Early use of immunosuppression has been suggested to prevent generalization of ocular myasthenia gravis (OMG), but high-quality evidence is limited in this regard. We examined whether treatment with prednisone and other immunosuppressants reduce the risk of generalization in OMG.

This is a retrospective study of consecutive adults with pure OMG who had a minimum 6 months of follow-up. The main outcome was the time to developing generalized symptoms. We used propensity scores to create matched data sets of patients treated with prednisone or any immunosuppressant vs controls. We also used unmatched models with inverse probability of treatment weights (IPTW) and variable exposure times. We used Cox proportional hazards model to estimate hazard ratio (HR) for generalization, comparing treated patients vs controls.

A total of 154 patients were included, with a mean follow-up of 87.4 ± 73 months since onset. Forty-three (28%) were generalized, and mean time to generalization from diagnosis was 24.2 ± 24.1 months. Patients who received prednisone had lower risk of generalization than controls, with pooled HR 0.43 (95% CI 0.19-1.06) for the matched model, HR 0.46 (95% CI 0.21-0.89) for the IPTW model, and for HR 0.44 (95% CI 0.23-0.81) for the time-dependent exposure model. Patients who received any immunosuppressant had lower risk of generalization, with HR 0.30 (95% CI 0.11-0.77), 0.32 (95% CI 0.14-0.70), and 0.35 (95% CI 0.15-0.80) for the matched, IPTW, and IPTW-varying exposure models, respectively.

Our study provides evidence that steroidal and nonsteroidal immunosuppression in patients with OMG is associated with a reduced risk of developing generalized symptoms over time. This supports the early use of immunosuppression in this population.

This study provides Class III evidence that treatment of OMG with corticosteroids or nonsteroidal immunosuppressants reduces the risk of generalization.