Background. Acute myeloid leukemia (AML) comprises a heterogeneous group of hematologic malignancies with a generally poor prognosis. Secondary AML (sAML) is clinically more aggressive and associated with worse outcomes. This study aimed to compare the efficacy of azacitidine plus venetoclax (AzaVen) in patients with De Novo AML (DN-AML) versus sAML.

Materials and Methods. We analyzed 162 patients with confirmed AML who received at least one cycle of the AzaVen protocol. Seventy-one patients had DN-AML, and 91 had sAML. Both groups were comparable at baseline with regard to demographic characteristics. According to WHO 2022 and ELN 2022 guidelines, sAML was defined by a history of prior cytotoxic or radiotherapy; a preceding hematologic disorder (myelodysplastic syndrome [MDS], chronic myelomonocytic leukemia [CMML], or Philadelphia chromosome–negative myeloproliferative neoplasm [Ph– MPN]); and/or the presence of biologic features, including ≥1 myelodysplasia-associated gene mutation (ASXL1, BCOR, EZH2, SF3B1, etc.) and/or cytogenetic abnormalities (del(5q), del(7q), complex karyotype, etc.). Patients who had received therapy for a preceding hematologic disease (except hydroxyurea) were excluded. Survival curves were estimated by the Kaplan–Meier method, and differences between groups were assessed using the log-rank test. Statistical analyses were performed in RStudio 4.4.0 with the dplyr, survival, and gtsummary packages.

Results. Complete remission (CR) rates were 76.1% (54/71) in DN-AML and 72.5% (66/91) in sAML (p = 0.72). Median time to CR was 1.49 months (IQR, 0.97–2.30) versus 1.41 months (IQR, 0.96–2.32) (p = 0.95). By three months, the cumulative probability of achieving CR was 89% in DN-AML and 86% in sAML (p = 0.81).

Remission duration was significantly longer in DN-AML, with a median of 11.3 months (IQR, 5.4–23.6) versus 7.3 months (IQR, 2.8–11.4) in sAML (p = 0.007). At six months, CR persisted in 88% of DN-AML and 76% of sAML patients (p < 0.001). Median time to relapse was not reached (>19.5 months) in DN-AML but was 10 months in sAML.

Median overall survival (OS) was 22.2 months (95% CI, 12.0–not reached) in DN-AML and 10.0 months (95% CI, 6.3–13.8) in sAML (p = 0.015). Median event-free survival was 12.2 months versus 6.6 months.

During the first six months, sAML patients required significantly more red blood cell transfusions than DN-AML patients (median 2,239 mL [IQR, 686–3,917] vs. 1,017 mL [IQR, 123–3,172]; p = 0.037), while platelet transfusion volumes did not differ significantly (median 1,776 mL [IQR, 260–4,185] vs. 825 mL [IQR, 0–4,018]; p = 0.20).

In an exploratory subgroup analysis by AML subtype, median overall survival varied substantially across groups. Patients with DN-AML had the longest median OS of 22.2 months (95% CI, 12.0–not reached). Among sAML, those with a history of prior cytotoxic chemotherapy or radiotherapy (n = 13) had a median OS of 11.0 months (95% CI, 3.6–not reached); patients with a preceding hematologic disorder (n = 39) had a median OS of 10.2 months (95% CI, 6.6–24.5); and patients with biologic features (n=39) had the shortest median OS of 8.1 months (95% CI, 5.1–17.7). The difference across subgroups did not reach statistical significance (log-rank p = 0.08).

In multivariable Cox analysis, sAML status was not an independent predictor of mortality (HR, 1.35; 95% CI, 0.60–3.06; p = 0.50), whereas minimal residual disease (MRD) positivity conferred a fourfold higher risk of death (HR, 3.95; 95% CI, 1.51–10.3; p = 0.005). Allogeneic hematopoietic stem cell transplantation (allo-HSCT) showed a protective but non-significant trend (HR, 0.40; 95% CI, 0.09–1.77; p = 0.20).

Conclusions. Our findings confirm that sAML patients have poorer outcomes than those with DN-AML when treated with AzaVen. Although sAML patients achieve remission and MRD negativity rates similar to DN-AML, remissions are less durable, reflecting a shallower depth of response. These patients warrant intensified supportive care and early evaluation for transplantation and investigational therapies, particularly if MRD persists.

This content is only available as a PDF.
2025
Sign in via your Institution