News

Study: Pediatric cancer patients have high rate of germline mutations in predisposition genes

Publish date: November 18, 2015

View on the News

Investigating cancer in children

The sequencing study by Zhang et al. found that 8.5% of 1,120 participants with pediatric cancer had pathogenic mutations in an autosomal dominant cancer-predisposition gene, and four of the mutations were mosaic (i.e., present in only a subset of normal cells, probably indicating the defect was not inherited). The four mosaic mutations probably would not be detected by standard genetic testing strategies.

Not surprisingly, more than 50% of the mutations were in the tumor suppressor gene TP53. A long list of other genes identified as potentially disease causing each occurred at a prevalence of less than 6%.

Although the study’s inclusion of certain high-risk childhood cancers could bias the results toward overestimating the proportion of germline cancer predisposition mutations, more likely the results are an underestimate. By evaluating mutations in only a small subset of candidate autosomal dominant genes, the findings to not reflect a thorough assessment of most genes in the genome. In addition, focusing only on the exome ignores DNA mutations in noncoding regions, especially in tissue-specific enhancers, which may have a role in cancer susceptibility. So-called epimutations may affect cancer susceptibility in a nonmendelian fashion. Finally, the ability to study the interaction of several of these events may contribute to our understanding of tumor initiation.

The study raises several important questions. Are children with mutations in APC, BRCA1, or BRCA2 at risk for childhood cancers? How does germline mosaicism influence disease penetrance, and how many of the mosaic mutations were inherited? How do the mutations identified interact with less well known gene mutations elsewhere in the genome to influence malignant transformation? How can the findings translate to the clinic?

The study highlights the fact that family history is insufficient to assess the likelihood of a cancer-predisposition syndrome in any patient with a newly diagnosed cancer.

Dr. John Maris is a pediatric oncologist in the division of oncology and Center for Childhood Cancer Research, Children’s Hospital of Philadelphia, and in the department of pediatrics at the Perelman School of Medicine, University of Pennsylvania. These remarks were part of an editorial accompanying the report (N Engl J Med. 2015 Nov. 18 doi: 10.1056/NEJMoa1508054). Dr. Maris reported having no disclosures.

FROM THE NEW ENGLAND JOURNAL OF MEDICINE

References

Children and adolescents with cancer were found to have a significantly higher rate of germline mutations in cancer predisposing genes compared with individuals with no known cancer; however, family history of cancer did not predict the presence of a predisposition syndrome for most patients.

Of the 1,120 children with cancer, 8.5% had mutations in predisposition genes, compared with 1.1% in the control group. Mutations in TP53 were most common (50 patients), followed by APC (6), BRCA2 (6), NF1 (4), PMS2 (4), RB1 (3), and RUNX1 (3). Germline TP53 mutations were present in 27 of 39 patients (69%) with adrenocortical tumors, 9 of 47 (19%) with hypodiploid acute lymphoblastic leukemia, and 1 of 4 (25%) with choroid plexus carcinoma (N Engl J Med. 2015 Nov. 18. doi: 10.1056/NEJMoa1508054).

Only 40% of the pediatric patients with pathogenic germline mutations had a family history of cancer, and in just half of those cases the history was consistent with a known cancer-predisposition syndrome. Among patients without predisposition germline mutations, a similar proportion (42%) had a family history of cancer.

“On the basis of these observations, family history cannot be the sole indication used to guide the provision of genetic testing,” wrote Jinghui Zhang, Ph.D., of the department of computational biology, St. Jude’s Research Hospital, Memphis, Tennessee, and her colleagues.

Unexpected germline mutations were found in several cases. Six patients with Ewing’s sarcoma had unexpected pathogenic germline mutations (TP53 in four patients, PMS2 in one and RET in one). Eight patents had heterozygous mutations in BRCA1, BRCA2, or PALB2, supporting the notion that mutations in these genes may play a role in pediatric as well as adult cancer. Other new associations included germline APC and SDHB mutations with neuroblastoma, and APC, VHL, CDH1, PTCH1, and SDHA germline mutations with leukemia.

The St. Jude–Washington University Pediatric Cancer Genome Project (PCGP) included 1,120 patients representing the major types of pediatric cancer, including 53% with leukemia, 22% with CNS tumors, and 26% with non-CNS tumors. Whole genomes were sequenced from 595 patients, whole exomes (coding regions only) from 456 patients, and both whole genomes and exomes from 69 patients. Whole exomes were sequenced from two control cohorts of individuals with no known cancer: 966 from the 1,000 Genome project and 723 from the National Database for Autism Research (cancer predisposition genes known to be associated with autism, NF1 and PTEN, were excluded from analysis with this control set).

The study sequenced whole genomes and exomes, but focused most of the analysis on 60 autosomal dominant cancer predisposition genes. Tumor types with the highest prevalence of germline mutations in these genes were non-CNS solid tumors (48 of 287 patients, 17%) and CNS tumors (21 of 245, 9%). Among patients with adrenocortical tumors, 69% had germline mutations. Despite inclusion of hypodiploid acute lymphoblastic leukemia, the lowest germline mutation prevalence was found in leukemia (26 of 588, 4%).

Dr. Zhang reported having no disclosures. One coauthor reported financial ties to an industry source.

Study: Pediatric cancer patients have high rate of germline mutations in predisposition genes

View on the News

References

Recommended Reading