Treatment phase stratified outcomes of core binding factor acute myeloid leukemia.

Abstract

e19055 Background: Mutations in core binding factor (CBF) are associated with a favorable prognosis in acute myeloid leukemia (AML). However, between 30-40% of patients with CBF-AML relapse, and relative outcomes with each subsequent line of therapy remain unclear. Additionally, the overall survival of CBF-AML compared to other favorable risk molecular aberrations — such as NPM1mut and CEBPA — is unknown in light of the addition of monoallelic CEBPA in ELN 2022. Therefore, the purpose of this study was to clarify the outcomes of CBF-AML relative to other favorable risk aberrations and to determine the relative overall survival associated with each subsequent relapse. Methods: We retrospectively analyzed 46 consecutive patients with CBF-AML treated from January 2013 to January 2022 at VCU Massey Cancer Center that underwent intensive chemotherapy (IC) with or without gemtuzumab ozogamicin. During the same study period, we analyzed a comparison cohort of 272 consecutive patients: 103 ELN 2017 intermediate-risk patients, 111 adverse-risk patients, eight patients with mono- or biallelic mutations in CEBPA, and 50 with ELN 2017 favorable-risk NPM1mut. All patients were treated with front-line IC. We analyzed survival by the Kaplan-Meier method and compared groups with the log-rank test. Results: We analyzed 46 patients with CBF-AML across 136 treatment phases. Patients with CBF-AML had superior overall survival compared to patients with NPM1mut (median overall survival not reached at a median follow-up time of 5.0 years versus 21.0 m., p = 0.028). There were no differences in survival between CBF-AML and AML with monoallelic or biallelic mutations in CEBPA at a median follow-up time of 3.7 years (not reached, p = 0.976). Expectedly, the overall survival for CBF-AML was superior to ELN 2017 intermediate (30.1 m., p = 0.011) and adverse (10.9 m., p < 0.0001) risk disease. Next, we analyzed survival trends in the relapsed or refractory setting for CBF-AML by lines of therapy, using ELN 2017 intermediate and adverse risk cohorts as a comparison. During the first relapse, CBF-AML was associated with a significantly superior overall survival compared to the adverse cohort (19.1 m. versus 10.0 m., p = 0.024) but not the intermediate cohort (14.5 m., p = 0.734). The absence of KITmut was not associated with improved survival compared to the ELN 2017 intermediate category at the first relapse ( p = 0.610). At the second relapse, the survival disparity disappeared between the CBF-AML and adverse risk groups (7.7 m. vs 4.0 m., p = 0.964). Conclusions: CBF-AML was associated with similar survival to mono- or biallelic mutations in CEBPA, and significantly prolonged overall survival compared to NPM1mut. Relative to a matched cohort of adverse-risk AML at the time of the first relapse, outcomes of relapsed CBF-AML remain favorable. The relative overall survival benefit of CBF-AML disappeared at the time of the second relapse.