Novel germline mutations and unclassified variants of BRCA1 and BRCA2 genes in Chinese women with familial breast/ovarian cancer

BRCA1 and BRCA2 are the most important genetic susceptibility genes for breast/ovarian cancer in
both Caucasian and Chinese populations. The spectrum and frequencies of mutations
in these two genes in Chinese women with familial breast/ovarian cancer have been
insufficiently explored to date. Moreover, the penetrance has not yet been investigated.
Due to the limited knowledge on hereditary breast/ovarian cancer, there is no genetic
counseling or testing services available in Mainland China.

Our results demonstrated that the frequency of BRCA1 and BRCA2 mutations among Chinese women with familial breast/ovarian cancer was 23.3 %. Similar
results have been reported in the Korean population 23], Hispanic population 24] and Africa American population 25]. However, the frequency observed in the current study is lower than that reported
in an Ashkenazi Jewish population, in which the frequency of BRCA1 and BRCA2 mutations was 69 % 25]. Compared with other reports about Chinese populations, the frequency found in our
cohort was the highest in patients with familial breast/ovarian cancer. Li et al.
9] used PCR-DHPLC assay to screen for BRCA1 and BRCA2 mutations in 241 women with familial breast cancer from northern or southern China
and found a frequency of 12.9 %. Although the PCR-DHPLC assay is cost-effective for
screening for genetic mutations, a considerable number of disease-associated mutations
may have been missed by this indirect detection method 26]. Zhang et al. 11] reported that the frequency of BRCA1 and BRCA2 mutations in northern Chinese familial breast cancer patients was 10.5 % (43/409)
based on PCR-sequencing assay. The enrolment criteria and mutation detecting assay
used in this were comparable with the criteria used in our study, but the reported
frequency was much lower than that observed in the present study. In their subgroup
analysis, the highest frequency was 23 % in the patients whose tumors had been diagnosed
at or before the age of 40 years. However, the frequency reached 33.3 % in this group
of patients in our cohort. Moreover, in the study conducted by Kwong et al., 12] the frequency of BRCA1 and BRCA2 mutations in high-risk breast/ovarian cancer patients was 15.3 % (69/651). These
authors also employed the conventional PCR-sequencing assay, and the patients were
recruited from southern China. The proportion of high-risk breast/ovarian cancer patients,
including familial breast cancer patients and early-onset cases and the frequency
of two-gene mutations were much lower in the early-onset patients than in the familial
breast cancer cases. Large genomic rearrangements account for 4–28 % of all BRCA1 and BRCA2 mutations 27], and such mutations have been found in Chinese women at a high risk for breast cancer
28]–32]. Because the PCR-sequencing assay cannot detect these rearrangements, the frequency
of mutations in our cohort might have been underestimated, and the frequency of BRCA1 and BRCA2 mutations in the eastern Chinese population could be significant.

Although several studies have reported that the BRCA2 mutations are more frequent than BRCA1 mutations in Asian population 11], 12], 33], 34], BRCA1 mutations seemed to be more prevalent in our cohort. This finding might be attributable
to two points. First, most studies have reported that BRCA2 mutations predominantly occur in relatively late-onset breast cancer patients compared
with BRCA1 mutations 11], 35], but the patients enrolled in our study were much younger than those in other studies,
which might have resulted in an underestimation of the contribution of BRCA2 mutations. Second, a greater number of recurrent mutations were found in BRCA1 than in BRCA2 in our study, which elevated the frequency of BRCA1 mutations.

In the present study, we found that 52.2 % (12/23) of the deleterious mutations were
novel; these mutations included five mutations in BRCA1 and seven mutations in BRCA2. In our previous systemic analysis of the spectrum of BRCA1 and BRCA2 mutations in Han Chinese women, we reported that 56.3 % (40/71) and 47.9 % (35/73)
of the BRCA1 and BRCA2 mutations were novel, respectively 22]. It seems that the spectrum of BRCA1 and BRCA2 mutations in Chinese women exhibit unique features. The BRCA2 mutation c.1-40delGA in our cohort was novel. Bakker et al. 36] found a BRCA2 c.1-40 G??A mutation in a Japanese Fanconi anemia family. The functional analysis
of these authors used a mouse embryonic stem cell-based assay that revealed that this
mutation caused aberrant splicing, reduced transcript levels and hypersensitivity
to DNA damaging agents, suggesting that this mutation was likely pathogenic. These
authors thought that this finding was relevant for mutation analysis in hereditary
breast and ovarian cancer syndrome families in a diagnostic setting. The mutation
c.1-40delGA, which deletes a guanine in intron 1 and an adenine in exon 2 and causes
the loss of the donor site of intron 1, should also be pathogenic.

Six BRCA1 and BRCA2 recurrent mutations were identified in multiple patients, and these accounted for
45.2 % (14/31) of the total patients with mutations. Of these mutations, one (c.3780_3781delAG)
was novel, another (c.5468-1del8) was recently reported in Chinese women 11], and the remaining four had been reported in the BIC database. Founder mutations
provide population-specific genetic risk assessment, and facilitate genetic mutation
screening. Thus far, few studies have suggested that putative founder mutations of
BRCA1 and BRCA2 might exist in Chinese women at a high risk for breast cancer, such as the c.981delAT
and c.5470_5477del8 mutations in BRCA1 and the c.3109C??T, c.7436_7805del370 and c.9097_9098insA mutations in BRCA29], 10], 12]. In our cohort, the BRCA1 c.5470_5477del8 mutation and BRCA2 c.3109C??T mutation were both recurrent, but no other three putative founder mutations
was found. Our haplotype analysis revealed that BRCA1 c.5154G??A and c.5468-1del8 mutations were the two putative founder mutations. Since
there are only two patients reported for each of the putative founder mutation, the
founder effects are needed to be investigated by larger sample size of patients. In
our previous study, we reported that the most common recurrent mutations in Chinese
women at high risk for breast cancer are c.5470_5477del8 in BRCA1 and c.3109C??T in BRCA222], which were reported to be the putative founder mutations. However, the study that
enrolled the greatest number of familial breast cancer patients from northern China
did not find these six putative founder mutations except the BRCA1 c.5468-1del8 mutation 11]. The discrepancy regarding the founder mutations in Chinese familial breast cancer
patients may be due to geographic differences. The characterization of BRCA1 and BRCA2 founder mutations and association between the founder mutations and breast cancer
risk should be studied in a large-scale Chinese population size.

Although, elevated mutation rates of BRCA1 and BRCA2 were found in patients who had been diagnosed at or before 40 years of age, no significant
differences were found between the BRCA1 mutation carriers, BRCA2 mutation carriers and non-carriers when compared to a mean age at diagnosis. The
inconsistent results implied that these observations did not withstand multiple comparisons
in our cohort. Breast cancer patients with family histories of ovarian cancer exhibited
the highest overall mutation rate of BRCA1 and BRCA2, which implied that BRCA1 and BRCA2 mutations are more likely to occur in families with a history of both breast and
ovarian cancer. This result is consistent with those of other studies 9], 11].

Eleven UVs were found in our study, and the potentials for these variants to disrupt
the functions of BRCA1 and BRCA2 varied according to the algorithm program used. The UVs accounted for nearly 1/3
of the total mutations/variants in this study. The risks of breast and ovarian cancer
in the UVs carriers might be as high as those in the carriers of the classical pathogenic
mutations. A variety of approaches have been used to investigate the clinical relevance
of these UVs. Co-segregation analysis is regarded as a robust approach because it
is directly related to the disease risk and is not affected by selection bias 37]. The absence of co-segregation provides strong evidence against pathogenicity. Unfortunately,
the samples required for us to perform co-segregation analysis of UVs and the deleterious
mutations in the multi-tumor families were not available.