A phase 1 study of orally administered 5-fluoro-2’-deoxycytidine with tetrahydrouridine in patients with refractory solid tumors

Purpose: The DNA methyltransferase (DNMT) inhibitor 5-fluoro-2’-deoxycytidine (FdCyd) combined with tetrahydrouridine (THU) yielded promising activity in patients with advanced solid tumors, but the intravenous administration schedule of FdCyd limited the clinical feasibility of this treatment program. Therefore, we developed an orally bioavailable formulation of FdCyd and determined the safety, recommended phase 2 dose (RP2D), pharmacokinetics, molecular pharmacodynamic (PD) effects, and antitumor activity of this agent combined with THU. Methods: Adult patients with advanced solid tumors received FdCyd and THU orally on an intermittent schedule in 21-day cycles; dose levels included once- or twice-daily dosing administered on the first 3–7 days (depending on the dose level) of weeks 1 and 2 of each cycle, with no administration on week 3. Dose escalation followed a standard 3 + 3 design; doses were increased until the target FdCyd maximum plasma concentration corresponding to DNMT inhibition in preclinical studies (1 µM) was reached, after which, the total dose was escalated by increasing the number of days and/or frequency of FdCyd-THU administration. Blood specimens were collected for pharmacokinetic analysis and circulating tumor cell (CTC) PD analyses. Paired pre- and on-treatment (cycle 1 week 3) tumor biopsies were collected during the expansion phase to assess changes in expression of DNMT1 and the epigenetically regulated tumor suppressor protein p16 by immunohistochemistry (IHC), as well as changes in genome-wide DNA promoter methylation. Results: Fifty-nine patients with solid tumors were enrolled. The RP2D was 160 mg FdCyd once daily combined with 3000 mg THU once daily on days 1–6 and 8–13 of each 21-day cycle. Dose-limiting toxicities (DLT) were grade 3 diarrhea and grade 3 refractory nausea, vomiting, and diarrhea; the most common grade 3–4 adverse events were hematological toxicities. The best response was prolonged stable disease (17 cycles). Active FdCyd plasma concentrations were achieved at doses of 60 mg and higher, and THU exposure was associated with DLT. One of the 7 patients (14%) with analyzable paired tumor biopsy specimens exhibited an appreciable increase in tumor p16 expression, and none had appreciable decreases in qualitative tumor DNMT1 levels. An increase in the proportion of p16-expressing cytokeratin-positive (CK⁺) CTCs was detected in 77% of patients (23 of 30) evaluable for CK⁺ CTC response, while that for vimentin-positive (V⁺) CTCs was 9% of patients (2 of 22) evaluable for V⁺ CTC response. Patients with paired biopsies and a best response of stable disease showed treatment-induced promoter hypomethylation for several epigenetically regulated genes, including tumor suppressor genes. Conclusion: We determined the RP2D for the combination of orally administered FdCyd and THU and measured prolonged stable disease and tumor suppressor gene hypomethylation in some patients, suggesting potential clinical benefit and molecular activity for this regimen in some patients. The paucity of tumor DNMT1 decreases and p16 re-expression are consistent with the lack of clinical response. However, it may also reflect the timing of on-treatment biopsies (following the 1-week break in FdCyd administration), since increases in p16-expressing CTCs were measured for the majority of CTC-assessable patients.