Content uploaded by Guangyu Shan
Author content
All content in this area was uploaded by Guangyu Shan on Dec 25, 2019
Content may be subject to copyright.
Human Mutation. 2019;1–13. wileyonlinelibrary.com/journal/humu © 2019 Wiley Periodicals, Inc.
|
1
Received: 1 October 2018
|
Revised: 18 September 2019
|
Accepted: 17 November 2019
DOI: 10.1002/humu.23965
RESEARCH ARTICLE
Comprehensive profiling of BRCA1 and BRCA2 variants in
breast and ovarian cancer in Chinese patients
Xianqi Gao*
|
Xiyan Nan*
|
Yilan Liu
|
Rui Liu
|
Wanchun Zang
|
Guangyu Shan
|
Fei Gai
|
Jingfeng Zhang
|
Lei Li
|
Gang Cheng
|
Lele Song*
Novogene Co. Ltd., Beijing, China
Correspondence
Gang Cheng and Lele Song, Novogene Co.,
Ltd., Zone A10 Jiuxianqiao North Road,
Chaoyang District, Beijing 100015, China.
Email: jeff.cheng@novogene.com (G. C.) and
songlele@sina.com (L. S.)
Abstract
The identification and interpretation of germline BRCA1/2 variants become
increasingly important in breast and ovarian cancer (OC) treatment. However, there
is no comprehensive analysis of the germline BRCA1/2 variants in a Chinese
population. Here we performed a systematic review and meta‐analysis on such
variants from 94 publications. A total of 2,128 BRCA1/2 variant records were
extracted, including 601 from BRCA1 and 632 from BRCA2. In addition, 414, 734, 449,
and 307 variants were also recorded in the BIC, ClinVar, ENIGMA, and UMD
databases, respectively, and 579 variants were newly reported. Subsequent analysis
showed that the overall germline BRCA1/2 pathogenic variant frequency was 5.7%
and 21.8% in Chinese breast and OC, respectively. Populations with high‐risk factors
exhibited a higher pathogenic variant percentage. Furthermore, the variant profile in
Chinese is distinct from that in other ethnic groups with no distinct founder
pathogenic variants. We also tested our in‐house American College of Medical
Genetics‐guided pathogenicity interpretation procedure for Chinese BRCA1/2
variants. Our results achieved a consistency of 91.2–97.6% (5‐grade classification)
or 98.4–100% (2‐grade classification) with public databases. In conclusion, this study
represents the first comprehensive meta‐analysis of Chinese BRCA1/2 variants and
validates our in‐house pathogenicity interpretation procedure, thereby providing
guidance for further PARP inhibitor development and companion diagnostics in the
Chinese population.
KEYWORDS
BRCA,BRCA1,BRCA2, breast cancer, Chinese, ovarian cancer, pathogenicity interpretation,
variant
1
|
INTRODUCTION
Pathogenic variants in the cancer predisposition genes (BReast CAncer
gene) BRCA1 and BRCA2 are strongly associated with the occurrence
and development of breast cancer (BC) and ovarian cancer (OC;
Balmana, Diez, & Castiglione, 2009; Narod, 2010; Trainer et al., 2010). A
recent extensive cohort study has indicated that the cumulative risk of
acquiring BC or OC by 80 years of age was 72% and 44%, for BRCA1
Abbreviations: ACMG, American College of Medical Genetics; AMP, Association for
Molecular Pathology; B, benign; BC, breast cancer; BRCA, BReast CAncer gene; CDS, coding
sequence; HER2, human epidermal growth factor receptor 2; HGVS, human genome
variation society; LB, likely benign; LP, likely pathogenic; LR, long‐range rearrangement;
NGS, next‐generation sequencing; OC, ovarian cancer; P, pathogenic; PARPi, poly (ADP‐
ribose) polymerase (PARP) inhibitor; TNBC, triple‐negative breast cancer; UTR, untranslated
region; VUS, variant of uncertain significance.
*Xianqi Gao, Xiyan Nan and Lele Song contributed equally to this study.
mutant carriers (carrier refers to heterozygote throughout the whole
study), and was 69% and 17% for BRCA2 mutant carriers, respectively
(Kuchenbaecker et al., 2017). In addition, distinctive prognoses were
found to be related to the BRCA1/2 mutation status in the respective
BC and OC patients. A prospective young‐onset BC study has shown no
significant difference in the 10‐year overall survival of BRCA1/2 mutant
carriers versus noncarriers (Copson et al., 2018). Interestingly, BRCA1/2
mutations have been associated with improved overall survival in OC
patients. The 5‐year overall survival of epithelial OC was 36% for
noncarriers, and 44% and 52% for BRCA1 and BRCA2 mutant carriers,
respectively (Bolton et al., 2012). Moreover, platinum‐based chemother-
apy has been found to be highly effective for the treatment of
metastatic triple‐negative breast cancer (TNBC) and OC in BRCA1/2
mutant carriers (Alsop et al., 2012; Isakoff et al., 2015).
A series of clinical trials have shown that poly (ADP‐ribose)
polymerase (PARP) inhibitors (PARPis) are effective at well‐tolerated
doses, providing an antitumor activity for cancers containing BRCA1/2
abnormalities. Commercial PARPis, including olaparib, niraparib, and
rucaparib, have been approved by the Food and Drug Administration
(FDA) for the treatment of OC (Audeh et al., 2010; Coleman et al.,
2017; Mirza et al., 2016; Pujade‐Lauraine et al., 2017; Swisher et al.,
2017) and BC (Robson, Goessl, & Domchek, 2017) at advanced stages.
Multiple clinical trials are currently undergoing to test PARPis in
monotherapy (e.g., olaparib in prostate cancer trial: NCT02987543) or
in combination with other therapeutic agents, such as immune
checkpoint inhibitors (e.g., rucaparib + atezolizumab in advanced
gynecologic cancer and TNBC trial: NCT03101280). The BRCA1/2
variant status has been frequently utilized as a selection criterion or
stratifying factor in these clinical trials. BRCA1/2 testing has been
extensively performed with FDA‐approved diagnostic products such
as FoundationFocus CDxBRCA and BRACAnalysis CDx.
The prevalence of germline BRCA1/2 variants shows a large
variation across different ethnicities, ranging from 6.5% to 25.0% in
BC and from 12.1% to 29% in OC (Han et al., 2013; Kim et al., 2016;
Robson et al., 1997; Weitzel et al., 2005). Due to remarkable
advances in next‐generation sequencing (NGS) technologies (Rain-
ville & Rana, 2014), it is now possible to investigate the prevalence of
different BRCA1/2 variants in patients of distinct ethnicities, as well
as their clinical significance (Sun et al., 2017; Wu et al., 2017). It is
known that the homologous recombination repair function can be
affected by pathogenic BRCA1/2 variants, which are randomly
distributed along with coding and noncoding regions, with no defined
hotspots (Weitzel et al., 2005). Contradicting interpretation of the
roles of many BRCA1/2 variants is largely found in the current
literature, mostly due to the discrepant understanding of the
functions and the clinical significance of these variants (Blackwood
& Weber, 1998; Oglesbee et al., 2018). To standardize a genetic
variant interpretation, the American College of Medical Genetics
(ACMG) and the Association for Molecular Pathology (AMP) jointly
issued revised guidelines for variant classification (Richards et al.,
2015). However, expert judgment on the quality of available
evidence is still warranted for a more precise interpretation. In the
context of BRCA1/2 gene variation, both “BRCA1/2 Gene Variant
Classification Criteria from ENIGMA Consortium”(Spurdle et al.,
2012) and “China expert consensus on BRCA variant interpretation”
(Panel members of China expert consensus on BCRA variant
interpretation, 2017) have been released. These guidelines have
contributed to the standardization of the BRCA1/2 variant inter-
pretation; however, the variant database, the bioinformatic pipeline
and the parameters used during their interpretation still need to be
validated in the practice.
To tackle some of the challenges indicated here, we have performed
an extensive review of both Chinese and English scientific publications
on Chinese germline BRCA1/2 variants. We have assembled and
analyzed all reported BRCA1/2 variants to better understand the
characteristics of BRCA1/2 variants in the Chinese population, and
compared them to other ethnicities. Furthermore, we have also
integrated different guidelines, along with our own evaluation and
evidence, to establish an interpretative procedure for BRCA1/2 variants
in the Chinese population. As a result, we have analyzed the distribution
of the pathogenic variants according to exon location, variant type, and
key domain. Our interpretative procedure and results were validated
through a comparison with well‐established public databases.
2
|
METHODS
2.1
|
Search strategy and selection criteria
Systematic search and review of relevant publications were performed
in accordance with the Preferred Reporting Items for Systematic
Reviews and Meta‐Analyses (PRISMA) guidelines (Liberati et al., 2009;
Figure S1). Two of our investigators independently searched PubMed,
Google Scholar, CNKI, and Wanfang databases and identified relevant
nonreview literature up to the end of November 2017, using a
combination of terms including “BRCA1/2”and “Chinese”or “China”or
“Hong Kong”or “Taiwan,”and “breast cancer”or “ovarian cancer”or
“prostate cancer”in English or Chinese, separately. All author names,
affiliations, recruitment period of patients and characteristics of each
subject, enrolled in the respective study, were properly checked to
allow the exclusion of redundant reports retrieved from the
aforementioned literature databases. If the same group of authors
published several reports, the recruitment period and the character-
istics of patients of each study were rechecked to exclude any
overlapping study. If so, the study that covered the largest subject
population was adopted while other reports with a smaller sample size
would be removed to minimize potential double‐counting. Studies on
specific subgroups, such as TNBC, hereditary (or familial) BC,
hereditary BC and OC and others, were classified into each subgroup
separately to avoid affecting the overall accuracy of BC or OC data.
2.2
|
Data extraction and quality assessment
All studies were categorized according to the first author and the
year of publication and the following information from extracted all
2
|
GAO ET AL.
collected articles: BRCA1/2 variant information (BRCA1/2 variant
designation, corresponding carrier number, and variant type), cancer
type, phenotypic subtype, sample size, number of BRCA1/2‐mutated
carriers, detection method, principle investigator, and province/area/
location.
The format standardization of BRCA1/2 variant nomenclature
was performed after the collection. The BRCA1/2 variant information,
including the physical location in the chromosome (locus), reference
and alteration sequences, was extracted to call the standard
designation in the human genome variation society (HGVS) nomen-
clature, which followed the HGVS checklist (den Dunnen, 2016).
NM_007294.3 (BRCA1) and NM_000059.3 (BRCA2) were used as
transcript reference sequences, while NP_009225.1 (BRCA1) and
NP_000050.2 (BRCA2) were used as protein reference sequences to
annotate the respective BRCA1/2 variants. In addition, the systematic
exon numbering of the BRCA1 gene, rather than the conventional
ones, was used in this study. Moreover, the format of BRCA1/2
variants, described according to the mRNA position in the original
studies, was transformed into HGVS nomenclature according to the
format of the cDNA position.
2.3
|
Interpretation procedure of BRCA1/2 variant
The interpretation of each BRCA1/2 variant in our study followed the
guidelines of the ACMG/AMP (Richards et al., 2015), the ENIGMA
BRCA1/2 Gene variant Classification Criteria (Spurdle et al., 2012) and
the China expert consensus on BRCA1/2 variant interpretation (Panel
members of China expert consensus on BCRA variant interpretation,
2017). We integrated these guidelines systematically and established
an ACMG guideline‐based interpretation procedure specifically for the
BRCA1/2 genes of the Chinese population. Specific details about the
procedures of BRCA1/2 variant interpretation are shown in Figure S2.
On the basis of this methodology, BRCA1/2 variants were classified
into five grades, according to their probability of pathogenicity:
Pathogenic (“P”), Likely Pathogenic (“LP”), Variant of Uncertain
Significance (“VUS”), Likely Benign (“LB”), and Benign (“B”;FigureS3).
Two‐grade classification (i.e., pathogenic or nonpathogenic) was also
used, in which the “P”and “LP”in 5‐grade classification were
considered as Pathogenic, and VUS, LB, and B in 5‐grade classification
were considered as Non‐Pathogenic. “Pathogenic”means that the
variant carrier should consider genetic counseling and risk‐reducing
surgery, or might benefit from PARPi or platinum‐based chemotherapy
treatment, and vice versa. This definition of “pathogenic”was applied
throughout this study unless specified otherwise.
2.4
|
Data analysis and statistics
The frequency of BRCA1/2 pathogenic variants in BC, OC, and each
phenotypic subgroup was calculated by pooling analysis across
different studies. The numbers of variant carriers from different
studies were pooled together to obtain the overall variant carrier
number and were ranked in descending order to identify the high‐
frequency variants. Variants without information of carrier number
in their original publications were considered as “one”carrier only.
The statistical data were then used to analyze the variant
frequency and distribution. The variant data from BIC were used to
calculate the high‐frequency variants in non‐Chinese ethnicities. BIC
was used because it was the only public database containing
information on both variant population frequency and ethnicity.
The distribution of pathogenic variants in each exon, intron, variant
type and the key domain was calculated accordingly. Cross analysis
between our collection and other well‐known public databases (BIC,
ENIGMA, ClinVar, and UMD) were performed by InteractiVenn. We
analyzed the interpretation consistency for each of them to validate
our BRCA1/2 variant interpretation procedure and results.
3
|
RESULTS
3.1
|
Summary of BRCA1/2 studies focusing on the
Chinese population (period of 1999–2017)
We have identified 94 eligible publications on Chinese BRCA1/2 variants,
including 43 English publications and 51 Chinese publications, 72 peer‐
reviewed papers, and 22 graduate theses (Figure 1 and Table S1). In the
greater China region, Hong Kong, and Taiwan, the earliest BRA1/2 papers
were published in 1999, followed by a rapid growth in the total number
of publications from 1999 to 2017 (Figure 1a), with Beijing, Shanghai,
Hong Kong, and Xinjiang at the top of the list (Figure 1b,c). The first
English BRCA1/2‐related paper coming from Mainland China was seen in
2000. In 2015, we witnessed a rapid growth of BRCA1/2 research papers
published from Mainland China.
Advancements in technology used to detect the BRCA1/2 gene
variants were similarly observed (Table S1). Sanger sequencing and
PCR were the main technologies used to detect the BRCA1/2 gene
variants since the 1990s. It was not until the year 2015 that the NGS
technology was widely used to facilitate the BRCA1/2 gene variant
research in clinical genetics in Mainland China. Most sequencing
methods covered both the whole coding region and exon–intron
boundaries of BRCA1/2 genes, leading to a whole coverage of
potential variants and more robust variant data quality during more
recent years.
3.2
|
Frequency and whole‐gene distribution of
BRCA1/2 variants in the Chinese population
Most Chinese BRCA1/2 studies focused on BC and OC, and some of
them included healthy subjects as a control group. Our comprehen-
sive review identified 2,970 people with BRCA1/2 germline variants.
From these, 2,128 BRCA1/2 variant records were extracted for
subsequent analyses. These extracted records originated from
35,178 Chinese people enrolled in a BRCA1/2 variant screening from
1999 to 2017. As shown in Figure 2, the overall BRCA1/2 pathogenic
GAO ET AL.
|
3
variant frequency was 5.7% in Chinese BC patients, and 21.8% in
Chinese OC patients.
We analyzed the high‐risk factors and molecular subtypes that
may affect the detection of the BRCA1/2 pathogenic variant in BC
patients. Most studies were defined by early‐age onset, family
history, male, and bilateral as high‐risk factors for BRCA1/2
pathogenic variation. Among these factors, early‐age onset BC and
BC with family history represented the two groups with the largest
number of patients (Table S2). As shown in Figure 2a, the overall
frequency of BRCA1/2 pathogenic variants for high‐risk BC patients
was 14.4%. Interestingly, we found that the frequency of BRCA1/2
pathogenic variants in early‐age onset BC was relatively low (7.4%).
Three groups of patients have the highest prevalence: those with a
family history (15.9%), male BC patients (14.5%), and bilateral BC
patients (16.6%). In contrast, the frequency of BRCA1/2 pathogenic
variants was only 2.8% in sporadic BC patients and 0.4% in Chinese
healthy subjects (control). The BC molecular subtype was a key
factor affecting the incidence of BRCA1/2 pathogenic variants
(Figure 2b). Among all molecular subtypes, TNBC, the most studied
molecular BC subtype in regards to BRCA1/2 detection, exhibited the
highest frequency of BRCA1/2 pathogenic variants (11.2%). In
contrast, the human epidermal growth factor receptor 2 (HER2)
positive BC subtype showed the lowest frequency (1.7%). The variant
frequency in Luminal A and B subtypes varied in different studies.
We found that Luminal A subtype exhibited small but a significantly
higher variant frequency (5.3%) than the Luminal B subtype (3.8%).
It has been reported that the frequency of BRCA1/2 pathogenic
variants in OC among non‐Chinese ethnicities is around 13% (Cancer
Genome Atlas Research Network, 2011). Our analysis has shown
that this frequency among Chinese OC patients was higher (21.8%)
(Figure 2c). In fact, one large sample size screening (826 patients)
indicated that this variant frequency in Chinese OC patients was
28.5% (Wu et al., 2017). Family history appeared to be a strong risk
factor since BRCA1/2 pathogenic variant was found in 41.0% of
patients with family history. The frequency of pathogenic variants
reached 34.0% in platinum‐sensitive OC patients, and 34.5% in
patients experiencing two or more lines of therapies, suggesting that
platinum‐sensitive and multiple therapeutic approaches may be also
considered high‐risk factors.
Comprehensive analyses of the type and distribution of BRCA1/2
variants in the Chinese population were further performed (Figure 3).
Chinese BRCA1/2 variants spread along most of the exons and
FIGURE 1 The number of publications on BRCA1/2 variants from China from 1999 to 2017. Panel a shows the numbers and comparison of
publications in mainland China (Chinese publications), mainland China (English publications), Hong Kong, and Taiwan (all English publications)
from 1999 to 2017. Panel b shows the distribution of publications throughout Chinese province or area. Panel c shows the detailed numbers for
each province. The color bar represents the number of publications in Chinese province or area in Panels b and c. Red, 15–20 papers; orange,
6–15 papers; yellow, 3–6 papers; light green, 2–3 papers; dark green, 1–2 papers; gray, 0 papers
4
|
GAO ET AL.
introns of BRCA1 and BRCA2 genes. No distinct dominant hotspot
variants or potential founder variants were identified, though there
were some variants with relatively higher population frequency than
others (Figure 3a,b). For instance, the c.5470_5477delATTGGGCA
(p.Ile1824AspfsTer3) variant in the BRCA1 gene and the c.3109C>T
(p.Gln1037Ter) variant in the BRCA2 gene ranked as the highest in
variant frequency in the Chinese population. We have also analyzed
the distribution of long‐range rearrangements (LRs) variants in the
Chinese population (Figure 3c). A total of 18 types of LRs were
reported among the Chinese BRCA1/2 variants (14 in BRCA1 and 4 in
BRCA2), where four of them were reported more than twice. The LRs
were more frequent in BRCA1 than those in BRCA2 (17 vs. 9 carriers),
and there were many more carriers with deletion variants than
duplication‐related variants (23 vs. 3 carriers), which were consistent
with the reports from Western populations (Richards et al., 2015).
The statistical data for each BRCA1 and BRCA2 variant type are
shown in Figure 3d,e, respectively. There were a total of 601 types of
variants reported in BRCA1, and 632 types reported in BRCA2 (895
variants were repeatedly reported in different articles out of a total
of 2,128 variants). Although the number of variant types of BRCA2
was slightly higher than that of BRCA1, the number of BRCA1 variant
carriers was higher than that in BRCA2 (1,847 vs. 1,405 carriers).
Frameshift variants were the most frequent variant types, followed
by missense, nonsense, intronic‐untranslated region (UTR) and
splicing variants (Panel D). These five types of variants accounted
for approximately 95% of variants and 97% of carriers.
3.3
|
Cross‐ethnicity comparison of BRCA1/2
variants
The high frequency of BRCA1/2 variants in the Chinese population
was analyzed and compared with those from other ethnicities. The
top 20 variants in BRCA1 and BRCA2 among the Chinese, Caucasian
non‐Jewish, Ashkenazi Jewish, Africans, and Mongolians are illu-
strated in Figure 4 (raw data in Table S3). Data from the BIC
database were used for comparison, since it included the most
comprehensive ethnicity information among all databases here
analyzed. With regard to BRCA1/2, we observed that frameshift,
nonsense, and missense mutations composed the main variant types
FIGURE 2 The frequency of germline BRCA1/2 pathogenic variant in BC and OC patients. Panels a and b show the pathogenic variant
frequency in BC, grouped by risk factors (Panel a) or BC molecular subtypes (Panel b). Panel c shows the pathogenic variant frequency in OC
grouped by family history, histological typing, platinum sensitivity, and therapeutic condition. BC, breast cancer; OC, ovarian cancer
GAO ET AL.
|
5
in the Chinese population, with no obvious founder variant identified
(Figures 4a and 4f). In contrast, data from the BIC database showed
that the distribution and frequency of variants among Caucasians
were different from that among the Chinese (Figures 4b and 4g).
Founder variants were identified in Ashkenazi Jewish, among which
(a) c.66_67delAG and c.5263_5264insC in BRCA1 and (b)
c.5946_5946delT in BRCA2 were the three most frequently reported
variants in the BIC database (Figures 4c and 4h). The top two BRCA1
FIGURE 3 The distribution of various types of BRCA1/2 variants on full‐length BRCA1/2 exons. Panels a and b show the distribution of six
types of variants on BRCA1 and BRCA2, respectively. The scheme of exons (blue bar at the bottom of each panel) is shown as the reference.
Panel c shows the distribution of long‐range rearrangement variants in BRCA1 and BRCA2. Panel d shows the number of different variants
(inner ring) and the number of variant carriers (outer ring). Colors represent different variant types and the legends are shown in Panels a and d,
respectively
6
|
GAO ET AL.
FIGURE 4 Cross‐ethnicity comparison of variant frequency and distribution. Panels a–e show the top 20 variant types of BRCA1 in
descending order for five populations (as labeled), and Panels f–j show the top 20 variant types of BRCA2 in descending order for five
populations (as labeled). Colors represent different variant types and the legends for colors are shown on the top of the figure. The details of
each variant are labeled under the x‐axis of each panel
GAO ET AL.
|
7
variants were also among the top ones in non‐Jewish Caucasians
(Figure 4b). Despite the limited data from BIC with regard to the
African (Figures 4d and 4i) and Mongolian populations (Figures 4e
and 4j), the top 20 BRCA1/2 variants showed a clear trend in these
two ethnicities, in which missense was the main variant type,
especially among the Africans.
3.4
|
Genetic distribution of pathogenic BRCA1/2
variants
The distribution of pathogenic BRCA1/2 variants in key domains along
with exons or introns, as well as understanding the roles of each
particular variant is of seminal importance for therapeutic advances in
BC and OC. For this, here we profile the distribution of pathogenic
variants in main exons and introns of BRCA1 and BRCA2 (Figure 5a,b).
A total of 57.1% and 59.6% of pathogenic variants were distributed in
Exon 10 of BRCA1 and Exon 10/11 of BRCA2, respectively. Among the
variants here collected, the overall pathogenic frequency was 69.9%
and 71.1% in BRCA1 and BRCA2 exons, respectively. Pathogenic
variants along introns were mostly restricted to splicing regions. The
number of variants in each exon was normalized according to the exon
length (Figure 5c,d). The average frequency of BRCA1 variants along
the whole coding sequence (CDS) region was 8.9%, while the average
frequency of pathogenic variants was 6.2%. For BRCA2,theaverage
frequency of variants was 5.9% in the CDS region, and the frequency
of pathogenic variants was 4.2%. The Exon 2 in BRCA1 and the Exon 5
in BRCA2 exhibited the highest number of variants, while Exon 18 in
BRCA1 and Exon 23 in BRCA2 exhibited the highest number of
pathogenic variants.
Proper clarification of the pathogenicity of respective BRCA1/2
variants is crucial for their interpretation. According to the 5‐grade
classification of pathogenicity from ACMG, we analyzed the patho-
genicity of variants in each variant type. The central panel of Figure 5e
shows the ratio of each type of variant among all types of variants
examined here. On the basis of this data, it is clear that the frameshift,
missense, and nonsense variants are the main variant types
(Figure S3). The peripheral panels show the ratio of variants in each
one of the five pathogenic classes. As indicated, the ratio of pathogenic
variants (“P”) was very high among the frameshift, nonsense, splicing,
and LR variants, possibly due to loss of function. In contrast, the
missense, intron‐UTR, synonymous and in‐frame variants were more
challenging to interpret. The majority of these variants were classified
as “VUS”due to conflict or lack of evidence. Furthermore, not only “P”
(pathogenic) or “LP”(likely pathogenic), but also “B”(benign) or “LB”
(likely benign) were found in missense, intron‐UTR and synonymous
variants, reiterating the importance of a more accurate interpretation.
Next, we focused on variant types with a lower frequency and
controversial pathogenicity prediction, including missense, synon-
ymous and in‐frame variants (Figure 5f,g). Although these account for
25% of BRCA1/2 variants, only 7.6% of these variant types were
regarded as pathogenic. By examining the distribution of these
variants at clinically important domains (as suggested by ENIGMA
[Spurdle et al., 2012]), we verified that 21.8% of these variants in
the key domains/motifs of BRCA1/2 were pathogenic, mostly in the
RING, BRCT and the exon boundary regions of BRCA1 as well as the
DBD and exon boundary regions of BRCA2 (Figure 5f,g).
3.5
|
Interpretation of Chinese BRCA1/2 variants
by comparative database analysis
To accurately interpret the pathogenicity of BRCA1/2 variants here
discussed, we analyzed single nucleotide variants and small indel
variants in our study and compared them with those of the public
BRCA1/2 database (BIC, ClinVar, ENIGMA, and UMD). LRs were
excluded from this comparison because BIC did not include this
particular variant type, and ClinVar lacked detailed information
regarding to LRs. As a result, 585 BRCA1 variants and 628 BRCA2
variants (in total, 1,213) were included in this comparative analysis.
The numbers of overlapping variants were 414, 734, 449, and 307 for
BIC, ClinVar, ENIGMA, and UMD, respectively (Figure 6a,b). The
nonoverlapping variants were novel variants found in the Chinese
population, and were not included in any public database.
Our interpretative comparison was performed one‐by‐one, based on
the 5‐gradeandthe2‐grade pathogenicity classifications. Since 159
variants in BIC were annotated as “pending”and 41 variants in ClinVar
were shown as “not provided,”these variants were excluded from our
comparison. A high consistency was achieved between our interpretation
and that from the public database (Figures 6c and 6f). The consistency
based on the 5‐grade classification (green squares) was 97.6%, 96.1%,
96.1%, and 91.2% for BIC, ClinVar, ENIGMA and UMD, respectively,
while the consistency based on the 2‐grade classification (both green and
yellow squares) was 98.4%, 99.1%, 100.0%, and 98.7% for BIC, ClinVar,
ENIGMA, and UMD, respectively. The highest consistency was achieved
with BIC in the 5‐grade classification, and with ENIGMA in the 2‐grade
classification. Table S4 shows details about the inconsistent interpretation
for each variant in this comparative analysis.
4
|
DISCUSSION
4.1
|
Characteristics of germline BRCA1/2 variants
in BC and OC in the Chinese population
There were not many Chinese studies focusing on BRCA1/2 variants
until 2015, when the number of published papers started to grow and
kept increasing in the following years. The testing of BRCA1/2 variants
has attracted more attention in the past 3 years, partially due to the
success of PARPis in clinical studies. The expansion of BRCA1/2
research in China allowed us to obtain a comprehensive profile of the
BRCA1/2 variants in the Chinese population, and to further review the
findings and provide insightful analysis on the characteristics of these
variants. On the basis of this study, a Chinese BRCA1/2 database is
expected to be designed, where future novel strategies on BRCA1/2
testing and PARPi treatment will be established.
8
|
GAO ET AL.
FIGURE 5 Interpretation of pathogenicity and distribution of pathogenic and nonpathogenic variants at full‐length BRCA1/2 genes. Panels a
and b show the number of variants in each exon and intron of BRCA1 and BRCA2 genes, respectively. Panels c and d show the number of variants
normalized to exon length for BRCA1 and BRCA2, respectively. Panel e shows the relative ratio of each type of pathogenicity in each type of
variant, and the central figure of Panel E shows the ratio of each type of variant. Panels f and g show the interpretation of pathogenicity of
missense, synonymous and in‐frame variants in key domains of BRCA1/21andBRCA1/22, respectively
GAO ET AL.
|
9
In this study, we confirmed the high‐risk factors for BC in the
Chinese population. We found that the overall frequency of
pathogenic BC variants in the Chinese population was similar to what
have been reported for the western population (Sun et al., 2017).
Patient groups prone to BC risk factors exhibited a significantly higher
frequency of pathogenic variants than the overall and sporadic groups.
More significant risk factors were related to family history, male and
bilateral BC, while the early‐onset group showed a small but still
significantly higher frequency of pathogenic variants. Some pathogenic
variants were still detected in healthy controls (at a very lowest
frequency) suggesting that a more detailed screening for BRCA1/2
pathogenic variants is still warranted to identify those individuals with
genetic high risk. The increased frequency of pathogenic BRCA1/2
variants in HER2‐negative BC, compared with other molecular types,
suggested some intrinsic properties of the malignancy, therefore
justifying the application of PARPis for this type of cancer.
The ratio of OC patients with germline BRCA1/2 pathogenic variants
appeared to be much higher than that of BC patients, which is consistent
with previous reports (Cancer Genome Atlas Research Network, 2011).
This suggests that PARPis could be applicable, at higher extent, for OC
patients. Interestingly, OC patients sensitive to platinum therapy
exhibited an increased frequency of pathogenic variants than the overall
group, supporting the use of “platinum sensitivity”as a criterion for
patient selection for PARPi therapy (Liberati et al., 2009). Meanwhile,
patients with two or more lines of therapy also exhibited a high
frequency of pathogenic variants. This is possibly due to the fact that
patients with BRCA1/2 pathogenic variants are selected for PARPi
therapy and may enter into multiple lines of therapy.
No obvious gene “hotspots”were found for the BRCA1/2 variants
identified in the Chinese population, although a few variants with
relatively higher frequency have been identified. Frameshift, mis-
sense, and nonsense variants were the top three BRCA1/2 variant
types, in which nearly 100% of frameshift and nonsense were
pathogenic. With the overall pathogenic frequency of 69.9% and
71.1% in BRCA1 and BRCA2 exons, respectively, our study suggests
that the large majority of BRCA1/2 variants analyzed here are
pathogenic. Therefore, new specific tests looking at exons and intron‐
exon boundaries will be necessary for a more precise clinical
diagnosis and treatment. It is also noteworthy that several LRs,
especially involving exon deletions, have been identified in both
BRCA1 and BRCA2. Moreover, it has been reported that LRs account
for approximately 6% of BRCA1/2 variants in Chinese patients
FIGURE 6 Comparison of pathogenicity interpretation between the Chinese BRCA1/2 database (this study) with public databases. Panels a
and b show the Venn diagram comparing the included number of variants in the Chinese BRCA1/2 database (this study), BIC, ClinVar, ENIGMA,
and UMD databases in BRCA1/21 and BRCA2, respectively. Panel c shows the detailed comparison of pathogenicity interpretation between the
Chinese BRCA1/2 database (this study) with the BIC, ClinVar, ENIGMA, and UMD based on the ACMG 5‐grade classification. The comparison
results are shown in different color squares, among which the green square represents identical interpretation, the yellow square represents
different interpretations with identical clinical actions, and the red square represents different interpretations with potential different clinical
actions. ACMG, American College of Medical Genetics
10
|
GAO ET AL.
(Kwong et al., 2015), and these mutations are definitely pathogenic
when leading to a frameshift. Since limited data about LRs have been
collected in our analysis, the application of BRCA1/2 LR detection
requires more detailed attention in the future.
4.2
|
Comparison of germline BRCA1/2 variants
reveals ethnicity‐specific diversity and heterogeneity
The types and numbers of the top‐ranked variants, among different
ethnicities, were found to be considerably diverse and heterogeneous.
With regard to BRCA1/2, no obvious “hotspots”variants could be
found in the Chinese, African, and Mongolian (non‐Chinese) popula-
tions, while a couple of “hotspots”were identified for Caucasians (non‐
Jewish) in BRCA1, and for the Jewish population in BRCA1/2.Data
suggested that each population would have a unique spectrum of
BRCA1/2 variants, though the Chinese and Mongolians may have a
consanguineous relationship. The most distinct variant distribution
was that of the Jewish, which had two predominant variants for
BRCA1 and one for BRCA2, besides much smaller carrier numbers for
other variants. This observation has been previously reported (Brody
& Biesecker, 1998; Kwa et al., 2014; Kwong et al., 2015; Pennington
et al., 2014; Schmidt et al., 2017; Tobias et al., 2000), confirming the
quality of our data with regard to the cross‐ethnicity comparisons. The
Caucasian (non‐Jewish) shared two predominant BRCA1 variants with
the Jewish, but their prevalence was not as obvious as that in the
Jewish population. These facts suggest that BRCA1/2 variants were
broadly distributed along with BRCA1/2 genes, without obviously
predominant hotspots in non‐Caucasian populations. This was possibly
due to the long‐term population migration and hybridization which,
therefore, did not establish a founder effect. Since the Jewish
traditionally do not marry people from other ethnic groups, and the
predominant variants are inherited within the population, the founder
effect was more obvious. The top‐ranked variants in Caucasians
overlapped with the Caucasian variants enriched in the worldwide
study of BRCA1/2 gene variants (Rebbeck et al., 2018).
It is also noteworthy that each variant type exhibited distinct
characteristics across different populations. Frameshift and missense
variants were the predominant types in Chinese and Caucasians
(including Jewish and non‐Jewish population), while missense
(nonframeshift) variants were the leading types in Africans and
Mongolians, especially in BRCA2. Due to the limited data acquired for
the African and Mongolian populations using the BIC database, the
interpretation of comparisons between Chinese and these population
groups needs to be cautiously investigated.
4.3
|
Significance of the first systematic
interpretation of pathogenicity for BRCA1/2 variants
in the Chinese population
In this study, we have performed the first systematic collection
and the pathogenicity interpretation of BRCA1/2 variants in the
Chinese population. With the accumulating success of PARPis in
treating OC, BC, and other tumors, the need for better testing and
interpretation of distinct BRCA1/2 variants is rapidly evolving. An
accurate interpretation requires both population‐specific database
and careful one‐by‐one examination of interpreting results, based
on ACMG guidelines. As the significance of many newly found
variants in the Chinese population has not been fully elucidated,
we have built our own Chinese BRCA1/2 variant database, which
included scientific and clinical evidence beyond those provided by
commercial databases. The accuracy of our interpretation was
verified by a one‐by‐one variant check, according to ACMG
guidelines. Altogether, this study will allow a better understanding
of the BRCA1/2 variant landscape in Chinese patients, as well as
facilitating a more accurate interpretation of these variants. Still,
due to the relatively small number of included variants compared
with those in databases of the western population, we will need to
continuously update the database as new information becomes
available. In addition, a large number of “VUS”variants is currently
present in our database (334 variants, which accounts for 27.1% of
collected variants). The interpretation of these variants will need
to be updated as new evidence is reported.
Despite the inaugural value of this systematic BRCA1/2 variant
analysis in the Chinese population, some limitations were still
noticed. First, the current data were collected from scientific
research papers, which do not necessarily represent the real
distribution in the Chinese population. As a result, this analysis
may have had an unexpected higher ratio of pathogenic variants
(69.9% in BRCA1 exons and 71.1% in BRCA2 exons). Second, the
selection criteria used in these studies were somewhat distinctive,
although most of them used the standard high‐moderate risk criteria.
Third, even though we attempted to minimize the potential bias, the
number of patients in each separate study might still be double‐
counted in our analysis, since the patients’details were not provided
in these studies. Fourth, the cross‐ethnicity variants were mainly
analyzed using the data provided by the BIC database, which
included insufficient data from African and Mongolian populations,
for variant spectrum analysis to reflect the real landscape. Fifth, it
should be noted that only a small proportion of BC and OC was
caused by BRCA1/2 germline variants, while the majority of the
cancers were sporadic and, therefore, without germline variants.
Tests for more comprehensive gene abnormalities are further
necessary for the establishment of therapeutic strategies. Finally,
ambiguous or contradictory interpretation of BRCA1/2 gene defects
still impede the clarification of pathogenicity for certain variants,
which, in turn, postpones the selection of proper therapeutic
methods. Future efforts should focus on the interpretation of
variants with contradictory interpretation as well as variants of
uncertain significance, based on new evidence and database profiling.
CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.
GAO ET AL.
|
11
DATA AVAILABILITY STATEMENT
In addition to submission into LOVD, the data are also available as a
supporting data file called “20191123 BRCA data submitted.xlsx.”
ORCID
Lele Song http://orcid.org/0000-0003-0296-2978
REFERENCES
Alsop, K., Fereday, S., Meldrum, C., deFazio, A., Emmanuel, C., George, J., …
Mitchell, G. (2012). BRCA mutation frequency and patterns of
treatment response in BRCA mutation‐positive women with ovarian
cancer: A report from the Australian Ovarian Cancer Study Group.
Journal of Clinical Oncology,30(21), 2654–2663.
Audeh, M. W., Carmichael, J., Penson, R. T., Friedlander, M., Powell, B.,
Bell‐McGuinn, K. M., …Tutt, A. (2010). Oral poly(ADP‐ribose)
polymerase inhibitor olaparib in patients with BRCA1 or BRCA2
mutations and recurrent ovarian cancer: A proof‐of‐concept trial.
Lancet,376, 245–251.
Balmana, J., Diez, O., & Castiglione, M. (2009). BRCA in breast cancer:
ESMO clinical recommendations. Annals of Oncology,20(Suppl 4), 19–20.
Blackwood, M. A., & Weber, B. L. (1998). BRCA1 and BRCA2: From
molecular genetics to clinical medicine. Journal of Clinical Oncology,16,
1969–1977.
Bolton, K. L., Chenevix‐Trench, G., Goh, C., Sadetzki, S., Ramus, S. J.,
Karlan, B. Y., …Pharoah, P. D., EMBRACE; kConFab Investigators;
Cancer Genome Atlas Research Network. (2012). Association
between BRCA1 and BRCA2 mutations and survival in women with
invasive epithelial ovarian cancer. Journal of the American Medical
Association,307(4), 382–390.
Brody, L. C., & Biesecker, B. B. (1998). Breast cancer susceptibility genes.
BRCA1 and BRCA2. Medicine,77, 208–226.
Cancer Genome Atlas Research Network. (2011). Integrated genomic
analyses of ovarian carcinoma. Nature,474, 609–615.
Coleman, R. L., Oza, A. M., Lorusso, D., Aghajanian, C., Oaknin, A., Dean, A.,
…Ledermann, J. A., ARIEL3 investigators. (2017). Rucaparib
maintenance treatment for recurrent ovarian carcinoma after
response to platinum therapy (ARIEL3): A randomised, double‐blind,
placebo‐controlled, phase 3 trial. Lancet,390, 1949–1961.
Copson, E. R., Maishman, T. C., Tapper, W. J., Cutress, R. I., Greville‐
Heygate, S., Altman, D. G., …Eccles, D. M. (2018). Germline BRCA
mutation and outcome in young‐onset breast cancer (POSH): A
prospective cohort study. Lancet Oncology,19(2), 169–180.
den Dunnen, J. T. (2016). Sequence variant descriptions: HGVS
nomenclature and mutalyzer. Current Protocols in Human Genetics,
90, 7.13.1–7.13.19.
Han, S. A., Kim, S. W., Kang, E., Park, S. K., Ahn, S. H., Lee, M. H., …Jung, S.
H. KOHBRA Research Group and the Korean Breast Cancer Society
(2013). The prevalence of BRCA mutations among familial breast
cancer patients in Korea: Results of the Korean Hereditary Breast
Cancer study. Familial Cancer,12,75–81.
Isakoff, S. J., Mayer, E. L., He, L., Traina, T. A., Carey, L. A., Krag, K. J., …
Ellisen, L. W. (2015). TBCRC009: A Multicenter Phase II Clinical Trial
of platinum monotherapy with biomarker assessment in metastatic
triple‐negative breast cancer. Journal of Clinical Oncology,33(17),
1902–1909.
Kim, Y. C., Zhao, L., Zhang, H., Huang, Y., Cui, J., Xiao, F., …Wang, S. M.
(2016). Prevalence and spectrum of BRCA germline variants in
mainland Chinese familial breast and ovarian cancer patients.
Oncotarget,7, 9600–9612.
Kuchenbaecker, K. B., Hopper, J. L., Barnes, D. R., Phillips, K. A., Mooij, T.
M., Roos‐Blom, M. J., …Olsson, H. (2017). Risks of breast, ovarian, and
contralateral breast cancer for BRCA1 and BRCA2 mutation carriers.
Journal of the American Medical Association,317, 2402–2416.
Kwa, M., Edwards, S., Downey, A., Reich, E., Wallach, R., Curtin, J., &
Muggia, F. (2014). Ovarian cancer in BRCA mutation carriers:
Improved outcome after intraperitoneal (IP) cisplatin. Annals of
Surgical Oncology,21, 1468–1473.
Kwong, A., Chen, J., Shin, V. Y., Ho, J. C., Law, F. B., Au, C. H., …Ford, J. M.
(2015). The importance of analysis of long‐range rearrangement of
BRCA1 and BRCA2 in genetic diagnosis of familial breast cancer.
Cancer genetics,208, 448–454.
Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche, P. C.,
Ioannidis, J. P., …Moher, D. (2009). The PRISMA statement for
reporting systematic reviews and meta‐analyses of studies that
evaluate health care interventions: Explanation and elaboration.
PLoS Medicine,6,1–28.
Mirza, M. R., Monk, B. J., Herrstedt, J., Oza, A. M., Mahner, S., Redondo, A.,
& Matulonis, U. A., ENGOT‐OV16/NOVA Investigators. (2016).
Niraparib maintenance therapy in platinum‐sensitive, recurrent
ovarian cancer. The New England Journal of Medicine,375, 2154–2164.
Narod, S. A. (2010). BRCA mutations in the management of breast cancer:
The state of the art. Nature Reviews Clinical Oncology,7, 702–707.
Oglesbee, D., Cowan, T. M., Pasquali, M., Wood, T. C., Weck, K. E., Long, T.,
& Palomaki, G. E. (2018). CAP/ACMG proficiency testing for
biochemical genetics laboratories: A summary of performance.
Genetics in Medicine,20,83–90.
Panel members of China expert consensus on BCRA variant interpreta-
tion. (2017). China expert consensus on BRCA variant interpretation.
Chinese Journal of Pathology,46, 293–297.
Pennington, K. P., Walsh, T., Harrell, M. I., Lee, M. K., Pennil, C. C., Rendi,
M. H., …Swisher, E. M. (2014). Germline and somatic mutations in
homologous recombination genes predict platinum response and
survival in ovarian, fallopian tube, and peritoneal carcinomas. Clinical
Cancer Research,20, 764–775.
Pujade‐Lauraine, E., Ledermann, J. A., Selle, F., Gebski, V., Penson, R. T.,
Oza, A. M., …Pautier, P., SOLO2/ENGOT‐Ov21 investigators. (2017).
Olaparib tablets as maintenance therapy in patients with platinum‐
sensitive, relapsed ovarian cancer and a BRCA1/2 mutation (SOLO2/
ENGOT‐Ov21): A double‐blind, randomised, placebo‐controlled,
phase 3 trial. The Lancet Oncology,18, 1274–1284.
Rainville, I. R., & Rana, H. Q. (2014). Next‐generation sequencing for
inherited breast cancer risk: Counseling through the complexity.
Current Oncology Reports,16, 371.
Rebbeck, T. R., Friebel, T. M., Friedman, E., Hamann, U., Huo, D., Kwong, A.,
…Nathanson, K. L. (2018). Mutational spectrum in a worldwide study
of 29,700 families with BRCA1 or BRCA2 mutations. Human Mutation,
39(5), 118–620.
Richards, S., Aziz, N., Bale, S., Bick, D., Das, S., Gastier‐Foster, J., …Rehm,
H. L., ACMG Laboratory Quality Assurance Committee. (2015).
Standards and guidelines for the interpretation of sequence
variants: A joint consensus recommendation of the American
College of Medical Genetics and Genomics and the Association for
Molecular Pathology. Genetics in Medicine,17, 405–424.
Robson, M., Dabney, M. K., Rosenthal, G., Ludwig, S., Seltzer, M. H., Gilewski,
T., …Offit, K. (1997). Prevalence of recurring BRCA mutations among
Ashkenazi Jewish women with breast cancer. Genetic Testing,1,47–51.
Robson, M., Goessl, C., & Domchek, S. (2017). Olaparib for metastatic
germline BRCA‐mutated breast cancer. The New England Journal of
Medicine,377, 1792–1793.
Schmidt, A. Y., Hansen, T. V. O., Ahlborn, L. B., Jonson, L., Yde, C. W., &
Nielsen, F. C. (2017). Next‐generation sequencing‐based detection of
germline copy number variations in BRCA1/BRCA2: Validation of a
one‐step diagnostic workflow. The Journal of Molecular Diagnostics,19,
809–816.
12
|
GAO ET AL.
Spurdle, A. B., Healey, S., Devereau, A., Hogervorst, F. B., Monteiro, A. N.,
Nathanson, K. L., …Goldgar, D. E., ENIGMA. (2012). ENIGMA‐‐
evidence‐based network for the interpretation of germline mutant
alleles: An international initiative to evaluate risk and clinical
significance associated with sequence variation in BRCA1 and
BRCA2 genes. Human Mutation,33,2–7.
Sun, J., Meng, H., Yao, L., Lv, M., Bai, J., Zhang, J., …Xie, Y. (2017).
Germline mutations in cancer susceptibility genes in a large series of
unselected breast cancer patients. Clinical Cancer Research,23,
6113–6119.
Swisher, E. M., Lin, K. K., Oza, A. M., Scott, C. L., Giordano, H., Sun, J., …
McNeish, I. A. (2017). Rucaparib in relapsed, platinum‐sensitive high‐
grade ovarian carcinoma (ARIEL2 Part 1): An international,
multicentre, open‐label, phase 2 trial. The Lancet Oncology,18,75–87.
Tobias, D. H., Eng, C., McCurdy, L. D., Kalir, T., Mandelli, J., Dottino, P. R., &
Cohen, C. J. (2000). Founder BRCA 1 and 2 mutations among a
consecutive series of Ashkenazi Jewish ovarian cancer patients.
Gynecologic Oncology,78, 148–151.
Trainer,A.H.,Lewis,C.R.,Tucker,K.,Meiser,B.,Friedlander,M.,&
Ward, R. L. (2010). The role of BRCA mutation testing in
determining breast cancer therapy. Nature Reviews Clinical
Oncology,7, 708–717.
Weitzel, J. N., Lagos, V., Blazer, K. R., Nelson, R., Ricker, C., Herzog, J., …
Neuhausen, S. (2005). Prevalence of BRCA mutations and founder
effect in high‐risk Hispanic families. Cancer Epidemiology, Biomarkers &
Prevention,14, 1666–1671.
Wu, X., Wu, L., Kong, B., Liu, J., Yin, R., Wen, H., …Liu, Y. (2017). The first
nationwide multicenter prevalence study of germline BRCA1 and
BRCA2 mutations in Chinese ovarian cancer patients. International
Journal of Gynecological Cancer,27, 1650–1657.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section.
How to cite this article: Gao X, Nan X, Liu Y, et al.
Comprehensive profiling of BRCA1 and BRCA2 variants in
breast and ovarian cancer in Chinese patients. Human
Mutation. 2019;1–13. https://doi.org/10.1002/humu.23965
GAO ET AL.
|
13
