To estimate the within-patient association between biologic therapy and hospitalization rates among adults with generalized myasthenia gravis (gMG).

Biologics are increasingly used in gMG. We aimed to provide real-world evidence of outcome improvement compared to non-biologics.

We retrospectively identified patients with gMG treated at Mass General Brigham treated with biologics defined as chronic administration of IVIg, SCIg, rituximab, eculizumab, ravulizumab, or efgartigimod. Outcome was gMG related hospitalization comparing pre- and post-biologic therapy. First 30-days after starting biologic were treated as washout. We constructed person-period data and fit patient fixed-effects (FE) Poisson model with log person-time offsets, clustering by patient and adjusting for Charlson-Comorbidity Index, MGFA Clinical Classification, time-since-diagnosis, and calendar quarter. The FE analysis was restricted to patients contributing both non-biologic and biologic time with complete covariates.

Seventy-seven patients with gMG met inclusion criteria. Median age at gMG diagnosis was 65 years [IQR 53–72] with 51% of females. MGFA-CC at diagnosis was I–III in 85%. Antibody data was available in 65 patients, 97% were AChR-positive and 3% MuSK-positive.

Patients contributed 184.28 person-years (PY) off-biologic with 41 hospitalizations (22.25 per 100 PY) and 380.09 PY on-biologic with 36 hospitalizations (9.47 per 100 PY); crude post/pre biologic incidence rate ratio of hospitalizations was 0.43. In the FE model, current biologic exposure was associated with a markedly lower hospitalization rate (incidence rate ratio 0.01, 95% CI 0.00–0.09; p<0.001). 58/77 (75.3%) patients had zero on-biologic hospitalizations. Per-patient person-time contributed amounted to (median [IQR]): 0.73 PY [0.19–3.12] off-biologic; 4.01 PY [1.61–6.67] on-biologic. Drugs contributing on-biologic time included IVIg (n=65), rituximab (n=15), eculizumab (n=11), efgartigimod (n=7), and ravulizumab (n=3).

In a within-patient design with lagging and temporal controls, biologic therapy was associated with fewer hospitalizations compared with non-biologics. The large effect-size reflects longer post-biologic follow-up and a high proportion of zero post-biologic hospitalizations.