Quizartinib with Decitabine +/- Venetoclax Is Highly Active in Patients (Pts) with FLT3-ITD Mutated (mut) Acute Myeloid Leukemia (AML): Clinical Report and Signaling Cytof Profiling from a Phase IB/II Trial

Background: Single-agent gilteritinib and quizartinib (quiz) met their primary endpoints of improved overall survival (OS) in relapsed/refractory (R/R) FLT3-mut AML in phase III studies, but CRc durations were short (4-11 months). Quiz demonstrated potent synergy with venetoclax (VEN) (a BCL-2 inhibitor) in AML cell lines and PDX models (Mali et al. Haematologica 2020).

Methods: Newly diagnosed or R/R AML with ECOG ≤2, and adequate organ function were eligible. The protocol initially evaluated safety of the doublet of decitabine (DAC) plus quiz regardless of FLT3 status [Fig 1A]. After the first 10 pts were treated on the doublet, the protocol was amended to add VEN (triplet) in a safety dose escalation lead in [Fig 1B]. CRc criteria were as published in the phase III ADMIRAL and QUANTUM-R studies.

Results: 21 pts including DAC10 + quiz (phase 1: N=10) and DAC10 + VEN + quiz (phase IB/II: n=11) were evaluable at the time of this report (Table 1).

DAC10 + Quiz

Of 10 pts treated with DAC10 + quiz (Table 1), CRc rate was 40% (1 CRp and 3 CRi). No FLT3 R/R WT (n=3) pts responded. Four of 7 FLT3-ITD mut pts (1 frontline and 5/6 R/R, who had all received ≥1 FLT3 TKI) achieved CRc (57%), with 3/6 FLT3-PCR negative at response. 1 pt experienced QTcF prolongation > 500 msec on quiz 40mg/day (resolved after holding quiz with no clinical cardiac events). Grade 3/4 toxicities in >/= 2 pts, irrespective of attribution, included infection (12), neutropenic fever (3), mucositis (3), diarrhea (2), and prolonged QTcF >450 (2). With a median (med) follow-up (f/u) of 12 months (mos), the med OS was 5.7 months in the 6 R/R FLT3-mut pts (Fig 2A). Three of 4 CRc pts proceeded to allogeneic stem cell transplantation (ASCT) and 2 were alive in remission at last f/u.

DAC10 + VEN + Quiz

Of 11 enrolled pts (Table 2), 10 treated prior to July 1 2020 were evaluable. CRc achieved in 9/10 (90%) (1 CR, 3 CRp, 5 CRi) with 5/9 responders FLT3-PCR negative. Excluding the 1 frontline evaluable pt, 8 of 9 R/R FLT3-ITD (88% with ≥1 prior FLT3 TKIs) pts achieved CRc. No pts developed a DLT with 30 mg/day quiz, however with the 40mg/day quiz 2 pts developed hematologic DLT (grade ≥3 neutropenia with a <5% cellular bone marrow lasting ≥42 days. Hence, quiz 30 mg/day dose was determined as RP2D for triplet. Grade 3/4 toxicities in >/=2 pts, irrespective of attribution, included infection (7) and neutropenic fever (4). No QTcF prolongations >500msec noted. 60-day mortality was 0. With a med f/u of 6 mos, med OS was not reached and 6-month OS was 86% (one death) (Fig 2B). Of 9 CRc pts, 5 remain on protocol in CRc, 2 underwent ASCT and are alive post-ASCT, and 2 have relapsed.

CyTOF (single-cell mass cytometry) analysis was performed in longitudinal samples collected on the triplet therapy. In a pt (baseline FLT3-ITD, NRAS, IDH1/2, SRSF2, DNMT3A, RUNX1, CSF3R) with a 3 week CRc duration, we identified two distinct blast populations at C1D1; monocytic (CD33+/CD34-/CD68+) and myeloid (CD33+/CD34+) (Fig 3A). The monocytic blasts had low BCL-2 and high MCL-1 at baseline, with modest reduction at EOC1 (from 25.8% to 10.8%) (Fig 3B-C). However, the myeloid blasts with high BCL-2 and lower MCL-1 showed significant reduction at EOC1 (from 52% to 5.5%) (Fig 3B-C). Heat-map demonstrated major downregulation in signaling pathways (MCL-1, STAT, BCL-XL, and MAPK) in the myeloid blasts at EOC1 of triplet therapy (Fig 3D). However, these signaling pathways remained active in the monocytic blasts at EOC 1 (Fig 3D). Pt relapsed prior to starting C2, driven by monocytic blasts (Fig 3C).On the other hand, in a pt (baseline FLT3-ITD only) who remains in CR (DOR 11 mos), both myeloid and monocytic populations had high BCL-2 at baseline (Fig 4A-B). Despite almost complete elimination of myeloid blasts (Fig 4C), monocytic blasts were detectable at a low level at C1D28 (Fig 4C). Importantly, converse to the prior pt, the active signaling pathways at diagnosis (MCL-1, STAT, MAPK, c-MYC) were significantly downregulated in both populations at EOC1 (Fig 4D). This patient underwent ASCT on D35 of therapy, and remains in CR.

Conclusion: DAC10 + venetoclax + quiz is highly active in R/R FLT3-ITD mut AML pts, with CRc rates of 90% and projected 6-month OS of 86%. CYTOF profiling may predict for response based on pre- and on-therapy apoptotic and signaling pathway profiles enabling potential optimal selection of future combinatorial or sequential approaches. Accrual to the triplet continues and updated clinical and correlative data will be presented.