Dual Bclxl and BCL2 Inhibition with Navitoclax (NAV), Venetoclax (VEN), and Decitabine (DEC) for Advanced Myeloid Neoplasms (MN): Safety and Biological Activity in a Phase 1 Study

Introduction: The combination of a DNA methyltransferase inhibitor (DNMTi) and the BCL2 inhibitor venetoclax (VEN) is effective for newly-diagnosed acute myeloid leukemia (AML). However, response rates of VEN/DNMTi are lower for other advanced myeloid neoplasms (MN). One resistance mechanism to VEN is increased cell dependence on the alternative anti-apoptotic protein BCLxL. We hypothesized that addition of the BCLxL/BCL2 inhibitor navitoclax (NAV) to VEN would increase apoptotic priming in advanced MNs and facilitate lower NAV dosing to reduce its platelet (PLT)-associated toxicity. We report results of a phase 1 study of NAV + VEN/decitabine (DEC) in advanced MNs.

Methods: This phase 1 study used a straight six-design. Eligible subjects were age ≥18 years, with PLT ≥25 x 10⁹/L, and diagnosed with 1) secondary AML with treatment of prior MN or therapy-related AML (s/t-AML), 2) accelerated/blast-phase myelofibrosis (AP/BP-MF), 3) myelodysplastic syndrome/myeloproliferative neoplasm (MDS/MPN) with ≥5% blasts, or 4) relapsed/refractory (R/R) MDS with ≥5% blasts. 1° objectives were to determine safety and recommended phase 2 dose (RP2D) of NAV + VEN/DEC. 2° objectives were to determine best overall response rate (ORR) and overall survival (OS).

Cycle (C) lengths were 28 days (D). DEC 20 mg/m²/day was administered IV D1-5 for all dose levels (DL). At DL0 (starting DL, open to all diagnoses), oral VEN 400mg/day and NAV 25mg/day were given D1-14. At DL1 (open to all diagnoses), NAV dose was increased to 50mg/day. At DL2 (BP-MF and AML only), VEN duration was increased to D21. For C1 only, NAV began D3 if PLT ≥10 x 10⁹/L after VEN dose ramp-up D1-3. ORR was defined as complete remission (CR) or CR with incomplete hematologic recovery (CRi) for AML (ELN 2022); CR, CR uni/bilineage (CRuni/bi), or CR with partial hematologic recovery (CRh) for MDS (IWG 2023); CR or marrow response for MDS/MPN (IWG 2015); and acute leukemia response-complete (ALR-C) for AP/BP-MF (MPN-BP 2012). Hematologic dose-limiting toxicity (DLT) during C1 was defined as grade 4 neutropenia or thrombocytopenia >D42 in absence of disease, or PLT <10 x 10⁹/L ≥7 days.

Results: Sixteen subjects were treated (4 s/t-AML, 3 AP/BP-MF, 4 chronic myelomonocytic leukemia [CMML], 1 MDS/MPN-unclassifiable [U], 4 R/R MDS). Median age was 72 years (range 50-80); 7 (44%) were female, and 3 (19%) were non-white. Four subjects previously received VEN; 1 had prior HSCT. The MDS/MPN-U subject withdrew after 3 NAV doses. Dose-escalation proceeded through DL0 (n=6), DL1 (n=7), and DL2 (n=1 only, due to discontinued NAV availability); dose-expansion occurred at DL1 (n=2). At DL1, 1 DLT occurred due to grade 4 ANC that resolved on D65. Frequent grade ≥3 TEAEs were neutropenia (69%), thrombocytopenia (63%), and febrile neutropenia (44%). One death occurred due to respiratory failure in a R/R MDS subject in C2.

A median of 2 cycles (range 1-not reached) was administered. ORR among evaluable subjects was 60% (9/15), including s/t-AML (ORR 2/4; both CR), AP/BP-MF (1/3; ALR-C), CMML (4/4; 1 CR and 3 marrow responses), and MDS (2/4; 1 CRh and 1 CRbi). To date, 5 subjects were bridged directly to HCT. At a median follow-up of 13 months (95% CI: 5.5-not reached), the 1-year OS was 35% (95% CI: 15%-83%) and median OS was 10 months (95% CI: 5.4-not reached).

BH3 profiling of pretreatment myeloblasts showed increased apoptotic priming in responding subjects (n=7; p=0.007) and BCLxL dependence (p=0.046) using the BH3 peptides BIM and HRK, respectively, compared to non-responders (n=6). Among 8 subjects with monoblasts (CD33+CD64+CD14+), all responders (n=4) had CMML and increased mitochondrial BCL2 and BCLxL dependence vs. non-responders (p=0.029). Based on this, single-cell RNA and chromatin profiling of 2 CMML subjects at 6 pre- and post-treatment timepoints was done, revealing transcriptomic and epigenomic signatures of a BCL2 and BCLxL-expressing, myeloid progenitor-like cell compartment that was eradicated on therapy. Finally, the immature platelet fraction was higher in subjects with baseline PLT <100 vs. ≥100 x 10⁹/L (p=0.035), but this difference diminished upon adding NAV on C1D3, suggesting on-target BCLxL inhibition.

Conclusion: NAV + VEN/DEC is safe for high-risk MN with manageable cytopenias and evidence of BCL-xL on-target activity. Initial RP2D in MDS or MDS/MPN is NAV 50 mg D1-14. Continued study of BCLxL inhibition is recommended for high-risk MNs, especially CMML.