Abstract
Introduction: Since the introduction of hypomethylating agent (HMA) and venetoclax combination therapy for acute myeloid leukemia (AML), retrospective analyses have demonstrated similar outcomes between patients who received this less intensive approach compared to those who received intensive therapy (Cherry et al. Blood Adv 2021; Matthews et al. Blood Adv 2022). Prior studies on the effect of these treatments on post-allogeneic hematopoietic cell transplant (alloHCT) outcomes have been limited by between-group differences in treatment period, alloHCT conditioning intensity, and GVHD prophylaxis (Winters et al. Transplant Cell Ther 2022; Short et al. Blood Adv 2023). Herein, we aim to compare alloHCT outcomes between patients who received intensive, cytarabine-based therapy vs. HMA/venetoclax for AML in a patient population that received a standardized conditioning and GVHD prophylaxis platform.
Methods: The cohort included patients with AML who underwent alloHCT in first remission from 2018-June 2024 at Johns Hopkins University using a uniform reduced-intensity conditioning (RIC) platform of fludarabine, cyclophosphamide, and 400 cGy total body irradiation and GVHD prophylaxis incorporating post-transplant cyclophosphamide (PTCy) 50 mg/kg on days +3 and +4. Patients were stratified by receipt of a remission induction regimen containing at least standard dose cytarabine or HMA/venetoclax. Baseline differences were evaluated via Mann-Whitney U and Fisher Exact tests for continuous and categorical variables, respectively. Overall survival (OS), non-relapse mortality (NRM), and relapse incidence (RI) were compared between induction groups by the Kaplan-Meier estimator with log-rank test and Gray's test for unadjusted analyses and Cox proportional hazards and subdistribution hazards methods for analyses adjusted for unbalanced baseline characteristics.
Results: A total of 253 patients met inclusion criteria. 200 received pre-alloHCT intensive, cytarabine-based remission induction therapy (70% 7+3 induction, 19.5% CPX-351), and 53 received HMA/venetoclax. Groups were statistically balanced by AML TP53 mutation presence (7.5% and 2.0%, respectively), primary refractory disease, Karnofsky performance status, HCT-CI, and graft source. Recipients of cytarabine-based induction were younger at time of alloHCT (median 56.5 vs. 69.2 years, p<0.001) and had younger donors (median 28.0 vs. 35.0 years, p=0.003). Pre-alloHCT multicolor flow cytometry (MFC) measurable residual disease (MRD) was less frequent in those who received cytarabine-based remission induction (3.5% vs. 11.3%, p=0.03). In unadjusted analysis, 2-year OS was 82.1% (95% CI: 76.9-87.7%) and 68.5% (95% CI: 56.8-82.7%) in cytarabine-based and HMA/venetoclax groups, respectively (p=0.01). This difference was attributed to a lower incidence of NRM in the cytarabine-based therapy group (2-year NRM: 8.0% [95% CI: 4.8-12.3%] vs. 25.3% [95% CI: 14.3-37.9%], p<0.001). RI was similar between groups (2-year RI: 16.1% [95% CI: 11.3-21.6%] vs. 12.3% [95% CI: 4.9-23.5%], respectively, p=0.47). In analyses adjusted for recipient and donor age, neither OS (cytarabine-based therapy hazard ratio (HR): 0.68 [95% CI: 0.38-1.23], p=0.20), NRM (HR: 0.61 [95% CI: 0.27-1.36], p=0.22), nor RI (HR: 1.11 [95% CI: 0.43-2.90], p=0.83) were statistically different between induction groups. Among the four subgroups in which at least ten patients received either therapy type, IDH mutations, secondary mutations, secondary cytogenetics, and alloHCT recipients age ≥ 65 years, post-alloHCT RI was not shown to be different based on remission induction therapy in unadjusted and adjusted analyses.
Conclusions: When contemporary alloHCT techniques of RIC and high-dose PTCy GVHD prophylaxis are used, differences in OS and NRM between those receiving intensive, cytarabine-based AML remission induction regimens and HMA/venetoclax therapy appear to be accounted for by recipient and donor age differences. Those receiving HMA/venetoclax were found to have similar post-alloHCT RI on unadjusted and adjusted analyses despite a higher incidence of pre-alloHCT MFC MRD, further supporting MFC MRD's diminished prognostic relevance after HMA/venetoclax. These data support further investigation of HMA/venetoclax AML induction compared to intensive regimens using propensity score-based approaches and randomized trials with particular interest in genetically-defined subgroups.
