5-(Benzyloxybenzyl)barbituric acid acyclonucleoside (BBBA) was recently synthesized as a potent and specific inhibitor of uridine Phosphorylase (EC 2.4.2 3), the enzyme responsible for the catabolism of 5-fluoro-2′-deoxyuridine (FdUrd) in many types of tumors that are deficient or have little thymidine Phosphorylase (EC 184.108.40.206) activity. The effect of BBBA on modulating the antitumor efficacy of FdUrd was evaluated in vitro, against the human colon carcinomas DLD-1 and HCT-15 grown in culture, and in vivo, against DLD-1 grown as xenografts in anti-thymocyte serum immunosuppressed mice. The concentrations of FdUrd that produced 50% growth inhibition after a 3-h exposure were 88 and 340 nM for HCT-15 and DLD-1, respectively. BBBA alone, at all concentrations tested, had no significant effect on the growth of DLD-1 and HCT-15 in culture. However, BBBA at 5,10,20, and 40 nM potentiated (P < 0.05) the cytotoxicity of FdUrd (340 nM; 3 h) against DLD-1 in culture by 20,33,55, and 63%, respectively. Similarly, BBBA at 10 and 20 nM potentiated the cytotoxicity of FdUrd (88 nM; 3 h) against HCT-15 in culture by 37 and 45%, respectively. In soft agar, BBBA (10 nM) also enhanced the cytocidal effect of FdUrd (10 and 32 nM) against DLD-1 by 41 and 55%, respectively, and against HCT-15 by 6 and 31%, respectively. Increasing BBBA dose to 20 nM enhanced further the FdUrd (10 and 32 nM) cytotoxicity against DLD-1 by 76 and 77%, respectively, and HCT-15 by 31 and 48%, respectively. BBBA also potentiated the chemotherapeutic efficacy of FdUrd in anti-thymocyte serum immunosuppressed mice bearing DLD-1 xenografts with no apparent host toxicity. At a low tumor burden (2.5 × 106 cells/mouse), 2 days treatment with FdUrd alone (50 mg/kg/ day × 2) did not result in significant reduction in tumor volume. Coadministration of BBBA at 5 and 10 mg/kg/day × 2 did not potentiate the efficacy of FdUrd over that achieved by FdUrd alone, but it significantly reduced the tumor volume by 27 and 32%, respectively, when compared with untreated controls. FdUrd alone at 150 mg/kg/day × 2 reduced the tumor volume by 29%. This reduction in tumor volume was enhanced 1.8-fold by coadministration of BBBA (10 mg/kg/day × 2). At a higher tumor burden (5 × 106 cells/mouse) and 4 days treatment, BBBA at 10 and 30 mg/kg/day × 4 reduced further the tumor volume produced by FdUrd alone (200 mg/kg/day × 4) by 1.2- and 1.4-fold, respectively. At a higher dose of FdUrd (400 mg/kg/day × 4), the potentiation by BBBA (10 and 30 mg/kg/day × 4) was 1.6- and 3.4-fold, respectively. Enzyme studies suggest that the lower sensitivity to FdUrd and the better potentiation of FdUrd cytotoxicity by BBBA in DLD-1 as compared to HCT-15 could be attributed to higher uridine Phosphorylase activity in DLD-1. There were no significant differences between DLD-1 and HCT-15 in the activities of other enzymes involved in FdUrd metabolism. Enzyme studies also indicated that DLD-1 and HCT-15, in contrast to host tissues, contain no thymidine Phosphorylase and have higher kinase activities towards FdUrd. Therefore, the enhancement of FdUrd antitumor efficacy by BBBA appears to be due to the specific inhibition of uridine phosphoryl ase. Such inhibition would selectively prevent catabolism and deactivation of FdUrd in the tumors but not in the host. The selective Inhibition of FdUrd catabolism along with the higher thymidine kinase activities in the tumors would channel the metabolism of FdUrd in the tumors towards anabolism and formation ofits active metabolite 5-fluoro-dUMP to prod uce the selective toxicity of FdUrd. These findings may lead to a more successful use of FdUrd in cancer chemotherapy, especially against tumors that lack thymidine phosphorylase. © 1995, American Association for Cancer Research. All rights reserved.