2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Saturday, December 12, 2026
05:30 p.m. - 07:30 p.m.
MCCNO - Hall F
Technology and Procedures, Assays, Molecular Testing, Omics Technologies, Pathology, Bioinformatics, Computational Biology, Emerging Technologies, Profiling, Treatment Considerations, Combination Therapy, Supportive Care, Biological Therapies, Antibody Therapy, Natural Killer (NK) Cell Therapies, Transplantation, (CAR)-T Cell Therapies, Immunotherapy, Non-Biological Therapies, Chemotherapy, Artificial Intelligence (AI), Machine Learning, Biological Processes, Apoptosis, Pathogenesis, Genomics, Immune Mechanism, Metabolism, Molecular Biology, Multi-Systemic Interactions, Clinical Practice (Health Services and Quality), Clinical Guidelines, Practice Models, Systems-Based Hematology, Diseases, Bone Marrow Failure Syndromes, Acquired Marrow Failure Syndromes , Genetic Disorders, Myeloid Malignancies, Acute Myeloid Malignancies, Chronic Myeloid Malignancies, Measurable Residual Disease, Research, Fundamental Science, AML, Translational Research, MDS, MPN, Clinical Research, Patient-Reported Outcomes, Real-World Evidence, Registries, Study Population, Human, Adult, Elderly, Pediatric, Young Adult
PAPER

1688 Machine learning-based prognostic stratification in older adults with AML treated with venetoclax-based regimens

Wenjing Shu, PhD, Jianfeng Li, PhD2, Xinjie Chen3, Wen-Yan Cheng4, Xiaojing Lin Sr, PhD5, Yuncan Cao, MD6, Jiong Hu, MD7, Shen Yang, MD8
1 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
2 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
3 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
4 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
5 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
6 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
7 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China
8 Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, shanghai, China

Background: Venetoclax (VEN)-based low-intensity regimens are standard for older adults with acute myeloid leukemia (AML) who are ineligible for intensive chemotherapy, yet responses and survival outcomes remain heterogeneous. We aimed to develop an integrated clinical-transcriptomic model to predict event-free survival (EFS), identify candidate VEN resistance biomarkers, and explore synergistic combination strategies for molecularly high-risk disease.

Methods: We retrospectively included 178 older adults with AML who were considered unfit for intensive chemotherapy and treated with VEN-based regimens at Ruijin Hospital between 2019 and 2024; RNA sequencing data were available for 157 patients. Candidate predictors included disease characteristics, Eastern Cooperative Oncology Group (ECOG) performance status, cytogenetics, mutational profile, VEN exposure intensity, and first-cycle VEN duration. Within each training set, clinical variables were screened by univariable Cox regression with false-discovery rate (FDR) correction, while transcriptomic features were reduced by high-variance filtering followed by supervised ranking based on absolute residual correlation. Seven survival-learning algorithms—elastic-net Cox, random survival forest, CoxBoost, gradient-boosted Cox, XGBoost-Cox, survival support vector machine, and SuperPC-Cox—were evaluated using repeated nested five-fold cross-validation. Performance was assessed using C-index, time-dependent area under the curve (AUC), Brier score, calibration, decision curves, and Kaplan–Meier stratification. Candidate genes were independently evaluated in an ex vivo cohort of 100 primary AML bone marrow mononuclear cell samples by correlating expression with VEN IC50. Combination indices were analyzed in RNA8-defined high-risk samples with reduced VEN sensitivity.

Results: After a median follow-up of 16 months, the complete remission rate was 68.0%, median overall survival was 16 months, and median EFS was 9 months; 1- and 2-year overall survival rates were 62% and 43%. The median VEN treatment durations for cycles 1 through 6 were 22, 21, 21, 14, 14, and 14 days, respectively. Neither 6-month cumulative VEN exposure nor first-cycle VEN duration, categorized as 7–13, 14–20, 21–27, or ≥28 days, was associated with survival in exploratory analyses. SuperPC-Cox showed the strongest overall performance. Cross-validated C-indices were 0.756 (95% CI, 0.705–0.802) for Clinical+RNA8, 0.739 (95% CI, 0.684–0.789) for RNA8, and 0.566 (95% CI, 0.492–0.642) for the clinical-only model. Higher ECOG score, monosomal karyotype, and TP53 mutations were associated with increased EFS risk, whereas NPM1 mutations were associated with lower risk; VEN exposure and first-cycle duration were not retained. RNA8 comprised the favorable genes LINC00861, EDA2R, and NELL2 and the adverse genes H1-3, SH3PXD2B, PODXL2, DPYSL2, and ARPIN-AP3S2. RNA8 achieved 12- and 24-month AUCs of 0.828 (95% CI, 0.752–0.904) and 0.786 (95% CI, 0.662–0.909), respectively. Patients with an RNA8-high signature had significantly inferior EFS (HR=3.77, P<0.001). In independent ex vivo validation, higher expression of SH3PXD2B, DPYSL2, and ARPIN-AP3S2 correlated with increased VEN IC50 (all P<0.05). Among RNA8-high, VEN-resistant samples, addition of homoharringtonine, camptothecin, idarubicin, oridonin, and selinexor to VEN+AZA produced the lowest median combination indices, suggesting enhanced ex vivo synergy.

Conclusions: The SuperPC-Cox model integrating clinical characteristics, mutational profiles, and gene expression significantly improved EFS risk stratification compared with clinical variables alone in older adults with AML receiving VEN-based regimens. VEN exposure intensity was not significantly associated with long-term survival in this cohort. Machine learning and ex vivo validation identified SH3PXD2B, DPYSL2, and ARPIN-AP3S2 as candidate VEN resistance biomarkers. Drug screening further nominated homoharringtonine, camptothecin, idarubicin, oridonin, and selinexor as potential partners to enhance VEN+AZA activity in resistant samples, warranting further preclinical and prospective clinical validation.

 

Wenjing Shu, PhD: No relevant conflicts of interest to declare.