Therapy-related AML (t-AML) is a myeloid neoplasm that evolves secondary to cytotoxic therapy for malignant or non-malignant diseases, e.g. chemotherapy and/or radiotherapy, due to DNA damage in hematopoietic progenitor cells. Approximately 5-15 % of adult AML patients are reported to have t-AML. Chemotherapy and/or radiotherapy may cause mutations, leading to clonal hematopoiesis with subsequent selection of resistant clones during AML therapy. Therefore, adverse clinical and genetic features accumulate in t-AML patients. There is ongoing discussion on the impact of t-AML on long-term outcome in AML and the optimal choice of consolidation therapy in these patients. In our article, we retrospectively analyzed clinical and biological characteristics in a large cohort of 1133 AML patients (225 t-AML patients and 908 de novo AML patients) with a median follow-up of 81.8 months.

Our comparative analysis revealed that patients with t-AML generally have a higher risk of death than patients with de novo AML which is most likely caused by a higher frequency of adverse genetic alterations, higher age and more comorbidities in these patients. This applies particularly to patients receiving intensive chemotherapy with curative intent but not to patients receiving palliative chemotherapy or best supportive care. Stratifying intensively treated patients according to the European Leukamia Net (ELN) 2010 risk groups, we showed inferior survival for t-AML in the ELN intermediate and adverse molecular/cytogenetic risk groups. However, in these latter ELN subgroups, t-AML patients were older and had signs of increased chemoresistance in terms of lower response rates. Thus, it seems conceivable that inferior survival in these t-AML subgroups is mediated by additional genetic aberrations that would have been detected by a more comprehensive molecular panel as proposed by the recent ELN 2022 recommendations. Furthermore, “hidden” mutational and clonal complexity due to clonal selection over time is likely to affect OS in t-AML. In contrast, our analysis led to a very interesting alignment of survival rates between t-AML and de novo AML in intensively treated patients of the favorable ELN risk groups, amongst intensively treated elderly AML patients as well as patients under palliative treatment.

Therefore, we conclude that long-term survival in t-AML is primarily determined by genetic and patient-related risk factors rather than by t-AML itself. This is supported by our multivariate analysis of risk factors in the entire cohort and within the different ELN risk groups. Thus, we assume that long-term survival in t-AML is comparable with de novo AML, if baseline characteristics and genetics are balanced. We conclude that t-AML should not be handled as a distinct entity and that therapeutic algorithms in t-AML, in particular with regard to allo-HSCT, should be guided by ELN genetic risk rather than classification as t-AML only. Our data support the WHO and ICC 2022 classifications, which include t-AML as diagnostic qualifier rather than a separate subcategory.