RESUMEN
BACKGROUND: Whether long-term effectiveness differs between anti-tumour necrosis factor (anti-TNF) agents is unknown. AIMS: To examine drug survival of first-line anti-TNF agents and identify predictors of discontinuation. To reduce channelling bias, we also compared drug survival of the second anti-TNF. METHODS: Biologic-naïve patients (N = 955) recorded in the Swedish IBD Quality Register (SWIBREG) were examined. We used propensity score matching, comparing drug survival over up to three years of follow-up. Cox regression estimated adjusted hazard ratios (aHRs) with 95% confidence intervals (CIs). RESULTS: In Crohn's disease, discontinuation because of lack/loss of response was 32% [95%CI = 26%-38%] for infliximab versus 16% [95%CI = 11%-21%] for adalimumab. Infliximab [vs adalimumab; aHR = 1.96; 95%CI = 1.20-3.21] and colonic disease (L2) [vs no L2; aHR = 2.17; 95% CI = 1.26-3.75] were associated with higher discontinuation rates, whereas normalised CRP at three months [aHR = 0.40; 95% CI = 0.19-0.81] with a lower rate. Consistently, patients who switched from adalimumab to infliximab (vs infliximab to adalimumab) had earlier discontinuation (P = 0.04). Concomitant use of immunomodulators was associated with a lower adverse drug reaction-mediated discontinuation rate [aHR = 0.46; 95% CI = 0.28-0.77], in part explained by fewer infusion reactions [aHR = 0.27; 95% CI = 0.08-0.89]. In ulcerative colitis, the probability of discontinuation because of lack/loss of response was 40% [95% CI = 33%-47%] for infliximab versus 37% [95% CI = 21%-53%] for adalimumab. Disease duration ≥10 years [aHR = 0.25; 95% CI = 0.10-0.58] and normalised CRP after three months [aHR = 0.39; 95% CI = 0.18-0.84] were associated with lower discontinuation rates. CONCLUSIONS: Clinical characterisation of patients may aid decision-making on anti-TNF treatment. The consistently shorter drug survival for infliximab (vs adalimumab) in Crohn's disease, suggests a potential difference between the two drugs.