King's College London

Objectives: To establish whether the decision to stop, continue or switch TNF inhibitor (TNFi) therapy to a biologic drug with an alternative mode of action following a serious infection (SI) impacts upon the risk of recurrent SI in patients with RA. Methods: Patients recruited to the British Society for Rheumatology Biologics Register-RA with at least one episode of SI while on TNFi were included. The biologic treatment decision following SI was considered. A multivariable adjusted Cox proportional hazards model was used to identify predictors of recurrent SI and whether biologic treatment choices influenced future SI risk. Results: In total, 1583 patients suffered at least one SI while on TNFi. Most patients (73%) were recorded as continuing TNFi 60 days after an index SI. The rate of recurrent SI was 25.6% per annum (95% CI: 22.5, 29.2%). The rate of recurrent SI was highest in patients who stopped their TNFi (42.6% per annum, 95% CI: 32.5, 55.7%) and lowest in those who switched biologic drug class (12.1% per annum, 95% CI: 3.9, 37.4%). Compared with patients stopping biologic therapy, patients who continued or switched drug class had significantly lower risk of recurrent SI (drug continuation hazard ratio = 0.54, 95% CI: 0.40, 0.74; drug switch hazard ratio = 0.29, 95% CI: 0.09, 0.95). Conclusions: Patients who continued or switched their TNFi post-index SI had a lower risk of recurrent SI infection compared with those who stopped the drug. This may be explained by better control of disease activity with reintroduction of biologic therapy, a driving factor for SI or alternatively channelling fitter patients to restart biologic therapy.