Prognostic factors of overall (OS) and relapse-free survival (RFS) for patients with acute myeloid leukemia (AML) in remission after intensive chemotherapy (IC): Multivariate analyses from the QUAZAR AML-001 trial of oral azacitidine (Oral-AZA).

Abstract

7014 Background: Demographic and disease factors influence outcomes for patients (pts) with AML. In the phase 3 QUAZAR AML-001 trial, Oral-AZA significantly prolonged OS and RFS vs. placebo (PBO) for pts with AML in first remission after IC (Wei, NEJM, 2020). Univariate analyses showed OS and RFS benefits with Oral-AZA vs. PBO across pt subgroups defined by baseline (BL) characteristics. MV analyses were performed to identify BL characteristics independently predictive of OS/RFS in QUAZAR AML-001, and to assess Tx effects of Oral-AZA vs. PBO on survival when adjusted for BL factors. Methods: Pts were aged ≥55 yrs with AML in complete remission (CR) or CR with incomplete count recovery (CRi) after induction ± consolidation. Within 4 months of CR/CRi, pts were randomized 1:1 to receive Oral-AZA 300 mg or PBO for 14d/28d cycle. Cox proportional hazards models were used to estimate Tx effects of Oral-AZA vs. PBO on OS and RFS, adjusting for BL age, sex, ECOG PS score, cytogenetic risk at diagnosis (Dx), prior MDS, geographic region, CR/CRi after induction (per investigator) and at BL (per sponsor), MRD status, receipt of consolidation, number of consolidation cycles, platelet count, and ANC. In a stepwise procedure, randomized Tx and BL variables were selected incrementally into a Cox model if P ≤ 0.25. After each addition, the contribution of the covariate adjusted for other covariates in the model was evaluated and retained in the model if P ≤ 0.15. Results : Oral-AZA Tx remained a significant independent predictor of improved OS (HR 0.70) and RFS (HR 0.57) vs. PBO after controlling for BL characteristics (Table). MRD status, cytogenetic risk, and pt age were each also independently predictive of OS and RFS. Response after induction (CR vs. CRi) and BL ANC were predictive of OS but not RFS, whereas prior MDS, CR/CRi at BL, and number of consolidation cycles were only predictive of RFS. Conclusions: Tx with Oral-AZA reduced the risk of death by 30% and risk of relapse by 43% vs. PBO independent of BL characteristics. Cytogenetic risk at Dx, MRD status, and pt age also independently predicted survival outcomes. Clinical trial information: NCT01757535. [Table: see text]