To Örebro University

Background: The real-world comparative safety of vedolizumab in inflammatory bowel disease (IBD) remains uncertain. We aimed to assess the risk of serious infection in IBD patients treated with vedolizumab, compared to (i) those treated with anti-tumour necrosis factor (TNF) treatment and (ii) the general population.

Methods: In this nationwide cohort study, treatment episodes were identified from Swedish health registers (from 1 May 2014 – 31 December 2020). Patients were considered exposed from initiation of treatment until 90 days after discontinuation of treatment. We used Cox regression with propensity score-matched cohorts to estimate hazard ratios (HRs) for incident serious infection, defined as infection requiring hospital admission.

Results: After propensity score matching, the cohorts were not materially different at baseline with regard to demographic, disease and treatment characteristics (Table 1). During 1376 treatment-episodes in patients with Crohn’s disease, there were 5.18 (95%CI: 3.98-6.63) serious infections per 100 person-years (PY) with vedolizumab vs 3.54 (95%CI: 2.50-4.85) per 100 PY with anti-TNF; HR 1.72 (95%CI: 1.12-2.65; Figure 1A). When examining site-specific infections in Crohn’s disease, vedolizumab was associated with an HR of 2.47 (95% CI: 0.96-6.39) for serious gastrointestinal infections. Compared to the rate of 0.75 (95%CI: 0.59-0.92) serious infections per 100 PY in the general population, vedolizumab demonstrated an increased HR of 7.00 (95%CI: 5.04-9.72).

Across 1294 episodes among patients with ulcerative colitis there were 3.74 (95%CI: 2.66-5.11) serious infections per 100 PY with vedolizumab vs 3.42 (95%CI: 2.31-4.89) per 100 PY with anti-TNF, corresponding to HRs of 0.80 (95%CI: 0.47-1.36, Figure 1B) within the initial 1.1 years of treatment and 2.03 (95%CI: 0.65-6.32) after 1.1 years (follow-up truncated due to non-proportional hazards). In ulcerative colitis, there was no statistically significant association between vedolizumab treatment and any of the site-specific serious infections. Compared to the rate of 0.69 (95%CI: 0.53-0.87) serious infections per 100 PY in the general population, vedolizumab showed an increased HR of 5.45 (95%CI: 3.67-8.11).

Conclusion: Vedolizumab was associated with higher hazard ratios of serious infections compared to anti-TNF in Crohn’s disease, but not in ulcerative colitis. Nonetheless, in both IBD subtypes vedolizumab exhibited increased hazard ratios compared to the general population. These results underscore the importance of heightened clinical awareness of infections in vedolizumab-treated patients and may help clinicians understanding the optimal positioning of vedolizumab.