PT  - JOURNAL ARTICLE
AU  - J T Wilmink
TI  - CT morphology of intrathecal lumbosacral nerve-root compression.
DP  - 1989 Mar 01
TA  - American Journal of Neuroradiology
PG  - 233--248
VI  - 10
IP  - 2
4099  - http://www.ajnr.org/content/10/2/233.short
4100  - http://www.ajnr.org/content/10/2/233.full
SO  - Am. J. Neuroradiol.1989 Mar 01; 10
AB  - Spinal CT scans and radiculograms of 100 patients who had undergone both examinations were studied with the aim of identifying morphologic CT features associated with compression of the intrathecal segment of the nerve root as demonstrated by radiculography. The interest for such a study lies in the fact that, in contradistinction to the distal, extrathecal root segment outlined by fat in the foramen, the proximal segment within the dural sac cannot be distinguished from the surrounding CSF by CT. CT features assessed consisted of deformation of the dural sac and displacement of the surrounding epidural fat. These features were compared with radiculographic signs of root involvement in the same location: kinking of the nerve root, local swelling of the root within the dural sac or the root sheath, and cutoff of root-sheath filling. In addition, a separate "expert opinion" verdict was given in each location as to the likelihood of compression of the intrathecal root segment on the basis of CT findings as confirmed or rejected by radiculography. A degree of correlation existed between CT and radiculographic features in clearly normal or abnormal locations, but there were some marked discrepancies. In borderline cases there were many discrepancies. The lack of agreement was confirmed by generally disappointing kappa values. The expert opinion, combining separate radiologic features into verdicts for CT and radiculography, did not lead to significantly better agreement as expressed by kappa, but less extreme discrepancies were seen. It proved to be possible, however, to distinguish locations with CT features likely to correlate well the radiculographic picture from those less likely to show good correlation. Such a distinction can guide the decision whether or not to perform confirmatory radiculography. False-positive locations were identified by questionnaire; CT and radiculographic features proved to differ only a matter of degree from the same features is presumably symptomatic locations. CT and radiculographic findings in compression of the intrathecal segment of the nerve root are complementary. Correlating the two studies is of limited value because they provide different anatomic information.