Predicting Postoperative C5 Palsy Using Preoperative Spinal Cord Rotation.

Abstract

The development of C5 nerve palsy after cervical decompression surgery has been well documented. The goal of this study was to determine whether preoperative spinal cord rotation could be used as a predictor of C5 palsy in patients who underwent posterior cervical decompression at C4-C6. The authors reviewed the records of 72 patients who had posterior decompression and 77 patients who had anterior decompression. With the patients undergoing anterior decompression used as a control group, magnetic resonance imaging scans were analyzed for area of the spinal cord, anterior-posterior diameter, and cord rotation relative to the vertebral body. The rate of C5 palsy was 7.3%. Average degrees of rotation were 3.83°±2.47° and 3.45°±2.23° in the anterior and posterior groups, respectively. A statistically significant association was detected between degree of rotation and C5 palsy. Point-biserial correlations were 0.58 (P<.001) and 0.60 (P<.001) in the anterior and posterior groups, respectively. With a diagnostic cutoff of 6°, the sensitivity and specificity of identifying patients with C5 palsy in the posterior group were 0.67 (95% confidence interval, 0.24-0.94) and 0.95 (95% confidence interval, 0.86-0.98), respectively. The results suggested that preoperative spinal cord rotation may be a valid predictor of C5 nerve palsy after posterior cervical decompression. With mild rotation defined as less than 6°, moderate rotation as 6° to 10°, and severe rotation as greater than 10°, the prevalence of C5 palsy in the posterior group was 2 of 65 for mild rotation, 3 of 6 for moderate rotation, and 1 of 1 for severe rotation.