Recent progress in genome-based prognostic prediction of pediatric acute myeloid leukemia

Abstract

The treatment outcomes for pediatric acute myeloid leukemia (AML) have been significantly enhanced based on risk-based therapy guided by genomic analysis and treatment response. Recently, progress in genomic analysis has led to genomic abnormalities identification that can be beneficial for prognostic predictions. For instance, in RUNX1::RUNX1T1-positive AML, the presence of KIT exon 17 mutations has been identified as a poor prognostic marker. Pediatric AML cases with KMT2A gene rearrangements have demonstrated varying prognoses, depending on the fusion partner. Furthermore, the analysis of acute megakaryocytic leukemia, a representative pediatric AML subtype, has revealed various genomic abnormalities, including CBFA2T3::GLIS2, NUP98::KDM5A, and KMT2A rearrangements, demonstrating their close association with prognosis. Recently, RNA-seq expression analysis has led to the development of prognostic models, such as the LSC17 and pLSC6 scores based on the gene expression levels of 17 and 6 genes, respectively, which are associated with leukemia stem cells. Single-cell analysis has been attempted, offering insights into cellular hierarchies and potential treatment strategies based on specific cellular fractions. In the future, integration of newly identified genomic abnormalities and prognostic models with treatment responses will result in novel risk stratification development, enabling more precise prognostic predictions and therapeutic development.