Updated genomic landscape for pediatric acute myeloid leukemia enables new treatment possibilities

The findings establish 23 distinct molecular categories (including 12 categories not covered by the current classification systems) that can be used to classify cases of pAML on an individual basis. This marks a vital step towards understanding the molecular background and improving treatment strategies in the future. The classification of these cancers and appreciation of their uniqueness is not just a helpful tool for doctors to diagnose patients more accurately. It is also critical to identify the most effective therapeutic route and to ultimately save lives.

Identifying the blind spot

Just over 4 per every 100,000 people develop AML yearly; the vast majority are adults. This conceals the effect of pediatric cases, which account for about 500 new cases each year. Due to this, a disparity in how we understand this cancer has grown. "Most of what we know about classifications of AML really comes from the adult field," said Jeffery Klco, MD, PhD, St. Jude Department of Pathology. "AML is more common in adults than it is in children."

Klco, along with first authors Masayuki Umeda, MD, PhD, and Jing Ma, PhD, first demonstrated the unique features of pAML in 2022 with the discovery of a novel alteration in the UBTF gene.

"That put forth a new entity in pediatric AML that had not been described before," explained Klco. "We decided to take a deeper dive into the overall classification of pediatric AML, recognizing that there are very significant differences between AML in adults and kids."

To do this, they amassed a cohort of 887 unique pAML cases and examined what was driving the cancers through transcriptome and gene profiling.

"A lot of the types of AML that we see here at St. Jude, you just don't see in adults," Klco emphasized.

AML classification provides biological insight and potential clinical guidance

The reasons why different features give rise to AML in children and adults have yet to be fully understood. The normal accumulation of mutations during aging may explain many adult AML cases. In contrast, fusion oncoproteins, aberrant proteins resulting in the fusion of two separate genes, are major drivers for many pAML cases and were shown to account for over 70% of cases in this new study. Although each molecular category has a unique driver, some show very similar transcriptional and mutational profiles, which stood out to the researchers. "Some categories show very similar transcriptional profiles, indicating that the background biology is similar and can be potentially treated by similar drugs," said Umeda.

This work offers a clear path forward for clinicians to identify distinct pAML sub-types and for larger collaborative groups to define cancer more accurately. "The World Health Organization [WHO] issued new classifications for hematological cancers in 2022, mostly based on adults," Klco explained. "We recognized that many of the recurrent alterations we found in pAML aren't even mentioned. Some of these sub-types have a significant impact on outcomes."

Correct classification will allow clinicians worldwide to understand their pAML patients better and offer guidance to treat cases as low-risk or high-risk. In fact, through their analysis, they determined a strong association between the new sub-types and clinical outcomes. "The study definitely fills a lot of gaps in the current classification of pAML," Ma confirmed. "It provides a risk stratification strategy that we hope will provide clinicians with a simpler road to accurate diagnosis and optimal treatment in the future."

Authors and funding

The study's other authors include Tamara Westover, Yonghui Ni, Guangchun Song, Jamie Maciaszek, Michael Rusch, Delaram Rahbarinia, Scott Foy, Michael Walsh, Priydarshini Kumar, Yanling Liu, Yiping Fan, Gang Wu, Xiaotu Ma, Lu Wang, Jeffrey Rubnitz, and Stanley Pounds of St. Jude; Benjamin Huang of the University of California San Francisco, Sharyn Baker of the Comprehensive Cancer Center, The Ohio State University, and Todd Alonzo of Keck School of Medicine, University of Southern California.

The study was supported by grants from The National Institutes of Health (P30 CA021765, Cancer Center Support Grant and Developmental Fund Award and U54 CA243124), the Fund for Innovation in Cancer Informatics, the Burroughs Wellcome Fund, the V Foundation, and ALSAC, the fundraising and awareness organization of St. Jude.