| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Musculoskeletal Imaging |
1 From the Departments of Radiology (S.C.W., M.E.S., L.P.) and Neurosurgery (G.J.P.), Thomas Jefferson University Hospital, 111 S 11th St, Rm 3350G, Philadelphia, PA 19107, and the Wilford Hall Medical Center, Lackland AFB, Tex (W.B.M.) Received March 17, 1999; revision requested May 3; revision received June 14; accepted July 19. Address reprint requests to M.E.S. (e-mail: mark.schweitzer@mail.tju.edu).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Imaging studies in 33 patients were evaluated, including 14 patients with spinal neuropathic arthropathy (12 radiographic, seven CT, and six MR studies) and 19 with disk space infection (13 radiographic, nine CT, and 12 MR studies). Potential imaging discriminators, including endplate sclerosis or erosions, osteophytes, spondylolisthesis, facet involvement (narrowing or erosions), vacuum disk, paraspinal soft-tissue mass, joint disorganization, and osseous joint debris, were recorded, as were MR imaging signal intensity and gadolinium-enhancement characteristics.
RESULTS: The most helpful findings for diagnosis of spinal neuropathic arthropathy were vacuum disk on radiographs and CT images, debris on radiographs and CT and MR images, disorganization on radiographs and CT and MR images, facet involvement on radiographs and CT and MR images, spondylolisthesis on CT and MR images, diffuse signal intensity patterns in vertebral bodies on MR images, and rim enhancement of disks on gadolinium-enhanced MR images. Findings that were not helpful included endplate sclerosis and erosions, osteophytes, paraspinal soft-tissue mass, and decreased disk height.
CONCLUSION: Vacuum disk, facet involvement, vertebral body spondylolisthesis, joint disorganization and debris, and gadolinium-enhancement patterns of vertebral bodies and disks may help differentiate spinal neuropathic arthropathy from infection.
Index terms: Computed tomography (CT), comparative studies • Magnetic resonance (MR), comparative studies • Spine, CT, 30.12111 • Spine, diseases, 30.78, 30.821 • Spine, infection, 30.21 • Spine, intervertebral disks, 30.78 • Spine, MR, 30.121411, 30.12143
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The clinical and radiologic distinctions between infection and neuropathic arthropathy are both important and difficult. Investigators (6) have described imaging findings that help differentiate these processes in the foot and ankle. As a consequence, in the present preliminary investigation, we compared the imaging characteristics of spinal neuropathic arthropathy with those of disk space infection to determine if these processes can be differentiated on the basis of imaging characteristics.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The 14 patients with spinal neuropathic arthropathy consisted of four men and 10 women aged 42–82 years (mean age, 63 years). Neuropathic arthropathy involved the thoracic spine in two patients (T7 through T8 in one, T10 through T11 in the other), the thoracolumbar junction in one patient (T12 through L1), and the lumbar spine in 11 patients (L1 through L2 in one, L2 through L3 in one, L3 through L4 in three, L4 through L5 in five, and L5 through S1 in one). The clinical history in these patients included diabetes mellitus in seven patients, paraplegia in five, polio in one, and congenital insensitivity to pain in one. All patients with spinal neuropathic arthropathy were followed up clinically for 6 months after imaging by a neurosurgeon and by one of the authors (G.J.P.). During follow-up, no patient developed clinical signs or symptoms of disk space infection.
The 19 patients with disk space infection consisted of six men and 13 women aged 45–91 years (mean age, 63 years). Disk space infection involved the thoracic spine in four patients (T8-9, T9-10, T10-11, and T11-12), the thoracolumbar junction in two patients (T12-L1), and the lumbar spine in 13 patients (L1-2 in three, L2-3 in two, L3-4 in two, L4-5 in three, and L5-S1 in three). The clinical history in these patients included diabetes mellitus in seven, paraplegia in six, previous back injury in four, polio in one, and scoliosis in one.
Imaging studies included conventional radiography, computed tomography (CT), and magnetic resonance (MR) imaging. Conventional radiographs were obtained by using the following standard projections: anterior-posterior (14 x 17-inch [35.6 x 43.2-cm] image), lateral (14 x 17-inch [35.6 x 43.2-cm] image), right posterior oblique (11 x 14-inch [27.9 x 35.6-cm] image), left posterior oblique (11 x 14-inch [27.9 x 35.6-cm] image), and lateral spot image at L5-S1 (8 x 10-inch [20.3 x 25.4-cm] image). A standard screen-film technique was used with fast-detail screens and film (Ultravision and Ultravision L, respectively; Sterling Diagnostic, Newark, Del) for a system speed of ASA 250. CT studies were obtained by using 3 x 3-mm transverse sections, a 130–150-mm scanning field, and sagittal reconstruction (Horizon; GE Medical Systems, Milwaukee, Wis). MR images were obtained with a 1.5-T unit (Signa; GE Medical Systems, Milwaukee, Wis). T1-weighted (500–600/9–13 [repetition time msec/echo time msec]) and T2-weighted (2,400-6,000/70-105) spin-echo images were obtained in the sagittal and transverse planes. In five patients with spinal neuropathic arthropathy and 11 patients with disk space infection, transverse T1-weighted spin-echo (500–550/11) and sagittal fat-suppressed T1-weighted spin-echo (500–550/11) MR images were obtained immediately after intravenous bolus injection of a standard dose (0.1 mmol per kilogram of body weight) of gadopentetate dimeglumine (Magnevist; Berlex, Wayne, NJ).
Among the 14 patients with spinal neuropathic arthropathy, conventional radiographs in 12 patients, CT scans in seven, and MR studies in six (including gadolinium-enhanced studies in five) were available for review. Among the 19 patients with disk space infection, conventional radiographs in 13 patients, CT scans in nine, and MR studies in 12 (including gadolinium-enhanced studies in 11) were available for review.
All imaging studies were interpreted randomly and independently by two experienced musculoskeletal radiologists (M.E.S., W.B.M.) in blinded fashion. The reviewers were given a predetermined set of imaging criteria to be evaluated. All images were evaluated for the presence endplate sclerosis, endplate erosions, osteophytes, decreased disk height, spondylolisthesis, facet involvement (narrowing or erosions), vacuum disk phenomenon, paraspinal soft-tissue mass, disorganization (ie, alterations in the articular contour resulting in incongruity of the intervertebral joint), and debris (ie, osseous fragments in or adjacent to the intervertebral joint). Signal intensity characteristics also were evaluated on T1- and T2-weighted MR images, as was enhancement of the vertebral bodies and intervertebral disks, when gadolinium-enhanced images were available. Findings were recorded as present or absent.
Statistical analysis of the results was limited by the small sample sizes; therefore, the results are reported as frequencies within the sample population. Frequencies reported in the text represent the results obtained by a single interpreter. However, results from both interpreters are listed in Tables 1–3. The sensitivities and specificities for spinal neuropathic arthropathy were calculated for each imaging finding. The sensitivity was calculated by dividing the frequency of the finding present in patients with spinal neuropathic arthropathy by the number of studies in patients with spinal neuropathic arthropathy and multiplying by 100. The specificity was calculated by dividing the frequency of the absence of the finding in patients with a disk space infection by the number of studies in patients with a disk space infection and multiplying by 100. Interobserver agreement statistics are not reported owing to the small sample sizes.
|
|
|
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
MR Imaging Findings
MR images were evaluated for findings similar to those on radiographic and CT images, as well as for T1- and T2-weighted signal intensity and gadolinium-enhancement characteristics of intervertebral disks and vertebral bodies (Table 3). As noted for the radiographic and CT results, similar frequencies of endplate sclerosis, endplate erosions, osteophytes, and paraspinal soft-tissue mass were present both in patients with spinal neuropathic arthropathy and in those with disk space infection. Again, debris and disorganization were present more frequently in patients with spinal neuropathic arthropathy than in those with disk space infection. Debris and disorganization were each present in two of six patients with spinal neuropathic arthropathy. In comparison with the 12 patients with disk space infection, debris was present in one patient, and disorganization was absent in all. Vertebral body spondylolisthesis and facet joint involvement were again more frequent in patients with spinal neuropathic arthropathy. Vertebral body spondylolisthesis was present in four of six patients with spinal neuropathic arthropathy and in one of 12 patients with disk space infection. Facet joint involvement was present in four of six patients with spinal neuropathic arthropathy, whereas no patient with disk space infection had facet joint involvement.
The results of the MR imaging evaluation of intervertebral disks are shown in Table 3. Both the patients with spinal neuropathic arthropathy and those with disk space infection frequently had a decrease in disk height. No distinguishing signal intensity characteristics were noted in the disks of either group. In patients with spinal neuropathic arthropathy, disks were uniformly isointense relative to muscle on T1-weighted images in five of six patients and isointense to fluid on T2-weighted images in five of six patients. In patients with disk space infection, disks were uniformly isointense to muscle on T1-weighted images in all 12 patients, whereas on T2-weighted images, disk signal intensity relative to that of fluid varied such that it was isointense in five patients, hypointense in five patients, and hyperintense in two patients. Gadolinium enhancement of disks differed between the groups such that rim enhancement was present more frequently in patients with spinal neuropathic arthropathy, and diffuse enhancement (Fig 4a) was more common in patients with disk space infection. A pattern of rim enhancement was present in four of five patients with spinal neuropathic arthropathy, whereas rim enhancement was present in two of 11 patients with disk space infection. Of the remaining nine patients with disk space infection, MR images in six showed diffuse enhancement, and MR images in three showed no enhancement.
|
|
|
|
|
|
The sensitivity and specificity for each MR imaging finding are listed in Table 3. The best discriminators on MR images were facet involvement, vertebral body spondylolisthesis, disorganization, debris, diffuse signal intensity in the vertebral body on both T1- and T2-weighted images, rim enhancement of the disk, and diffuse enhancement of the vertebral body. Less specific MR imaging findings were endplate erosion, soft-tissue mass, osteophytes, decreased disk height, low-signal-intensity vertebral body on T1-weighted images, and high-signal-intensity vertebral body on T2-weighted images.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The pathologic changes seen in neuropathic joints produce radiographic changes that may be difficult to differentiate from those of other disease processes. The distinction between neuropathic joint and infection in the foot of a patient with diabetes provides an example of the difficulty in differentiating these processes, since both have been shown to manifest with similar clinical and radiographic findings (11–14). As in the foot, a similar diagnostic dilemma may exist in the differentiation of spinal neuropathic arthropathy from infection. Previous observers (3,5,15,16) have reported that conventional radiographic and CT findings of spinal neuropathic arthropathy mimic findings of discitis and osteomyelitis. It has been suggested (3,5,15–17), however, that MR imaging may help differentiate neuropathic changes from infection. To our knowledge, the present work included the largest series of imaging studies in patients with spinal neuropathic arthropathy, which were used to determine if imaging findings can help differentiate spinal neuropathic arthropathy from disk space infection.
Our findings were similar to those classically described (18) as the "six D's" of the neuropathic joint: distention, density, debris, disorganization, dislocation, and destruction. We found that facet involvement, spondylolisthesis (antero- or retrospondylolisthesis), vacuum disk, osseous debris, joint disorganization, vertebral body signal intensity pattern on T2-weighted MR images, and gadolinium-enhancement pattern on MR images were criteria suggestive of the diagnosis of spinal neuropathic arthropathy and may help differentiate it from disk space infection.
For example, vertebral body spondylolisthesis was present on radiographs in six of 13 patients, absent on CT studies in all patients, and present on MR studies in only one of 12 patients with disk space infection. However, spondylolisthesis was present on radiographs in eight of 12 patients, on CT studies in five of seven patients, and on MR studies in four of six patients with spinal neuropathic arthropathy. Vacuum disk and facet involvement were uncommon in patients with disk space infection but were commonly seen on radiographs, CT scans, and MR images in patients with spinal neuropathic arthropathy. The rare finding of vacuum disk in patients with disk space infection supports previous findings (19) that the vacuum disk phenomenon is uncommon in spinal infection, with few exceptions. Debris and disorganization also were seen more commonly on radiographs, CT scans, and MR images in patients with spinal neuropathic arthropathy than in those with disk space infection. These findings were similar to those in previous reports (16,17), but our findings also included endplate erosion, vertebral body spondylolisthesis (antero- or retrospondylolisthesis), vacuum disk, and facet involvement.
MR imaging findings were evaluated in detail. Although the vertebral disks in both groups showed a similar decrease in height, as well as changes in signal intensity on T1- and T2-weighted images, the gadolinium-enhancement patterns of disks differed between groups: Images in patients with spinal neuropathic arthropathy frequently showed rim enhancement, whereas those in patients with disk space infection showed diffuse disk enhancement. Vertebral bodies in both groups of patients showed similar low signal intensity on T1-weighted images and high signal intensity on T2-weighted images. However, the high signal intensity on T2-weighted images was more frequently seen as diffuse in patients with spinal neuropathic arthropathy and at the endplates in patients with disk space infection. Gadolinium-enhanced MR images showed a pattern of enhancement similar to the distribution of high signal intensity on T2-weighted images, with diffuse vertebral body enhancement in patients with spinal neuropathic arthropathy and endplate enhancement in those with disk space infection. Our MR imaging findings were similar to those of Kapila and Lines (16) and Arnold et al (20), who reported decreased signal intensity in the disk and adjacent vertebral bodies on T1-weighted images and increased signal intensity on T2-weighted images. Our findings, however, differed from those of Park et al (17), who reported decreased signal intensity on T2-weighted images.
Seabold et al (12) suggested that the increased signal intensity on T2-weighted images in patients with spinal neuropathic arthropathy represents an acute stage of the disease in which inflammation and reparative changes are predominant. In this setting, bone marrow edema predominates over sclerotic changes in bone, and recent-onset rapidly progressive neuropathy cannot be distinguished from infection (12). Thus, the patients with spinal neuropathic arthropathy in our series may represent cases of recent onset or rapidly progressive disease. However, increased signal intensity may simply be due to the possibility that patients with spinal neuropathic arthropathy who are referred for imaging are often symptomatic, with either a new onset of symptoms or a rapidly worsening disease process; hence, they may be more likely to have bone marrow edema and inflammation. It is possible that subsequent T2-weighted MR studies in these patients would show low signal intensity as marrow edema is replaced by sclerosis. Although we did not quantify T2-weighted signal intensity, Craig et al (11) have suggested that the higher the signal intensity on T2-weighted images, the more likely the process represents osteomyelitis.
Recently, Wolansky et al (21) reported a potential pitfall in the MR imaging diagnosis of infectious spondylitis. In infectious spondylitis, the normal lipid-rich bone marrow is replaced by a water-based infiltrate, which causes a loss of the normal chemical shift artifact at the interface between intervertebral disk and vertebral body and renders the endplate more conspicuous. Because a classic finding of infectious spondylitis is loss of endplate definition, the "pseudosparing" represents a potential pitfall. In our study, endplate erosions and/or sclerosis were not useful MR imaging findings to help discriminate spinal neuropathic arthropathy from disk space infection. Thus, the MR imaging appearance of pseudosparing, although pertinent to the diagnosis of infectious spondylitis, has only a minor role in facilitating differentiation of spinal neuropathic arthropathy from disk space infection.
The best discriminators on radiographs were vacuum disk, debris, and disorganization; facet involvement was less useful. Similarly, the best discriminators on CT images were vacuum disk, facet involvement, disorganization, and debris, as well as vertebral body spondylolisthesis. The best discriminators on MR images were facet involvement, disorganization, vertebral body spondylolisthesis, debris, diffuse signal intensity in the vertebral body on T1- and T2-weighted images, disk rim enhancement on gadolinium-enhanced images, and diffuse enhancement of the vertebral body.
Our study was limited somewhat by the criteria used to diagnose spinal neuropathic arthropathy. In our study, spinal neuropathic arthropathy was diagnosed on the basis of bone biopsy specimens that yielded negative cultures and no histologic evidence of infection, while disk space infection was diagnosed on the basis of positive cultures and/or histologic evidence of infection. These criteria alone do not account for the potential false-negative results reported (8) for 32%–58% of bone biopsy cultures and 16% of bone biopsy histopathologic evaluations. However, no patient with spinal neuropathic arthropathy subsequently developed clinical signs or symptoms of disk space infection during the 6-month neurosurgical follow-up.
Owing to the preliminary nature of our study, an additional limitation was the small sample size of 14 patients with spinal neuropathic arthropathy. This precluded comparative statistical analysis of the results.
In summary, we compared the imaging findings from conventional radiography, CT, and MR imaging of spinal neuropathic arthropathy with the imaging findings of disk space infection. We identified several imaging criteria that may be useful for the differentiation of spinal neuropathic arthropathy from disk space infection. Facet involvement (narrowing or erosions) and vacuum disk were common findings in patients with spinal neuropathic arthropathy but were rare in patients with disk space infection; thus, the presence of the former is suggestive of spinal neuropathic arthropathy. Other findings such as vertebral body spondylolisthesis, osseous joint debris, joint disorganization, and, on MR images, gadolinium enhancement in the periphery of the disk (rim enhancement) and diffusely in the vertebral body were more frequently seen in cases of spinal neuropathic arthropathy; thus, their presence is suggestive of spinal neuropathic arthropathy. Endplate erosion, endplate sclerosis, osteophyte formation, loss of disk height, paraspinal soft-tissue mass, and signal intensity characteristics on T1- and T2-weighted MR images were not useful for the differentiation of spinal neuropathic arthropathy from disk space infection. The reported (16,17) observation that T2-weighted MR images can help differentiate spinal neuropathic arthropathy from disk space infection was not supported by our results. To further evaluate the MR imaging findings in patients with spinal neuropathic arthropathy, it may be useful to compare such findings over time to determine if the characteristics on T2-weighted images correlate with stage of disease.
|
Footnotes |
|---|
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. H. Hong, J.-Y. Choi, J. W. Lee, N. R. Kim, J.-A. Choi, and H. S. Kang MR Imaging Assessment of the Spine: Infection or an Imitation? RadioGraphics, March 1, 2009; 29(2): 599 - 612. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. B. Smith, A. G. Kane, A. H. Sholes, and J. H. Freeman Resolution of an Aggressive Ideopathic Diskitis AJNR Am. J. Neuroradiol., May 1, 2005; 26(5): 1008 - 1011. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Koenigsberg, P. Black, S. H. Faro, and J. Rykken Lumbar Disk Pseudotumor: An Unusual Presentation of Lumbar Spinal Fracture and Stenosis AJNR Am. J. Neuroradiol., November 1, 2002; 23(10): 1662 - 1664. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
