Linkage analysis of Dutch families at high risk for breast cancer.

Abstract

This study focused on the 16 Dutch families with three o r more cases of breast cancer (Cleton et al, 1983; Bailey-Wilson et al, 1986). These families were ascertained through a single known affected index case and two other affected first degree relatives, but the identities of the two other affected first degree relatives were not specified. The families were grouped prior to any analysis because their small size precluded estimation of the age of onset parameters separately for each family. The families were classified into 3 major groups based on epidemiologic criteria and into 2 major groups based on a genetic criterion. The epidemiologic criteria were age at breast cancer diagnosis and presence or absence among biological relatives of other cancers at high incidence in the Netherlands (stomach, bladder, and liver). The genetic criterion was based on segregation models I and I1 proposed by Go et a1 (1983), which were models of dominant Mendelian inheritance with lognormally distributed age of onset. (The median age of onset for susceptibility allele carriers was 48.2 years in Model I and 64.3 years in Model 11, and sporadic cases were allowed under both models). according to the model that yielded the larger likelihood conditional on the phenotype of the index case, provided that likelihood was larger than the likelihood for a model of no transmission from one generation to the next. on the epidemiologic and genetic criteria were similar (Table 1).