Résumé
This meta-analysis was designed to elucidate whether preoperative signal intensity changes could predict the surgical outcomes of patients with cervical spondylosis myelopathy on the basis of T1-weighted and T2-weighted magnetic resonance imaging images. We searched the Medline database and the Cochrane Central Register of Controlled Trials for this purpose and 10 studies meeting our inclusion criteria were identified. In total, 650 cervical spondylosis myelopathy patients with (+) or without (-) intramedullary signal changes on their T2-weighted images were examined. Weighted mean differences and 95g% confidence intervals were used to summarize the data. Patients with focal and faint border changes in the intramedullary signal on T2 magnetic resonance imaging had similar Japanese Orthopaedic Association recovery ratios as those with no signal changes on the magnetic resonance imaging images of the spinal cord did. The surgical outcomes were poorer in the patients with both T2 intramedullary signal changes, especially when the signal changes were multisegmental and had a well-defined border and T1 intramedullary signal changes compared with those without intramedullary signal changes. Preoperative magnetic resonance imaging including T1 and T2 imaging can thus be used to predict postoperative recovery in cervical spondylosis myelopathy patients.