Identification of recurrent BRCA1 mutation and its clinical relevance in Chinese Triple‐negative breast cancer cohort

Abstract Triple‐negative breast cancer (TNBC) accounts for 15–20% of all newly diagnosed breast cancers, and is enriched for germline mutation of BRCA. In Asian patients diagnosed with breast cancer, 268 deleterious mutations of BRCA1 and 242 of BRCA2 have been identified so far, including a reported BRCA1 frameshift mutation (rs80350973), apparently found only in Asian people, with a low prevalence of 0.3–1.7% in different breast cancer cohorts. Here, we reported the high prevalence (7.2%) of rs80350973 among 125 Chinese patients with TNBC, which implies its mutational predilection for certain breast cancer subtypes. Although its low prevalence had not indicated any particular clinical significance in previous studies, our results associated rs80350973 mutation with cell checkpoint malfunction, and was found to be more common in TNBC patients with high Ki‐67 indices (P = 0.004). As Ki‐67 overexpression is a predictor of poor prognosis in TNBC, inclusion of this mutation into genetic assessments may improve the clinical management of Chinese patients with TNBC.


Introduction
Triple-negative breast cancer (TNBC) is defined by little or no expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). This clinical subtype accounts for 15-20% of all newly diagnosed breast cancers (BCs) [1]. TNBCs possess highly unstable genomes; they are frequently diagnosed in young women and have a worse prognosis than other BC subtypes [2,3]. Furthermore, no promising regimen has so far been found for TNBC. Conventional chemotherapy remains the one effective treatment modality, but its long-term clinical outcomes are unsatisfactory [4,5]. Earlier studies of germline BRCA1 mutations in TNBC have shown wide variation in their prevalence, with a range of 10-40%. These mutations are associated with family history [6,7], risk of recurrence [8,9], and sensitivity to DNA-damaging agents [10,11].
Among Asians, 268 deleterious mutations of BRCA1 and 242 of BRCA2 so far have been documented in patients diagnosed with BC [12]. Several studies of Chinese patients with TNBC found that the prevalence of BRCA1 germline mutations varied from 18.6% to 36.8% [12]. The high frequency of certain mutations in BRCA1/2 has been widely studied, to optimize genetic testing strategies for those at high risk for BC. On the other hand, clinical trials suggest that BRCA1/2 mutations in TNBCs could respond better to platinum-based chemotherapy or other ORIGINAL RESEARCH Identification of recurrent BRCA1 mutation and its clinical relevance in Chinese Triple-negative breast cancer cohort DNA-damaging therapies [13,14]. Notably, current studies mostly focus on identifying novel mutations of BRCA1/2 and their prevalence in TNBC. However, little is known about the high prevalence of certain BRCA1/2 mutations and their clinical relevance in TNBC [12]. In this study, we screened recurrent germline mutations, mainly against the BRCA1 gene, and sought their clinical relevance in Chinese patients with TNBCs. The results of our work may help to interpret the clinical significance of certain recurrent mutations of BRCA1.

Study cohort
A total 125 TNBC of 1300 newly diagnosed BC patients were recruited at our hospital, between January 2013 and June 2015. All subjects gave written informed consent. The inclusion criteria were as follows: (1) patients whose disease was pathologically shown to be negative for ER, PR, and HER2/ErbB2, and (2) for whom complete clinical, pathological and follow-up data were available. The exclusion criteria were as follows: (1) patients without complete follow-up data, and (2) those who suffered nontumor-mediated death. Clinical data were obtained from review of medical records and patient interviews by clinical physicians. This study was approved by the medical ethics committee of our hospital. We initially performed whole-exon mutations screening of 26 patients with TNBC, using panel-based next-generation sequencing (NGS) analysis of 55 susceptibility genes; frequencies of some recurrent mutations were later detected in the larger TNBC cohort using mutation site targeting PCR amplification. Amplified products were submitted to Sanger sequencing.

Panel-based NGS analysis
The first step for NGS technology was use of the TruSeq Custom Amplicon method to design oligo probes that are specific for all coding sequences and intron/exon boundaries of coding exons from the 55 genes that affect BC susceptibility (Table S1), using Illumina Design-Studio (Illumina, Inc., San Diego, CA). For each 150-bp sequence of the target region, a pair of oligo probes were synthesized to hybridize with the 5′ and 3′ ends of the sequence at one end (the other end was complementary to the PCR primers). These oligo probes were used to construct a library containing the necessary nucleotide sequences. The target regions were determined by selecting all exons of the 55 susceptibility genes; however, to include sections of the intron-exon regions, the regions also included 50 nucleotides upstream and downstream of each exon.
Sequencing was performed using the NGS MiSeq Illumina sequencer (Illumina, Inc.). Obtained sequences were aligned to the reference genome (GRCh37/hg19) using MiSeq Reporter software (Illumina, Inc.), which detected discrepancies determining their type, such as deletions, insertions and SNPs. The sequences were analyzed using MiSeq software. As an acceptance threshold value, we selected a Q-score of 30, which corresponds to a 1: 1000 error rate.

Genotyping
Since rs80350973 is a heterozygous mutation, we design a pair of primers to amplify the germline DNA fragment containing this mutational site; unaligned sequences from the deletion site cause "double peaks" in Sanger sequencing, whereas the wild-type will not.
Amplified products were submitted to sequencing in an ABI 3730 automated sequencer using the Big Dye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA) as described by the manufacturer. All analyses were performed in duplicate. Discrepancies in Sanger sequencing outcome between the wild type and rs80350973 mutation are shown in Figure S1.

Immunohistochemistry of Ki-67
In this study, Ki-67 expression was quantified by a visual grading system and was determined by counting 1,000 tumor cells using the Olympus Image Analyzer (magnification 400×). Ki-67 immunoreactivity (Ki-67 index) was recorded as a continuous variable based on the proportion of positive tumor cells (0-100%), regardless of staining intensity. All cases were histopathologically confirmed independently by two experienced pathologists according to ASCO/CAP 2010 criteria.

Statistical analysis
The statistical significance of associations between the BRCA1 frameshift mutation and clinicopathological features of TNBC patients were evaluated using Pearson's χ2 tests or Fisher's exact tests as appropriate. The Mann-Whitney test was used to analyze associations between the Ki-67 index and BRCA1 frameshift mutations. Disease-free survival (DFS) was defined as the time from the date of diagnosis to first recurrence (not including second primary malignancies) or death from BC without a recorded relapse. Lengths of DFS for patients with different axillary lymph node status (N0-2 vs. N3), and Ki-67 index levels (≤0.7 vs.>0.7, and ≤0.14 vs. >0.14) were plotted with Kaplan-Meier curves and compared with log-rank tests. DFS was compared between rs80350973 mutation carriers and noncarriers using a timedependent covariate Cox regression model, which took lymph node status (N0-2 vs. N3) and Ki-67 indices into consideration. All statistical tests were two-sided; P < 0.05 was considered significant. Data analyses were calculated using SPSS ® 19.0 software (SPSS Inc., Chicago, IL).
Germline mutations of 55 genes that affect BC susceptibility All coding regions of 55 genes that affect BC susceptibility were sequenced in 26 patients with TNBC. Gene-based case-control association analysis using Asian population information obtained from the 1000 Genomes Project database. Only deleterious mutations were included in this study. Of all 26 patients with TNBC 6 (23.1%) carried four different BRCA1 mutations, three (11.5%) carried three different PALB2 mutations, three (11.5%) carried three different BRCA2 mutations, and the other three (11.5%) carried three different CDH1 mutations. Notably, four patients carried the same CHEK2 mutation (c.1246A>G). Deleterious mutation sites for each BC-related gene detected in these patients are listed in Table S2.

Validation of germline mutation rs80350973
The rs80350973 mutation was further validated in an additional 100 TNBC patients using PCR and Sanger sequencing. In total, aside from a failed Sanger sequencing as a result of an unqualified DNA sample, this BRCA1 frameshift mutation was detected in 9 (7.2%) of 125 TNBC patients.
Clinicopathological features of the 125 TNBC patients with or without the rs80350973 mutation are shown in Table 3. The rs80350973 mutation carriers and noncarriers did not significantly differ with regard to age at diagnosis, histological type, tumor size, lymph node status, TNM stage, family history, and visceral metastasis status at relapse. However, a higher Ki-67 index was seen in rs80350973 mutation carriers than in noncarriers (P = 0.004, Table 3). Kaplan-Meier survival analysis showed that patients with N3 stage lymph node status had significantly shorter DFS than did those with N0-2 status (P = 0.024, Fig. 1A). Although the subgroups with high versus low Ki-67 indices (≤0.14 vs. >0.14, as recommended by the St. Gallen International Expert Consensus of 2011) showed no statistical discrepancy (P = 0.975, Fig. 1B), shorter DFS was seen in the Ki-67 index >0.7 group than in the ≤0.7 group (P = 0.028, Fig. 1C). Statistical discrepancies in DFS between rs80350973 mutation carriers and noncarriers were not observed with regard to other clinicopathological features. A covariate Cox regression model showed no statistical difference between rs80350973 mutation carriers and noncarriers with regard to DFS (P = 0.335, Fig. 1D).

Discussion
Medical consultations for patients with newly diagnosed BCs increasingly use genetic counseling and testing. The National Comprehensive Cancer Network (NCCN) recently recommended BRCA1/2 testing among women with BC diagnoses who met well-established criteria [15]. In 2010, the NCCN guideline suggested that BRCA1/2 testing was indicated for women with TNBC who were younger than 40 years, based on emerging evidence that the triplenegative phenotype was associated with hereditary cancer syndromes regardless of family history [16,17]. In 2013, the NCCN guidelines officially included women younger than 60 years with TNBC among those for whom genetic testing was appropriate [18].
Earlier studies with enormous cohorts have focused on the prevalence of BRCA1/2 mutations in different patient subgroups classified by race, ethnicity or geographical factors [18][19][20][21]. Other TNBC susceptibility genes such as BARD1 and CHEK2 were also detected at relatively high frequency [22,23] which concords with our findings (Table 2). However, little is known about recurrent germline mutations of BRCA1/2 and PolyPhen was used to predicts possible impact of single amino acid substitution on the structure and function of a human protein using straightforward physical and comparative considerations. 2 The allele frequency of certain recurrent mutations in Asian dependent on the publication data from 1000 Genomes Project data. N/A, Not applicable; TNBC, Triple-negative breast cancer.
its clinical significance in Chinese patients with TNBC. Aside from mutations at specific identifiable sites, their clinical interpretation may be challenging. In this study, we focused on a recurrent germline mutation of BRCA1 and its clinical relevance in Chinese patients with TNBC. We observed that a frameshift BRCA1 mutation (rs80350973) occurs at high frequency (9/125, 7.2%). This mutation was first reported by Suter et al. [24]. in 2004, who identified it in two of 645 (0.3%) sporadic BC patients in Shanghai, China, but not in 319 unaffected health controls or in 342 patients with benign breast disease [24]. The rs80350973 mutation was also identified in one of 60 (1.7%) Korean women with early-onset BC (age ≤40 years) [25] and in one of 70 (1.4%) Chinese women with early-onset BC (age ≤35 years) [26]. In 2008, a larger cohort study of Chinese Han nationality was carried out by Li et al. [27]., aimed at identifying of recurrent BRCA1/2 mutations. They found the rs80350973 mutation in 4 of 489 (0.8%) women with family histories of BC and/or early-onset BC. Among these four carriers, two were diagnosed when younger than 35 years, one of whom had a family history of gastric cancer; the other two carriers had family histories of BC. Two rs80350973 mutation carriers (0.5%) were also found among 426 sporadic BC cases; both were diagnosed when older than 35 years, and neither reported family histories of breast or ovary cancer [27]. According to the studies mentioned above, rs80350973 mutation prevalence remained at a low level among Chinese women with sporadic BC (0.3-0.5%), and showed no significant correlation with family cancer history or early age of onset.
Our results show, for the first time, the high prevalence (7.2%) of rs80350973 mutation among 125 Chinese women with TNBC, which indicates its prevalence in this molecular subtype of BC.
So far, studies have shown that this mutation, which is considered deleterious, is only found in Asian people [12,27]. However, its clinical relevance in TNBC remained unclear. According to ClinVar and BIC database, the rs80350973 mutation induces a frameshift mutation in exon 24 of the BRCA1 gene, which could disrupt the BRCT domain (AA1784-AA1863). This BRCT domain modulates interactions between BRCA1 and proteins that are phosphorylated in response to DNA damage [28]; it has also been shown to bind directly to DNA doublestrand breaks (DSB) [29] and affects execution of the cell-cycle checkpoint [30,31]. Because the BRCT domain could affect proliferation through the cellular checkpoint, disruption of this domain might lead to uncontrolled cell growth. Our data showed that the BRCA1 rs80350973 mutation is only associated with high Ki-67 index in Chinese patients with TNBC (P = 0.004). A previous study showed that overexpression of Ki-67 was associated with BRCA1 mutation [32] and was an indicator of poor prognosis in TNBC [33,34]. Thus, we sought to determine if the rs80350973 mutation could contribute to shorter DFS. However, we failed to find a statistical difference in DFS between rs80350973 mutation carriers and noncarriers (P = 0.335).
In conclusion, this study found a high mutational prevalence of rs80350973 in Chinese patients with TNBC, indicating a mutational prevalence of this variant. Bioinformatics predictions further revealed that this mutation might disrupt the BRCT domain. As the BRCT domain critically affects DSB recognition and execution of checkpoint function, we supposed that rs80350973 mutation contributes to uncontrolled cell growth. This hypothesis

Supporting Information
Additional supporting information may be found in the online version of this article: Figure S1. Identification of BRCA1 rs80350973 mutation in 100 TNBCs using Sanger sequencing. Figure S2. Location of rs80350973-induced secondary structure variation in BRCA1 protein. Table S1. Panel list of 55 genes that affect susceptibility to breast cancer. Table S2. Exome sequencing of 55 genes that affect susceptibility to breast cancer in 26 patients with TNBC.