Use of new sequencing technology to improve clinical diagnostics of hereditary colorectal cancer
MetadataShow full item record
Backgroud: Identification of a germline pathogenic variant that causes increased risk and aggregation of CRC in a family is important for the clinical management of the family members. Next generation sequencing technology (NGS) has revolutionised research and diagnostics within the field of hereditary colorectal cancer (CRC), providing the opportunity to perform comprehensive genetic testing to discover new CRC genes and to analyse several genes in parallel in routine diagnostics. Due to overlapping phenotypes among hereditary CRC syndromes, genetic testing using multigene-panels has emerged as an attractive alternative strategy compared to sequential testing of one gene after the other based on clinical criteria, which has been the traditional approach to assess hereditary cancer risk. Over the past years, numerous NGS-efforts have aimed at identifying novel hereditary CRC genes, utilising exome or whole-genome sequencing. Consequently, several novel CRC susceptibility genes have recently been proposed. Some of these genes already have enough convincing evidence to be included in routine genetic testing, while other genes need additional validating evidence and are therefore regarded as candidate genes. Despite identification of several novel hereditary CRC genes, only a fraction of the heritability in familial CRC is currently explained. It is therefore likely that there are still some undiscovered high-penetrance CRC susceptibility genes. Aims: The main aims of the three studies included in this thesis were to establish NGS in a diagnostic setting and to use NGS to improve the detection rate of genetic causes of familial CRC. Method: In the first study, a workflow based on amplicon enrichment and 454-sequencing technology (Roche) was developed and assessed in terms of suitability for diagnostic sequencing of genes involved in Lynch Syndrome. In the second and the third study, Illumina technology was utilised for sequencing analysis. The second study involved exome sequencing to find the genetic cause for predisposition in a family with high accumulation of CRC and other cancers. In the third study, we sequenced 274 familial CRC patients, previously tested for Lynch Syndrome, using a multigene-panel targeting 112 genes. The gene-panel included both well-known and candidate CRC susceptibility genes. Results: The NGS workflow based on 454 technology for routine genetic testing of genes involved in Lynch Syndrome was developed, optimised and established. The workflow had to work around a methodological problem of incorrect basecalling in homopolymeric regions and was evaluated to be suited for routine diagnostics. In the second study, a novel POLE variant, c.1373A>T p.(Tyr458Phe), was found to be causal and lead to the hereditary CRC syndrome polymerase proofreading- associated polyposis. This POLE variant predisposed to a broad tumour phenotype including CRC in the family investigated. We found the variant to be located in the exonuclease domain at an active site important for exonuclease function. Previous functional studies have demonstrated that the variant is pathogenic. In the third study, we identified 16 pathogenic or likely pathogenic variants (class 4 and 5) and 19 variants of unknown clinical significance in well-established cancer susceptibility genes. We also found 37 interesting variants in CRC candidate genes. Conclusion: The NGS workflow for analysing genes involved in Lynch Syndrome was adopted in the diagnostic sequencing activity at the Department of medical genetics, St. Olavs Hospital. However, due to methodological problems with 454-sequencing and tough competition in the NGS market, the technology is now outperformed, terminated and no longer relevant. The patients included in study II and III had previously been subjected to genetic testing without the cause for their increased CRC risk being identified. Using exomeand gene-panel sequencing, we succeeded in finding the genetic cause for predisposition in several of these individuals and families. This implies that comprehensive genetic testing, using exome sequencing or multigene-panels, is more effective than the traditional approach of phenotype-directed sequential testing of a few genes. Some of the variants of unknown clinical significance identified in well-established cancer susceptibility genes and some of the variants identified in the candidate genes might prove to be pathogenic, if subjected to further research. However, at this stage, we cannot confirm or refute the potential predisposing effect of pathogenic variants in any of the candidate CRC genes.