FENOX Trial (Comparative Effectiveness of Fexuprazan Co-therapy in Patients Receiving Non-Vitamin K Antagonist Oral Anticoagulants)

Background Non-vitamin K antagonist oral anticoagulants (NOACs) are recommended for stroke prevention in non-valvular atrial fibrillation (AF). Although NOACs substantially reduce intracranial hemorrhage, upper gastrointestinal bleeding (UGIB) remains a frequent and clinically consequential complication. Proton pump inhibitors (PPIs) may reduce UGIB risk; however, concerns regarding long-term safety and pharmacodynamic variability persist. Fexuprazan, a potassium-competitive acid blocker (P-CAB), provides rapid and sustained acid suppression independent of acid activation and CYP2C19 metabolism. No randomized trial has evaluated P-CAB therapy for prevention of UGIB in anticoagulated patients. Methods FENOX is a multicenter, prospective, randomized, open-label, blinded-endpoint (PROBE) superiority trial. Approximately 1,000 high-risk patients with non-valvular AF initiating NOAC therapy will be randomized 1:1 to receive fexuprazan plus NOAC therapy or NOAC therapy alone. High-risk enrichment includes advanced age, renal impairment, concomitant antiplatelet therapy, prior ulcer disease, or elevated HAS-BLED score. The primary endpoint is clinically relevant upper gastrointestinal bleeding (CR-UGIB) at 12 months, defined according to ISTH criteria. All events will be adjudicated by an independent blinded Clinical Events Committee. Primary analyses will follow the intention-to-treat principle using time-to-event methods. Results The planned sample size provides 80% power to detect a 50% relative risk reduction in CR-UGIB, assuming a 12-month incidence of 10% in the control group. Interim safety monitoring will be conducted under independent oversight. Conclusion FENOX is the first randomized trial designed to evaluate a P-CAB-based gastroprotective strategy for prevention of clinically relevant UGIB in high-risk patients receiving NOAC therapy. By integrating high-risk enrichment, pragmatic design, and blinded endpoint adjudication, the study aims to provide rigorous