PURPOSE: To study the magnetic resonance (MR) imaging features of intradural nerve roots in avulsion injury of the cervical nerve roots. MATERIALS AND METHODS: Presurgical gadolinium-enhanced MR imaging was performed in 27 consecutive patients. The images were evaluated for enhancement of the intradural nerve roots and enhancement of the spinal cord surface at the root entry zone ("root stump" enhancement) by two radiologists, who were blinded to the clinical information. Intradural nerve roots were classified as enhanced, normal, or not visible. Root stumps were classified as enhanced or normal. The results were compared with the results of nerve root continuity studies (exploratory surgery with intraoperative recording of somatosensory evoked potentials or neurologic examination). RESULTS: At the 250 nerve levels (500 nerve roots) studied, 12 enhanced nerve roots and 42 enhanced root stumps were found. Seven enhanced nerve roots and 38 enhanced root stumps were related to preganglionic injury. Results of the kappa test for interobserver concordance on findings of intradural nerve roots and root stumps were 0.55 and 0.93, respectively. CONCLUSION: Although visualization of nerve roots is a problem, enhanced MR imaging is a noninvasive means of obtaining anatomic information and unique physiologic information on injured intradural nerve roots.
PURPOSE: To study the magnetic resonance (MR) imaging features of intradural nerve roots in avulsion injury of the cervical nerve roots. MATERIALS AND METHODS: Presurgical gadolinium-enhanced MR imaging was performed in 27 consecutive patients. The images were evaluated for enhancement of the intradural nerve roots and enhancement of the spinal cord surface at the root entry zone ("root stump" enhancement) by two radiologists, who were blinded to the clinical information. Intradural nerve roots were classified as enhanced, normal, or not visible. Root stumps were classified as enhanced or normal. The results were compared with the results of nerve root continuity studies (exploratory surgery with intraoperative recording of somatosensory evoked potentials or neurologic examination). RESULTS: At the 250 nerve levels (500 nerve roots) studied, 12 enhanced nerve roots and 42 enhanced root stumps were found. Seven enhanced nerve roots and 38 enhanced root stumps were related to preganglionic injury. Results of the kappa test for interobserver concordance on findings of intradural nerve roots and root stumps were 0.55 and 0.93, respectively. CONCLUSION: Although visualization of nerve roots is a problem, enhanced MR imaging is a noninvasive means of obtaining anatomic information and unique physiologic information on injured intradural nerve roots.