Genome-wide scanning for linkage in 56 Dutch breast cancer families selected for a minimal probability of being due to BRCA1 or BRCA2

Background

A conventional model of inherited breast cancer susceptibility is that disease risks are affected by mutations in a small number of genes causing a high risk of the disease and by a larger number of lower risk gene variants probably interacting together [1]. Model-based linkage analysis in multiple-case families, followed by positional cloning, led to the identification of BRCA1 and BRCA2. The cancer risks conferred by mutations in these genes are now well established, but together they explain only approximately 25% of the overall excess familial risk. Families with at least four cases of breast cancer and at least one case of ovarian cancer can be attributed largely to BRCA1. Multiple-case families with at least one case of male breast cancer are mainly due to BRCA2. But the majority of families with four or five cases of female breast cancer diagnosed before the age of 60 are not due to BRCA1 or BRCA2. This has been taken as evidence that one or more moderate-risk to high-risk breast cancer susceptibility genes still remain to be identified [2].

Methods

There have been several linkage claims since BRCA1 and BRCA2 were identified, but none of these have been replicated in other, often larger, studies. Most studies, however, are heavily underpowered to detect a new breast cancer locus by linkage in the presence of substantial genetic heterogeneity. The Breast Cancer Linkage Consortium is currently compiling genome-wide linkage data on approximately 200 families in which the role of BRCA1 or BRCA2 has been excluded with >90% certainty. This comprises the largest post-BRCA1/2 linkage search effort in breast cancer families in the world to date. These families are characterised by the presence of at least three cases of breast cancer diagnosed before the age of 60, and no cases of ovarian or male breast cancer.

Results

Here we present the Dutch contribution to this effort, including GENEHUNTER analysis of 56 families in which 208 patients were genotyped at 410 microsatellite markers. Allowing for heterogeneity, there were two regions, on 1q and 15q, that gave lod scores of 1.40 (α = 0.12) and 1.19 (α = 0.29), respectively. Lod scores for all other markers were <1.0. In non-parametric analysis, lod scores >1.0 were found for one or more markers at 4p, 6q, 7q, 9p, 15q, and 21q. All of these regions are currently being followed up by additional flanking marker typings.