P0937 Role of Interleukin-8 in Predicting Long-Term Response to Vedolizumab in Ulcerative Colitis Patients

Abstract

Listen

Translate article

Abstract Background Vedolizumab is effective for moderate-severe Ulcerative Colitis (UC), but predictive markers for long-term response are limited. This study aimed to profile circulating cytokines linked to inflammatory bowel disease (IBD) pathophysiology to identify immune predictive markers of vedolizumab response in UC patients. Secondary objectives included analyzing associations between baseline characteristics—such as smoking, disease duration, prior biologic exposure, and initial faecal calprotectin (FCP) and C-reactive protein (CRP) levels—and treatment response. Methods This retrospective, single-centre cohort study included 28 UC patients with inflammation objectified by clinical and biochemical data who initiated vedolizumab from 2015 to 2022. Blood samples were collected before treatment, and responses were evaluated at follow-up. Clinical remission (pMAYO < 3 with no subscore >1) with biochemical response (≥50% reduction in FCP and/or CRP) defined positive response. Endoscopic response was assessed in patients with available colonoscopy (endoscopic Mayo 0-1). Cytokines were measured via Magpix® Luminex with a Milliplex® T Cell Magnetic Bead Panel including 13 cytokines (IFN-gamma, IL8, IL12, IL13, IL17A, IL10, IL2, IL21, IL23, IL7, TNF-alpha, IL6, IL-1β). Results Twenty-eight patients with UC (100% pancolitis or left colitis) and S2+S3 94.7% were included in the study. 85.2% werebioexposed (infliximab in 75%). 50% of patients received concomitant corticosteroid therapy at the start of vedolizumab. Vedolizumab response was seen in 13 patients (46.4%) after a median overall treatment duration in both cohorts of 2.79 years (IQR7.55). The responder group was on vedolizumab treatment for a median of 5.58 (IQR 6.9) years vs 1.17 (IQR 5.13) years in the nonrespondergroup, p=0.005. There were no baseline differences between responders in pMayo or in FCP and CRP values. Activesmoking, disease duration until vedolizumab initiation and previous exposure to other biologics were not associated withresponse/non-response to vedolizumab treatment. Vedolizumab responders had significantly lower median interleukin 8 (IL8) levels at baseline than vedolizumab non-responders(median IL8 1.67 pg/mL versus 3.03 pg/mL, p=0.016). No statistically significant differences in baseline concentrations were found inany of the other interleukins analysed. Conclusion In our series of UC patients treated with vedolizumab, patients with long-term response to the drug show lower plasma IL8 levels than non-responders at baseline. Therefore, IL8 could serve as a predictive marker of long-term response to vedolizumab inUC patients. These results should be confirmed in prospective studies with larger sample sizes.