BACKGROUNDCervical spondylotic myelopathy (CSM) is managed by conservative or surgical measures. While surgery is often performed in cases of longstanding or severe CSM, there is a lack of evidence concerning its efficacy. Transcranial magnetic stimulation (TMS) is a quick, safe, painless and non-invasive technique to study conduction in the descending corticospinal pathways in the spinal cord. The conduction time from the motor cortex to the anterior horn cell [central motor conduction time (CMCT)] is a measure of the integrity of corticospinal pathways. We have previously established the role of TMS in diagnosis and screening of CSM. In this study, we further investigate the use MEPs obtained with TMS in the outcome prediction of severe CSM patients requiring operative intervention.

METHODSWe prospectively evaluated 46 consecutive patients (mean age, 57.6 years; range, 36 to 84 years; 28 men) presenting with clinical features of CSM over a 2-year period. Disease duration ranged from 6 to 24 months. A total of 45 healthy controls were studied for comparison. All patients underwent clinical scoring. Patients' initial clinical score (S1) and postoperative scoring at 6 months (S2) were based on a modified Japan Orthopedic Association Scoring Scale. A Modified Recovery Rate (MRR) was calculated based on the formula: (S2 - S1/17 - S1) x 100. We regarded a good surgical outcome as MRR of 50 or above. This was depicted as MRR50. The patients were separated into 4 groups according to the degree of cord compression by degenerative osteo-cartilaginous elements at the most significant level on MRI. TMS studies were performed before surgery. Each investigator was blinded to the results of the other investigators.

RESULTSThe upper limb (UL) CMCT (r = -0.507, P <0.0005) and lower limb (LL) CMCT (r = - 0.452, P = 0.002) were significantly and negatively correlated with S1. Similarly, UL MEP amplitude (r = 0.494, P <0005) and LL MEP amplitude (r = 0.305, P = 0.039) were significantly correlated with S1. Surgery consisted of anterior or posterior decompression with cervical laminoplasty, performed by an experienced team of orthopaedic surgeons. No significant intraoperative or postoperative complications were documented. Surgery resulted in significantly improved clinical scoring (unpaired t test, P <0.0005). No correlation between clinical scoring with patients' age, disease duration, severity or levels of cord compression on MRI was found. ULCMCT and MEP amplitude abnormality were significantly associated with improvement in clinical scoring after surgery (Mann-Whitney test, P <0.05). The UL CMCT was the independent predictor of a good clinical outcome after surgery (odds ratio, 9.09; P = 0.011).

CONCLUSIONSIn early CSM, lateral corticospinal tracts are first to be affected. It is thus possible that UL CMCT abnormality reflect more severe affectation of the corticospinal tracts placed relatively more medially in the cervical cord. Surgical intervention may have then effectively relieved the clinically significant compression, leading to a better outcome. This was further corroborated by our finding of negative correlation of S1 with UL CMCT, suggesting that patients who were clinically more severe were also electrophysiologically more abnormal, and subsequently benefited more from surgical decompression relative to patients with normal UL CMCT. This the largest series, to our knowledge, showing for the first time that UL CMCT abnormality obtained with TMS is an independent predictor of good surgical outcome in severe CSM.