The aim of the current study was to estimate the prevalence of BRCA mutations in a series of unselected female breast cancer cases from Greece. We identified a deleterious BRCA1 or BRCA2 mutation in 4.7% of the women with breast cancer. The most common mutation was the G5331A mutation, accounting for 1.6% of the cancers diagnosed and 33% of all mutations. Four additional mutations were identified; two in BRCA1 and two in BRCA2. Other studies of unselected populations have reported relatively low BRCA mutation frequencies (range: BRCA1 0-7%; BRCA2 1-3%) (reviewed in ).
One other hospital-based study has evaluated the frequency of BRCA1 (but not BRCA2) mutations among unselected breast cancer cases in Greece . The authors evaluated the prevalence of four specific BRCA1 mutations. They previously reported that these account for 54% of all mutations detected in both genes . They found that 26 of the 987 patients (2.6%) carried one of four mutations in BRCA1: 5382insC (1.3%), 5331G>A (0.4%), and two large genomic rearrangements involving deletions of exons 20 (0.3%) and 24 (0.6%). 5382insC has been previously described in numerous populations while the latter three mutations have only been described in Greek populations. Further, 14 of the 26 (54%) carriers had an early age of diagnosis (<40 years) and 10% of all the early-onset patients carried one of the four mutations. Given that the original data was based on the testing of 287 Greek families with a history of breast and/or ovarian cancer, this was not a study of unselected patients.
Nearly all the other prior studies evaluating BRCA mutation frequencies among patients in Greece have included women with familial breast and/or ovarian cancer (i.e., those with a family history) while the women in the current study were unselected for a personal or family history of cancer [11–16]. In the most comprehensive of these studies, the authors identified 26 BRCA mutations (15 BRCA1 and 11 BRCA2) in 287 Greek breast/ovarian cancer families using dHPLC followed by direct sequencing . Four BRCA1 mutations accounted for more than half of the identified mutations, while the remaining 22 mutations were a mixture of unique mutations occurring at a low frequency suggesting a genetically heterogenous population . The 5382insC and G5331A mutations in exon 20 of BRCA1 accounted for 46% of the families. We did not detect the 5382insC frameshift mutation in our study population. Although the 5382insC mutation is known as a founder mutation in the Ashkenazi Jewish population, Hamel et al. recently estimated that this mutation likely arose approximately 1,800 years ago in Scandinavia or northern Russia and infiltrated the Ashkenazi Jewish population 400-500 years ago . Further, the authors concluded that this is likely the most common mutation in many European countries.
Relatively few BRCA mutations account for BRCA-associated cancers among homogenous populations such as Iceland and Poland, while more heterogeneous populations such as in Canada or the United States tend to display a broad mutation spectrum [18, 22, 23]. Founder mutations are specific mutations that appear repeatedly in ethnically defined groups because of a shared common ancestry. Such mutations have been identified in Icelandic , Polish , French-Canadian , and Ashkenazi Jewish , among other, populations. If genetic testing is to be offered to high-risk individuals, it is more economical to screen for founder mutations rather than comprehensive genetic testing which is considerably more costly.
Based on our and prior findings [14, 16, 28], it appears that G5331A mutation in BRCA1 may be a founder mutation in the Greek population. Previous studies indicate that the 5382insC frameshift [11, 13, 14, 16, 21] is likely a founder mutation as well, although this was not the case in the current study. In addition to these two mutations, many others reported several novel mutations, deletions or genomic rearrangements as well as unclassified variants and [11, 12, 15, 16, 19] suggesting a genetically heterogeneous population.
A major strength of the current study were the inclusion of unselected breast cancer cases thus allowing for an accurate estimation of the proportion of cases due to BRCA mutations. Further, genetic testing for BRCA1 and BRCA2 is highly sensitive . Nonetheless, we were limited by the relatively small sample size (n = 127). The majority of the aforementioned studies were limited by the inclusion of familial cases, the relatively small sample sizes (n range 25-287) and the screening of one rather than both BRCA genes.