Reassessment of the diagnostic value of selective lumbosacral radiculography

In nerve root infiltration (NRI) consisting of neural blockage and radiculography, response to the nerve root block has usually been thought to be diagnostically significant. However radiculography has not been statistically evaluated. The purpose of this paper is to assess the value of selective radiculography of patients with group 1 response (typical pain reproduced by needle placement and then relieved by nerve root block) according to Dooley's criteria. We studied selective radiculography in a consecutive series of 88 patients with lumbo-sacral radicular pain who showed group 1 response in NRI. The accuracy of the preoperative nerve root block and radiculography in 88 nerve roots (L5,S1) were correlated with the intraoperative findings. Selective radiculograms were classified into three groups; normal (absence of block), partial block, and complete block. The tilting angle of all nerve roots was measured. We found the symptomatic root at the same level of the nerve root block in all 88 patients. Selective radiculograms showed five normal roots, 15 roots with incomplete block and 63 roots with complete block. Fifteen radiculograms had abnormal tilting angles. The accuracy of radiculography was 84% in the canal zone and 100% in the intra and extraforaminal zones. If the L5 nerve root angle was more than 60(o), an intra or extraforaminal lesion was strongly suggested (P<0.01). Radiculography of patients with group 1 response is useful for detecting compressed sites in the symptomatic nerve root, particularly for detecting lesions in the intra and extraforaminal zones.