AUTHORS
Li Fan, PhD, Jian Wang2, Rhonda E. Ries, MA3, Jack H. Peplinski, MSc4, Timothy I Shaw, PhD5, Ramzi Alsallaq6, Yuan Feng7, Yanling Liu8, Pandurang Kolekar, PhD9, Hanxia Li10, Soheil Meshinchi, MD, PhD11, Xiaotu Ma, PhD12, Xin Zhou13
AFFILIATIONS
1 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
2 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
3 Fred Hutch Cancer Center, Seattle, United States
4 Fred Hutch Cancer Center, Seattle, United States
5 H.Lee Moffitt Cancer Center and Research Institute, Tampa, United States
6 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
7 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
8 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
9 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
10 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
11 Fred Hutch Cancer Center, Seattle, United States
12 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
13 St. Jude Children's Research Hospital, Memphis, Tennessee, United States
ABSTRACT
Background
Pediatric acute myeloid leukemia (AML) is a genetically heterogeneous disease, yet comprehensive resources integrating genomic, transcriptomic, clinical, and longitudinal relapse data remain limited. To address this gap, we developed PedAML Portal, an interactive multi-omic resource for pediatric AML.
Methods
We analyzed 4,087 tumor specimens from 2,479 pediatric AML patients, including whole-genome sequencing (WGS) from 1085 patients, RNA sequencing (RNA-seq) from 2,456 patients, and matched diagnosis-relapse pairs from 143 patients. The cohort comprised 2,272 pediatric AML and 206 Down syndrome (DS-AML) patients. Findings from genomics and transcriptomics analyses and clinical annotations were integrated into an interactive web portal.
Results
The portal provides an interactive platform for exploring genomic alterations, transcriptomic programs, RNA splice junctions, clinical outcomes, and disease evolution through integrated visualization and analytical tools, including mutation lollipop plots, co-mutation matrices, subtype-specific genomic landscapes, GRIN2 genome-wide driver discovery, transcriptome UMAP embeddings, RNA-seq-derived cell-type abundance estimation, differential expression, pathway enrichment, survival analyses, longitudinal diagnosis-relapse comparisons, and customizable cohort analyses.
Among 866 pediatric AML patients with matched genomic and transcriptomic data of pediatric AML, 90% were classified into one of 23 molecular subtypes defined by recurrent genomic alterations, including KMT2A-rearranged (23%), RUNX1::RUNX1T1 (13%), CBFB::MYH11 (9.5%), NUP98 rearrangements (7.7%), UBTF tandem duplications (3.8%) and NPM1 (9%). Across these patients, we identified 187 recurrent driver genes and 116 distinct oncogenic fusion types, including rare oncogenic fusion types in AML, such as SFPQ::ZFP36L2 (n=2), PSPC1::ZFP36L1 (n=1) and TNRC18::XPO1 (n=1). Among 206 DS-AML patients, 96% harbored GATA1 mutations. Together, these data establish a comprehensive resource for integrated genomic and transcriptomic analyses of pediatric AML.
Consistent with previous reports, WT1 mutations frequently co-occurred with NUP98::NSD1, UBTF -TD, and FLT3-ITD. Integrated longitudinal analyses revealed clonal evolution during relapse. WT1 mutations were significantly more frequent at relapse than at diagnosis (22.4% vs. 11.9%; Fisher's exact test, P = 0.027). Most relapse-associated WT1 alterations were hotspot frameshift mutations and were observed across molecular subtypes including NPM1 and NUP98::NSD1. Shared WT1 mutations were also observed in conjunction with loss of heterozygosity at relapse, suggesting a potential second-hit mechanism. For FLT3, RAS pathway genes, TP53 and KIT, both gains and losses of mutations occurred between diagnosis and relapse, with losses occurring more frequently than gains. Overall, pediatric AML exhibited a distinct pattern of genomic evolution compared with that reported in pediatric ALL.
Transcriptomic analyses identified relapse-associated proliferative and inflammatory programs, accompanied by significant upregulation of genes involved in cell-cycle progression, myeloid differentiation, and leukemic cell survival. Analysis of RNA splice junctions identified recurrent AML-specific and subtype-specific junctions absent from normal hematopoietic tissues and other pediatric cancers, highlighting candidate biomarkers and potential therapeutic targets.
Conclusions
PedAML Portal represents one of the most comprehensive integrated genomic and transcriptomic resources for pediatric AML. By integrating comprehensive multi-omic datasets with interactive analytical and visualization tools, the portal enables systematic exploration of molecular heterogeneity, disease evolution, and clinical associations while accelerating biomarker discovery, hypothesis generation, and precision medicine research in pediatric AML.
DISCLOSURES
Li Fan, PhD: No relevant conflicts of interest to declare.