Preoperative Magnetic Resonance Imaging Is Associated With Baseline Neurological Status and Can Predict Postoperative Recovery in Patients With Cervical Spondylotic Myelopathy

Study Design. A blinded observational study of consecutive patients, prospectively enrolled and followed up to 1 year postoperatively. Objective. To assess whether quantitative assessment of preoperative magnetic resonance imaging (MRI) predicts baseline patient status and postoperative neurological recovery. Summary of Background Data. Factors that can predict neurological recovery in patients with cervical spondylotic myelopathy (CSM) postoperatively are of great interest. Currently, the literature regarding the significance of MRI signal changes in relation to prognosis is inconsistent and confl icting. Methods. A total of 57 consecutive patients with CSM were studied preoperatively and 1 year postoperatively. Modified Japanese Orthopaedic Association (mJOA), Nurick Grade, SF-36, neck disability index, 30-meter walk cadence (Wc) and time (Wt), grip strength, and Berg Balance Scale were administered at baseline and 1 year after surgery. Preoperative status and postoperative recovery were assessed in relation to quantitative MRI measurements pre-and postoperatively using univariate and multivariate analysis. Results. Low T1 signal change preoperatively was associated with a lower mJOA (P = 0.0030), higher Nurick Grade (P = 0.0298), decreased grip (P = 0.0152), impaired Wt, Wc (P <= 0.0001) and poor Berg Balance Scale (P = 0.0005) at baseline. Focal high T2 signal was associated with lower mJOA scores and higher Nurick Grade compared with diffuse T2 (P = 0.0035 P = 0.0079) or no T2 signal (P = 0.0680 P = 0.0122). Preoperative segmentation of T2 signal, showed a significant increase in Wt, Wc, and Berg Balance Scale (P = 0.0266; P = 0.0167; P = 0.0042). Preoperative T1 signal was associated with lower postoperative grip (P = 0.0260), greater Wt, Wc (P = 0.0360, P = 0.0090). Preoperative focal T2 signal had a significant association with poorer postoperative Wt, Wc (P = 0.0220) and Nurick Grade (P = 0.0230). Preoperative maximal cord compromise was negatively correlated with postoperative SF-36 mental score (P = 0.0130). Conclusion. MRI signal changes are predictive of baseline neurological status and postoperative recovery. MRI indicators of poorer outcome include the presence of low T1 signal, focal increased T2 signal and segmentation of T2 signal changes.