Segregation analyses in Ashkenazi Jewish families: Evidence of an additional breast cancer susceptibility gene, and evaluation of cancer risks in BRCA1/2 mutation carriers

The genes BRCA1 and BRCA2 account for two-thirds of hereditary breast cancer. Using segregation analysis, families of cases without BRCA1/2 mutations were studied for statistical evidence of another major breast cancer gene. This was carried out in a community-based sample of Jewish probands previously tested for the presence of three founder mutations. Complete data was available for 602 female first-degree relatives of 231 breast cancer probands with no founder BRCA1/2 mutation; 78 of the relatives had breast cancer.; Segregation analysis, used to evaluate the likelihood of various genetic and non-genetic models, rejected sporadic, environmental, arbitrary and decreasing Mendelian genetic models. A Mendelian recessive model fit better than dominant and co-dominant models, though none of these could be rejected. Cumulative incidence curves predicted by the recessive and co-dominant models fit observed incidence among first-degree relatives well. The assumption of Mendelian transmission of a major recessive gene is compatible with the data.; The recessive model predicts that 4% of women would carry the high-risk genotype, with 85% of them developing breast cancer by age 70. Heterogeneity exists between the study families and the 114 BRCA1/2 families from the same study population, implying that the recessive effect is not due to undetected BRCA1/2 mutations. The study adds support for an additional major recessive locus affecting breast cancer susceptibility.; An additional segregation analysis was used to estimate breast and ovarian cancer risks in families of the 114 BRCA1/2 mutation carriers from the same Jewish study population. These included 254 first-degree female relatives, 59 of whom had breast cancer. The cumulative incidence curve for mutation carriers predicted by the segregation analysis corresponds very closely to previous estimates of cancer risks in this population made using the kin-cohort method. Segregation analysis predicted lower rates of breast cancer in non-carriers than the kin-cohort method. A test of heterogeneity between BRCA1 and BRCA2 carriers showed no significant difference in the breast cancer risks attributable to the two genes. The kin-cohort method, which allows for estimation of genotype-specific risks from genotyped probands who provide a family history of disease, appears to approximate risk estimates obtained from segregation analysis.