Rare germline variants in childhood cancer patients suspected of genetic predisposition to cancer

Identification of cancer-predisposing germline variants in childhood cancer patients is important for therapeutic decisions, disease surveillance and risk assessment for patients, and potentially, also for family members. We investigated the spectrum and prevalence of pathogenic germline variants in selected childhood cancer patients with features suggestive of genetic predisposition to cancer. Germline DNA was subjected to exome sequencing to filter variants in 1048 genes of interest including 176 known cancer predisposition genes (CPGs). An enrichment burden analysis compared rare deleterious germline CPG variants in the patient cohort with those in a healthy aged control population. A subset of predicted deleterious variants in novel candidate CPGs was investigated further by examining matched tumor samples, and the functional impact of AXIN1 variants was analyzed in cultured cells. Twenty-two pathogenic/likely pathogenic (P/LP) germline variants detected in 13 CPGs were identified in 19 of 76 patients (25.0%). Unclear association with the diagnosed cancer types was observed in 11 of 19 patients carrying P/LP CPG variants. The burden of rare deleterious germline variants in autosomal dominant CPGs was significantly higher in study patients versus healthy aged controls. A novel AXIN1 frameshift variant (Ser321fs) may impact the regulation of β-catenin levels. Selection of childhood cancer patients for germline testing based on features suggestive of an underlying genetic predisposition could help to identify carriers of clinically relevant germline CPG variants, and streamline the integration of germline genomic testing in the pediatric oncology clinic.

Investigation of clinically relevant germline variants detected by next-generation sequencing in patients with childhood cancer: a review of the literature

Genetic predisposition is an important underlying cause of childhood cancer, although the proportion of patients with childhood cancer carrying predisposing pathogenic germline variants is uncertain. This review considers the pathogenic or likely pathogenic germline variants reported by six studies that used next-generation sequencing to investigate genetic predisposition in selected cohorts of patients with childhood cancer and used incompletely overlapping gene sets for analysis and interpretation. These six studies reported that 8.5%-35.5% of patients with childhood cancer carried clinically relevant germline variants. Analysis of 52 autosomal dominant cancer predisposition genes assumed common to all six studies showed that 5.5%-25.8% of patients with childhood cancer carried pathogenic or likely pathogenic germline variants in at least one of these genes. When only non-central nervous system solid tumours (excluding adrenocortical carcinomas) were considered, 8.5%-10.3% of the patients carried pathogenic or likely pathogenic germline variants in at least one of 52 autosomal dominant cancer predisposition genes. There was a lack of concordance between the genotype and phenotype in 33.3%-57.1% of the patients reported with pathogenic or likely pathogenic germline variants, most of which represented variants in autosomal dominant cancer predisposition genes associated with adult onset cancers. In summary, germline genetic testing in patients with childhood cancer requires clear definition of phenotypes and genes considered for interpretation, with potential to inform and broaden childhood cancer predisposition syndromes.