Next generation sequencing identifies a pattern of novel germline variants in early-onset colorectal cancer

Early-onset colorectal cancer (EOCRC) incidence is increasing rapidly worldwide. However, the majority of EOCRCs are not substantiated by germline variants in the main colorectal cancer (CRC) predisposition genes (the "DIGE" panel). To investigate a potential genetic transmission of EOCRC (dominant, recessive and oligogenic hypotheses) and thus identify potentially novel EOCRC-specific predisposition genes, we conducted an analysis of 585 cancer pathway genes on an EOCRC patient cohort (n=87 patients diagnosed at [≤] 40 years of age, DIGE-) with or without a CRC family history. By comparing this germline variant spectrum to the GnomAD cancer-free database, we identified high impact variants (HVs) in 15 genes significantly over-represented in the EOCRC cohort. Among the 32 unrelated patients with a CRC family history (i.e. with a potentially dominant transmission pattern), nine presented HVs in ten of the genes tested, four of these genes had a DNA repair function. A potentially recessive transmission of EOCRC in patients without a CRC family history cannot be supported by our results nor can an oligogenic transmission. We subsequently sequenced these 15 genes in a cohort of 82 late-onset CRCs (cancer diagnosis [≥]50 years, DIGE-) and found variants in 11 of these genes to be specific to EOCRC. To evaluate whether variants in these 11 genes would allow to specifically detect EOCRC patients, we screened our patient database (n=6482), which only contained 2% of EOCRCs (DIGE-), and identified two other EOCRC cases diagnosed after the constitution of our cohort, with individual HVs in RECQL4 and NUTM1. Altogether, we showed that 37.5% and 18.75% of heterozygous NUTM1 and RECQL4 HVs of our database were diagnosed with EOCRC. Our work has identified a pattern of germline gene variants not previously associated with EOCRC. This paves the way to addressing the contribution of these variants to EOCRC risk and oncogenesis. Author SummaryEarly-onset colorectal cancer (diagnosed at [≤] 40 years of age) is a rare disease that can in part be explained by a hereditary genetic predisposition. To identify novel gene variants potentially associated with EOCRC risk, we analysed a panel of 585 genes in 87 patients with early-onset colorectal cancer unexplained by conventional genetic tests. This first analysis highlighted 15 genes of interest. To evaluate if this genetic profile is specific to early onset, we sequenced these 15 genes in a population of late-onset colorectal cancers (diagnosed after 50 years of age). Variants in 11 of these genes were specific to the early-onset population. To assess if this genetic pattern allows to identify other early-onset cases, we screened these genes in our whole database of 6482 patients and identified two new early-onset cases. Our results need to be confirmed, and validated in larger cohorts but pave the way for future research into early-onset colorectal cancer and the possibility of improving screening or treatment options for these patients and their family members.