Phase 1 Trial Testing the Novel Combination Therapy of Entrectinib and ASTX727 in TP53 Mutated Relapsed/Refractory Acute Myeloid Leukemia Patients

Introduction:

The TP53 apoptotic network and mitochondrial functionality have been established as drivers of response to the BCL2 inhibitor, venetoclax, in acute myeloid leukemia (AML). We previously identified that TP53-knockout AML cell lines were venetoclax-resistant but sensitive to NTRK inhibitors (Nechiporuk et al. Cancer Discovery 2019). These findings were corroborated in primary AML samples and animal models, justifying the conduct of a clinical investigation of NTRK inhibition in TP53-mutated (mTP53) AML.

Methods:

We developed a Phase 1 study combining the pan-NTRK inhibitor, entrectinib, with oral decitabine/cedazuridine (DEC-C or ASTX727) for an initially planned sample size of 12 patients (pts). Study objectives were to identify a maximum tolerated dose (MTD) and assess safety and efficacy of this novel all-oral therapy in pts with relapsed or refractory (R/R) mTP53 AML, predicted to be sensitive based on our preclinical data.

A Bayesian toxicity probability interval ("Keyboard") dose-finding design was planned to select an MTD among 3 dose levels of entrectinib – 200 mg (Dose Level [DL] -1), 400 mg (DL1), and 600 mg (DL2) – with the first dosing cohort starting at DL1 and a target dose-limiting toxicity (DLT) rate of 20%. A 7-day entrectinib monotherapy lead-in (Cycle 0 [C0]) was followed by 28-day cycles comprising daily entrectinib (at a patient’s assigned DL) and 1 tablet at fixed dose (35 mg DEC/100 mg C) of ASTX727 on days 1-5. MTD-evaluable pts had a DLT and/or exposure to >14 days of entrectinib in Cycles 0-1 and all Cycle 1 ASTX727 doses. DLTs were defined for C0 and C1 as study drug-related grade ≥3 non-hematologic toxicities (with some exceptions for toxicities that quickly resolved to a lower grade) and grade 4 neutropenia or thrombocytopenia persisting at Day 42 in the absence of leukemia. Adverse events (AEs) were analyzed for pts with any exposure to both entrectinib and ASTX727 (the safety set) while clinical response was evaluated in pts with at least one post-baseline response assessment (the efficacy set). Correlate studies including plasma inhibitory assays (PIA), mutation burden evaluation, and ex vivo drug profiling were performed on blood and marrow samples obtained from pts at pre-treatment, C0D7, and C1D28.

Results:

Thirteen pts were enrolled (1 was replaced, per protocol, for not taking any ASTX727): 12 met safety set criteria, 11 were MTD-evaluable (3 treated at DL-1 and 8 at DL1; DL2 has not been reached), and 7 were efficacy-evaluable. Among safety set participants, 9 (75%) were male, median age was 69 years, 92% had ECOG status 0-1, and 67% had at least 3 lines of therapy. One patient (assigned to DL1) had a postural blood pressure drop that was a DLT; observed DLT rate was 12.5% at this dose level. Six pts (50%) had grade 3+ treatment-related AEs, with 5 having grade 3 febrile neutropenia (4 of whom had active disease at time of onset). Among efficacy set pts, 3 (43%) had stable disease at C1D28 and one (14%) – assigned to DL1 – received 7 cycles of therapy and achieved a complete remission with response lasting 5 months. This responding patient was heavily pre-treated (4 lines) and enrolled on this study following an early relapse post-allogeneic stem cell transplant (allo-SCT).

PIA analysis performed on 10 pts after 7 days of entrectinib monotherapy indicated that three pts (2 in DL1, 1 in DL-1) lacked sufficient drug levels in plasma to inhibit the kinase activity of an Trk-expressing cell line (Joshi et al. Blood 2020), potentially explaining each one’s early progressive disease. Variant allele frequency (VAF) comparisons of pre-treatment and C1D28 samples identified four out of seven evaluable pts (57%) with TP53 VAF decreases, two of whom had increased VAFs in other mutations, suggesting differential impact on mTP53 cells. Ex vivo drug screens of mononuclear cells showed pre-treatment dose-dependent sensitivity to entrectinib, without additive effect of ASTX727 on cell viability.

Conclusion:

Combined therapy of entrectinib and oral DEC-C was safe for pts with R/R mTP53 AML, a nearly fatal diagnosis. Notably, we report a durable response in one patient who relapsed early after allo-SCT, suggesting a role for this approach in augmenting a graft vs. leukemia effect. We also observed on-target effects on mTP53 VAF in pts receiving this novel regimen. The study is ongoing and an MTD has not yet been identified. More data and correlative findings will be presented at the meeting.