Putative pathogenic germline variants associated with hereditary cancer syndromes identified through clinical tumor-only sequencing

Background: Clinical tumor sequencing using next-generation sequencing (NGS) is widely applied to identify targetable variants and molecular biomarkers in cancer patients. Due to time and cost constraints, most institutions perform targeted gene panel sequencing only on tumor tissue, rather than on both tumor and matched normal samples. However, tumor-only sequencing cannot reliably distinguish true germline variants carried by the patient, particularly those associated with hereditary cancer syndromes. Methods: We retrospectively collected clinical tumor-only sequencing data from cancer patients at Seoul National University Hospital (SNUH) between December 2022 and May 2025. All data were generated using targeted NGS with a customized SNUH FiRST Cancer Panel consisting of 336 cancer-related genes. Potential pathogenic germline variants (PPGVs) were classified based on the recommendations of the ESMO Precision Medicine Working Group and the patient's clinical and family history. Finally, we confirmed true pathogenic germline variants using Sanger sequencing or targeted NGS methods with normal genomic DNA from peripheral blood mononuclear cells. Results: A total of 2,513 tumor samples underwent clinical targeted cancer panel sequencing during the examined period. Of those samples, 239 PPGVs were identified in 221 samples (8.8%) of various cancer types. While most cancers were breast or ovarian, different types of cancer were also associated with PPGVs, including those of the colon, pancreas, bladder, kidneys, prostate, bile ducts, and uterus. Of the 239 PPGVs, 129 (54.0%) were validated through germline confirmation testing. Of those, 88 (68.2%) were confirmed as true pathogenic germline variants, while the remaining 41 (31.8%) were somatic. Most germline variants were found in genes involved in homologous recombination repair (HRR), such as BRCA1, BRCA2, PALB2, BRIP1, RAD51D, CHEK2, and RAD50; or in genes involved in mismatch repair (MMR), such as MLH1, MSH6, MSH2, and PMS2. Among the PPGVs, 85.0% of BRCA1 variants were confirmed as germline, whereas only 12.5% of PTEN variants were verified to be germline. The germline variants showed an average variant allele frequency (VAF) of 58.1%, which is sufficiently high to be considered germline. However, the somatic variants exhibited a mean VAF of 51.5%, making them difficult to distinguish from true germline variants without additional confirmatory testing. Conclusions: Clinical tumor-only sequencing can not only identify therapeutically actionable mutations but also infer PPGVs associated with hereditary cancers. Therefore, careful attention to PPGVs is essential when interpreting tumor-only sequencing results for the accurate diagnosis and prevention of hereditary cancer syndromes.