Paradoxical reactions and biologic agents: a French cohort study of 9303 patients.

BACKGROUND: Paradoxical reactions (PRs) are defined as the occurrence during biologic therapy of a pathological condition that usually responds to these drugs. OBJECTIVES: To estimate the incidence of PRs and identify risk factors. METHODS: Multicentre study of the database for the Greater Paris University Hospitals, including biologic-naive patients receiving anti-tumour necrosis factor-α, anti-interleukin-12/23, anti-interleukin-17 or anti-α4ß7-integrin agents for psoriasis, inflammatory rheumatism or inflammatory bowel disease (IBD). We used natural language processing algorithms to extract data. A cohort and a case-control study nested in the cohort with controls selected by incidence density sampling was used to identify risk factors. RESULTS: Most of the 9303 included patients (median age 43·0, 53·8% women) presented an IBD (3773, 40·6%) or a chronic inflammatory rheumatic disease (3708, 39·9%), and 8489 (91·3%) received anti-TNF-α agents. A total of 297 (3·2%) had a PR. The global incidence rate was 7·6 per 1000 person-years [95% confidence interval (CI) 6·8-8·5]. The likelihood of PR was associated with IBD [adjusted odds ratio (aOR) 1·9, 95% CI 1·1-3·2, P = 0·021] and a combination of at least two inflammatory diseases (aOR 6·1, 95% CI 3·6-10·6, P < 0·001) and was reduced with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) and corticosteroids (aOR 0·6, 95% CI 0·4-0·8, P = 0·002; and OR 0·4, 95% CI 0·2-0·6, P = 0·002, respectively). CONCLUSIONS: The likelihood of PRs was associated with IBD or a combination of a least two inflammatory diseases. More studies are needed to assess the benefit of systematically adding csDMARDs for such high-risk patients. What is already known about this topic? Most published studies about paradoxical reactions concern paradoxical psoriasis in patients receiving anti-tumour necrosis factor-α agents. Few data are available for other paradoxical reactions and the most recent biologics. What does this study add? Risk of paradoxical reactions was increased with inflammatory bowel disease and a combination of at least two inflammatory diseases. Risk of paradoxical reactions was reduced with conventional synthetic disease-modifying antirheumatic drugs or corticosteroid therapy, which could be added for high-risk patients.