Real-World Multicenter Evaluation of Venetoclax and Azacitidine for Molecular Relapse after First Line Intensive Chemotherapy in Patients with NPM1 Mutated or Core Binding Factor (CBF) AML. a Filo Study

Introduction: Molecular follow-up (FU) of measurable residual disease (MRD) by real-time quantitative PCR (qPCR) is well established in AML patients (pts) with NPM1^mut, RUNX1::RUNX1T1, or CBFB::MYH11 transcripts. Pts who fail to achieve molecular complete remission (Mol-CR) after intensive chemotherapy (IC) have a higher risk of relapse and death. Pts who achieve Mol-CR have a better prognostic but should be monitored according to ELN MRD guidelines since up to 30% of them may present molecular relapse (Mol-Rel) during follow-up (Heuser, Blood 2021). In those pts, there is growing interest to use pre-emptive treatment before overt morphologic relapse (Orvain, Leukemia 2024). However, there is no consensus on the best treatment approach. Venetoclax–based low intensity therapy has shown promising efficacy in molecular failure of NPM1-mutated AML (Jimenez-Chillon, Blood Adv 2024).

Here, we report results of venetoclax plus azacitidine (VEN/AZA) in patients with NPM1^mut or CBF AML in Mol-Rel.

Methods: Inclusion criteria were : NPM1^mut or CBF AML, first morphologic CR following IC, no alloHCT in first line,regular monitoring of MRD by qPCR in bone marrow (BM) and/or blood (PB), Mol-Rel (defined according to ELN criteria), at least 1 cycle of VEN/AZA for Mol-Rel. Pts with Mol-Rel followed by morphologic relapse before VEN/AZA or those who relapsed after alloHCT were not included. At time of Mol-Rel confirmation, pts received off label venetoclax 400 mg/d (d1 to d7, 14, 21 or 28 according to centers) without ramp-up and azacitidine 75 mg/m²/d subcutaneously (d1-7 or d1-5, d8-9). Antifungal and antibiotic prophylaxis and G-CSF were used according to each center.

Results: 72 pts from 10 FILO centers treated with VEN/AZA for Mol-Rel between Sept 2019 to July 2024 were included. Main characteristics at diagnosis were: female (54%), median age 57y (IQR 45-65), median WBC 36 G/L (IQR 15-80), NPM1^mut (n=68, 94%), CBF (n=4, 6%), and FLT3-ITD^mut (n=22, 31%). All achieved first CR after one induction chemotherapy and received a median of 3 IDAC/HDAC or 6 mini-consolidations cycles (pts > 60y) as consolidation. Only 5 pts (7%) did not achieve negative or low-level (LL) MRD in PB after consolidation.

Median time between CR1 and Mol-Rel was 10 months (IQR 8-15) and between Mol-Rel (first sample) and VEN/AZA was 49 days. Median number of VEN/AZA cycles was 2 (IQR 2-3).

During first VEN/AZA cycle, 1 (3%), 14 (37%), 6 (16%), and 16 (42%) pts received 7, 14, 21, 28 days of VEN, respectively, with 56 (78%) treated as outpatient, 16 (22%) receiving posaconazole prophylaxis and 28 (41%) G-CSF. Thirty-eight pts (53%) had grade 3-4 neutropenia. However, only 13 (18%) had febrile neutropenia and 7 (10%) a documented infection.

After the first cycle, 52 (72%) were evaluated for MRD in PB, with 11 (21%) achieving MRD negativity and 22 (42%) LL MRD. After the second cycle, 45 (80%) were evaluable for MRD in PB, with 22 (49%) achieving MRD negativity and 15 (33%) LL MRD.

After a median of 2 cycles (IQR 1-2,5) 53 pts (73%) proceeded to alloHCT, including 50 (94%) with morphologic CR and 25 (49%) with Mol-CR. Most patients (n=33, 62%) received reduced-intensity sequential conditioning and the donor was often matched-unrelated (46%). Among the pts who did not undergo alloHCT (n=19), 12 (63%) did not because of comorbidity or age. Those patients received a median number of 7 (IQR 3-10) VEN/AZA cycles.

With a median FU from Mol-Rel of 20 months, median overall survival (OS) was not reached with a 1y-OS of 82%. 57 pts (84%) maintained molecular response in PB (negative or LL MRD) at last news. In patients who proceeded to alloHCT (median FU from alloHCT of 16 months), median OS was not reached with a 1y-OS of 84 % and 48 pts (91%) were in molecular response in PB at last news. In pts who did not proceed to alloHCT, median OS was not reached and 1y-OS of 73%. There were 10 morphologic relapses (3 and 7 in allografted and non allografted patients). Among the eleven patients who died (7 after alloHCT), 5 did while in continuing CR.

Conclusion: MRD monitoring and treatment of sub-clinical relapses is becoming a major challenge in AML, since treatments applied in a situation where the tumour burden is still small are likely more effective and better tolerated. Although more data in CBF-AML are needed, this study shows that VEN/AZA is a safe and effective treatment in NPM1mutated AML pts with Mol-Rel and provides benchmark data for the design of prospective trials in this specific situation.