Guarcello, Vincenzo

2557 Background: Irreversible inhibition of dihydropyrimidine dehydrogenase (DPD) by EU blocks 5-FU catabolism allowing for oral 5-FU administration with complete bioavailability. Unfortunately, phase III trials with co-administered EU/5-FU showed inferiority vs. 5-FU/leucovorin, and were discontinued. We recently reported that competitive inhibition of human uridine phosphorylase (UP) and thymidine phosphorylase (TP) 5-FU-anabolic enzymes by EU is an important mechanism potentially responsible for clinical failure of the combined EU/5- FU regimen. We hypothesize that EU inhibition of UP and TP is transient, while that of DPD is prolonged, allowing for novel schedule dependent optimization of EU/5-FU dosing regimens with improved efficacy. Methods: In this phase I study, five patients received a single oral dose (2 mg, 5 mg or 10 mg) of EU 12–14 hours prior to scheduled resection of primary/metastatic colorectal cancer. Dosage was as follows: Two patients received the 2 mg dose, one patient received the 5 mg dose and two patients received the 10 mg dose. Matched normal and tumor tissue biopsies were immediately snap frozen and subsequently UP, TP and DPD activity was measured in vitro via HPLC detection of [6- 14C]-5-FU catabolites/anabolites. Peripheral blood mononuclear cell (PBMC) DPD activity was determined at baseline prior to EU administration, 30 min prior to surgery (Day 1), and on Days 2, 5 and 14 following EU administration. Results: At 12–14 hours following EU administration, there was an absence of inhibition of UP and TP, while DPD was significantly inhibited in matched tumor and normal tissue. Importantly, PBMC DPD activity was significantly inhibited by EU on Day 1 (12–14 hours after EU administration) and Day 2 (36 hours after EU administration) at 0 ± 0% and 17 ± 11% (mean ± SD) of baseline, respectively. Conclusions: These data demonstrate a differential recovery time of EU mediated inhibition of UP and TP compared to DPD, which permits future schedule dependent optimization of EU/5-FU therapy. No significant financial relationships to disclose.

Dihydropyrimidine dehydrogenase (DPD) is one of the factors that determine the efficacy and toxicity of 5-fluorouracil. Variations in DPD activity may result from alterations at the transcriptional level of the DPYD gene. Heterogeneity in DPYD expression has been reported, but the molecular mechanisms responsible for this remain unclear. We investigated methylation of the DPYD promoter as a mechanism for transcriptional regulation of DPYD in the RKO colorectal cancer cell line. We demonstrate that the active transcription machinery for DPYD is present in RKO cells, but promoter binding of Sp1, a transactivator of DPYD, was inhibited, which on subsequent examination was shown to be associated with dense promoter methylation. Treatment with 5-aza-2′-deoxycytidine alone or the combination of 5-aza-2′-deoxycytidine and trichostatin A induced demethylation of the promoter and markedly increased the DPYD mRNA level in RKO cells but not in unmethylated WiDr cells. Furthermore, in vitro methylation of the DPYD promoter decreased promoter activity. These data suggest an important role for methylation in DPYD suppression. The transcriptional suppression of DPYD by methylation may be responsible for the increased 5-fluorouracil sensitivity observed in some patients. This may also provide insight into the mechanism underlying the downregulation of DPYD in some colorectal cancers.