A large retrospective study from China has provided one of the most detailed real-world pictures yet of how patients with acute myeloid leukemia respond to venetoclax-based treatment, and it comes with a bonus: a statistical model that can predict, before therapy even finishes its first cycle, which patients are likely to achieve a complete remission. The research, published in Annals of Hematology by a team led by Anan Wang and Bei Liu at the First Hospital of Lanzhou University, followed 216 patients treated between March 2020 and May 2025, spanning both newly diagnosed cases and the far harder problem of relapsed or refractory disease.
Acute myeloid leukemia is an aggressive cancer of the blood and bone marrow in which abnormal myeloid cells proliferate uncontrollably, crowding out healthy blood-forming tissue. For decades, the backbone of treatment was intensive chemotherapy, which many patients, particularly older adults, could not tolerate. The arrival of venetoclax changed that calculus. Venetoclax is a selective inhibitor of BCL-2, a protein that leukemia cells exploit to evade programmed cell death. By binding BCL-2, venetoclax restores the apoptotic pathway, priming malignant cells to self-destruct. When combined with hypomethylating agents such as azacitidine or with low-dose cytarabine, it has produced remission rates in clinical trials that were unthinkable for unfit or elderly patients a decade ago.
The Lanzhou study captured how this class of drugs performs outside the tightly controlled environment of a randomized trial. Among the 165 patients in the cohort with newly diagnosed acute myeloid leukemia, 74.5 percent achieved what clinicians call composite complete remission, a measure that includes complete remission itself as well as complete remission with incomplete blood count recovery. That figure aligns closely with results from pivotal registration trials, suggesting that the drug combination retains its efficacy when delivered in routine clinical practice, where patients are often older, carry more comorbidities, and present with more heterogeneous disease biology than trial enrollees.
The picture was markedly different for the 51 patients with relapsed or refractory disease, whose leukemia had either returned after prior treatment or had never responded to it. In this group, only 39.2 percent achieved composite complete remission. That gap is clinically meaningful and underscores a central challenge in modern hematology: while venetoclax has transformed first-line therapy, salvage treatment for patients whose disease has escaped earlier regimens remains one of the most difficult problems in the field. Leukemia cells that survive initial therapy often carry mutations and survival adaptations that blunt the effectiveness of BCL-2 inhibition, and the study’s numbers quantify that resistance in a real-world setting.
Survival analysis revealed which factors determined long-term outcomes in each group. Among patients with newly diagnosed disease, three variables emerged as independent predictors of shorter overall survival: failure to achieve composite complete remission, older age, and the presence of an NF1 mutation. The NF1 gene encodes neurofibromin, a negative regulator of the RAS signaling pathway, and mutations in it are associated with dysregulated cell growth and, in some contexts, with resistance to certain targeted therapies. Its identification here as an adverse marker adds to a growing body of evidence that the mutational landscape of a patient’s leukemia shapes how much benefit venetoclax-based regimens can deliver.
In the relapsed or refractory group, the determinants of response were different. A lower pretreatment white blood cell count and a higher bone marrow blast percentage were associated with achieving composite complete remission, an apparently counterintuitive pairing that the authors report as an observational finding from multivariable analysis. Meanwhile, failure to achieve remission in this group was strongly linked to higher mortality, reinforcing the point that for relapsed patients, the first salvage attempt is often the best chance to regain disease control. Together, these findings suggest that clinicians managing relapsed disease should weigh disease burden and inflammatory parameters carefully when selecting and timing venetoclax-based salvage therapy.
The most technically ambitious component of the study was the construction of a predictive model for early remission. The researchers assembled a multivariable model using six predictors: AML status, meaning whether disease was newly diagnosed or relapsed or refractory; age; bone marrow blast percentage; platelet count; Eastern Cooperative Oncology Group performance status, a standard measure of a patient’s functional fitness; and TET2 mutation status. TET2 is an epigenetic regulator frequently mutated in myeloid malignancies, and prior research has suggested that TET2-mutant leukemia cells may be particularly sensitive to BCL-2 inhibition, making its inclusion in the model biologically coherent rather than merely statistical.
The model’s performance was quantified using the area under the receiver operating characteristic curve, a metric that expresses how well a model discriminates between patients who respond and those who do not, with 0.5 representing chance and 1.0 representing perfect prediction. The apparent AUC was 0.841, with a 95 percent confidence interval of 0.785 to 0.897, indicating good discriminative ability. Crucially, the team did not rely on a single fit of the data. They internally validated the model using 1,000 bootstrap resamples, a resampling technique that repeatedly refits the model on random subsets of the data to estimate how much its performance is inflated by overfitting. This kind of validation is considered best practice in clinical prediction modeling and lends credibility to the reported statistics.
The model’s predictive reach extended beyond remission itself. When the researchers applied a fixed CRc score to survival outcomes, it yielded time-dependent AUCs of 0.788 at six months and 0.695 at twelve months for overall survival. In practical terms, this means the score captured early treatment trajectory well, with its discriminatory power gradually attenuating over the first year, as would be expected for a model built around baseline and early-treatment variables. The authors are appropriately cautious about these results, noting that the model may be useful for estimating early treatment response but that external validation in independent cohorts is needed before it can be applied routinely in clinical decision-making.
The significance of this work lies less in any single number than in its demonstration of how precision oncology is maturing. Venetoclax-based therapy is now a global standard for many patients with acute myeloid leukemia, yet response remains unpredictable at the individual level, and the drug carries costs, both financial and toxicological, including the risk of tumor lysis syndrome and severe cytopenias. A validated tool that integrates disease status, age, blast burden, platelet count, performance status, and clonal genetics could help oncologists identify patients unlikely to benefit from a first cycle, prompting earlier consideration of alternative strategies or clinical trial enrollment. For now, the Lanzhou cohort offers clinicians a rigorous real-world benchmark: roughly three in four newly diagnosed patients and two in five relapsed patients can expect a remission with venetoclax-based regimens, and a growing statistical toolkit is making those odds calculable before the first dose is even given.
Subject of Research: Venetoclax-based therapy outcomes and remission prediction modeling in acute myeloid leukemia
Article Title: Retrospective evaluation of venetoclax-based regimen in newly diagnosed and relapsed/refractory acute myeloid leukemia: Efficacy determinants, survival outcomes, and model validation
Article References: Wang, A., Liang, H., Ren, C., Wang, Y., Guo, J., Li, J., Yu, X., & Liu, B. (2026). Retrospective evaluation of venetoclax-based regimen in newly diagnosed and relapsed/refractory acute myeloid leukemia: Efficacy determinants, survival outcomes, and model validation. Annals of Hematology. https://doi.org/10.1007/s00277-026-07258-9
Image Credits: AI Generated
DOI: 10.1007/s00277-026-07258-9
Keywords: acute myeloid leukemia, venetoclax, composite complete remission, BCL-2 inhibitor, prediction model, overall survival, NF1 mutation, TET2 mutation, relapsed refractory AML, retrospective study, hematology, bootstrap validation
Cite Scienmag News
APA MLA Chicago
Nathaniel Bowman. (October 3, 2026). Venetoclax Combination Therapy Shows Strong Remission Rates in Leukemia Study. Scienmag. https://scienmag.com/venetoclax-combination-therapy-shows-strong-remission-rates-in-leukemia-study/
Nathaniel Bowman. “Venetoclax Combination Therapy Shows Strong Remission Rates in Leukemia Study.” Scienmag, 3 October 2026, https://scienmag.com/venetoclax-combination-therapy-shows-strong-remission-rates-in-leukemia-study/. Accessed 3 October 2026.
Nathaniel Bowman. “Venetoclax Combination Therapy Shows Strong Remission Rates in Leukemia Study.” Scienmag. October 3, 2026. https://scienmag.com/venetoclax-combination-therapy-shows-strong-remission-rates-in-leukemia-study/
Copy citation Download RIS
Tags: acute myeloid leukemiaacute myeloid leukemia remissionBCL-2 inhibitorBCL-2 inhibitor in leukemiabootstrap validationchemotherapy alternatives for AMLcomposite complete remissionhematologyhypomethylating agents in leukemialeukemia treatmentleukemia treatment in older adultsNF1 mutationoverall survivalprediction modelpredictive remission modelingreal-world leukemia studyrelapsed or refractory AMLrelapsed refractory AMLretrospective studyretrospective study on AMLTET2 mutationvenetoclaxVenetoclax combination therapyvenetoclax-based therapy
