Neurnradiologv

Neuroradiology(1981) 22: 57-60

© Springer-Verlag 1981

Originals CT Scan of the Spine for Herniated Discs A. N. Gulati, R. Weinstein and E. Studdard Department of Radiology,GeorgeWashingtonUniversityMedicalCenter,Washington,DC, USA

Summary. The high resolution conversion kit introduced for the EMI CT 5005 scanner has twice the spatial resolution of the usual scan mode and has proven to be quite useful in evaluation of the spinal canal. The present study was performed to determine the capability of the high resolution system to diagnose herniated lumbar disc without intrathecal contrast. HNP was diagnosed in 83% of the cases which had a positive myelogram, 70% of which had operation and all were proven positive. Key words: High resolution - Computed tomography Spinal canal - Herniated intervetebral disc

Introduction The use of the high resolution mode for the EMI CT 5005 scanner doubles the spatial resolution of that unit from 5.5 line pairs/cm to 11 line pairs/cm, which is perhaps higher than any presently available commercial scanner operating without zoom reconstruction of raw data. CT scanning with water soluble contrast such as metrizamide 1 makes delineation of herniated discs in the lumbar region possible. However, this technique is merely an extension of myelography, and remains a noxious procedure. The following study was performed to determine the overall capability of the high resolution system to diagnose herniated lumbar disc without metrizamide. CT scans of intervertebral disc spaces were performed with and without metrizamide in the subarachnoid space and the findings were correlated with standard myelograms and operative results. Amipaque/Winthrop

Materials and Methods Ten patients who were felt clinically to have a lumbar herniated nucleus pulposus (HNP) and had positive findings, at least at one level on standard lumbar myelography with metrizamide, were included in this study. High resolution CT scans of the appropriate levels of the spine with and without metrizamide in the subarachnoid space were performed in all cases. Myelography was performed in a routine fashion after instillation of 10 to 12 cc of metrizamide (220 mg of iodine per cc) into the subarachnoid space at the L2_3 level via a 22 gauge needle. The disc levels were marked on the skin with ink at the time of the fluoroscopy so that they could easily be localized at the time of CT examination. This method proved quite satisfactory. If scans were performed prior to myelography a standard AP film of the lumbosacral spine was used for localization purposes. Most scans without contrast were obtained 36 to 48 h after the performance of myelography when the metrizamide had been absorbed. All CT scans were performed on the EMI 5005 scanner with a high resolution technique utilizing 25 cm wedges; 5 mm sections were obtained in nine patients, but 8 mm overlapping scans were obtained in one patient. On the average, six scans were required to evaluate an intervertebral space and adjoining vertebral bodies. Standard CT examination of the spinal canal reveals the dural sac as a rounded or slightly oval density within the canal, outlined by a variable amount of fat which separates the sac from the vertebral bodies, discs, and posteriorly, from the neural arches and ligaments. Nerve root sleeves arise anterolaterally and are seen as soft tissue densities separated from the dural sac, in the anterolateral spaces (Fig. 1). HNP or disc fragments are recognized as soft tissue densities projecting into the anterior part of the spinal canal,

0028-3940/81/0022/0057/$01.00

58

Fig. 1 a and b. 5 mm contiguous baseline scans through lower part of body of Ls. Note dural sac (*) surrounded by variable amount of epidural fat. Note also a nerve sleeve (--,) arising from the dural sac in scan a, and lying in the anterolateral recess in lower scan b

Fig. 2a and b. CT of spine at L4_5disc space. a with, b without metrizamide in dural sac (*). Note ( • ) HNP as soft tissue density obliteraing epidural fat and nerve sleeve and also effacing dural sac

Fig. 3a and b. CT of spine at L34 disc space. a with and b without metrizamide in dural sac (*). Note (m) bulging disc indenting dural sac and obliterating epidural space anteriorly and anterolaterally. Note also (---~)hypertrophied ligamentum flavum, contributing to narrowing of spinal canal

usually at the level of the intervertebral disc. A free fragment, however, may be seen rostral or caudal to the disc space. Depending on the size of the herniation, displacement, partial or complete obliteration of the epidural fat occurs. Additionally, there may be nonvisualization of the root sleeve, and effacement of the dural sac itself (Fig. 2). Bulging discs project posteriorly from the entire intervertebral disc space, and

appear as a convex density, with slightly greater Hounsfield units (HU) than that of the subarachnoid space. They produce effacement of the epidural fat, lack of visualization of the nerve sleeve, and indentation of the dural sac (Fig. 3). Calcification may be seen in the herniated or bulging disc. We have also noted gas in a herniated disc in one patient whose disc showed a vacuum phenomenon [1].

59 Table 1. Myelographic, CT, and operative findings in 10 cases Case

Age

Sex

Myelogram

CT without metrizamide

CT with metrizamide

Operative findings

1 2 3

48 39 66

m m f

L5 S~ hnp a ft. L4-5 hnp It. L4-5 hnp rt. Ls-Si normal L4-5 hnp lt. L5-$1 normal L4-5 hnp It. L4-5 hnp lt. Ls-SI hnp lt. Ls-S~ hnp It. L,_5 hnp Ls-S~ normal Ls-SI bulging disc L4-5 hnp It, L3-4 bulging disc

+ + _ b + + _+ + + + + + + +

+ +

+ + ~_

4

33

m

5 6 7 8 9

43 43 28 38 36

m m m f m

10

79

f

-

-

+ + + + + + + + + +

b

no operation + + + + + no operation no operation no operation

~' herniated nucleus pulposus ~ Artifacts on CT

Tabelle 2. Correlation between myelogram and computed tomography (CT) Myelogram

negative positive

CT without metrizamide 3 12

negative positive negative questionable

CT with metrizamide 3 10 1 1

negative positive negative questionable

3 I1 1 0

Results

Table 1 shows the myelographic, CT and surgical findings in ten patients. The correlation between the myelogram and CT with and without metrizamide, of the 15 intervertebral disc spaces studied in the ten patients is shown in Table 2. Of the 15 disc spaces evaluated, the myelogram was negative in three. CT, with and without metrizamide, was also normal at those three levels. The remaining 12 levels had positive myelographic findings, but the noncontrast CT was positive only in ten (83%). One space was definitely negative, while another was questionable. The negative pre-metrizamide CT findings were due to significant degradation by artifacts from bowel gas, which also caused the post metrizamide CT to be interpreted as normal at that space. If the scans rendered suboptimal by artifact are excluded, no false negative or false positive results were obtained. Findings on CT before and after metrizamide were generally in agreement, except for one patient where disc herniation was questionable before instillation of metrizamide, but was positive after metrizamide had been instilled. The

regular myelogram was positive in this latter case. Seven out of l0 cases with positive CT (non-metrizamide) findings of H N P underwent surgery and all were proven positive.

Discussion

CT scan is proving to be increasingly useful in the evaluation of the lumbar spinal area, both for evaluation of the contents of the canal, and also its size. Generalized and focal narrowing of the spinal canal, or stenosis [2-4] can be usefully shown, and facet disease and fractures [5, 6] are also easily evaluated. As the spatial and contrast resolution of CT scanners improve, their role in the evaluation of the spinal canal is expanding, but few reports attesting to the usefulness of CT as the only diagnostic procedure for H N P are available. Meyer et al. [7], and Williams, et al. [8], using the GE CT 8800 scanner, have indicated that the CT scan alone may be reliable for diagnosis of disc disease and recent reports by Ethier [9] and Naidich [10] suggest that the CT 5005 scanner with the high resolution system may also be useful for this purpose. Our work tends to Support this conclusion in that 10 out of 12 myelographically proven disc lesions were shown by CT scanning without the use ofintrathecal contrast material. Additionally, no false negatives were encountered. While the sample size is quite small and definitive recommendations may not be made, it appears that the following suggestions are useful. If indeterminate or indefinite clinical signs of an H N P are present, and an optimal, nondegraded CT scan shows a normal

60

disc space, no further studies need to be done. Specifically, myelography need not be performed. If the clinical findings are definite, further studies, including myelography, should be performed. Finally, when clinical findings correspond to a positive noncontrast CT diagnosis of herniated nucleus pulposus at the appropriate level, surgery may be performed without preoperative myelography. Acknowledgments.We thank our Special Procedure technicians for their help in the conduct of this study, and Diane Hankla and Lilly Desy for secretarial assistance.

6. Miller EM (1980) The role of CT in the evaluation of spinal trauma. Presented at the 18th Annual Meeting of the American Society of Neuroradiology, 16-21 March 1980 7~ Meyer GA, Haughton VA, William AL (1979) Diagnosis of herniated lumbar disc with computed tomography. N Engl J Med 301:1166-1167 8. Williams AL, Haughton VM, Syvertsen A (1980) Computed tomography in the diagnosis of herniated nucleus pulposus. Radiology 135: 95-99 9. Ethier R, King DG, Melancon D, Gelanger G, Taylor S, Thompson C (1979) Development of high resolution computed tomography of the spinal cord. J Comput Assist Tomogr 3: 433438 10. Naidich TP, King DG, Moran CJ, Sagel SS (1980) Computed tomography of the lumbar thecal sac. J Comput Assist Tomogr 4:3743

References 1. Gulati AN, Weinstein ZR (1980) Gas in the spinal canal in association with the lumbosacral vacuum phenomenon: CT findings. Neuroradiology 20: 191-192 2. Hammerschlag SB, Wolpert SM, Carter BL (1976) Computed tomography of the spinal cord. Radiology 121 : 361-367 3. Lee BCP, Kazam E, Newman AD (1978) Computed tomography of the spine and spinal cord. Radiology 128:95-102 4. Hirschy JC, Leue WM, Berniger WH, Hamilton RG, Abbott GF (1980) CT of the lumbosacral spine: importance of tomographic planes parallel to vertebral end plate. AJNR 1: 551-556 5. Colley DP, Dunsker SB (1978) Traumatic narrowing of the dorsolumbar spinal canal demonstrated by computed tomography. Radiology 129:95-98

Received: 4 September 1980 in revised form: 1 July 1981

Dr. A. N. Gulati Department of Radiology George Washington University Medical Center 901 23rd Street, NW Washington, DC 20037 USA