Prediction of Clinical Response to Venetoclax Plus Decitabine in AML Using a 7-Day Ex Vivo Assay

INTRODUCTION: The combination of venetoclax (VEN) and a hypomethylating agent (HMA) such as decitabine (DEC) is frequently used in acute myeloid leukemia (AML) for patients unfit for standard chemotherapy. However, there is a subset of patients who do not clinically respond to such combination therapies. To identify VEN + HMA non-responders prior to treatment we developed an assay on our automated, high throughput (HTS) ex vivo drug sensitivity platform to predict how a given patient will clinically respond to DEC + VEN within an ongoing, open label, phase II clinical trial (NCT03404193) for patients with newly diagnosed elderly/unfit or relapsed/refractory (R/R) AML. To maximize the predictive power of our assay, blast and T cells of baseline samples were phenotyped, using multiparameter flow cytometry. Identifying non-responding patients before they are treated is important as it would enable them to avoid a treatment failure, unnecessary toxicity, and prioritize an alternative treatment. Importantly, this approach would define a patient population with a high unmet medical need, warranting the development of new therapeutics specifically for this population.

METHODS: In NCT03404193, participants received DEC IV over 1 hour on Days 1-10 for induction and on Days 1-5 after achieving complete remission (CR) or complete remission with incomplete count recovery (CRi) during consolidation/maintenance. Participants received VEN PO daily on Days 1-28 of Cycle 1 and on Days 1-21 of subsequent cycles. VEN could be stopped on Day 21 if bone marrow analysis showed blast clearance or hypocellularity. Treatment cycles repeated every 4-6 weeks for up to 24 cycles in the absence of disease progression or unacceptable toxicity. Fig. 1A outlines the treatment schedule for NCT03404193. The primary objective is overall response rate (ORR). Peripheral blood (PB) specimens were obtained at baseline and 3±1 days after treatment begin in Cycle 1. Specimens were shipped on ice and processed within 1-4 days of the blood draws. Following red blood cell (RBC) lysis, nucleated cells were seeded in multi-well plates and treated with VEN in combination with DEC. Cells were cultured for a total of 7 days. Ex vivo blast reduction and frequencies of T cell subtypes were evaluated via multiparametric flow cytometry analysis and compared to clinical response.

RESULTS: 7 days after initiation of the DEC + VEN ex vivo treatment of baseline samples, blast counts from Responders (Rs, N=2) were markedly reduced compared to those from Non-Responders (NRs, N=4) (Fig. 1B). Furthermore, baseline samples from Rs (N=9) exhibited higher levels of both CD4+ and CD8+ naïve T cells than from NRs (N=7) (Fig. 1C). To assess whether the kinetics of blast reduction in vivo correlates with increased or decreased levels of certain immune cell subsets we re-defined Rs and NRs based on their baseline-normalized blast counts 2 or 4 days after treatment begin in Cycle 1. Rapid Responders, enriched for subjects who experienced CRs or CRi's, were defined as subjects with blast levels ≤50% 2 days, or ≤33% 4 days, after treatment begin. Rapid Responders (N=6) exhibited higher levels of CD4+ T cells at the evaluation time point (2-4 days after treatment begin) than subjects with slower blast reductions or increases in their blast counts (N=4) (Fig. 1D).

CONCLUSIONS: Baseline samples may predict clinical response to DEC + VEN in ex vivo drug treatment and immunophenotyping assays and may consequently serve to identify non-responders early, allowing them to switch to a potentially more effective treatment regimen before the lack of clinical response to DEC + VEN would become apparent. Rapid reduction of blasts in vivo coupled with increases in CD4 T cell levels may be a prognostic indicator of good clinical response. Limitations of our laboratory study include small sample sizes. Efforts are underway to validate the trends reported here.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal