| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Emergency Radiology |
1 From the Department of Clinical Radiology (A.S., J.E., M.R.) and Clinic of Internal Medicine III (R.B.), University Hospital Ludwig-Maximilians-University Munich, Grosshadern, Marchioninistrasse 15, 81377 Munich, Germany; Institute of Legal Medicine (R.P.) and Institute of Anatomy (S.P.M.), Ludwig-Maximilians-University Munich, Germany; and Department of Radiology, Veterans Administration Medical Center, San Diego, Calif (D.R.). Received January 24, 2001; revision requested February 21; revision received May 7; accepted May 21. Address correspondence to A.S. (e-mail: axel.staebler@ikra.med.uni-muenchen.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The cervical spines of 10 adult accident victims underwent radiography and MR imaging, with T1-weighted, fast spin-echo T2-weighted, and four gradient-echo pulse sequences. The frozen specimens were cut into 3-mm-thick slices (sagittal plane) and photographed, and microfocus radiographs were obtained. Imaging findings were compared with the anatomic and pathologic findings.
RESULTS: Eight of the 10 specimens had 28 posttraumatic lesions: three fractures (two missed at the initial MR imaging reading), 10 facet joint capsule lesions with bleeding, five soft-tissue and ligament lesions, eight disk lesions, and two spinal cord lesions. Radiography depicted one lesion (4%). Two partial ruptures of the anterior annulus fibrosus were depicted at only MR imaging. Initially, 11 of 28 lesions were detected on MR images; retrospectively, 17 of 28 lesions were correlated with anatomic findings.
CONCLUSION: Soft-tissue and intervertebral disk and ligament injuries account for 89% (25 of 28) of posttraumatic cervical spine lesions detected on postmortem images. Occult lesions, including apophyseal joint injuries, were found in clinically noninjured cervical spines. MR imaging was limited in the depiction of discrete lesions when T1-weighted non–fat-saturated, fast spin-echo T2-weighted, and gradient-echo pulse sequences were used.
Index terms: Magnetic resonance (MR), comparative studies, 31.121411, 31.121412 • Spinal cord, injuries, 341.41, 341.42, 341.43 • Spine, injuries, 31.41, 34.42, 34.43 • Spine, radiography, 31.11 • Trauma, 31.40
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Imaging
Radiographs of each specimen were obtained in the anteroposterior, lateral, and oblique projections. Screen speed was 200 (Ultra Vision; Dupont, Bad Homburg, Germany), and exposure parameters were 10 mAs and 48 keV.
MR images were obtained with a 1.5-T MR imager (Magnetom Vision; Siemens Medical Systems, Erlangen, Germany) with use of a temporomandibular joint coil. Specimens were centered in the coil and immobilized with foam pads. The imaging protocol consisted of six sequences in the sagittal plane with a section thickness of 3 mm. The field of view for all sequences was 80 x 80 mm, which resulted in an in-plane resolution of 0.31 x 0.31 mm. MR imaging was performed with the following sequences: T1-weighted spin echo (repetition time msec/echo time msec of 500/20); T2-weighted fast spin echo (4,200/17, 119); T2-weighted gradient-recalled echo fast low-angle shot, or FLASH, two-dimensional (561/15, 30° flip angle); gradient-recalled echo fast low-angle shot two-dimensional (608/15, 90° flip angle); fast low-angle shot three-dimensional fat saturated (53/11, 60° flip angle, with frequency-selective fat presaturation); and double-echo steady-state three-dimensional (26.8/9, 40° flip angle). In a preliminary study for sequence optimization, the short inversion time inversion-recovery, or STIR, sequence was removed from the imaging protocol because of an insufficient signal-to-noise ratio.
After imaging, the specimens were deep frozen at -80°C. Then, each specimen was cut into 3-mm-thick slices along the sagittal plane, and all the slices were photographed. A contact radiograph of each slice was obtained with a microfocus tube and screenless film (Faxitron; Hewlett-Packard, Palo Alto, Calif).
Some of the lesions were selected for histopathologic examination. After being embedded in methacrylate, the undecalcified specimens were sectioned (3 µm thick), and Giemsa staining was performed. For evaluation, a low-power magnification (x10) was used.
Evaluation
Only completely intact vertebrae with undestroyed intervertebral disks on both sides and completely intact intervertebral segments were evaluated. Because the field of view was restricted to 80 mm, only three to four segments of the middle and lower cervical spine were available for evaluation. Thirty-one images of intact intervertebral segments were correlated with anatomic sections.
All radiographs, MR images, and microfocus radiographs were evaluated by two observers (A.S., J.E.), who were experienced in musculoskeletal imaging. During the same session, the consensus readings started with the radiographs, followed by the microfocus radiographs. Criteria for evaluation were cortical disruption for fractures, widening of the intervertebral disk space, and intervertebral joint or spinous process distance for diskoligamentous injuries. For evaluation of the anatomic sections, which was performed later, the readers were blinded to the pathologic findings.
The MR imaging evaluation was performed 4 weeks later by the same two observers in consensus. The anterior and posterior longitudinal ligaments were evaluated for continuity and increased signal intensity on T1-weighted, T2-weighted, and gradient-recalled echo MR images. The intervertebral disks were reviewed for increased signal intensity on T2-weighted MR images and signal intensity changes on gradient-recalled echo MR images, as indicators for bleeding. The bone structures were evaluated for cortical disruption as an indication of fracture and for signal intensity increase on T2-weighted MR images in the bone marrow as an indirect sign of bone lesions. The intervertebral joints and surrounding soft tissues on T2-weighted MR images were also examined for fluid accumulations.
For pathologic evaluation, the photographs of the slices were magnified and studied by an expert in forensic medicine (R.P.). He performed a careful review of all anatomic structures, with special focus on hemorrhage in the prevertebral muscles; rupture or bleeding of the anterior or posterior longitudinal ligaments; rupture or bleeding of the anterior or posterior annulus fibrosus or nucleus pulposus; bleeding into the uncovertebral and facet joints; and bleeding into the posterior muscles, ligamenta flava, ligamenta interspinosa, and ligamentum supraspinale. The spinal canal and spinal cord were also evaluated for bleeding.
After the pathologic evaluation, a slice-by-slice comparison with direct correlation of the anatomic findings in the slices with the findings on the MR images was performed by the three readers (A.S., J.E., R.P.).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Findings in our comparison study confirmed these results. Twenty-five (89%) of 28 lesions detected during the study were injuries to soft-tissue structures, including the longitudinal ligaments, the disks, and the apophyseal joint capsules. Only three of the 28 lesions were fractures; two were missed on the radiographs. These findings are consistent with results in the studies by Jónsson et al (3) and Saternus (8), who found that fractures contributed to 20.0%–22.7% of the injuries overall (3,8).
Hyperextension injuries of the cervical spine are common and are associated with a risk of spinal cord compromise of variable degree even in the presence of normal radiographs (10–13). Degenerated segments with fissuring of the disk expose the longitudinal ligament to increased loading forces and increased risk of hyperextension injury (10,14). The spectrum of cervical hyperextension injuries ranges from muscle sprains to serious lesions, such as diskoligamentous disruption (12). In hyperextension injuries, both the anterior and posterior longitudinal ligaments are stretched or disrupted, the anterior longitudinal ligament more severely, which causes hemorrhage beneath the prevertebral fascia and occasional avulsion fragments from the anterior margin of the vertebra (12). Diagnosis of hyperextension injuries may be difficult because they often show only subtle radiographic abnormalities, even in severe or unstable lesions (10,15,16).
In young persons, a special form of lesion in the intervertebral segment can occur, with characteristic separation of the intervertebral disk from the vertebral end plate (2,17,18). After a severe neck injury, complete horizontal rupture of a disk was not observed in the presence of an intact nucleus pulposus (8). The annulus fibrosus itself can remain intact in a hyperextension injury, and a partial or complete avulsion of the fibro-osseous junction, including the hyaline cartilage end plate, can occur at the ossified ring apophysis. In the cadaver of a 19-year-old accident victim, we documented the imaging appearance of an incomplete avulsion of the anterior annulus fibrosus, with water signal intensity in the defect (Fig 2). These two lesions were not identified at the initial evaluations of the anatomic sections owing to the lack of bleeding. A similar injury was documented by Davis et al (12) at MR imaging in a patient. Incomplete lesions of the annulus fibrosus may also occur in whiplash injuries.
Incomplete avulsion of the anterior annulus fibrosus is responsible for the vacuum phenomenon observed by Reymond et al (19), Resnick et al (20), and Bohrer (21). This vacuum cleft might be called an annulus cleft to differentiate it from a nucleus cleft, which occurs in the center of the disk (16,22). This annulus cleft vacuum phenomenon after hyperextension injury of the cervical spine is located in the anterior portion of the annulus fibrosus and not in the center of the nucleus pulposus. The high amount of water inside an intact nucleus pulposus makes the occurrence of a vacuum unlikely. In our study, an anterior vacuum phenomenon depicted on lateral MR images of the cervical spine in extension in a young patient without preexisting degenerative changes probably represents this type of partial rupture of the annulus (Fig 2). The anterior longitudinal ligament remained intact, and stability was preserved.
Hyperflexion injuries of the cervical spine can cause partial or complete disruption of the posterior cervical complex, which comprises the posterior articulations stabilized by the joint capsule, interspinous and supraspinous ligaments, and ligamenta flava (9). No clear border exists between maximal physiologic flexion and partial subluxation on lateral radiographs obtained in flexion and extension, which allows different interpretations. It is also difficult to decide whether the lesion is complete and unstable or only a partial tear without instability (9,23).
Because anterior subluxation can represent an unstable cervical spine injury, diagnosis of instability is crucial (24). Soft-tissue damage in anterior cervical subluxation may be severe, and late progressive displacement can occur with persisting instability and varying neurologic sequelae (25,26). MR imaging has proved to be more accurate than conventional radiography in the depiction of a wide spectrum of neck injuries, because it is capable of direct depiction of disruption of the joint capsule of the apophyseal joints, the ligamenta flava, and the inter- and supraspinous ligaments (4,27). In our series, the most frequently injured sites were the apophyseal joints, and the most frequently missed lesions were in the posterior elements.
The conditions of MR imaging during our study were superior to those associated with in vivo studies in several respects. First, motion artifacts were not a concern in the investigations of the specimens. Second, three acquisitions were performed with all sequences to gain a high signal-to-noise ratio for improved image quality, which resulted in an acquisition time of 6 minutes 3 seconds to 13 minutes 37 seconds. Such long acquisition times in a clinical setting may result in degradation of image quality due to motion. Finally, removal of major parts of the soft tissues of the neck enabled the use of a temporomandibular joint coil. In addition, a field of view of 80 x 80 mm, which results in an in-plane resolution of 0.31 mm, is not suitable for MR imaging investigation of the whole cervical spine, which would necessitate use of a field of view of 140 x 140 mm or greater. The signal-to-noise ratio with the head and neck surface coil is less than that with a temporomandibular joint coil.
The major limitation of our imaging protocol was the lack of a highly sensitive water sequence, such as the short inversion time inversion-recovery or T2-weighted fat-saturated sequences. As a result of the pulse arrangement, these sequences generally have a low signal-to-noise ratio. Because signal is dependent on temperature, the decreased temperature of the specimen (19°C) compared with the in vivo temperature (37°C) may account for the inferior signal-to-noise ratio.
In conclusion, this correlation study of postmortem MR imaging, conventional radiographic, microfocus radiographic, and pathologic findings of cervical spines in accident victims confirmed the high frequency of associated cervical spine injuries after severe head injuries. In our study, nearly 89% (25 of 28) of the posttraumatic lesions were soft-tissue lesions; only three of the 28 lesions were fractures, and only one was visible on the radiographs.
An incomplete hyperextension injury with incomplete rupture of the anterior annulus fibrosus was depicted at MR imaging. We presumed that this type of injury was responsible for a vacuum phenomenon (annulus cleft), which can be seen occasionally on lateral images of the cervical spine in extension obtained after acceleration injuries of the cervical spine.
Findings in our study showed the limitations of MR imaging for depicting posttraumatic lesions of the lower cervical spine with gradient-recalled echo, T1-weighted, and non–fat-saturated fast spin-echo T2-weighted sequences. Inversion-recovery sequences with a short inversion time or fat-saturated T2-weighted sequences might be more effective for depicting many of the soft-tissue injuries in patients. Our study involved direct comparison of posttraumatic cervical spine lesions with imaging findings. Various occult lesions, including many apophyseal joint injuries, which can also occur in survivors of vehicle accidents, were found on postmortem MR images obtained in clinically noninjured cervical spines.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Denaro, R. Papalia, L. Denaro, A. Di Martino, and N. Maffulli Cervical spinal disc replacement J Bone Joint Surg Br, June 1, 2009; 91-B(6): 713 - 719. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Harris Jr and A. Stabler Apophyseal Joint and Intervertebral Disk Injuries Caused by Hyperflexion and Hyperextension * Dr Stabler responds: Radiology, September 1, 2002; 224(3): 932 - 932. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
