Mucosal α4β7+ Lymphocytes and MAdCAM+ Venules Predict Response to Vedolizumab in Ulcerative Colitis

Abstract

Abstract Background Therapeutic strategies for patients with ulcerative colitis (UC) are based on patient- and disease-related factors in combination with drug characteristics but fail to predict success in individual patients. A considerable proportion of UC patients do not respond to the biological vedolizumab. Therefore, pretreatment biomarkers for therapeutic efficacy are urgently needed. Mucosal markers related to the integrin-dependent T lymphocyte homing could be potent predictors. Methods We prospectively included 21 biological- and steroid-naive UC patients with moderate-to-severe disease activity planned to escalate therapy to vedolizumab. At week 0, before initiating treatment, colonic biopsy specimens were obtained for immunophenotyping and immunohistochemistry. Clinical and endoscopic disease activity were determined at week 16 after 4 infusions of vedolizumab. In addition, we retrospectively included 5 UC patients who were first treated with anti-tumor necrosis factor α before receiving vedolizumab to compare with biological-naive patients. Results Abundance of α4β7 on more than 8% of all CD3+ T lymphocytes in colonic biopsies at baseline was predictive for responsiveness to vedolizumab (sensitivity 100%, specificity 100%). The threshold for the proportion of MAdCAM-1+ and PNAd+ of all venules in the biopsies predictive for responsiveness to vedolizumab was ≥2.59% (sensitivity 89%, specificity 100%) and ≥2.41% (sensitivity 61%, specificity 50%), respectively. At week 16, a significant decrease of α4β7+CD3+T lymphocytes was demonstrated in responders (18% [12%-24%] to 8% [3%-9%]; P = .002), while no difference was seen in nonresponders (4% [3%-6%] to 3%; P = .59). Conclusions UC responders to vedolizumab have a higher percentage of α4β7+CD3+ T lymphocytes and a higher proportion of MAdCAM-1+ venules in colonic biopsies than nonresponders before initiating therapy. Both analyses could be promising predictive biomarkers for therapeutic response and may lead to more patient tailored treatment in the future.