There are several approaches to identifying low-penetrance candidate genes for breast cancer. In one approach, it is assumed that missense variants of genes for which truncating mutations are clearly pathogenic might also be deleterious. If the missense allele demonstrates high penetrance (i.e. like truncating mutations), then it will be relatively straightforward to establish the association when allele frequency is high. If the penetrance of the missense variant is low, however, then the association may be missed if only a small number of cancers is studied and the variant may falsely be classified as a neutral polymorphism. We are in the process of establishing a large database of breast cancer cases and ethnic-matched controls in order to evaluate the pathogenicity of the common founder alleles of the most important cancer susceptibility genes. Several deleterious founder alleles have been identified in BRCA1, but to date, no founder mutation in BRCA2 has been identified [1–3]. There are a few common variant alleles in BRCA2 in Poland; one of these (C5972T) changes the amino acid sequence of BRCA2 from threonine to methionine at codon 1915. It lies within the range of the BRC encoded by exon 11 that are thought to be involved in binding to RAD51 [4]. We sought to determine whether this common missense BRCA2 variant plays a role in breast cancer susceptibility.

We found 13 C5972T homozygotes (11 patients and 2 controls). The OR associated with the homozygous state was 4.7 (95% CI: 1.1 to 21.4). Among 11 cases of breast cancer diagnosed in homozygotes, 10 were diagnosed between ages 40 and 50 years. The 11 homozygotes had predominantly intraductal (ductal carcinoma in situ (DCIS) with micro-invasion, 4 cases), ductal G3 (4 cases), lobular (2 cases), and ductal G2 (1 case) cancers. None of the first- or second-degree relatives of the homozygotes were diagnosed with breast or ovarian cancer.

The C5972T allele of the BRCA2 gene appears to be over-represented in women with early onset breast cancer in Poland compared to controls. Although the effect was modest in the overall data set, it was stronger for women who had predominantly intraductal breast cancer and for women who were homozygous for the variant allele. Existing reports on the C5972T variant do not record any case of homozygote TT genotypes [6, 7]. It is of interest that the heterozygote state was associated with early onset breast cancer and the homozygote state with later onset cancer (above 40 years of age). To our knowledge, this has not been seen before and there is no compelling explanation; however, we believe that the sizes of these subgroups are small and require confirmation in other studies.

Our study population consisted of DCIS cases with a micro-invasive component only. To our knowledge, this is the first genetic association reported for intraductal cancer with micro-invasion. Our data set did not include cases of DCIS alone and we were unable to assess the effect of the BRCA2 C5972T allele in women with predominantly invasive cancer and associated DCIS. These topics will be the focus of future studies. We did not see other pathological features that distinguished BRCA2-associated tumors from tumors in non-carriers. This situation is not different from that described for BRCA2 carriers in general [8–10].

Several previous studies have shown that DCIS is a characteristic feature of cancers in BRCA2 carriers [11–14]. In the case of BRCA2-associated breast cancers, it appears that loss of heterozygosity (LOH) occurs as a means of silencing the second allele. In one family with a germline BRCA2 mutation, three cases of invasive breast cancer and one case of DCIS were observed [13]. LOH at the BRCA2 locus was demonstrated in all four cases. The authors suggest that DCIS and LOH in the BRCA2 locus may be common early steps in the development of invasive breast cancer in BRCA2 carriers. The LOH data support the concept that BRCA2 behaves as a tumor suppressor gene. In support of this hypothesis are our observations of a greater breast cancer risk in women with two inherited 5972T variants and a lower frequency of positive family history among 5972C/T heterozygotes.

In our study, we observed no association between the examined variant and increased family history. This may be explained by the modest association of the examined variant with breast cancer risk and probable recessive pattern of inheritance (the highest OR values were observed in homozygotes, all of them with negative breast cancer family history).

Actually, at least some of the 5972C/T heterozygotes may be compound heterozygotes, with the second allele being another BRCA2 variant.

We now provide epidemiological and clinical evidence demonstrating the deleterious nature of the C5972T allele. To our knowledge, functional studies have not been performed on this variant. Few other deleterious missense variants in BRCA2 have been identified.

The HH genotype of the non-conservative amino acid substitution polymorphism N372H in the BRCA2 gene was reported to be associated with a 1.3-fold to 1.5-fold increase in the risk of both breast and ovarian cancer [15–18], but there have been negative studies as well [19, 20].

Our control group was drawn both from the newborns of 10 Polish cities and from the adult population in the region of Szczecin; however, the frequency of alleles was similar in the newborn (5.7%) and adult (5.8%) populations. There was no evidence that the genotype frequencies of the C5972T variants deviated from those expected from the Hardy-Weinberg equilibrium for the combined control group (p = 0.79) or for newborn and adult control groups (p = 0.59 and p = 0.73). There was no statistical difference in the C5972T allele frequencies in newborns recruited from the Szczecin metropolitan region compared to other Polish cities (data not shown). According to existing literature data, the frequency of the C5972T variant is similar in other populations such as North American, Austrian and Oceanian [6, 7].

The heterozygosity of the BRCA2 C5972T variant appears to be a neutral polymorphism in women above the age of 40 but may be deleterious prior to age 40. The risk associated with this allele was modest (OR = 1.4) in young women. Strong associations were not observed until cases were subclassified by their histological appearance or by homozygous state. It is important to confirm these observations in other populations. Large, well-controlled studies are needed to establish the full range of risks associated with other BRCA2 alleles before they can definitely be identified as neutral. Identification of breast cancer susceptibility genes that are associated with modest penetrance requires very large association studies, unless the high-risk alleles are very common.