The outcomes of elderly patients and relapsed/refractory patients with either acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS) remain poor. Venetoclax is a selective BCL-2 inhibitor that recently demonstrated impressive activity when combined with hypomethylating agents decitabine or 5-azacitidine; however, outcomes remained more modest among patients with unfavorable risk cytogenetics or with TP53 mutations. Building on recent findings that increasing the schedule of decitabine from days 1-5 to days 1-10 of 28 day cycles was associated with improved overall responses and survival among patients with unfavorable- risk cytogenetics and TP53 mutations, we hypothesize that concomitant use of venetoclax and 10-day decitabine will improve the response and survival rates, especially in patients with high-risk karyotypes or TP53 mutations. We designed a Phase II trial that will enroll four parallel, open-label cohorts, each with 40 patients consisting of high-risk AML or MDS patients, either with advanced age or with relapsed/refractory disease. The primary objective is to determine the composite overall response rate; secondary objectives include determining disease-free and overall survival, and the impact of high-risk karyotypes on response and survival. In addition, two molecular hypotheses will be tested. First, that clearance of exome-defined founding clone mutations from the peripheral blood provides a consistent and quantified response end-point that circumvents many sources of inter-patient response variability, and that combination venetoclax and decitabine will be associated with increased rate and depth of mutation clearance vs. single-agent decitabine. Determining whether a hypomethylating doublet has improved outcomes vs. single-agent has been challenging for all but large randomized studies, partially due to clinical confounders such as hemodilute aspirates and poor count recovery in older and heavily pre-treated patients. This novel approach to response determination isolates anti-leukemic activity from other factors, thus improving statistical power of the study. Second, we will determine whether responses to combination venetoclax and decitabine correlate with leukemic dependence on BCL-2 activity or on other anti-apoptotic proteins. We will apply dynamic BH3 profiling to determine the dependence of AML blasts on BCL-2, BCL-XL or MCL-1 and correlate these results with response, survival, and mutation patterns. Further, we will apply CyTOF analysis to serial bone marrow samples obtained during therapy and at relapse. We will quantify leukemia stem cell subpopulations and the expression of BCL-2 family proteins within bulk AML cells vs. leukemia stem cells. These data will determine whether combination venetoclax and decitabine leads to elimination of both bulk and leukemia stem cell populations, and whether sensitivity within subpopulations corresponds with intracellular levels of BCL-2 family proteins. Collectively, these studies will evaluate clinical responses and molecular outcomes of a novel combination therapy and will identify prognostic biomarkers and resistance mechanisms.
High-risk Acute Myeloid Leukemia and Myelodysplastic Syndrome are aggressive hematologic cancers associated with a very poor prognosis, with less than 20% of patients surviving beyond 1 year. Venetoclax, a BCL-2 inhibitor is one of the most exciting drugs in AML therapy. In this project we designed a Phase II clinical trial combining venetoclax with an optimized 10-day decitabine regimen in unfavorable risk categories of AML and MDS, with goals to improve outcomes of these patients by accelerating molecular responses and to further characterize molecular determinants of response and resistance.