Paper Details

ABSTRACT

Background
Acute myeloid leukemia (AML) remains associated with poor outcomes, especially in elderly, high-risk, or unfit patients. Venetoclax (VEN) combined with azacitidine (AZA) has become a standard for elderly/unfit patients, while standard “7+3” induction remains the cornerstone for young/fit patients. However, relapse and resistance are still major challenges. Homoharringtonine (HHT) exerts synergistic anti-leukemic effects with VEN and AZA, and may delay VEN resistance. The VAH regimen (VEN+AZA+HHT) has shown promising activity in relapsed/refractory AML. We therefore aimed to evaluate the efficacy and safety of VAH versus VA (VEN+AZA) and versus IA in newly diagnosed AML.

Methods
We retrospectively reviewed newly diagnosed AML patients treated at Henan Provincial People's Hospital from January 2021 to October 2024. Patients receiving VAH, VA, or IA were included. VAH consisted of venetoclax (100 mg d1, 200 mg d2, 400 mg d3–14), azacitidine (75 mg/m² d1–7), and homoharringtonine (1.3 mg/m² d1–7). VA comprised venetoclax (100 mg d1, 200 mg d2, 400 mg d3 for 21–28 days) plus azacitidine (75 mg/m² d1–7). IA comprised idarubicin (10–12 mg/m² d1–3) plus cytarabine (100 mg/m² d1–7). Following 1:1 propensity score matching (PSM), we analyzed efficacy after one induction cycle. Primary endpoint was complete remission with incomplete blood count recovery (CRc). Secondary endpoints included overall response rate (ORR), minimal residual disease (MRD) negativity, relapse-free survival (RFS), overall survival (OS), and safety.

Results
VAH vs. VA
After PSM, 80 patients were analyzed (VAH vs. VA, n=40 each). Baseline characteristics were balanced (all P>0.05). Median age: VAH 57 vs. VA 61 years (P=0.141); male: 70% vs. 55% (P=0.166); ECOG≥2: 40% vs. 60% (P=0.074); ELN 2022 risk: favorable 25% vs. 37.5%, intermediate 37.5% vs. 27.5%, adverse 37.5% vs. 35% (P=0.438). Compared with VA, VAH yielded significantly higher ORR (95% vs. 80%, P=0.043) and CRc (87.5% vs. 57.5%, P=0.003). Among responders, MRD negativity was numerically higher with VAH (77.1% vs. 56.5%, P>0.05). In high-risk patients, VAH achieved significantly higher CRc (93.3% vs. 42.9%, P=0.005) and MRD negativity (85.7% vs. 33.3%, P=0.037); intermediate/low-risk patients showed comparable outcomes. In patients with FLT3-ITD/TKD, NPM1, or DNMT3A mutations, VAH showed higher CRc trends. With median follow-up 19 months in VAH, median OS was 13 months and RFS 9 months. In patients receiving >3 cycles, VAH showed significantly longer OS [24 vs. 13 months] and RFS [23 vs. 10 months] (both P<0.05). Adverse events (AEs) were comparable.

VAH vs. IA
After PSM, 56 patients were analyzed (VAH vs. IA, n=28 each). Baseline characteristics were balanced (all P>0.05). Median age: VAH 56 vs. IA 53 years (P=0.382); male 67.9% vs. 64.3% (P=0.260); ECOG≥2 32.1% vs. 25% (P=0.340); ELN risk: favorable 21.4% vs. 32.1%, intermediate 42.9% vs. 21.4%, adverse 35.7% vs. 46.4% (P=0.097). VAH showed significantly higher ORR (92.8% vs. 67.9%, P=0.019); CRc was numerically higher (85.7% vs. 67.9%, P>0.05). MRD negativity among responders was numerically higher (75% vs. 52.6%,P>0.05). Subgroup analysis showed that high-risk patients in the VAH group had significantly higher ORR (100% vs. 61.5%, P=0.046). Intermediate-risk patients in the VAH group also had higher ORR (91.7% vs. 33.3%, P=0.022). VAH showed significantly higher ORR in patients with >50% bone marrow blasts (100% vs. 63.2%, P=0.013), and in those with platelets ≤100×10?/L  at diagnosis (CRc 88.5% vs. 63.6%, ORR 96.2% vs. 63.6%, both P<0.05). Patients with NPM1, DNMT3A, and FLT3-ITD/TKD mutations showed CRc improvements with VAH (P>0.05). Median OS: VAH 14 vs. IA 19 months (P>0.05); median RFS: 13 vs. 9 months (P>0.05). VAH had significantly higher granulocyte and hemoglobin nadirs than IA (both P<0.05), suggesting milder hematologic toxicity. Non-hematologic AEs were manageable in both groups.

Conclusions
(1) VAH demonstrated significantly higher ORR and CRc compared with VA, with particularly pronounced benefit in high-risk patients, and comparable safety. (2) VAH showed significantly higher ORR and favorable trends in CRc and MRD negativity compared with IA, with advantages in high/intermediate-risk patients, high blast burden, and thrombocytopenia, and potentially milder hematologic toxicity. (3) VAH may represent a promising novel induction option across both intensive and low-intensity settings, warranting further prospective validation.