Imaging of Vertebral Trauma (Cambridge Medicine)

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] < 15); drowning, immersion, or diving accident; fall greater than 10 feet (3 m); head or severe facial injury; known thoracic or lumber fracture; rigid spine disease (diffuse idiopathic skeletal hyperostosis [DISH] or ankylosing spondylitis); and any conscious patient who complains of paresthesias or burning in the extremities [6]. Vandemark stressed that only one of these criteria needed to be present to indicate that the patient was at high risk for a cervical injury and, therefore, imaging was needed. A decade later, as CT came to be used more frequently for screening for cervical injury in trauma patients, Hanson and colleagues published a slightly different set of indications based on CT using mechanistic and clinical parameters to determine high risk [7]. The mechanistic indicators included high speed, defined as 35 mph or greater; a death at the crash scene; and a fall of greater than 10 feet (3 m). Clinical criteria included closed head injury, neurologic symptoms referable to the cervical region, neck pain or tenderness, and pelvic or multiple extremity fractures [7]. Note the similarities with Vandermark’s criteria. In contradistinction to those earlier studies, which looked at high-risk factors, two additional studies were designed to identify factors that would indicate low probability of cervical injury. The first of these was the National Emergency X-radiography Utilization Study (NEXUS), which reviewed the data from 34 000 patients seen at multiple trauma centers [8]. These investigators concluded that there are five clinical signs that would indicate low risk of cervical injury: normal alertness, no intoxication, no midline tenderness, no focal neurologic deficits, and no painful distracting injuries. I refer to these at the 5 No’s. The second study, by the same group of Canadian researchers who formulated the Ottawa Rules for ankle and knee injuries [3,4] examined data from 8924 patients in 10 large Canadian trauma centers. They formulated what they called the Canadian C-Spine Rule (CCR), applying their findings only to patients who were alert (GCS 15) and clinically stable, with no severe distracting injuries [9]. This “rule” relies on the answers to three questions relating to the traumatic incident and to the victim: (1) Are there any high-risk factors that mandate imaging? (2) Are there any low-risk factors that will allow the safe assessment of the cervical range of motion? (3) Can the patient actively rotate the head 45° to the left and right? They define three high-risk factors: age over 65 years, presence of

46

paresthesias in the extremities, and “dangerous mechanism of injury.” This last parameter is further defined as a fall of greater than 3 feet (1 m) or five stairs (a distance considerably smaller than in the Vandemark criteria), axial loading to the head, diving accident, high-speed motor vehicle crash (>  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 < 15) Multiple fractures Drowning or diving accident Significant head or facial injury Age > 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 <1.0 [5,14]. These lines and measurements are easier to construct on sagittal reconstructed CT images using a PACS system (Fig. 8.6C). An alternative method was described by Lee and associates [15]. In this method (Fig. 8.6B,D), a line (the descending limb) is drawn from the basion to the midpoint of the spinolaminar line of C2. A second line (the ascending limb) is drawn from the posteroinferior corner of the body of C2 to the opisthion. Under normal circumstances, the descending limb should touch the dens or be within 5  mm of it. The ascending limb should pass through or be within 5  mm of the spinolaminar line of C1. Again, these lines are easy to construct on the sagittal CT images (Fig. 8.6D).

8 Radiologic “footprints”: the ABCS

A

B

C

D

Unfortunately, both the Powers ratio and the Lee method have pitfalls, and because of this, they have been supplanted by CT. Both methods depend on accurately locating the opisthion, or posterior margin of the foramen magnum. This landmark may not be readily visible on lateral cervical radiographs because of overaerated mastoids, rotation, poor positioning, or a combination of these factors. This is not a problem on CT. Both methods also require the posterior arch of the atlas to be fused. Furthermore, while both methods work well with anterior subluxations or dislocations, they tend to be inaccurate in posterior displacements. For these reasons, Harris and coworkers [6,11–13] devised an easier method that requires drawing one line and making two measurements. An additional advantage to Harris’ method is that it works on both CT and radiographs, for which it was originally devised. The Harris method entails drawing the posterior axial line (PAL), a cephalad extension of the posterior vertebral body line of C2. The basion–axial interval is the distance between the basion and the PAL (Figs. 8.7 and 8.8). This distance should not be less than 6 mm or more than 12 mm in adults or children [6,11,12]. The basion–axial interval is a reliable measurement independent of the slope of the clivus, anomalies of the posterior elements of C1 or C2, and inclination of the dens. In addition, it is not affected by flexion or extension.

Fig. 8.6 Occipito-atlantal disruption. (A) The Powers ratio is the ratio of the length of a line BC (drawn from the basion [B] to the midpoint of the arch canal line [C] on the posterior arch of the atlas) to the length of a line OA (from the opisthion [O] to the midpoint of the posterior surface of the anterior arch [A] of the atlas). In disruption here, the ratio BC/OA is 1.2 (normal is < 1.0). (B) Lee method show the descending line (D–D’) touching the tip of the dens. The ascending line (A–A’) is 16 mm anterior to the posterior arch of the atlas, instead of touching the arch. (C,D) Although cumbersome, the Powers ratio and Lee method can also be performed on sagittal reconstructed CT images, as here in another patient.

The second measurement in the Harris method is the basion–dens interval. This is the distance between the basion and the most cephalad tip of the dens. The maximum distance in adults and children over the age of 13 is also 12  mm. The basion–dens interval is not accurate in children under 13 years of age because of incomplete ossification and fusion of the dens [6,11–13]. Both of these measurements may be made on sagittal CT reconstructed images. An important landmark on the lateral radiograph of the cervical column is Harris’ ring (Fig. 8.9) [10]. This “ring” is not a real structure but rather results from overlap of images of several portions of the axis. The upper arc represents the superior articular facets; the posterior arc represents the posterior vertebral body line; the inferior arc represents the lower margin of the transverse foramen; and the anterior arc represents the pedicle and anterior body cortex. Low dens or body fractures of C2 frequently disrupt Harris’ ring. In many cases, this may be the only radiographic manifestation of the fracture. This is illustrated in the next section on abnormalities of bony integrity. There is one more important relationship in the upper cervical region that should be recognized: the posterior atlantoaxial relationship. Under normal circumstances, the ratio of the flexion interspinous atlanto-axial distance divided by the

129

8 Radiologic “footprints”: the ABCS

A

B

C

Fig. 8.7 Occipito-atlantal disruption. Harris method. (A) Drawing showing the tip of the basion (*) 6–12 mm from a line drawn along the posterior aspect of the dens. (B) Lateral radiograph showing the dens–basion distance is 18 mm (normal is 6–12 mm). (C) Sagittal reconstructed CT image in another patient shows similar findings.

130

Fig. 8.8 Occipito-atlantal dislocation. Lateral radiograph shows gross widening of the craniovertebral junction (*). The skull is dislocated anteriorly; note the “naked” facet of the lateral mass of C1.

Fig. 8.9 Normal lateral cervical radiograph showing the four components of Harris’ ring: 1, superior articular facet; 2, posterior vertebral body line; 3, inferior margin of transverse foramen; 4, portion of anterior vertebral body.

height of the atlantal spinolaminar line should be 2 or less [16]. In absolute distances, the space should not exceed 18 mm. On frontal views (Fig. 8.10), alignment can be assessed by looking at the lateral margins of the vertebrae, the pedicles, the spinous processes, the uncinate processes in the cervical region, and the facet joints in the thoracic and lumbar regions [5,6]. Under normal circumstances, lines drawn along each of these margins should be straight and uninterrupted. There should be less than 2 mm of difference from one level to the next for analogous structures. In the thoracic and lumbar regions, the difference in the transverse distance between the pedicles (interpedicle distance) should be less than 2 mm from level to level. Any deviation of more than 2 mm should be considered abnormal (Fig. 8.11). Similarly, the differences in the vertical interpedicle distances should be symmetric and less

than 2 mm at each level. When the spinous processes are used to assess alignment, both rotation off the midline and widening of the interspinous space are noteworthy [5,6,18–22]. The frontal view is ideal for this purpose. As on the lateral view, there are additional considerations in the craniocervical region. Under normal circumstances, the inion of the occipital bone (internal occipital protuberance), the dens, and the spinous processes of C2 and C3 should align [5,6]. Minor degrees of rotation result in differences in the distance between the dens and the lateral masses of C1. When this is encountered, it is important to check the other midline alignment points in this area to determine whether the abnormality reflects rotation rather than ligamentous disruption. On CT, this asymmetry is commonly a result of patient positioning. However, if a patient with this finding complains of pain or is

8 Radiologic “footprints”: the ABCS

A

A

B

Fig. 8.10 Normal frontal views: cervical (A) and lumbar (B). The spinous processes should be in normal alignment, and the distances between them (* in A) should be uniform. The spaces between the articular pillars should also be uniform. The interpedicle distances (double arrows) are uniform, within 2 mm of the neighbors. The vertical interpedicle distances should also be uniform. These same parameters apply to the thoracic region as well as the lumbar.

B

Fig. 8.11 Burst fracture of L1. (A) Frontal radiograph shows widening of the interpedicle distance (double arrow). (B) Axial CT image shows sagittal cleavage fractures through the body and lamina (arrows).

tender in the craniocervical junction, MR imaging should be performed to rule out a ligamentous injury. The lateral masses of C1 normally align with the body and lateral aspect of C2 (Fig. 8.12). In children, there can be up to 3 mm of disparity in the lateral atlanto-axial border because of differences in growth rate between the two bones (Fig.  8.13) [3,4,8,23]. In adults, up to 2 mm of lateral atlanto-axial overlap can be considered normal if the patient has an anomaly of C1 (Fig. 8.14) [8]. Malalignment by more than 3 mm in children or adults is usually encountered in a Jefferson-type burst fracture (Fig. 8.15) [5,6,8,24,25]. There are two additional considerations on frontal radiographs. One should always look at the occipital condyles in the atlanto-axial views to make sure that they are smooth and uninterrupted. Fractures of these structures, which were

Fig. 8.12 Normal atlanto-axial (open-mouth) view. The lateral alignment between the lateral masses of C1 and the body of C2 should be uniform. Up to 2 mm of unilateral or bilateral lateral offset is allowed in the presence of a congenital anomaly of C1. Minor differences in the width of the space between the dens and lateral masses of the atlas may occur with rotation. In this instance, they are equal (*).

thought to be uncommon, may occasionally be detected on radiographs [26]. Use of CT now shows us that they are not that unusual. The second consideration involves the transverse processes. Fractures of these structures are often associated with vertebral artery and cranial nerve injuries in the cervical region. In the thoracic and lumbar regions, they may be associated with visceral injury. Again, transverse process fractures are easily demonstrated by CT. Oblique views, although infrequently used for trauma patients today, may also be used to assess alignment (Fig. 8.16). The anterior and posterior vertebral body margins should normally align. A line drawn along the anterior or posterior margins of the pedicles should be smooth and uninterrupted, as should a line drawn across the laminae. In addition, the facet joints should be symmetric in their alignment, in much the

131

8 Radiologic “footprints”: the ABCS

A

Fig. 8.13 Normal atlanto-axial malalignment in children. Radiographs of two children show malalignment of the lateral masses of C1 with respect to the body of C2 (arrows). This occurs because of differences in the growth rates of the two bones and the disparity is often striking. As the child grows, the relationship normalizes.

B

A

B

A

B

C

D

same way that shingling overlaps on a roof (imbrication) [5]. In the cervical region, the intervertebral foramina and their margins should be symmetric. Alignment abnormalities can be easily detected because of the disruptions they produce in the structures or lines described above. The most common abnormality is anterolisthesis, which is found frequently in flexion injuries (Figs. 8.17 and 8.18)

132

Fig. 8.14 Bilateral atlantoaxial offset in a patient with failure of fusion of the posterior arch of C1. (A) Frontal tomogram shows malalignment of the lateral masses of the atlas on the body of C2 (arrows). (B) Lateral radiograph shows absence of the spinolaminar line at C1. Note the normal spinolaminar line at C2 (arrow) and hypertrophy of the anterior arch of the atlas (*). Fig. 8.15 Jefferson fracture of C1. (A) Open-mouth view shows bilateral lateral atlanto-axial malalignment (arrows). (B) Lateral view shows fractures in both the anterior and posterior arches of C1 (arrows). (C,D) Axial CT images show fractures of the anterior and posterior arches (arrows).

[19,27]. The findings often indicate combined abnormalities, and anterolisthesis is usually accompanied by widening of the interlaminar or interspinous space and widening of the facet joints (Figs. 8.17 to 8.19). For example, in a patient with unilateral facet lock (Fig. 8.20), in addition to anterolisthesis, the spinous processes of the abnormal vertebra and those above are rotated toward the side of the locked facet on the frontal

8 Radiologic “footprints”: the ABCS

A

Fig. 8.16 Normal oblique views. (A) Cervical. There is normal alignment of the vertebral bodies, articular pillars, and transverse process images. The intervertebral foramina are uniform. (B) Lumbar. There is normal alignment of the vertebral bodies, pedicles, and articular facets. The pars interarticularis (*) is intact.

B

A

Fig. 8.17 Hyperflexion sprain at C4–C5, showing kyphotic angulation and widening of the interlaminar (interspinous) distance (*). There is slight anterolisthesis of C4 on C5.

B

Fig. 8.18 Unilateral facet lock C6 on C7. (A) Lateral radiograph shows anterolisthesis of C6 on C7 (large arrow). Note the abrupt narrowing of the space between the articular pillars and the spinolaminar line at C6 (small arrows). (B) Frontal radiograph shows the spinous processes from C6 and above rotated to the left, the side of the lock, and widening of the interspinous space (double arrow).

133

8 Radiologic “footprints”: the ABCS

A

B

C

Fig. 8.19 Hyperflexion sprain C5–C6. (A) Lateral radiograph shows slight anterolisthesis of C5 on C6 with widening of the posterior C5 disc space as well as the interlaminar (interspinous) space (*). (B) Frontal view shows widening of the interspinous space (double arrow). (C) The T1-weighted sagittal MR image shows rupture of the posterior longitudinal ligament (arrow).

134

A

B

A

B

Fig. 8.21 Unilateral L5–S1 facet lock on the right. (A) Frontal radiograph shows rotation of the spinous processes (long line) to the right. The spinous process of S1 (short line) is midline. (B) Lateral radiograph shows anterolisthesis of L5 on S1. Note the “naked” facets (arrows). (C) The CT axial image shows the point of locking on the right (arrow). (With permission from Daffner [29].)

Fig. 8.20 Unilateral facet lock at C5–C6. (A) Lateral radiograph shows anterolisthesis of C5 on C6. There is a small displaced fragment of bone from the inferior aspect of the dislocated facet of C5 (arrow). (B) Frontal radiograph shows widening of the interspinous space between C5 and C6 (double arrow). There is subtle right shift of the spinous processes of C5 and above, indicating that the side of locking is on the right.

C

8 Radiologic “footprints”: the ABCS

A

B

Fig. 8.22 Unilateral facet lock C4–C5. (A) Lateral radiograph show anterolisthesis of C4 on C5, with a “bowtie sign” (arrows) as a result of the rotation of the vertebrae. (B) Axial CT image shows the locked facet (arrow) and a “reverse hamburger bun sign.”

Fig. 8.23 Unilateral facet lock at C5–C6. Lateral radiograph shows anterolisthesis, abrupt duplication of the facet images at C5 and above (arrows), and abrupt narrowing of the laminar space between the spinolaminar line and the pillar images (arrowheads). The pillars have a “bowtie” appearance, best appreciated at C4, as a result of the dislocation.

condition from simple positional rotation. One other sign of unilateral facet lock is abrupt narrowing of the laminar space between the spinolaminar line and the articular pillar (Figs. 8.22 and 8.23) [27,32]. On CT, the normal facets at any level should resemble a hamburger bun with the flat articular surface central (Figs. 8.24 and 8.25) [33]. Unilateral or bilateral facet jump results in an abnormal appearance to the facet joint with the convex, non-articular surfaces in approximation and the flat articular surfaces on top and bottom. We call this the “reverse hamburger bun sign” (Fig. 8.26) [33].Unilateral jumped and locked facets are the most frequently missed cervical injuries for which lawsuits are filed. In most cases, the radiologist thought the study “looked funny.” Box 8.1 lists the radiographic findings of unilateral facet lock. Fig. 8.24 Drawing showing the basis for the “hamburger bun sign” and “reverse hamburger bun sign.” On a CT scan, the facet joints normally resemble a hamburger. Facet dislocation produces reversal of the “bun halves.” (With permission from Daffner and Daffner [33].)

radiograph [5,6,18,19,27–30]. These findings may be found anywhere unilateral facet lock occurs (Fig. 8.21) [27–31]. In addition to anterolisthesis, the lateral view in such a patient shows duplication of the facet images above the level of lock, to give the appearance of a bowtie (Fig. 8.22) [27,32]. Below the level of lock, there is the normal overlap of both facet images. Above this level, two distinct facet images are seen (Figs. 8.20 and 8.23). The transition is abrupt, which differentiates this

Box 8.1. Radiographic findings in unilateral facet lock Anterolisthesis Wide interlaminar space Abrupt duplication of pillar images ”Bowtie” sign Spinous process rotated toward side of lock ”Reverse hamburger bun sign” on CT Narrow laminar space

Additional abnormalities may be detected by observing the facet images and the spinolaminar line. If a patient has a dislocation with locking of the facets, and if the spinolaminar

135

8 Radiologic “footprints”: the ABCS

A

B

Fig. 8.25 “Hamburger bun sign.” (A) All-American hamburger. (B) Axial CT image shows the resemblance of the normal facet joints (arrows) to a hamburger.

A

B

C

Fig. 8.26 “Reverse hamburger bun sign” (arrows) in facet dislocation. (A) Reverse bun halves. (B,C) Axial CT images in a patient with bilateral facet lock shows the “reverse hamburger bun sign” (arrows).

line is in its normal position, bilateral fractures are present in either the pedicles or laminae. It is easy to understand this finding because the spinolaminar line would be disrupted if the posterior arch were intact (Fig. 8.27). Degenerative anterolisthesis or retrolisthesis may cause diagnostic difficulty in the traumatized patient. Lee and associates [34] showed that the articular facets are “ground down” in degenerative slippage, with accompanying narrowing of the facet joint space. In traumatic subluxation, the articular facets are either normal or fractured and the facet joint spaces are abnormally widened (Figs. 8.17 and 8.19). As a rule, most degenerative listheses are 2 mm or less. Flexion and extension views are recommended to determine whether there is any motion at these abnormal levels; the abnormality usually is fixed. Lateral flexion injuries in the cervical region may result in an isolated crush fracture to an articular pillar. [35] This unusual injury can be detected by observing rotation of the pillar images on lateral, frontal (and oblique) radiographs (Fig.  8.28), [5,35,36] or, ideally, by CT (Figs. 8.28 and 8.29).

136

Kyphotic angulation and loss of lordosis are additional findings that may suggest underlying injury, particularly if associated with other findings [31]. Kyphotic angulation and loss of lordosis more often result from positioning or from spasm linked to soft tissue injury, however, and caution is needed in interpreting such findings. Flexion and extension films are recommended for these patients. These views may have to be repeated if kyphosis persists when the patient’s pain and spasm have subsided [37]. A common injury is the hyperflexion sprain, in which the posterior ligaments are ruptured [5,6,38]. In these patients, supine lateral radiographs, and even CT images, at first glance appear normal. However, subtle findings of anterolisthesis, kyphotic angulation, and widening of the interlaminar or interspinous space may be found on radiographs and CT (Fig.  8.30). An MR image shows that this injury is significant in that there is demonstrable ligament damage with or without disc herniation (Fig. 8.31). Hyperflexion sprain is to be differentiated from positional loss of lordosis or lordosis

8 Radiologic “footprints”: the ABCS

A

B

C

Fig. 8.27 Bilateral facet lock. (A) Lateral radiograph shows anterolisthesis of C5 on C6. (B) Sagittal reconstructed CT image shows locking of the facets. A fragment of the inferior facet of C5 remains in its normal anatomic position (arrow). (C) Axial CT image shows the bilateral locked facets and “reverse hamburger bun signs” (arrows). A

B

Fig. 8.28 “Floating” pillar. (A) Lateral radiograph shows elevation of a duplicated pillar image (upper arrows). The normal position of the opposite pillar is shown (lower arrow). (B) Axial CT image shows fractures of the right pedicle and lamina with rotation of the pillar (arrows), accounting for the “floating” appearance.

A

B

C

Fig. 8.29 Fractured articular pillar of C5 on the left in a patient with unilateral facet lock at C5–C6. Sagittal reconstructed (A), axial (B), and coronal reconstructed (C) CT images show the pillar fracture (arrow). Note the locking (arrowhead in A). Note rotation of the spinous process to the left in B.

137

8 Radiologic “footprints”: the ABCS

caused by simple muscle spasm without ligamentous injury. In these patients, there is an abnormal vertebral curvature with no disruption of the spinolaminar line or widening between the spinous processes (Fig. 8.32). Similarly, loss of lordosis in the lumbar area may be the result of muscle spasm rather than underlying injury; in most cases, if an injury has occurred, there are other radiographic manifestations. Finally, torticollis in the neck produces alignment abnormalities, but this is a nonspecific sign and is usually the result of muscle spasm. Box 8.2 summarizes alignment abnormalities.

A

B

Fig. 8.31 Hyperflexion sprain (same patient as in Fig. 8.30). Sagittal T2-weighted MR image shows rupture of the posterior longitudinal ligament (arrow).

138

Box 8.2. Alignment abnormalities Disruption of anterior or posterior vertebral body lines Disruption of spinolaminar line Jumped or locked facets Rotation of spinous processes Wide interlaminar space Wide interpedicle space Wide predental space Acute kyphotic angulation Loss of lordosis Torticollis

Fig. 8.30 Hyperflexion sprain. (A) Lateral radiograph shows kyphotic angulation at C4–C5 and widening of the interlaminar (interspinous) distance (*). (B) Frontal radiograph shows the wide interpinous space (double arrow).

Fig. 8.32 Reversal of lordosis caused by muscle spasm. The normal lordotic curve is lost. No other structures are malaligned in this otherwise normal lateral cervical column. Sagittal reconstructed CT images can show similar findings.

8 Radiologic “footprints”: the ABCS

Bony integrity abnormalities Any disruption in a bone indicates a fracture. There are subtle findings that aid in diagnosis, however. Four of these signs are disruption of the posterior vertebral body line [2,27], wide interpedicle distance [5,6,21,27,39,40], disruption of the C2 body ring [10], and disruption of the arcuate lines of the sacrum [41]. As described above, the posterior border of the vertebral body is represented by a single vertical line in the cervical and upper thoracic regions (Fig. 8.33A) and a single vertical line with central interruption in the lower thoracic and lumbar regions (Fig. 8.33B) [2]. This central interruption is the site of the basivertebral vein as it traverses the vertebral body. Any other interruption, duplication, displacement, rotation, angulation, or absence of this line is abnormal. This is particularly important in patients who have suffered flexion or rotary injuries; any of these findings indicates either a burst fracture or a rotary fracture. Burst and rotary fractures usually produce retropulsion of one or more fragments from the posterior aspect of the vertebral body (Figs. 8.34 to 8.38). Disruption of the posterior vertebral body line on a radiograph indicates the extent of injury and, when combined with additional alignment abnormalities, is an indicator of an unstable injury [2,5,20,27,31,42]. Burst injuries tend to explode the vertebral body and separate it from its posterior elements. One of these manifestations, particularly in the thoracolumbar region, is widening of the transverse interpedicle distance (Figs. 8.39 and 8.40) [18,19,35,36,39], which can be easily discerned by measuring the distance between the pedicles. This is thought by Martijn and Veldhuis [43] to be a reliable finding of severe trauma. Widening also indicates posterior disruption, since the

A

B

lamina must be cleft in order for this to occur. In most cases, additional findings indicative of burst injury are present, including comminution of the vertebral body, disruption of the posterior vertebral body line, and occasionally widening of the interspinous distance [2,5,19,21,27]. McGrory and colleagues [44] described a new sign of thoracolumbar burst fracture in which they measured the angle of the posterior vertebral body cortex with the vertebral endplate. They found that any angular measurement of more than 100° indicates displacement of the posterior vertebral body line [44]. They did not give the lower limit measurement. The increased angulation is a function of the displacement of the posterior vertebral body line. On a lateral radiograph of C2, the ringlike density (Harris’ ring) at the base of the dens is often disrupted in low dens and axis body fractures (Figs. 8.41 and 8.42) [6,10,45,46]. These fractures are often difficult to diagnose on radiographs because the bony fragments are seldom displaced. The dens does not contribute to the ring. Hence, most dens fractures do not disrupt the ring unless they extend into the body. CT easily identifies these fractures. There is some confusion about the classification of dens fractures, particularly regarding the widely used AndersonD’Alonzo classification [47]. These researchers originally described three types of fracture. Type I was reported as an avulsion fracture of the tip of the dens as a result of tension of the alar ligaments between the atlas and the axis. Type II is a horizontal fracture at the base of the dens. Type III is a fracture of the base of the dens that extends into the body. A careful review of the original paper by Anderson and D’Alonzo and their illustrations strongly suggests that both of the type I fractures presented were, in fact, pseudofractures (see Chapter 10). One case is a terminal ossicle, a normal variant, which was Fig. 8.33 Normal posterior vertebral body lines (arrows). (A) Cervical. The lines are unbroken. (B) Lumbar. There is interruption of the center of the line by a nutrient vessel.

139

8 Radiologic “footprints”: the ABCS

A

B

C

Fig. 8.34 Burst fracture of T12. (A) Lateral radiograph shows compression of L1, kyphotic angulation, and posterior displacement of a fragment from the posterior vertebral body line (arrow). (B) Sagittal reconstructed CT image shows similar findings, with the displaced fragment (arrow) impinging the vertebral canal. (C) Axial CT image shows the displaced fragments (*) in the canal.

A

B

C

Fig. 8.35 Burst fracture of L2. (A) Lateral radiograph shows compression of the vertebral body and displacement of a fragment (arrow) into the vertebral canal. (B) Sagittal reconstructed CT image shows the canal compromise by the displaced fragment (arrow). (C) Axial CT image shows the displaced fragment (*) narrowing the canal.

140

8 Radiologic “footprints”: the ABCS

A

Fig. 8.36 Burst fracture of C6. (A) Lateral radiograph shows duplication of portions of the posterior vertebral body line (arrows). (B) Axial CT image shows fragmentation of the body of C6 and displacement of a fragment into the canal on the left (arrow). This fragment accounted for the double body lines seen in A.

B

A

B

C

Fig. 8.37 Subtle compromise of the vertebral canal in a burst fracture of L3. (A) Axial CT image shows flattening of the posterior border of the vertebral body (arrow). (B) Axial CT image slightly higher shows the normal concavity of the posterior vertebral body (arrow). (C) Sagittal reconstructed CT image shows the coronal fracture of the vertebral body and displacement of the lower posterior body line into the vertebral canal (arrow).

A

B

Fig. 8.38 Double posterior vertebral body line in a patient with a hanged-man fracture of C2. (A) Lateral radiograph shows two vertical lines representing the posterior border of C2 (arrows). There is anterolisthesis of C2 on C3. (B) Axial CT image shows fracture through the posterior body of C2 as well as through the lamina on the right (arrowhead). This combination of injuries has allowed the main portion of the body of C2 to slip forward, creating the double posterior body image (arrows).

141

8 Radiologic “footprints”: the ABCS

A

B

C

Fig. 8.39 Burst fracture of L1 (same patient as in Fig. 8.34). (A) Frontal radiograph shows the wide interpedicle distance (double arrow). (B,C) Coronal reconstructed (B) and axial (C) CT images show a sagittal fracture of the body of L1 (arrows). This finding can occur only as a result of sagittal fractures of the body and the posterior elements, as in C.

A

C

142

Fig. 8.40 Burst fracture of L1. (A) Frontal radiograph shows compression of the body of L1 and widening of the interpedicle distance (double arrow). (B) Sagittal reconstructed CT image shows typical burst pathology with displacement of a bone fragment (arrow) into the vertebral canal. (C) Axial CT image shows two bone fragments in the canal (arrows). (D) Axial CT image slightly lower shows sagittal fractures through the body and lamina (arrows).

B

D

8 Radiologic “footprints”: the ABCS

A

A

C

Fig. 8.41 Fracture of body of C2. (A) Lateral radiograph shows Harris’ ring to be indistinct (?). (B) Axial CT image shows severe comminution of the body (arrows).

B

B

Fig. 8.42 Fracture of body of C2. (A) Lateral radiograph shows disruption of the posterior body line near the base of the dens (arrow). Harris’ ring is indistinct (?). (B) Sagittal reconstructed CT image shows fractures of the anterior and posterior body of C2 (arrows). (C) Axial CT image shows fracture through the right side of the body of C2 (arrows).

143

8 Radiologic “footprints”: the ABCS

A

C

B

Fig. 8.43 Dens fracture with anterior displacement. (A) Lateral radiograph shows malalignment of the spinolaminar lines of C1 and C2 (arrows). Harris’ ring is indistinct. There is prevertebral soft tissue swelling (*). (B) Sagittal reconstructed CT image shows the dens fracture and anterior displacement (large arrow). Note the malalignment of the spinolaminar line (small arrows). (C) Axial CT image shows comminuted fractures through the body of C2, accounting for the changes in Harris’ ring.

reported to go on to “non-union.” The other is a Mach band caused by the image of the occiput on the upper tip of the dens. Not surprisingly, this “fracture” subsequently was shown to have “healed.” A review by Burke and Harris [46] supports the observation of many radiologists (myself included) at large trauma centers that the type I dens fracture does not exist. Furthermore, type III dens fractures should be referred to as axis body fractures. True dens fractures are of two varieties: low, at the base, and high, above the base. They may be associated with anterolisthesis (Fig. 8.43) or retrolisthesis (Fig. 8.44). Occasionally, on a frontal open-mouth radiograph or coronal CT reconstruction, dens fractures show lateral angulation. This is determined by measuring the angle between a horizontal line drawn between the most lateral aspects of the superior margins of the superior articular facets and a vertical line drawn to bisect the shaft of the dens. A dens angle of 87 to 90° is normal. Any deviation in excess of 5° from the vertical (or less than 85° from the horizontal) should raise the suspicion of a dens fracture, and CT should be performed, if it has not already been obtained [48]. Smoker and Dolan [49] described another sign of a fracture of the body of C2, the “fat” C2 sign. This sign results from a shift of fracture fragments to produce apparent sagittal widening of the body of C2 in relation to C3. Vertical fractures of the body of the axis are most likely to produce this finding [49,50]. The fat

144

C2 sign is a valuable finding for the identification of subtle C2 fractures (Figs. 8.45 and 8.46), which can be confirmed by CT. Fractures of the upper sacrum are common findings in pelvic fractures. They are frequently overlooked, however, because of the superimposition of intestinal gas and contents and because of the complex configuration of the sacrum. Jackson and coworkers [41] described abnormalities of the arcuate lines of the sacrum that would indicate fractures. These archlike structures are easily visible on frontal radiographs of the pelvis and abdomen and represent the inferior surfaces of the bony struts that form the roofs of the anterior sacral foramina. Generally, the arcs of the first three segments are easily identified, but sometimes only the first two are visible. Fractures to the body of the sacrum result in disruption of these arcuate lines (Figs. 8.47 and 8.48). Changes include frank disruption, angulation, and obliteration; CT will confirm the presence of a fracture [41,51]. Box 8.3 summarizes abnormalities of bony integrity. Box 8.3. Bony integrity abnormalities Obvious fracture Disruption of “ring” of C2 Fat C2 sign Widening of interpedicle distance Disruption of posterior vertebral body line Disruption of sacral arcuate lines

8 Radiologic “footprints”: the ABCS

A

B

A

B

A

B

Fig. 8.44 Dens fracture with posterior displacement. (A) Sagittal reconstructed CT image shows a fracture at the base of the dens (*) with posterior displacement. (B) Sagittal STIR MR image shows impingement of the dens fragment on the spinal cord with extensive hemorrhage in the cord (arrows).

Fig. 8.45 Fat C2 sign. (A) Lateral radiograph shows the width of C2 (21.9 mm) to be greater than that of the other cervical vertebrae (18.2 mm) but a fracture cannot be seen. (B) Axial CT image shows bilateral fractures (arrows). Minimal separation of the fragments caused C2 to appear larger than its mates.

C

Fig. 8.46 Fat C2 sign. (A) Lateral radiograph shows an increase in the width of the body of C2 (arrows). There is disruption of Harris’ ring. (B) Sagittal reconstructed CT image shows displacement of a fragment from the posterior body of C2, which resulted in the “fat” appearance (arrows). (C) Axial CT image shows the coronal comminuted fracture that allowed widening of the body of C2.

145

8 Radiologic “footprints”: the ABCS

A

C

A

B

Fig. 8.47 Sacral arcuate lines. (A) Detail of a pelvic radiograph shows the normal sacral arcuate lines (arrows). There are anomalies of L5. (B) Disrupted arcuate lines (arrows) in a patient with a pelvic fracture. (C) Disrupted arcuate lines (arrows) in another patient. There is a fracture of the right pubic bone (arrowhead).

B

Cartilage (joint) space abnormalities Injury to the vertebral column may change the width of the joint spaces. Increases in joint spaces are more common than decreases and include widening of the intervertebral disc space, of the interlaminar or interspinous distance, of the interpedicle distance, of the predental space, and of the facet joints [27]. All of these findings indicate severe ligamentous disruption. Conversely, joint spaces may also be narrowed. The most common traumatic manifestation of this occurs with flexion injuries, in which the disc space above the compressed

146

Fig. 8.48 Subtle sacral fracture. (A) Pelvic radiograph shows disruption of the sacral arcuate lines on the left (arrow). (B) Pelvic CT shows a buckle fracture of the anterior sacral cortex (arrow).

vertebra is narrowed (Fig. 8.49). Degenerative disc disease, not trauma, is the most common cause of disc space narrowing, however. Widening of the intervertebral disc space is a common result of extension injuries [52] (Figs. 8.50 and 8.51). This radiographic manifestation indicates damage to the anterior longitudinal ligament and to the disc itself. Any wide disc space must be viewed with suspicion, especially when found in association with degenerative changes in an older patient [27]. If there are no other radiographic findings of vertebral injury, carefully

8 Radiologic “footprints”: the ABCS

Fig. 8.49 Compression fracture of L1. The anterior aspect of the body of L1 is flattened and the T12 disc space is narrowed (*).

A

Fig. 8.50 Extension injury C5–C6. There is widening of the C5 disc space (*) and retrolisthesis.

B

supervised flexion and extension films should be obtained to determine whether the wide disc space is a fixed deformity. If these cannot be performed successfully, MR imaging is the procedure of choice. Two other phenomena may be observed in association with traumatic disc space changes. The first is the vertebral annulus vacuum [37]. This is a small linear lucency near the anterosuperior or anteroinferior margin of a (cervical) disc. Such lucencies were once believed to indicate degeneration of the annulus fibrosus. Bohrer and Chen [37] demonstrated two types of annulus vacuum. The first results from degenerative disease and is generally associated with osteophyte formation and calcification along the annulus margin, usually along the anterosuperior vertebral corners. If these corners are deformed, the vacuum may indicate an old injury. Small annulus vacuums (Fig. 8.52) adjacent to normal-appearing vertebral

Fig. 8.51 Extension injury in a patient with diffuse idiopathic skeletal hyperostosis (“broken DISH”). (A) Lateral radiograph shows widening of the C3 and C4 disc spaces (arrows). (B) Sagittal reconstructed CT image shows the widened disc spaces are associated with body fractures (arrows).

corners are more likely to be the result of acute trauma. This injury is called a vertebral disc bond injury [37]. In children and adolescents, the same mechanisms that produce the annulus vacuum in adults may result in avulsion of the cervical ring apophysis [53]. Flexion mechanisms are more likely to avulse the superior apophysis and produce bow-shaped deformities along the anterosuperior endplate. Extension mechanisms produce inferior avulsions that result in osteophyte formation when healed [53]. Widening of the interlaminar or interspinous space is technically not a joint abnormality. Nevertheless, this finding typically accompanies severe flexion injuries and is associated with abnormalities in the facet joints themselves (Figs. 8.53 and  8.54) [5,18,20,27,54,55]. In addition to being visible on frontal and lateral radiographs, such widening is apparent on CT scans by the absence of the spinous process on two

147

8 Radiologic “footprints”: the ABCS

148

Fig. 8.52 Annulus vacuum (arrow) caused by an extension injury.

Fig. 8.53 Widening of the interspinous space (double arrow) in a patient with a hyperflexion sprain at C6–C7.

Fig. 8.54 Widening of the interspinous space (*) in a patient with a Chance-type fracture. Note the similarity to the findings in Fig. 8.53. There is a fracture of the transverse process on the left (arrow).

Fig. 8.55 “Naked” facets (arrow). Axial CT image shows absence of any additional posterior elements (*) in this patient with a Chance-type fracture. (Same patient as in Fig. 8.54.)

contiguous sections (Fig. 8.55) [54]. This finding should be reported whenever encountered on thorax–abdomen–pelvis CT studies. As mentioned above, the distance between the pedicles should not vary from one vertebral level to the next by more than 2 mm. This applies to both the transverse plane and the vertical plane. An increase in the transverse distance is an indication that the vertebra has been disrupted in the sagittal plane with resultant bilateral lateral separation of the major fragments. This typically occurs in burst fractures (Figs. 8.56 and 8.57; see also Fig. 8.40) and occasionally in rotary injuries (Fig. 8.58). Widening of the interpedicle distance is also easily

observable on frontal radiographs, and this finding usually accompanies widening of the facet joints of the involved vertebrae (Figs. 8.57 and 8.59). The typical findings of a burst injury are usually present, including fragmentation, widened interspinous space, disrupted posterior vertebral body line, retropulsion of bone fragments, and narrowing of the superior disc space. Disruption in the sagittal plane typically results from a flexion injury. Rotational and shearing injuries (see Chapter 7) often produce similar-appearing changes. The predental space may be widened from injuries involving the “check” ligaments between C1 and C2. The predental space is measured from the anterior aspect of the dens to the

8 Radiologic “footprints”: the ABCS

A

B

A

B

posterior aspect of the anterior arch of the atlas. This measurement should not exceed 3 mm in adults or 5 mm in children (Fig. 8.60) [5,17,19]. Widening of the predental space is found most commonly in patients with severe rheumatoid arthritis in whom synovial proliferation has resulted in destruction or disruption of the ligamentous attachments about the dens (Fig. 8.61). True traumatic widening of the space is unusual in adults (Fig. 8.62) but more common in children. The facet joints are often disrupted with severe flexion injuries [19,27,54]. In the typical case, there is forward subluxation or dislocation of these facets (Fig. 8.63). Complete disruption may result in unilateral (Fig. 8.64) or bilateral (Fig.

Fig. 8.56 Burst fracture of L3. (A) Frontal radiograph shows widening of the interpedicle distance (double arrow). (B) Axial CT image shows a fragment displaced into the vertebral canal (*) and a fracture through the lamina (arrow).

Fig. 8.57 Burst fracture of L1. (A) Frontal radiograph shows widening of the interpedicle distance (double arrow). (B) Axial CT image shows comminution of the vertebral body, retropulsion of bone fragments into the vertebral canal, and widening of the facet joint on the left (arrow).

8.65) facet lock. Occasionally, the facet abnormality may be the only indication of an injury (Fig. 8.66). Lee and Woodring [36] observed that lateral flexion mechanisms in the cervical region may produce sagittally oriented fractures of the articular pillars. Only two thirds could be seen on lateral films; all were seen on frontal films. There were two varieties: type I (42%), with a fracture through the articular pillar and lateral subluxation of the fragment; and type II (58%), in which fractures of the pedicle and lamina produced a “floating” pillar (Fig. 8.66). Pillar fractures are easily demonstrated by CT and have a high propensity (45%) to produce not only localized neurologic findings [36] but also vascular injuries (Fig. 8.67).

149

8 Radiologic “footprints”: the ABCS

A

Fig. 8.59 Burst fracture of L1. Frontal radiograph shows widening of the T12–L1 facet joint on the left (arrow). The interpedicle distance is also wide.

A

150

Fig. 8.58 Rotary injury of L1. (A) Frontal radiograph shows widening of the interpedicle distance as well as of the interspinous distance (double arrow). The T12–L1 facet joint on the left is wide (arrow). (B) Axial CT image shows severe comminution of the vertebral body, with a “naked” facet on the right (arrow) and widening of the facet joint on the left (arrowhead).

B

Fig. 8.60 Wide predental space (*) in a patient with os odontoideum (arrow).

B

Fig. 8.61 Wide predental space in a patient with rheumatoid arthritis. This is the most common etiology of this finding. (A) Lateral radiograph shows anterolisthesis of C1 on C2 with disruption of the spinolaminar line (arrows) in addition to the wide predental space (*). (B) Sagittal reconstructed CT image shows settling of the atlas on the axis and protrusion of the dens (arrow) through the foramen magnum. Note the shift of the anterior (A) and posterior (P) arches of the atlas with severe canal compromise and the wide predental space (*).

8 Radiologic “footprints”: the ABCS

A

B

Fig. 8.62 Wide predental space (*) following trauma. (A) Lateral radiograph. (B) Axial CT image.

Fig. 8.63 Flexion teardrop injury at C5. Lateral radiograph shows widening of the facet joints of C5–C6 (arrows).

Fig. 8.64 Unilateral facet lock (arrow) of C6–C7. Lateral radiograph shows slight anterolisthesis of C6 on C7 with abrupt duplication of the pillar images.

In general, facet abnormalities occur with other findings that indicate a flexion injury. “Naked” facets indicate severe ligamentous damage [19–21,54]. This is often a manifestation of a severe flexion injury in the thoracolumbar region (Figs. 8.54 and 8.55). Finally, one must be aware of the relationships of the craniovertebral junction. Injuries to this region were once believed to be uniformly fatal. A typical history in such a situation is of a victim with no other apparent injuries who died at the accident scene or who suffered respiratory arrest at the scene. In fact, survival is possible with a number of less-severe injuries with occipito-atlantal subluxations. As mentioned, of the three

methods of determining the relationships of the craniovertebral junction, the dens–basion interval method of Harris [11,12] is the most reliable and easiest to perform (Fig. 8.68). Box 8.4 summarizes cartilage or joint space abnormalities. Box 8.4. Cartilage or joint space abnormalities Wide predental space Wide or narrow intervertebral disc space Wide apophyseal joints ”Naked” facets Wide interspinous or interlaminar distance Abnormal dens–basion line

151

8 Radiologic “footprints”: the ABCS

Fig. 8.65 Anterior dislocation of C6–C7 with bilateral facet locking (arrow).

A

A

D

152

Fig. 8.66 Lateral flexion injury with “floating” pillar of C5. (A) Lateral radiograph shows duplication of the pillar image at C5 (arrows). In addition, there is a fracture of the spinous process of C6. (B) Axial CT image shows fractures of the pedicle and lamina on the right (arrows). Note the asymmetric pillar image (P) as a result of the rotation.

B

B

C

Fig. 8.67 Vertebral artery injuries in two patients with pillar fractures. (A) Axial CT image from a CT angiogram shows pedicle and lamina fractures (arrowheads) and a “floating” pillar on the left. The lucency in the left vertebral artery (arrow) is a clot. (B) Coronal reconstructed CT image shows the clot (arrow) on the left. (C) Axial CT image in another patient shows a pillar fracture on the left (arrow). (D) Axial T2-weighted MR image shows high signal within the left vertebral artery on the left (arrow) representing a dissection.

8 Radiologic “footprints”: the ABCS

A

B

Soft tissue abnormalities Injury to the vertebral column produces various soft tissue abnormalities, which often are the first indication of an occult fracture or subluxation on radiographs [5,6,18,27,38]. In an era when CT is used as the primary screening tool, soft tissue findings have a less important role than previously. Nonetheless, they still interest us. Significant soft tissue abnormalities include widening of the retropharyngeal space, widening of the retrotracheal space, displacement of the prevertebral fat stripe, soft tissue mass near the craniovertebral junction, deviation of the trachea or larynx, widening of the paraspinal soft tissues, and loss of the psoas stripe. The first five of these signs occur exclusively in the cervical region. Paraspinal soft tissue changes occur in the thoracic area, and loss of a psoas stripe is observed within the lumbar region [5,6,18]. The retropharyngeal space is really a potential space, which becomes visible when it is filled with fluid. It is measured from the anteroinferior aspect of the body of C2 to the posterior aspect of the pharyngeal air column (Fig. 8.69). This space should never exceed 7 mm in adults and children. In establishing this limit, magnification and technical factors are taken into account. Measurements are based on a horizontal beam lateral radiograph with a 40 inch (1 m) focal film distance. This distance can be directly measured from a sagittal reconstructed CT image. Disruptive injuries in this region produce widening of this space even in the absence of more obvious skeletal abnormalities (Fig. 8.70) [5,6,18]. Harris has observed that the contour of the soft tissue changes are as important as the actual measurements [6]. It should be kept in mind that the retropharyngeal fascia continues uninterrupted from the posterior pharynx to the mediastinum. The fascial planes that make up the retropharyngeal space serve as a potential transit route for fluid in either direction from the neck to the thorax.

Fig. 8.68 Occipito-atlantal dislocation in two patients. (A) Lateral radiograph shows the dens–basion interval to be 18 mm (normal is 6–12 mm). (B) Sagittal reconstructed CT image shows similar findings in another patient. Note the increased distance between the basion and the tip of the dens, which should also not exceed 12 mm.

The retrotracheal space is measured from the anteroinferior aspect of the body of C6 to the posterior tracheal wall (or shadow of the endotracheal tube) (Fig. 8.71). This space should never exceed 14 mm in children and 22 mm in adults. Again, measurements are based on standard 40  inch (1  m) lateral radiographs made with horizontal beam technique. As with widening of the retropharyngeal space, retrotracheal space widening accompanies severe disruptive injuries (Fig. 8.72) and is a clue on radiographs to an adjacent injury [5,6,18,31,55]. This sign has largely been rendered obsolete by CT. Injury to the craniovertebral junction frequently produces a prevertebral hematoma that appears as a soft tissue mass anterior to C1 and C2 (Figs. 8.73 and 8.74). This sign is a reliable indicator of injury when the history supports that diagnosis and other radiographic findings are lacking [5,6,38]. Again, Harris emphasized that the contour of the soft tissues is more important than the actual measurements [38]. The prevertebral fat stripe courses caudally along the anterior surfaces of the vertebral bodies from C2 through C6 (Fig. 8.75) [5,6,18,56,57]. This thin band of fatty tissue parallels the anterior longitudinal ligament to the level of C6, where it gradually deviates anteriorly and inferiorly toward the base of the neck. Displacement of this line from its normal location is a reliable indicator that underlying injury has produced a hematoma (Fig. 8.76) [31]. Again, cervical CT has rendered this sign a historic curiosity. Tracheal or laryngeal deviation is a nonspecific sign that may occur under various circumstances besides trauma. The most common causes of tracheal deviation are rotation and thyroid enlargement. The airway structures are not as securely fixed in the neck and are subject to motion by both posttraumatic hematoma and infection. A diagnosis of injury should not be based solely on this sign (Figs. 8.77 and 8.78).

153

8 Radiologic “footprints”: the ABCS

A

B

Fig. 8.69 Normal retropharyngeal space. (A) Sagittal drawing showing the fascial spaces of the neck. Note that the retropharyngeal space continues uninterrupted from the posterior pharynx to the mediastinum, allowing fluid from either region to accumulate in both directions. (B) Axial drawing. (C) Lateral radiograph shows the normal space, where it is measured at C2 (arrow).

C

A

B

C

Fig. 8.70 Wide retropharyngeal space as a result of fractures of C1 and C2. (A) Lateral radiograph shows widening of the retropharyngeal space with a convex contour superiorly (small arrows). There is a fracture of the posterior arch of the atlas (large arrow). (B) Axial CT image shows fractures of the anterior and posterior arches of the atlas (arrows). (C) Sagittal reconstructed CT image shows a dens fracture (arrow) in addition to the prevertebral soft tissue swelling (*). Again, note the convex shape to the retropharyngeal hematoma (arrowheads).

154

8 Radiologic “footprints”: the ABCS

A

Fig. 8.72 Burst fracture of C6. (A) Lateral radiograph shows retrotracheal soft tissue fullness (*) and irregularity of the superior body of C6. (B) Sagittal reconstructed CT image shows the burst fracture of C6 with retropulsion of a fragment into the vertebral canal.

Fig. 8.71 Normal retrotracheal space (*), as measured at C6. Abnormalities of this space are less important for finding occult injuries when CT is performed.

A

B

B

C

Fig. 8.73 Jefferson fracture of C1. (A) Lateral radiograph shows massive prevertebral soft tissue swelling (*). There is widening of the predental space (arrowhead) and anterolisthesis of the atlas on the axis (arrow). (B) Open-mouth view shows offset of the lateral masses of C1 on C2 (arrows). Note the widening of the spaces between the dens and the lateral masses of C1. (C) Axial CT image shows comminuted fractures of both sides of the anterior arch of the atlas.

Paraspinal soft tissue changes are an indication that an occult injury may have occurred in the thoracic region, particularly in obese individuals with high thoracic fractures (Figs. 8.79 and 8.80). Frequently, these changes will be encountered on a portable chest radiograph. This sign is by no means specific for injury and also occurs in the presence of infections, neoplasms, or extramedullary hematopoiesis. Loss of the psoas stripe is another nonspecific finding that occasionally is useful in diagnosing an occult lumbar injury. Generally, there is some other indication that injury has occurred (Fig. 8.81). Loss of the psoas stripe may also occur with infections or spasm of any etiology. Although the soft tissue abnormalities described above are helpful in directing attention to the possibility of an underlying

vertebral injury, they may result from other etiologies besides traumatic injury. The most common pitfall occurs in patients who have undergone endotracheal or nasogastric intubation (Fig. 8.82). Irritation of the pharyngeal tissues during intubation can produce dramatic and rapid swelling of these tissues because of the rich vascular venous plexus of these tissues. Consequently, retropharyngeal soft tissue swelling, if it is the only abnormality present, should not be interpreted as an indication of underlying cervical injury. Retropharyngeal abscess (Fig. 8.83) is another cause of wide prevertebral soft tissues. However, it usually does not cause any significant diagnostic difficulty because of the clinical setting. Massive retropharyngeal or retrotracheal soft tissue swelling as the result of vascular or facial injury may suggest

155

8 Radiologic “footprints”: the ABCS

A

Fig. 8.74 Dens fracture. (A) Lateral radiograph shows retropharyngeal soft tissue swelling (*) as well as anterior displacement of the posterior arch of the atlas (arrow). (B) Sagittal reconstructed CT image shows the dens fracture (arrows) as well as anterior displacement of the dens, which accounted for the shift in position of the posterior arch of C1 shown in A.

B

A

B

Fig. 8.75 Normal prevertebral fat stripe (arrows). Fig. 8.76 Displacement of the prevertebral fat stripes (arrows) in two patients by prevertebral hematomas. (A) A compression fracture of C5. (B) A dislocation of C5 on C6.

A

156

B

Fig. 8.77 Tracheal and esophageal deviation in a patient with an extension sprain at C6–C7. (A) Frontal radiograph shows deviation of the trachea (arrows) and nasogastric tube to the right and widening of the C6 disc space (*). (B) Lateral radiograph shows prevertebral soft tissue swelling (*), an avulsion fracture off the base of C6 (arrow), and widening of the C6 disc space (**).

8 Radiologic “footprints”: the ABCS

A

B

A

B

Fig. 8.78 Tracheal deviation in a patient with a burst fracture of C5. (A) Frontal radiograph shows the trachea deviated to the right (arrows) by hematoma. (B) Lateral radiograph shows posterior bowing of the posterior vertebral body line of C5 (arrow). There is a teardrop fragment of the inferior body of C5.

Fig. 8.79 Widening of the paraspinal line in a patient with a fracture of T9. (A) Frontal chest radiograph shows widening of the paraspinal stripes (arrows) in the lower thoracic region. (B) Detailed view of the lower thoracic area shows compression of T9 and left lateral listhesis of T8. There is a large paraspinal hematoma (arrows). (C) Lateral radiograph shows compression of T9. (D) Axial CT image shows disruption of the facet joints (arrows).

C

D

157

8 Radiologic “footprints”: the ABCS

A

Fig. 8.80 Paraspinal widening in a patient with a T5–T6 fracture–dislocation. (A) Frontal chest radiograph shows a dextroscoliosis of the mid thoracic region and paraspinal widening on the right (arrow). (B) Axial CT image shows fractures of the body of T6 and paraspinal hematoma (arrows).

B

A

B

Fig. 8.81 Loss of right psoas stripe in a patient with a Chance-type injury of L2 (*). The left psoas is normal (arrows).

Fig. 8.82 Effect of intubation on retropharyngeal soft tissues. (A) Lateral radiograph before intubation shows the soft tissues to be normal. (B) Following intubation there is significant soft tissue swelling (*) in the same region.

possible cervical injury because of the history of massive trauma. The soft tissue planes between the thorax and neck freely connect, and blood from an aortic injury may dissect cephalad (Fig. 8.84). Similarly, blood from a severe facial injury may dissect distally (Fig. 8.85). As a rule, cervical injury seldom produces the massive soft tissue swelling found in patients with vascular injury (or infection for that matter). What is interesting, however, is the converse phenomenon, in which widening of the superior mediastinal structures as the result of vertebral injury mimics the findings that occur in aortic injury [58]. This is one reason why thoracic vertebral CT is crucial for all victims of severe trauma. Crying infants or small children may also have prevertebral soft tissue swelling. This is usually not a significant problem

158

from a diagnostic standpoint. The cervical radiographs of children generally show either gross abnormalities (fracture, dislocation, physeal injury) or are normal. The subtle radiographic findings used to diagnose injuries in adults are not present in children. When confronted with wide soft tissues in a pediatric cervical radiograph, it is best to repeat the lateral view when the child is quiet (Fig. 8.86). Patients may be encountered who have vascular migration of the carotid arteries into the retropharyngeal soft tissues (Fig. 8.87) [59,60]. This finding can be disturbing, particularly in an elderly patient who has suffered trauma. In most elderly individuals with this abnormality, vascular calcification is present within the prevertebral soft tissues on the lateral view. Normally, this calcification overlies the vertebrae and is not visible on this

8 Radiologic “footprints”: the ABCS

A

B

A

B

Fig. 8.83 Retropharyngeal abscess. (A) Lateral radiograph shows massive prevertebral soft tissue swelling (*). (B) Axial CT image with contrast enhancement shows the prevertebral abscess (*).

C

Fig. 8.84 Retropharyngeal and retrotracheal hematoma as a result of aortic injury. (A) Lateral radiograph shows massive prevertebral soft tissue swelling (*); the retropharyngeal space measures 17 mm; the retrotracheal space measures 30 mm. (B) Frontal chest radiograph shows widening and irregularity of the superior mediastinum and the trachea is deviated to the right. (C) Aortogram shows evidence of disruption of the descending aorta at the arch, with multiple areas of irregularity representing hematoma and pseudoaneurysm formation (arrows). As a rule, cervical injury seldom produces the massive soft tissue swelling found in patients with vascular injury.

view. On frontal views, calcified vessels normally are found on either side of the vertebrae (Fig. 8.88). In transposition, they are not. This problem can be solved by performing CT. Finally, the simple act of swallowing while a radiograph or CT scan is being obtained can produce widening of the retropharyngeal and retrotracheal soft tissues (Fig. 8.89). In these instances a repeat examination or CT will show that there is no evidence of injury. Box 8.5 summarizes soft tissue abnormalities.

Box 8.5. Soft tissue abnormalities Wide retropharyngeal space Wide retrotracheal space Displacement of prevertebral fat stripe Soft tissue mass in craniocervical junction Deviation of trachea or larynx Paraspinal soft tissue mass

159

8 Radiologic “footprints”: the ABCS

Fig. 8.86 Retropharyngeal soft tissue swelling (*) in a crying child.

Fig. 8.85 Retropharyngeal soft tissue swelling in a patient who suffered a gunshot wound to the face.

A

B

D

160

C

Fig. 8.87 Vascular transposition of the carotid vessels. (A) Lateral radiograph shows retropharyngeal soft tissue swelling with overlying calcified vessels (arrows). (B) Frontal radiograph shows calcified vessels overlying the vertebrae on the right (arrows). (C) Contrast-enhanced CT image shows that the internal carotid arteries (*) have migrated to the prevertebral area. (D) Axial CT myelogram on another elderly patient shows vascular migration (*) into the prevertebral region.

8 Radiologic “footprints”: the ABCS

A

A

Fig. 8.88 Normal appearance of carotid vessels. (A) Lateral radiograph shows calcified carotid vessels overlying the vertebrae (arrow). (B) Frontal radiograph shows the calcified vessels (arrows) lying lateral to the vertebrae.

B

Fig. 8.89 Effect of swallowing. (A) Lateral radiograph shows widening of the prevertebral soft tissues (*). (B) A second radiograph made minutes later shows normal soft tissues.

B

Significant signs and their significance This chapter has described numerous radiographic signs useful for the diagnosis of vertebral injury. Many are based on the work of Clark and coworkers [18], who originally described 12 signs of cervical vertebral trauma. They called these signs “significant” since their presence indicated the possibility of an underlying and often occult cervical injury. The signs were divided into three categories: • abnormal soft tissues ∙ wide retropharyngeal space ∙ wide retrotracheal space ∙ displaced prevertebral fat stripe ∙ tracheal or laryngeal displacement • abnormal vertebral alignment ∙ loss of lordosis ∙ acute kyphotic angulation

•

∙ torticollis ∙ wide interspinous space ∙ rotation of vertebral bodies abnormal joints ∙ wide predental space ∙ abnormal disc ∙ wide facet joints.

The Clark study merely reported the findings and stated their usefulness in diagnosing vertebral injuries [18]. Indeed, these signs have been amply illustrated in this book and others [5,18]. The original report, however, did not mention the frequency of occurrence of each of the 12 signs, their reliability, or common combinations of signs. Furthermore, the report prompted us to question how common the same signs were in a similar patient population without cervical injury. To answer these questions, we reviewed the images of 100 patients

161

8 Radiologic “footprints”: the ABCS

with proven cervical injuries and compared their radiographic findings with those of an age-matched group of control patients who had cervical radiographs for suspected trauma. The control patients had no symptoms referable to the cervical region but were brought to the hospital in cervical restraining collars that had been applied at the accident scene. All were subsequently shown to have no radiographic evidence of injury. The images were obtained under the same conditions for each group. In addition to the 12 significant signs, the patients were studied for the presence of anterolisthesis or retrolisthesis [31]. Our review showed that the following signs never occurred under normal circumstances: wide interspinous or interlaminar space, wide facet joints, wide retropharyngeal space, displaced prevertebral fat stripe, rotation, and wide predental space. Loss of lordosis occurred slightly more often in the group with proven cervical injuries than did anterolisthesis or retrolisthesis. In the control group, listhesis was always associated with cervical spondylosis. Kyphotic angulation occurred about equally in the two groups. In the group with injuries, loss of lordosis, kyphotic angulation, and anterolisthesis or retrolisthesis never occurred alone. Disc space widening was four times as common in the injury group and was associated with extension injury. Tracheal deviation occurred twice as often in the injury group. In the control group, tracheal deviation was the result of patient rotation. In summary, the most reliable radiographic signs of underlying cervical injury were: • • • •

wide interspinous (interlaminar) space wide facet joints wide retropharyngeal space prevertebral fat stripe displacement.

The following findings never occurred as the sole indication of injury: • • •

loss of lordosis kyphotic angulation tracheal deviation.

These findings occurred in higher incidence in the trauma group and may be found in normal patients, and their presence alone should not be interpreted as evidence of injury. When combined with other findings, however, loss of lordosis, kyphotic angulation, and tracheal deviation are reliable indicators of underlying injury. The most reliable combination of significant signs is: • • • •

wide interspinous (interlaminar) space wide facet joint loss of lordosis kyphotic angulation.

Narrow disc spaces and anterolisthesis or retrolisthesis are more likely the result of degenerative disc disease when encountered as isolated findings. A wide disc space should always suggest an extension injury, particularly if found in conjunction with degenerative changes in an older patient. Supervised flexion

162

Table 8.1 Incidence of significant signs

Sign

Injury group (%) Control group (%)

Loss of psoas stripe

63

52

Wide interspinous space

43

0

Wide facet joint

39

0

Anterolisthesis or retrolisthesis

36

24

Wide retropharyneal space

31

3

Kyphotic angulation

21

17

Displaced prevertebral fat stripe

18

0

Tracheal deviation

13

7

Narrow disc space

24

32

Wide disc space

8

2

Rotation

5

50

Wide predental space

3

0

Wide retrotracheal space

2

1

Torticollis Source: from Daffner and Verma [31].

1

1

and extension views should be performed if there are no other signs of injury. In the absence of other signs of injury, flexion and extension films should also be obtained for all patients with minor degrees of anterolisthesis or retrolisthesis to determine stability. Finally, the presence of any significant sign on a radiograph should prompt close scrutiny for additional subtle evidence of underlying injury [31]. Table 8.1 summarizes the significant signs.

The rules of 2S Throughout this book, a number of measurements are given in which 2 mm is the maximum distance allowed for that particular diagnostic parameter. These can be summarized as the “rules of 2s.” The distance of 2 mm has been determined from observation of thousands of radiographs with technical parameters as well as clinical findings taken into consideration. Although exceptions exist, the rules of 2s remain valid in most cases. • The maximum allowable difference between the interlaminar or interspinous spaces of three contiguous levels is 2 mm. • The maximum allowable difference in the interpedicle distance between two contiguous levels is 2 mm. This applies not only to the transverse interpedicle distance but also to the vertical interpedicle distance. • In the atlanto-axial region, 2 mm is the maximum allowable unilateral or bilateral lateral atlanto-axial offset between the lateral masses of C1 and the body of C2 on the frontal view. This is most likely to occur in the presence of some form of arch anomaly of C1 [8]. Jefferson fractures produce offset of 3 mm or more.

8 Radiologic “footprints”: the ABCS

•

•

•

Lateral flexion and extension views, particularly in the cervical region, may produce up to 2 mm of anterolisthesis or retrolisthesis. This is simply the result of ligamentous laxity of the anterior or posterior longitudinal ligaments and is believed to be of no clinical significance as long as the spinolaminar line remains intact. The normal facet joints should never exceed 2 mm in width, particularly in flexion. Any increase is usually the result of posterior ligamentous disruption. Finally, there are normal differences in the heights of the anterior and posterior portions of the thoracic and lumbar vertebral bodies. The anterior margins of thoracic vertebral bodies are 2 mm shorter than the posterior margins. This accounts for the thoracic kyphosis. Conversely, the posterior portions of the lumbar vertebral bodies are up to

References 1.

Daffner RH, Hackney DB. ACR Appropriateness Criteria on suspected spine trauma. J Am Coll Radiol 2007; 4:762–775. 2. 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. 3. Cattell HS, Filtzer DL. Pseudosubluxation and other normal variations of the cervical spine in children: a study of one hundred and sixty children. J Bone Joint Surg 1965;47A:1295–1309. 4. Swischuk LE. Emergency Imaging of the Acutely Ill or Injured Child, 4th edn. Philadelphia, PA: Lippincott Williams & Wilkins, 2000, pp. 532–598. 5. Gehweiler JA Jr., Osborne RL Jr., Becker RF. The Radiology of Vertebral Trauma. Philadelphia, PA: WB Saunders, 1980. 6. Harris JH Jr., Mirvis SE. The Radiology of Acute Spinal Trauma, 3rd edn. Baltimore, MD: Williams & Wilkins, 1996. 7. Dolan KD. Cervicobasilar relationships. Radiol Clin North Am 1977;15:155–166. 8. Gehweiler JA Jr., Daffner RH, Roberts L Jr. Malformations of the atlas vertebra simulating the Jefferson fracture. AJR Am J Roentgenol 1983;140:1083–1086. 9. Greenberg AD. Atlantoaxial dislocations. Brain 1968;91:665–684. 10. Harris JH Jr., Burke JT, Ray RD, et al. Low (type III) odontoid fracture: a new radiographic sign. Radiology 1984;153:353–356.

2 mm shorter than the anterior portions; this accounts for the lumbar lordosis. Box 8.6 summarizes the rules of 2s. Box 8.6. Rules of 2s 2 mm is the normal upper limit of difference for: interspinous space or interlaminar space interpedicle distance (transverse and vertical) unilateral or bilateral lateral atlanto-axial offset anterolisthesis or retrolisthesis with flexion or extension facet joint width differences in height of anterior and posterior thoracic and lumbar vertebral bodies.

11. Harris JH Jr., Carson GC, Wagner LK. Radiologic diagnosis of traumatic occipitovertebral dissociation. 1. Normal occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 1994;162:881–886. 12. 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. 13. Harris JH Jr. The cervicocranium: its radiographic assessment. Radiology 2001;218:337–351. 14. Powers B, Miller MD, Kramer RS, et al. Traumatic anterior atlantooccipital dislocation. Neurosurgery 1979; 4:12–17. 15. Lee C, Woodring JH, Goldstein SJ, et al. Evaluation of traumatic atlantooccipital dislocations. AJNR Am J Neuroradiol 1987;8:19–26. 16. Lovelock JE, Schuster JA. The normal posterior atlantoaxial relationship. Skeletal Radiol 1991;20:121–123. 17. Shapiro R, Youngberg AS, Rothman SL. The differential diagnosis of traumatic lesions of the occipito-atlanto-axial segment. Radiol Clin North Am 1973; 8:33–38. 18. Clark WM, Gehweiler JA Jr., Laib R. Twelve significant signs of cervical spine trauma. Skeletal Radiol 1979;3:201–205. 19. Daffner RH, Deeb ZL, Rothfus WE. “Fingerprints” of vertebral trauma:

20.

21.

22.

23.

24.

25.

26.

27.

a unifying concept based on mechanisms. Skeletal Radiol 1986; 15:518–525. Daffner RH, Deeb ZL, Goldberg AL, et al. The radiologic assessment of posttraumatic vertebral instability. Skeletal Radiol 1990;19:103–108. Gehweiler JA Jr., Daffner RH, Osborne RL, Jr. Relevant signs of stable and unstable thoracolumbar vertebral column trauma. Skeletal Radiol 1981; 7:179–183. Naidich JB, Naidich TP, Garfein C, et al. The widened interspinous distance: a useful sign of anterior cervical dislocation in the supine frontal projection. Radiology 1977;123:113–116. Suss RA, Zimmerman RD, Leeds NE. Pseudospread of the atlas: false sign of Jefferson fracture in young children. AJNR Am J Neuroradiol 1983;4: 183–186. Jefferson G. Fracture of the atlas vertebra: report of four cases, and a review of those previously recorded. Br J Surg 1920;7:407–422. Lee C, Woodring JH. Unstable Jefferson variant atlas fractures: an unrecognized cervical injury. AJNR Am J Neuroradiol 1992;12:1105–1110. Malham GM, Ackland HM, Jones R, et al. Occipital condyle fractures: incidence and clinical follow-up at a level 1 trauma centre. Emerg Radiol 2009;16:291–297. Daffner RH, Daffner SD. Vertebral injuries: detection and implications. Eur J Radiol 2002;42:100–116.

163

8 Radiologic “footprints”: the ABCS

28. Braakman R, Vinken PJ. Uilateral facet interlocking in the lower cervical spine. J Bone Joint Surg 1967; 49B:249–257. 29. Daffner RH. Dislocation at L-5–S-1 with unilateral facet lock. Skeletal Radiol 1989;18:489–490. 30. Connolly PJ, Esses SI, Heggeness MH, et al. Unilateral facet dislocation of the lumbosacral junction. Spine 1992; 17:1244–1248. 31. Daffner RH, Verma SV. The significant signs of cervical vertebral trauma: a reassessment. Appl Radiol 1995; 24:31–35. 32. Young JWR, Resnik CS, DeCandido P, et al. The laminar space in the diagnosis of rotational flexion injuries of the cervical spine. AJR Am J Roentgenol 1989;152:103–107. 33. Daffner SD, Daffner RH. Computed tomography diagnosis of facet dislocations: the “hamburger bun” and “reverse hamburger bun” signs. J Emerg Med 2002;23:387–394. 34. Lee C, Woodring JH, Rogers LF, et al. The radiographic distinction of degenerative slippage (spondylolisthesis and retrolisthesis) from traumatic slippage of the cervical spine. Skeletal Radiol 1986;15:439–443. 35. Schaaf RE, Gehweiler JA Jr., Powers B, et al. Lateral hyperflexion injuries of the spine. Skeletal Radiol 1978; 3:73–78. 36. Lee C, Woodring JH. Sagittally oriented fractures of the lateral masses of the cervical vertebrae. J Trauma 1991; 31:1638–1643. 37. Bohrer SP, Chen YM. Cervical spine annulus vacuum. Skeletal Radiol 1988; 17:324–329. 38. Harris JH Jr., Yeakley JS. Radiographically subtle soft tissue injuries of the cervical spine. Curr Probl Diagn Radiol 1989;18:167–190.

164

39. Daffner RH. Thoracic and lumbar vertebral trauma. Orthop Clin North Am 1990;21:463–482. 40. Kricun ME, Kricun R. Fractures of the lumbar spine. Semin Roentgenol 1992; 27:262–270. 41. Jackson H, Kam J, Harris JH Jr., et al. The sacral arcuate lines in upper sacral fractures. Radiology 1982;145:35–39. 42. 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. 43. Martijn A, Veldhuis EFM. The diagnostic value of interpediculate distance assessment on plain films in thoracic and lumbar spine injuries. J Trauma 1991;31:1393–1395. 44. McGrory BJ, VanderWilde RS, Currier BL, et al. Diagnosis of subtle thoracolumbar burst fractures: a new radiographic sign. Spine 1993;18: 2282–2285. 45. Benzel EC, Hart BL, Ball PA, et al. Fractures of the C-2 vertebral body. J Neurosurg 1994;81:206–212. 46. Burke JT, Harris JH Jr. Acute injuries of the axis vertebra. Skeletal Radiol 1989; 18:335–346. 47. Anderson LD, D’Alonzo RT. Fractures of the odontoid process of the axis. J Bone Joint Surg 1974;56A:1663–1674. 48. Thomeier WC, Brown DC, Mirvis SE. The laterally tilted dens: a sign of subtle odontoid fracture on plain radiography. AJNR Am J Neuroradiol 1990;11: 605–608. 49. Smoker WRK, Dolan KD. The “fat” C2: a sign of fracture. AJNR Am J Neuroradiol 1987;8:33–38. 50. Bohay D, Gosselin RA, Contreras DM. The vertical axis fracture: a report on three cases. J Orthop Trauma 1992; 6:416–419.

51. Montana MA, Richardson ML, Kilcoyne RF, et al. CT of sacral injury. Radiology 1986;161:499–503. 52. Cintron E, Gilula LA, Murphy WA, et al. The widened disk space: a sign of cervical hyperextension injury. Radiology 1981;141:639–644. 53. Jonsson K, Niklasson J, Josefsson PO. Avulsion of the cervical spinal ring apophyses: acute and chronic appearance. Skeletal Radiol 1991; 20:207–210. 54. Templeton PA, Young JWR, Mirvis SE, et al. The value of retropharyngeal soft tissue measurements in trauma of the adult cervical spine. Skeletal Radiol 1987;16:98–104. 55. O’Callaghan JP, Ullrich CG, Yuan HA, et al. CT of facet distraction in flexion injuries of the thoracolumbar spine: the “naked” facet. AJNR Am J Neuroradiol 1980;1:97–102. 56. Whalen JP, Woodruff CL. The cervical prevertebral fat stripe: a new aid in evaluating the cervical prevertebral soft tissue space. AJR Am J Roentgenol 1970; 109:445–451. 57. Penning L. Prevertebral hematoma in cervical spine injury, incidence and etiologic significance. AJR Am J Roentgenol 1981;136:553–561. 58. Dennis LN, Rogers LF. Superior mediastinal widening from spine fractures mimicking aortic rupture on chest radiographs. AJR Am J Roentgenol 1989;152:27–30. 59. Daffner RH, Kennedy SL, Fix TJ. The retropharyngeal prevertebral soft tissues revisited. Emerg Radiol 1996;3:247–252. 60. Fix TJ, Daffner RH, Deeb ZL. Carotid transposition: another cause of wide retropharyngeal soft tissues. AJR Am J Roentgenol 1996;167:1305–1307.

Chapter

9

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

Introduction The prevalence of vertebral injuries in children has been reported at less than 1–2% of all pediatric fractures [1]. Spinal column injuries in children are rare compared with similar injuries in adults. Children account for less than 10% of all spinal column injuries; however, the mortality rate of spinal column injuries in children is significantly higher than in adults [2–4]. Children differ from adults both in the type and the outcome of spinal injuries. Clearing the spine in children can be especially difficult because of the relative rarity of these injuries, the inability of young children to communicate, and the presence of various normal variants such as synchondroses and pseudosubluxation. In addition, the vulnerability of children to the deleterious effects of radiation exposure requires a rational approach to the imaging of spine trauma in this population. The cervical spine is the most common site of spinal injuries in children, especially those younger than eight years of age [5]. This chapter reviews the epidemiology of pediatric cervical spine injury, the normal development of the cervical spine, and its normal variants in children. We propose an algorithm for imaging of cervical spine trauma in children and discuss specific cervical spine injuries that are more common in the pediatric population. This is followed by a brief discussion of thoracic and lumbar trauma in the pediatric age group. We hope that understanding of the unique anatomic and biomechanical features of the pediatric spine and knowledge of common injuries of children will aid radiologists in the evaluation of pediatric vertebral trauma.

Etiology The most common cause for spinal cord injury in the entire pediatric age group is the motor vehicle crash [6]. Up to 68% of children sustaining spinal cord injury in motor vehicle accidents have been reported to not be wearing seat belts or be otherwise properly restrained [6]. When evaluated by age group, pedestrian accidents and falls are the most common etiology in children under eight years of age, while older children are more often injured as passengers in motor vehicle crashes or in sports injuries, such as football, diving, and wrestling [4,5,7,8].

Pediatric cervical injuries Unique features The cervical spine is the most common site for spinal trauma in children accounting for up to 79% of all vertebral injuries in children [5,9–11]. Pediatric patients with cervical spine injuries can be divided into two distinct groups: those aged eight years and under and those nine years or older [7]. While cervical injuries in the older group of children are similar to those seen in adults, the younger group differs in the type, level, and frequency of injuries and their outcome. Developmental anatomy and biomechanical factors in the pediatric spine help to explain these differences between young children and adults. A child’s cervical spine does not fully take on the characteristics of the adult spine until eight years of age, at which time the pattern of injury also changes to the pattern seen in adults [12]. The prevalence of cervical spine injury in children eight years or younger is much less than in older children – 15% of 122 pediatric cervical spine injuries in one series [9]. This is likely related to decreased exposure to trauma in young children and the relative plasticity of the pediatric vertebral column. However, if a cervical spine injury does occur in a child under eight years of age, it is much more likely to involve the region between the occiput (C0) and C3; older children tend to have pancervical injuries (Fig. 9.1) [7,12,13]. In a study of 227 children with spinal column injuries, 87% of those under eight years of age had an injury of C3 or higher and had a higher risk of dying from the injury. In contrast, older children, like adults, predominantly had injuries caudal to C4 and none died [14]. The anatomic and biomechanical features of the immature cervical spine make the upper segments of the cervical spine and the craniocervical junction particularly susceptible to injury. The fulcrum of cervical movement is located higher in young children than in adolescents and adults (C5–C6) [15,16]. The fulcrum for flexion is at C2–C3 in infants, C3–C4 by age five years, C4–C5 at 10 years, and C5–C6 by 15 years of age [17]. A disproportionately large head size on a small body subjects the upper cervical spine to high torque under acceleration/deceleration stress. In addition, the spine of children under eight years of age has more physiologic mobility, resulting from relatively weak neck muscles, ligamentous

165

9 Vertebral injuries in children

Fig. 9.1 Relationship between patient’s age, level of bone injury (solid lines) and neurologic deficit (broken lines). (With permission from Hill et al. [9] .)

A

B

laxity, horizontal orientation of the cervical facet joints, wedgeshaped vertebral bodies, and poorly developed uncinate processes (Fig.  9.2) [12,18]. This laxity accounts for a higher incidence of subluxations without fractures and spinal cord injury without radiographic abnormality (SCIWORA) in this population [19]. Autopsy studies have shown that while the spinal column of infants can be stretched up to 2 inches (5 cm), the cord undergoes shearing if stretched beyond 0.25 inches (0.6  cm) [20]. This elasticity of the spinal column compared with the cord helps to explain the higher frequency of cord injury in the absence of osseous injuries. Like other osseous injuries in the appendicular skeleton of children, osseous cervical injuries in children tend to be avulsions or epiphyseal separations (e.g., fractures through the basilar synchondrosis of C2), rather than true fractures [21,22]. Although the prevalence of cervical spine injuries in children under eight years of age is low, they are more likely to have neurologic deficits than their older counterparts. In a series of 179 children with spinal trauma, Osenbach and Menezes noted the incidence of neurologic deficit to be 62% in children under eight compared with 47% in older children [5]. However, because of their tremendous healing potential, children with incomplete neurologic lesions have a better outcome than

166

Fig. 9.2 Normal development of the cervical spine. Cervical spine lateral view in a four-day-old (A) and a nine-year-old (B) child. Vertebral bodies are wedge shaped in the infant (straight arrow in A) and mature to a rectangular shape in the older child. Note the more horizontal orientation of the facets in the infant compared with the older child (wavy arrows in A and B).

adults, with up to 90% showing significant and 60% showing full recovery [4,9]. While the mortality rate of children with head injuries is significantly less than for adults, this is not true for cervical spine injury [23]. In a study of 216 children with spinal injuries, Hamilton et al. [4] reported a mortality rate of 28% in children compared with 11% in adults.

Imaging studies and protocols Although cervical spine injuries in the pediatric population are rare, most physicians have a low threshold for requesting CT of the cervical spine to avoid the potentially serious and costly consequences of missing a cervical spine injury. Studies have documented an increasing utilization of CT in the emergency setting, in spite of an increasing awareness of the deleterious effects of radiation exposure in children [24,25]. A North Carolina study showed that even though patient volume in the emergency department had not changed much over the previous six years the number of cervical spine CTs performed in the same period had increased by 366% (Fig. 9.3) [25]. This increase has been most prominent amongst adolescents aged 13–17 years, a population that is able to communicate

9 Vertebral injuries in children

effectively to be able to provide a good history and physical examination. The National Emergency X-radiography Utilization Study (NEXUS) validated a decision instrument to classify patients as low probability of injury if they met all of the following five criteria: no midline cervical tenderness, no focal neurologic

Fig. 9.3 Increase in CT utilization for C-spine clearance ( – ) in the trauma setting compared with increase in patient volume (- - -). (Modified with permission from Broder et al. [25].)

deficit, normal alertness, no intoxication, and no painful distracting injury [26]. In the NEXUS study, there were 3065 children under 18 years of age, and the decision instrument performed well in children and could have potentially resulted in 20% fewer imaging studies [27]. Although this instrument was primarily developed for use in the adult population, its use has been endorsed in children to minimize health care costs and radiation exposure. A recently published study has shown that CT of the cervical spine exposes the thyroid gland to 90–200 times more radiation than multiple conventional radiographs, significantly increasing the risk for thyroid cancer in young children [28]. The establishment of protocols to clear the pediatric spine has been shown to decrease the time required for that purpose, reduce the need for neurosurgical consultation, and reduce the number of missed injuries [29–31]. Although there is no universal agreement on how to clear the cervical spine in pediatric trauma patients, we recommend the following approach (Fig. 9.4). Imaging is not required in children under eight years of age who are able to communicate and are negative for all NEXUS criteria. If a patient meets any of the NEXUS criteria and a head CT has been ordered, the scan is extended to include C2 level. If a head CT has not been ordered, evaluation with anterior–posterior and lateral radiographs of the cervical spine is performed. The open-mouth view has been shown to be of limited value in children younger than five years of age because of the difficulty of obtaining it in those with limited understanding and cooperation [32]. If the initial evaluation is abnormal, further evaluation with CT is considered. In the presence of neurologic

1. Mental status abnormal? (GCS <14 or age/developmental factors) 2. Neck pain? 3. Neurological deficit? 4. Intoxication? 5. Distracting injuries? 6. Unable to effectively communicate?

Yes to any above and no head CT ordered

Yes to any above and head CT ordered

AP and lateral plain films of the cervical spine (portable films are okay)

Plain films abnormal: leave collar; obtain CT occiput to T3

CT abnormal: leave collar on; consult spine service

Plain films normal

Bring head CT down to C2

CT normal: re-examine patient and proceed with clearance. Document in chart date and time

CT abnormal: leave collar on; consult spine service

No to all above

Passive range of motion

If non-tender and pain free; clear cervical spine. Document in chart date and time

Reassess patient

Patient now asymptomatic with mental status improvement

Patient remains symptomatic, or history of or presence of neurologic deficit giving concern for SCIWORA

Re-examine patient and proceed with clearance. Document in chart date and time

Consult spine specialist; MRI to rule out ligament injury if patient is “safe to travel”

MRI abnormal: leave collar on

MRI normal: remove collar. Document date and time in chart

Fig. 9.4 Protocol for evaluation of pediatric cervical spine trauma. GCS, Glasgow Coma Scale; SCIWORA, spinal cord injury without radiologic abnormality.

167

9 Vertebral injuries in children

deficits, further evaluation with MRI is performed to evaluate for soft tissue injuries. Children older than eight years are stratified into low-risk and high-risk strata based on the NEXUS criteria. Radiographic evaluation is performed for clearing of the cervical spine in the low-risk group, while the high-risk group is evaluated with CT. If the child remains unresponsive for more than 48 hours, MRI is recommended to exclude cord or ligamentous injury even if the CT is negative [30].

Radiographic pitfalls Several normal anatomic variants and congenital anomalies may be misinterpreted as injuries in children. Some of the common ones are discussed below. Physiologic laxity of the cervical spine in children may give the false impression of subluxation and is referred to as pseudosubluxation. This occurs in young children most commonly at the C2–C3 level though it can also be seen at the C3–C4 level (Fig. 9.5) [33,34]. The posterior spinolaminar line of Swischuk helps to differentiate between true laxity and pseudosubluxation. This line is drawn along the posterior arches of C1 and C3 and the posterior arch of C2 should fall within 2 mm of it in normal children [34]. In young children the lateral masses of C1 may appear laterally offset relative to the lateral masses of C2 on the openmouth view or coronal reformatted CT images, giving the impression of a Jefferson fracture (Fig. 9.6) [35]. This results from a discrepancy in the growth rates of the atlas and the axis, as the former follows a neural growth pattern while the latter follows a somatic growth pattern [36,37]. Secondary ossification centers of the spinous processes and ring apophyses of the vertebral bodies should not be confused with fractures. The

A

synchondrosis located between ossification centers should be differentiated from fracture lines. The typical location, smooth and corticated borders of synchondroses, and absence of the “jigsaw puzzle effect” help to differentiate these from fractures. The retropharyngeal soft tissues frequently show marked variation in young children. The normal thickness of the retropharyngeal space is less than 7  mm and the retrotracheal space is less than 14 mm; however, if the radiograph is taken with the neck in flexion or during expiration, marked buckling of the soft tissues may be seen. A repeat radiograph with the neck appropriately extended and during inspiration will show reversal of this finding (Fig. 9.7). Crying is another cause. Children under 16 years of age often have localized kyphosis of the midcervical spine, which disappears with extension [33]. Wedging of the cervical vertebral bodies, especially at the C3 level, is a normal variant in children up to seven years of age (Fig. 9.2). It represents a normal developmental stage between the oval vertebral bodies of an infant and the rectangular bodies of an adolescent and should not be mistaken for a compression fracture [38]. As many as 20% of children under seven years may have overriding of the anterior arch of the atlas on the odontoid process [33,39]. As the tip of the odontoid process is cartilaginous, up to two thirds of the C1 arch may appear to be above the odontoid process (Fig. 9.7B).

Normal relationships and measurements at the craniovertebral junction There are only a few normal landmarks that need to be identified to perform evaluation of the craniovertebral junction. A lateral view of the cervical spine including the skull base is

B

Fig. 9.5 Pseudosubluxation of C2 on C3. (A) Lateral view obtained with neck flexion shows a step-off of C2 on C3 (arrow). However, the posterior spinolaminar line of Swischuk (dotted line) shows normal alignment of C2 with C1 and C3 posterior arches. (B) Repeat radiograph with neck straightening shows normal alignment.

168

Fig. 9.6 Pseudospread of atlas. Opeen--mo outh view in a three-year-old child demonstrattes latteral offsett of the atlas lateral masses (arrows), com mpaared with the lateral masses of C2.

9 Vertebral injuries in children

A

required to assess the relationships at this junction. Though originally described on radiographs, these relationships can be demonstrated with greater detail on CT and MR images, because of the lack of superimposing structures. In addition, MRI can help to evaluate the soft tissue and vascular consequences of abnormal osseous relationships, such as cord compression and vascular dissection. The normal landmarks and craniometric lines are summarized in Fig. 9.8. Wackenheim clivus baseline (referred to by some as the basilar line) is constructed by drawing a line along the clivus

A

D

Fig. 9.7 Normal buckling of the retropharyngeal soft tissues in a child. (A) Lateral view of the cervical spine in flexion demonstrates prominence of the retropharyngeal soft tissues (*), which has become smaller in extension (B).

B

B

E

and extending it inferiorly into the upper cervical spinal canal (Fig. 9.8B) [40]. This line normally falls tangent to the posterior aspect of the tip of the dens and is very useful for assessment of traumatic injuries at the craniovertebral junction. If the line falls too far posterior to the dens, posterior craniocervical dislocation may be present; if the line intersects the body or base of the dens, anterior craniocervical dislocation may be diagnosed. The atlanto-dental interval (also called the predental interval) is the normal space between the posteroinferior

C

Fig. 9.8 Anatomic landmarks for radiographic evaluation of the craniovertebral junction. (A) Sagittal reconstruction of cervical spine CT in the midline: 1, basion; 2, opisthion; 3, dens; 4, posterior arch of C1; 5, posterior axial line. (B) Wackenheim clivus line: line drawn along the clivus falls at a tangent to the dens. (C) Basion–axis interval (black line) is the distance between the basion and a line drawn along the posterior margin of the axis (PAL); it can vary between 4 and 12 mm. The basion–dens interval is the distance between the basion and the tip of the dens (white line), which should be less than 12 mm. (D) Powers method uses the ratio of the distance between the basion (B) and the posterior arch of C1 (S) to the distance between the opisthion (O) and anterior arch of C1 (A); it should be less than 1. (E) Condylar gap is the distance between the occipital condyle and superior facet of the atlas (black line); this should be less than 5 mm. (With permission from Khanna and El-Khoury [11].)

169

9 Vertebral injuries in children

margin of the anterior arch of C1 and the adjacent anterior surface of the dens. It normally measures up to 5 mm in children up to the age of eight years and 3 mm in older children [41]. Widening of this space is found with atlanto-axial instability, as occurs in children with Down syndrome. The basion–dens interval is the distance between the basion and the tip of the dens and can measure up to 12  mm. This index may not be reliable in young children whose dens is not yet fully ossified [42,43]. The relationship between the basion and the dens can also be evaluated using the basion–axial interval, described by Harris and colleagues in 1994 (Fig. 9.8C) [43]. This is the distance between the basion and a line drawn along the posterior aspect of the dens (the posterior axial line). The basion can lie up to 12 mm anterior to the posterior axial line or up to 4 mm posterior to it. The Powers ratio is the ratio of the distance between the basion and the midpoint of the spinolaminar line of C1 to the distance between the opisthion and the midpoint of the posterior margin of the anterior tubercle of C1 (Fig. 9.8D). The normal measurement of this ratio is 0.77 ± 0.09 and a ratio of more than 1 is indicative of atlanto-occipital dislocation [44]. The chief limitation of the Powers method is its low sensitivity in detecting longitudinal distraction and posterior dislocation [45]. As mentioned in Chapter 8, this method is seldom used today. Kaufman’s method relies on measuring the gap of the atlanto-occipital junction (Fig. 9.8E). This should measure less than 5  mm in all children regardless of age and gender [46]. Widening of the atlanto-occipital junction raises concern for distraction at the craniovertebral junction. Of the above mentioned lines and relationships, the Wackenheim clivus baseline is the most useful in the evaluation of the craniocervical relationship and should always be mentally drawn [47]. The atlanto-dental interval should also be visually assessed to exclude atlanto-axial subluxation or instability. The other three methods are helpful if there is a specific concern for craniocervical distraction injury.

anterior dislocation, posterior dislocation, and longitudinal distraction (Fig. 9.9). Retropharyngeal swelling may be the first indication of this injury but is a nonspecific finding [45]. A variety of craniometric methods have been described for the radiographic evaluation of OAD. Some of these include the Wackenheim clivus baseline (Fig. 9.10), Kaufman’s method (condylar gap method) [46], the Powers ratio [44], the basion–dens interval [42], the basion–posterior axial line method (Harris method) [52], and the Lee (X-line) method [45]. The X-line or occipital–axial lines method was described by Lee and colleagues in 1987 [45]. A line drawn from the basion to the midpoint of the C2 spinolaminar line should fall tangential to the posterosuperior aspect of the dens, while a second line drawn from the opisthion to the posteroinferior corner of the body of C2 should intersect the highest point on the C1 spinolaminar line (Fig. 9.11). In children with suspected OAD, the recommended craniometric technique is the distance between the occipital condyle and the superior facet of C1 (Kaufman method) measured on lateral Fig. 9.9 Occipito-atlantal dislocation. Line drawings of the three types of atlantooccipital dislocation. Type I, anterior dislocation; type II, distraction–separation; type III, posterior distraction. (With permission from Hosalkar et al. [51].)

Cervical injuries in children Occipito-atlantal dislocation Occipito-atlantal dislocation (OAD) is rare, but a common cause of fatal cervical spine injuries. Studies of traffic fatalities have revealed 20 to 35% incidence of OAD [48,49]. Though almost uniformly fatal in the past, a high index of suspicion, improved patient transportation, advances in imaging technology, and improvements in management have led to improved survival in patients with this complex and often fatal injury [50,51]. The cause is most commonly motor vehicle crashes or a pedestrian struck by a motor vehicle. Concomitant intracranial injuries are often present because of the high energy involved in this injury. The small size of the occipital condyles and an essentially horizontal plane of the occipito-atlantal joint makes children twice as likely to sustain OAD than adults [49]. Occipito-atlantal dislocation can be classified into three types:

170

Fig. 9.10 Atlanto-occipital dislocation (type I, anterior dislocation) in a 16-year-old victim of a high-speed motor vehicle collision. Lateral cervical spine view demonstrates disruption of the Wackenheim clivus baseline (dotted line).

9 Vertebral injuries in children

Fig. 9.11 The Lee (X-line) method. The craniovertebral junction: a line drawn from the basion to the midpoint of the C2 spinolaminar line should touch the tip of the dens, and a line drawn from the opisthion to the posteroinferior border of the C2 body should pass through the cranial part of the C1 spinolaminar line. (With permission from Khanna and El-Khoury [11].)

radiographs or reformatted CT images. This distance ranges between 1.5 and 3.5 mm in normal children, and a measurement of more than 5 mm should raise concern for longitudinal distraction [45,46]. In a blinded assessment of these methods applied to 104 patients (adult and pediatric), the condylar gap method has been shown to have the highest sensitivity (1.0) with a negative predictive value of 1.0 and a positive predictive value of 0.45 [53]. The efficacy of all these methods is dependent on the visualization of the appropriate landmarks on lateral views of the cervical spine, which can be challenging in trauma patients. Variable magnification on cross-table lateral views can also affect the accuracy of various measurements [42]. Pang and colleagues [54] have suggested a variation of the Kaufman method whereby measurement of the occipital condyle and C1 facet interval (CCI) on sagittal and coronal CT reconstructions is the most accurate method of detecting OAD. In their study in normal children, the mean interval was 1.28 mm with no measurement exceeding 1.95 mm. They have suggested using a CCI of 4  mm on reformatted CT slices to detect OAD with the highest accuracy [55]. Imaging with MRI will demonstrate associated soft tissue injuries, including ligamentous disruption; neurovascular injuries, such as cervicomedullary or cord laceration/transection; hematomas; and vascular injuries such as vasospasm or dissection of the vertebral or internal carotid arteries.

the other (cock-robin posture). Atlanto-axial rotatory fixation is a pathologic fixation–subluxation of the atlas on the axis in a position that is within the normal range of rotation. Hence, the term ‘fixation’ is preferred to ‘subluxation’ as the joint is fixed within the normal range of rotation. The atlanto-axial joint is the most active joint in the body, with a normal range of motion of approximately 45° to either side in both adults and children [57,58]. This range of motion is allowed by the fact that the facet joints at C1 and C2 are nearly horizontal, which results in instability at the atlanto-axial articulation. The atlanto-axial joint is stabilized by the transverse and the alar ligaments. The transverse ligament prevents excessive anterior shift of C1 on C2 while the alar ligament checks excessive rotation. Atlanto-axial subluxation has been classified into four types by Fielding and Hawkins (Fig. 9.12) [59]. In type I, C1 is rotated on C2 with the dens as the pivot and without anterior displacement (Fig. 9.13). In type II, C1 is rotated on C2 with the lateral mass as the pivot and with less than 5 mm of anterior displacement. In type III, C1 is rotated on C2 with more than 5 mm of anterior displacement of C1 on C2, implying deficiency of the transverse and alar ligaments (Fig. 9.14), and in type IV there is posterior displacement of C1 on C2. Type I is the most common type in children and generally resolves spontaneously. Though the amount of rotation is within the physiologic range, rotation of C1 on C2 can become fixed or limited. The exact etiology of the limited rotation remains unclear, though Fielding and colleagues have suggested that this may be a result of swollen capsular and synovial tissues [59]. Types III and IV are

Atlanto-axial rotatory fixation Atlanto-axial rotatory fixation may be caused by minor and sometimes unnoticed trauma. Other more common causes include inflammation in infection, rheumatologic causes, and surgical interventions such as mastoid surgery [56]. Patients typically present with neck discomfort, limited range of rotation, and with the head rotated in one direction and tilted in

Fig. 9.12 Fielding’s classification of atlanto-axial rotatory subluxation. Type I, rotatory subluxation with the dens as the pivot and no anterior displacement; type II, rotatory subluxation with the lateral mass as the pivot and less than 5 mm of anterior displacement; type III, rotatory subluxation of C1 on C2 with more than 5 mm of anterior displacement of C1; type IV, rotatory subluxation of C1 on C2 with posterior displacement of C1 on C2. (With permission from Fielding and Hawkins [59].)

171

9 Vertebral injuries in children

A

B

C

D

A

B

much less common and have a higher risk of neurologic deficits through cord compression or compromise of the vertebral artery as it courses through the C1 and C2 transverse foramen. The radiographic evaluation of C1–C2 rotatory subluxation can be difficult because of overlapping structures [59]. On the lateral radiograph, the posterior arches of C1 may not superimpose because of the tilt of the atlas on the dens. Evaluation of the atlanto-dental interval may be limited as it is obscured by the anteriorly rotated lateral mass of C1. The open-mouth view demonstrates asymmetry of the lateral atlanto-axial joint spaces. The lateral mass of C1 that is rotated anteriorly appears closer to the midline while the contralateral lateral mass appears further from the midline, with widening of the space

172

Fig. 9.13 Atlanto-axial rotatory fixation with occipito-atlantal counter-rotation in a 17-yearold boy. (A) Axial CT image demonstrates C1–C2 rotatory subluxation (Fielding type I). C1 is rotated to the left, C2 is rotated to the right. (B) Axial image shows offset of the left lateral mass of C1 to the right with respect to the occipital condyle (arrow), consistent with associated C0– C1 counter-rotation. (C) Sagittal reconstructed CT image on the right side shows posterior offset of the C1 lateral mass (L) with respect to C2, but normal alignment with the occipital condyle (C). (D) Sagittal reconstructed image on the left side demonstrates posterior subluxation of the occipital condyle C with respect to the superior facet of the atlas (L). (With permission from Khanna and El-Khoury [11].)

Fig. 9.14 Atlanto-axial rotatory subluxation (type III). (A) Lateral view of the cervical spine demonstrates widening of the atlanto-dental interval (*) with failure of the posterior arches to align. (B) Threedimensional reconstructed CT images show the anterior dislocation of C1 on C2 with the right facet as the pivot.

between the lateral mass and the dens. These radiographic findings are nonspecific and can be seen in normal children secondary to head tilting present in other causes of torticollis. The CT images will demonstrate rotation of C1 on C2; however, the extent of rotation cannot be used to differentiate atlanto-axial rotatory fixation from torticollis of other etiologies or from normal patients scanned with their heads rotated. Dynamic tests have been used to evaluate for a fixed deformity. Dynamic CT performed at rest and repeated with maximum voluntary head turning to the contralateral side will demonstrate failure of C1 and C2 to rotate independent of each other in children with rotatory fixation (Fig. 9.15) [58]. The thyroid gland should not be included in the field of view to keep the

9 Vertebral injuries in children

A

B

D

E

radiation exposure to the gland to a minimum. Our scanning protocol for pediatric cervical spines is beam collimation of 0.75 mm on a 16-slice scanner with axial slices reconstructed at 0.3  mm increments to achieve isotropic imaging. The use of 16-slice scanners and automated tube current modulation helps to reduce radiation exposure even when scanning at submillimeter slice thickness. Viewing the images at thicker slabs (7–20 mm or more) by generating fused images on a workstation helps to illustrate the relationships between the occiput, atlas, and axis (Fig. 9.13). The alignment between the occiput and C1 must also be checked to exclude an associated counterrotation of the occiput on the atlas (Figs. 9.13 and 9.14) [60]. The duration of torticollis is an important prognostic factor in determining the risk of recurrence [45,61]. If the duration of symptoms is less than three weeks, most children will respond to conservative management. However, if the deformity has been present for longer than this, the likelihood of failure to correct the deformity with traction, or risk of recurrence after correction, is much higher [62]. Some authors have suggested that chronic subluxation results in deformity of the C2 facet, which results in a higher risk of recurrent fixation [63]. Inclination of the atlas by more than 20°, as determined on CT, may also represent a poor prognostic sign for the correction of atlanto-axial fixation [63]. This underscores the importance of prompt diagnosis and timely use of appropriate imaging studies in the evaluation of atlanto-axial rotatory fixation.

C

Fig. 9.15 Dynamic CT for evaluation of atlanto-axial rotatory fixation. (A–C) Axial CT images obtained with the head turned to the right demonstrate rotatory subluxation of C1 on C2 with the right lateral mass acting as the pivot (type II). (D,E) Repeat scan with the maximum voluntary turning of the head to the left demonstrates no change in alignment of C1 relative to C2, confirming a diagnosis of atlantoaxial rotatory fixation. (With permission from Khanna and El-Khoury [11].)

Atlanto-axial subluxation may be rarely accompanied by occipito-atlantal counter-rotation [64]. The subluxation at the occipito-atlantal joint may occur secondary to the same insult that caused the atlanto-axial subluxation, or it can occur as an effort to compensate for the atlanto-axial subluxation [60,65,66]. The child or parent might report apparent improvement in the severity of torticollis secondary to the development of compensatory C0–C1 subluxation. Multidetector CT images with multiplanar reformations demonstrate the rotation of C1 on C2 with counter-rotation at C0–C1 resulting in near anatomic alignment of the head with C2 and the lower cervical spine.

Atlas fractures Fractures of the C1 ring are much less common than atlantoaxial subluxation or AOD. The atlas is protected from injury by the elasticity of the atlas synchondrosis, which acts as a buffer; the flexibility of the pediatric neck; and the plasticity of the skull. A C1 burst fracture (Jefferson fracture) occurs secondary to axial loading forces, such as diving accidents or motor vehicle crashes resulting in impaction of the head against the car top. Because of the mechanism of injury, Jefferson fractures are found more often in teenagers than in young children. A similar fracture can occur in young children secondary to a fall on the top of the head. The open-mouth view may show offset of the lateral masses of C1 on C2. This should be differentiated

173

9 Vertebral injuries in children

from pseudospread caused by differential growth of the atlas and axis (Fig. 9.8). If the offset is more than 6 mm, rupture of the transverse ligament should be suspected, making this an unstable injury [17]. Unlike adults who generally break the C1 ring in two to four places, children may have an isolated fracture involving one synchondrosis – the weakest point of the atlas in a child [67]. Disruption of the synchondrosis without displacement may make detection difficult on CT, but detection of fluid at the synchondrosis with MRI helps in establishing the diagnosis [68].

Dens fractures Dens synchondrosis fractures are one of the more common types of cervical spine fracture in children younger than eight years of age [69]. The dens fuses with the axis body at the subdental synchondrosis at around six years of age. The combination of a weak zone through the synchondrosis and the relatively large head size makes the synchondrosis prone to injury. A total of 55 cases of dens synchondrosis fractures in children have been reported in the literature, with a mean age of 2.8 years [70]. Lateral radiographs most often show anterior angulation of the dens. The fractured dens tends to displace anteriorly; consequently, most patients do not have neurologic deficits (Fig. 9.16). In children, these fractures have a good outcome without growth disturbance as the synchondrosis is located below the level of the vascular supply to the dens [71]. Dens fractures in older children and adults typically occur above the synchondrosis and can interrupt the blood supply to the odontoid, resulting in complications such as non-union and pseudoarthrosis. Fractures of the axis body and of the neural arches are rare in children. The usual mechanism is hyperextension and may Fig. 9.16 Dens fracture. Lateral view of the cervical spine demonstrates fracture through the synchondrosis of the dens (arrow), with anterior displacement of the dens.

174

be secondary to non-accidental trauma [72]. The hanged-man fracture is characterized by bilateral pars interarticularis fractures and anterior subluxation of C2 on C3 (Fig. 9.17). This fracture typically widens the vertebral canal and, therefore, neurologic deficits are rare.

Subaxial cervical injuries Subaxial cervical injuries are injuries involving the C3 through C7 level. Although these are uncommon in young children, they are the predominant site of cervical spine injury in older children and adolescents [73]. The most common site of these injuries is between the C5 and C7 levels [73,74]. As in adults, four patterns of injury may occur in the subaxial cervical spine: fractures, fracture with dislocation, dislocation, and purely ligamentous injury. The most common fractures in adolescents are vertebral body compression fractures, or facet fractures and dislocations caused by hyperflexion [75]. These injuries are extremely unusual in children below the age of eight years. In the series of Ehara and colleagues [13], the youngest child with a facet injury and compression fracture was 14 years of age. The clinical and radiographic patterns of lower cervical spine injuries are similar to those in adults and will not be discussed further as they have been well described elsewhere in this book.

Spinal cord injury without radiographic abnormality Spinal cord injury without radiographic abnormality (SCIWORA) is defined as presence of objective signs of myelopathy secondary to trauma with no abnormality noted on radiographs or CT. The SCIWORA syndrome was first described by Pang and colleagues [19] in a series of 24 children who were radiographically normal but had neurologic Fig. 9.17 Hangedman fracture in a three-year-old boy in a motor vehicle collision. Cross-table lateral view of the cervical spine demonstrates fracture through the lamina of C2 (arrow). (With permission from Khanna and El-Khoury [11].)

9 Vertebral injuries in children

findings. The reported incidence of SCIWORA varies widely, but it likely accounts for 10 to 20% of all pediatric cord injuries [76]. As with other cervical spine injuries, the frequency, severity, and level of SCIWORA in children under nine years of age differ from older children. The occurrence of SCIWORA can account for up to 63% of cervical spine injuries in children under nine years of age, while the reported incidence in older children is 20% [77]. Younger children are also more likely to have involvement of the upper cervical spine as a result of SCIWORA with more neurologic deficits and poor prognosis for recovery. The potential mechanisms involved in the pathogenesis of SCIWORA in children include hyperextension, hyperflexion, distraction, and spinal cord ischemia. The elasticity of the pediatric spine makes it susceptible to momentary dislocation and spontaneous reduction, resulting in a normal radiographic appearance. Although radiography and CT are by definition negative in patients with SCIWORA, MR images can demonstrate a variety of neural and extraneural findings. The neural findings could vary from cord edema, to minor or major hemorrhage, to complete cord transection. The extraneural findings are more useful for assessing the stability of the spinal column and include ligamentous disruption, edema or hemorrhage in the paraspinal muscles, disc edema or herniation, or epidural/subdural hematomas (Fig. 9.18; see Fig. 6.5) [78]. With the increasing use of MRI for evaluation of spinal injuries, some authors have suggested limiting the usage of the term SCIWORA to those situations where there is no abnormality on MRI [79].

A

B

Thoracolumbar spine Injury to the thoracic or lumbar spine in children accounts for only 0.6% to 0.9% of all spinal injuries [80]. The majority of thoracolumbar spinal trauma in children occurs in those aged 14 to 16 years [81]. That having been said, the thoracic spine is the second most common site of injury in children younger than 14 years of age [6,82]. The most striking difference between the pediatric and adult thoracolumbar spine is the higher cartilage and water content in the former. The pediatric disc and vertebral body act as much better shock absorbers, resulting in a lower incidence of bony injury [83]. The most frequently injured levels of the spine are T5 to T12 followed by T12 to L2 [84]. Compression fractures of the thoracic spine are the most common injury seen in children and adolescents, followed by burst fractures and Chance-type fractures [85]. Compression fractures are common because of the wedge shape of the immature vertebral body and the presence of a natural kyphosis [81]. Radiographically anterior wedging of the vertebral body is seen. A difference in height of more than 3 mm between the anterior and posterior cortices of the vertebral body is considered significant and helps to differentiate between normal wedging and a compression fracture. Restoration of vertebral body height occurs in children who have active vertebral physes if the amount of wedging is less than 30° [84]. Burst fractures are found almost exclusively in the older child. These typically occur in the lumbar spine and thoracolumbar region secondary to axial loading forces. These fractures involve the anterior and middle column. Cross-sectional

Fig. 9.18 Spinal cord injury without radiographic abnormality (SCIWORA) in a three-year-old child with hemiplegia involved in a motor vehicle collision. Cervical spine plain films and CT (not shown) were normal. Sagittal T1-weighted (A) and T2-weighted (B) MR images demonstrate an intramedullary hematoma at the T2 level (arrow).

175

9 Vertebral injuries in children

imaging is essential to evaluate for retropulsed bone fragments, generally from the posterosuperior corner, which may cause spinal canal impingement. Chance-type fracture or seat-belt injury (a flexion–distraction injury) can also occur in children. Children in motor vehicle accidents are at a higher risk of Chance-type fractures, for two reasons: children have a higher center of gravity because of their larger head–body ratio, and the immature iliac crests can lead to improper positioning of the seat belt over the abdomen rather than the pelvis. Unlike adults, Chance-type fractures in children may involve the growth plate of the vertebral body. Pediatric Chance-type fractures have been classified into three types based on MRI findings: type I, with physeal injury through the superior growth plate with posterior lesion through the soft tissues above the pedicle or superior facet fracture; type II, with fracture through the vertebral body and posterior elements; and type III, with physeal injury through the inferior growth plate with posterior lesion through the soft tissues below the pedicle or inferior facet fracture (Fig. 9.19) [86]. Up to 70% of children with Chance-type fractures can have associated intraabdominal visceral and/or vascular injuries [87,88]. Another injury that is unique to the pediatric population is a slipped vertebral apophysis. This is analogous to disc herniation in the adults and most frequently occurs at the L4 level [89]. Scheuermann disease [90] is the result of minor trauma at the discovertebral junction in adolescent boys and girls.

A

176

B

The pathophysiology relates to herniation of disc material into the vertebral body. Three patterns are recognized: central herniation into the vertebral body (Schmorl node, Fig. 9.20A), marginal herniation through the vertebral endplate to produce a “limbus” deformity (Fig. 9.20B), and multiple small herniations into the vertebral endplate to produce irregularity, and often a kyphotic deformity (Fig. 9.20C). In most instances, the disease is diagnosed as an incidental finding on a lateral chest or thoracic spine radiograph or a sagittal reconstructed CT image. The findings include disc space narrowing, disc margin irregularity, limbus deformities, single or multiple Schmorl nodes, or kyphosis (Fig. 9.21) [90].

Summary Pediatric cervical spine injuries in children under eight years of age are rare but carry a high risk of morbidity and mortality. The upper cervical spine is the region most susceptible to injury in this patient population, followed by the thoracic spine. As the vertebral column matures, the spectrum of injury in older children approximates that of adults. Knowledge of the normal development of the pediatric spine, its normal radiographic appearance, and specific injuries seen in children should facilitate the appropriate use of diagnostic modalities and aid the accurate interpretation of pediatric spine imaging.

Fig. 9.19 Chance-type fracture of the lumbar spine in a two-year-old child who was restrained by a lap belt in the back seat of a car that collided head on with another car. (A) Lateral view of the lumbar spine shows widening of the posterior elements at L1–L2 with widening of the posterior aspect of the intervertebral disc space (arrow). (B) Sagittal T2-weighted image shows disruption of the posterior ligament complex with the fracture extending through the inferior growth plate (arrow).

9 Vertebral injuries in children

A

B

Fig. 9.20 Abnormalities of the discovertebral junction. (A) Schmorl node from central herniation. (B) “Limbus” deformity from marginal disc herniation with fracture. (C) Scheuermann disease with kyphosis. Multiple marginal small herniations produce the deformity.

C

A

B

Fig. 9.21 Scheuermann disease. (A) Lateral radiograph shows slight anterior wedging and disc space narrowing and irregularity along the margins (arrows). (B) Lateral radiograph in another patient with similar findings (arrows). The deformities have resulted in kyphosis.

177

9 Vertebral injuries in children

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

178

Galano GJ, Vitale MA, Kessler MW, Hyman JE, Vitale MG. The most frequent traumatic orthopedic injuries from a national pediatric inpatient population. J Pediatr Orthop 2005; 25:39–44. Hachen HJ. Spinal cord injury in children and adolescents: diagnostic pitfalls and therapeutic considerations in the acute stage. Paraplegia 1977; 15:55–64. Kewalramani LS, Kraus JF, Sterling HM. Acute spinal-cord lesions in a pediatric population: epidemiological and clinical features. Paraplegia 1980;18:206–219. Hamilton MG, Myles ST. Pediatric spinal injury: review of 61 deaths. J Neurosurg 1992;77:705–708. Osenbach RK, Menezes AH. Pediatric spinal cord and vertebral column injury. Neurosurgery 1992;30:385–390. Vitale MG, Goss JM, Matsumoto H, Roye DP Jr. Epidemiology of pediatric spinal cord injury in the United States: years 1997 and 2000. J Pediatr Orthop 2006;26:745–749. Hill SA, Miller CA, Kosnik EJ, Hunt WE. Pediatric neck injuries. A clinical study. J Neurosurg 1984;60:700–706. Anderson JM, Schutt AH. Spinal injury in children: a review of 156 cases seen from 1950 through 1978. Mayo Clin Proc 1980;55:499–504. Hadley MN, Zabramski JM, Browner CM, Rekate H, Sonntag VK. Pediatric spinal trauma. Review of 122 cases of spinal cord and vertebral column injuries. J Neurosurg 1988;68:18–24. Leonard M, Sproule J, McCormack D. Paediatric spinal trauma and associated injuries. Injury 2007;38:188–193. Khanna G, El-Khoury GY. Imaging of cervical spine injuries of childhood. Skeletal Radiol 2007;36:477–494. Bailey DK. The normal cervical spine in infants and children. Radiology 1952; 59:712–719. Ehara S, El-Khoury GY, Sato Y. Cervical spine injury in children: radiologic manifestations. AJR Am J Roentgenol 1988;151:1175–1178. Nitecki S, Moir CR. Predictive factors of the outcome of traumatic cervical spine fracture in children. J Pediatr Surg 1994; 29:1409–1411. Townsend EH Jr., Rowe ML. Mobility of the upper cervical spine in health and disease. Pediatrics 1952;10:567–574.

16. Penning L. Normal movements of the cervical spine. AJR Am J Roentgenol 1978;130:317–326. 17. Bonadio WA. Cervical spine trauma in children: Part II. Mechanisms and manifestations of injury, therapeutic considerations. Am J Emerg Med 1993; 11:256–278. 18. Sullivan CR, Bruwer AJ, Harris LE. Hypermobility of the cervical spine in children; a pitfall in the diagnosis of cervical dislocation. Am J Surg 1958; 95:636–640. 19. Pang D, Wilberger JE Jr. Spinal cord injury without radiographic abnormalities in children. J Neurosurg 1982;57:114–129. 20. Leventhal HR. Birth injuries of the spinal cord. J Pediatr 1960;56:447–453. 21. Sherk HH, Schut L, Lane JM. Fractures and dislocations of the cervical spine in children. Orthop Clin North Am 1976; 7:593–604. 22. Lawson JP, Ogden JA, Bucholz RW, Hughes SA. Physeal injuries of the cervical spine. J Pediatr Orthop 1987; 7:428–435. 23. Tepas JJ, III, DiScala C, Ramenofsky ML, Barlow B. Mortality and head injury: the pediatric perspective. J Pediatr Surg 1990;25:92–95; discussion 96. 24. Broder J, Warshauer DM. Increasing utilization of computed tomography in the adult emergency department, 2000–2005. Emerg Radiol 2006; 13:25–30. 25. Broder J, Fordham LA, Warshauer DM. Increasing utilization of computed tomography in the pediatric emergency department, 2000–2006. Emerg Radiol 2007;14:227–332. 26. 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. National Emergency X-Radiography Utilization Study Group. N Engl J Med 2000; 343:94–99. 27. Viccellio P, Simon H, Pressman BD, et al. A prospective multicenter study of cervical spine injury in children. Pediatrics 2001;108:E20. 28. Jimenez RR, Deguzman MA, Shiran S, Karrellas A, Lorenzo RL. CT versus plain radiographs for evaluation of C-spine injury in young children: do benefits outweigh risks? Pediatr Radiol 2008;38:635–644.

29. Anderson RC, Scaife ER, Fenton SJ, et al. Cervical spine clearance after trauma in children. J Neurosurg 2006; 105(5 Suppl):361–364. 30. Frank JB, Lim CK, Flynn JM, Dormans JP. The efficacy of magnetic resonance imaging in pediatric cervical spine clearance. Spine 2002;27:1176–1179. 31. Platzer P, Jaindl M, Thalhammer G, et al. Clearing the cervical spine in critically injured patients: a comprehensive C-spine protocol to avoid unnecessary delays in diagnosis. Eur Spine J 2006; 15:1801–1810. 32. Swischuk LE, John SD, Hendrick EP. Is the open-mouth odontoid view necessary in children under 5 years? Pediatr Radiol 2000;30:186–189. 33. Cattell HS, Filtzer DL. Pseudosubluxation and other normal variations in the cervical spine in children. A study of one hundred and sixty children. J Bone Joint Surg Am 1965;47:1295–1309. 34. Swischuk LE. Anterior displacement of C2 in children: physiologic or pathologic. Radiology 1977;122: 759–763. 35. Suss RA, Zimmerman RD, Leeds NE. Pseudospread of the atlas: false sign of Jefferson fracture in young children. AJR Am J Roentgenol 1983;140:1079–1082. 36. Budin E, Sondheimer F. Lateral spread of the atlas without fracture. Radiology 1966;87:1095–1098. 37. Tulsi RS. Growth of the human vertebral column. An osteological study. Acta Anat (Basel) 1971;79:570–580. 38. Swischuk LE, Swischuk PN, John SD. Wedging of C-3 in infants and children: usually a normal finding and not a fracture. Radiology 1993;188:523–526. 39. Loder R. The cervical spine. In Morrissy RT, Weinstein SL, eds. Lovell and Winter’s Pediatric Orthopedics, 6th edn. Philadelphia, PA: Lippincott-Raven, 2005, pp. 871–919. 40. Wackenheim A. Roentgen Diagnosis of the Craniovertebral Region. New York: Springer, 1974. 41. Locke GR, Gardner J, Van Epps EF. Atlas–dens interval (ADI) in children: a survey based on 200 normal cervical spines. Am J Roentgenol Radium Ther Nucl Med 1966;97:135–140. 42. Wholey MH, Bruwer AJ, Baker HL Jr. The lateral roentgenogram of the neck; with comments on the atlanto– odontoid–basion relationship. Radiology 1958;71:350–356.

9 Vertebral injuries in children

43. Harris JH, Carson GC, Wagner LK. Radiologic diagnosis of traumatic occipitovertebral dissociation: 1. Normal occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 1994; 162:881–886. 44. Powers B, Miller MD, Kramer RS, Martinez S, Gehweiler JA Jr. Traumatic anterior atlanto-occipital dislocation. Neurosurgery 1979;4:12–17. 45. Lee C, Woodring JH, Goldstein SJ, et al. Evaluation of traumatic atlantooccipital dislocations. AJNR Am J Neuroradiol 1987;8:19–26. 46. Kaufman RA, Carroll CD, Buncher CR. Atlantooccipital junction: standards for measurement in normal children. AJNR Am J Neuroradiol 1987;8:995–999. 47. Harris JH, Mirvis SE. The Radiology of Acute Cervical Spine Trauma, 3rd edn. Baltimore, MD: Williams & Wilkins, 1996. 48. Adams VI. Neck injuries: I. Occipitoatlantal dislocation – a pathologic study of twelve traffic fatalities. J Forensic Sci 1992;37:556–564. 49. Bucholz RW, Burkhead WZ. The pathological anatomy of fatal atlantooccipital dislocations. J Bone Joint Surg Am 1979;61:248–250. 50. Bloom BM, Powell BP. Surviving atlanto-occipital dislocation. Emerg Med Australas 2007;19:379–382. 51. Hosalkar HS, Cain EL, Horn D, et al. Traumatic atlanto-occipital dislocation in children. J Bone Joint Surg Am 2005; 87:2480–2488. 52. Harris JH Jr., Carson GC, Wagner LK, Kerr N. 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. 53. Dziurzynski K, Anderson PA, Bean DB, et al. A blinded assessment of radiographic criteria for atlantooccipital dislocation. Spine 2005; 30:1427–1432. 54. Pang D, Nemzek WR, Zovickian J. Atlanto-occipital dislocation, part 1: normal occipital condyle–C1 interval in 89 children. Neurosurgery 2007; 61:514–521; discussion 521. 55. Pang D, Nemzek WR, Zovickian J. Atlanto-occipital dislocation, part 2: the clinical use of (occipital) condyle–C1 interval, comparison

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

with other diagnostic methods, and the manifestation, management, and outcome of atlanto-occipital dislocation in children. Neurosurgery 2007;61: 995–1015; discussion 1015. Roche CJ, O’Malley M, Dorgan JC, Carty HM. A pictorial review of atlantoaxial rotatory fixation: key points for the radiologist. Clin Radiol 2001;56: 947–958. Hohl M. Normal motions in the upper portion of the cervical spine. J Bone Joint Surg Am 1964;46:1777–1779. Kowalski HM, Cohen WA, Cooper P, Wisoff JH. Pitfalls in the CT diagnosis of atlantoaxial rotary subluxation. AJR Am J Roentgenol 1987;149:595–600. Fielding JW, Hawkins RJ. Atlantoaxial rotatory fixation (fixed rotatory subluxation of the atlanto-axial joint). J Bone Joint Surg Am 1977;59:37–44. Clark CR, Kathol MH, Walsh T, El-Khoury GY. Atlantoaxial rotatory fixation with compensatory counter occipitoatlantal subluxation. A case report. Spine 1986;11:1048–1050. Phillips WA, Hensinger RN. The management of rotatory atlanto-axial subluxation in children. J Bone Joint Surg Am 1989;71:664–668. Subach BR, McLaughlin MR, Albright AL, Pollack IF. Current management of pediatric atlantoaxial rotatory subluxation. Spine 1998;23:2174–2179. Ishii K, Chiba K, Maruiwa H, et al. Pathognomonic radiological signs for predicting prognosis in patients with chronic atlantoaxial rotatory fixation. J Neurosurg Spine 2006;5:385–391. Washington ER. Non-traumatic atlantooccipital and atlanto-axial dislocation; a case report. J Bone Joint Surg Am 1959; 41A:341–344. Altongy JF, Fielding JW. Combined atlanto-axial and occipito-atlantal rotatory subluxation. A case report. J Bone Joint Surg Am 1990;72:923–926. Bouillot P, Fuentes S, Dufour H, Manera L, Grisoli F. Imaging features in combined atlantoaxial and occipitoatlantal rotatory subluxation: a rare entity. Case report. J Neurosurg 1999;90(2 Suppl):258–260. Hagino T, Ochiai S, Tonotsuka H, et al. Fracture of the atlas through a synchondrosis of the anterior arch complicated by atlantoaxial rotatory fixation in a four-year-old child. J Bone Joint Surg Br 2006;88:1093–1095.

68. Judd DB, Liem LK, Petermann G. Pediatric atlas fracture: a case of fracture through a synchondrosis and review of the literature. Neurosurgery 2000; 46:991–994; discussion 994–5. 69. Sherk HH, Nicholson JT, Chung SM. Fractures of the odontoid process in young children. J Bone Joint Surg Am 1978;60:921–924. 70. Fassett DR, McCall T, Brockmeyer DL. Odontoid synchondrosis fractures in children. Neurosurg Focus 2006;20:E7. 71. Blockey NJ, Purser DW. Fractures of the odontoid process of the axis. J Bone Joint Surg Br 1956;38B:794–817. 72. Lam FC, Mehta V, Fox R. Spondylolisthesis of C2 in an eightweek-old infant: long term followup. Can J Neurol Sci 2007;34:372–374. 73. Dogan S, Safavi-Abbasi S, Theodore N, et al. Pediatric subaxial cervical spine injuries: origins, management, and outcome in 51 patients. Neurosurg Focus 2006;20:E1. 74. Kokoska ER, Keller MS, Rallo MC, Weber TR. Characteristics of pediatric cervical spine injuries. J Pediatr Surg 2001;36:100–105. 75. McGrory BJ, Klassen RA, Chao EY, Staeheli JW, Weaver AL. Acute fractures and dislocations of the cervical spine in children and adolescents. J Bone Joint Surg Am 1993;75:988–995. 76. Congress of Neurological Surgeons. Management of acute central cervical spinal cord injuries. Neurosurgery 2002; 50(Suppl):S166–S172. 77. Pang D. Spinal cord injury without radiographic abnormality in children, 2 decades later. Neurosurgery 2004; 55:1325–1342; discussion 1342–1343. 78. Grabb PA, Pang D. Magnetic resonance imaging in the evaluation of spinal cord injury without radiographic abnormality in children. Neurosurgery 1994;35:406–414; discussion 414. 79. Yucesoy K, Yuksel KZ. SCIWORA in MRI era. Clin Neurol Neurosurg 2008; 110:429–433. 80. Slotkin JR, Lu Y, Wood KB. Thoracolumbar spinal trauma in children. Neurosurg Clin N Am 2007; 18:621–630. 81. Dogan S, Safavi-Abbasi S, Theodore N, et al. Thoracolumbar and sacral spinal injuries in children and adolescents: a review of 89 cases. J Neurosurg 2007; 106(Suppl):426–433.

179

9 Vertebral injuries in children

82. Ruge JR, Sinson GP, McLone DG, Cerullo LJ. Pediatric spinal injury: the very young. J Neurosurg 1988;68:25–30. 83. Ferguson RL. Thoracic and lumbar spinal trauma of the immature spine. In Herkowitz HN, Rothman RH, Simeone FA, eds. Rothman-Simeone, The Spine. Philadelphia, PA: Saunders-Elsevier, 2006, pp. 603–612. 84. Clark P, Letts M. Trauma to the thoracic and lumbar spine in the adolescent. Can J Surg 2001;44:337–345.

180

85. Santiago R, Guenther E, Carroll K, Junkins EP Jr. The clinical presentation of pediatric thoracolumbar fractures. J Trauma 2006; 60:187–192. 86. de Gauzy JS, Jouve JL, Violas P, et al. Classification of Chance fracture in children using magnetic resonance imaging. Spine 2007;32:E89–E92. 87. Mulpuri K, Reilly CW, Perdios A, Tredwell SJ, Blair GK. The spectrum of abdominal injuries associated with Chance fractures in pediatric

patients. Eur J Pediatr Surg 2007; 17:322–327. 88. Swischuk LE, Jadhav SP, Chung DH. Aortic injury with Chance fracture in a child. Emerg Radiol 2008;15:285–287. 89. Sovio OM, Bell HM, Beauchamp RD, Tredwell SJ. Fracture of the lumbar vertebral apophysis. J Pediatr Orthop 1985;5:550–552. 90. Resnick D. Diagnosis of Bone and Joint Disorders, 4th edn. Philadelphia, PA: W.B. Saunders, 2002, pp. 3725–3730.

Chapter

10

Vertebral stability and instability Richard H. Daffner

One of the problems facing clinicians who deal with victims of vertebral injury is determining whether the vertebral column is stable. The subject of stability is a source of controversy among spine surgeons as well as radiologists. Physicians dealing with vertebral injuries need to know specific information: the presence and location of all fractures, whether there is encroachment of the vertebral canal or neural foramina by bone or herniated disc fragments, and the results of imaging assessment of vertebral stability. As a rule, the decision of whether or not operative intervention is needed rests on the issues of encroachment and vertebral stability. Stability of the vertebral column depends on the integrity of the major skeletal components, the intervertebral discs, facet joints, and the ligaments (Fig. 10.1) [1–3]. Stability is defined as the ability of the vertebral column to maintain its normal alignment, to provide support for the head and torso, and to protect the neural elements under normal physiologic stress [3]. Unstable injuries are those that have the potential to cause progressive neurologic deterioration or skeletal deformity under normal physiologic motion or loading [3]. If the instability is in the craniocervical junction, death may be a consequence. For a lesion to be considered unstable, enough damage must have occurred to allow abnormal motion at, above, or below the site(s) of injury. Disruption of any one of the elements mentioned above may not necessarily produce an unstable injury; it is generally a combination of abnormalities that results in instability [4–7]. The unstable vertebral column requires surgical intervention to obtain the best functional outcome. The recognition and treatment of posttraumatic vertebral instability, however, is, as mentioned above, both difficult and controversial [8,9]. The understanding of the concepts of vertebral stability and instability has evolved during the past half century. In 1960, Roaf [10] demonstrated that simple hyperflexion or hyperextension could not produce rupture of the normal vertebral ligaments. For rupture to occur, rotation or translation (shearing) forces had to be present. His findings gave rise to the concept of the “two-column” vertebral column, which uses the posterior longitudinal ligament as the dividing point. Everything anterior to the posterior longitudinal ligament was called the anterior column, and everything posterior was the posterior column [10]. In an attempt to form a unifying concept of vertebral instability of the thoracolumbar region, Holdsworth [7]

proposed that disruption of the posterior ligamentous complex alone was sufficient to produce instability. He believed that the posterior column was the key to stability. He did not take into account posterior extrusion of bone fragments impinging on the vertebral canal or disruption of the posterior aspect of the intervertebral disc. Although Holdsworth’s concept was held in high regard for many years, a number of experimental studies revealed that rupture of the posterior longitudinal ligament and a portion of the annulus fibrosus were necessary to produce instability [2,4,5,11]. This paved the way for Denis [4,5] to develop a new biomechanical theory and classification of vertebral injury in 1983. Denis’ classification was based on the concept of dividing the vertebral apparatus into three distinct anatomic columns rather than the traditional two (Fig. 10.2): • the anterior column extends from the anterior longitudinal ligament to a vertical line drawn through the junction point of the middle and posterior third of the intervertebral disc • the middle column extends from this line to the posterior longitudinal ligament • the posterior column extends from the posterior longitudinal ligament through the supraspinous ligament and, therefore, includes all of the posterior skeletal and ligamentous structures. Denis was able to demonstrate that instability would result only when disruption occurred in two contiguous zones or, more specifically, when the middle zone was disrupted. This can occur only in conjunction with anterior or posterior disruption; middle zone disruption never occurs alone. What imaging findings indicate instability? In 1981, Gehweiler and colleagues [12] studied a group of patients with acute thoracolumbar vertebral injury and found that all of the unstable lesions had at least one of the following signs: • vertebral displacement (Fig. 10.3) • widened interlaminar or interspinous spaces (Figs. 10.4 and 10.5) • widening of the apophyseal (facet) joints (Figs. 10.6 and 10.7) • widening and elongation of the vertebral canal manifest as widening of the interpedicle distance in transverse and vertical planes (Figs. 10.8 and 10.9) [4].

181

10 Vertebral stability and instability

Fig. 10.1 Schematic drawing of the ligaments of the vertebral column in the sagittal plane. In addition, there are ligaments between the facet joints (not shown). ALL, anterior longitudinal ligament; IS, intraspinous ligament; LF, ligamentum flavum; PLL, posterior longitudinal ligament; SS, supraspinous ligament.

Fig. 10.2 Denis’ three-column concept of vertebral stability. See text for description. A, anterior column; M, middle column; P, posterior column.

A

B

C

Fig. 10.3 Displacement. (A) Lateral radiograph in a patient with a hanged-man fracture of C2. There is anterolisthesis of the body of C2 and a massive prevertebral hematoma (*). (B) Thoracic dislocation of T11 on T12 in an abused child. (C) Sagittal reconstructed CT image in the same patient shows facet lock (arrow).

A

182

B

Fig. 10.4 Flexion sprain C4–C5. (A) Lateral radiograph shows widening of the interlaminar distance (*). (B) Frontal radiograph shows the widening also (double arrow).

10 Vertebral stability and instability

Fig. 10.5 Wide interspinous space (double arrow) in a patient with a flexion distraction injury at T12– L1. Note the wide facet joint (large single arrow) and fracture through the pedicle (small single arrow) on the left.

A

B

In 1990, Daffner and colleagues [13] added disruption of the posterior vertebral body line as a fifth sign (Figs. 10.10 and 10.11). An additional sign that we now recognize is widening of the disc space. Only one of these signs needs to be present to make a radiographic diagnosis of an unstable injury. Vertebral displacement of more than 2  mm implies injury to the major ligamentous and articular structures (Fig. 10.12). As discussed in Chapter 8, 2 mm is given as the upper limits of normal motion at any level. Many patients with degenerative changes in the vertebral column have up to 2  mm of anterolisthesis or retrolisthesis (Fig. 10.13). In these patients, if lateral flexion and extension radiographs show no further motion, the deformity is considered fixed or “stable.” In the unstable situation, extensive skeletal and ligamentous damage must occur to produce displacement of an entire vertebra or major portion of one. Hence, all three zones are involved. Widening of the interlaminar or interspinous space can occur only when there is injury to the posterior ligamentous structures, the facet joints, and the posterior aspect of the

C

Fig. 10.6 Wide facet joints (arrows) in three patients. (A) Lateral radiograph in a patient with a flexion sprain at C4–C5. Note the wide interlaminar distance (*). (B) Axial CT image in a patient with a thoracic subluxation. (C) Axial CT image in a patient with a distraction injury. Note how wide the facet joints are in this patient.

A

B

Fig. 10.7 “Naked facets” in a patient with T12–L1 distraction dislocation. (A) Frontal radiograph shows the unpaired (“naked”) bilateral facets (black arrows). There is widening of the interspinous space (*) and a fracture of the transverse process of L1 on the left (white arrow). (B) Axial CT image shows the “naked” facet on the right (white arrow) and an area devoid of posterior elements (*).

183

10 Vertebral stability and instability

A

Fig. 10.8 Wide interpedicle distance in a burst fracture of L1. (A) Frontal radiograph shows the wide interpedicle distance (double arrow). (B) Axial CT image shows sagittal fractures anteriorly and posteriorly (arrows).

B

A

B

C

Fig. 10.9 Wide interpedicle distance in a burst fracture of L1. (A) Coronal reconstructed CT image shows a sagittal fracture (arrow) through the body of L1. (B) Coronal reconstructed image slightly posterior shows widening of the interpedicle distance (double arrow). (C) Axial CT image shows sagittal fractures through the body and lamina.

A

184

B

Fig. 10.10 Abnormal posterior vertebral body line (PVL) in a burst fracture of C6. (A) Lateral radiograph shows duplication of the PVL (arrows). (B) Axial CT image shows a sagittal fracture through the body of C6 with malalignment of the right and left bone fragments (arrows). This was responsible for the duplicated PVL on the radiograph.

10 Vertebral stability and instability

B

A

Fig. 10.12 Cervical dislocation with bilateral facet lock C5–C6.

A

B

Fig. 10.11 Abnormal posterior vertebral body line (PVL) in a burst fracture of L1. (A) Lateral radiograph shows compression of the body of L1 and posterior displacement of a fragment of bone from the PVL (arrow). (B) Axial CT image shows the displaced bone fragment (arrow) narrowing the vertebral canal.

Fig. 10.13 Minimal retrolisthesis (arrow) at C3 on C4 in a patient with cervical spondylosis.

Fig. 10.14 Chance-type fracture of L3. (A) Frontal radiograph shows horizontal fractures through the body and pedicles (black arrows) as well as the left transverse process (white arrow). (B) Lateral radiograph shows posterior distraction through the pedicles (arrows).

185

10 Vertebral stability and instability

A

Fig. 10.15 Flexion sprain C5–C6. (A) Lateral radiograph shows widening of the interlaminar (interspinous) distance (*). (B) Frontal radiograph shows interspinous widening (double arrow).

B

annulus fibrosus. Widening of the interlaminar spaces indicates disruption of the middle and posterior zones [1,4,5,13]. Not surprisingly, these findings are most common in flexion– distraction injuries (Figs. 10.14 and 10.15) and rotary injuries. Widening of the facet joints is another indicator of severe posterior ligamentous damage. It generally occurs in conjunction with widening of the interlaminar space. In order to produce facet joint widening, the injury must also extend as far anteriorly as the posterior longitudinal ligament and the annulus fibrosus [2,5,11,13]. There are several radiographic variations of wide facet joints. Widening may occur in hyperflexion sprains (Fig. 10.15) or in hyperextension sprains (Fig. 10.16). Severe distraction produces “naked” facets that can be recognized on radiographs (Fig. 10.17) as well as on a CT scan by the absence of posterior structures on more than one contiguous slice (Fig. 10.18). “Naked” facets can easily be observed on thoracic or abdominal CT examinations. Wide

A

186

B

Fig. 10.16 Wide facets (arrows) in an extension sprain of C5–C6. Note the wide disc space, the “fingerprint” of an extension injury.

Fig. 10.17 “Naked facets” in a flexion dislocation. (A) Frontal radiograph shows widening of the interspinous distance between T12 and L1 (double arrow). (B) Axial CT image shows the “naked” facets and absence of posterior elements of the adjoining vertebra (*).

10 Vertebral stability and instability

A

A

A

Fig. 10.18 Posterior distraction in a patient with a Chance-type fracture of L1. (A) Frontal radiograph shows a horizontal fracture that splits the spinous process of L1 into two parts (arrows). (B) Axial CT image shows the posterior elements to be missing (*).

B

Fig. 10.19 Sagittal cleavage burst fracture of L1. (A) Frontal radiograph shows widening of the interpedicle distance (double arrow). (B) Axial CT image shows the sagittal cleavage fractures through the body and lamina.

B

B

Fig. 10.20 Rotary injury of L1. (A) Lateral radiograph shows compression of the vertebral body and displacement of a segment of the posterior vertebral body line (arrow). (B) Axial CT image shows a large fragment of bone (*) in the vertebral canal. Note the concentric pattern of bone fragments anteriorly.

187

10 Vertebral stability and instability

A

B

C

D

Fig. 10.21 Shearing injury of L2. (A) Frontal radiograph shows a “windswept” appearance of L2 (arrows). (B) Lateral radiograph shows disruption of the posterior vertebral body line (arrows). (C) Axial CT image shows a “windswept” appearance of the vertebra (double arrow) and displacement of a fragment into the vertebral canal (white arrow). (D) Sagittal reconstructed CT image shows a “windswept” appearance (long black arrows) in the sagittal plane as well and retropulsion of a bone fragment into the vertebral canal (white arrow).

facet joints are most commonly found in distraction injuries, dislocations, and rotary injuries. Widening of the vertebral canal (wide interpedicle distance) implies that an injury has occurred to the entire vertebra in the sagittal plane. The most useful approach in identifying the extent of this injury is CT (Fig. 10.19), which can reveal the fractures in both the vertebral body and the lamina. A wide interpedicle distance implies injury to all three zones. Disruption of the posterior vertebral body line is an indicator of injury to both the anterior bony structures and the posterior longitudinal ligament. Although this radiographic

188

finding was initially reported only in burst injuries [1], the sign is also observed in distraction injuries from the use of lap-type seat belts (Fig. 10.14), as well as in injuries caused by rotary (Fig. 10.20) or shearing (Fig. 10.21) mechanisms. Widening of the intervertebral disc space is an infrequent indicator of instability. This finding is most often encountered in extension injuries and implies disruption of the disc space and the posterior ligamentous complex. In the cervical region, this finding may appear after traction has been applied. Table 10.1 summarizes the six imaging signs of instability and the zone disruptions.

10 Vertebral stability and instability

OBTUNDED TRAUMA PATIENT CT on admission –

+

+

Keep C-collar; consult spine service

Day 3 MRI

–

Spine clear: collar off

Fig. 10.22 Protocol for spine “clearance” at Allegheny General Hospital.

Table 10.1 Vertebral instability

Radiographic sign

Zonesa

Displacement

A, M, P

Wide interlaminar space

M, P

Wide facet joint

M, P

Wide vertebral canal Disrupted posterior vertebral body line Wide disc space

A, M, P A, M

Fig. 10.23 Thoracic burst fracture. The CT images (not shown) suggested minimal damage to the posterior vertebral body line. The sagittal STIR MR image shows rupture of the posterior longitudinal ligament (arrow) as well as cord edema and a posterior epidural hematoma.

A, M, (P)

A, anterior; M, middle; P, posterior. a The middle zone is disrupted in all of the signs.

The previous discussion dealt with patients in whom there were recognizable abnormalities on either radiographs or CT studies. What of the patient with “normal” studies? Chapter 4 contained an in-depth discussion of the various methods of “clearing” the spine in unconscious patients [14–23]. Should such patients undergo additional examinations? The consensus in the literature favors using multislice CT as the main tool for examining the vertebral column at all levels [14,18,20,21]. Indeed, Hogan’s article [21] strongly urges this modality and the authors feel it is highly reliable for identifying subtle injuries that may indicate instability [21–23]. Several authors recommend using MRI to look for ligament damage [19,24–27], and some authors favor a combined approach [28,29]. We use combined imaging for patients who remain unconscious after 48 hours in the hospital (Fig. 10.22). In the majority of patients with unstable injuries, however, there will be CT or radiographic abnormalities. Finally, there seems to be agreement in the literature that flexion–extension radiography and/or fluoroscopy is not only inefficient at identifying ligament injuries (and hence instability) but is also not cost-effective [14–18]. Furthermore, it may be downright dangerous!

If the findings of instability are present, it is the radiologist’s duty to inform the referring physician of the situation. The use of the terms stable and unstable in a radiologic report is discouraged, however, because many referring physicians will elect to treat a “mildly” unstable injury conservatively. The use of the term unstable in a radiographic report indicates that operative intervention is mandatory. If surgery is not performed, and there is subsequent deterioration of the patient, an adversarial relationship may develop between the radiologist and the surgeon. It is better to report the individual findings and later communicate the implications verbally to the surgeon rather than to use the word stable or unstable in a written report. The strict imaging criteria that define instability in the cervical or lumbar regions are not always applied in the thoracic region, where the ribs and sternum restrict inherent motion. The association of multiple rib fractures (with or without flail chest) or sternal fractures indicates that the stabilizing influence of the thoracic structures is no longer present [8,9,13]. When these findings occur, they must be reported and emphasized to the attending surgeon, since they will influence his or her decision about surgical intervention. It is for this reason that it is important that thoracolumbar CT studies derived from the thorax–abdomen–pelvic scan have sagittal reconstructed images that include the sternum.

189

10 Vertebral stability and instability

The subject of “stable” burst fractures needs further elaboration. Many burst fractures in the thoracolumbar region are treated conservatively with bracing rather than by surgical intervention. In these instances, the surgeon has determined that the posterior longitudinal ligament is intact. If there are no neurologic deficits, and the degree of displacement of the fragment from the posterior vertebral body is minimal, spine surgeons may attempt to treat these injuries conservatively. In theory, by applying a hyperextension brace, the height that was lost by the fractures is restored and decompression occurs. Gaines and Humphreys [8] have advocated non-operative management for two to four months in this situation. Although these researchers relied extensively on CT scanning for their decision, the use of vertebral MRI can actually demonstrate the integrity or disruption of the posterior longitudinal ligament (Fig. 10.23).

References 1.

2.

Panjabi MM, Hausfield JN, White AA. A biomechanical study of ligamentous stability of the thoracic spine in man. Acta Orthop Scand 1981;52:315–326.

3.

White AA, Panjabi MM. Clinical Biomechanics of the Spine, 2nd edn. Philadelphia, PA: JB Lippincott, 1990.

4.

Denis F. The three-column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 1983;8:817–831.

5.

Denis F. Spinal instability as defined by the three-column spine concept in acute spinal trauma. Clin Orthop Rel Res 1984; 189:65–76.

6.

Bedbrook GM. Stability of spinal fractures and fracture dislocations. Paraplegia 1971;9:23–32.

7.

Holdsworth F. Fractures, dislocations, and fracture–dislocations of the spine. J Bone Joint Surg 1970;52A:1534–1551.

8.

Gaines RW, Humphreys WG. A plea for judgment in management of thoracolumbar fractures and fracture–dislocations: a reassessment of surgical indications. Clin Orthop Rel Res 1984;189:36–42.

9.

190

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.

Jacobs RR, Casey MP. Surgical management of thoracolumbar spinal injuries: general principles and controversial considerations. Clin Orthop Rel Res 1984;189:22–35.

The “fingerprint” approach to vertebral injury described in Chapter 7 may provide the attending surgeon with the necessary information for determining the management of patients with vertebral injury. Simple flexion and extension injuries are, by definition, stable, because the major skeletal and posterior ligamentous structures remain intact. Any injury that involves distraction or dislocation is inherently unstable, because abnormal motion may occur about the site of disruption. The presence of distraction or dislocation indicates that severe skeletal disruption, ligamentous disruption, or both have occurred. Burst, shearing, and rotary fractures disrupt all three columns and are inherently unstable. Disruption of the posterior vertebral body line from any mechanism also implies that an unstable situation exists.

10. Roaf R. A study of the mechanics of spinal injuries. J Bone Joint Surg 1960; 42B:810–823. 11. Panjabi MM, White AA, Johnson R M. Cervical spine mechanics as a function of transection of components. J Biomech 1975;8:327–336. 12. Gehweiler JA Jr., Daffner RH, Osborne RL Jr. Relevant signs of stable and unstable thoracolumbar vertebral column trauma. Skeletal Radiol 1981; 7:179–183. 13. Daffner RH, Deeb ZL, Goldberg AL, et al. The radiologic assessment of posttraumatic vertebral stability. Skeletal Radiol 1990;19:103–108. 14. Daffner RH, Hackney DB. ACR Appropriateness Criteria on suspected spine trauma. J Am Coll Radiol 2007; 4:762–775. 15. Anglen J, Metzler M, Bunn P, Griffiths H. Flexion and extension views are not costeffective in a cervical spine clearance protocol for obtunded trauma patients. J Trauma 2002;52:54–59. 16. Bolinger B, Shartz M, Marion D. Bedside fluoroscopic flexion and extension cervical spine radiographs for clearance of the cervical spine in comatose trauma patients. J Trauma 2004;56:132–136. 17. Freedman I, van Gelderen D, Cooper J, et al. Cervical spine assessment in the unconscious trauma patient: a major trauma service’s experience with passive flexion–extension radiography. J Trauma 2005;58:1183–1188. 18. Padayachee L, Cooper J, Irons S, et al. Cervical spine clearance in unconscious

traumatic brain injury patients: dynamic flexion–extension fluoroscopy versus computed tomography with threedimensional reconstruction. J Trauma 2006;60:341–345. 19. D’Alise MD, Benzel EC, Hart BL. Magnetic resonance imaging evaluation of the cervical spine in the comatose or obtunded trauma patient. J Neurosurg 1999;91:54–59. 20. Diaz JJ, 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;59:897–904. 21. 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. 22. 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. 23. 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. 24. Emery SE, Pathria MN, Wilber RG, et al. Magnetic resonance imaging of posttraumatic spinal ligament injury. J Spinal Disord 1989;2:229–233.

10 Vertebral stability and instability

25. 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. 26. 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.

27. 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. 28. 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. 29. American College of Radiology. ACR Appropriateness Criteria. Suspected Spine Trauma. Reston, VA: American College of Radiology, 2009.

191

Chapter

11

Normal variants and pseudofractures Richard H. Daffner

The preceding chapters discussed using imaging to diagnose fractures and dislocations of the vertebral column. This chapter explores entities that are often confused with fractures and that may be misdiagnosed as fractures. It is just as important for the radiologist and referring physician to be familiar with normal variants and pseudofractures that simulate fractures as it is for them to be able to recognize the fractures themselves [1,2]. This chapter explores five groups of findings that may be misinterpreted as fractures: • normal variants and anomalies • Mach bands • subjective contours • overlapping extraneous materials • old fractures. The list of entities that may mimic fractures is lengthy, so discussion is limited to the salient features of the most common variants and pseudofractures. A few general comments regarding pseudofractures are in order. The majority of acute fractures have sharply defined, irregular margins that will fit together like pieces of a jigsaw puzzle (Fig. 11.1). This principle is valid even when fragments are displaced. In this situation, it is possible to mentally “move” the fragments together. Old, ununited fractures or accessory ossicles have smooth, rounded margins and do not have this “jigsaw puzzle effect” (Fig. 11.2). Some accessory ossicles begin life as acute fractures and later remodel when they fail to unite. The common limbus deformity of the lumbar spine is a good example of this (Fig. 11.3). Similarly, lucent lines caused by vessels have parallel sclerotic margins.

Normal variants and anomalies Human skeletal development provides numerous opportunities for variants and anomalies to occur. Keats and Anderson, in their Atlas of Normal Roentgen Variants That May Simulate Disease [2], devote more than 200 pages to over 500 vertebral variants and anomalies. Another work devotes 150 pages to the same subject [3]. Many of these entities may be confused with fractures or dislocations. Since a complete catalog of such anomalies is beyond the scope of this book, this portion of the chapter describes and illustrates some of the more common

192

variants that are most likely to cause diagnostic difficulty when encountered in patients who are suspected of having sustained vertebral injuries. Many of these variants are the result of normal skeletal vertebral development, which was discussed in Chapter 2.

Cervical column Most of the normal variants found in the cervical column occur in children. Cattell and Filtzer [4] illustrated a number of normal variants in a group of 160 children. They found three major categories of normal variants: (1) variations owing to displacement that resemble subluxation [5,6], (2) variations of curvature that resemble spasm or ligamentous injury [4–7], and (3) variations resulting from skeletal growth centers that resemble fractures [4,8]. The researchers reported that these variations could be encountered in as many as 20% of the normal population [4]. Most of the variants encountered in the cervical column in both adults and children occur at C1 and C2. Anomalies of the atlas vertebra are most common, and foremost of these is failure of fusion of the posterior arch of the atlas [9]. This is easily recognized on the lateral radiograph by the absence of the spinolaminar line at Cl (Fig. 11.4). Although failure of fusion is rarely a cause for concern, when encountered on a frontal view, it could be misinterpreted as a fracture of the posterior arch. On CT, arch defects typically have smooth, often sclerotic borders that do not fit together like pieces of a jigsaw puzzle (Fig.  11.5). In all cases of failure of fusion of the posterior arch of C1, the anterior arch is hypertrophied. This relationship is true for all anomalies of the atlas. Other arch anomalies range from simple failure of fusion of one of the rings of the atlas to complete agenesis (Fig. 11.6). On frontal radiographs, arch anomalies may suggest a vertical dens fracture (Fig. 11.7). Computed tomography shows the true nature of the abnormality. Arch anomalies of the atlas may, however, be associated with unilateral or bilateral lateral atlanto-axial offset of up to 2 mm (Fig. 11.8) [9]. These anomalies must be differentiated from the offset encountered with a bursting fracture of Jefferson. As a rule, Jefferson fractures have lateral atlanto-axial offsets greater than 3 mm (Fig. 11.9) [9].

11 Normal variants and pseudofractures

A

Fig. 11.1 Jigsaw puzzle effect. (A) Drawing of a jigsaw puzzle. Each piece fits precisely into its designated space, regardless of its position. (B) Drawing of a dens fracture showing that the fracture fragments, even if displaced, have the appearance that they will fit together like a jigsaw puzzle. (C) Coronal reconstructed CT image of a dens fracture shows the pieces fitting together like a jigsaw puzzle (arrow). (D) Sagittal reconstructed CT image of another patient with dens fracture shows anterior displacement. The fractured ends (arrows) can be mentally moved into anatomic position.

D

C

A

B

B

C

Fig. 11.2 Accessory ossicles. (A) Drawing of an os odontoideum shows round, smooth borders of the ends of the bones. They will not fit together like a jigsaw puzzle. Lateral radiograph (B) and axial CT image (C) show an accessory spinous process of C2 (O). Note the round, smooth borders and absence of the jigsaw puzzle effect.

Another arch anomaly of the atlas is a cleft through the sulcus of the vertebral artery (Fig. 11.10). Careful inspection of the defect often shows a sclerotic margin around the cleft. Such clefts are often bilateral [9]. Differential growth rates between the atlas and axis in young children often result in bilateral lateral atlanto-axial offset (pseudospread ) (Fig. 11.11) [4,6,10–12]. The degree

of offset may be quite striking. This is a normal physiologic variant resulting from the disparity of “neural” growth of C1 and the “somatic” growth of C2, and not a Jefferson fracture, which is rare before the teenage years [4,10–12]. The relationships between the atlas and axis should be normal in teenagers; by 16 years of age they should have an adult configuration.

193

11 Normal variants and pseudofractures

A

Fig. 11.3 Limbus deformities of L2 and L4. (A) Lateral radiograph shows fragmentation of the anterior superior margins of L2 and L4 (arrows). (B) Axial CT image shows bony fragment (*) without the jigsaw puzzle effect and a Schmorl node (arrow).

B

A

B

Fig. 11.4 Failure of fusion of the posterior arch of the atlas. (A) Lateral radiograph shows absence of the spinolaminar line at C1 (*). Note the hypertrophy of the anterior arch. (B) Axial CT image shows the absent posterior arch (white arrow) and failure of fusion of the anterior arch (black arrow).

There is another normal developmental finding that may mimic subluxation. Kattan [13] described an entity in which differential growth rates of the atlas and axis resulted in apparent backward “displacement” of the spinolaminar line at C2 (Fig. 11.12). With this anomaly, the spinolaminar lines of C1 and C3 and below align, as do the anterior and posterior vertebral body lines of all the vertebrae (including C2). Linear spurlike deformities of the medial aspect of the lateral masses of the atlas are normal variants that should not be mistaken for lateral mass fractures. These are often bilateral and symmetric (Fig. 11.13) [1,2,14].

194

The normal ossification center of the tip of the dens usually appears at around two years of age and fuses by age 12 (Fig. 11.14). This terminal ossicle should not be mistaken for a fracture of the tip of the dens. The so-called type I dens fracture, which reportedly involves the tip of the dens and is considered rare, is, in reality, an unfused apophysis [1–3]. In young children, synchondroses may be misinterpreted as fractures. The synchondrosis of the base of the dens is a lucency extending horizontally across C2 (Fig. 11.15) [5,6]. Sclerotic margins are often found in conjunction with this abnormality. Swischuk and colleagues [8] have described the appearance of

11 Normal variants and pseudofractures

A

B

Fig. 11.5 Failure of fusion of anterior and posterior arches of C1. (A) Openmouth radiograph shows a vertical cleft through the arch of the atlas (arrow). (B) Axial CT image shows clefts anteriorly (arrow) and posteriorly (*). Note the smooth lines and absence of the jigsaw puzzle effect. The lateral radiograph (not shown) demonstrated absence of the spinolaminar line of C1.

A

B

D C

Fig. 11.6 Arch anomalies of the atlas. (A) Complete agenesis of posterior arch. (B,C) Partial agenesis (* in B, arrow in C). (D) Axial CT image of the patient in C shows the partially developed posterior arch (*) as well as failure of fusion of the anterior arch (arrow).

195

11 Normal variants and pseudofractures

A

B

Fig. 11.7 Anterior and posterior arch clefts of C1. (A) Frontal atlanto-axial view shows a vertical lucency (arrow) over the center of the dens. (B) Axial CT image shows the anterior cleft (*) and the posterior cleft (arrow). None of these show a jigsaw puzzle effect.

A

B

Fig. 11.8 Atlanto-axial offset in a patient with failure of fusion of the posterior arch of C1. (A) Lateral radiograph shows absence of the spinolaminar line of C1. (B) Frontal tomogram shows bilateral lateral atlanto-axial offset (arrows). The degree of offset is 2 mm or less. (With permission from Gehweiler et al. [9].)

A

Fig. 11.9 Jefferson fracture of C1. (A) Open-mouth view shows bilateral lateral atlanto-axial offset of greater than 3 mm (arrows). (B) Axial CT image shows fractures of the anterior and posterior arches of C1 (arrows).

196

B

11 Normal variants and pseudofractures

Fig. 11.10 Cleft in the atlas through the sulcus of the vertebral artery (arrow). This is another anomaly that may be associated with atlanto-axial lateral offset. (With permission from Gehweiler et al. [9].)

A

the synchondrosis between the dens and the neural arch. They observed that this structure is never visible on the lateral view but is clearly seen on oblique views, as well as on CT. They believed that the fact that the synchondrosis was not visible on the lateral radiograph allowed one to differentiate it from a true hanged-man fracture, which is very apparent on the lateral view [8]. Smith and coworkers [15] described a case of persistent wide synchondrosis of C2 on the lateral view giving the appearance of a hanged-man fracture. A careful analysis of the figures accompanying this article, however, suggests that this synchondrosis may have, in fact, been a real fracture without significant displacement. Normally, the clefts at the base of the dens develop in a vertical or V-shaped fashion. These also are remnants of the synchondrosis and should not be mistaken for a fracture, particularly when they are unilateral (Fig. 11.16) [2,6,14]. They

B

Fig. 11.11 Pseudospread of the atlas in two children caused by the disparity of growth rates between C1 and C2. The offset (arrows) was great enough in the child shown in B to require a CT scan to confirm that no fracture was present.

A

B

Fig. 11.12 Backward “displacement” of the spinolaminar line of C2 as a result of differential growth rates, shown in two different children. The spinolaminar lines of C1 and C3 (small arrows) do not align with that of C2 (arrowhead). Note the remnant of the dens synchondrosis (large arrow in A).

197

11 Normal variants and pseudofractures

Fig. 11.13 Vertical clefts in the lateral masses of C1 (arrows).

A

B

Fig. 11.14 Normal ossification center of the tip of the dens in a child (arrows). (Courtesy of T. Keats, MD, Charlottesville, VA, USA.)

A

B

C

Fig. 11.15 Normal dens synchondrosis (arrows), shown in three young children. This common normal variant should never be misinterpreted as a fracture. Fractures through this region always result in anterior or posterior displacement.

may be unilateral or bilateral; CT generally shows the true nature of these clefts and solves the problem. Os odontoideum is an abnormality in which the dens is not united to the body of the axis (Fig. 11.17). In this entity, the bony margin of the os and the body of C2 are smooth and rounded. Characteristically, there is overgrowth of the anterior arch of the atlas to compensate for the increased stress placed on that structure by failure of union of the dens with the body of C2 [1,2,16–19]. Two etiologies are proposed for os odontoideum. The first is that it is a congenital failure of fusion of the dens synchondrosis [18]. Most cases of os odontoideum, how-

198

ever, are believed to result from old trauma [16,19]. Ample evidence supports both theories of etiology. In most instances, os odontoideum is recognized when an adult suffers trauma and the abnormality is encountered as an incidental finding. On occasion, neurologic signs result from instability. Os odontoideum, like most accessory ossicles, can be recognized by its smooth borders [16,19] and by the absence of the “jigsaw puzzle” effect, in which the ossicles in question would not fit together in the manner of a jigsaw puzzle (Fig.  11.2). Fractures, by comparison, have irregular, jagged edges that would fit together like a jigsaw puzzle if apposed (Fig. 10.1).

11 Normal variants and pseudofractures

A

B

Fig. 11.16 Clefts at the base of the dens (arrows). (A) Coronal reconstructed CT image. (B) Axial CT image. Occasionally these may be unilateral and suggest a fracture.

A

B

D

E

Ununited ring apophyses (Fig. 11.18) may be misinterpreted as a fracture. These are commonly found along the superior and inferior surfaces of the anterior aspect of the vertebral bodies in adolescents and young adults [1,2,6]. Their location and smooth, round borders are key findings to allow correct interpretation of them. The cervical vertebral column in children is much more mobile than that in adults. Even with minor degrees of flexion, pseudosubluxation may occur (Fig. 11.19) [4,5,10,11].

C

Fig. 11.17 Unstable os odontoideum (O). (A) Lateral extension radiograph shows retrolisthesis of the spinolaminar line of C1 in relation to that of C2 (arrows). (B) Lateral flexion radiograph shows motion of C1 on C2 (arrows). Note the hypertrophy of the anterior arch of the atlas. (C,D) Sagittal (C) and coronal (D) reconstructed CT images show the os (O) is displaced cephalad and to the right. (E) Sagittal inversion recovery MR image shows a bursa (arrow) to have developed between the os and the body of the axis.

The spinolaminar line is important in establishing whether a true subluxation has occurred. In pseudosubluxation, the spinolaminar line is intact. In adults, a similar phenomenon may be encountered in patients with degenerative disc disease (Fig. 11.20), in whom loss of disc height may result in anterolisthesis or retrolisthesis of the body of the vertebra. The spinolaminar line, however, remains relatively intact. In most cases, flexion and extension views demonstrate the deformity to be fixed.

199

11 Normal variants and pseudofractures

A

B

Fig. 11.19 Pseudosubluxation of C2 on C3. Lateral radiograph shows the body of C2 positioned anterior to that of C3 (arrow). The spinolaminar line (solid line) is normal.

Finally, intercalary bones within the anterior longitudinal ligament at the level of the intervertebral disc space should not be misinterpreted as an acute fracture (Fig. 11.21). These ossifications occur commonly in older patients and are usually the result of old trauma. Smooth margins and a lack of a point of origin are clues to their identity. Similarly, ossification in the nuchal ligament (ligamentum nuchae; Fig. 11.22) should not be misinterpreted as a fracture of the spinous process. Monu and associates [20] have described an abnormality of the predental space in which it is V-shaped. This is the result

200

Fig. 11.18 Ring apophyses (arrows) in a teenager (A) and an adult (B), showing an unfused apophysis. Note the lack of a jigsaw puzzle effect.

Fig. 11.20 Retrolisthesis caused by degenerative disease. Posterior osteophytes (arrows) at the C5 disc level encroach upon the vertebral canal. There is narrowing of the C5 disc space and slight posterior displacement of the spinolaminar line (arrowheads) above C6.

of slight posterior tilt of the dens, is not abnormal, and is a function of normal development. The dens tilt angle is measured from a line along the posterior vertebral body line of the dens and the line along the anterior dens contour. The normal angle is up to 35°, with a mean of 17.5°. A second measurement, the predens space angle, is measured between the posterior border of the anterior arch of the atlas and the anterior margin of the dens. This angle can measure up to 22°, with a mean of 6°. These measurements are valid not only on radiographs but also on sagittal CT reconstructed images. These variants and others are illustrated in Keats and Anderson’s atlas [2].

11 Normal variants and pseudofractures

Occasionally, one encounters irregularities of the articular pillars of the lower cervical vertebrae. This may be found on frontal cervical or chest radiographs in patients with significant hyperlordosis of the cervical column, or on pillar views. Smith and Abel [21] and Kattan and Pais [7] thought that these were normal variants. I believe these deformities are of old traumatic origin and are of no clinical consequence. Lawson [19] described abnormalities that were considered clinically significant anatomic variants. In almost every instance, they had smooth, rounded, sclerotic borders. Many of these anomalies are the residua of old injury. Cervical intercalary bones are a prime example.

The use of CT as the primary screening tool for vertebral injuries now allows us to identify the vertebral vasculature, particularly when thin sections are obtained using multidetector units. These vessels frequently perforate through the cortical surface of the vertebral body and may have the appearance of a fracture (Figs. 11.23 and 11.24). Careful review of sagittal and coronal reconstructed images will frequently show the course of these normal vessels. Furthermore, the margins are frequently sclerotic. Finally, congenital fusions and block vertebrae may result in rotation of one or more spinous processes (Fig. 11.25). Similarly, failure of fusion may suggest a fracture (Fig. 11.26). However, the unfused portions of the spinous processes typically have sclerotic margins, indicating that the abnormality is not acute. The best modality is CT for identifying this finding. Recognizing the nature of the anomaly is the key to making a correct diagnosis.

Thoracic column

Fig. 11.21 Intercalary bones (arrows) in the anterior longitudinal ligament.

A

B

The variants in the thoracic column that are most likely to result in a misdiagnosis of a fracture are ununited ring apophyses and Scheuermann disease [2]. United thoracic ring apophyses are encountered on the anterosuperior and anteroinferior aspects of the vertebral bodies (Fig. 11.27), similar to their cervical counterparts. They frequently have sclerotic margins and occur in younger people [2,3]. These findings should be sufficiently characteristic to permit a proper diagnosis. Scheuermann disease, posttraumatic osteochondrosis of the ring apophyses, produces irregularity and occasionally fragmentation along the bony disc plate (Fig. 11.28) [22]. In virtually every case, the disc space at the involved segment is narrowed as a result of the old trauma and occasionally there is herniation of disc material anteriorly. CT is necessary to establish the correct diagnosis by showing extensive involvement along the disc

Fig. 11.22 Nuchal ligament ossification (* in A, arrow in B). This should not be misinterpreted as a spinous process fracture. (A) Lateral radiograph. (B) Sagittal reconstructed CT image in another patient.

201

11 Normal variants and pseudofractures

A

B

C

Fig. 11.23 Vascular groove in an articular pillar. (A) Axial CT image shows a lucency (arrow) in the posterior pillar. Note the sclerotic margins. (B) Axial CT image slightly higher shows no fracture. (C) Sagittal reconstructed CT image shows that the lucency results from a perforating vessel (arrow). These findings are common on CT images of the spine.

A

Fig. 11.24 Vascular channels (arrows) in vertebral bodies. These are common CT findings. (A,B) Cervical. (C) Lumbar.

B

C

margin (Fig. 11.29). Schmorl nodes, particularly in younger patients, are another manifestation of discovertebral trauma [22]. Scheuermann disease frequently results in a kyphotic deformity (Figs. 11.28 and 11.29).

Lumbar column Ununited lumbar ring apophyses, like their cervical and thoracic counterparts, may also be misinterpreted as fractures of the edges of vertebral bodies (Fig. 11.30). These abnormalities often appear similar to the vertebral edge separation (limbus vertebra) that occurs as the result of herniation of disc material through the vertebral body (Figs. 11.31 to 11.33). They are variants of Schmorl node formation (Fig. 11.34) [22]. Smooth sclerotic borders are key to differentiate these abnormalities from simple compression fractures (Fig. 11.35). Occasionally, a vertebral edge separation occurs along the posterior surface

202

11 Normal variants and pseudofractures

Fig. 11.25 Rotation of spinous process of C7 (arrow) caused by congenital fusion of C7 and T1.

A

Fig. 11.27 Normal thoracic ring apophyses (arrows) along the superior and inferior body margins.

of the vertebral body (Figs. 11.36 and 11.37). This is, in fact, a variation of classic lumbar disc herniation. As a group, os odontoideum, accessory ossicles, intercalary bones, Scheuermann disease, limbus vertebrae, and Schmorl node variants may be considered the result of old or repetitive trauma. In most instances, they are encountered as incidental findings in patients who have undergone imaging after traumatic events. As mentioned above, the presence of sclerotic rounded margins and the absence of the jigsaw puzzle effect should enable one to differentiate these old injuries from acute injuries.

B

Fig. 11.28 Scheuermann disease. Injury to the ring apophyses has caused anterior wedging of several vertebral bodies, irregularity along the disc margins (arrows), and narrowing of the involved disc spaces. Kyphotic deformity has resulted from the wedging.

Fig. 11.26 Failure of fusion of the posterior arches of C7 and T1 (arrows). (A) Frontal radiograph showing the failure (arrows). (B) Axial CT image shows smooth margins along the cleft.

Fig. 11.29 Scheuermann disease. Sagittal reconstructed CT image shows anterior wedging, irregularity along the disc margins and multiple Schmorl nodes. There is a kyphotic curve of 35° between T8 and L1.

203

11 Normal variants and pseudofractures

Fig. 11.30 Normal lumbar ring apophyses (arrows).

A

C

204

Pseudofractures Pseudofractures belong to a group of illusory phenomena that result from overlapping images, differences in background illumination, subjective contour formation, and parallax effect. The value of any imaging study depends on the ability of the observer to correctly assess each component contributing to that image. A single radiograph represents a composite of images of all of the structures contained within the plane of the study. Both CT and MR images represent a single slice of a study and contain tissue from a small section of the body. As a rule, the majority of the pseudofractures that will be discussed below are not encountered on cross-sectional imaging studies. However, these studies are plagued by false images that result from partial volume averaging. This occurs when a portion of one anatomic structure is contained in a section containing other structures. It is most likely to occur when the patient is not straight in the CT gantry or in patients with scoliosis (Fig. 11.38). Thin sections (1  mm) and sagittal and coronal

Fig. 11.31 Limbus deformities. (A,B) Lumbar. (C,D) Cervical. (A) Sagittal reconstructed CT image shows fragments (arrows) along the anterior superior margins of L2 and L4. (B) Axial CT image shows a Schmorl node (arrow) and anterior fragmentation. (C) Lateral radiograph shows a limbus fragment (arrow) of the anterior superior margin of the vertebral body. (D) Axial CT section shows the fragmentation (arrow). There is no jigsaw puzzle effect in either of these patients.

B

D

11 Normal variants and pseudofractures

A

A

Fig. 11.32 Anterior and posterior limbus deformities. (A) Sagittal reconstructed CT image shows an anterior superior limbus deformity (arrow) and a posterior superior deformity (arrowhead) of the vertebra below. (B) Axial CT image shows the posterior fragment (arrow) encroaching the vertebral canal. This is essentially a herniated disc with an attached bone fragment.

B

B

reconstructions can help to rule out fractures suggested by the lucencies produced by partial volume averaging. Illusory phenomena result in images that are often falsely interpreted as resulting from significant pathologic abnormalities. The parallax phenomenon results from the location of real structures in relationship to the radiographic beam. It poses no diagnostic problems in vertebral imaging and is not responsible for any pseudofractures. All of the other phenomena are the result of overlapping images of adjacent structures [23–25]. Image perception is a complex process that depends on input from optical, anatomic, physiologic, and psychologic factors, and, ultimately, perceptual learning. Optical factors

Fig. 11.33 Anterosuperior limbus deformity of L4. (A) Lateral radiograph shows the displaced fragment (arrow). (B) Sagittal T1-weighted MR image shows disc material (arrow) extruded between the limbus fragment and the remainder of the body of L4.

include the size of the object and its contour, the type of illumination (direct or background), color, and adjacent shading. Anatomic factors include the information-gathering structures (cornea, iris, and lens), the receptor system (rods and cones of the retina), and the processing system (optic nerve, optic tract, and visual cortex). The physiologic factors include the sensitivity of the rods and cones to stimulation and the effect of lateral inhibition. The psychologic factors are those that make it possible to interpret visual data and are based on previous learning experiences. The final component of image interpretation is that of perceptual learning. The novice to radiologic interpretation sees the same image on the study that an experienced radiologist sees. The learning process, however, teaches the student

205

11 Normal variants and pseudofractures

A

Fig. 11.34 Schmorl nodes. (A) Cervical lesion (arrow) caused by intra-osseous herniation of disc material. The lesion has a sclerotic margin. (B) Multiple thoracic lesions in a patient with Scheuermann disease. Schmorl nodes are encountered commonly in the thoracic and lumbar vertebrae. There is often associated discogenic sclerosis of the vertebral body adjacent to the lesion. The disc spaces of the involved Schmorl nodes are typically narrowed, as in these examples.

B

Fig. 11.35 Simple compression fracture compresses the anterior superior endplate of the vertebra (arrow). The posterior vertebral body line (arrowheads) is intact.

to cull out the important information from the irrelevant and to collate study findings with a pathologic process that can account for those changes [23–25].

Mach bands Mach bands are a perceptual phenomenon in which bright and dark “lines” appear at the borders of structures of different optical (radiographic) density. They are observed in any situation in which a half shadow or penumbra is cast and consequently are found in radiologic studies along the borders of structures of different radiographic density [23–27]. Ernst Mach (1838–1916) was an Austrian physicist–philosopher– psychologist who accidentally discovered the phenomenon in 1865. His original formulation predated, by nearly a century, the discovery of the neural inhibitory interactions within the

A

B

C

Fig. 11.36 Posterior vertebral edge separation and disc herniation. (A) Sagittal reconstructed CT image shows displacement of a bone fragment from the posterior inferior margin of L4 (arrow) into the vertebral canal. (B) Axial CT image at bone window shows typical limbus pathology with canal encroachment (arrow) and Schmorl node formation. (C) Axial CT image at soft tissue window shows the associated herniated disc (*).

206

11 Normal variants and pseudofractures

A

B

A

B

retina and other portions of the nervous system that give rise to the phenomenon [23–27]. Mach bands result from the process of lateral inhibition. A histologic section through the retina reveals an intricate meshwork of nerve fibers. In addition to the primary afferent neurons from each receptor, a fine network of fibers runs from the primary neuron to its neighbors (Fig. 11.39). Impulses passing along these fibers result in an inhibitory effect on the neighboring primary neurons. The effect of this inhibitory response is illustrated in Fig. 11.40. If receptor X is stimulated and its response alone is recorded, that response is as shown on

Fig. 11.37 Posterior inferior vertebral edge separation. (A) Lateral radiograph shows a fragment of bone (arrow) displaced into the vertebral canal. The disc space is narrow. (B) Sagittal T1-weighted MR image shows herniated disc material encroaching the thecal sac (arrow).

Fig. 11.38 Scoliosis. (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) Axial CT image shows partial volume averaging, displaying portions of L1 and L2 simultaneously. The border between the two vertebrae could be misinterpreted as representing a fracture.

the shaded portion of Fig. 11.40. If receptor X and receptor 1 are both stimulated, and the response from receptor X alone is recorded, that response is shown as X + 1. Similarly, stimulating receptor X and any of the other receptors and recording the response from receptor X only produces a gradation of responses. Figure 11.40 illustrates two points: the closer two receptors are, the greater the inhibitory effect; and the greater the separation between receptors, the less powerful the inhibitory effect [23–27]. The actual formation of negative (dark) and positive (light) Mach bands is shown in Fig. 11.41. In this illustration, two sets

207

11 Normal variants and pseudofractures

Fig. 11.39 Schematic drawing of the neural pathways through the retina. The heavy lines represent the primary afferent neurons to the optic nerve. The small lines represent an interlacing network of fibers that carry inhibitory impulses to each of the neighboring primary neurons. Impulses passing along these fibers result in lateral inhibition. (Modified from Daffner [23].)

208

Fig. 11.40 Effect of lateral inhibition. See text for details. (Modified from Daffner [23].)

Fig. 11.41 Mach band formation. See text for details. (Modified from Daffner [23].)

of light receptors that are separated by a barrier are exposed to low-intensity (X) and high intensity (2X) light sources. Because of the barrier, no overlap occurs between receptors receiving low-intensity stimulation and those receiving high-intensity stimulation. Line A is the expected response curve that would be recorded if lateral inhibition did not occur. The response to the receptors receiving higher stimulation is higher. Because of lateral inhibition, however, the actual response curve is less than expected and is represented by line B. The difference in expected response and actual response is greater on the side receiving the greater stimulation. This is because the greater the amount of stimulation, the greater the degree of lateral inhibition. Conversely, the less intense the stimulation, the less pronounced the inhibition.

Adjacent to the midline, there is a dip in response for receptor L, which received low stimulation, and a peak in responses for receptor H, which received high stimulation. A mathematical model can designate the degree of inhibition for each receptor on the low side as i  +  i  =  2i; for receptors receiving high stimulation, the amount of inhibition is 2i  + 2i  = 4i. The border receptors (L and H) are also inhibited by fibers from their neighbors. Receptor L receives inhibitory impulses not only from its neighbors but also from receptor H. Therefore, the total inhibitory effect on receptor L is i  + 2i  = 3i. Because the recorded response from receptor L is less than that of its neighbors, a dip is recorded on the curve and this appears as a negative or dark Mach band. Conversely, on the high intensity side, receptor H is inhibited by its neighbors as well as by

11 Normal variants and pseudofractures

receptor L. The total inhibitory effect here is also 3i. Since this is less than the total inhibitory effect on the other receptors on that side, the recorded response is greater and appears as a spike or a positive or light Mach band [23–27].

Clinical applications Most radiologists are familiar with the contrast-enhancing effect Mach bands have in various areas of the body. As mentioned above, the formation of Mach bands is favored in any situation in which there is overlap of the images of structures of different radiographic density. Mach bands actually aid in the interpretation of radiographs in many instances by providing apparent enhanced borders of structures [2,3]. In the skeleton, however, Mach bands can cause considerable diagnostic difficulty. They are responsible for a large number of pseudofractures because bone is of the optimal density to allow Mach bands to be generated. The pseudofractures produced by Mach bands result from overlapping images of normal bone, osteophytes, or soft tissue structures on radiographs only and most

A

are encountered in the vertebral column [23–25,27,28]. They do not occur on CT or MR images. Fortunately, now that CT is the recommended screening method for vertebral trauma, pseudofractures produced by Mach bands have been relegated to a historical curiosity, for the most part. The best known skeletal Mach band occurs at the base of the dens, where a thin, lucent line may be misinterpreted as a fracture (Figs. 11.42 and 11.43). The overlapping structures that are most likely to produce this phenomenon are the posterior arch of the atlas, the occiput [23–25,27], skin folds in the neck, and, occasionally, air over the back of the tongue. The lateral radiograph of the same region is most useful for resolving the problem. When a Mach band is present on the frontal view, there is no evidence of fracture on the lateral view, Harris’ ring is intact [29], the body of C2 is not “fat” [30], and there are no abnormal soft tissue changes in the predental region (Fig. 11.43). In many instances, the Mach band can be traced to the image of the arch of the atlas as it extends laterally beyond the dens (Fig. 11.43). Prior to the era of CT screening for cervical

B

Fig. 11.42 Mach band at the base of the dens. (A) Open-mouth radiograph shows a lucent line at the base of the dens (arrow). (B) Lateral radiograph shows no fracture of the dens and no disruption of the spinolaminar line (arrows) to indicate there is no motion. Dens fractures typically have anterior or posterior displacement.

A

B

Fig. 11.43 Mach band at the base of the dens. (A) Open-mouth radiograph shows a lucent line at the base of the dens (large arrows). Note that this line continues with the image of the arch of the atlas laterally (small arrow). (B) Frontal radiograph of articulated atlas and axis shows the relationship of the posterior arch of the atlas (arrows) to the dens (D). Lm, lateral mass.

209

11 Normal variants and pseudofractures

A

B

Fig. 11.44 Dens fracture. (A) Open-mouth radiograph shows a lucent line at the base of the dens (arrow). Note the similarity in appearance with Fig. 11.42A. (B) Lateral radiograph shows posterior displacement of the dens (D) and C1 on C2, manifest by disruption of the spinolaminar lines (arrows).

A

B

Fig. 11.45 Dens fracture. (A) Open-mouth radiograph shows a vague lucency at the base of the dens (arrow). There is displacement of C1 on C2 on the right (arrowhead). (B) Lateral radiograph shows anterior displacement of the dens (D) and C1 on C2, manifest by disruption of the spinolaminar lines (arrows).

trauma, tomography was often necessary to rule out a fracture. Axial CT is not as useful for this purpose because the lucency in question often lies in the plane of the scan. However, as mentioned above, these Mach bands are not encountered when CT is the prime imaging modality. Most true fractures of the dens show anterior or posterior displacement on lateral radiographs (Figs. 11.44 and 11.45). In almost every instance of a dens fracture, Harris’ ring is disrupted and there is soft tissue swelling in the predental region. Overlapping images of the incisor teeth may produce another Mach band of the dens or body of C2 (Fig. 11.46). This abnormality is not difficult to diagnose when the overlapping images can be traced. The lateral radiograph is normal. On an oblique view of the cervical column, overlapping images of the posterior arches of the vertebrae may result in Mach band for-

210

mation at any level (Fig. 11.47). Again, CT has eliminated both of these problems. Lower in the cervical column, the images of the uncinate processes abutting the undersurface of the vertebral bodies may result in a spurious “lucency” within the vertebral disc on a lateral radiograph (Fig. 11.48). This lucency is probably the second most common Mach band encountered in the vertebral column and is constant in its location [31]. It should not be confused with the less common disc bond injury (Fig. 11.49), which results in a lucency adjacent to the lower anterior aspect of the intervertebral disc space. Prominent transverse processes whose images overlie the vertebral bodies may produce Mach bands on lateral radiographs of the cervical column (Fig. 11.50). Occasionally, the

11 Normal variants and pseudofractures

A

B

Fig. 11.46 Mach bands producing lucencies in C2 (arrows) as the result of overlapping incisor teeth.

A

B

Fig. 11.47 Mach band of the posterior arch of C2. (A) Supine oblique radiograph shows a lucency across the posterior arch of C2 on the left (arrow). This results from the overlapping images of the arches of the axis in this position. (B) Lateral radiograph shows no evidence of fracture. Using CT as the primary screening imaging modality eliminates this problem.

overlap of normal structures from articular pillars produces Mach bands running across a vertebral body (Fig. 11.51). Similarly, spurs from uncinate processes (Fig. 11.52) or large osteophytes projecting off the anterior or lateral margins of the vertebral bodies (Fig. 11.53) may produce pseudofractures running horizontally or obliquely across the middle to posterior portions of vertebral bodies [25,32]. These findings are extremely common, particularly in elderly people. I have never seen a horizontally oriented cervical fracture in this location. Use of CT eliminates this problem also. The overlap of air within the recesses of the larynx may be misinterpreted as a fracture of the cervical vertebra. This

is not a Mach band. A curious pseudofracture produced by overlapping images of the sternum on the thoracic vertebrae (Fig.  11.54) is a true Mach band. By carefully following the contour of the lucency, it is possible to determine that it continues with the sternum. As always in difficult cases, CT may be necessary for resolution.

Subjective contours Subjective contours are a curious psychophysiologic phenomenon in which a geometric “figure” is mentally constructed by integrating information from partial lines [24,25,33]. Under

211

11 Normal variants and pseudofractures

Fig. 11.48 Mach band produced by the proximity of the uncinate process and the vertebral body above. Horizontal lucencies appear beneath the bodies of C3 and C4 (arrows). There is no predental soft tissue swelling to indicate injury. This finding is fairly common on lateral radiographs and is a background density effect, a Mach band variant.

212

A

B

A

B

normal circumstances, contours are perceived because of sudden changes in the brightness or color of adjacent areas that stimulate the retina. The phenomenon works for curvilinear structures as well as for straight lines. The eye takes in these differences and the “mind’s eye” organizes the information into recognizable “structures.” The term subjective contours is used because the outlines are perceived but are not real. They exist as a real presence only in the observer’s visual experience. Subjective contour images have certain characteristics in common. Firstly, the area contained by a subjective contour appears to be more prominent or brighter than the background (Fig. 11.55) [34]. This most likely is the result of lateral inhibition similar to that occurring in Mach band formation. Secondly, the region subtended by the contours appears to be superimposed over adjacent structures (Fig. 11.56). Thirdly, subjective contours may be generated in the absence of straight lines, because geometric regularity is not a prerequisite for their formation. Finally, close examination of subjective con-

Fig. 11.49 Disc bond injuries. Linear lucencies are present just above the anterior superior aspects of the involved vertebrae (arrow). This is a true hyperextension injury. Compare these findings with Fig. 11.48.

Fig. 11.50 Mach band produced by the transverse process image superimposed over the vertebral body. (A) Lateral radiograph shows a curvilinear lucency extending across the posterior vertebral body (arrow). (B) Photograph of articulated cervical vertebrae shows the position of the transverse process (arrows) in relation to the body of the superior vertebra (arrow pair). (With permission from Daffner et al. [30].)

11 Normal variants and pseudofractures

A

B

Fig. 11.51 Mach bands produced by overlapping articular pillar images (arrows). (A) Cervical. (B) Lumbar. In each instance, the lucency is contiguous with the adjacent facet.

A

B

A

B

Fig. 11.52 Mach bands of vertebral body produced by uncinate spurs. Two different patients with degenerative changes show curvilinear horizontal lucencies through the bodies of C5 (arrow). This is a common finding in older individuals with cervical spondylosis and occurs most often at C5–C7. Cervical fractures do not occur in this plane.

Fig. 11.53 Mach band produced by osteophytes in the thoracic region. (A) Frontal radiograph shows a curvilinear lucency (arrows) along the inferior body of T9. (B) Lateral radiograph shows large anterior osteophytes with a pseudoarthrosis (arrow).

213

11 Normal variants and pseudofractures

A

C

B

Fig. 11.54 Pseudofractures produced by overlap of portions of the sternum. (A,B) Detail from chest radiographs shows horizontal lucencies (arrows) across thoracic vertebral bodies, caused by overlap of the image of the sternal angle of Louis. (C) Curvilinear lucency across the left side of the body of T3 (solid arrows) results from the overlap of the manubrium. The margin of the right side of the manubrium (open arrow) does not overlap the vertebra.

A

B

Fig. 11.55 Subjective contours. A square appears to be superimposed on another square as well as on the circles. The effect occurs even when the images are reversed. Furthermore, in A, the center white square appears to be brighter than the background. In B, the center black square appears to be darker than the background. There are no squares or circles. (With permission from Daffner et al. [34].)

tours makes them disappear because they have no physical basis. Magnifying the area in question also results in the disappearance of the contour. If the entire area is viewed without magnification, the contour reappears [24,25,33]. From a clinical standpoint, subjective contour formation is most often encountered on chest radiographs, where it produces “cavities” or “nodules.” The one area of the vertebral column in which subjective contour formation is most likely to occur is

214

on oblique radiographs of the lumbar region. On such radiographs, the proximity of a sclerotic pars interarticularis and the normally sclerotic line of the adjacent pedicle result in the false perception of a lucency across the pars (Fig. 11.57). In this region, subjective contour formation is enhanced by the presence of background contrast effect, another psychophysiologic phenomenon related to Mach band formation in which differences in background density affect the perception of adjacent

11 Normal variants and pseudofractures

Fig. 11.56 Subjective contours. The shaded area of each circle gives rise to the image of a transparent rectangle superimposed on those circles. There is no rectangle. (With permission from Daffner et al. [34].)

A

Fig. 11.57 Subjective contour showing a false pars defect of L5. Oblique radiograph shows a lucency across the pars (arrows) produced by the proximity of a sclerotic pars and the margin of the pedicle.

B

Fig. 11.58 Subjective contour pseudofracture produced by overlying bowel gas. (A) Lateral radiograph suggests fractures (arrows) through the body of L3. On closer inspection, the lucencies are contiguous with overlying bowel gas. (B) Repeat lateral radiograph made minutes after A shows the “fracture” lines to have disappeared.

structures [31]. Overlying bowel gas is often a source of subjective contour formation (Fig. 11.58). In these instances a repeat radiograph or CT will solve the dilemma. Finally, carotid calcification overlying a cervical vertebra produces a pseudofracture as the result of subjective contour formation (Fig. 11.59). Here, CT will be required to clear the areas in question.

Extraneous materials Extraneous materials lying in or on a patient may produce pseudofractures on radiographs, a problem not found on CT

studies. A cervical restraining collar is the object most likely to produce difficulties on vertebral radiographs (Fig. 11.60) [25,35]. Such radiographs should be carefully examined to determine whether there are other clues to indicate the cause of the lucency. If the patient is wearing a cervical restraining collar, the easiest solution is to carefully remove the collar, stabilize the neck in some other way, and repeat the radiograph. A similar procedure should be followed if it is necessary to remove extraneous clothing or other foreign material (Fig. 11.61).

215

11 Normal variants and pseudofractures

A

B

Fig. 11.59 Subjective contour pseudofracture (arrow) resulting from overlying vascular calcification of C2.

Fig. 11.60 Pseudofracture appearance produced by a plastic rivet on a cervical restraining collar. (A) Lateral radiograph shows an oblique lucency (large arrow) suggesting a teardrop fracture of C2. On careful inspection, there is a curvilinear density (small arrows) containing this lucency. (B) The lucency disappears after the collar is removed. (With permission from Daffner and Khoury [35].)

Old fractures

Fig. 11.61 Pseudofracture (arrow) of L3 produced by overlying intravenous tubing.

216

Healed vertebral fractures produce residual deformities that may be misinterpreted as being the result of acute fracture. One may encounter two situations in which old fractures cause diagnostic difficulties. The first is the presence of an old ununited fracture. As a rule, this presents little diagnostic difficulty because a sclerotic line along the ununited fracture is present immediately adjacent to the lucency. As with any

11 Normal variants and pseudofractures

A

B

C

Fig. 11.62 Subtle motion artifact producing pseudofracture. (A) Sagittal reconstructed CT image shows a horizontal lucency across the base of the dens (arrow). (B) Axial CT image shows indistinct margins because of motion. (C) Lateral radiograph shows no evidence of fracture. Motion artifacts are usually more apparent on axial CT images than in this patient.

A

B

C

Fig. 11.63 Motion artifact suggesting a displaced fracture of the dens. (A) Sagittal reconstructed CT image shows apparent disruption of the dens at its base (arrow) with posterior displacement. (B) Axial CT image shows gross motion. (C) Sagittal T2-weighted MR image shows no evidence of fracture.

217

11 Normal variants and pseudofractures

imaging study, an old examination should be obtained for comparison if available. One should be aware, however, that a new fracture may occur through the site of a previous injury. In this situation, the normally smooth sclerotic border is lost and a jigsaw puzzle effect is present.

A

The second situation in which old injury is encountered has been discussed previously in regard to Scheuermann disease, Schmorl nodes, and limbus lesions. All of these abnormalities are the result of varying degrees of discovertebral trauma. As with other old injuries, however, each of these abnormalities shares similar radiographic or CT findings of sclerotic lines and reactive new bone formation. It may be necessary to obtain old studies for comparison or, if these are not available, to perform CT if it has not been done. In the latter case, MRI may be necessary to show either the fracture and/or the soft tissue changes an acute injury produces.

Motion artifacts The widespread use of CT for screening patients with suspected vertebral fractures has engendered a number of technical abnormalities that may give the appearance of pseudofractures. Two of these are partial volume averaging effect and motion artifacts. Partial volume averaging effect was previously discussed in relation to scoliosis (Fig. 11.38). Motion artifacts are more likely to pose a diagnostic problem, particularly if the degree of motion is not severe. There is a tendency among radiologists on viewing vertebral CT images, particularly from the high-end scanners (64 slice), to rely on the sagittal and coronal reconstructed images. Motion artifacts may appear as fractures (Figs. 11.62 and 11.63) or like dislocations (Fig. 11.64). There are several clues that the abnormalities are caused by motion Fig. 11.64 Motion artifact suggesting dislocation. (A) Sagittal reconstructed CT image shows apparent posterior displacement of C4 and widening of the C3 disc space. (B) Axial CT image shows gross motion. (C) Lateral radiograph shows no abnormalities.

B

218

C

11 Normal variants and pseudofractures

artifacts and not injury. Firstly, the axial images will show gross evidence of motion (Figs. 11.63 and 11.64). Secondly, the sagittal and coronal images will show a linear shift of data along the suspected fracture or dislocation line. Careful attention to the scout views may also be used to determine if a motion artifact is present. If these methods cannot resolve the situation, I recommend that radiographs be obtained.

Table 11.11 (cont.) Findings

Pseudofracture

Mach bands  Normal structures

Arch of atlas on dens Posterior vertebral arches Transverse processes Uncinate processes

Table 11.11 Common vertebral pseudofractures

Articular pillars

Findings

Teeth on vertebrae

Pseudofracture

Sternum on thoracic spine

Variants and anomalies  True variants

 Osteophytes Failure of fusion

Uncinate spurs  Disc margin spurs

Cleft sulcus of vertebral artery of atlas Spurlike deformities of lateral masses of atlas

 Soft tissues

Skin folds in the neck

 Gas

Air over the back of the tongue

Subjective contours

Accessory ossification centers Synchondrosis at the base of dens Ununited ring apophysis

Pseudo pars interarticularis defect Vascular calcification Overlying bowel gas

Extraneous materials

Immobilization collars

Os odontoideum

Clothing

Nuchal ligament ossification

Life-support equipment

Normal physes

Old fractures

Vascular grooves

Partial volume averaging

 Possible result of old or repetitive trauma

Motion artifacts Os odontoideum Accessory ossicles Intercalary bones Scheuermann disease “Limbus” vertebrae Schmorl node variants

References 1.

2.

3.

4.

Kattan KR. “Trauma” and “No-trauma” of the Cervical Spine. Springfield, IL: Charles C Thomas, 1975. Keats TE, Anderson MW. Atlas of Normal Roentgen Variants that may Simulate Disease, 8th edn. Philadelphia, PA: Mosby, 2007, pp. 155–372. Freyschmidt J, Wiens J, Brossmann J, Sternberg A. Freyschmidt’s “Kohler/ Zimmer.” Borderlands of Normal and Early Pathological Findings in Skeletal Radiography, 5th edn. New York: Thieme, 2003, pp. 597–748. Cattell HS, Filtzer DL. Pseudosubluxation and other normal variations in the cervical spine in children: a study of one hundred and sixty children. J Bone Joint Surg 1965; 47A:1295–1309.

5.

6.

7.

8.

9.

Swischuk LE. Anterior displacement of C2 in children: physiologic or pathologic? A helpful differentiating line. Radiology 1977;122:759–763. Swischuk LE. Emergency Imaging of the Acutely Ill or Injured Child, 3rd edn. Baltimore, MD: Williams & Wilkins, 1994, pp. 711–718. Kattan KR., Pais MJ. Some borderlands of the cervical spine. Skeletal Radiol 1982;8:1–6. Swischuk LE, Hayden CK Jr., Sarwar M. The dens-arch synchondrosis versus the hangman’s fracture. Pediatr Radiol 1979; 8:100–102. Gehweiler JA Jr., Daffner RH, Roberts L Jr. Malformations of the atlas vertebra simulating the Jefferson fracture. AJR Am J Roentgenol 1983;140:1083–1086.

10. Harrison RB, Keats TE, Winn HR, et al. Pseudosubluxation of the axis in young adults. J Can Assoc Radiol 1980; 31:176–177. 11. Jacobson G, Bleecker HH. Pseudosubluxation of the axis in children. AJR Am J Roentgenol 1959;82:472–481. 12. Suss RA, Zimmerman RD, Leeds NE. Pseudospread of the atlas: false sign of Jefferson fracture in young children. AJNR Am J Neuroradiol 1983;4:183–186. 13. Kattan KR. Backward “displacement” of the spinolaminal line at C-2: a normal variation. AJR Am J Roentgenol 1977; 129:289–290. 14. Kattan KR. Two features of the atlas vertebra simulating fractures by tomography. AJR Am J Roentgenol 1979; 132:963–965.

219

11 Normal variants and pseudofractures

15. Smith JT, Skinner SR, Shonnard NH. Persistent synchondrosis of the second cervical vertebra simulating a hangman’s fracture in a child. J Bone Joint Surg 1993;75A:1228–1230. 16. Dagirmanjian A, Daffner RH. Radiographic assessment of the atlantoaxial articulation in patients with os odontoideum. Radiologist 1994; 1:165–170. 17. Fielding JW, Griffin PP. Os odontoideum: an acquired lesion. J Bone Joint Surg 1974;56A:187–190. 18. Fielding JW, Hensinger RN, Hawkins RJ. Os odontoideum. J Bone Joint Surg 1980; 62A:376–383. 19. Lawson JP. Clinically significant radiologic anatomic variants of the skeleton. AJR Am J Roentgenol 1994; 163:249–255. 20. Monu J, Bohrer SP, Howard G. Some upper cervical spine norms. Spine 1987; 12:515–519. 21. Smith GR, Abel MS. Anatomical variations on the articular masses of

220

22.

23.

24.

25. 26. 27.

28.

the seventh cervical vertebra simulating fracture. Clin Radiol 1977;28:181–186. Martel W, Seeger JF, Wicks JD, Washburn RL. Traumatic lesions of the discovertebral junction in the lumbar spine. AJR Am J Roentgenol 1976; 127:457–464. Daffner RH. Visual illusions affecting perception of the roentgen image. CRC Crit Rev Diagn Imaging 1983;20:79–119. Daffner RH. Visual illusions in interpretation of the radiographic image. Curr Probl Diagn Radiol 1989; 18:62–87. Daffner RH. Skeletal pseudofractures. Emerg Radiol 1995;2:96–104. Ratliff F. Contour and contrast. Sci Am 1972;226:90–110. Daffner RH. Pseudofracture of the dens: Mach bands. AJR AmJ Roentgenol 1977; 128:607–612. Lane EJ, Proto AV, Phillips TW. Mach bands and density perception. Radiology 1976;121:9–17.

29. Daffner RH, Gehweiler JA Jr. Pseudovacuum of the cervical intervertebral disc: a normal variant. AJR Am J Roentgenol 1981;137:737–739. 30. Daffner RH, Deeb ZL, Rothfus WE. Pseudofractures of the cervical vertebral body. Skeletal Radiol 1986;15:295–298. 31. Harris JH Jr., Burke JT, Ray RD, et al. Low (type III) odontoid fracture: a new radiographic sign. Radiology 1984;153: 353–356. 32. Smoker WRK, Dolan KD. The “fat” C2: a sign of fracture. AJNR Am J Neuroradiol 1987;8:33–38. 33. Kanizsa G. Subjective contours. Sci Am 1976;234:48–52. 34. Daffner RH, Gehweiler JA Jr., Rodan BA. Subjective contours and illusory roentgenographic images. Appl Radiol 1984;13:95. 35. Daffner RH, Khoury MB. Pseudofractures due to Nec–Loc cervical immobilization collar. Skeletal Radiol 1987;16:460–462.

Index

ABCS (“footprints” of vertebral injuries), 126–159 alignment and anatomy abnormalities, 126–138 bony integrity abnormalities, 139 cartilage (joint) space abnormalities, 146 soft tissue abnormalities, 153–159 accessory ligaments, 26 ACR Appropriateness Criteria, 46–47, 53 aging, Wolff ’s law, 41 alar ligaments, 63 alignment and anatomy abnormalities, 126–138 box, 138 amphiarthrodial joints, 28 angiography, CT, 61 ankylosing spondylitis, 82, 89 annulus fibrosus, 29 failure, 43 anterior vertebral body line, pediatrics, 126 anterior/middle/posterior vertebral columns, 181 anterior/posterior longitudinal ligaments, 81 intercalary bones, 200 ossification in DISH, 81 anterior/posterior vertebral columns, 181 anterolisthesis, 103, 132–136 and dens fracture, 144 aortic injury, 158 apophyseal (facet) joints, 28, 31, 38 articular pillar, fractures, 97, 136 articular pillars, 126, 201 Mach bands, 211 atlanto-axial articulation pediatrics, 171, 171–173 atlanto-axial offset, pediatrics, 193 atlanto-axial overlap, 131 atlanto-axial ratio, 129 atlanto-axial rotation with occipito-atlantal rotary subluxation, 118 pediatrics, 171–173

atlanto-dental interval, pediatrics, 169 atlanto-occipital junction see occipito-atlantal junction atlas, 16 anomalies, 192–194 failure of fusion of posterior arch, 126, 192 pediatric variants, 192–194 spur-like deformities, 194 development, 12 pediatric fractures, 173–174 posterior tubercle, 126 relationship of occiput to dens, 126 superior articular facets, 16 synchondroses, 12 axis, 41 articular processes, 43 basion–axial interval, 129, 170 development, 12 Harris’ ring, 139 pedicles, 16 synchondroses, 12 Baastrup disease, 31 basion–axial interval, 129 pediatrics, 170 basion–dens interval, 129 pediatrics, 170 basivertebral vein, 139 bending force, 36 biomechanical features of pediatric cervical injuries, 165 biomechanical principles of vertebral motion, 36–44 definitions, 36 injuries, load spectrum, 43 pattern of motion, 36 bony integrity abnormalities, 139–145 box, 144 bowel gas, 215 burst fractures, 97, 139–139 crush-cleavage fracture, 93 interpedicle distance, 94, 138, 139, 148–148 Jefferson-type, 131, 192 MRI, 79, 89

pediatrics, 175–176 posterior vertebral body line, 94, 138, 139–139 “stability,” 190 types, 92–94 vertebral endplate angle, 139 vs rotational grinding injuries, 113 vs shearing injuries, 120 burst–rotary injury, 93 Canadian C-Spine Rule, 46 carotid arteries calcification, 215 migration, 158 cartilage (joint) space abnormalities, 146–152 box, 151 central cord syndrome, 107 cervical anatomy, 12–13 anomalies, 20 osteophytes, 30 anterior/posterior longitudinal ligament, 81 anterior/posterior tubercles, 126 apophyseal (facet) joints, 28 articular pillars, 16 atypical, 16 costal processes, 16 interpedicle distance, 130 pedicles, 15 spinolaminar line, 57 spinous processes, 16 transverse processes, 16 uncinate processes, 15 spurious lucency, 210 vertebral foramen, 16 cervical column variants, 192–201 categories, 192 cervical development, 12–13 cervical imaging CT, 59–66 lateral views, table, 54 radiography, 53–57 vs CT, 47–48 risk-based indications, 46–47 table of criteria, 47 see also imaging techniques

cervical injuries avulsion of ring apophysis, 147 classifications, 88 clay shoveler fracture of spinous process, 92 “fat” C2 sign, 144 loading force, 1 major and minor classification, 49–50 “military” posture, 126 neck sprains, 56 pediatrics, 165 subaxial, 174 wedging of vertebral bodies, 168 transverse processes, 131 whiplash, 55–56 cervical lordosis, 57 reversal of normal, 57 spinolaminar line, 57 cervical rotation, 38 cervicothoracic junction CT, 63–66 radiography, 53 Chance-type fractures, 90 pediatrics, 176 check ligaments, 31 clay shoveler fracture, 92 clivus baseline, pediatrics, 169 coccyx, 26 cock-robin posture, 171 collar, 215 comatose patient, clearing the spine, 50, 165, 167–168, 189 compression fractures, 97 MRI, 73 osteoporotic, 38 computed tomography (CT), 59–66 alignment and anatomy abnormalities, 126 canal encroachment, 59 craniovertebral and cervicothoracic junctions, 63–66 CT angiography, 61 efficacy, costs and radiation dose, 48–50 fractures in plane of scan, 64

221

Index

computer tomography (CT) (cont.) intervertebral foramen encroachment, 59 level calibration, 64 motion artifacts, 64, 216–218 multidetector CT, 59 with myelography, 61 partial volume averaging effect, 64 picture archiving and computer storage, 59 radiation exposure, 48–50, 49–49 pediatrics, 165 scout film annotation, 64 scout views, 59 vs radiography, 47–48 weight of patient, 64 congenital fusions and block vertebrae, 201 conjoint vertebrae, 20 coupling, 36 craniovertebral junction, 126, 151–151 basion–axial interval, 129 basion–dens interval, 129 CT, 63–66 hematoma, 153 occult fractures, 63 pediatrics, 168–170 see also occipito-atlantal joint cruciform ligament, 31 crush-cleavage fracture, 93 CT see computed tomography cystic myelopathy, 66 MRI, 77 degenerative disc disease, 199 degrees of freedom, 36 dens, 13 absence of fusion to body of C2 and anterior arch of atlas, 126 atlanto–dental interval, pediatrics, 169 basion–dens interval, 129, 170 and lateral mass of atlas, asymmetry, 33 Mach bands, 194, 209 ossification center, 194 predens space angle, 200 predental space, 126, 200 pediatrics, 169 synchondroses, 194 tilt angle, 144, 200 dens fracture, 107, 139–144 classification, 139–144 high/low, 144 MRI, 74 pediatric fractures, 174, 194 diarthrodial joints, 28 diffuse idiopathic skeletal hyperostosis (DISH), 81, 82–83, 89, 108–108

222

disc spaces see intervertebral disc spaces DISH see diffuse idiopathic skeletal hyperostosis dislocation (luxation), 4, 96 extension–dislocation, 112 distraction force, 36 distraction injuries, 94–96 Down syndrome, 170 dynamic flexion–extension fluoroscopy, 50 dynamics, 36 dysraphism, 20 elastic zone, 36 Wolff ’s Law, 41 extension injuries, 103–113 definition, 5 dislocation, 112 “fingerprints,” 124 teardrop fracture, 111 extension radiography see flexion–extension radiography extensor muscles, 43 extraneous materials, 215 facet joints, 28 duplicated (bowtie), 135 flexion injuries, 149–151 fractures, 82 jumped/locked, 103, 135 unilateral (box), 135 naked facets, 94 and stability, 82 widening, 148, 181, 186–188 failure zone, Wolff ’s law, 41 flexion injuries, 89–103 classification, 92 definition, 5 facet joints, 149–151 “fingerprints,” 92 flexion interspinous atlanto-axial distance, 129 flexion–extension radiography, 47, 55, 189 contraindications, 189 flexor muscles, 38 fractures, 1–2 definition, 1 healed/old, 216 see also specific regions and types fragmentation, 97 functional spinal unit, 36 compression, 43 range of allowable motion, 40 hamburger bun sign, 135 reversed, 135 hanged-man fracture, 106, 174, 197 Harris method, relationship of occiput to atlas, 129 Harris’ ring, 139 head traction, 50

hematoma craniovertebral junction, 153 MRI, 74 hemorrhage, MRI, 73 hyperextension dislocation, 111 hyperextension sprain, 107 hyperflexion sprain, 94, 136 iliac crests, determining lumbar levels, 26 image perception extraneous materials, 215 healed/old fractures, 216 parallax phenomenon, 205 pseudofractures, 192–218, 204–206 subjective contours, 211–215 imaging techniques children, 56–57 clearing the comatose patient, 50, 167–168, 189 cost containment, 45–46 efficacy and costs of CT, 48–50 image perception, 205 indications, 45–51 Ottawa Rules, 45 protocols, 45–46 recommendations, 51 risk-based indications, 46–47 ACR Appropriateness Criteria, 46–47, 53 Canadian C-Spine Rule, 46 NEXUS, 46, 168 table of criteria, 47 Vandemark, 46 see also specific types imbrication, 126 incisor teeth, Mach bands, 210 inion of occipital bone (internal occipital protuberance), 130 injuries see vertebral injuries instantaneous axis (center) of rotation, 41–43 intercalary bones, 200 interlaminar space, 126 see also interspinous space interpedicle distance, 130 burst fractures, 94, 138, 139, 148–148 rotary fractures, 138, 139 variation, 148 widening of vertebral canal, 181, 188–188 interspinous ligament, 31 interspinous space, 129, 130 ratio between C1 and C2, 126 widening, 103, 139, 181, 183–186 intervertebral disc spaces narrowing, 103 widening, 106, 146, 183 intervertebral discs, 30 bond injury, 147, 210 degeneration, 199

herniation, 77, 203 vertebral endplate, 38 Jefferson-type burst fractures, 131, 192 jigsaw puzzle sign, 79 pseudofractures, 192 joint spaces narrowing, 146 widening, 146 joints and ligaments, 28–29 types amphiarthrodial, 28 diarthrodial, 28 Kimmerle anomaly, 12 kinematics, 36 kinetics, 36 Klippel–Fiel anomaly, 20 kyphosis, thoracic, 21 kyphotic angulation, 136 laryngeal deviation, 153 lateral inhibition, Mach bands, 207–210, 211 Lee method, 32–33 relationship of occiput to atlas, 128–129 ligamenta flava, 81 ligaments, 28, 38–40 alar, 63 rupture, 1 stretching, 38 ligamentum nuchae ossification, 200 limbus deformity of lumbar spine, 192 limbus fragments, 216 lumbar column, 202–203 lumbar imaging anomalies, determining lumbar levels, 26 CT, 47, 59 indications, 47 osteoarthrosis (Baastrup disease), 31 radiography, 57–59 spondylolysis, 25 lumbar lordosis loss of, 138, 139 lumbar vertebral anatomy, 26 anomalies, 25 congenital absence of pedicle/lamina, articular process, 33 limbus deformity, 192 lumbarization of first sacral segment, 25 presence of first rib, 25 sacralization of L5, 25 articular processes, 126 columns, 202–203 mammillary processes, 25 nutrient foramen, 126 pars interarticularis and styloid processes, 25

Index

lumbar vertebral injuries classifications, 88 pediatrics, 175–176 lumberjack paraplegia, 119 Luschka joints, 29 Mach bands, 140, 206–211 lateral inhibition, 207–210, 211 transverse processes, 210 major injuring vectors, 36 mammillary processes, 25 metallic objects, 72 titanium instruments, 74 methemoglobin, T1-shortening effect, 74 motion artifacts, 216–218 CT, 64 motor vehicle crashes, 9, 47, 89, 89–90 pediatrics, 165 MRI, 72–83 imaging parameters, 72–73 three-dimesional time-offlight examination, 73 axial gradient/spin echo, 73 sagittal (turbo) spin echo, 73 STIR, 73 indications, 50 pathologic aspects, 73–83 acute spinal cord injury, 73–76 bony and ligamentous injury, 77–83 chronic spinal cord injury, 77 dens fracture, 74 extradural cord or nerve root compression, 77 intramedullary hemorrhage, 73 patient backboard removal, 72 positioning, 72 safety, 72 recommendations, 72 superior sagittal sinus, presaturation band, 73 T1-shortening effect of methemoglobin formation, 74 muscles, flexors, extensors, and rotators, 38 myelography, 66–70 with CT, 61 indications, 72 myelomalacia, MRI, 77 myelopathy, 66, 77 cystic myelopathy, 66 MRI, 77 neck sprains, 56 neural arch, ossification centers, 12 neural arch fractures, 79

neurocentral synchondroses, 12 neutral zone, Wolff ’s law, 36 NEXUS (National Emergency X-radiography Utilization Study), 46 pediatrics, 168 normal variants and pseudofractures, 192–218 notochordal remnants, 29 nuchal ligament (ligamentum nuchae), 16 ossification, 200 nucleus pulposus, 200 occipital bone, inion, 130 occipital condyles fractures, CT, 63 views, 131 occipito-atlantal junction, 32 dislocation, 83, 128, 170–171 pediatrics, 170–171 counter-rotation, 173 see also craniovertebral junction occipito-atlantal membrane, 32 occipito-atlantal rotary fixation, 116–118 subluxation, with atlanto-axial rotation, 118 occipito-atlanto-axial fusions, 20 occult fractures cervica, l49 craniovertebral junction, 63 odontoid process, 16 pediatrics, 168 opisthion, Powers ratio and Lee method, 128, 129–129 os odontoideum, 198 ossification anterior/posterior longitudinal ligaments, 81 ligamentum nuchae, 200 ossification centers, 12 dens, 194 neural arch, 12 osteoarthrosis (Baastrup disease), 31 osteophytes, 107, 211 annulus vacuums, 147 cervical vertebrae, 30 osteoporotic compression, 36 parallax phenomenon, 205 paraspinal soft tissue changes, 155 partial volume averaging effect, 216 scoliosis, 216 pediatrics, 165–176 cervical injuries, 165 biomechanical features, 165 neurological deficits, 166 comatose patient, clearing the spine, 50, 165, 167–168 dens, predental space, 126, 200

imaging techniques, 56–57, 168 radiography, 48 normal variants and pseudofractures, 192–218 pseudosubluxation, 126, 168–168, 199 radiation exposure, 165 SCIWORA, 166 soft tissue swelling, 155 thyroid gland, 172 picture archiving and computer storage (PACS), 59 pillar fractures see articular pillar fractures plastic zone, 36 Wolff ’s law, 41 posterior axial line, 129 posterior vertebral body line burst fractures, 94, 138, 139–139 marker of integrity of vertebral canal, 103, 183, 188–188 pediatrics, 126 Powers ratio pediatrics, 170 relationship of occiput to atlas, 128 presaturation band, superior sagittal sinus, 73 prevertebral fat stripe, 153 pseudofractures, 192–218, 204–206 healed/old fractures, 216 jigsaw puzzle sign, 192 Mach bands, 140, 206–211 motion artifacts, 216–218 table, 218 pseudosubluxation, pediatrics, 126, 168–168, 199 psoas stripe, loss, 155 radiation exposure pediatrics, 165 thyroid gland, 172 radiography, 53–59 ACR Appropriateness Criteria, 46–47, 53 alignment and anatomy abnormalities, 126 cervical vertebrae, 53–57 atlantoaxial region, 54 flexion–extension radiography, 47 lateral views, 54 minimum, 53–57 supine/trauma projection, 54–55 cervicothoracic junction, 53 lateral views, table, 54 lumbar vertebrae, 57–59 pediatrics, 48 pitfalls, 168 rule of 2s, 162 significant signs, 161–162

combination, 162 thoracic vertebrae, 57–59 vs CT, 47–48 retrolisthesis, 106, 136 and dens fracture, 144 retropharyngeal abscess, 155 retropharyngeal space, 153, 153 pediatrics, 168 retropharyngeal swelling, 155 retrotracheal space, 153 rheumatoid arthritis, 149 ribs absence of twelfth, 25 articulation, 27 rigid spine disease, 46 ring apophyses, 199 road accidents see motor vehicle crashes rotary atlanto-axial fixation injuries, 116–118 rotary fractures, interpedicle distance, 138, 139 rotation, 36 rotational injuries, 113–119, 139 definition, 5 “fingerprints,” 124 grinding injuries, 113 vs burst fractures, 113 rotator muscles, 38–40 rule of 2s, 162 box, 163 sacral arcuate lines, 26 sacroiliac joints, 33 sacrum, 25 fractures, 144 lumbarization of first sacral segment, 25 pelvic sacral foramina, 26 sagittal sinus, presaturation band, 73 Scheuermann disease, 176, 201–202 Schmorl nodes, 202 pediatrics, 176 SCIMRA, SCIWORA see spinal cord injury scoliosis, 204 partial volume averaging effect, 216 shearing force shearing injuries, 119–123 definition, 5 “fingerprints,” 123 vs burst fracture, 120 short tau inversion recovery (STIR), 73 significant signs combination, 162 radiography, 161–162 table, 162 soft tissue abnormalities, 153–159 box, 159 soft tissue changes, paraspinal, 155

223

Index

soft tissue swelling, pediatrics, 155 spina bifida, 20 spinal cord injury MRI, 73–76, 77–77 bone fragment, 77 cystic myelopathy vs myelomalacia, 77 extradural cord or nerve root compression, 77 with minimal radiographic abnormalities (SCIMRA), 76 without radiologic abnormality (SCIWORA), 76, 89, 166 pediatrics, 174–175 spine anterior/posterior columns see vertebral columns clearing, comatose patient, 50, 165, 167–168, 189 stability see vertebral stability/ instability spinolaminar line, 57, 126, 135–136, 194 pediatrics, 168, 194, 199–199 spinous processes, 16 clay shoveler fracture, 92 rotation, 201 spondylolysis, 108 spondylosis, degenerative, 82 sprains, 1 hyperextension sprain, 107 hyperflexion sprain, 94, 136 neck sprains, 56 stability/instability see vertebral columns STIR see short tau inversion recovery subjective contours, 211–215 thoracic nodules, 214 subluxations, 1 occipito-atlantal rotary subluxation, 118 pseudosubluxation, pediatrics, 126, 168–168, 199 supraspinous ligament, 31 swallowing, 159 synchondroses, 12

224

dens, 194 pediatrics, 194–198 syndesmophytes, 107 syringomyelia, 66 teardrop fractures, 93–94 fragmentation, 103 tectorial membrane, 31 teeth, Mach bands, 210 thoracic column, 201–202 thoracic imaging CT, 59 CT, 47 indications, 47 radiography, 57–59 thoracic kyphosis, 21 thoracic nodules, subjective contours, 214 thoracic ring apophyses, 199 thoracic vertebral anatomy, 21 anomalies, 25 absence of twelfth rib, 25 articular processes, 126 articulation of ribs, 21 atypical, 21 column, 201–202 nutrient foramen, 126 spinous processes, 21 typical, 21 uncinate processes, 23 thoracic vertebral injuries classifications, 88 pediatrics, 175–176 thoracolumbar junction, 38 thorax–abdomen–pelvis body scan, 59 thyroid gland, 172 titanium instrumentation, 74 torticollis, 138, 139–139 duration, 173 pediatrics, 171–173 tracheal deviation, 153 translation force, 36 transverse processes cervical, 16 cervical injuries, 131 Mach bands, 210 uncinate processes cervical, 15, 29 spurious lucency, 210

spurs, 211 thoracic, 23 uncovertebral joints, 29 variants see normal variants and pseudofractures vector, major, 20 vertebra prominens, 20 vertebral anatomy, 13 anterior/posterior columns, 181, 181–181 block vertebrae, 201 congenital fusions and block vertebrae, 201 conjoint vertebrae, 20 vascular channels, 21 see also specific regions vertebral annulus vacuum, 147 vertebral arch, 13 fractures, “fingerprints,” 113 vertebral body prevertebral fat stripe, 153 wedging, pediatrics, 168, 175–175 see also burst fractures vertebral canal, 13 marker of integrity, posterior vertebral body line, 103, 183 widening (interpedicle distance), 181, 188–188 vertebral columns anterior/posterior columns, 181, 181–181 lumbar column, 202–203 stability/instability three column concept, 181 two column concept, 181 thoracic column, 201–202 vertebral development, 12–13 congenital fusions and block vertebrae, 201 vertebral disc bond injury, 147, 210 vertebral displacement, 181, 183–183 vertebral edge separation, 29 vertebral endplate, 38 angle with body cortex, burst fractures, 139 fractures, 43

herniation, pediatrics, 176 vertebral injuries, 9, 88–123 biomechanical features of pediatric cervical injuries, 165 biomechanical principles of vertebral motion, 36, 44 combined imaging, 189 combined injuries, 97 “fingerprints,” 124 “footprints,” 126–159 levels, 89 load spectrum, 43 major vector, 43 mechanisms, 88–123 background, 89 predictable and reproducible patterns, 88–89, 89–89 types 1–4 dislocation (luxation), 4 extension injuries, 5 occult instability, 1 rotational injuries, 5 shearing injuries, 5 subluxations, 1, 118 see also specific regions and types vertebral motion biomechanical principles, 36 motion artifacts, 64, 216–218 range of allowable motion, 40 table, 40 vertebral processes, 13 vertebral stability/instability, 181–190 vertebral stability/instability see vertebral columns vertebral stability/instability, signs, summary table, 189 vertebral vasculature, 201 artery, cleft through sulcus, 193 vascular channels, 21 Wackenheim clivus baseline, pediatrics, 169 whiplash, 55–56, 83, 95 Wolff ’s law, 41 aging, 41 elastic zone, 36 neutral zone, 36