A prospective study of "undiagnosed" isolated myelopathy: I. Case selection and clinical features

We aimed to carry out a study of isolated, unexplained myelopathy which avoided as far as possible selection and referral bias. We used a prospective design with case ascertainment from both neurological and neurosurgical referrals. We wanted to establish the sensitivity, predictive value and possible redundancy of currently available tests as assessed by the proportion of such patients who finally developed clinically definite multiple sclerosis at follow up. Although MRI scanning of the spinal cord is now the investigation of choice for myelopathy, the case ascertainment for the present studies was completed, and follow-up commenced, before MRI became available to patients in our catchment region. We believe that this was an important factor in the ascertainment of a representative sample of undiagnosed myelopathies. Of 2104 patients having myelography in a 36-month period, 473 had signs of a myelopathy without lower motor neurone involvement Sixty-nine of these 473 patients (15%, 95% CI 11.8-18.2) eventually satisfied the entry criteria for the prospective trial. Fourteen of the 69 patients (20%) presented with acute or subacute cord syndromes, while 55 (80%) had chronic symptoms of more than 3 months duration. Approximately two-thirds of the chronic cases were progressive and one-third relapsing and remitting. Overall there was a female predominance (3:2) and a significant excess of HLA-DR2 antigen compared with a healthy control population (p < 0.0001). In this series the patients with chronic unexplained myelopathy were younger and had shorter symptom durations (mean 46.5 months) and less disability (mean EDSS 3.0) than cases in other comparable published series. The neurological findings, sex distribution and HLA characteristics were similar to those found in patients with clinically definite MS.