P320 Confocal Laser Endomicroscopy to predict the efficacy of vedolizumab as first-line biotherapy in ulcerative colitis

Abstract Background There is no biomarker to guide first-line of biotherapy in ulcerative colitis (UC). The main objective this multicenter pilot study was demonstrate the feasibility identifying a4b7 and/or TNF-a expressing cells by dual-band confocal endomicroscopy (CLE) into mucosa predict response vedolizumab as biotherapy. NCT02878083. Methods Patients with moderate severe UC, naïve were prospectively included and received vedolizumab. evaluated clinically, biologically, endoscopically histologically at week 22. Nonresponding patients subsequently treated adalimumab 30. Prior initiation vedolizumab, during endoscopy, 8 biopsies taken for extemporaneous ex vivo analysis CLE, histology microscopy. Fresh coupled FITC Alexa-Fluor analyze their respective fluorescence (figure) one step Dual-band CLE system (Mauna Kea Technologies, France). number surface positive stained areas visualized counted compared between responders non-responders. Blood samples collected before each infusion used immune (LT, LB, NK, monocytes) FACS trough levels. Frozen analyzed microscopy (Opal Akoya multiplexing; Tyramide signal amplification) after labeling an anti-a4b7 antibody. Results 19 out 20 initially planned included. Clinical remission, endoscopic improvement (Mayo ≤ 1), remission 0) W22 respectively 58%, 58% 45%. 18 (95%) could be CLE. vedolizumab-FITC numerically superior responder week22; contrast, difference found marked adalimumab-Alexa-fluor. A threshold value 6 0 discriminated non-responders 22 a Se 78% Sp 85%. No circulating cell population has been identified levels not predictive response. In microscopy, 15 have analyzed; percentage a4b7-binding different (n=9) (n=6) Conclusion This confirms identification mucosal TNF-a. suggest that only proportion binding indicative