AUTHORS
Christopher A Eide, Stephen E Kurtz, PhD2, Andy Kaempf, MS3, Nicola Long4, Alexandria Miller5, Daniel Bottomly, MS6, Shannon McWeeney, PhD7, Curtis Lachowiez, MD8, Sudarshan Anand9, Stanley WK Ng10, Jean Wang, MD11, John E. Dick, PhD12, Jeffrey Tyner, PhD13
AFFILIATIONS
1 Oregon Health & Science University, Portland, Oregon, United States
2 Oregon Health & Science University, Portland, Oregon, United States
3 Oregon Health & Science University, Portland, Oregon, United States
4 Oregon Health & Science University, Portland, Oregon, United States
5 Oregon Health & Science University, Portland, Oregon, United States
6 Oregon Health & Science University, Portland, Oregon, United States
7 Oregon Health & Science University, Portland, Oregon, United States, Oregon Health & Science University, Portland, Oregon, United States
8 Oregon Health & Science University, Portland, Oregon, United States
9 Oregon Health & Science University, Portland, Oregon, United States
10 UC Irvine School of Medicine, Irvine, California, United States
11 University Health Network, Toronto, Ontario, Canada, University of Toronto, Toronto, Ontario, Canada
12 University Health Network, Toronto, Ontario, Canada, University of Toronto, Toronto, Ontario, Canada
13 Oregon Health & Science University, Portland, Oregon, United States, Oregon Health & Science University, Portland, Oregon, United States, Oregon Health & Science University, Portland, Oregon, United States
ABSTRACT
Introduction: Acute myeloid leukemia (AML) is characterized by considerable genetic heterogeneity resulting from various mutations and chromosomal translocations that contribute to disease and influence duration of response to treatments (Siegel, Cancer Statistics 2020). The standard-of-care for fit patients with minimal comorbidities is intensive chemotherapy (IC) induction. For older and less fit patients ineligible for IC, standard-of-care is a hypomethylating agent combined with the BCL2 inhibitor venetoclax (HMA+Ven). Although response rates with HMA+Ven induction in newly diagnosed (ND) AML patients are nearly 70%, median progression-free and overall survival are 10 and 15 months (mos), respectively (DiNardo, Blood 2020). Recent findings have also shown promising activity of HMA+Ven induction in younger, fit patients (Fathi, Blood 2025). These outcomes highlight the importance of defining the features of patients who best respond to treatment.
AML risk classification based on European LeukemiaNet (ELN) 2022 guidelines, used in predicting outcomes for patients treated with IC, relies on the presence of cytogenetic and mutational features. However, the prognostic value of these risk categories is not retained for patients treated with HMA+Ven, leading to development of new risk classifiers for these patients. A 4-gene mutation classifier (mPRS; Dohner, Blood 2024), which stratifies overall survival (OS) for HMA+Ven across low, intermediate, and high benefit groups, was validated in an independent cohort (Bataller, Blood Adv 2024). Subsequent risk model refinements include ELN2024 Less-Intensive (Dohner, Blood 2024), ELN2024 Refined (Lachowiez, Blood 2024), Beat AML Genetic Risk (Hoff, Blood Neoplasia 2025) and PRISM-3 (Lachowiez, JCO 2026). However, the use of mutational profiling to guide HMA+Ven treatment is imperfect due to sparsity and heterogeneity of mutational events. The incorporation of gene expression has the potential to further improve prognostic stratification for AML.
Methods: Primary specimens from ND AML patients receiving HMA+Ven induction were assessed using a rapid qPCR-based assay (Myelo-ID). Bulk RNA isolated from peripheral blood or bone marrow (BM) mononuclear cells at diagnosis were evaluated for BCL2 expression and classified as high or low based on the cohort’s median expression level. Composite Complete Remission (cCR) was defined as achieving a response of CR, CRh, or CRi while on HMA+Ven. Patients with early death (<60 days from HMA+Ven start) but no post-treatment BM exam were considered non-responders. OS and event-free survival (EFS) were measured from start of HMA+Ven. An EFS event was the earliest of the following: failure to achieve <5% blasts in BM while on HMA+Ven for up to 2 cycles, disease relapse, or death. Patients without an EFS event were censored on their last known alive date. Median survival was estimated by the Kaplan-Meier (KM) method and median follow-up (FUP) by reverse KM. With risk classifiers and BCL2 expression (low vs high) as predictors, cCR was modeled with logistic regression while EFS and OS were analyzed using Cox regression, with area under the ROC curve and Harrell’s C-index used to quantify discrimination, respectively. C-indices were compared across models using bootstrap resampling (1000 replications).
Results: In our 100-patient cohort, median age was 71 (range 19-86), 57% were male, 71% had de novo disease, and ELN2024 Less-Intensive risk was favorable in 50%, intermediate in 35%, and adverse in 15%. cCR rate was 67%, median EFS was 7.1 mos, and median OS was 13.6 mos over a median FUP of nearly 3 years. Patients with high levels of BCL2 had considerably longer OS (median 21.2 vs 7.4 mos, HR=0.50 [95% CI=0.31-0.82], p=0.006) and EFS (median 13.1 vs 4.6 mos, HR=0.52 [95% CI=0.32-0.82], p=0.006). Adding BCL2 expression markedly increased the ability to predict each clinical outcome compared to published risk models (mPRS, ELN2024 Less-intensive, ELN2024 Refined, Beat AML Genetic Risk, PRISM-3). With ELN2024 Less-Intensive as the comparator, including a patient’s BCL2 level significantly increased discrimination of OS (C-index 0.638 vs 0.590, p=0.036) and EFS (0.630 vs 0.576, p=0.020).
Conclusion: A qPCR-based method assessing BCL2 gene expression provides a rapid companion diagnostic test for AML patients eligible for HMA+Ven. Incorporating BCL2 level in the prospective evaluation of ND AML patients enhances the prognostic value of risk models.
DISCLOSURES
Christopher A Eide: No relevant conflicts of interest to declare.