Using genomics to refine pediatric AML risk stratification

Abstract In the past 20 years, advances in genomic technologies have greatly improved our understanding of pediatric acute myeloid leukemia (AML). Today, cytogenetic tests can detect structural changes in approximately 75% of cases and remain a main tool for assessing risk. Recent technologies, such as next-generation sequencing, are revealing additional structural alterations (cryptic fusions) and mutations that often cooperate to influence disease biology and treatment response. This evolving genetic landscape has identified unique childhood subtypes of AML defined by specific fusions, such as NUP98::NSD1, CBFA2T3::GLIS2, and varied KMT2A rearrangements, which are linked to distinct clinical outcomes. Emerging data also point to the poor prognosis associated with certain subtypes of NPM1, like the NPM1-D isoform. Additionally, mutations in genes like WT1, DNMT3A, and TP53, the latter of which are rare in childhood AML, may influence patients' outcomes, particularly when occurring in combination. Targeted therapies, including FLT3, BCL2, and menin inhibitors, are beginning to reshape treatment, offering more personalized approaches. However, integrating these drugs effectively into the patient's treatment strategy remains challenging due to the genetic complexity and rarity of pediatric AML. Key issues ahead include identifying which genetic features truly affect outcomes, using this information to personalize therapy, predicting who will benefit from targeted drugs, and choosing the best markers to track disease response over time. Looking forward, collaborative efforts are urgently needed to validate pediatric-specific biomarkers, test novel drug combinations, and link genetic data to clinical outcomes to design trials and future treatment strategies.