An update on APC I1307K homozygosity: observations from a large multigene panel testing cohort

Numerous studies have demonstrated that Ashkenazi Jewish individuals heterozygous for APC c.3902T > A; p.Ile1307Lys have 1.7-fold increased odds of colorectal adenocarcinoma (CRC). The cancer risk for homozygotes is unknown, and published data is based on small sample sizes. Here, we aim to quantitatively evaluate the odds of CRC in homozygotes from a clinical diagnostic setting. Analyses comparing the demographic and clinical histories among heterozygotes (n = 340) or homozygotes (n = 74) were compared to individuals with MSH6-associated Lynch syndrome (n = 372) and individuals from a pan-cancer panel of up to 90 genes who were negative for APC I1307K, MSH6 (likely) pathogenic (LP/P) variants, and other reportable LP/P variants after exclusions, henceforth referred to as wildtype (WT, n = 19,809). Comparisons included Fisher's exact or two-sample t-test and multivariate logistic regression. Homozygotes demonstrated a trend toward increased odds of CRC compared to WT (OR: 2.48; p = 0.074), and there was no difference between heterozygotes and WT (OR: 0.94; p = 0.83). MSH6-positive individuals had increased odds of CRC relative to homozygotes (OR: 3.54; p < 0.0001) but this was not significant in the logistic regression (OR: 1.19; p = 0.78). While homozygotes but not heterozygotes demonstrated an increase in CRC OR compared to WT, there was no statistically distinct odds of CRC between heterozygotes and homozygotes in this cohort. While the ORs herein support that the observed increase in CRC for homozygotes was more modest than that observed for MSH6-related Lynch syndrome in this cohort, current sample size and selection bias limit precise risk estimation and do not allow for distinction between modestly increased risk and no risk of CRC.