OBJECTIVE The objective of this study was to investigate the effect of prophylactic trimethoprim-sulfamethoxazole (TMP-SMX) on the incidence of serious infections in patients with antineutrophil cytoplasmic antibody-associated vasculitis (AAV).

METHODS This multicenter cohort study was designed to emulate a target trial that studied 296 patients with AAV who were treated with rituximab (RTX) or cyclophosphamide (CYC) as induction therapy. Patients were grouped based on the administration of TMP-SMX within 14 days following induction therapy (n = 240 and 56 patients in prophylaxis and control groups, respectively) (intention-to-treat) and also as a time-dependent exposure (time-varying). Inverse probability weighting was applied to minimize the baseline imbalance between the two groups. The primary outcome was one-year incidence of serious infection.

RESULTS During the 252.1 person-years of observation, 77 cases of serious infections were recorded in 65 patients with a fatality rate of 18.5%. Most serious infections (n = 66, 85.7%) occurred within the first 180 days of observation. The prophylaxis group showed a significantly lower incidence of serious infections than the control group (hazard ratio [HR] 0.48 [95% confidence interval (CI) 0.32-0.72]). However, this beneficial effect of TMP-SMX was only significant during the first 180 days (HR 0.41 [95% CI 0.22-0.76]) and not thereafter (HR 3.67 [95% CI 0.46-29.43]) (interaction P = 0.044). This result was also consistent with the time-varying analysis result. Based on one case of severe adverse drug reaction related to TMP-SMX, the number needed to harm was 127.4, whereas the number needed to treat to prevent one serious infection was 8.0.

CONCLUSION Prophylactic TMP-SMX significantly reduced the risk of serious infections in patients with AAV, particularly during the first six months of induction therapy with RTX or CYC.