Germline variation in DNA damage repair genes in children with cancer: Unveiling SMARCAL1 as novel osteosarcoma predisposition gene
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Recent large-scale sequencing studies reveal that up to 18% of children with cancer harbor pathogenic variants (PV) in known cancer predisposing genes (CPG). However, many patients exhibit features of an underlying predisposition but lack CPG PV. This observation suggests the presence of unidentified cancer predisposing variants and/or genes. To address this gap, we interrogated genes implicated in DNA damage repair (DDR) as a source of novel cancer predisposing signals that could account for this missing heritability.
Methods
We examined germline data from 5,993 childhood cancer cases and 14,477 adult non-cancer controls for damaging variants among 189 DDR genes. Damaging variants were defined as rare (allele frequency <0·05% in the gnomAD v2.1 non-cancer subset), nonsense, frameshift, canonical splice site, and missense with REVEL scores >0·7. Enrichment of damaging variants was determined using logistic and firth regression models. Putative novel CPG were replicated using data from 1,494 additional childhood cancer cases in three independent cohorts.
Findings
Pan-cancer analyses revealed enrichment of germline TP53 variants (false discovery rate [FDR] logistic =0·0066, FDR Firth =0·0064). Cancer-specific analyses demonstrated significant associations for TP53 in adrenocortical carcinoma (37%, FDR logistic <0·0001, FDR Firth =0) and high-grade glioma (2·4%, FDR logistic =0·0022, FDR Firth =0·1082), as well as BARD1 in neuroblastoma (1·2%, FDR logistic =0·0341, FDR Firth =0·2682). Three novel gene-tumor associations were identified, including POLL in Ewing sarcoma (1·7%, FDR logistic =0·0319, FDR Firth =0·3101), SMC5 in medulloblastoma (1·6%, FDR logistic =0·0005, FDR Firth =0·0499) and SMARCAL1 in osteosarcoma (2·6%, FDR logistic =0·0250, FDR Firth =0·2180). Among these new putative CPG, enrichment of SMARCAL1 PV in osteosarcoma was replicated across all three additional pediatric cancer cohorts (2·5%, P Fisher <0·0001). All osteosarcoma tumors (n=3) with whole genome sequencing data available exhibited deletion of the wild-type SMARCAL1 allele.
Interpretation
We report a catalogue of germline SMARCAL1 variants in pediatric osteosarcoma and demonstrate potential molecular effects in tumor formation through the presence of second somatic SMARCAL1 hits. The contribution of germline POLL and SMC5 variants in Ewing sarcoma and medulloblastoma, respectively, warrants further investigation. Our study highlights the power of unbiased genomic investigations to uncover novel CPG, providing insights into tumor biology and creating opportunities for therapeutic intervention, surveillance, and prevention.
Funding
Funding was provided by the American Lebanese Syrian Associated Charities and US National Institutes of Health, German Cancer Research Center, German Cancer Consortium.
Research in context
Evidence before this study
We performed a PubMed search using the terms “childhood cancer” OR “pediatric cancer” AND “germline mutation” OR “germline variant” AND “predisposition gene” AND “sequencing” AND “DNA damage repair” OR “DNA damage response” between Jan 1, 2020, and April 1, 2025, to investigate the spectrum of genes mutated in the germline of children with cancer. Genes participating in DNA damage repair (DDR) are often mutated in childhood tumors, highlighting the integral role of defective DNA repair in tumor formation. Further, pathogenic variants (PV) in DDR genes underlie several highly penetrant cancer predisposition syndromes and are implicated in the increased risk of treatment-related subsequent malignant neoplasms among adult survivors of childhood cancer. To the best of our knowledge, a comprehensive assessment of DDR genes across children with primary cancers has not yet been performed.
Added value of this study
The use of a robust case-control design, stringent germline variant filtering criteria, and integration of tumor data provided a rigorous framework with which to identify novel associations between germline PV in DDR genes and childhood cancers, including POLL in Ewing sarcoma, SMC5 in medulloblastoma, and SMARCAL1 in osteosarcoma. Among these, enrichment of SMARCAL1 PV in osteosarcoma was replicated across three additional pediatric cancer cohorts with loss-of-heterozygosity at the SMARCAL1 locus observed in all osteosarcoma tumors with whole genome sequencing available. Altogether, these findings provide compelling evidence linking SMARCAL1 germline variation with pediatric osteosarcoma.
Implications of all the available evidence
Identifying SMARCAL1 as a novel osteosarcoma predisposing gene has significant biological and clinical implications. This information provides new insights into the pathogenesis of osteosarcoma, an aggressive and often fatal childhood cancer, and may enable development of more effective therapies. Incorporating germline genetic testing for SMARCAL1 into clinical practice could inform the institution of novel surveillance strategies to detect incipient osteosarcoma tumors at the earliest and most curable stages, thereby improving outcomes for affected children and families. Future research should focus on inheritance patterns, penetrance, and therapeutic vulnerabilities of SMARCAL1 -deficient tumors, including their potential sensitivity to agents that inhibit DNA damage repair.
