A Phase I Trial of a Pharmacodynamically-Conceived Decitabine and Thioguanine Combination in Patients with Advanced Myeloid Malignancies

Background: Using a chemosensitivity screening assay, we previously demonstrated that decitabine and thioguanine combinations can rescue therapeutic efficacy in primary leukemia cells isolated from patients with relapsed and refractory acute myeloid leukemia (AML). Although both decitabine and thioguanine have single-agent anti-leukemic activity, they have not previously been used concurrently. To test the safety and preliminarily assess possible additive/synergistic activity of this combination, we are performing a Phase I dose escalation trial of thioguanine given with decitabine.

Patients and Methods: Patients with untreated AML over 60 years who are not suitable candidates for standard induction therapy, as well as those with relapsed/refractory AML, advanced myelodysplastic syndrome, and chronic myelomonocytic leukemia (CMML) are eligible. Three thioguanine dose levels are being evaluated: 80, 120, and 160 mg/m²/day, given in two divided doses on Days 1-12 of each induction course for up to two cycles and Days 1-7 of maintenance cycles. Decitabine 20 mg/m² is administered IV on Days 3-12 during induction cycles and on Days 3-7 during maintenance cycles. In addition to standard safety measures and clinical outcomes, the biologic activity of the combination is assessed by patient specific pharmacodynamic measures. These measures include an in vitro chemosensitivity assay, genome-wide analysis of DNA methylation changes, and BH3 profiling in order to measure the degree to which samples are primed to undergo apoptotic cell death.

Results: Six patients (median age, 69 yrs; range, 66–83 yrs) with newly diagnosed AML (n=2), relapsed/refractory AML (n=3), and newly diagnosed CMML (n=1) have been treated to date with thioguanine at the 80 mg/m² dose. Dose-limiting toxicity was seen in one patient who developed acute renal failure (ARF) requiring hemodialysis. Infectious complications were the most common toxicity and included two episodes of grade 3 neutropenic colitis in one patient, grade 3 pseudomonal bacteremia, grade 3 staphylococcal bacteremia, and a dental infection requiring extraction. No other grade 3–4 non-hematologic toxicities were observed. Three patients remain on active treatment three to seven months from the initiation of therapy. Two patients were removed from study due to disease progression (n=1) and grade 4 ARF noted above (n=1). A patient with CMML was removed to undergo allogeneic hematopoietic stem cell transplantation after achieving a hematologic remission with red cell transfusion independence and platelet count normalization. The median number of cycles administered was 3 (range, 1-5). Bone marrow blast reductions to less than 5% were seen in all 4 evaluable patients with AML and included 1 CR and 1 CRi. The overall response rate in this high risk patient population was 67% (4 of 6 patients). Importantly, the patient who achieved a CR was 83 years old with relapsed disease following 4 previous cycles of single-agent decitabine therapy, demonstrating that this regimen can rescue previous hypomethylating agent failures. Clinical activity was well-correlated with the in vitro cytotoxicity assays. The chemosensitivity assay results for thioguanine on pre-treatment mononuclear cells directly correlated with clinical outcome in both response to therapy and clinical resistance. BH3 profiling was also performed on pre-treatment myeloblasts. Significant apoptotic priming corresponded to good initial clinical response. In addition, in the single patient who responded and subsequently relapsed, decreased apoptotic priming was observed in the relapsed sample, suggesting that the thioguanine/decitabine regimen applies selective pressure for reduction in apoptotic priming. Detailed analysis of treatment-related changes in DNA methylation will be presented and compared to prior work showing decitabine-induced hypomethylation in CD34+ leukemic cells during single agent therapy with decitabine.

Conclusions: Combination decitabine and thioguanine has been well-tolerated and has shown surprising anti-leukemic activity at the lowest dose level studied. Importantly, the in vitro chemosensitivity testing and BH3 profiling accurately predicted the observed clinical responses while also confirming the etiology of the loss of therapeutic response. Accrual to this trial continues at higher dose levels to determine the maximum tolerated dose.