In 1971, Dr. Alfred G. Knudson proposed his “two-hit” hypothesis to explain the role of recessive tumor suppressor genes in dominantly inherited cancer syndromes. He postulated that the first hit was inherited, and the second hit was acquired and triggered tumorigenesis.1 This hypothesis was subsequently confirmed by the demonstration of a loss of heterozygosity at 13q14 in retinoblastomas2 and the cloning of the first tumor suppressor gene RB1.3 Since then, numerous genes associated with inherited cancer predisposition syndromes have been discovered. Mutations in these genes lead to disease through several mechanisms, including, but not limited to, inactivation of tumor suppressor genes. The frequency of germline mutations in cancer-predisposition genes in children and adolescent patients had formerly not been determined across a broad range of tumor types.
To determine this frequency, Dr. Jinghui Zhang and colleagues performed whole-genome and/or whole-exome sequencing of constitutional DNA purified from blood samples from 1,120 children with a variety of cancers. The median age of the patients was 6.9 years (range, 8 days to 19.7 years). Using a candidate-gene analysis approach that largely focused on 60 known autosomal-dominant or 29 autosomal-recessive cancer-predisposition genes, they found that 8.5 percent of the patients (95 of 1,120) harbored pathogenic or likely pathogenic mutations in their candidate genes compared with 1.1 percent (11 of 966) and 0.6 percent (4 of 723) in two cancer-free control groups. The most commonly mutated genes included TP53 (Li-Fraumeni syndrome), APC (familial adenomatous polyposis), BRCA2 (familial ovarian-breast cancer), NF1 (neurofibromatosis), PMS2 (hereditary colorectal cancer), RB1 (hereditary retinoblastoma), and RUNX1 (congenital thrombocytopenia with predisposition to myelodysplastic syndrome/acute myeloid leukemia). Among those patients with a monoallelic germline mutation, 66 percent (61 of 93) harbored a second hit within the tumor genome as shown by loss of heterozygosity or mutational inactivation of the second allele.
The prevalence of pathogenic or likely pathogenic germline mutations in autosomal dominant cancer–predisposition genes was highest among patients with non–central nervous system (non-CNS) solid tumors (16.7%; 48/287), followed by those with CNS tumors (8.6%; 21 of 245). The histologic subtypes of CNS tumor that were most often associated with germline mutations included choroid plexus carcinoma (25%; 1 of 4), medulloblastoma (13.5%; 5 of 37), high-grade glioma (9.1%; 9 of 99), low-grade glioma (7.9%; 3 of 38), and ependymoma (6.0%; 4 of 67). The prevalence of germline mutations varied among patients with different subtypes of non-CNS solid tumors: 69.2 percent for adrenocortical tumor (27 of 39), 17.9 percent for osteosarcoma (7 of 39), 13.3 percent for retinoblastoma (2 of 13), 10.9 percent for Ewing’s sarcoma (5 of 46), 7.0 percent for rhabdomyosarcomas (3 of 43), and 4.0 percent for neuroblastoma patients (4 of 100). The incidence in leukemia patients was 4.4 percent (26 of 588). Eight patients harbored germline mutations in the adult-onset cancer-predisposition genes BRCA1, BRCA2, and PALPB2. These genes are not typically sequenced in pediatric cancer patients as they are thought to predispose to adult cancer. Only one of these patients’ tumors exhibited evidence for mutational inactivation or loss of the second allele. Importantly, the investigators uncovered a family history of a first- or second-degree relative with cancer in only 40 percent of the 95 patients with a putative heritable mutation who had charts documenting a family history (18 of 43 records).
The study likely underestimates the proportion of pediatric cancer patients harboring a heritable cancer-susceptibility mutation as the candidate-gene approach excluded analysis of most of the genes in the genome. Additionally, our understanding of mutations in noncoding regions of the genome is limited, and thus pathogenic mutations in these regions could have been missed.
In Brief
This is the most comprehensive study to date of the genetics of childhood cancer predisposition. It demonstrates the clinical utility of next-generation sequencing to identify inherited cancer predisposition in pediatric and young adult patients. The quality and depth of sequencing allowed identification of mutations that would have been missed by standard Sanger sequencing, including the identification of germline mosaicism. This broad genetic approach to diagnosis highlights how patients with inherited cancer predisposition may present with atypical findings, that is, cancers not commonly expected within the phenotypic spectrum of their disorder and in the absence of a strong family history of cancer. This work also raises a number of important questions, including whom to test; do BRCA1, BRCA2, or PALB2 mutations contribute to pediatric cancer; what additional factors contribute to cancer development; and would the inclusion of more inherited marrow failure and inherited predisposition to leukemia/myelodyplastic syndrome genes in the analyses have altered the conclusions? More work is needed to address these and other questions as we sort out how best to incorporate germline sequencing into clinical care.
References
Competing Interests
Dr. Keel indicated no relevant conflicts of interest.
