Identification of cellular hierarchy in paediatric acute myeloid leukaemia: The Japan Children's Cancer Group trial ( <scp>JCCG AML</scp> ‐12)

<jats:title>Summary</jats:title> <jats:p> RNA sequencing from 262 patients with paediatric acute myeloid leukaemia (AML) (JPLSG AML‐12) was deconvoluted employing adult single‐cell RNA‐sequencing signatures and Zeng's method, which defined five cellular hierarchy subtypes: primitive, leukaemic stem/progenitor cell (LSPC)‐Cycle, ProMono‐like, granulocyte–monocyte progenitor (GMP)‐like and intermediate. Principal component analysis revealed two main axes that distinguish paediatric from adult AML, with notable LSPC‐Cycle and ProMono‐like phenotype enrichment. LSPC‐Cycle (&gt;25% cycling stem‐like cells) had proliferative and quiescent LSPCs, frequent French‐American‐British (FAB)‐M7 and worst prognosis (overall survival odds ratio vs. GMP‐like: 11.33). Morphology was related to the primitive (FAB‐M0), GMP‐like (M2/M4) and ProMono‐like (M5) groups. Genomic patterns aligned with hierarchy: <jats:italic>CBFA2T3::GLIS2</jats:italic> , <jats:italic>MYB::GATA1</jats:italic> and <jats:italic>MECOM</jats:italic> high expression in LSPC‐Cycle; <jats:italic>CBFB::MYH11</jats:italic> , <jats:italic>NUP98::NSD1</jats:italic> and <jats:italic>NPM1</jats:italic> in the intermediate; and <jats:italic>DEK::NUP214</jats:italic> and <jats:italic>NUP98::KDM5A</jats:italic> were concentrated in the primitive group. bZIP <jats:italic>CEBPA</jats:italic> and <jats:italic>FLT3</jats:italic> ‐ITD mutations clustered in the intermediate and primitive groups. Most of <jats:italic>RUNX1::RUNX1T1</jats:italic> clustered in the intermediate group, whereas <jats:italic>KMT2A::MLLT3</jats:italic> hierarchy differed with <jats:italic>MECOM</jats:italic> expression level. Paediatric AML comprised more primitive cells, rarely with mono‐like/cDC‐like dominance. LSPC‐Cycle, <jats:italic>FLT3</jats:italic> ‐ITD and <jats:italic>NUP98::KDM5A</jats:italic> were considered independent prognostic factors in multivariate analysis. Findings indicate the prognostic relevance of cellular hierarchy and the importance of integrating hierarchy‐specific molecular profiles for improved risk stratification and treatment formulation. </jats:p>