Rabia Shahswar

Venetoclax and azacitidine (ven-aza) is a standard of care therapy for newly diagnosed acute myeloid leukemia (AML) patients unfit to receive intensive therapy, and may also be appropriate for certain fit patients. Models predicting response to ven-aza currently focus predominantly on genetics. Here, we present two multimodal fusion models: a custom ResNet-34-based model and a pathology foundation model fine-tuned for bone marrow smear (BMS) images that combine clinical and genetic variables with whole slide images. Training on 193 newly diagnosed AML patients who received ven-aza, we obtained an area-under-the-receiver-operating-characteristic (AUROC) of 0.79 and 0.82 for the ResNet-based and fine-tuned foundation models in a heldout test set, and 0.76 and 0.73 in an external validation set, predicting best response to this therapy. Sensitivity analysis showed that fusing image data with clinical and genetic variables outperformed training on either modality alone. In univariable and multivariable analyses, image-based classification was an independent predictive marker, second only in effect size to IDH1/2 mutation status, outperforming other genetic alterations. Image-based explainability showed high model attention on myeloblasts, suggesting previously unrecognized cell-specific patterns associated with response. Our findings suggest BMS images complement genetic biomarkers to refine outcome prediction in ven-aza treatment and incentivize AI-supported outcome prediction, including bone marrow morphology.

MAC-Score predict response within mutational subgroups

Cell line sensitivity and additional patient data

Subpopulation specific Bcl-2 family levels and time on HMA/Venetoclax

Comparison of MAC-Score and BH-3 profiling based response prediction

Bcl-2 family gene expression in LSCs predicts in vitro response to HMA/VEN

7 Tables showing: molecular and clinical patient data, cell line info, antibody panels and overview of assays used

Transcriptomic and functional characterization of LSC-like cells

Subpopulation specific Bcl-2 family levels and response to HMA/Venetoclax

Chronic kidney disease (CKD) before treatment and renal function decline during treatment are common in elderly patients receiving hypomethylating agents (HMAs) and venetoclax (VEN). This retrospective multicenter study of 130 newly diagnosed older acute myeloid leukemia (AML) patients evaluated the prognostic impact of renal function before and during the first treatment cycle. A total of 56 patients (43%) had CKD, and 49 (38%) developed acute kidney injury (AKI) during treatment. AKI occurrence was associated with laboratory tumor lysis syndrome (LTLS). CKD before treatment showed a trend towards a lower overall response rate (ORR, OR 0.5, p = 0.07) in multivariable analysis (MVA) and inferior relapse-free survival (RFS, HR 2.16, p = 0.06) in univariate analysis (UVA), but not MVA (RFS, HR 1.93, p = 0.15). CKD did not affect overall survival (OS) or event-free survival (EFS). AKI during HMA/VEN therapy was associated with significantly higher 30-day and 60-day mortality rates and emerged as an independent prognostic factor for inferior OS (HR 1.86, p = 0.01) and EFS (HR 1.81, p = 0.007). RFS did not differ significantly by AKI status. Sepsis was a more frequent cause of death in patients with vs. without AKI (33% vs. 5%). In conclusion, kidney function is a key prognostic factor in HMA/VEN-treated patients, warranting further study on treatment adjustments and supportive care.