Efficacy and Safety of Venetoclax in Combination with Gilteritinib for Relapsed/Refractory FLT3-Mutated Acute Myeloid Leukemia in the Expansion Cohort of a Phase 1b Study

Background: The FMS-like tyrosine kinase 3 (FLT3) inhibitor gilteritinib (Gilt) improves survival compared with standard salvage chemotherapy in relapsed/refractory (R/R) FLT3-mutated (FLT3^mut+) acute myeloid leukemia (AML); however, relapses are common and long-term survival remains poor. Combination therapies of FLT3 tyrosine kinase inhibitors (TKIs) with agents that induce apoptosis have demonstrated preclinical synthetic lethality and enhanced cytotoxicity against FLT3^mut+ clones. This may delay or prevent drug resistance. We present updated data from the dose-escalation and expansion cohorts of an ongoing, multicenter, open-label, Phase 1b trial (NCT03625505) of Gilt in combination with B-cell lymphoma 2 (BCL-2) inhibitor venetoclax (Ven) in patients (pts) with R/R FLT3^mut+ AML.

Methods: Pts with R/R AML with ≥1 prior line of therapy and Eastern Cooperative Oncology Group (ECOG) performance status of 0–2 were eligible. The expansion cohort required FLT3^mut+ R/R AML at enrollment. Pts received the recommended Phase 2 dose (RP2D) of Ven 400 mg in combination with Gilt 120 mg daily in 28-day cycles, following a 3-day ramp-up of Ven. Pts with prior Gilt were excluded from expansion, and none were enrolled in dose escalation; prior Ven or other FLT3 inhibitor was permitted. Concurrent use of moderate or strong CYP3A4 inhibitor was allowed. Interruptions of Ven or both drugs simultaneously were permitted for management of adverse events (AEs), and subsequent cycles could include dose adjustments to minimize myelosuppression or other drug-related AEs. The primary endpoint was modified composite complete remission (mCRc; complete response [CR] + CR with incomplete platelet recovery [CRp] + CR with incomplete blood count recovery [CRi]) + morphologic leukemia-free state (MLFS). Responses were graded by the International Working Group (IWG) for AML criteria (with adapted CRi criteria as presented in the ADMIRAL trial) for comparison with composite CR (CRc) used in Gilt development. All criteria for response used by IWG were maintained in these response definitions. Secondary endpoints included CR + CR with partial hematologic recovery (CRh) and duration of response (DOR) of mCRc.

Results: At data cutoff, April 15, 2020, 39 R/R pts started therapy at the RP2D (internal tandem duplications [ITD] only [n=32]; tyrosine kinase domain [TKD] mutation only [n=5]; both TKD and ITD [n=2]). Median age was 63 years (range 23–85) and the majority had an ECOG status of 1–2 (87.2%). Most pts (n=29; 74.4%) had received ≥2 prior lines of therapy, including ≥1 FLT3 TKI (n=25, 64.1%; midostaurin, n=19; sorafenib, n=5; crenolanib, n=1). Twelve pts (30.8%) had prior stem-cell transplant (SCT); 3 (7.7%) had received prior Ven.

Grade ≥3 treatment-emergent AEs (TEAEs) were reported in 37 pts (94.9%), and 28 (71.8%) reported serious AEs. The only Grade 3/4 nonhematologic TEAE reported in >20% of pts was febrile neutropenia (48.7%). Four pts (10.3%) experienced a fatal AE, including colitis, disease progression, multiple organ dysfunction syndrome, and bronchopulmonary aspergillosis (all n=1). AEs leading to Ven and Gilt interruptions occurred in 19 pts (48.7%) and 18 pts (46.2%), respectively, and AEs leading to dose reductions occurred in 2 pts (5.1%) for each treatment. Six pts (15.4 %) and 5 pts (12.8%) discontinued Ven and Gilt, respectively, owing to AEs.

mCRc was achieved by 83.8% of pts (n=31/37; Table). With median duration of follow-up of 2.9 (range 0.8−11.0) months in all pts, median event-free survival was 5.1 (95% confidence interval: 2.8, 8.3) months but administrative censoring after short follow-up for a large number of recently enrolled patients limits interpretation; median DOR has not been reached. Eight pts (21.6%), all ITD only, proceeded to SCT. CR + CRh was achieved by 7 pts (18.9%).

Conclusions: Ven plus Gilt achieved a very high overall rate of marrow and blood blast elimination and mCRc rate (84%) in this expansion cohort of heavily pretreated FLT3^mut+ pts, the majority of whom had prior FLT3 TKI exposure. Using similar response assessment criteria, the high mCRc rate here suggests substantially greater antileukemic activity from Ven plus Gilt compared with single-agent Gilt. Cytopenias were prominent but manageable with dose interruption/modification on subsequent cycles. Nonhematologic toxicities were modest, and the combination was well tolerated.