This retrospective cohort study aimed to delineate disease trajectories, antibody subtypes, and long-term outcomes in a large French Myasthenia gravis (MG) cohort.

We retrospectively analyzed 119 MG patients followed over 18 year at the Mondor University Hospital neuromuscular reference center. Demographics, clinical, antibody status, thymic pathology, treatment patterns, and long-term outcome were collected. Outcomes included generalization, relapse, remission, and mortality. Kaplan-Meier, Cox regression, and logistic regression identified predictors.

The cohort included 77% (N= 92) anti-AChR, 5% (N=6) anti-MuSK, and 17.6% (N=21) seronegative patients. Generalized MG accounted 76% (N=90) , ocular MG 24% (N=29). Relapses occurred in 72% (N=86). About half of ocular MG later generalized, often after two years, with higher generalization risk in female and late-onset MG. Severity at diagnosis and of the first relapse independently predicted long-term burden. Remission (MGFA 0–I) was achieved in 42% (N=50), more frequently in ocular and relapses-free patients. Seronegative patients had longer relapse-free intervals and a higher proportion of sustained ocular disease. At last follow-up, 31% (N=37) used anticholinesterase monotherapy, 62% (N=74) received corticosteroids (45% discontinued, N=33), and 80% (N=95) received azathioprine or mycophenolate. Rituximab, FcRn and complement inhibitors were used in 15% (N=18). Mortality was comparable to the general population.

Early disease severity and occurrence of the first relapse were strong predictors of long-term outcomes in MG, underscoring the importance of early, risk-adapted management. Ongoing vigilance for generalization remains essential, particularly in ocular-onset and elderly-onset cases. In this real-world cohort, both conventional therapies and newer agents targeting FcRn or complement pathways were associated with high remission rates and facilitated corticosteroid tapering.