Abstract
BACKGROUND: Infection is a major cause of mortality in patients with systemic lupus erythematosus (SLE). Therefore, minimizing the risk of infection is an important clinical goal to improve the long-term prognosis of SLE patients. Treatment with ≥7.5 mg prednisolone (PSL) or equivalent has been reported to increase the risk of infections. However, it remains unclear whether <7.5 mg PSL or equivalent dose affects the risk of infection in SLE patients. This study evaluated the association between the occurrence of infection in patients with SLE and low-dose glucocorticoid (GC) usage, especially <7.5 mg PSL or equivalent, to explore the GC dose that could reduce infection occurrence. METHODS: This prospective cohort study included patients from the Japanese multicenter registry of patients with SLE (defined as ≥4 American College of Rheumatology 1997 revised criteria) over 20 years of age. The PSL dose was categorized as PSL 0-2.5, 2.6-5.0, 5.1-7.5, and 7.6-15.0 mg. The primary outcome was infection requiring hospitalization. We conducted a multivariable analysis using time-dependent Cox regression analysis to assess the hazard ratio of infection occurrence compared with a dose of 0-2.5 mg PSL or equivalent in the other three PSL dose groups. Based on previous reports and clinical importance, the covariates selected were age, sex, and concurrent use of immunosuppressants with GC. In addition, two sensitivity analyses were conducted. RESULTS: The mean age of the 509 SLE patients was 46.7 years; 89.0% were female, and 77.2% used multiple immunosuppressants concomitantly. During the observation period, 52 infections requiring hospitalization occurred. The incidence of infection with a PSL dose of 5.0-7.5 mg was significantly higher than that in the PSL 0-2.5 mg group (adjusted hazard ratio: 6.80, 95% confidence interval: 2.17-21.27). The results of the two sensitivity analyses were similar. CONCLUSIONS: Our results suggested that the use of 5.0-7.5 mg PSL or equivalent could pose an infection risk in SLE patients. This finding indicates that PSL dose should be reduced to as low as possible in SLE patients to avoid infection.