P603 Vedolizumab trough levels at Week 6 predict endoscopic and clinical remission in inflammatory bowel disease

Abstract

BackgroundVedolizumab (VDZ) is an anti-integrin monoclonal antibody for the treatment of inflammatory bowel disease. There appears to be a positive correlation between drug levels and clinical outcomes, however, data on endoscopic outcomes are scarce. We investigated the impact of induction VDZ trough levels (TL) on clinical and endoscopic outcomes in our real-world cohort.MethodsWe prospectively identified all patients treated with VDZ in a tertiary referral hospital: 50 patients (25 Crohn’s disease [CD], 25 ulcerative colitis [UC]), mostly resistant to anti-TNF agents with a mean disease duration of 10 years (SD 7.8). VDZ TL were measured at Weeks 2, 6, 10, and 14 using an in-house developed ELISA assay (KU Leuven, Belgium). The main clinical outcome was a composite of partial (need for dose optimisation to q4 weeks) or complete drug failure (withdrawal of drug due to inefficacy). Thirty patients (13 CD, 17 UC), including all treatment failures, underwent endoscopy between Weeks 30 and 52 to assess for mucosal healing (modified Rutgeerts score <4; Mayo endoscopic subscore 0/1), evaluated centrally by a blinded endoscopist. Potential independent predictors of clinical (CRP, albumin, disease phenotype, steroid use) and endoscopic outcomes were subjected to univariate analysis and a Cox or logistic regression model respectively. Predictive cut-off values were identified using ROC curve analysis, categoric variables were compared with the χ2 test.ResultsAfter a median follow-up of 44 weeks (IQR 30) 23 of 50 patients (46%) discontinued treatment or required dose optimisation. The only independent significant predictor were TL at Week 6 (HR 0.54 per 10 µg/ml increase in TL, 95% CI 0.35–0.90; p = 0.02). Patients with TL below the median (18 µg/ml) were more likely to experience drug failure (60% vs. 32%; p = 0.04) (Figure 1). Mucosal healing was achieved by 8 of 30 (27%) patients, again TL at Week 6 were a significant predictor (HR 2.84 per 10 µg/ml increase in TL, 95% CI 1.10–6.19; p = 0.04). Patients with TL above the median (18 µg/ml) were more likely to have mucosal healing (47% vs. 7%; p = 0.01) (Figure 1). A Week 6 TL below 28.2 µg/ml predicted complete or partial drug failure with 81% sensitivity and 70% specificity (AUC = 0.70; p = 0.03), while a Week 6 TL above 33.6 µg/ml predicted mucosal healing with 25% sensitivity and 94% specificity (AUC = 0.80; p = 0.02).View largeDownload slideImpact of Week 6 vedolizumab trough levels on clinical and endoscopic outcomes. Clinical failure–drug discontinuation or need for optimisation; mucosal healing–modified Rutgeerts below i2b, SES-CD below 4, Mayo endoscopic subscore 0 or 1.View largeDownload slideImpact of Week 6 vedolizumab trough levels on clinical and endoscopic outcomes. Clinical failure–drug discontinuation or need for optimisation; mucosal healing–modified Rutgeerts below i2b, SES-CD below 4, Mayo endoscopic subscore 0 or 1.ConclusionsVedolizumab trough levels after 6 weeks of treatment independently predicted drug failure and endoscopic remission, highlighting their possible role in future treatment algorithms.