Targeted sequencing of 36 known or putative colorectal cancer susceptibility genes.

BACKGROUND: Mutations in several genes predispose to colorectal cancer. Genetic testing for hereditary colorectal cancer syndromes was previously limited to single gene tests; thus, only a very limited number of genes were tested, and rarely those infrequently mutated in colorectal cancer. Next-generation sequencing technologies have made it possible to sequencing panels of genes known and suspected to influence colorectal cancer susceptibility.

METHODS: Targeted sequencing of 36 known or putative CRC susceptibility genes was conducted for 1231 CRC cases from five subsets: (1) Familial Colorectal Cancer Type X ( n  = 153); (2) CRC unselected by tumor immunohistochemical or microsatellite stability testing ( n  = 548); (3) young onset (age <50 years) ( n  = 333); (4) proficient mismatch repair (MMR) in cases diagnosed at ≥50 years ( n  = 68); and (5) deficient MMR CRCs with no germline mutations in MLH1, MSH2, MSH6, or PMS2 ( n  = 129). Ninety-three unaffected controls were also sequenced.

RESULTS: Overall, 29 nonsense, 43 frame-shift, 13 splice site, six initiator codon variants, one stop codon, 12 exonic deletions, 658 missense, and 17 indels were identified. Missense variants were reviewed by genetic counselors to determine pathogenicity; 13 were pathogenic, 61 were not pathogenic, and 584 were variants of uncertain significance. Overall, we identified 92 cases with pathogenic mutations in APC , MLH1 , MSH2 , MSH6 , or multiple pathogenic MUTYH mutations (7.5%). Four cases with intact MMR protein expression by immunohistochemistry carried pathogenic MMR mutations.

CONCLUSIONS: Results across case subsets may help prioritize genes for inclusion in clinical gene panel tests and underscore the issue of variants of uncertain significance both in well-characterized genes and those for which limited experience has accumulated.