Defining Why Cancer Develops in Children
A partir d'échantillons tumoraux prélevés sur 1 120 patients atteints de cancer et âgés de moins de 20 ans, de l'analyse du génome entier (chez 595 patients) et de l'exome entier (chez 456 patients), cette étude identifie la présence de mutations constitutionnelles dans des gènes de prédisposition au cancer
It has been more than 40 years since Knudson ushered in the modern era of cancer genetics with his seminal “two hit” hypothesis to explain familial cancer syndromes and provide a mathematical model for the biallelic inactivation of tumor-suppressor genes — with the first hit being inherited and the second occurring somatically to initiate tumorigenesis.1 Notably, he studied rare childhood cancers to develop his initial model, which proved to be remarkably accurate for both pediatric-onset and adult-onset heritable cancers.
Since the identification in 1986 of loss-of-function mutations in the gene RB1 that cause hereditary retinoblastoma,2 more than 100 cancer-predisposition genes have been identified,3 and our understanding of how common and rare DNA variations influence cancer susceptibility has also grown exponentially through the use of genomewide association studies. However, fundamental questions remain, not the least of which is how to use this information to improve human health. Thus, it is timely that the burgeoning power of cancer genomics be applied to address the question of why cancer develops in children, especially in those without a family history of the disease.
Zhang and colleagues now report in the Journal their initial observations from sequencing constitutional DNA purified from blood samples from 1120 children with a variety of cancers.4 These sequences were obtained as part of a larger project to define the landscape of somatic mutations in the tumor specimens, but in the current study, the investigators sought to define the prevalence of heritable mutations that would probably explain the index cancers. Although the investigators started with whole-genome or whole-exome (coding regions only) sequencing data (or both), they performed a candidate-gene analysis that was largely focused on 60 known autosomal dominant cancer-predisposition genes, leaving the majority of the genome for follow-up analyses.
New England Journal of Medicine , éditorial, 2014