A new study by researchers at the German Cancer Research Center (DKFZ), the Hopp Children’s Cancer Center Heidelberg (KiTZ), and Heidelberg University Hospital has uncovered compelling evidence that the most aggressive forms of a childhood brain cancer known as medulloblastoma originate during fetal development. Published in Nature, the findings offer a deeper understanding of tumor initiation, suggesting that the earliest genetic disruptions occur long before diagnosis and that key oncogenic mutations arise only later in disease progression.
Medulloblastoma, a malignant tumor of the central nervous system, is the most common brain cancer in children. Among its various subgroups, groups 3 and 4 are particularly aggressive and difficult to treat. Until now, it was unclear which genetic events initiate these tumors and when they occur. Using high-resolution single-cell technologies, the researchers were able to reconstruct the timeline of tumor evolution at unprecedented detail.
The team analyzed thousands of individual cells from patient tumor samples using single-nucleus RNA sequencing (snRNA-seq), single-nucleus ATAC sequencing (snATAC-seq), and spatial transcriptomics. They discovered that large-scale chromosomal copy number variations (CNVs) are the earliest detectable genetic changes. These chromosomal alterations often precede the well-known oncogenic amplifications of MYC, MYCN, and PRDM6, which the study found to be subclonal and late-arising.
“Our data suggest that large-scale CNVs act as initiating events,” the authors wrote, noting that such early changes likely occur during the first trimester of gestation or in the first year of life. These aberrations often affect chromosomes 4, 7, 8, 10, 11, and 17, and appear in the precursors of unipolar brush cells, a highly specialized cerebellar neuron type previously implicated in medulloblastoma.
Importantly, the study challenges the assumption that amplification of MYC or MYCN directly drives tumor initiation. Instead, these alterations emerge as the disease progresses, potentially explaining why tumors harboring such mutations often resist therapy and relapse.
By integrating somatic mutation clocks and spatial transcriptomic data, the researchers modeled tumor development over time. They estimated that while large-scale CNVs initiate tumorigenesis early, clinical detection often occurs years later. This prolonged latency highlights a potential window for early intervention if sensitive diagnostic tools can be developed.
Lena Kutscher, co-author of the study and junior group leader at the DKFZ, emphasized the diagnostic implications in a press release: “If we succeed in developing sufficiently sensitive methods to detect these early changes, for example as DNA fragments in the blood, this could form the basis for early detection in newborns and infants.”
The findings underscore the power of single-cell multi-omics in revealing cancer’s hidden origins and call for a reevaluation of diagnostic and therapeutic strategies for pediatric brain cancer. As the authors concluded, understanding when and where medulloblastoma begins may be key to stopping it before it starts.
