Introduction: The Shift in Acute Myeloid Leukemia Management
For decades, the standard of care for acute myeloid leukemia (AML) was defined by intensive induction chemotherapy. However, for a significant portion of the patient population—particularly the elderly or those with substantial comorbidities—intensive treatment was often intolerable. The landscape of hematologic oncology changed dramatically with the emergence of Venetoclax (VEN) in combination with hypomethylating agents (HMA), such as azacitidine or decitabine. This combination has demonstrated high rates of complete remission and improved overall survival (OS) in patients previously deemed ineligible for intensive therapy.
Despite these advances, clinicians have long observed a frustrating variability in patient outcomes. Some patients achieve durable, multi-year remissions, while others progress rapidly despite initial response. Existing prognostic frameworks, primarily the European LeukemiaNet (ELN) risk classifications, were largely validated in the context of intensive chemotherapy. There is a pressing clinical need for tools that specifically account for the genomic landscape of AML in the era of BCL-2 inhibition.
Study Design: Analyzing the Genomic Landscape
To address this gap, researchers at the Moffitt Cancer Center analyzed a substantial cohort of 506 patients with newly diagnosed AML who received frontline treatment with HMA/VEN. The goal was to construct a robust, evidence-based prognostic model by integrating mutational and cytogenetic data.
The researchers employed multivariate analysis to determine which specific genetic lesions independently impacted overall survival. After identifying these variables, they developed a point-based scoring system to facilitate clinical decision-making. The model was then subjected to external validation using an independent cohort of 126 patients to ensure its generalizability across different clinical settings.
Key Findings: Identifying High-Risk and Favorable Markers
The Impact of Somatic Mutations
The study identified several key mutations that significantly reduced the efficacy of HMA/VEN therapy. Mutations in TP53, KRAS, JAK2, U2AF1, and CBL were found to be independent predictors of inferior overall survival. TP53 mutations, in particular, remain a significant hurdle in AML therapy, often associated with complex karyotypes and extreme resistance to DNA-damaging agents and BCL-2 inhibitors alike. The inclusion of signaling pathway mutations like KRAS and JAK2 suggests that alternative survival pathways may allow leukemic blasts to bypass BCL-2 dependence.
Conversely, mutations in IDH1 and IDH2 were identified as favorable prognostic markers. This finding is consistent with the biological understanding that IDH-mutant AML creates a metabolic state characterized by an accumulation of 2-hydroxyglutarate, which enhances BCL-2 dependency, making these cells particularly sensitive to Venetoclax.
Cytogenetic Abnormalities
Beyond single-gene mutations, the model integrated broader chromosomal changes. Inferior OS was predicted by the presence of:
– del(7q) or monosomy 7
– del(17p), monosomy 17, or isochromosome 17q
– del(20q)
– MECOM rearrangements
These cytogenetic markers often reflect deep-seated genomic instability, which correlates with a more aggressive disease biology and a higher propensity for clonal evolution under treatment pressure.
Risk Stratification and Survival Outcomes
The resulting point-based system successfully stratified patients into three distinct risk categories with vastly different survival trajectories. In the primary cohort, the outcomes were as follows:
– Low-Risk Group: Median OS of 54.2 months.
– Intermediate-Risk Group: Median OS of 22.3 months.
– High-Risk Group: Median OS of 7.5 months.
The statistical separation between these groups was highly significant (p < 0.0001). Furthermore, the model achieved a C-index of 0.648, indicating a strong predictive performance for a clinical prognostic tool.
Validation and Comparative Performance
A critical component of this research was the comparison of the Moffitt model against existing HMA/VEN-specific prognostic tools. While existing models have provided some utility, the new model demonstrated superior discrimination, particularly in distinguishing between low- and intermediate-risk patients. The researchers noted a 31.9-month separation in survival between these two groups (p = 0.002), a distinction that other models often fail to capture clearly. The external validation cohort confirmed these findings, maintaining a clear prognostic separation with a C-index of 0.626.
Expert Commentary: Clinical Implications and Future Directions
This model provides clinicians with a powerful framework for personalized medicine in AML. By identifying high-risk patients early—those with a median survival of only 7.5 months—physicians can more aggressively pursue clinical trials, novel combinations, or early transition to allogeneic hematopoietic stem cell transplantation (HSCT) where feasible.
However, the study also highlights the persistent challenge of the ‘High-Risk’ genomic profile. The poor outcomes for patients with TP53 mutations or MECOM rearrangements suggest that even the potent HMA/VEN combination is insufficient for these subsets. Future research must focus on identifying the specific resistance mechanisms at play in these patients, perhaps by targeting the MCL-1 pathway or utilizing immune-based therapies.
One limitation noted is that while the model is highly effective at baseline, it does not yet incorporate dynamic data such as Measurable Residual Disease (MRD) status after one or two cycles of therapy. Integrating genomic risk at baseline with MRD response could lead to an even more refined ‘dynamic prognostic’ tool in the future.
Conclusion: A New Tool for Precision Hematology
The integration of mutational and cytogenetic data into a cohesive prognostic model represents a significant advancement in the management of AML. By moving beyond a ‘one-size-fits-all’ approach to HMA/VEN therapy, this model allows for more accurate expectations and more tailored therapeutic strategies. As the field moves toward more personalized care, such validated tools will be essential in navigating the complex genomic landscape of myeloid malignancies.
References
Drekolias D, Tuz Zahra F, Fileni C, Sallman DA, Mo Q, Chan O, Zhang L, Vincelette ND, Yu X, Sammut R, Moon J, Park J, Umasangtongkul SA, Lledo FM, Razabdouski TN, Cheng CH, Qin D, Mai K, Ball S, Shallis RM, Xie Z, Kuykendall AT, Padron E, Sweet K, Walker AR, Komrokji RS, Lancet JE, Niyongere S, Cluzeau T, Yun S. Prognostic Model Combining Mutational and Cytogenetic Profiles in Acute Myeloid Leukemia Treated with Venetoclax and Hypomethylating Agents. Blood Cancer Discov. 2026 Feb 11. doi: 10.1158/2643-3230.BCD-25-0193. PMID: 41671569.