. Ordinary flexion produces motion about a fulcrum through the middle of the vertebral body. Excessive compression (curved arrow) results in fractures of the anterior and superior portions of the vertebral body. As the force continues, the fracture propagates posteriorly, ultimately producing fragments that can be displaced into the vertebral canal in a burst injury. In addition, there is distraction of the posterior elements with subsequent tearing of the soft tissue structures (straight arrow). A single mechanism can produce disruption of more than one vertebral compartment, resulting in a spectrum of injuries.

Fig. 7.2 Flexion mechanism in an unrestrained driver of a motor vehicle. On impact, the victim is thrown forward. The chest is impaled on the steering column, and the knees strike the dashboard. This mechanism is sufficient to produce flexion injuries in the lumbar vertebrae. If the sternum or ribs fracture, concurrent thoracic vertebral fractures can occur. In addition, if the head pitches forward, a cervical flexion injury may result as contact is made with the windshield.

90

produce additional injuries if the victim is thrown from the vehicle (“ejection” injuries). In another mechanism involving occupants of a motor vehicle, particularly one that rolls over, the victim’s head strikes any other solid object within the vehicle and hyperflexes the neck. In most of these individuals, a devastating vertebral injury could easily have been avoided by using seat belts. Cervical injuries produced by flexion are usually clustered between C4 and C7. Flexion injuries to the thoracolumbar region are also found in unrestrained drivers of motor vehicles who strike the steering column, which serves as a fulcrum for flexion. Air bags generally prevent this type of injury. Once a person is thrown from a motor vehicle, flexion injuries can occur at any level within the vertebral column when the victim strikes a solid object and the body flexes. This mechanism accounts for many of the multilevel injuries (cervicothoracic, cervicolumbar, thoracolumbar) that have been observed by a number of investigators [4,11]. Occupants of motor vehicles who wear lap-type seat belts without the shoulder harnesses may suffer a unique type of distraction fracture. Although originally described by Chance and Smith, these are generally referred to as Chance-type fractures. In these injuries, the lap belt becomes the fulcrum of flexion at the anterior abdominal wall, and the vertebra is literally ripped in two through a horizontal plane (Figs. 7.3 and 7.4) [17,18]. The thoracolumbar region is most commonly involved. These injuries occasionally result in severe neurologic deficit. A similar injury is produced when an individual traveling at high speed (e.g., in a fall or while skiing) strikes a solid object with the upper abdomen and the trunk forcibly flexes on that fulcrum. Motorcyclists suffer a characteristic fracture in the upper thoracic region when they are thrown over the handlebars and strike a solid object. In most instances, the area of contact is in the upper thorax between the scapulae. These injuries typically involve dislocations between T2 and T6 (Figs. 7.5 and 7.6) [19].

Fig. 7.3 Three types of distraction flexion injury caused by the use of lap-type seat belts. (A) Smith fracture. (B) Chance fracture. (C) Pure horizontal fracture. (From Daffner RH. Injuries of thoracolumbar vertebral column. In Dalinka MK, Kaye JJ, eds. Radiology in Emergency Medicine. New York: Churchill Livingstone, 1984, with permission.)

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.4 Chance fracture of L4. (A) Sagittal reconstructed CT image shows compression of the anterior superior body of L4 (arrow) and a horizontal fracture through the spinous process (arrowhead). (B) Coronal reconstructed CT image shows bilateral laminar fractures (arrows). (C) Abdominal CT section shows pneumoperitoneum (arrow) caused by the ruptured duodenum. (D) Sagittal STIR MR image shows edema in the body of L3 (*) in addition to the extensive soft tissue changes produced in and about L4. D

Individuals who dive into shallow water and strike their heads may suffer a devastating injury of the lower cervical region, usually at the C5–C7 level (Figs. 7.7 and 7.8). In these situations, the weight of the body provides the axial loading force that causes the damage [4]. Another form of flexion injury occurs in people who jump or fall from a height and land on their feet [4]. In addition to calcaneal fractures, the resultant forward flexion with axial loading of the upper torso generally produces burst fractures in the thoracolumbar region. Individuals with histories of a fall or with known bilateral calcaneal fractures or pylon fractures of the ankle should have CT or radiographs of the thoracolumbar region. Pelvic vertical shear injuries also result from this mechanism, and therefore pelvic radiographs should also be obtained.

Fig. 7.5 Mechanism of flexion injury in motorcyclists. On impact with a solid object, the rider is thrown over the handlebars. Forward flexion occurs in the upper thoracic region. (From Daffner et al. with permission [19].)

91

7 Mechanisms of injury and their “fingerprints” 

A

A

B

B

Flexion injuries characteristically involve the vertebral bodies, apophyseal (facet) joints, and the posterior ligaments. Fractures of the bony posterior elements are secondary to injuries to these structures. An exception is the “clay shoveler” fracture of the spinous process of the lower cervical column, which occurs as an isolated injury (Fig. 7.9). Flexion injuries can be divided into five categories: simple, burst, distraction, dislocation, and combined [16]. Simple injuries can be defined as compression of the vertebral endplates with anterior wedging of the vertebral body. Such injuries spare

92

Fig. 7.6 High thoracic fracture–dislocation in a motorcyclist. The patient is paraplegic. (A) Chest radiograph shows widening of the paraspinal soft tissues (arrows). (B) Sagittal STIR MR image shows fragmentation of T2 with retropulsion of a large bone fragment. There is evidence of cord hemorrhage (*) at T1 and edema. Note the extensive prevertebral hemorrhage (H) as well as widening of the interspinous space posteriorly (arrow).

Fig. 7.7 Mechanism of a flexion injury in a diving accident. (A) The diver’s head strikes the bottom with resultant forced flexion and increase in axial load. (B) This mechanism typically produces a teardrop fracture, usually at C5.

the posterior arch and the posterior ligaments. The disc space above is characteristically narrowed. These injuries are rarely associated with neurologic deficit (Figs. 7.10 and 7.11) and require no operative intervention. Burst fractures are those in which the vertebra has been exploded by compressive forces. This results in comminution of the vertebral body, retropulsion of bone fragments into the vertebral canal, and cleavage of the posterior arch [16,20–24]. Most burst fractures result in neurologic deficits of varying severity. Most will require surgical stabilization.

7 Mechanisms of injury and their “fingerprints” 

A

B

Fig. 7.9 Clay-shoveler fracture of C7. (A) Lateral radiograph shows the fracture (arrow) with downward displacement of the distal portion of the spinous process of C7. (B) Frontal radiograph shows an “extra” spinous process (arrow) representing the lower fragment of the C7 spinous process. Fig. 7.8 Teardrop fracture of C4 from a diving accident. Note the teardrop fragment (*). There is posterior subluxation of C5 on C6. The facet joints of C5–C6 (arrows) are widened. The interspinous space is also slightly widened.

A

B

There are several types of burst fracture that result from different degrees of flexion and axial loading. The most common variety (type A) is the result of pure axial loading. This produces comminuted fractures of both endplates with flattening of the vertebral body and retropulsion of the entire posterior cortex (Figs. 7.12 and 7.13) [21,24]. Type B burst injuries present with comminuted fractures of the upper endplate, retropulsion of the upper posterior vertebral body line, and a sagittal split of the lower part of the body as well as of the lamina. As a result of the sagittal splitting, both anteriorly and posteriorly, the interpedicle distance is widened (Figs. 7.14 and 7.15). This fracture is often referred to as a crush–cleavage fracture and is the result of combined flexion and axial loading [21,22,24]. Less common is the type C burst fracture, in which

Fig. 7.10 Simple compression fracture of L3. (A) Lateral radiograph shows interruption of the anterior portion of the body of L3 (large arrow). The posterior vertebral body line (small arrow) is intact. Note depression of the superior endplate. (B) T1-weighted sagittal MR image shows low signal in the body of L3. The posterior vertebral body line (arrow) is intact.

there is a comminuted fracture of the lower endplate and retropulsion of the lower posterior vertebral body line (Fig. 7.16). This fracture is also the result of flexion and axial loading [21,24]. Willen and coworkers [24] described a type D fracture that they called a burst–rotary injury. In reality, this is a pure rotary injury, which is discussed below. Finally, type E fractures are burst–lateral flexion injuries [24]. Two variations of burst fracture can occur in the cervical region [23,25]. These injuries are called flexion teardrop injuries. Torg and associates [23] differentiated between the two types by calling one a simple teardrop fracture, in which a triangular fragment of bone is displaced from the anteroinferior margin of a vertebral body. These injuries are isolated to the body only and produce no neurologic deficits. The second

93

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.11 Simple compression fracture of T12. (A). Lateral radiograph shows loss of height of T12, buckling of the anterior superior margin (large arrow), and central depression of the superior endplate (small arrow). (B) Sagittal reconstructed CT image shows the fractures. The posterior vertebral body line is intact (arrow). (C) Axial CT image shows the posterior vertebral body maintains its normal concavity (arrow).

Fig. 7.12 Mechanism of a burst injury. Forward flexion and axial loading contribute to the injury.

variety is a typical burst fracture with a teardrop fragment as well as retropulsion of the posterior vertebral body line and widening of the facet joints and interlaminar (interspinous) space (Fig. 7.17). Sagittal fractures of the vertebral body and lamina frequently accompany this type. These injuries are associated with severe neurologic deficit [23,25].

94

One cannot overemphasize the importance of abnormalities of the posterior vertebral body line in the diagnosis of burst fractures [21,26]. Any displacement, rotation, angulation, duplication, or absence of this structure is abnormal. These findings are not pathognomonic of burst fractures but they occur in sufficient frequency to suggest the diagnosis. McGrory and colleagues [26] found that the angle of the posterior vertebral body line as measured from the top and bottom of the vertebral endplates should not exceed 100° or be less than 80°. The interpedicle distance is measured directly on CT or on frontal radiographs from the sclerotic medial borders of the pedicles. The difference in the measurement between two contiguous levels should never exceed 2  mm. The sagittal cleavage variety of the burst fracture produces widening of this distance. This type of fracture also produces widening of the facet joints at the involved level. These findings are the result of the vertebra being split along the sagittal plane anteriorly through the vertebral body, as well as posteriorly through the lamina. Distraction injuries are of two varieties [16]. The more common of the two is manifested by widening of the interlaminar or interspinous space and interfacet distance without frank dislocation (Figs. 7.18 and 7.19). The hyperflexion sprain (Fig. 7.18) is the most common distraction injury [22,27]. There may be associated fractures caused by avulsion of bony fragments. Thoracolumbar or lumbar distraction injuries produce “naked facets,” a characteristic finding that may be seen on abdominal or vertebral CT scans as a result of the facet distraction. As a rule, when one does not see the posterior elements on more

7 Mechanisms of injury and their “fingerprints” 

A

B

C

D

Fig. 7.13 Burst fracture of C6. (A) Lateral radiograph shows compression of the body of C6 as well as duplication of the posterior vertebral body line (arrows). (B) Sagittal reconstructed CT image shows posterior displacement of a fragment to encroach the vertebral canal (arrow). (C,D) Axial CT images show severe comminution of the body of C6 with sagittal split. There is also fracture of the articular pillar on the right.

than one contiguous CT section, a distraction injury should always be suspected (Fig. 7.20). This type of distraction injury produces neurologic deficits in a large percentage of patients. A related distraction injury is produced by ligamentous damage that is less severe than that of the hyperflexion sprain, the socalled whiplash injury. The second type of distraction injury is associated with

horizontally oriented fractures through the vertebral body, pedicles, articular pillars, laminae, and/or spinous processes. These are the Chance-type injuries and, as mentioned, are frequently the result of accidents involving lap-type seat belts (Fig. 7.21) [16–18]. Chance-type injuries infrequently produce neurologic findings as a result of vertebral canal decompression. Patients with injuries from lap-type seat belts frequently have

95

7 Mechanisms of injury and their “fingerprints” 

A

B

D

C

E

Fig. 7.14 Burst fracture of L1. (A) Lateral radiograph shows kyphotic angulation and loss of height of L1. A fragment of the posterior vertebral body line is displaced into the vertebral canal (arrow). (B) Frontal radiograph shows widening of the interpedicle distance (double arrow). (C) Sagittal reconstructed CT image shows the bone fragment in the vertebral canal (arrow), correlating with the findings in A. (D) Axial CT image shows the retropulsed bone fragment (*) in the vertebral canal. (E) Axial CT image shows the sagittal fractures through the body and lamina (arrows), which resulted in widening of the interpedicle distance.

intraabdominal visceral injuries and should undergo body CT evaluation if they have not already done so. All distraction injuries require surgical stabilization. Dislocation involves a loss of bony continuity at the articular surfaces (Figs. 7.22 and 7.23). Flexion–dislocation injuries

96

are the result of severe distraction forces. These may or may not be associated with fractures [16]. In the cervical region, unilateral or bilateral facet lock is a common manifestation of dislocation (Fig. 7.22). These injuries also result in a high incidence of neurologic deficit, plus there is a high incidence

7 Mechanisms of injury and their “fingerprints” 

A

C

B

D

of vertebral artery injury (Fig. 7.24) [28,29]. These injuries also require surgical intervention. Combined injuries are those with features of more than one of these categories. Unilateral facet lock (Fig. 7.25) is a combined lateral flexion–distraction injury. It is discussed in detail in Chapter 8.

Fig. 7.15 Burst fracture of L1. (A) Axial CT image shows fragments of the posterior body displaced into the vertebral canal (*). (B) Axial CT image shows sagittal fractures of the body and lamina (arrows). (C) Sagittal reconstructed CT image shows the retropulsed fragments in the vertebral canal (*). There is an old limbus deformity of the superior margin of L4. (D) Coronal reconstructed CT image shows widening of the interpedicle distance (double arrow).

Articular pillar fractures are common in flexion mechanisms, particular if the patient’s head is turned at the time of impact. These injuries are also frequently associated with a high incidence of injury to the vertebral arteries (Figs. 7.26 and 7.27) [28,29].

97

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.16 Burst fractures of L2 and L3. (A,B) Sagittal reconstructed CT images show typical burst pathology of L3. The bone fragment in the canal (arrows) is from the inferior margin of L2. (C) Axial CT image shows significant spinal stenosis from the displaced fragments (arrows) from the posterior inferior margin of L2. This injury is a variant of the limbus injury.

A

D

98

B

C

Fig. 7.17 Burst fracture of C5. (A) Lateral radiograph shows fragmentation of the body of C5 and posterior bowing of the posterior vertebral body line (arrow). (B) Sagittal reconstructed CT image shows a fragment of the posterior body of C5 in the vertebral canal (arrow). (C) Axial CT image shows a sagittal split of the body of C5 (large arrow) as well as a fracture of the lamina on the right (small arrow). (D) Coronal reconstructed CT image shows the sagittal split of the body of C5 (arrow).

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.18 Hyperflexion sprain. (A) Lateral radiograph shows reversal of lordosis at C4–C5 and widening of the interlaminar (interspinous) distance (*) between C4 and C5. (B) Frontal radiograph shows the wide interspinous distance (double arrow). (C) Sagittal T1-weighted MR image shows a tear of the posterior longitudinal ligament at C4–C5 (arrow). A

B

C

Fig. 7.19 Distraction injury L1–L2. (A) Frontal radiograph shows widening of the interspinous space (double arrow). Note the “naked” facets. (B) Lateral radiograph shows perching of the facets (arrow). There is a small avulsion off the posterior inferior margin of L1. (C) Axial CT image shows the absence of posterior elements of the adjacent vertebra (*) and “naked” facets. A

B

C

Fig. 7.20 Distraction fracture–dislocation L1–L2. (A) CT scout view shows L1 pulled away from L2. Note the “naked” facets of L2 (*). (B,C) Axial CT images show absence of posterior elements of adjacent vertebrae (* in B), “naked” facets, and fractures of the vertebral body.

99

7 Mechanisms of injury and their “fingerprints” 

A

B

C

D

Fig. 7.21 Chance-type fractures. (A) Frontal radiograph shows horizontal fractures through the body and left transverse process (arrows) of L3. (B) Lateral radiograph shows horizontal fracture through the pedicle (arrow). (C) Sagittal reconstructed CT image in another patient shows compression of the vertebral body (arrow) and a horizontal fracture through the spinous process (arrowhead). (D) Coronal reconstructed CT image shows the horizontal posterior element fractures (arrows).

A

D

100

B

C

E

Fig. 7.22 Dislocation C5–C6 with bilateral facet lock. (A) Lateral radiograph shows anterolisthesis of C5 on C6 with facet perch (arrow). (B) Sagittal reconstructed CT image shows the dislocation with widening of the interspinous space (*). (C) Sagittal image more laterally shows the facet lock (arrow). (D) Axial CT image shows the locked facet on the right (arrow). (E) Sagittal T2-weighted MR image shows a large fragment of herniated disc impinging the spinal cord (arrow). Remarkably, the patient was neurologically intact.

7 Mechanisms of injury and their “fingerprints” 

A

B

C

D

Fig. 7.23 L4 fracture–dislocation. (A) Lateral radiograph shows severely comminuted fractures of L4 with anterolisthesis of major fragments. (B,C) Axial CT images show the severe comminution and obliteration of the vertebral canal. (D) Sagittal three-dimensional volumetric reconstructed image shows the severity of the injury.

A

C

Fig. 7.24 Unilateral facet lock with vascular injury. (A) Lateral radiograph shows anterolisthesis of C3 on C4. The point of facet locking (large arrow) is visible. Note the “bow tie” appearance of the rotated articular pillars (small arrows). (B) Sagittal reconstructed CT image shows the locked facet (arrow). (C) Axial CT angiogram image of C2 shows contrast in the right vertebral artery (arrow) and no contrast on the left (*). (D) Sagittal reconstructed CT angiogram image shows hematoma (arrow) in the vertebral artery at the point of lock.

B

D

101

7 Mechanisms of injury and their “fingerprints” 

A

B

C

D

Fig. 7.25 Unilateral facet lock C4–C5. (A) Lateral radiograph shows anterolisthesis of C4 on C5 (arrow). (B) Frontal radiograph shows rotation of spinous process of C4 to the right, the side of the lock, while the spinous process of C5 is midline (arrows). (C) Sagittal reconstructed CT image shows a severe comminuted fracture of the articular pillar of C5 with locking of the C4 facet on the fracture (arrow). (D) Axial CT image shows the comminuted pillar fracture of C5 on the right.

102

A

B

A

B

Fig. 7.26 Articular pillar fracture. (A) Axial CT image shows severe comminution of the right pillar with impingement of the neural foramen (arrow). (B) Sagittal reconstructed CT image shows locking of the facet on the fractured pillar (arrow).

Fig. 7.27 Articular pillar fracture with vascular injury. (A) Axial CT angiogram image shows fractures of the pedicle and lamina on the left (arrowheads). There is a hematoma in the left vertebral artery (arrow). Compare with the right. (B) Sagittal reconstructed CT angiogram image shows the hematoma (arrow) in the vertebral artery.

7 Mechanisms of injury and their “fingerprints” 

Fig. 7.28 Simple compression fracture (arrow) of L2. The posterior vertebral body line is intact.

A

A

B

B

Injuries include • compression (Fig. 7.28), fragmentation (Fig. 7.29), and burst fracture of the vertebral bodies (Figs. 7.30 and 7.31) • teardrop fragments of the anteroinferior margins of the vertebral bodies (Fig. 7.32) • widening of the interlaminar or interspinous spaces (Fig. 7.33) • anterolisthesis (Fig. 7.34) • disruption of the posterior vertebral body line (Figs. 7.30 and 7.31) • jumped or locked facets (Figs. 7.35 and 7.36) • narrowing of the intervertebral disc spaces, usually above the level of involvement (Figs. 7.28 and 7.33) [4,7]. Once again, these findings may be found at any level of the vertebral column. Note the similarity among the findings in Fig. 7.30, a cervical injury and Fig. 7.31, a lumbar injury.

Fig. 7.29 Burst fracture of L3 with fragmentation. (A) Axial CT image shows severe comminution of the body of L3. There is widening of the facet joint on the left (arrow). (B) Sagittal reconstructed CT image shows fragmentation of the body of L3 and retropulsion of a bone fragment from the posterior inferior margin of L2 (arrow).

C

Fig. 7.30 C7 burst fracture. (A) Lateral radiograph shows compression of the body of C7 and retropulsion of a bone fragment (arrow) into the vertebral canal. (B) Sagittal reconstructed CT image shows the canal encroachment (arrow). (C) Sagittal STIR MR image shows the canal encroachment. The spinal cord is intact.

103

7 Mechanisms of injury and their “fingerprints” 

A

D

B

C

E

Fig. 7.31 L1 burst fracture. (A) Lateral radiograph shows compression of L1 and retropulsion of a fragment from the upper posterior vertebral body line (large arrow). Note the normal position of the lower body line (small arrow). (B) Frontal radiograph shows widening of the interpedicle distance of L1 (double arrow). (C,D) Sagittal reconstructed (C) and axial CT (D) images show the retropulsed fragment in the canal (arrow in C, * in D). (E) Axial CT image shows the sagittal fractures (arrows) through the body and lamina to account for the wide interpedicle distance.

104

7 Mechanisms of injury and their “fingerprints” 

A

B

A

B

D

Fig. 7.32 Cervical teardrop fractures (arrow) of the anterior inferior vertebral bodies in two patients (A,B). Note the wide interlaminar (interspinous) space in B (*).

C

Fig. 7.33 Distraction injury at T12–L1 showing wide interspinous space. (A) CT scout view shows cephalad displacement of T12 with perching of the facets on the right (arrow). (B) Sagittal reconstructed CT image shows the posterior distraction (*). (C) Axial CT image shows no posterior elements (*) and naked facets. (D) Sagittal STIR MR image shows rupture of the posterior longitudinal ligament (arrow) as well as extensive posterior hemorrhage.

105

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.34 Dislocation at T11–T12 in a patient with diffuse idiopathic skeletal hyperostosis (“broken DISH”). (A) Lateral radiograph shows anterolisthesis of T11 on T12. Note the fragmentation of the superior body of T12 (arrow). (B,C) Sagittal reconstructed CT images show bilateral facet locking (arrow).

A

B

C

Fig. 7.35 Unilateral jumped and locked facet. (A) Lateral radiograph shows duplication of the articular pillars (“bowtie sign”) at C4 (*). The point of locking is visible (arrow). (B) Sagittal reconstructed CT image shows the lock on a fractured articular pillar (arrow). (C) Axial CT image shows the point of locking on the left. Note a “reverse hamburger bun sign” (arrow).

Extension injuries Extension injuries are common in the cervical region but rare in the thoracic and lumbar regions [16]. They are the result of varying degrees of backward bending, with the articular pillars serving as the fulcrum of motion. Consequently, extension injuries disrupt anterior structures. The main radiographic abnormality (fingerprint) encountered is widening of the disc space below the level of injury, frequently associated with avulsion fractures of the anterosuperior lip of the vertebral bodies. Retrolisthesis commonly occurs when both the entire disc and the anterior and posterior longitudinal ligaments are disrupted. In severe injuries, the articular pillars are crushed and the facet joints are dislocated [2,4].

106

Widening of the intervertebral disc space rarely occurs under normal circumstances. The finding of a wide disc space, particularly in an older individual with extensive degenerative changes that have produced disc space narrowing at other levels, should alert the radiologist to the possibility that an extension injury may be present. When a wide disc space is encountered, in a patient without or with neurologic signs, MR imaging is the procedure of choice. In the cervical region, two of the most commonly encountered mechanisms of extension injury are motor vehicle crashes and falls. In a motor vehicle crash, the neck of an unrestrained driver may hyperextend as the chest strikes the steering wheel, producing a traumatic spondylolysis of the posterior arch of C2 (“hanged-man” fracture; Figs. 7.37 to 7.39) [2,4,27,30,31].

7 Mechanisms of injury and their “fingerprints” 

A

D

B

C

E

Fig. 7.37 Mechanism of the hanged-man injury in a motor vehicle crash. The unrestrained driver pitches forward, impaling the thorax on the steering wheel. If the face strikes the windshield before the vertex of the head, the head is forced backward in hyperextension to produce the cervical injury.

Fig. 7.36 Cervical dislocation with bilateral facet lock. (A) Lateral radiograph shows anterolisthesis of C5 on C6. (B,C) Sagittal reconstructed CT images show the facet lock (arrows). (D) Axial CT image shows the lock manifest as bilateral “reverse hamburger bun signs” (arrows). (E) Sagittal inversion recovery MR image shows herniation of the C5 disc impinging the canal (arrow).

Fractures of the dens with posterior dislocation may also occur with extension. (Anterior fracture–dislocation of the dens usually occurs with a primary or secondary flexion mechanism.) It is unusual for either of these dens injuries to have any associated neurologic findings unless an epidural hematoma is compressing the spinal cord. Clinically, these injuries may produce nothing more than upper neck stiffness, dysphagia, or torticollis. The patients may seek medical evaluation days or weeks after injury. A third extension-type injury of the cervical region occurs at a lower level. Anatomically, these injuries range from simple hyperextension sprains, in which the anterior ligaments are disrupted along with disc-bond injury (Figs. 7.40 and 7.41) [4,16,32,33], to severe fracture–dislocation (Fig. 7.42). In either case, the patient may experience severe neurologic compromise, usually a central cord syndrome. This is particularly true in the elderly, in whom osteophytes or syndesmophytes project into the vertebral canal and narrow it. In these individuals, relatively mild extension trauma may result in

107

7 Mechanisms of injury and their “fingerprints” 

A

B

Fig. 7.38 Hanged-man fracture of C2. (A) Lateral radiograph shows anterolisthesis of C2 on C3. There is duplication of the posterior vertebral body line of C2 (arrows). (B) Axial CT image shows a fracture through the posterior body of C2 on the left and a laminar fracture on the right. Note the position of the fragments of the posterior body of C2 (arrows). The displacement of the right side of the body of C2 accounted for the double posterior body line on the radiograph.

A

C

108

B

Fig. 7.39 Hanged-man fracture of C2. (A) Lateral radiograph shows duplication of the posterior vertebral body line of C2 (arrows). Note the wide disc space (*), the hallmark of an extension injury. (B) Sagittal reconstructed CT image shows the coronal cleavage fracture of the posterior body of C2. The arrows show the reason for the duplication of the posterior body line. (C) Axial CT image shows the coronal fracture of C2 (arrows). In addition, there are fractures of the posterior arch of C1 (arrowheads).

severe neurologic compromise. The typical clinical picture is of an elderly patient with quadriplegia in whom the only significant radiologic finding is cervical spondylosis or DISH (Fig. 7.40). Typically, they will also have a bruise or laceration on the chin as a sign of the mechanism of injury. Extension injuries are unusual in the thoracic and lumbar regions. They may occur, however, in several scenarios of hyperextension, such as when a person falls and lands backward over a solid object (Fig. 7.43) or is struck from behind by a large object. Patients with DISH or ankylosing spondylitis can suffer extension injuries through the fused vertebrae even with relatively minor trauma (Figs. 7.44 to 7.46) [16,34,35].

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.40 Extension sprain. (A) Mechanism of injury in an elderly individual. Forced extension of the cervical column produces severe cord compromise as the result of compression of the spinal cord by osteophytes (or syndesmophytes). This is a common mechanism of injury in an elderly patient with severe neurologic compromise in whom the only radiographic finding is evidence of degenerative disease. (B) Lateral radiograph shows retrolisthesis of C3 on C4 with widening of the C3 disc space (*). This patient is quadriplegic. (C) Autopsy specimen from the same patient shows widening of the C3 disc space (*) and hemorrhage in the spinal cord (arrow).

A

B

C

Fig. 7.41 Extension sprain. (A) Lateral radiograph shows extensive prevertebral soft tissue swelling (*). (B) Sagittal reconstructed CT image shows retrolisthesis of C6 on C7 with widening of the C6 disc space. There is a small avulsed bone fragment off the anterior superior body of C7 (arrow). Note the prevertebral soft tissue swelling (*). (C) Sagittal STIR MR image shows rupture of the anterior (large arrow) and posterior (small arrow) longitudinal ligaments, retrolisthesis of C6 on C7, and prevertebral soft tissue swelling (*).

Extension injuries can be divided into three categories: simple, distraction, and dislocation [1–4,6,16,30,31]. Simple injuries are defined as avulsion of the anterosuperior portion of the vertebral body. They produce minimal radiographic findings and generally no neurologic deficit unless the patient has an underlying degenerative condition (Fig. 7.47). In that

situation, severe neurologic compromise, most likely “central cord syndrome” occurs. Distraction injuries result in widening of the intervertebral disc space with or without an avulsion fracture of the vertebral body below (Fig. 7.48) [16]. The hanged-man fracture of C2 with separation of the fracture fragments is also an example

109

7 Mechanisms of injury and their “fingerprints” 

A

B

D

C

Fig. 7.42 Hyperextension fracture–dislocations. (A) Lateral cervical radiograph shows anterior dislocation of C6 on C7. The spinolaminar line, however, remains intact (arrows). (B) Sagittal reconstructed CT image shows the same findings. (C) Scout view of a lumbar dislocation shows findings identical to those seen in the cervical dislocation in A. Note the preserved spinolaminar line (arrow). (D) Sagittal reconstructed CT image shows the same findings.

A

Fig. 7.43 Mechanism of extension injury of the thoracolumbar column. Most of these injuries occur when the individual falls and lands across a fixed object. Similar injuries may occur in individuals who are thrown from motor vehicles or horses and strike solid objects.

110

B

Fig. 7.44 Extension fractures in patients with ankylosing spondylitis. (A) Cervical fracture–dislocation (arrow). (B) L1–L2 injury. Note the wide disc space (arrow).

7 Mechanisms of injury and their “fingerprints” 

A

A

B

C

B

of a distraction injury (Figs. 7.38 and 7.39). The incidence of neurologic findings will depend on the degree of distraction. Extension dislocation (Fig. 7.42) results in a loss of bony continuity at the articular surfaces. These injuries almost always produce severe neurologic deficit. Two entities that deserve closer scrutiny for differentiation are the hyperextension dislocation (Fig. 7.48) and the extension teardrop fracture (Fig. 7.49). In the hyperextension dislocation, the horizontal length of the avulsed fragment exceeds its vertical height. These patients have severe neurologic deficits. By

Fig. 7.45 Extension injury in a patient with ankylosing spondylitis. (A) Lateral radiograph shows buckling of the anterior cortex of C7 (arrow). (B) Sagittal reconstructed CT image shows anterior and posterior disruption of the fused spine (arrows) and widening of the disc space. (C) Sagittal T2-weighted sagittal MR image shows hemorrhage and swelling in the spinal cord (arrow).

Fig. 7.46 Extension injury in diffuse idiopathic skeletal hyperostosis (DISH). Lateral radiograph (A) and sagittal reconstructed CT image (B) show fractures through the fused bony mass (arrowheads) (“broken DISH”).

comparison, in the extension teardrop fracture, the avulsed fragment has a vertical height equal to or greater than the length of the horizontal component [27]. These patients rarely have neurologic findings. The findings should be apparent on CT as well as radiographs. In summary, the fingerprints of an extension injury include the following: • widening of the disc space below the level of injury (Fig. 7.41) • triangular avulsion fractures of the anterosuperior lip of vertebral bodies (Fig. 7.41)

111

7 Mechanisms of injury and their “fingerprints” 

A

A

Fig. 7.47 Extension sprain C5–C6. (A) Lateral radiograph shows retrolisthesis of C5 on C6 with a small avulsed fragment of bone from the anterior inferior margin of the body of C5 (arrow). Note the wide disc space (*). (B) Sagittal reconstructed CT image shows the avulsed fragment (arrow).

B

B

C

Fig. 7.48 Extension dislocation. (A) Lateral radiograph shows reversal of lordosis at C4–C5 and widening of the C4 disc space (*). (B) Sagittal reconstructed CT image shows a large fragment of the posterior inferior margin of C4 displaced in the vertebral canal (arrow). Note the wide disc space (*). (C) Sagittal STIR MR image shows complete transection of the spinal cord and extensive prevertebral (*) and posterior (**) hemorrhage. The plane of the injury extends through the C4 disc space (arrow).

112

7 Mechanisms of injury and their “fingerprints” 

• •

retrolisthesis (Figs. 7.41 and 7.47) neural arch fractures (Fig. 7.39). In the less common extension fracture–dislocation injury involving the articular pillars and vertebral arches, there are two fingerprints: • anterolisthesis with normal interlaminar or interspinous spaces • normal spinolaminar lines (Fig. 7.42) [4,7].

Fig. 7.49 Extension teardrop injury C6–C7. There is avulsion of a bone fragment from the anterior inferior margin of C6 (black arrow). Note that the fragment’s height exceeds its length. There are also fractures of the spinous processes of C5 and C6 (arrows). There is widening of the disc space (*). The patient had no neurologic findings.

A

Rotary (torque) injuries Rotary injuries are the result of rotational or torsion force applied about the long axis of the vertebral column. Rotary injuries occur primarily in two areas of the spine. The lesssevere variety are found at the craniovertebral junction as rotary subluxation/fixation of C1 on C2. The more severe type is found at the thoracolumbar junction, where they are frequently associated with a flexion component as a consequence of torsional loading or compression in that region [16]. The usual mechanism is that of a heavy blow in the shoulder region that compresses the vertebral column while deflecting and twisting the lower torso laterally. This produces disruption of the posterior ligament complex and consequent dislocation of the facet joints or facet fracture. The injury may be sustained when the victim is struck by a large falling object, in a fall, or, most commonly in our practice, by ejection from a motor vehicle and secondary impact on a solid object (Fig. 7.50) [4,7]. Clinically, a bruise or skin injury in the vicinity of the shoulder or scapula is a clue that this injury may be present. The reason most of these injuries are at the thoracolumbar junction relates to the anatomy of the facet joints, which restrict motion in the region, as discussed in Chapter 3. These injuries are all highly disruptive and typically produce severe neurologic compromise. All require surgical stabilization. The mechanism described above produces the most common type of rotary injury encountered at the Allegheny Trauma Center. They are also the most disruptive and result in the involved vertebra literally being pulverized. For this reason, the more descriptive term rotary grinding injury is often used in discussing them. The typical radiographic manifestations of a rotary grinding injury include dislocation and rotation of fragments (Fig. 7.51). Fractures of the transverse processes, ribs, or both are common. Typically, there is avulsion of a triangular fragment of bone from the anterior superior margin of the vertebra, giving that structure the appearance of a soft drink B

Fig. 7.50 Mechanisms of rotary injuries.

113

7 Mechanisms of injury and their “fingerprints” 

A

C

B

Fig. 7.51 Rotary fracture–dislocation of L3. (A) Lateral radiograph shows compression of the body of L2 with a loose fragment of the anterior superior margin that resembles a torn soft drink can (arrow). (B) Frontal radiograph shows right lateral dislocation of the major portion of L3. There is a transverse process fracture of L2 on the left (arrow). (C) Axial CT image shows severe comminution of the body of L3 resembling a burst fracture. However, the transverse process fracture (arrow) and wide facet joint (*) on the left indicate the true nature of this injury.

A

B

C

Fig. 7.52 Rotary injury of L3. (A) Sagittal reconstructed CT image shows a severely comminuted fracture of the body of L3 with the ripped can top appearance anterosuperiorly (arrow) and retrolisthesis. (B,C) Axial CT images show the severe comminution, transverse process fractures on the left (arrow) and widening of the facet joint on the right (* in B).

can that has had its top ripped off (Fig. 7.52) [16]. The posterior vertebral body line is commonly disrupted and often cannot be recognized as a discrete structure. The CT findings are also characteristic and consist of severe fragmentation with a concentric distribution of the fragments (Figs. 7.51 and 7.52). Typically, the facet joints are disrupted, one being displaced forward and one backward (Fig. 7.53). On MR examination, there is severe damage not only to the vertebra but also to the soft tissues, particularly posteriorly (Fig. 7.54). Rotary grinding injuries are frequently confused with burst fractures, since they often self-reduce after the patient is immobilized in the supine position. It is important to be able to distinguish between these injuries, since the treatments differ

114

significantly. In order for a spine surgeon to produce a mechanically stable vertebral column after these injuries, the rotational component of the injury must be corrected. Burst fractures typically occur about the sagittal plane of the vertebral column. Surgical stabilization of these injuries is oriented about that sagittal plane. Rotary injuries, by comparison, require stabilization not only in the sagittal plane but also in the coronal and axial planes. Therefore, if a rotary injury is misinterpreted as a burst injury, it is possible that the rotational component will not be corrected, stability will not be achieved, and collapse and perhaps further neurologic damage may occur (Fig. 7.55). The salient imaging features for the differentiation of rotary grinding injuries from burst fractures are shown in

7 Mechanisms of injury and their “fingerprints” 

A

C

Figs. 7.56 and 7.57. Rotary injuries have a greater degree of separation of fragments and a greater tendency to dislocate than do burst fractures. In rotary injuries, the normal radiographic anatomy is severely distorted; in burst injuries, the vertebral components, while separated, are clearly recognizable. Burst fractures typically produce widening of the interpedicle distance as a result of sagittal cleavage. Displaced fragments from the posterior vertebral body line tend to be located along the sagittal plane. Fractures of the transverse

B

Fig. 7.53 Rotary injury of L3. (A) Axial CT image shows severe comminution and canal encroachment, suggesting a burst fracture. However, there is a transverse process fracture on the left (arrow). (B) Axial CT image slightly lower shows the concentric fracture pattern (arrowheads), a transverse process fracture on the right, and widening of the facet joint on the left (arrow). In this patient, the injury vector was left to right. (C) Sagittal reconstructed CT image shows involvement of not only L3 but also L2.

Fig. 7.54 Rotary injury of L3. Sagittal STIR MR image in the same patient as in Fig. 7.53 shows bone fragments displaced into the vertebral canal (arrow) as well as evidence of hemorrhage of the conus and surrounding soft tissues.

115

7 Mechanisms of injury and their “fingerprints” 

A

B

C

Fig. 7.55 Rotary injury of L1 assumed to be a burst fracture. (A) Lateral radiograph shows pedicle screw and rod fixation spanning T12 to L2. (B) Lateral radiograph one month later shows retrolisthesis of T12 on L1 and loss of height of L1. (C) Lateral radiograph one month later shows an extensive repair to have been performed that accounts for providing stability in the sagittal, coronal, and axial planes. There is already evidence of construct failure at L2. One pedicle screw has backed out and the other has become disconnected from the rod (arrows).

A

B

C

Fig. 7.56 Rotary fracture of L3. (A) Lateral radiograph shows severe comminution and loss of height of the body of L3. A fragment of bone is displaced into the vertebral canal (large arrow). Note the simple fracture of the body of L4 (small arrow). (B) Frontal radiograph shows comminution of the body of L3 with left laterolisthesis of a portion. Note the transverse process fracture on the left (arrow). (C) Axial CT image shows the comminution, transverse process fracture on the left (arrowhead) and widening of the facet joints asymmetrically (arrows). In this patient, the injuring vector was from right to left. Compare with the burst fracture in Fig. 7.57.

processes, ribs, or both are a frequent component of rotary injuries but do not occur in burst injuries. On CT, rotary injuries typically produce a concentric distribution of fragments and facet disruptions, as described above. Burst fractures, by comparison, typically have a linear sagittal distribution of displaced fragments on CT. Finally, on MR imaging, rotary injuries have severe posterior soft tissue damage, whereas the damage typically is confined only to the involved vertebra in burst fractures.

116

A second type of rotary injury occurs at the occipitoatlanto-axial region [36–39]. These are the rotary atlanto-axial fixation injuries first described by Fielding and Hawkins in 1977 [37]. The injury results from disruption of the transverse ligament of the atlas and also of the alar ligaments, which ordinarily prevent excessive rotation of the atlas on the axis. Fielding and Hawkins [37] described four types of this abnormality, of which the most common variety involves rotary fixation without dislocation of the atlas (Fig. 7.58). The other

7 Mechanisms of injury and their “fingerprints” 

A

B

C

D

Fig. 7.57 Burst fracture of L4. (A) Lateral radiograph shows comminution of the body with posterior bowing of bone into the vertebral canal (arrow). (B) Frontal radiograph shows widening of the interpedicle distance of L4 (double arrow). (C) Axial CT image shows the comminuted body fracture and canal encroachment (*). There is a laminar fracture on the left. In this instance, the force vectors were along the sagittal plane. (D) Sagittal reconstructed CT image shows comminution, loss of height, and displacement of bone fragment into the vertebral canal (arrow). Compare with the rotary injury in Fig. 7.56.

A

B

D

E

C

F

Fig. 7.58 Rotary atlanto-axial fixation. The patient was unable to straighten his head. (A) CT scout view shows gross rotation of the head. (B) CT axial image of C1 shows gross rotation to the left. (C) CT axial image slightly lower shows dislocation of the lateral mass of C1 on the left. (D) Sagittal reconstructed CT image shows dislocation of both lateral masses (arrows), which sit nearly 90° to C2. (E,F) Three-dimensional volumetric reconstructed CT images show the dislocated lateral masses (arrow).

117

7 Mechanisms of injury and their “fingerprints” 

varieties involve anterior or posterior displacement of the atlas [37–39]. These occur much less commonly. Pure rotary atlanto-axial dislocation is extremely rare. In this instance, the atlas is rotated on the axis more than 45°, with resultant locking of the lateral masses of the atlas over the superior articular surfaces of the axis. This condition should not be confused with the more common atlanto-axial rotary fixation (Fig. 7.59).

A

C

Two additional rotary abnormalities may also be encountered at the craniovertebral junction. The first is the unusual combination of atlanto-axial rotation with occipito-atlantal rotary subluxation [36]. The second is the more common rotary subluxation of the axis (Fig. 7.60). The key imaging findings in this entity are alignment of the external occipital protuberance, dens, and spinous process of C3 with rotation of the

B

D

Fig. 7.59 Rotary atlanto-axial fixation. (A,B) Axial CT images shows C1 rotated to the right and C2 rotated to the left. A small portion of the lateral mass of C1 is visible on the left (arrow in B). (C,D) Coronal reconstructed CT images show the naked facets of the lateral masses of C1 (arrows). This patient is not as severely injured as the patient in Fig. 7.58.

118

7 Mechanisms of injury and their “fingerprints” 

A

C

B

Fig. 7.60 Rotary atlanto-axial subluxation. (A,B) Axial CT images show rotation of C1 and C2 to the left. However, the rotation of the axis is more pronounced. (C) Axial CT image at the atlanto-axial junction shows the degree of rotation of C1 on C2 to be 34° (normal is 22.5° or less).

Shearing injuries

spinous process of the axis. This entity should not be confused with the asymmetry often found between the dens and the lateral masses of C1 that occurs as a result of the patient being turned in the CT gantry. There may also be minor degrees of rotation of C1 on C2. However, if the amount of the rotation is less than 22°, you can be assured that the abnormality is caused by positioning. In summary, the fingerprints of rotary injuries include the following (Figs. 7.51 and 7.52): • severe fragmentation of vertebrae, including rotation and dislocation of the fragments • fractures of the transverse processes, ribs, or both • fracture or dislocation of the facets and pillars • disruption of the posterior vertebral body line • circular array of fragments on CT • spinous process fracture.

Shearing injuries are the result of horizontally or obliquely directed forces in which axial loading is not a factor. They may occur in combination with flexion or extension injuries. In most instances, the lower portion of the body is fixed, and the vertebral column, the unfixed portion, absorbs the horizontal or oblique force and moves with it. Typically, the patient is struck with a large object, suffers a fall, or, most commonly, is ejected from a motor vehicle and suffers secondary impact (Fig. 7.61) [4,7,18,40–42]. Shearing injuries may be combined with rotary injuries and, like them, are extremely disruptive. Most shearing injuries result in severe neurologic compromise. All require surgical stabilization in all three planes. Shearing injuries usually occur in the thoracolumbar region because of the limits placed on motion other than flexion and extension. In the cervical region, they are most likely to occur at the craniovertebral junction in the form of occipito-atlantal dislocation (Fig. 7.62). Denis and Burkus [41] described a group of 12 patients who had shearing thoracolumbar fracture–dislocations associated with extension mechanisms. They called these injuries “lumberjack paraplegia,” since most were incurred while the victim was harvesting timber and struck by falling trees or limbs. The remainder of the patients suffered identical injuries as the result of motor vehicle crashes in which they were thrown from a vehicle, run over by a tractor, or pinned between a tree and a moving vehicle. The typical radiographic features of shearing injuries of the thoracolumbar region include horizontal or oblique distraction

119

7 Mechanisms of injury and their “fingerprints” 

B

A

Fig. 7.61 Mechanisms of shearing injuries.

A

B

Fig. 7.62 Occipito-atlantal dislocation. (A) Sagittal reconstructed CT image shows the dens–basion distance to be 17 mm (normal is 6–12 mm). (B) Sagittal STIR MR image shows rupture of the apical ligaments (black arrow), cord hemorrhage (white arrow), and massive prevertebral soft tissue swelling (*).

and dislocation (Figs. 7.63 and 7.64). The involved vertebrae often have a “windswept” appearance on CT as well as on radiographs (Figs. 7.65 and 7.66). The linear plane of the shearing force is usually apparent. Fractures of the transverse processes, ribs, or both are typically present. There are usually localized pillar and vertebral body fractures on one side [4,7,18]. If flexion is an associated component of a shearing injury, angulation is present at the sites of injury. The posterior vertebral body line is also disrupted. Like rotary injuries, shearing injuries may be confused with burst fractures. As mentioned above, shearing injuries

120

have a greater tendency for lateral dislocation and lateral displacement. Burst fractures show widening of the interpedicle distance, but this is usually minimal. Any displacement in a burst fracture is usually manifested as retropulsed fragments from the posterior vertebral body line. Fractures of the transverse processes, ribs, or both are common in shearing injuries but are rare in burst injuries. The linear oblique, or windswept, appearance of a shearing injury is characteristic on plain radiographs and CT scans. This is in contrast to the linear sagittal distribution of a burst fracture on a CT scan. Figures 7.67 and 7.68 contrast the two injuries.

7 Mechanisms of injury and their “fingerprints” 

A

B

C

D

Fig. 7.63 Shearing injury at L1. (A) The CT scout image shows the upper spine displaced to the left and the lower segments to the right (arrows). The line shows the normal expected position. (B,C) Axial CT images show a “windswept” appearance to the vertebral bodies, severe comminution, canal encroachment, and transverse process fractures on the left (arrow). (D) Sagittal T2-weighted MR image shows fractures of L1 and L2, displacement of bone fragment into the vertebral canal, as well as cord hemorrhage (arrow). The patient was paraplegic.

A

D

B

C

Fig. 7.64 Shearing T11–T12 fracture– dislocation. (A,B) Coronal (A) and sagittal (B) reconstructed CT images show dislocation of T11 to the right and posteriorly. (C) Axial CT image shows the posterior dislocation of T11 and a large fragment of bone filling the vertebral canal at T12. (D,E) Sagittal (D) and coronal (E) T2weighted MR images show cord transection (arrow in D). Note the extensive posterior hemorrhage.

E

121

7 Mechanisms of injury and their “fingerprints” 

A

B

Fig. 7.65 Shearing injuries produce a “windswept” appearance. (A) Photograph of windswept trees. (B) Shearing injury at L1, showing windswept pattern. The arrow indicates the direction of the injuring vector.

Fig. 7.66 Windswept vertebra, CT appearance. The arrow indicates the direction of the injuring vector.

A

B

C

D

Fig. 7.67 Shearing injury with dural tear. (A) Frontal radiograph shows a windswept appearance to the spine (arrows). (B) Lateral radiograph shows the margins between L1 and L2 to be indistinct, indicating overlapping bone fragments. (C) Axial image from a CT myelogram shows widening of the facet joint on the left (*) and extravasation of contrast on the right from a dural tear (arrow). (D) Coronal reconstructed CT myelogram image shows the windswept appearance and the contrast extravasation (arrow).

122

7 Mechanisms of injury and their “fingerprints” 

A

D

B

C

E

Fig. 7.68 Burst fracture of L1. (A) Frontal radiograph shows widening of the interpedicle distance (double arrow). (B) Lateral radiograph shows kyphotic angulation and retropulsion of a bone fragment into the vertebral canal (arrow). (C) Sagittal reconstructed CT image shows the bone fragment in the canal (arrow) as well as compression of L1. There is no dislocation. (D,E) Axial CT images show typical burst pathology with sagittal displacement of bone fragments in D (arrow) and sagittal cleavage in E (arrows). Compare with Figs. 7.64 and 7.67.

123

7 Mechanisms of injury and their “fingerprints” 

In summary, the fingerprints of shearing injuries are as follows (Figs. 7.63 and 7.67): • lateral distraction and lateral dislocation • windswept appearance • fractures of the transverse processes, ribs, or both • linear oblique (windswept) array of fragments on CT scans.

Table 7.1 “Fingerprints” of vertebral trauma Mechanism

Fingerprints

Flexion

Compression, fragmentation and burst fracture of vertebral bodies Teardrop fragments Wide interlaminar (interspinous) space Anterolisthesis Disrupted posterior vertebral body line Locked facets Narrow disc space above involved vertebrae

Extension

Wide disc space below involved vertebrae Triangular avulsion fracture anteriorly Retrolisthesis Neural arch and/or pillar fracture Anterolisthesis with normal interspinous space and spinolaminar line

Rotation

Rotation Dislocation Disrupted posterior vertebral body line Facet and/or pillar fractures or dislocation Transverse process and/or rib fractures Circular array of fragments on CT

Shearing

Lateral distraction Lateral dislocation “Windswept” appearance Transverse process and/or rib fractures

References 1.

2.

3.

4.

5.

6.

124

Whitley JE, Forsyth HF. The classification of cervical spine injuries. AJR Am J Roentgenol 1960;83:633–644. Holdsworth FW. Fractures, dislocations, and fracture–dislocations of the spine. J Bone Joint Surg 1970;52A:1534–1551. Roaf R. International classification of spinal injuries. Paraplegia 1972;10: 78–84. Gehweiler JA Jr., Osborne RL Jr., Becker RF. The Radiology of Vertebral Trauma. Philadelphia, PA: WB Saunders. 1980. Allen BL Jr., Ferguson RL, Lehmann TR, et al. A mechanistic classification of closed, indirect fractures and dislocations of the lower cervical spine. Spine 1982;7:1–27. Ferguson RL, Allen BL Jr. A mechanistic classification of thoracolumbar spine

fractures. Clin Orthop Rel Res 1984: 189:77–88. 7. Daffner RH, Deeb ZL, Rothfus WE. “Fingerprints” of vertebral trauma: a unifying concept based on mechanisms. Skeletal Radiol 1986:15:518–525. 8. Clark WM, Gehweiler JA Jr., Laib R. Twelve significant signs of cervical spine trauma. Skeletal Radiol 1979;3:201–205. 9. Kiwerski J. The influence of the mechanism of cervical spine injury on the degree of spinal cord lesion. Paraplegia 1991;29:531–536. 10. Miller MD, Gehweiler JA, Martinez S, et al. Significant new observations on cervical spine trauma. AJR Am J Roentgenol 1978;130:659–663. 11. Calenoff L, Chessare JW, Rogers LF, et al. Multiple level spinal injuries: importance of early recognition. AJR Am J Roentgenol 1978;130:655–669.

12. Gupta A, El Masri WS. Multilevel spinal injuries: incidence, distribution, and neurological patterns. J Bone Joint Surg 1989;71B:692–695. 13. Powell JN, Waddell JP, Tucker WS, et al. Multiple-level noncontiguous spinal injury. J Trauma 1989;29:1146–1151. 14. Daffner RH, Goldberg AL, Evans TC, et al. Cervical vertebral injuries in the elderly: a 10-year study. Emerg Radiol 1998;5:38–42. 15. Ong AW, Rodriguez A, Kelly R, et al. Detection of cervical spine injury in alert, asymptomatic geriatric blunt trauma patients: who benefits from radiologic imaging? Am Surgeon 2006; 72:773–777. 16. Daffner RH, Daffner SD. Vertebral injuries: detection and implications. Eur J Radiol 2002;42:100–116.

7 Mechanisms of injury and their “fingerprints” 

17. Chance GQ. Note on type of flexion fracture of the spine. Br J Radiol 1948; 21:452–453. 18. Smith WS, Kaufer H. Patterns and mechanisms of lumbar injuries associated with lap seatbelts. J Bone Joint Surg 1969;51A:239–254. 19. Daffner RH, Deeb ZL, Rothfus WE. Thoracic fractures and dislocations in motorcyclists. Skeletal Radiol 1987; 16:280–284. 20. Atlas SW, Regenbogen V, Rogers LF, et al. The radiographic characterization of burst fractures of the spine. AJR Am J Roentgenol 1986;147:575–582. 21. Daffner RH, Deeb ZL, Rothfus WE. The posterior vertebral body line: importance in the detection of burst fractures. AJR Am J Roentgenol 1987; 148:93–96. 22. Lindahl S, Willén J, Nordwall A, et al. The crush-cleavage fracture: a “new” thoracolumbar unstable fracture. Spine 1983;8:559–569. 23. Torg JS, Pavlov H, O’Neill MJ. The axial load teardrop fracture: a biomechanical, clinical, and roentgenographic analysis. Am J Sports Med 1991;19:355–364. 24. Willen JAG, Gaekwad UH, Kakulas BA. Burst fractures in the thoracic and lumbar spine: a clinico-neuropathologic analysis. Spine 1990;14:1316–1323. 25. Kim KW, Chen HH, Russell EJ, et al. Flexion teardrop fracture of the cervical spine: radiographic characteristics. AJNR Am J Neuroradiol 1988;9:1221–1228.

26. McGrory BJ, VanderWilde RS, Currier BL, et al. Diagnosis of subtle thoracolumbar burst fractures: a new radiographic sign. Spine 1993;18: 2282–2285. 27. Harris JH Jr., Mirvis SE. The Radiology of Acute Spinal Trauma, 3rd edn. Baltimore, MD: Williams & Wilkins, 1996. 28. Biffl WL, Ray CE Jr., Moore EE, et al. Noninvasive diagnosis of blunt cerebrovascular injuries: a preliminary report. J Trauma 2002;53:850–856. 29. Cothren CC, Moore EE, Biffl WL, et al. Cervical spine fracture patterns predictive of blunt vertebral injury. J Trauma 2003;55:811–813. 30. Schneider RC, Livingston KE, Cave AJE, et al. “Hangman’s fracture” of the cervical spine. J Neurosurg 1965;22: 141–154. 31. Seljeskog EL, Chous SN. Spectrum of the hangman’s fracture. J Neurosurg 1976;3:45–48. 32. Cintron E, Gilula LA, Murphy WA, et al. The widened disk space: a sign of cervical hyperextension injury. Radiology 1981;141:639–644. 33. Roaf R. A study of the mechanics of spinal injuries. J Bone Joint Surg 1960; 428:810–823. 34. Hendrix RW, Melany M, Miller F, et al. Fracture of the spine in patients with ankylosis due to diffuse skeletal hyperostosis: clinical and imaging

35.

36.

37.

38.

39.

40.

41.

42.

findings. AJR Am J Roentgenol 1994; 162:899–904. Woodruff FP, Dewing SB. Fracture of the cervical spine in patients with ankylosing spondylitis. Radiology 1963; 80:17–21. Altongy JF, Fielding JW. Combined atlanto-axial and occipito-atlantal rotatory subluxation. J Bone Joint Surg 1990;72A:923–926. Fielding JW, Hawkins RJ. Atlantoaxial rotary fixation: fixed rotatory subluxation of the atlanto-axial joint. J Bone Joint Surg 1977;59A:37–44. Klein DM, Kuhn JP. Problems in the radiographic diagnosis of atlanto-axial rotation deformity. Concepts Pediatr Neurosurg 1985;5:26–33. Ono K, Yonenobu K, Fuji T, et al. Atlantoaxial rotatory fixation: radiographic study of its mechanism. Spine 1985;10:602–608. De Oliveira JC. A new type of fracture– dislocation of the thoracolumbar spine. J Bone Joint Surg 1978;60A:481–488. Denis F, Burkus JK. Shear fracture– dislocations of the thoracic and lumbar spine associated with forceful hyperextension (lumberjack paraplegia). Spine 1992;17:156–161. Jeanneret B, Ho PK, Magerl F. Burst-shear-flexion–distraction injuries of the lumbar spine. J Spinal Disorders 1993;6:473–481.

125

Chapter

8

Radiologic “footprints” of vertebral injury: the ABCS Richard H. Daffner

The initial assessment of any patient with suspected vertebral trauma should be by CT or by radiography if CT is unavailable [1]. It is from these studies that the diagnosis is made. Special studies such as MR imaging should be used to delineate the full extent of injury by revealing additional findings not shown on CT or radiographs [1]. Chapter 7 discussed the mechanisms of injury and the radiographic or CT “fingerprints” they produced. This chapter considers the radiologic “footprints” of the injury. Footprints lead you to the injury; fingerprints tell you what kind of injury has occurred and alert you to the possible extent of that injury. The interpretation of any imaging examination demands that a logical system be followed. I prefer to use the ABCS for evaluating vertebral trauma: • A: alignment and anatomy abnormalities • B: bony integrity abnormalities • C: cartilage (joint) space abnormalities • S: soft tissue abnormalities This chapter presents a detailed discussion of the various radiographic and CT findings given in the previous chapters and shows how they are integrated in the overall diagnosis of traumatic lesions of the vertebral column. Although many of the comments that follow pertain to the cervical column, the principles apply to all areas. Furthermore, the principles apply to radiographs as well as to CT studies.

Alignment and anatomy abnormalities Chapter 2 discussed the detailed anatomy of the bones, joints, and ligaments of the vertebral column. A thorough knowledge of normal anatomy and its variants is a prerequisite for interpretation of any imaging study. Normal alignment may be determined on all radiographs as well as on multiplanar reconstructed CT images. Of all of the views used in the evaluation of the vertebral column, the lateral (sagittal reconstructed CT) view is the most important for assessing alignment. Indeed, the first images most radiologists view on a vertebral CT are the sagittal reconstructions; on MR, they are the sagittal T2-weighted images. This is true throughout the vertebral column. Normal markers of alignment on the lateral view include the anterior and posterior margins of the vertebral bodies, the spinolaminar line, the

126

articular pillars and their facet joints, and the interlaminar or interspinous distance (Fig. 8.1). The posterior vertebral body line should be smooth and uninterrupted in the cervical and upper thoracic regions; in the lower thoracic and lumbar regions, it is interrupted centrally by a nutrient foramen. Any rotation, angulation, displacement, duplication, or absence of this line is abnormal [2]. Under normal circumstances, a line drawn along the anterior or posterior margins of the vertebral body should be smooth and uninterrupted. A notable exception is in the cervical column of children, in whom pseudosubluxation occurs because of the disparate growth rates of the various portions of the vertebral column (Fig. 8.2) [3,4]. Similarly, a line connecting the junction of the laminae with the spinous processes (the spinolaminar line or the arch canal line) should be smooth and unbroken. Even with the pseudosubluxation of childhood, however, there should be no disruptions in the spinolaminar line [4]. On a perfectly positioned lateral radiograph, the articular pillars in the cervical region and the articular processes in the thoracic and lumbar regions should be symmetric and should appear as single images with superimposition of the facets. The facets should align like shingles on a roof (imbrication) [5]. Minor degrees of rotation (as evidenced by malalignment of the mandibular image in the cervical region) may result in double facet images. This usually does not present a diagnostic problem. The space between the spinous processes at the level of the spinolaminar line or between the laminae themselves (the interspinous space and the interlaminar space, respectively) should be symmetric and should not vary by more than 2 mm from one level to the next in the neutral or flexed position. In the cervical region, straightening of the neck resulting from the “military” posture usually does not result in dramatic changes in these spaces (Fig. 8.3) [5,6]. In the craniocervical region (Fig. 8.4), there are special considerations related to the anatomy of the area [7–17]. The anterior arch of the atlas bears a constant relationship to the dens. The predental space between these structures should be  no wider than 3  mm in an adult and 5  mm in a child [5,17]. The posterior arches of the atlas merge in the midline to form the  posterior tubercle. This creates a dense arc that aligns with the spinolaminar line. In people in whom fusion

8 Radiologic “footprints”: the ABCS

A

B

A

B

A

B

Fig. 8.1 Normal lateral radiographs. (A) Cervical region. The anterior and posterior margins of the vertebral bodies are uniformly aligned. The posterior vertebral margin is uninterrupted. The spinolaminar line is smooth and unbroken. The interlaminar (interspinous) spaces are uniform. The facet joints are symmetric and the spaces between the posterior margins of the articular pillars and the spinolaminar line are also symmetric. (B) Lumbar region. The anterior margins of the vertebral bodies are smooth and uninterrupted. The posterior vertebral body line is interrupted centrally by a nutrient foramen. There is a mild lordotic curve.

Fig. 8.2 Pseudosubluxation. (A) Apparent malalignment of C2 on C3 resulting from disparity of growth rates between the two vertebrae in a young child. The body of C2 projects anterior to that of C3 (large arrow). The spinolaminar line (small arrows) is uniform. (B) “Physiologic offset” in an adult of C3 on C4 and C4 on C5 caused by reversal of lordosis. The spinolaminar line is normal (arrows).

Fig. 8.3 Reversal of cervical lordosis resulting from a “military” posture, in which the patient’s chin is tucked downward. (A) Lateral radiograph shows the mandible (*) overlying the upper spine as a result of the positioning. There are no abnormalities of the spinolaminar line, facet joints, or soft tissues. There are degenerative changes from C5 downward. (B) Sagittal reconstructed CT image shows the same findings.

127

8 Radiologic “footprints”: the ABCS

A

B

D

E

A

C

Fig. 8.4 Predental space. (A) Normal space (arrow) in an adult. (B) Normal child in whom there is “apparent” widening (*). The spinolaminar line (arrows) is normal. (C–E) Wide predental space. Lateral radiograph (C) shows the widening (*) in association with disruption of the spinolaminar line (arrows). Sagittal reconstructed (D) and axial (E) CT images show the widening (*) and canal compromise to advantage.

B

has not occurred in the posterior arch of the atlas, this dense arc is absent and there is hypertrophy of the anterior arch (Fig. 8.5). Alignment then depends on assessment of the anterior structures. One of the most difficult areas to assess radiographically in the past was the craniocervical junction. Numerous methods had been devised to allow one to make a diagnosis of occipito-atlantal subluxation or dislocation. Two of these will be mentioned only for their historical significance, since CT has rendered them obsolete as radiographic methods. However, they still are valid on sagittal CT reconstructed images. The first of these is the Powers ratio [5,14], which was

128

Fig. 8.5 Failure of fusion of the posterior arch of the atlas. (A) Lateral radiograph shows absence of the spinolaminar line (?) of C1; C2 and C3 have normal spinolaminar lines (arrows). There is hypertrophy of the anterior arch (*). (B) Axial CT image shows the dysraphism (*). There is partial fusion of the anterior arch.

determined by measuring the distances from four points (Fig. 8.6A,C). The first line is drawn from the basion (B) to the midpoint of the arch canal line (C) on the posterior arch of the atlas (line B–C). The second line is drawn from the opisthion (O) to the midpoint of the posterior surface of the anterior arch (A) of the atlas (line O–A). Under normal circumstances, the ratio BC/OA is