Retrospective Validation of Clinical Decision Support Tools for Predicting Effectiveness Outcomes in Inflammatory...

Background/Objectives: Two clinical decision support tools (CDSTs) have been developed to predict treatment effectiveness of vedolizumab in Crohn's disease (CD) and ulcerative colitis (UC). This study aimed to validate the CDSTs with real-world data from a Dutch teaching hospital. Methods: Patients with CD or UC treated with vedolizumab between October 2014 and July 2023 were included. IBD patients treated with ustekinumab were included to study the specificity of the CDSTs. The primary outcomes were rates of clinical remission (CREM), biochemical remission (BioREM), composite (either clinical or biochemical) remission (CompREM) and corticosteroid-free clinical remission (CSF-CREM) at week 14, 30 and 54. Results: 32 CD patients and 41 UC patients treated with vedolizumab were included, as well as 28 CD patients treated with ustekinumab and 41 UC patients treated with vedolizumab. Among UC patients treated with vedolizumab, rates of CREM, CompREM and CSF-CREM at week 54 were statistically significantly higher in the group with high probability of response compared to the group with low + intermediate probability of response (CREM 33.3% vs. 81.8% p = 0.003; CompREM 33.3% vs. 86.4% p < 0.001; CSF-CREM 22.2% vs. 77.3% p = 0.001). For both CD patients treated with vedolizumab and CD patients treated with ustekinumab, no statistically significant differences were found between low + intermediate and high probability groups for all outcomes. Conclusions: The CDST for UC is able to predict various effectiveness outcomes for treatment with vedolizumab and can therefore help in the selection of optimal treatment for patients with UC. For CD patients treated with vedolizumab, the CDST could not predict effectiveness outcomes.