Identification of deleterious germline <i>CHEK2</i> mutations and their association with breast and ovarian cancer

Abstract

Germline mutations in checkpoint kinase 2 ( CHEK2 ), a multiple cancer‐predisposing gene, increase breast cancer (BC) risk; however, risk estimates differ substantially in published studies. We analyzed germline CHEK2 variants in 1,928 high‐risk Czech breast/ovarian cancer (BC/OC) patients and 3,360 population‐matched controls (PMCs). For a functional classification of VUS, we developed a complementation assay in human nontransformed RPE1‐ CHEK2 ‐knockout cells quantifying CHK2‐specific phosphorylation of endogenous protein KAP1. We identified 10 truncations in 46 (2.39%) patients and in 11 (0.33%) PMC ( p = 1.1 × 10 −14 ). Two types of large intragenic rearrangements (LGR) were found in 20/46 mutation carriers. Truncations significantly increased unilateral BC risk (OR = 7.94; 95%CI 3.90–17.47; p = 1.1 × 10 −14 ) and were more frequent in patients with bilateral BC (4/149; 2.68%; p = 0.003), double primary BC/OC (3/79; 3.80%; p = 0.004), male BC (3/48; 6.25%; p = 8.6 × 10 −4 ), but not with OC (3/354; 0.85%; p = 0.14). Additionally, we found 26 missense VUS in 88 (4.56%) patients and 131 (3.90%) PMC ( p = 0.22). Using our functional assay, 11 variants identified in 15 (0.78%) patients and 6 (0.18%) PMC were scored deleterious ( p = 0.002). Frequencies of functionally intermediate and neutral variants did not differ between patients and PMC. Functionally deleterious CHEK2 missense variants significantly increased BC risk (OR = 3.90; 95%CI 1.24–13.35; p = 0.009) and marginally OC risk (OR = 4.77; 95%CI 0.77–22.47; p = 0.047); however, carriers low frequency will require evaluation in larger studies. Our study highlights importance of LGR detection for CHEK2 analysis, careful consideration of ethnicity in both cases and controls for risk estimates, and demonstrates promising potential of newly developed human nontransformed cell line assay for functional CHEK2 VUS classification.