Importance: CHEK2 pathogenic and likely pathogenic variants (PVs) are common, and low-risk (LR) variants, p.I157T, p.S428F, and p.T476M, are even more common. Biallelic CHEK2 PVs are associated with specific cancer phenotypes, including early age at onset of breast cancers. Whether biallelic LR variants are associated with cancer predisposition is unknown.
Objective: To characterize the cancer phenotype among individuals with biallelic CHEK2 variants, specifically those that have been associated with lower cancer risk in the heterozygous state.
Design, setting, and participants: This retrospective observational cohort study examining cancer phenotype by CHEK2 genotype was conducted at a single diagnostic genetic testing laboratory. Of 36 821 individuals who underwent genetic testing, 3783 (10.3%) with CHEK2 PVs or LR variants were ascertained from July 1, 2012, to September 30, 2019. Analyses were conducted from September 2022 to January 2024.
Exposures: Cancer phenotype among individuals with 2 LR variants and those with 1 PV and 1 LR variant was compared with cancer phenotype among individuals with wild type (WT) (n = 33 034), single LR variant (n = 1566), single PV controls (n = 2167), and 2 PVs (n = 21). Cancer phenotypes were investigated for any cancer, multiple primary cancers, female breast cancer, and bilateral female breast cancers.
Main outcomes and measures: Cancer phenotype of CHEK2 2 LRs and 1 PV and 1 LR.
Results: Of 36 821 individuals, 92.1% were female, and the median age at testing was 53 years (IQR, 44-63 years); 3787 (10.3%) were identified as having a CHEK2 PV or LR variant. There were 13 individuals with 2 LR variants and 20 with 1 PV and 1 LR variant. Among those with 2 LR variants, prevalence of any cancer (76.9%) and breast cancer (60.0%) were similar to those with WT (any cancer, 69.8%; breast cancer, 52.7%) and those with a single LR variant (any cancer, 70.9%; breast cancer, 57.5%). Among participants with 1 PV and 1 LR variant, 95.0% had a prior cancer diagnosis, a higher rate than among those with a single PV (76.8%), but the difference was not statistically significant. Among female individuals with 1 PV and 1 LR variant, 86.7% had a breast cancer diagnosis, compared with 67.1% with a single PV, although these differences were not statistically significant.
Conclusions and relevance: In this cohort study, individuals with 2 LR variants in CHEK2 had a cancer phenotype similar to those with a single LR variant and similar to WT controls. Individuals with 1 PV and 1 LR variant may have a more penetrant cancer phenotype than individuals with a single PV. Future studies focused on CHEK2 LR variants will aid in better understanding whether these variants are genetic modifiers associated with cancer risk.