Objective: To evaluate Pneumocystis jirovecii pneumonia (PJP) infection risk in patients with systemic lupus erythematosus (SLE) in Taiwan. Methods: We identified 24,367 patients with SLE from the National Health Insurance research database between 1997 and 2012 and compared the PJP incidence rates (IRs) with those in 243,670 age- and sex-matched non-SLE controls. PJP risk in the patients was evaluated using a Cox multivariate proportional hazards model. Results: The SLE patients exhibited a significantly higher PJP risk than the controls, with an IR of 2.63 per 10,000 person-years and IR ratio of 27.65 (95% confidence interval 17.2–45.3; P < 0.001). Male sex (hazard ratio [HR] 2.42, P < 0.01), end-stage renal disease (ESRD; HR 1.74, P = 0.01), recent use of mycofenolate mofetil (MMF; HR 4.43, P < 0.001), intravenous steroid pulse therapy (HR 108.73, P < 0.001), and average oral dosage of >7.5 mg/day prednisolone or equivalent treatment (HR 4.83, P < 0.001) were associated with PJP in SLE, whereas hydroxychloroquine use reduced its risk (HR 0.51, P = 0.01). Of note, cyclophosphamide was not associated with PJP infection in the multivariate Cox proportional hazard model. Conclusion: Patients with SLE have a considerably high PJP risk. Cyclophosphamide does not increase PJP risk. Male sex, ESRD, MMF use, intravenous steroid pulse therapy, and oral prednisolone or equivalent treatment (>7.5 mg/day) are risk factors for PJP, whereas hydroxychloroquine use reduces PJP risk.
ASJC Scopus subject areas