CPX-351 selectively benefits patients with AML and myelodysplasia-related mutations in the pivotal randomized trial

CPX-351 was approved for the treatment of acute myeloid leukemia (AML) using now-outdated definitions of AML with myelodysplasia-related changes (AML-MRC) and therapy-related AML. We evaluated whether the overall survival (OS) benefit of CPX-351 over standard cytarabine plus anthracycline (7+3) therapy is confined to molecularly defined AML subgroups by performing DNA sequencing in 184 patients enrolled in the pivotal phase 3 randomized trial. Patients were categorized hierarchically based on gene mutations: (1) TP53-AML, (2) DDX41-AML, (3) myelodysplasia-related AML (AML-MR) defined by the World Health Organization fifth edition criteria, or (4) other-AML. TP53-AML was subclassified as single (TP53single) or multihit (TP53multi) based on the number of alleles altered via mutation, deletion, or copy neutral loss of heterozygosity. Two-year OS differed significantly across molecular subgroups: TP53-AML (7%), AML-MR (19%), other-AML (37%), and DDX41-AML (70%) (P< .001). CPX-351 improved survival in patients with AML-MR compared with 7+3 (median, 9.7 vs 6.8 months; P = .037), with no benefit in TP53-AML or other-AML. For patients undergoing transplantation, CPX-351 improved 2-year survival (76% vs 27%; P< .01), an effect primarily observed in AML-MR. Multivariable analysis confirmed the independent association with survival of both CPX-351 and hematopoietic cell transplantation in AML-MR. TP53multi demonstrated significantly worse survival than TP53single (median, 3.8 vs 7.0 months; P = .004). The OS benefit of CPX-351 observed in the trial was driven by AML-MR with no benefit of CPX-351 in TP53-AML, in which the primary prognostic factor was allelic state. This trial was registered at www.clinicaltrials.gov as #NCT01696084.