Genetic Evaluation of Rare Cancer Syndromes Using a Combined Analysis of the Germline and Tumor Genome Data

Abstract

Research in oncology is often conducted by assessing the germline and somatic information separately. The tumor genome can help identify cancer types and therapeutic targets. Conversely, the germline genome can reveal cancer predisposition risk. The separate assessment of germline and somatic genomes can lead to challenges in understanding the full extent of tumor biology and tumor etiology. This thesis study hypothesizes that investigating a combination of germline and somatic information is more informative than assessing tumor-only or germline-only information. To test this hypothesis, 203 variants were randomly selected and assigned to three different cohorts based on genomic origin: 1) Cohort 1 included 42 somatic variants without any germline data; 2) Cohort 2 included 63 germline variants without any somatic data; and 3) Cohort 3, or Germline Plus, included 98 germline variants with combined somatic data. All variants were annotated using variant type, protein effect, genomic region, and population allele frequency. Allele frequency values were obtained informatically from the publicly available gnomAD database. Comprehensive variant analysis was performed to assess loss of heterozygosity (LOH), and whether LOH was sporadically or constitutionally driven. For each variant and gene, the associated cancer type was assessed. The annotated and analyzed variants were aggregated into a comprehensive database to enable the assessment of the informative status for each variant. The variant interpretation was performed using a combination of LOH status and cancer-type information. The determination of cancer etiology was performed to distinguish inherited versus non-inherited cancer types. The results showed that a combined assessment of the germline and somatic information is more informative. The Germline Plus Cohort produced the most informative results, as 98% of variants were informative. Somatic-only analysis showed that 57% of variants were informative. Germline-only analysis without any tumor information did not provide full details of tumor etiology. The findings of this research project support the hypothesis that the combined assessment of somatic and germline genetics sheds light on the etiology of cancer and tumor biology.