Journal: Blood advances
This retrospective cohort study analyzed 2,684 intensively treated adults with newly diagnosed AML, comparing:
- Clinically defined secondary AML after antecedent MDS
- Molecularly defined “secondary-type” AML (st-AML)
- Molecularly defined MDS/AML with 10–19% blasts (st-MDS/AML)
- Other (presumed de novo) AML
Targeted NGS of 54 recurrently mutated genes was performed at diagnosis in all patients and in 436 patients in complete remission (CR) for measurable residual disease (MRD) assessment.
Key findings:
1. Genetic profile of secondary-type disease
- Enriched mutations in clinically secondary AML: ASXL1, BCOR, EZH2, SF3B1, SRSF2, STAG2, U2AF1, and ZRSR2.
- Additional association: ETV6 was significantly associated with clinically secondary AML.
- These mutations defined the molecular signature for both st-MDS/AML and st-AML.
2. St-MDS/AML and st-AML behave similarly
- No overall survival (OS) difference was observed between molecularly defined st-MDS/AML (10–19% blasts) and st-AML (≥20% blasts).
- This supports their consideration as a single biological category.
3. Prognostic impact versus ELN 2022 risk groups
- Patients with molecularly defined st-AML (including st-MDS/AML) had significantly inferior survival:
- 5-year OS ~40% vs ~70% in ELN favorable-risk.
- 5-year OS ~40% vs ~49% in ELN intermediate-risk.
- This supports secondary-type genetics as a high-risk feature, even within otherwise favorable or intermediate ELN risk categories.
4. MRD assessment
- MRD based only on secondary-type mutations (e.g., myelodysplasia-related genes) in CR did not predict relapse.
- MRD positivity for non-DTA mutations (i.e., excluding typical age-related/clonal hematopoiesis genes like DNMT3A, TET2, ASXL1) in CR was strongly associated with higher cumulative incidence of relapse in st-AML (SHR 3.25; P < .001).
Clinical implications:
- Molecularly defined st-AML (including st-MDS/AML) constitutes a distinct high-risk AML subgroup with characteristic myelodysplasia-related mutations, including ETV6.
- Blast percentage (10–19% vs ≥20%) is less relevant than the secondary-type mutational profile for prognosis.
- NGS-based MRD in this population is prognostically meaningful only when focusing on non-DTA mutations; secondary-type mutations alone are not reliable MRD markers.
- Risk stratification and post-remission strategies for patients with secondary-type genetics should account for this high-risk biology and MRD status.