AUTHORS
Sudhesh Kumar, MD, Momna Warraich, MD2, Mahnoor Fatima, MD3, Mahsa Rezasoltani, MD4, Kristen McCullough, PharmD5, Aref Al-Kali, M.D.6, Hassan B Alkhateeb, M.D.7, Kebede Begna, M.D.8, Abhishek Avinash Mangaonkar, MBBS9, Aasiya Matin, MBBS10, Antoine N. Saliba, M.D.11, Mark R. Litzow, MD12, William Hogan, M.B., B.Ch.13, Mithun Shah, MD, PhD14, Mrinal M. Patnaik, MD, MBBS15, Animesh Pardanani, PhD, MBBS16, Talha Badar, MD17, James Foran, FRCPC18, Jeanne Palmer, MD19, Cecilia Arana Yi, M.D.20, Ayalew Tefferi, MD21, Naseema Gangat, MBBS22
AFFILIATIONS
1 Mayo Clinic, Rochester, United States
2 Mayo Clinic, Rochester, United States
3 Mayo Clinic, Rochester, United States
4 Mayo Clinic, Rochester, United States
5 Mayo Clinic, Rochester, United States
6 Mayo Clinic, Rochester, United States
7 Mayo Clinic, Rochester, United States
8 Mayo Clinic, Rochester, United States
9 Mayo Clinic, Rochester, United States
10 Mayo Clinic, Rochester, United States
11 Mayo Clinic, Rochester, United States
12 Mayo Clinic, Rochester, United States
13 Mayo Clinic, Rochester, United States
14 Mayo Clinic, Rochester, United States
15 Mayo Clinic, Rochester, United States
16 Mayo Clinic, Rochester, United States
17 Mayo Clinic, Jacksonville, United States
18 Mayo Clinic, Jacksonville, United States
19 Mayo Clinic, Scottsdale, United States
20 Mayo Clinic, Scottsdale, United States
21 Mayo Clinic, Rochester, United States
22 Mayo Clinic, Rochester, United States, Mayo Clinic, Rochester, United States
ABSTRACT
Background: Patients with newly diagnosed acute myeloid leukemia (ND-AML) treated with venetoclax + hypomethylating agent (Ven-HMA) display gender-specific differences in survival not accounted for by current prognostic models, including the Mayo Genetic Risk Model (Am J Hematol, 2025) and the European LeukemiaNet (ELN) 2024 model (Blood, 2024). We aimed to characterize these differences and develop gender-stratified prognostic models.
Methods: We retrospectively studied patients with ND-AML treated with frontline Ven-HMA at Mayo Clinic (Rochester, MN; Jacksonville, FL; Scottsdale, AZ) between November 2018 and October 2025. Transplant-censored survival was calculated from the date of Ven-HMA initiation to death, with censoring at the time of allogeneic stem cell transplantation or last follow-up. Cox proportional hazards identified gender-specific independent predictors that were integer-weighted into risk categories. Model performance was assessed by time-dependent area under the receiver operating characteristic curve (AUC) at 12, 24, and 36 months vs the Mayo Genetic Risk Model and the ELN 2024 model. Statistical analyses were performed using R.
Results: We included 540 patients with ND-AML: 339 males (M; 62.8%) and 201 females (F; 37.2%). Baseline characteristics were comparable: median age (M 75 vs. F 75 years), de novo AML (M 55.2% vs. F 48.3%), and ELN 2022 adverse-risk karyotype (M 35.1% vs. F 35.3%). Risk-model distributions (low/intermediate/high) were similar by gender: Mayo- M 8.0/49.9/42.2% vs. F 9.5/52.2/38.3%; ELN- M 59.3/18.3/22.4% vs. F 61.2/18.4/20.4%. Mutation frequencies were similar except for male-enriched ASXL1 (M 63 [18.6%] vs. F 16 [8.0%]; p = 0.001), U2AF1 (M 22 [6.5%] vs. F 1 [0.5%]; p = 0.001), SRSF2 (M 66 [19.5%] vs. F 22 [10.9%]; p = 0.011), and SETBP1 (M 13 [3.8%] vs. F 1 [0.5%]; p = 0.022).
Complete remission with or without count recovery (CR/CRi) was similar (M 63.7% vs. F 67.7%; p = 0.40), but transplant-censored survival was longer in females: median M 11.2 vs. F 17.4 months (p = 0.012), with 1-/2-/3-year rates 46.9%/26.5%/19.1% in M vs. 62.0%/34.4%/23.3% in F.
In males, five independent predictors of transplant-censored survival emerged: SETBP1 mutations (hazard ratio [HR] 2.98, p = 0.0005), ELN-adverse karyotype (HR 1.93, p = 0.0006), and TP53 mutations (HR 1.72, p = 0.007) as adverse; DDX41 (HR 0.29, p = 0.007) and IDH2 mutations (HR 0.55, p = 0.025) as protective. In females, NPM1 (HR 0.42, p = 0.015) and IDH2 mutations (HR 0.40, p = 0.022) emerged as protective. TP53 mutations and ELN-adverse karyotype were independent adverse predictors when modeled separately (TP53 mutation HR 1.93, p = 0.002; ELN-adverse karyotype HR 1.74, p = 0.005) but neither remained significant together. TP53 mutation was selected given its higher HR (1.93 vs. 1.74) and lower p-value (0.002 vs. 0.005).
Predictors were integer-weighted (+1 each) into gender-specific risk scores. The male 5-variable score (DDX41-wild-type [WT] + ELN-adverse + TP53-mutated + IDH2-WT + SETBP1-mutated) was stratified into low-risk (1 point; n = 37; median survival 39.8 months; 1-/2-/3-year 83/69/53%), intermediate-risk (2 points; n = 166; 15.1 months; 59/33/23%), and high-risk (≥3 points; n = 136; 6.0 months; 20/2/2%). Pairwise HRs were 2.38 (1.38–4.10, p = 0.002) for low- vs. intermediate-risk and 6.95 (3.97–12.17, p < 0.0001) for low- vs. high-risk. The female 3-variable score (TP53-mutated + NPM1-WT + IDH2-WT) was stratified into low-risk (0–1 points; n = 43; 58.6 months; 83/62/57%), intermediate-risk (2 points; n = 117; 17.4 months; 61/33/18%), and high-risk (3 points; n = 41; 8.0 months; 41/10/3%). Pairwise HRs were 2.93 (1.64–5.25, p < 0.001) for low- vs. intermediate-risk and 5.62 (2.97–10.66, p < 0.0001) for low- vs. high-risk.
Time-dependent AUC at 12/24/36 months favored the gender-specific risk scores (M 0.74/0.80/0.74; F 0.65/0.70/0.76) over the Mayo Genetic Risk Model (M 0.72/0.78/0.75; F 0.65/0.65/0.71) and the ELN 2024 model (M 0.65/0.71/0.65; F 0.64/0.65/0.61).
Conclusion: Current prognostic models for ND-AML on Ven-HMA do not account for gender and overlook mutations that lead to different survival outcomes despite similar CR rates. Because distinct predictors emerged in males and females, gender-stratified risk scores more accurately predicted survival within each gender. Additional studies are needed to validate these findings.
DISCLOSURES
Sudhesh Kumar, MD: No relevant conflicts of interest to declare.