Published in:
Open Access 01-12-2019 | Glucocorticoid | Research article
Pneumocystis pneumonia in patients with rheumatic diseases receiving prolonged, non-high-dose steroids—clinical implication of primary prophylaxis using trimethoprim–sulfamethoxazole
Published in: Arthritis Research & Therapy | Issue 1/2019
Login to get accessAbstract
Objectives
To investigate the incidence of pneumocystis pneumonia (PCP) and its risk factors in patients with rheumatic disease receiving non-high-dose steroid treatment, along with the risks and benefits of PCP prophylaxis.
Methods
This study included 28,292 treatment episodes with prolonged (≥ 4 weeks), non-high-dose steroids (low dose [< 15 mg/day, n = 27,227] and medium dose [≥ 15 to < 30 mg/day, n = 1065], based on prednisone) over a 14-year period. Risk factors for PCP and prophylactic effect of trimethoprim–sulfamethoxazole (TMP-SMX) were investigated if the 1-year incidence rate (IR) of PCP in each dose group was > 0.1/100 person-years. Cox regression with LASSO was used for analysis.
Results
One-year PCP IR in the low-dose group was 0.01 (95% CI 0.001–0.03)/100 person-years, and only the medium-dose group showed eligible PCP IR for further analysis. In the medium-dose group, prophylactic TMP-SMX was administered in 45 treatment episodes while other episodes involved no prophylaxis (prophylaxis group vs. control group). In 1018.0 person-years, 5 PCP cases occurred exclusively in the control group, yielding an IR of 0.5 (0.2–1.2)/100 person-years. Concomitant steroid-pulse treatment and baseline lymphopenia were the most significant risk factors for PCP. Treatment episodes with at least one of these factors (n = 173, high-risk subgroup) showed higher 1-year PCP IR (3.4 (1.1–8.0)/100 person-years), while no PCP occurred in other treatment episodes. TMP-SMX numerically reduced the risk (adjusted HR = 0.2 (0.001–2.3)) in the high-risk subgroup. The IR of adverse drug reactions (ADRs) related to TMP-SMX was 41.5 (22.3–71.6)/100 person-years, including one serious ADR. The number needed to treat with TMP-SMX to prevent one PCP in the high-risk subgroup (31 (17–226)) was lower than the number needed to harm by serious ADR (45 (15–∞)).
Conclusion
Incidence of PCP in patients with rheumatic diseases receiving prolonged, medium-dose steroids depends on the presence of risk factors. Prophylactic TMP-SMX may have greater benefit than potential risk in the high-risk subgroup.