OA18 Smoking and high BMI are associated with reductions in TNF inhibitor response in psoriatic arthritis: results from 12 European countries

Abstract Background/Aims In axial spondyloarthritis, it has been shown that while some adverse lifestyle factors (smoking and high body mass index (BMI)) are less likely to respond to TNF inhibition (TNFi), those who drink alcohol are more likely to respond to treatment - a result currently unexplained, and unlikely to be causal. We aimed to investigate these associations in psoriatic arthritis (PsA) and to quantify the impact of lifestyle factors on TNFi treatment response. Methods Adults with PsA commencing TNFi were identified via the European Spondyloarthritis Research Collaboration Network (https://eurospa.eu). Participants were required to have data on lifestyle factors (smoking, alcohol consumption, and/or BMI) recorded within ±30 days of commencing TNFi. The relationship between lifestyle factors and treatment outcome (ACR-50 response criteria at 12 months) was assessed using logistic regression, adjusted for age, sex, country, year of TNFi initiation, disease duration and baseline disease activity (DAS28). The analysis was then repeated for other treatment response criteria: ACR-20, ACR-70, DAPSA28-remission, DAPSA68-remission and Minimal Disease Activity. Results From 12 PsA registries, 9409 participants were included, 21% of whom were current smokers, 54% drank alcohol, 38% were overweight and 29% were obese. Other baseline characteristics are shown in Table 1. Among smokers, 18% met ACR-50 response criteria at 12 months, versus 26% of non-smokers, an association that remained after adjusting for potential confounders (adjusted odds ratio: 0.59; 95%CI: 0.48-0.72). Participants who were overweight (adjusted odds ratio: 0.80; 0.64-0.99) or obese (0.57; 0.44-0.72) were also less likely to achieve treatment response, compared to patients of normal weight. Similar results were seen across other outcome measures. The proportion who achieved ACR-50 response did not differ between patients who did/did not consume alcohol (both 22%). Conclusion Smoking and high BMI were associated with substantial decreases in the likelihood of TNFi response. The addition of non-pharmacological therapy, including the modification of adverse lifestyle factors, may offer the potential to enhance treatment outcomes. For example, if 100 patients were able to give up smoking when commencing TNFi, an additional eight would achieve ACR-50 at 12 months. Previous findings regarding alcohol and treatment response in axSpA were not replicated in PsA. Disclosure G.T. Jones: None. O.