A Chemo-Free Bridge-to-Transplant Strategy with Venetoclax and Azacitidine for NPM1-Mutated Acute Myeloid Leukemia in Molecular Failure

Background: In patients (pts) with NPM1-mutated acute myeloid leukemia (AML) fit for intensive chemotherapy, the persistence, reoccurrence, or progression of NPM1^mut transcript identifies poor chemotherapy responders that require allocation to allogeneic stem cell transplantation (alloSCT). Currently, there is no standard of care for pts in molecular failure and the proper bridging strategy to alloSCT still needs to be defined. The combination of Venetoclax (VEN) and Azacitidine (AZA) offers an alternative chemo-free strategy with favorable toxicity profile in NPM1^mut AML, a molecular subset exquisitely sensitive to VEN.

Aims: To characterize a series of NPM1^mut AML pts in molecular failure who received VEN-AZA as bridge-to-transplant treatment strategy.

Methods: We explored the off-label combination of Venetoclax (VEN) and Azacitidine (AZA) in 9 NPM1^mut fit AML pts in molecular failure after intensive chemotherapy eligible for alloSCT, treated at 4 Italian Hematology Institutions. Disease was monitored on bone marrow (BM) with NPM1^mut-specific qRT-PCR assay. The quantitative detection limit of the assays was 0.01%. Complete response with minimal residual disease negativity (CR-MRD_neg) was defined as ratio NPM1^mut/ABL x 100 transcript < 0.01 copies

Results: Of the 9 pts analyzed 4 (44%) were female, median age at diagnosis was 56 years (range 41-66) and 58 years (range 43-68) at VEN-AZA therapy. Eight of the 9 pts had ELN2017 favorable-risk AML (2 pts with concomitant FLT3-ITD^low and 1 FLT3-TKD) and 1 pt had intermediate-risk AML (FLT3-ITD^high). Pts received in median 4 (range 2-5) cycles of front line chemotherapy. Seven pts were treated for molecular relapse and 2 for molecular persistence/progression. All pts received full VEN dosage (400 mg/die or 100 mg/die if concomitant strong CYP3A4 inhibitor) and AZA 75 mg/m²/die s.c. in 5- or 7-day schedule. All pts but 1 performed VEN ramp-up. Pts received a median of 3 cycles (range 1-4). No AZA dose reduction occurred. Median VEN duration during cycle (C)1 was 25 days (range 14-28), 21 days (range 7-28) in C2, 24.5 days (range 7-28) in C3 and 25.5 days (range 19-28) in C4. VEN dosage reductions were due to hematological toxicity ≥ G3 in 60% of cases and in 40% for logistic reasons. The observed ≥ G3 toxicities were neutropenia in 4/9 (44%) pts, 1 thrombocytopenia and 1 febrile neutropenia, managed out-patient. No tumor lysis syndrome was observed.

Best response to VEN-AZA was CR MRD_neg (NPM1 ^mut undetectable) in 7/9 (78%) pts, reached after a median of 2 cycles (range 1-4). Importantly, no pts relapsed ongoing therapy; 1 pt, notably the FLT3-ITD^high pt, in CR MRD_pos presented molecular progression after C3.

Median time from VEN-AZA start to alloSCT was 4.3 months (range 1.7-8.1). At pre-transplant evaluation 6/9 (67%) pts were in CR MRD_neg. Eight of 9 pts received myeloablative conditioning and 1 pt received reduced-intensity conditioning.

With a median follow-up from alloSCT of 9.3 months (range 2.4-24.0) all patients are alive and 8/9 (89%) in CR MRD_neg. Four pts presented NPM1^mut reoccurrence after alloSCT: in 2 pts CsA tapering achieved molecular negativity, 1 pt underwent prophylactic therapy with sorafenib and 1pt is currently in CR MRD_pos. In no pts FLT3 mutations were detected and, most importantly, no patients experienced morphological relapse.

Conclusions: VEN-AZA demonstrated to be a safe and effective bridging regimen for NPM1^mut AML pts in molecular failure, allowing to achieve deep responses in the majority of pts at the expense of low toxicity. These preliminary encouraging results will be explored in the GIMEMA AML2521 trial in the same patient subset (clinicaltrials.gov NCT04867928).