Predicting Severe Infection and Effects of Hypogammaglobulinemia During Therapy With Rituximab in Rheumatic and Musculoskeletal Diseases

OBJECTIVE: To evaluate predictors of serious infection events (SIEs) during rituximab (RTX) therapy and effects of hypogammaglobulinemia on SIE rates, and humoral response and its persistence after discontinuation of RTX in the treatment of rheumatic and musculoskeletal diseases (RMDs). METHODS: A retrospective longitudinal study of 700 RMD patients treated with RTX in a single center was conducted. Immunoglobulin levels were measured at baseline and at 4-6 months after each treatment cycle. Baseline predictors of SIEs were assessed using multivariable logistic regression; for RTX cycles 2-4, a mixed-effects logistic regression model was used. RESULTS: A total of 507 patients (72%) had rheumatoid arthritis, 94 (13%) had systemic lupus erythematosus, 49 (7%) had antineutrophil cytoplasmic antibody-associated vasculitis, and 50 (7%) had other RMDs. The number of SIEs recorded was 281 in 176 patients (9.8 per 100 person-years). Predictors of SIEs included non-RTX-specific comorbidities (previous history of SIE, cancer, chronic lung disease, diabetes mellitus, and heart failure), higher corticosteroid dose, and RTX-specific factors, including low IgG (<6 gm/liter) both at baseline and during treatment, RTX-associated neutropenia, higher IgM, and longer time to RTX re-treatment, but not B cell count or depletion status. Of 110 patients with low IgG, SIE rates were higher in those with low IgG at baseline (16.4 per 100 person-years) and in those who acquired low IgG during or after RTX treatment (21.3 per 100 person-years) versus those with normal IgG (9.7 per 100 person-years). Five of 8 patients (63%) had impaired humoral response to pneumococcus and hemophilus following vaccination challenge, and only 4 of 11 patients (36%) had IgG normalized after switching biologic disease-modifying antirheumatic drugs. CONCLUSION: Immunoglobulin levels should be monitored at baseline and before each RTX cycle to identify patients at risk of SIEs. Individualized risk-benefit assessment should be undertaken in those with lower IgG as this is a consistent SIE predictor and may increase infection profiles when RTX is switched to different therapies.