AUTHORS
Jesus D Gonzalez-Lugo, MD, Alexander Joseph Ambinder, MD, MPH2, Justice Ameyi, M.S.3, Deedra Nicolet, MS4, Krzysztof Mrózek, MD, PhD5, Tyler Ross, MD6, Michael Walker7, Alice Scott Mims, MD8, Tara L. Lin, MD9, Ann-Kathrin Eisfeld, MD10
AFFILIATIONS
1 University of Kansas Cancer Center, Kansas City, Kansas, United States
2 Johns Hopkins University, Baltimore, Maryland, United States
3 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
4 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
5 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
6 University of Kansas Cancer Center, Kansas City, Kansas, United States
7 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
8 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
9 University of Kansas Cancer Center, Kansas City, Kansas, United States
10 The Ohio State University Comprehensive Cancer Center, Columbus, Ohio, United States
ABSTRACT
Background:
Therapy-related acute myeloid leukemia (t-AML) is generally associated with poor outcomes. Current ELN and NCCN guidelines recommend that favorable-risk t-AML, including core-binding factor (CBF) AML, CEBPA bZIP-mutated AML, and NPM1-mutated/FLT3-ITD-negative AML, be managed according to genetic risk rather than therapy-related etiology. However, retrospective studies have reported inferior survival in CBF t-AML, while data for other favorable-risk subgroups remain limited. We compared outcomes of favorable-risk t-AML and de novo AML.
Methods:
We retrospectively analyzed 393 adults with newly diagnosed favorable-risk AML (2022 WHO/ICC criteria) treated at three US centers in the MARROW Consortium (2005–2026). Patients received intensive chemotherapy (IC; 7+3 or CPX-351; n=327) or lower-intensity therapy (LIT; hypomethylating agent ± venetoclax; n=66). Therapy-related AML was defined as AML following prior chemotherapy and/or radiotherapy. Endpoints were composite complete remission (CRc), event-free survival (EFS), and overall survival (OS).
Results:
Among IC-treated patients, 289 (88%) had de novo AML and 38 (12%) had t-AML. Patients with t-AML were older (median, 57 vs 50 years, p=0.01), whereas allogeneic transplantation rates were similar (37% vs 36%, p=1.0). CRc did not differ between t-AML and de novo AML overall (86.8% vs 78.9%; p=0.29) or in CBF AML (92.9% vs 88.2%, p=0.74). Likewise, relapse rates were similar overall (45.5% vs 29.8%, p=0.11) and across favorable-risk molecular subgroups. OS was shorter in t-AML than de novo AML (median, 8.4 vs 14.6 years, p=0.04), whereas EFS was not significantly different (1.2 vs 3.0 years, p=0.17). Among patients aged <60 years, OS and EFS were comparable between groups. No significant differences in OS or EFS were observed in either CBF subtype, inv(16) or t(8;21). Although patients with NPM1-mutated t-AML had shorter OS than those with de novo AML (1.2 vs 7.0 years, p=0.02), EFS was similar, but the analysis was limited by a small sample size. In the LIT cohort, OS and EFS were comparable between t-AML and de novo AML.
Conclusions:
Favorable-risk t-AML patients treated with intensive chemotherapy had remission rates, relapse rates, and EFS comparable to those of de novo AML. Despite shorter OS in the overall IC cohort, survival was similar in younger patients and, importantly, in both CBF subgroups, contrasting with prior reports of inferior outcomes in CBF t-AML. The observed OS difference in NPM1-mutated t-AML requires validation in larger cohorts. These findings support current guideline recommendations to risk-stratify and treat favorable-risk t-AML according to its genetic profile rather than therapy-related etiology.
DISCLOSURES
Jesus D Gonzalez-Lugo, MD: Kura Oncology - Consultancy (Includes expert testimony)