Molecular Characterization of Chemoresistant Cell Populations Improves Risk Stratification and Therapy Response Prediction in Pediatric AML [pAML PDX Bulk RNA-seq]

Most pediatric acute myeloid leukemia (pAML) patients achieve complete remission after chemotherapy. However, relapse remains common, and nearly 40% ultimately die of the disease. Prognosis is guided by cytogenetic biomarkers and measurable residual disease (MRD) after one chemotherapy cycle, at which point the highest-risk patients are referred to targeted experimental therapies and stem cell transplantation (SCT). Because these therapies are most effective before relapse, accurate early risk prediction is essential for improving outcomes. To address help this need, we analyzed paired diagnosis–relapse samples from 33 pAML patients at single-cell resolution and identified chemoresistant pAML populations whose abundance at diagnosis significantly improved risk prediction. Incorporation of these populations into a risk model uncovered a previously unrecognized subgroup—representing 20% of pAML cases—with event-free survival (EFS) below 40%, accounting for half of the deaths among patients that were not prescribed SCT after a cycle of chemotherapy. Furthermore, molecular characterization of these chemoresistant populations revealed therapeutic targets and targeted interventions relevant to most high-risk patients, paving the way for more effective treatments for high risk pAML patients. Overall design: We profiled paired diagnosis and relapse samples from pAML PDX models using bulk RNA sequencing technology.