Early prediction of refractory myasthenia gravis based on response to treatment within the first year of diagnosis.

Background: The long lead time required to diagnose refractory myasthenia gravis (MG) can prevent early attempts at new treatments. Herein, we aimed to predict refractory MG by comparing treatment responses during the early stages between patients with refractory and non-refractory MG.

Methods: We retrospectively investigated the medical records of patients with acetylcholine receptor antibody-positive generalized MG. Refractory MG was defined as (1) inability to lower prednisolone below 20 mg/day, (2) requiring regular intravenous immunoglobulin or plasmapheresis, or (3) Myasthenia Gravis Activities of Daily Living score ≥ 6 after treatment with corticosteroids and ≥ 1 oral immunosuppressant for ≥ 12 months.

Results: Of 133 patients, 29 (21.8%) and 104 (78.2%) had refractory and non-refractory MG, respectively. The median duration from the diagnosis of MG to that of refractory MG was 31 months. During the first year following diagnosis, the proportion of patients who achieved a lowest effective prednisolone dose of ≤ 20 mg was lower in the refractory group than in the non-refractory group (69.0% vs. 92.0%). Conversely, the proportions of patients who initiated oral immunosuppressive therapy (58.6% vs. 22.5%), required rescue interventions (48.3% vs. 22.1%), or experienced a myasthenic crisis (34.5% vs. 14.4%) were significantly higher in the refractory group. The risk of refractory MG was 18.7-fold higher in patients who failed to achieve lowest effective prednisolone dose under 20 mg/day and started oral immunosuppressant during the first year than those with neither of these conditions.

Conclusions: The refractory group displayed distinct clinical features during the early disease stages, including frequent crises and higher levels of immunosuppressive treatment. These features can be used for early prediction of refractory MG.