Comparative Efficacy of Rebamipide, Pantoprazole, and Their Combination for Gastroprotection in Patients with Atrial Fibrillation Receiving Direct Oral Anticoagulants: The Randomized REGATA Study.

Abstract

Background: Patients with atrial fibrillation (AF) receiving direct oral anticoagulants (DOAC) are at an elevated risk of gastrointestinal (GI) mucosal damage and bleeding. Proton pump inhibitors (PPI) are conventional gastroprotectors, but their efficacy in the lower GI tract is limited. This study evaluated the efficacy and safety of rebamipide monotherapy and combination pantoprazole-rebamipide (P + R) therapy compared to pantoprazole monotherapy in this vulnerable population. Methods: In the randomized, parallel-group, open-label, single-center REGATA trial, 210 AF patients receiving DOAC were randomized (1:1:1) to three gastroprotective regimens (n = 70 each) for 24 weeks: pantoprazole monotherapy, rebamipide monotherapy, or their combination. Efficacy analyses were performed in the per-protocol population (n = 118: pantoprazole, n = 36; rebamipide, n = 39; P + R, n = 43). The primary and secondary outcomes included composite clinical GI events (bleeding classified by ISTH/BARC, ulcers/erosions, and severe dyspepsia) and biomarkers of mucosal barrier function (fecal zonulin and calprotectin). Results: The incidence of primary efficacy endpoint events was 13.9% (n = 5; 95% CI: 3.1-25.7) in the pantoprazole group, 10.3% (n = 4; 95% CI: 2.3-21.1) in the rebamipide group, and 9.3% (n = 4; 95% CI: 2.1-18.6) in the P + R group, successfully confirming the non-inferiority hypothesis for both experimental arms within the pre-specified margin of 10% (p < 0.05). Multivariable regression analysis showed that compared with pantoprazole monotherapy, rebamipide monotherapy demonstrated an odds ratio OR of 0.71 (95% CI: 0.18-2.88, p = 0.630), while the P + R therapy was associated with an OR of 0.64 (95% CI: 0.16-2.57, p = 0.525). No major GI bleeding events or fatalities occurred; a single episode of GI bleeding (ISTH minor/BARC Type 2) was recorded in the pantoprazole arm (2.8%). Fecal calprotectin levels significantly decreased in the rebamipide group (by 37.6%, from 139 to 78 µg/g; p < 0.001) and the P + R group (by 50.0%, from 135 to 61 µg/g; p < 0.001), while no significant changes were observed in the pantoprazole group (p = 0.530). Concurrently, fecal zonulin levels, reflecting intestinal epithelial permeability, fell by 45.1% in the rebamipide arm (from 434 to 225 ng/mL; p < 0.001) and by 47.1% in the P + R arm (from 347 to 182 ng/mL; p < 0.001) compared to a minimal 3.0% fluctuation in the pantoprazole group. Conclusions: While demonstrating non-inferiority regarding primary clinical endpoints, both rebamipide monotherapy and combination (P + R) therapy achieved some advantages over pantoprazole monotherapy in decreasing subclinical GI inflammation and restoring intestinal epithelial barrier integrity.