TNF-alpha inhibitor treatment patterns in patients with rheumatic diseases and those with inflammatory bowel disease

Background: TNFα inhibitors are the first line biological treatment for patients suffering from rheumatic diseases (RD) and inflammatory bowel disease (IBD). Limited information is available about long-term treatment patterns of patients starting a TNFα inhibitor and whether these differ between patients suffering from RD and IBD. Objectives: To compare treatment patterns of patients with RD and IBD starting a TNFα inhibitor. Methods: Included were all patients starting (i.e. no prior use of a biological) with aTNFα inhibitor (ATC code: L04AB) between 1 July 2012 and 1 July 2017 at a Dutch general teaching hospital (the Spaarne Gasthuis, Haarlem/Hoofddorp) and with a RD or IBD diagnosis. All patients were followed for at least one year. Outcomes at one year of follow-up were: continuous use of the first TNFα inhibitor, switch to a different TNFα inhibitor or to a biological with another mode of action, or discontinuation. In addition, median duration of first TNFα inhibitor treatment were compared for patients with RD and IBD using the Kaplan Meier method. Data were analyzed by Pearson's chi square and Kruskal Wallis test. Results: 646 patients were included (median age 46 years, 84% female), of which 63.9% (n = 413) received a TNFα inhibitor for RD and 36.1% (n = 233) for IBD. After 1 year, 60.1% of patients continuously used their first TNFα inhibitor, 13.4% switched to another biological and 26.5% discontinued treatment. Significantly less RD patients continued their TNFα inhibitor compared to IBD patients (54.4% versus 70.0%, RR 0.78, 95% CI 0.69-0.88) and RD patients discontinued treatment more frequently than IBD patients (33.0% versus 15.0%, RR 2.19, 95% CI 1.57-3.06). 12.6% of RD patients and 15.0% of IBD patients had switched, most patients (71.1% of RD switchers and 91.4% of IBD switchers) to a second TNFα inhibitor. The median treatment duration of the first TNFα inhibitor was significantly (p <0.01) lower for RD patients (437 days, IQR 686 days) when compared with IBD patients (728 days, IQR 988 days). Conclusions: RD patients discontinue their first TNFα inhibitor significantly more often than IBD patients and have a shorter duration of treatment, patterns of switching are equal in both indications. These findings show the importance of underlying disease in classification of exposure to TNFα inhibitors and this should be taken into account in future pharmacoepidemiologic studies.