Association of Loss-of-Function Variants With Colorectal Cancer Risk in a Cohort of Over 950,000 Individuals.

Purpose: is a proposed colorectal cancer (CRC) predisposition gene, with only four families with putative loss-of-function (pLoF) variants published. The prevalence, phenotypic spectrum, and clinical management recommendations for heterozygotes remain unknown.

Methods: Retrospective review of approximately 950,000 individuals undergoing multigene panel testing (MGPT) for cancer predisposition identified 36 individuals with pLoF variants in . Clinical features, tumor prevalence, and age at onset were compared with Lynch syndrome (LS) and wild-type (WT) cohorts.

Results: Thirty-six individuals (0.004%) had 28 unique pLoF variants in and a total of 42 primary tumors, 78.6% of which were CRC with a median age of diagnosis of 50 years. A total of 16.7% of individuals had multiple CRC primaries, and a majority (71.4%) were mismatch repair-proficient (pMMR). Signet ring cell (SRC) pathology was significantly enriched compared with WT ( = <.0001). The odds ratio for CRC was significantly higher than that for the -related LS cohort (odds ratio [OR], 45.3 [95% CI, 21.3 to 101] and OR, 16.9 [95% CI, 14.9 to 19.2], respectively).

Conclusion: is associated with early-onset, pMMR CRC enriched for SRC pathology. Comparison of prevalence showed a statistically significant two-fold increase in the odds of developing CRC compared with an LS cohort. These data confirm the gene-disease relationship between and CRC and support clinical management guidelines similar to -related LS.