ALBERT

Description: Introduction and objectives: During an adaptive immune response antigen-specific T cells rapidly proliferate and differentiate into cytotoxic T lymphocytes. Most of these cells undergo apoptosis but some develop into high-affinity memory CD8+ T cells. The BCL-2 family of proteins regulates apoptosis and has a critical role in development and maintenance of the immune system. Venetoclax (Venclexta™, ABT-199) is a selective BCL-2 inhibitor that increases tumor cell apoptosis, and is approved by the FDA for patients with chronic lymphocytic leukemia (CLL) or small lymphocytic lymphoma (SLL), with or without 17p deletion, who have received at least one prior therapy. Given the critical role of BCL-2 in the regulation of the immune system, we hypothesized that venetoclax may affect the anti-tumor activity of immune checkpoint inhibitors. Results: To interrogate the effects of venetoclax on T cells we initially performed a series of in vitro studies using human lymphocytes treated with clinically relevant doses of the drug. As previously reported (Khaw et al., Leukemia; 28(6):1207-1215, 2014), human peripheral blood mononuclear cells (PBMCs) treated with venetoclax exhibited a dose-dependent decrease in the number of B-cells and T-cells (CD4+ and CD8+ T-cells). Upon further characterization of the surviving T cells, we found that while the proportion of naïve T-cells decreased with increasing venetoclax concentrations, the proportion of memory T-cells increased, specifically CD8+ and CD4+ T effector memory cells (Figure 1). We next examined the effects of venetoclax on T-cell function in vitro in response to immune stimulation with or without immune checkpoint blockade. To address this we performed a mixed lymphocyte reaction (MLR) assay, in which primary monocyte-derived dendritic cells from one donor were cultured with CD4+ T-cells from another donor. In the MLR reaction we observed that venetoclax reduced CD4+ T-cell viability in a dose-dependent manner, but it did not limit T-cell proliferation of surviving cells. Venetoclax did not affect IFNγ secretion within these surviving cells and, more importantly, did not reduce the effects of the checkpoint inhibitor nivolumab (Figure 2). To test the effects of venetoclax on antigen-specific T cells, we performed a cytomegalovirus (CMV) recall assay where PBMCs from CMV-positive human subjects were incubated with CMV antigen and the activity of T cells was measured by IFNg secretion. Although venetoclax treatment reduced the total number of cells, IFNg production from antigen-specific CMV+ T cells remained comparable to DMSO control and combining venetoclax with nivolumab did not affect the anti-PD-1 response (Figure 3). Finally, to investigate the effects of venetoclax in combination with anti-PD-1 therapy in vivo we used the murine syngeneic tumor model MC38. Venetoclax did not impair the efficacy of anti-PD-1, and in some studies increased efficacy relative to either anti-PD-1 or venetoclax monotherapy alone. To determine whether the efficacy of the venetoclax-anti-PD-1 combination is immune-mediated, we transplanted immunodeficient mice with MC38 cells and repeated the same treatment regimens. The lack of efficacy in any of the treatment arms indicates that the contribution of venetoclax to efficacy in this solid tumor model is immune-mediated (Figure 4). Conclusions: These data suggest that venetoclax treatment results in loss of naïve but not memory T cells. Venetoclax did not affect the viability, the induction or frequency of memory T cells. In human in vitro experiments and in an in vivo syngeneic tumor model venetoclax did not antagonize the therapeutic effect of anti-PD-1. Contrary to our initial hypothesis, we find that modulation of the immune system by venetoclax may support its potential use for immune-based cancer therapy, as memory T-cells can rapidly acquire effector and cytotoxic function to eliminate cancer cells. Taken together, we provide evidence that venetoclax in combination with immune checkpoint inhibitors should be further explored as a therapy for cancer patients. All authors are employees of AbbVie. The design, study conduct, and financial support for this research were provided by AbbVie. AbbVie and Genentech participated in the interpretation of data, review, and approval of the publication. Disclosures Mathew: AbbVie Inc.: Employment. Haribhai:AbbVie Inc.: Employment. Kohlhapp:AbbVie Inc.: Employment. Duggan:AbbVie Inc.: Employment. Ellis:AbbVie Inc.: Employment. Riehm:AbbVie Inc.: Employment. Robinson:AbbVie Inc.: Employment. Shi:AbbVie Inc.: Employment. Bhathena:AbbVie Inc.: Employment. Leverson:AbbVie Inc: Employment, Equity Ownership, Patents & Royalties. Pappano:AbbVie Inc.: Employment. Donawho:AbbVie Inc.: Employment. Uziel:AbbVie Inc.: Employment.

Description: Background Acute myeloid leukemia (AML) is a highly heterogeneous disease with multiple prognostic indicators of outcomes to conventional treatments; including age, cytogenetic risk, and presence of genomic abnormalities (NPM1, TP53, FLT3 and IDH1/2). Additionally, overexpression of BCL-2 is a predictor for poor response to chemotherapy and can lead to therapeutic resistance in AML (Campos et. al. Blood, 1993). Venetoclax (Ven), an oral selective BCL-2 inhibitor, in combination with hypomethylating agents (HMA) or low dose cytarabine (LDAC), was recently approved for treatment naive patients (pts) with AML who were not fit for standard induction therapy on the basis of two phase 1b/2 studies (DiNardo et.al, Lancet Onclogy, 2018; Wei et.al., J Clin. Oncol., 2019). Here, we present clinical outcome results from the phase 1b/2 study populations in subgroups defined by molecular markers and correlations with BCL-2 family expression. Methods: The data cut-off dates were 31-Aug-2018 (VEN+HMA study M14-358, NCT02203773) and 22-Aug-2018 (VEN+LDAC study M14-387 NCT02287233). Pt responses were assessed according to the IWG criteria for AML (Cheson et.al., J Clin. Oncol.,2003). The presence of AML associated mutations (NPM1, IDH1, IDH2, TP53 and FLT3) were determined centrally in bone marrow aspirates (BMA) at baseline. BCL-2 (BCL2) gene expression was defined centrally by quantitative polymerase chain reaction (qPCR) using the deltaCt (ΔCt) method. BCL2 expression normalized to a reference gene (2-ΔCt) was evaluated for pts with 〉 50% AML blasts in baseline BMA (median 75%; range 50-99%). The composite complete remission (CR) and CR with incomplete marrow recovery (CRi) rate, overall survival (OS), time to first response (TTR), and duration of response (DOR), and BCL-2 gene expression were correlated with presence of baseline molecular markers. Results: Data from 209 pts who received Ven 400 mg or 600 mg in combination with either HMA or LDAC are reported here. The median age was 74 years (range: 61-90 years) and 61% (n=127) were male (Table 1). Of the 167 pts with centrally assessed molecular markers, IDH1 or IDH2 mutations were detected in 25.7%, NPM1 mutations in 15.6%, TP53 mutations in 22.2%, and FLT3 mutations in 18.0% pts. Pts may have more than one of these mutations (i.e co-mutations of NPM1 and IDH1or IDH2, NPM1 with FLT3, or NPM1 with TP53) identified in baseline BMA. The CR/CRi rates were 83.7% for pts with IDH1/IDH2 mutations, 84.6% for pts with NPM1 mutations, 59.5% for pts with TP53 mutations, and 53.3% for pts with FLT3 mutations (Table 2). The median OS was not reached (NR) for pts with IDH1, IDH2 or NPM1 mutations, while the median OS for pts with detectable TP53 mutations or FLT3 mutations was 8.9 mos and 12.4 mos respectively. The remissions (CR or CRi) were rapid (median TTR between 1.1 mos to 1.8 mos) for each group (Table 2) and durable responses were observed for pts with IDH1, IDH2, NPM1, or FLT3 mutations (median DOR was either NR for IDH and NPM1 and 19.9 mos for FLT3). Pts with TP53 mutations are known to have poor prognosis and the DOR in this cohort was 5.6 mos. The median time on study was 11.6 mos (range: 0.3-44 mos). The range of BCL2 mRNA expression (2-ΔCt) in bone marrow blasts was 0.03 to 2.93, with a median of 0.86. The Cox hazard ratio for OS was 1.06, p=0.8565 based on BCL2 above the median (〉 0.86 2-ΔCt). The multivariate analysis included age, cytogenetic risk, AML type (primary vs secondary) and trial (M14-387 vs M14-358). Albeit not statistically significant and small sample sets, pts with either IDH1 or IDH2 mutations (N = 21) tended to have higher BCL2 expression (median 2-ΔCt 1.1 ; range 0.19 -2.93) while pts with TP53 mutations (N = 7) had lower levels of BCL2 expression (median 2-ΔCt 0.55; range 0.03 - 1.27). Determination of baseline expression of other family members (BCL2L1, MCL1, BCL2A1, BCLw, NOXA, etc) is ongoing and additional univariate and multivariate analyses will be presented at the meeting. Conclusions: These analyses demonstrate that VEN + HMA or LDAC has efficacy across multiple molecular markers in AML. This activity is rapid and durable, and is observed across different levels of BCL2 expression in AML blasts. Disclosures Chyla: Abbvie, Inc: Employment, Other: Stock or options. Harb:AbbVie Inc: Employment, Other: Stock/stock options. Mantis:AbbVie Inc: Employment, Other: Stock/stock options. Riehm:AbbVie Inc.: Employment, Other: Stock/stock options. Ross:AbbVie: Employment, Other: Stock/stock options. Sun:AbbVie: Employment, Other: Stock/stock options. Huang:AbbVie Inc: Employment, Other: Stock/stock options. Jiang:AbbVie Inc.: Employment, Other: Stock/stock options. Dail:Genentech: Employment, Equity Ownership. Peale:Genentech Inc.: Employment, Other: Stock/stock options. Potluri:AbbVie, Inc.: Employment, Other: Stock/stock options. Hayslip:AbbVie Inc: Employment, Other: Stock/stock options. OffLabel Disclosure: Venetoclax is a BCL-2 inhibitor that is FDA-approved in some indications. This presentation will focus on venetoclax for treatment in acute myeloid leukemia, which is not an approved indication.