Abstract Background The clinical decision support tool (CDST) was developed from GEMINI II to help inform patient selection for intravenous (IV) vedolizumab (VDZ) in Crohn’s disease (CD). The tool’s applicability in the subcutaneous (SC) formulation of VDZ was tested post-hoc in VISIBLE-2 and also appears valid. Our aim was to examine its predictive performance in a real-world cohort of patients treated with either IV or SC VDZ. Methods We retrospectively collected data on all CD patients commenced on VDZ from January 2015 to October 2023 in two tertiary IBD centres. Baseline data was retrieved to calculate the CDST: prior exposure to tumour necrosis factor (TNF) antagonists, prior surgery for CD, prior fistulating disease, baseline albumin and C reactive protein (CRP). All patients with a minimum of 12 months follow up were included. Patients were categorised into the SC group if they switched to SC within 6 months of commencing VDZ. Our primary outcome was treatment persistence at 12 months. In addition, we looked at clinical, biochemical and endoscopic outcomes at 12 months, requirement for dose escalation and progression to CD-related surgery. Kaplan Meier analysis with logrank test for trend was used for treatment persistence. Categorial outcomes and continuous variables were compared using the chi square test for trend and ANOVA respectively. Results 408 patients were included, 278 (68%) received IV VDZ and 130 (32%) SC. 336 (82%) had prior exposure to TNF antagonists, 107 (26%) had fistulating disease, 191 (47%) had prior resectional surgery for CD. The CDST stratified patients into 3 groups by probability of treatment response: low (n=35, 9%), intermediate (n=152, 37%) and high (n=221, 54%). The low and intermediate groups were combined in the analysis of treatment persistence due to lower numbers in the former. In both IV and SC cohorts, CDST was able to predict greater treatment persistence in the high probability group versus the low/intermediate group. The high probability cohort was also statistically more likely to achieve mucosal healing (p<0.0001) and surgery-free survival (p<0.0001) at 12 months. Conclusion This is to our knowledge the first application of CDST to a real-world cohort of CD patients receiving either IV and/or SC VDZ. Our data demonstrates that the tool has predictive ability for long term treatment persistence regardless of the method of administration, as well as outcomes at 12 months. Our study supports a role for CDST in guiding treatment selection in clinical practice.
Read full abstract