P506: ACHIEVEMENT OF MEASURABLE RESIDUAL DISEASE CLEARANCE IS A STRONGER PREDICTOR OF PATIENT OUTCOME THAN TREATMENT INTENSITY IN NEWLY DIAGNOSED PATIENTS WITH ACUTE MYELOID LEUKEMIA

Abstract

Background: Modern therapy for acute myeloid leukemia (AML) can consist of either intensive or low-intensity regimens selected based on patient (pt) age/comorbidities. Measurable residual disease (MRD) has emerged as a strong independent prognostic factor in AML. It is unknown whether pts who achieve similar levels of MRD clearance experience comparable outcomes irrespective of treatment intensity. Aims: To establish the relative prognostic contribution of treatment intensity and MRD status on overall survival (OS) and relapse-free survival (RFS) in newly diagnosed AML pts who have achieved a first response. Methods: We conducted a retrospective chart review to identify non-CBF, non-APL AML pts treated at our institution between 2010 and 2021 with intensive (IA; int/high-dose cytarabine + anthracycline-based, without venetoclax) or low-intensity (Lo + VEN; low-dose cytarabine/hypomethylating agent-based, with venetoclax) regimens who had achieved CR/CRi/MLFS and undergone MRD testing by multiparameter flow cytometry at time of first response. Differences in baseline pt characteristics were evaluated using the Chi-square or Wilcoxon-Mann-Whitney tests for categorical and continuous variables, respectively. The Kaplan-Meier method was used to estimate median OS and RFS, with pts censored at the time of stem cell transplantation (SCT). Multivariate analysis was performed using a Cox proportional hazards model. Results: We identified 635 pts meeting inclusion criteria. Baseline characteristics are shown in Table 1. Compared to the IA-treated pts (n=385), pts treated with Lo + VEN (n=250) were significantly older, more likely to be male, less likely to achieve MRD clearance, and more likely to be classified as adverse risk by ELN 2017. The SCT rate in the Lo + VEN cohort was half that of the IA cohort (25% vs 50%). Median OS was 51m, 24.6m, 15m, and 9.9m in the IA MRD(-), Lo + VEN MRD(-), IA MRD(+), and Lo + VEN MRD(+) groups, respectively. Median RFS was 27.8m, 15.4m, 7.3m, and 5.2m in the IA MRD(-), Lo + VEN MRD(-), IA MRD(+), and Lo + VEN MRD(+) groups, respectively. Pts within the same MRD category (+ or -) had significantly higher OS and numerically higher RFS if treated with IA versus Lo + VEN. When the analysis was restricted to pts aged ≥ 60 years (n=56 in IA, n=239 in Lo + VEN; Fig. A-B), the differences within MRD categories did not reach statistical significance (median OS NR in IA MRD(-) vs 24.6m in Lo + VEN MRD(-), p=0.08; median OS 13m in IA MRD(+) vs 10.6m in Lo + VEN MRD(+), p=0.50; median RFS NR in IA MRD (-) vs 15.4m in Lo + VEN MRD(-), p=0.07; median RFS 25.1m in IA MRD(+) vs 5.3m in Lo + VEN MRD(+), p=0.24). The SCT rate was 29% (87/295) in the age ≥ 60 cohort. Given the confounding effect of unbalanced pt characteristics between the IA and Lo + VEN cohorts, we performed a multivariate analysis on the full population (n=635) taking into consideration the effects of age, treatment intensity, MRD status, and ELN category on OS and RFS (all significant by univariate analysis). In the multivariate analysis for OS, ELN adverse risk was the strongest predictor (HR 2.09, p<0.001), followed by MRD(+) status (HR 1.68, p<0.001), while treatment intensity and age were not significantly predictive. For RFS, ELN adverse risk (HR 2.31, p<0.001) and MRD(+) status (HR 1.98, p<0.001) were also the only two significant predictors. Image:Summary/Conclusion: In newly diagnosed AML patients, MRD status at time of first response and ELN risk are stronger predictors of pt outcome than intensity of therapy received.