ABSTRACT: BACKGROUND:In recent years, proton pump inhibitors (PPIs) have been investigated at high-dose to modulate tumor microenvironment acidification thus restoring chemotherapeutic sensitivity. This is the first trial to study activity and safety of repurposing high dose rabeprazole combined with metronomic capecitabine (mCAP). METHODS:A phase II study in which patients with gastrointestinal cancer, refractory to standard treatments, who had a life expectancy >3 months, were blind randomized 1:1 to mCAP, 1500 mg/daily, continuously with or without rabeprazole 1.5 mg/kg bid, three days a week. The primary endpoint was 3-months progression-free survival (PFS). The secondary endpoints were clinical benefit (CB) and overall survival (OS). Safety and plasma concentrations of capecitabine and its metabolites (5'-DFUR and 5-FU) were also evaluated. RESULTS:Sixty-seven (median age 69 years; 63% male; 84% colorectal cancer, 76% ECOG-PS ? 1; 84% pretreated with two or more lines of chemotherapy) out of 90 patients screened for eligibility, were randomized to receive mCAP+rabeprazole (n = 32) vs. mCAP (n = 35). All patients were evaluable for response. No significant difference between mCAP+rabeprazole vs. mCAP, in terms of 3-months PFS rate (HR = 1.43, 95%CI 0.53-3.85; p = 0.477), median PFS (HR = 1.22, 95%CI 0.75-2.00, p = 0.420), CB (RR = 0.85, 95%CI 0.29-2.44; p = 0.786) and median OS (HR = 0.89, 95%CI 0.54-1.48; p = 0.664) was observed. However, a 3-year OS rate of 10% and 12% was reported in the mCAP-rabeprazole and mCAP groups, respectively. Overall, no grade 3 or 4 toxicity occurred but grade 1 or 2 adverse event of any type were more frequently in the mCAP+rabeprazole group than in the mCAP (OR 2.83, 95%CI 1.03-7.79; p = 0.043). Finally, there was not statistically significant difference in the plasma concentration of capecitabine and its metabolites between the two groups. CONCLUSIONS:Although the adjunct of high dose rabeprazole to mCAP was not shown to affect mCAP activity, as PPI are being investigated worldwide as drugs to be repositioned in cancer treatment and also considering the limited sample size as well as the favorable safety profile of the combination in the present study, further clinical investigations are desirable.