P548: BEMCENTINIB COMBINED WITH LOW-DOSE CYTARABINE IS EFFICACIOUS AND WELL TOLERATED IN RELAPSED AML PATIENTS UNFIT FOR INTENSIVE CHEMOTHERAPY. UPDATES FROM THE ONGOING PHASE II TRIAL (NCT02488408)

Abstract

Background: The new standard of care (SOC) in newly-diagnosed AML patients (pts) unfit for intensive chemotherapy (IC) due to age or co-morbidities yields favourable efficacy. However, beyond 1st line, these pts have limited treatment options, with a dismal mOS of 2.4 months at relapse, highlighting a significant unmet need for new treatments in this pt population. Bemcentinib (BEM) is a first-in-class, orally bioavailable, highly selective AXL-inhibitor. AXL (a receptor tyrosine kinase) represents an important target in AML, as its expression on AML cells and stem cell compartment is associated with poor prognosis, resistance to chemotherapy and decreased antitumor immune response. Aims: The ongoing BGBC003 PhII trial studied the safety and efficacy of the BEM+LDAC combination in relapsed (REL) AML pts unfit for IC; in addition, a comprehensive translational biomarker analysis was pursued. Here, we present preliminary efficacy and safety data and include initial translational multiomics data from bone marrow mononuclear cells (BMMNC). Methods: Pts received BEM at 200 mg/OD with 3 loading doses at 400 mg/OD and LDAC at SOC schedule. Study endpoints included overall survival (OS), objective response (OR), clinical benefit rate (CBR) (OR+unchanged [UC]) and exploratory biomarker analyses. Longitudinal BMMNC samples (n=32) from 13 pts underwent scRNA-seq and CITEseq (Chromium 10x genomics; TotalSeq, Biolegend). Pts were stratified by best response: CR, CRi, PR for Responders; UC, PD for Non-Responders. Cell type annotations were inferred from expression of known markers at RNA and protein level and for identification of malignant cells copy number variation was included. Results: As of 13 Dec 2021, 22 REL AML pts (10/22 in 1st relapse) were treated with BEM+LDAC. Median age was 75.5yrs [66-86], median prior lines of therapy 1 [1-8] and adverse cytogenetic risk in 7/22 (32%). Only 19/22 pts were evaluable for efficacy. The overall REL population (n=19) demonstrated a composite CR (CRc=CR+CRi) of 26% (5/19) and CBR of 79% (14/19); AML pts in 1st relapse (10/19) showed a CRc of 30% (3/10) and CBR of 70% (7/10) whereas REL pts with time on treatment (ToT) > 3 months (11/19) achieved CRc of 45% (5/11) and CBR 91% (9/11); median ToT of 12.9 months for CR/CRi pts; 1 pt remains on treatment. Median OS was 6.2 months in the overall REL pts (n=19), 7.4 months in the AML pts in 1st relapse (10/19) and 11.3 months in REL pts with ToT >3 months. Late onset responses suggest an immunological mechanism of action (iMOA) and may contribute to a longer ToT and survival. CITEseq analysis identified differences in the immune compartment, underscoring an iMOA associated with response to BEM+LDAC. Furthermore, increased gene set enrichment of TNFα signalling via NfkB at screening in malignant blasts of responders emphasizes the influence of the TNFα signalling pathway as an interesting field of further research. The safety profile of BEM+LDAC in the REL AML pts (n=19) is comparable with the known safety profile of LDAC. TRAEs of ≥G3 observed in ≥10% of pts were anaemia (33% BEM+LDAC; 22% BEM), and ECG QT prolonged (11% BEM+LDAC; 11% Bem). No G5 TRAEs reported. Summary/Conclusion: BEM+LDAC is efficacious and well tolerated in REL AML pts unfit for IC, with promising survival benefit compared to historical data. Translational research showed activation of CD8+T cells and B/Plasma cells in response to treatment, indicating that BEM elicits activation of two major adaptive immune cell populations responsible for anti-AML immune responses. BEM+LDAC warrants further evaluation in this population.