P647 Tuberculosis in patients treated with vedolizumab: clinical trial and post-marketing case series

Abstract

Background: Treatment with anti‐tumour necrosis factor‐alpha (anti‐TNFa) agents has been shown to increase the risk of reactivating latent infections, especially tuberculosis (TB) [1]. Vedolizumab (VDZ) is a humanised monoclonal antibody that targets 0(4‐7 inte‐grin and selectively blocks gut‐specific lymphocyte trafficking. This gut selectivity may be associated with a lower risk of reactivating latent infections compared with anti‐TNFa agents, which cause systemic immunosuppression. Here we describe the frequency of TB with VDZ therapy in clinical trials and the post‐marketing (PM) setting. Methods: Safety data from the GEMINI 1 and 2 studies (VDZ vs placebo, in ulcerative colitis [UC] and Crohn's disease [CD], re‐spectively), the ongoing GEMINI open‐label extension (OLE) study (VDZ only, both UC and CD; data cut‐off: 21 May 2015) and PM data from the VDZ Global Safety Database (May 2014‐31 August 2016) were assessed. TB infections were classified according to Med‐DRA. Results: In GEMINI 1, 2 and OLE, 6 TB events were reported in 5 patients (serious: n=4; non‐serious: n=1; Table 1), with 4 TB events considered treatment‐related. Patients with TB were in the Czech Republic (n=1), India (n=1), Republic of Korea (n=1), Russian Federation (n=1) and Slovakia (n=1). All TB events resulted in treatment discontinuation, as per study protocol. In the context of 66,390 patient‐years of VDZ therapy in the PM setting, 5 patients (Table 2) reported TB (serious: n=4; non‐serious: n=1): 2 patients perma‐[Table Presented] nently discontinued treatment, 2 discontinued and later restarted, and 1 was not reported (discontinuation recommended in product label). Of the 5 patients in the PM setting, 2 had previously received anti‐TNFα agents (infliximab: n=1; adalimumab: n=1). The PM TB events occurred in the US (n=3), France (n=1) and Germany (n=1). Conclusions: The frequency of TB in VDZ clinical trials and the PM setting was low. Clinical trial events occurred in countries where TB incidence is higher than in the US, while the origin of PM reports is likely to be explained by the greater exposure in those countries. Limitations associated with PM safety reporting (e.g. incomplete co‐morbidity and co‐medication data), which make confirming a causal association between drug and event difficult, must be considered when interpreting the PM results.