Objective: Our objective was to describe the incidence and risk factors of legionellosis associated with tumor necrosis factor (TNF)-α antagonist use.
Methods: From February 1, 2004, to January 31, 2007, we prospectively collected all cases of legionellosis among French patients receiving TNF-α antagonists in the Research Axed on Tolerance of Biotherapies (RATIO) national registry. We conducted an incidence study with the French population as a reference and a case-control analysis with four control subjects receiving TNF-α antagonists per case of legionellosis.
Results: Twenty-seven cases of legionellosis were reported. The overall annual incidence rate of legionellosis for patients receiving TNF-α antagonists, adjusted for age and sex, was 46.7 (95% CI, 0.0-125.7) per 100,000 patient-years. The overall standardized incidence ratio (SIR) was 13.1 (95% CI, 9.0-19.1; P < .0001) and was higher for patients receiving infliximab (SIR, 15.3 [95% CI, 8.5-27.6; P < .0001]) or adalimumab (SIR, 37.7 [95% CI, 21.9-64.9; P < .0001]) than etanercept (SIR, 3.0 [95% CI, 1.00-9.2; P = .06]). In the case-control analysis, exposure to adalimumab (OR, 8.7 [95% CI, 2.1-35.1]) or infliximab (OR, 9.2 [95% CI, 1.9-45.4]) vs etanercept was an independent risk factor for legionellosis.
Conclusions: The incidence rate of legionellosis for patients receiving TNF-α antagonists is high, and the risk is higher for patients receiving anti-TNF-α monoclonal antibodies than soluble TNF-receptor therapy. In case of pneumonia occurring during TNF-α antagonist therapy, specific urine antigen detection should be performed and antibiotic therapy should cover legionellosis.
Trial registry: ClinicalTrials.gov; No.: NCT00224562; URL: www.clinicaltrials.gov.