Advertisement

Probability of detecting germline BRCA1/2 pathogenic variants in histological subtypes of ovarian carcinoma. A meta-analysis

Open AccessPublished:October 23, 2021DOI:https://doi.org/10.1016/j.ygyno.2021.10.072

      Highlights

      • Probability of finding germline BRCA1/2 PVs varies widely among histological subtypes of ovarian carcinoma (OC).
      • Germline BRCA1/2 PVs are most frequently detected in high-grade serous OC patients.
      • Limiting testing to high-grade serous histology will be insufficient to identify all OC patients with germline BRCA1/2 PVs.

      Abstract

      Background

      Histology restricted genetic predisposition testing of ovarian carcinoma patients is a topic of debate as the prevalence of BRCA1/2 pathogenic variants (PVs) in various histological subtypes is ambiguous. Our primary aim was to investigate the proportion of germline BRCA1/2 PVs per histological subtype. Additionally, we evaluated (i) proportion of somatic BRCA1/2 PVs and (ii) proportion of germline PVs in other ovarian carcinoma risk genes.

      Methods

      PubMed, EMBASE and Web of Science were systematically searched and we included all studies reporting germline BRCA1/2 PVs per histological subtype. Pooled proportions were calculated using a random-effects meta-analysis model. Subsets of studies were used for secondary analyses.

      Results

      Twenty-eight studies were identified. The overall estimated proportion of germline BRCA1/2 PVs was 16.8% (95% CI 14.6 to 19.2). Presence differed substantially among patients with varying histological subtypes of OC; proportions being highest in high-grade serous (22.2%, 95% CI 19.6 to 25.0) and lowest in clear cell (3.0%, 95% CI 1.6 to 5.6) and mucinous (2.5%, 95% CI 0.6 to 9.6) carcinomas. Somatic BRCA1/2 PVs were present with total estimated proportion of 6.0% (95% CI 5.0 to 7.3), based on a smaller subset of studies. Germline PVs in BRIP1, RAD51C, RAD51D, PALB2, and ATM were present in approximately 3%, based on a subset of nine studies.

      Conclusion

      Germline BRCA1/2 PVs are most frequently identified in high-grade serous ovarian carcinoma patients, but are also detected in patients having ovarian carcinomas of other histological subtypes. Limiting genetic predisposition testing to high-grade serous ovarian carcinoma patients will likely be insufficient to identify all patients with a germline PV.

      Keywords

      1. Introduction

      Recognition of heredity in ovarian carcinoma (OC) patients is crucial to reduce cancer risks among patients and family members and it may facilitate treatment decisions. About 20–25% of all OCs are caused by an underlying heritable tumor risk syndrome [
      • Norquist B.M.
      • et al.
      Inherited mutations in women with ovarian carcinoma.
      ,
      • Walsh T.
      • et al.
      Mutations in 12 genes for inherited ovarian, fallopian tube, and peritoneal carcinoma identified by massively parallel sequencing.
      ]. This proportion consists mainly of women harboring a germline pathogenic variant (PV) in the BRCA1 or BRCA2 gene [
      • Norquist B.M.
      • et al.
      Inherited mutations in women with ovarian carcinoma.
      ,
      • Walsh T.
      • et al.
      Mutations in 12 genes for inherited ovarian, fallopian tube, and peritoneal carcinoma identified by massively parallel sequencing.
      ]. Germline PVs in RAD51C, RAD51D, BRIP1, PALB2, ATM, and the mismatch repair genes (MLH1, MSH2, MSH6, PMS2) also confer a moderately increased risk for OC [
      • Lu H.-M.
      • et al.
      Association of breast and ovarian cancers with predisposition genes identified by large-scale sequencing.
      ,
      • Pavanello M.
      • et al.
      Rare germline genetic variants and the risks of epithelial ovarian cancer.
      ], but occur less frequently [
      • Norquist B.M.
      • et al.
      Inherited mutations in women with ovarian carcinoma.
      ,
      • Suszynska M.
      • et al.
      Large-scale meta-analysis of mutations identified in panels of breast/ovarian cancer-related genes - providing evidence of cancer predisposition genes.
      ]. Germline testing plays a central role in recognition of heredity in OC patients.
      Tumor DNA testing can be used as an efficient and effective prescreen to stratify germline testing and treatment options [
      • Vos J.R.
      • et al.
      Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
      ,
      • de Jonge M.M.
      • et al.
      Linking uterine serous carcinoma to BRCA1/2-associated cancer syndrome: a meta-analysis and case report.
      ]. Tumor testing detects both germline PVs and somatic PVs (present in tumor DNA but absent in blood). The increasing importance of tumor DNA testing is underlined by developments in treatment options. For example, poly ADP ribose polymerase inhibitor (PARPi) therapy has proven to be beneficial for patients with a tumor BRCA1/2 PV [
      • Miller R.E.
      • et al.
      ESMO recommendations on predictive biomarker testing for homologous recombination deficiency and PARP inhibitor benefit in ovarian cancer.
      ], either germline or somatic. Tumor DNA testing (i.e. a Tumor-First approach) detects individuals eligible for treatments options, and can simultaneously function as a prescreen to tailor genetic counseling and germline testing to patients at higher risk [
      • Vos J.R.
      • et al.
      Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
      ].
      Universal germline or tumor testing of all OC patients has increasingly become the norm [
      • Vos J.R.
      • et al.
      Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
      ,
      • Arts-de Jong M.
      • et al.
      Germline BRCA1/2 mutation testing is indicated in every patient with epithelial ovarian cancer: a systematic review.
      ,
      • Konstantinopoulos P.A.
      • et al.
      Germline and somatic tumor testing in epithelial ovarian Cancer: ASCO guideline.
      ]. However, OC is a heterogeneous disease and histological subtypes display varying molecular genetic landscapes and distinct precancerous lesions. Selection of histological subtypes for germline testing and tumor DNA testing (as prescreen) to reduce costs and optimize recognition of hereditary OC is still a topic of debate. Former studies and reviews have demonstrated that high-grade serous OC is the cancer associated with germline BRCA1/2 PVs [
      • Arts-de Jong M.
      • et al.
      Germline BRCA1/2 mutation testing is indicated in every patient with epithelial ovarian cancer: a systematic review.
      ,
      • Eccles D.M.
      • et al.
      Selecting patients with ovarian Cancer for germline BRCA mutation testing: findings from guidelines and a systematic literature review.
      ]. This is supported by the detection of serous tubal intraepithelial carcinomas (STICs) during prophylactic risk reducing salphingo-oophorectomy in individuals with a BRCA1/2 PV [
      • Piek J.M.
      • et al.
      Dysplastic changes in prophylactically removed fallopian tubes of women predisposed to developing ovarian cancer.
      ,
      • Labidi-Galy S.I.
      • et al.
      High grade serous ovarian carcinomas originate in the fallopian tube.
      ]. This association with distinct histology raises the question whether genetic predisposition testing could be executed more efficiently by restricting testing to certain histological subtypes of OC.
      Histology restricted genetic predisposition testing is highly dependent on the accuracy of histology typing and the proportion of PVs detected per histological subtype. In 2014 and 2020, the World Health Organization (WHO) published new criteria for histological subclassification of OC [
      • Duska L.R.
      • Kohn E.C.
      The new classifications of ovarian, fallopian tube, and primary peritoneal cancer and their clinical implications.
      ]. The accuracy of histology typing was of concern in older classification systems, but the WHO 2014 and 2020 are more robust [
      • Peres L.C.
      • et al.
      Histotype classification of ovarian carcinoma: a comparison of approaches.
      ,
      • Meinhold-Heerlein I.
      • et al.
      The new WHO classification of ovarian, fallopian tube, and primary peritoneal cancer and its clinical implications.
      ]. Therefore, the proportion of PVs per histological subtype needs re-evaluation. Here, we performed a systematic review and meta-analysis of recent literature (>2015) with the primary aim to investigate the proportion of germline BRCA1/2 PVs per histological subtype of OC. Secondarily, we evaluated (i) proportion and histology of somatic BRCA1/2 PVs and (ii) proportion of germline PVs in moderate risk genes for OC (BRIP1, RAD51C, RAD51D, ATM, PALB2, MLH1, MSH2, MSH6, PMS2) and OC histology.

      2. Methods

      This systematic review was conducted and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [
      • Moher D.
      • et al.
      Preferred reporting items for systematic reviews and Meta-analyses: the PRISMA statement.
      ].

      2.1 Search strategy and study selection

      Databases PubMed/Medline, EMBASE and Web of Science were systematically searched for studies published from 1 January 2015 to 5 November 2020. A comprehensive search strategy was constructed using medical subject headings (MeSH), Embase subject headings (Emtree), keywords and synonyms related to three aspects: (i) ovarian carcinoma, (ii) BRCA, and (iii) germline/tumor testing. The complete search strategies are provided in supplement S1. Searches were restricted to English language and timeframe of publication 2015 (after introduction of WHO 2014 histology classification system) till “current”. All references were uploaded in Endnote reference management program (Endnote™ X9). Manual removal of duplicates and selection of articles was performed by two reviewers (VW and MvB) independently, achieving agreement after discussion or by consultation of a third reviewer (NH).

      2.2 Eligibility criteria

      Selection of articles was performed according to predefined inclusion and exclusion criteria. Articles were included if all information required for computing the prevalence of germline BRCA1/2 PVs per histological subtype of OC was provided. Germline BRCA1/2 PVs were defined as class 4 and 5 variants, and OC was defined by the WHO 2014 and 2020 guidelines [
      • Peres L.C.
      • et al.
      Histotype classification of ovarian carcinoma: a comparison of approaches.
      ,
      • Meinhold-Heerlein I.
      • et al.
      The new WHO classification of ovarian, fallopian tube, and primary peritoneal cancer and its clinical implications.
      ]. Articles were excluded when the population did not consist of OC patients, when the number of OC patients was unclear, when no germline testing was performed, when testing was restricted to pre-specified (founder) mutations, or when the information on histology was insufficient to compute proportions per subtype. Solely articles written in English language and investigating human subjects were included. In case of overlapping cohorts, only the study with most patients was included. Review articles, case-reports, opinion pieces and letters to editors were excluded, similar to conference abstracts.

      2.3 Critical appraisal

      The quality of selected studies was rated using an adapted version of the critical appraisal tool for prevalence studies from the Joanne Briggs Institute [
      • Munn Z.
      • et al.
      Methodological guidance for systematic reviews of observational epidemiological studies reporting prevalence and cumulative incidence data.
      ]. The standard appraisal tool consisting of nine categories was adapted to enable scoring specifically for this systematic review. The adapted version is provided in supplement S2. Here, six (out of 13) items were considered to be essential (‘answered with yes’) to be included in the quantitative analysis: 1) sample frame broader than serous ovarian carcinomas, 2) total population size >50, 3) serous histology subdivided in high- and low-grade, 5) histological subgroup ‘non- high-grade serous’ specified, 5) germline and somatic PVs are distinguishable, and 6) variants of uncertain significance (VUS) and pathogenic variants are distinguishable. Scoring was performed by two reviewers (VW and MvB) independently, achieving agreement after discussion. The total critical appraisal score was the number of items answered with ‘yes’, which had no further consequences.

      2.4 Outcomes and data-extraction

      Our primary outcome was defined as the proportion of germline BRCA1/2 PVs per histological subtype of OC. Our secondary outcomes were: (i) proportion of somatic BRCA1/2 PVs and (ii) proportion of germline PVs in other risk genes for OC (BRIP1, RAD51C, RAD51D, PALB2, ATM, MLH1, MSH2, MSH6, PMS2). We defined somatic BRCA1/2 PVs as variants that are present in tumor DNA but absent in normal tissue or blood. We extracted data from eligible studies using a data extraction table consisting of predefined topics: bibliographical data, population data, methodological data, and outcome data. Data on histology of tested population and data on histology of BRCA1/2 positive cases were essential items in data extraction. We recorded whether tumor testing was performed in addition to germline testing and which genes other than BRCA1/2 were tested. Data extraction was split between two reviewers, who cross-checked each other's work. In case any discrepancies in original articles were identified during data extraction, we considered data from tables to be most reliable.

      2.5 Data analysis and statistics

      We performed meta-analyses of the proportion of germline and somatic BRCA1/2 PVs in all OCs. Additionally, we performed meta-analyses of the proportion of germline BRCA1/2 PVs per histological subtype of OC: high-grade serous (HGS), endometrioid, clear cell, low-grade serous (LGS), mucinous, carcinosarcoma, and ‘other’. The group ‘other’ was a merge of the histological types seromucinous, transitional cell, Brenner, undifferentiated, mixed, and other. We did not calculate an average proportion for (ovarian) carcinoma not specified, adenocarcinoma not specified and serous carcinoma not specified. Data analysis was performed at study level.
      Pooled proportions were calculated by a random intercepts logistic regression model (GLMM) using a maximum likelihood estimation (ML) [
      • Stijnen T.
      • Hamza T.H.
      • Özdemir P.
      Random effects meta-analysis of event outcome in the framework of the generalized linear mixed model with applications in sparse data.
      ,
      • Schwarzer G.
      • et al.
      Seriously misleading results using inverse of freeman-Tukey double arcsine transformation in meta-analysis of single proportions.
      ]. Heterogeneity across studies was estimated using the I2 statistic (<25% low level of heterogeneity, 50% moderate level of heterogeneity, >75% high level of heterogeneity) [
      • Higgins J.P.T.
      • et al.
      Measuring inconsistency in meta-analyses.
      ]. Subgroup analysis was undertaken based on ethnicity (country where study was performed) to assess potential differences.
      Data was examined for the presence of outliers, defined as studies in which the individual confidence interval (CI) did not overlap with the meta-analysis CI. Outliers were not excluded from analyses, but these articles were screened for potential reasons for variation. In addition, the data was examined for influential studies, defined as studies for which exclusion leads to changes in result of the meta-analysis [
      • Viechtbauer W.
      • Cheung M.W.-L.
      Outlier and influence diagnostics for meta-analysis.
      ]. Also, removing studies one-by-one was performed and the effect on pooled proportion was evaluated to assess sensitivity of the meta-analysis model.
      All analyses were conducted using statistical software R version 3.6.2 (2019-12-12) using the packages “meta” and “metafor”.

      3. Results

      3.1 Study selection

      Database searches generated a total of 4756 records, of which 2941 remained after removal of duplicates. Exclusion based on screening of title and abstracts (n = 2708) and full text (n = 135) resulted in 98 articles who were subjected to a critical appraisal. Then, another 69 articles were excluded which resulted in a total of 29 articles. During data extraction two articles were merged as they presented results of the same cohort [
      • Harter P.
      • et al.
      Prevalence of deleterious germline variants in risk genes including BRCA1/2 in consecutive ovarian cancer patients (AGO-TR-1).
      ,
      • Hauke J.
      • et al.
      Deleterious somatic variants in 473 consecutive individuals with ovarian cancer: results of the observational AGO-TR1 study (NCT02222883).
      ]. The critical appraisal scores of the included articles are presented in supplement S3. In total, this systematic review and meta-analysis is based on 28 studies (Fig. 1).

      3.2 Characteristics of the included studies

      Table 1 presents an overview of the study characteristics and main outcome data of the included studies. The included studies were conducted either in Asian or European countries, with the exception of one study which was conducted in the United States [
      • Norquist B.M.
      • et al.
      Inherited mutations in women with ovarian carcinoma.
      ]. Details on selected OC patients are also provided; most studies included all OC patients, with mucinous ovarian carcinoma being the predominant exclusion criterium. The number of included OC patients in individual studies ranged from 56 to 1915 (patients with known germline or somatic mutation status and known histology of OC). In total, we include 11,351 OC patients from 28 studies. The individual study results on total number of BRCA1/2 PVs and BRCA1/2 PVs per histological subtype of OC are presented in Table 1. Table 1 also marks that nine studies performed somatic tumor testing in addition to germline testing, this subset was used to analyze the proportion and histology of somatic BRCA1/2 PVs. Another subset of nine studies could be used to evaluate proportion and histology of other risk genes for OC as they tested for these in addition to BRCA1/2.
      Fig. 1
      Fig. 1Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart of study selection. After several rounds of selection, 28 articles remained from the initial database searches.
      Abbreviations: OC: ovarian carcinoma, HGS: high-grade serous, LGS: low-grade serous, PV: pathogenic variant, VUS: variant of uncertain significance.
      Table 1General study characteristics and primary outcome data.
      StudyGeneral study characteristicsBRCA1/2 PVs in total and per histological subtype of OC
      Number of patients with BRCA PVs in total population and per histological subtype: HGS, E, CC, LGS, M, CS, other. Number of patients with BRCA PVs (n) per number of tumors tested (N). ‘other’ is combined group from: seromucinous, transitional cell, undifferentiated, mixed and other. — indicates histological subtype is not present in this cohort. Excluded from this table are: (ovarian) carcinoma not specified, serous not specified and borderline tumors.
      RefAuthor, yearCountryIncluded OC patientsTest
      Test: only germline testing (G), germline and tumor testing (B), or tumor testing and subsequent germline testing when test was positive (S). *marks germline testing not performed for all patients with positive tumor test, these patients were excluded from our analysis.
      MGPTotal (n/N)HGS

      (n/N)
      E

      (n/N)
      CC

      (n/N)
      LGS

      (n/N)
      M

      (n/N)
      CS (n/N)Other (n/N)
      [
      • Ataseven B.
      • et al.
      Prevalence of BRCA1 and BRCA2 mutations in patients with primary ovarian cancer - does the German checklist for detecting the risk of hereditary breast and ovarian cancer adequately depict the need for consultation?.
      ]
      Ataseven, 2020GermanyAllGYes127/545125/4351/290/230/330/161/10/8
      [
      • Bu H.
      • et al.
      BRCA mutation frequency and clinical features of ovarian cancer patients: a report from a Chinese study group.
      ]
      Bu, 2019ChinaAllGNo117/50697/3981/177/236/332/134/22
      [
      • Choi M.C.
      • et al.
      Germline mutations of <em>BRCA1</em> and <em>BRCA2</em> in Korean ovarian cancer patients: finding founder mutations.
      ]
      Choi, 2015KoreaAllGNo18/7018/440/60/90/90/2
      [
      • Enomoto T.
      • et al.
      The first Japanese nationwide multicenter study of BRCA mutation testing in ovarian cancer: CHARacterizing the cross-sectionaL approach to ovarian cancer geneTic TEsting of BRCA (CHARLOTTE).
      ]
      Enomoto, 2019JapanAllGNo93/63478/2748/1204/1871/50/192/29
      [
      • Flaum N.
      • et al.
      Mainstreaming germline BRCA1/2 testing in non-mucinous epithelial ovarian cancer in the north west of England.
      ]
      Flaum, 2020UKNon-mucinousGNo89/48186/4272/210/140/81/11
      [
      • George A.
      • et al.
      Implementing rapid, robust, cost-effective, patient-centred, routine genetic testing in ovarian cancer patients.
      ]
      George, 2016UKNon-mucinous,

      partial age < 65
      GNo33/20732/1731/220/20/60/4
      [
      • Harter P.
      • et al.
      Prevalence of deleterious germline variants in risk genes including BRCA1/2 in consecutive ovarian cancer patients (AGO-TR-1).
      ,
      • Hauke J.
      • et al.
      Deleterious somatic variants in 473 consecutive individuals with ovarian cancer: results of the observational AGO-TR1 study (NCT02222883).
      ]
      Hauke, 2019 &

      Harter, 2017
      GermanyAllBYesg: 95/473

      s: 29/473
      g: 86/373

      s: 23/373
      g: 4/29

      s: 5/29
      g: 0/6

      s: 0/6
      g: 1/16

      s: 0/16
      g: 0/6

      s: 0/6
      g: 2/18

      s: 0/18
      [
      • Hirasawa A.
      • et al.
      Prevalence of pathogenic germline variants detected by multigene sequencing in unselected Japanese patients with ovarian cancer.
      ]
      Hirasawa, 2017JapanAllGYes27/23022/742/582/710/30/181/6
      [
      • Kim S.I.
      • et al.
      Germline and somatic BRCA1/2 gene mutational status and clinical outcomes in epithelial peritoneal, ovarian, and fallopian tube cancer: over a decade of experience in a single institution in Korea.
      ]
      Kim, 2020KoreaAllBNog: 13/56

      s: 3/56
      g: 13/51

      s: 3/51
      g: 0/1

      s: 0/1
      g: 0/3

      s: 0/3
      g: 0/1

      s: 0/1
      [
      • Kowalik A.
      • et al.
      Somatic mutations in BRCA1&2 in 201 unselected ovarian carcinoma samples – single institution study.
      ]
      Kowalik, 2019PolandAllS*Nog: 35/193

      s: 6/193
      g: 28/116

      s: 5/116
      g: 1/21

      s: 1/21
      g: 1/9

      s: 0/9
      g: 5/32

      s: 0/32
      g: 0/6

      s: 0/6
      g: 0/9

      s: 0/9
      [
      • Lertkhachonsuk A.-A.
      • et al.
      Prevalence of tissue BRCA gene mutation in ovarian, fallopian tube, and primary peritoneal cancers: a multi-institutional study.
      ]
      Lertkhachonsuk, 2020ThailandHGS, HGE, clear cellS*Nog: 14/138

      s: 9/138
      g: 13/76

      s: 7/76
      g: 0/4

      s: 0/4
      g: 1/55

      s: 2/55
      g: 0/3

      s: 0/3
      [
      • Lhotova K.
      • et al.
      Multigene panel germline testing of 1333 Czech patients with ovarian cancer.
      ]
      Lhotova, 2020Czech RepublicAll, incl. BorderlineGYes288/1120152/47818/901/1512/855/434/90
      [
      • Li W.
      • et al.
      Germline and somatic mutations of multi-gene panel in Chinese patients with epithelial ovarian cancer: a prospective cohort study.
      ]
      Li, 2019ChinaAllBYesg: 14/62

      s: 4/62
      g: 13/48

      s: 4/48
      g: 0/3

      s: 0/3
      g: 0/5

      s: 0/5
      g: 0/1

      s: 0/1
      g: 0/1

      s: 0/1
      g: 0/4

      s: 0/4
      [
      • Manchana T.
      • Phowthongkum P.
      • Teerapakpinyo C.
      Germline mutations in Thai patients with nonmucinous epithelial ovarian cancer.
      ]
      Manchana, 2019ThailandNon-mucinousGYes20/11219/490/281/240/60/4
      [
      • Morgan R.D.
      • et al.
      Prevalence of germline pathogenic BRCA1/2 variants in sequential epithelial ovarian cancer cases.
      ]
      Morgan, 2019UKNon JewishGNo103/55790/4755/292/180/100/41/60/2
      [
      • Norquist B.M.
      • et al.
      Inherited mutations in women with ovarian carcinoma.
      ]
      Norquist, 2016USAAll, partial selection on FIGO stageGYes280/1915240/14987/774/584/700/161/221/9
      [
      • Peixoto A.
      • et al.
      Tumor testing for somatic and germline BRCA1/BRCA2 variants in ovarian cancer patients in the context of strong founder effects.
      ]
      Peixoto, 2020PortugalNon-mucinousSNog: 18/135

      s: 8/135
      g: 17/95

      s: 5/95
      g: 0/9

      s: 2/9
      g: 0/10

      s: 0/10
      g: 0/14

      s: 0/14
      g: 1/4

      s: 1/4
      g: 0/3

      s: 0/3
      [
      • Plaskocinska I.
      • et al.
      New paradigms for BRCA1/BRCA2 testing in women with ovarian cancer: results of the genetic testing in epithelial ovarian Cancer (GTEOC) study.
      ]
      Plaskocinska, 2016UKHGS, HGEGNo18/32317/1920/200/5
      [
      • Rahman B.
      • et al.
      Mainstreamed genetic testing for women with ovarian cancer: first-year experience.
      ]
      Rahman, 2019UKNon-mucinous, high gradeGNo18/12217/1000/90/51/50/3
      [
      • Rivera D.
      • et al.
      Implementing NGS-based <em>BRCA</em> tumour tissue testing in FFPE ovarian carcinoma specimens: hints from a real-life experience within the framework of expert recommendations.
      ]
      Rivera, 2020ItalyNon-mucinousBNog: 12/66

      s: 7/66
      g: 12/59

      s: 6/59
      g: 0/1

      s: 1/1
      g: 0/3

      s: 0/3
      g: 0/1

      s: 0/1
      g: 0/2

      s: 0/2
      [
      • Rumford M.
      • et al.
      Oncologist-led BRCA ‘mainstreaming’ in the ovarian cancer clinic: a study of 255 patients and its impact on their management.
      ]
      Rumford, 2020UKNon-mucinousGNo34/25534/1970/250/140/80/11
      [
      • Rust K.
      • et al.
      Routine germline BRCA1 and BRCA2 testing in patients with ovarian carcinoma: analysis of the Scottish real-life experience.
      ]
      Rust, 2018ScotlandNon-mucinous, partial selection on familyGYes

      RAD51
      114/599102/5190/90/140/50/70/1
      [
      • Sakamoto I.
      • et al.
      BRCA1 and BRCA2 mutations in Japanese patients with ovarian, fallopian tube, and primary peritoneal cancer.
      ]
      Sakamoto, 2016JapanAllGNo12/9512/570/60/100/170/5
      [
      • Seo J.H.
      • et al.
      Prevalence and oncologic outcomes of BRCA1/2 mutation and variant of unknown significance in epithelial ovarian carcinoma patients in Korea.
      ]
      Seo, 2019KoreaAllGNo88/31084/2540/60/152/201/101/5
      [
      • Shi T.
      • et al.
      BRCA1 and BRCA2 mutations in ovarian cancer patients from China: ethnic-related mutations in BRCA1 associated with an increased risk of ovarian cancer.
      ]
      Shi, 2017ChinaAllGNo153/916114/6139/491/510/60/381/11
      [
      • Sugino K.
      • et al.
      Germline and somatic mutations of homologous recombination-associated genes in Japanese ovarian cancer patients.
      ]
      Sugino, 2019JapanNon- neoadjuvant chemotherapyBYesg: 13/207

      s: 13/207
      g: 10/50

      s: 6/50
      g: 1/39

      s: 2/39
      g: 1/99

      s: 5/99
      g: 0/6

      s: 0/6
      g: 1/13

      s: 0/13
      [
      • Vos J.R.
      • et al.
      Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
      ]
      Vos, 2020The NetherlandsAllS*Nog: 25/298

      s: 19/298
      g: 21/188

      s: 12/188
      g: 0/15

      s: 3/15
      g: 1/19

      s: 0/19
      g: 0/17

      s: 1/17
      g: 0/17

      s: 1/17
      g:1/10

      s:0/10
      g: 2/12

      s: 1/12
      [
      • Wu X.
      • et al.
      The first Nationwide multicenter prevalence study of germline BRCA1 and BRCA2 mutations in Chinese ovarian Cancer patients.
      ]
      Wu, 2017ChinaAllGNo235/823186/6017/303/373/182/102/31/4
      Abbreviations: UK: United Kingdom, USA: United States of America, OC: ovarian carcinoma, MGP: multi-gene panel, MA: meta-analysis, PV: pathogenic variant, g: germline, s: somatic, HGS: high-grade serous, (HG)E: (high-grade) endometrioid, CC: clear cell, LGS: low-grade serous, M: mucinous, CS: carcinosarcoma.
      a Test: only germline testing (G), germline and tumor testing (B), or tumor testing and subsequent germline testing when test was positive (S). *marks germline testing not performed for all patients with positive tumor test, these patients were excluded from our analysis.
      b Number of patients with BRCA PVs in total population and per histological subtype: HGS, E, CC, LGS, M, CS, other. Number of patients with BRCA PVs (n) per number of tumors tested (N). ‘other’ is combined group from: seromucinous, transitional cell, undifferentiated, mixed and other. — indicates histological subtype is not present in this cohort. Excluded from this table are: (ovarian) carcinoma not specified, serous not specified and borderline tumors.

      3.3 Proportion of germline BRCA1/2 PVs per histological subtype

      Meta-analysis of 28 studies resulted in an estimated proportion of 16.8% (95% CI 14.6 to 19.2) for germline BRCA1/2 PVs in a population of OC patients as is illustrated in Fig. 2A. Considerable (high level) heterogeneity was present (I2 = 88%), representing differences in results among studies. Subgroup analysis for studies conducted in Asian versus European/American countries (predominantly Asian versus predominantly Caucasian women) revealed no differences in proportions and heterogeneity between these two groups (supplement S4). More specifically, the estimated proportion of germline BRCA1 PVs was 10.7% (95% CI 8.8 to 12.9, I2 = 88%) and the proportion of BRCA2 PVs was 5.5% (95% CI 4.7 to 6.3, I2 = 51%).
      Fig. 3A presents a pie chart of the OC histological subtypes of all patients and Fig. 3B presents a pie chart of the OC histological subtypes of women with BRCA1/2 PVs. It appears that OC patients with a germline BRCA1/2 PV are relatively more likely to develop HGSOC compared to the general OC population in this meta-analysis: in women with BRCA1/2 PVs this percentage is 91% (1738 / 1907), whereas around 75% (7914 / 10,487) of all OCs are of HGS histology.
      Fig. 2
      Fig. 2Forest plots presenting meta-analyses of proportions of germline (A) and somatic (B) BRCA1/2 PVs in all histological subtypes of ovarian carcinoma combined. Somatic BRCA1/2 PVs are defined as those that are present in tumor DNA but absent in normal tissue or blood.
      Fig. 3
      Fig. 3Pie-charts representing the presence of histological subtypes in all ovarian carcinoma patients (A), patients with germline BRCA1/2 PV (B), patients with somatic BRCA1/2 PV (C), patients with germline PV in BRIP1, RAD51C, RAD51D, PALB2 and ATM (D) and patients with germline PV in MLH1, MSH2, MSH6, PMS2 (E). This excludes the presence of the following histological subtypes: ovarian carcinoma not specified, adenocarcinoma not specified, serous not specified, borderline & unknown (excluded: A n = 864, B n = 198, C n = 2, D n = 41, E n = 6).
      Proportions of germline BRCA1/2 PVs per histological subtype of OC are presented in Table 2 and Fig. 4. Presence of germline BRCA1/2 PVs varied substantially among patients having various histological subtypes of OC. The proportion of germline BRCA1/2 PVs was highest in patients with HGSOC (22.2%, 95% CI 19.6 to 25.0), and also the proportion in ovarian carcinosarcoma patients was found to be relatively high (11.9%, 95% CI 5.8 to 22.6). The probability of detecting a germline BRCA1/2 PV was lower in patients having endometrioid OC (5.8%, 95% CI 3.3 to 9.9), LGSOC (5.2%, 95% CI 2.3 to 11.3), clear cell OC (3.0%, 95% CI 1.6 to 5.6) or mucinous OC (2.5%, 95% CI 0.6 to 9.6). Thus, for all OC patients, irrespective of the histological subtype, there is a probability that the patient is carrying a germline BRCA1/2 PV, but the number of patients needed to test to identify one BRCA1/2 PV vary substantially as is presented in Table 2.
      Table 2Meta-analyses of proportion germline BRCA1/2 PVs per histological subtype of OC.
      HistologyNumber of studiesPositiveTotalPooled proportion (%)95% CI (%)Prediction Interval (%)
      Prediction interval reflects the range in which proportions are expected to be found in future research.
      Hetero-geneity (I2)Numbers needed to test to find 1 PV (95% CI)
      High-grade serous281738791422.219.6 to 25.011.6 to 38.288%5 (4 to 6)
      Carcinosarcoma1097711.95.8 to 22.63.6 to 32.30%9 (5 to 18)
      Endometrioid27677645.83.3 to 9.91.0 to 26.80%18 (11 to 31)
      Low-grade serous23344225.22.3 to 11.30.8 to 27.00%20 (9 to 44)
      Clear cell27297943.01.6 to 5.60.0 to 48.417%34 (18 to 63)
      Mucinous17112442.50.6 to 9.60.1 to 31.40%40 (11 to 167)
      Other25192727.04.5 to 10.74.4 to 10.90%15 (10 to 23)
      a Prediction interval reflects the range in which proportions are expected to be found in future research.
      Fig. 4
      Fig. 4Meta-analysis of pooled proportion of germline BRCA1/2 PVs per histological subtype of ovarian carcinoma. The pooled proportion including 95%CI is presented, and individual study results.
      Heterogeneity was low in all non-HGS histological subtypes, indicating here that the CIs of individual study estimates overlap, potentially due to rarity of these subtypes. We assessed the sensitivity of the meta-analysis by removing studies one-by-one and analyzing their influence on the pooled proportion. The pooled proportion of germline BRCA1/2 PVs in patients with non-HGSOC were more sensible to the effect of removing single studies compared to HGS. Therefore, the pooled proportion of germline BRCA1/2 PVs in patients with non-HGSOC are more uncertain, especially for carcinosarcoma, LGS, and mucinous OC. This uncertainty in the pooled proportion is also visible from the wider CIs of the pooled proportions. Despite this uncertainty in the estimate proportion, germline BRCA1/2 PVs were detected in all histological subtypes of OC. Also, in seven studies that incorporated an extra round of pathological revision in their study design, germline BRCA1/2 PVs were identified in all histological subtypes [
      • Norquist B.M.
      • et al.
      Inherited mutations in women with ovarian carcinoma.
      ,
      • Enomoto T.
      • et al.
      The first Japanese nationwide multicenter study of BRCA mutation testing in ovarian cancer: CHARacterizing the cross-sectionaL approach to ovarian cancer geneTic TEsting of BRCA (CHARLOTTE).
      ,
      • Hirasawa A.
      • et al.
      Prevalence of pathogenic germline variants detected by multigene sequencing in unselected Japanese patients with ovarian cancer.
      ,
      • Lertkhachonsuk A.-A.
      • et al.
      Prevalence of tissue BRCA gene mutation in ovarian, fallopian tube, and primary peritoneal cancers: a multi-institutional study.
      ,
      • Rivera D.
      • et al.
      Implementing NGS-based <em>BRCA</em> tumour tissue testing in FFPE ovarian carcinoma specimens: hints from a real-life experience within the framework of expert recommendations.
      ,
      • Shi T.
      • et al.
      BRCA1 and BRCA2 mutations in ovarian cancer patients from China: ethnic-related mutations in BRCA1 associated with an increased risk of ovarian cancer.
      ,
      • Sugino K.
      • et al.
      Germline and somatic mutations of homologous recombination-associated genes in Japanese ovarian cancer patients.
      ].

      3.4 Proportion and histology of somatic BRCA1/2 PVs

      Meta-analysis of a subset of nine studies that performed tumor testing in addition to germline testing indicated that the estimated proportion of somatic (non-germline) BRCA1/2 PVs in a population of OC patients was 6.0% (95% CI 5.0 to 7.3), as illustrated in Fig. 2B. Heterogeneity does not seem to play a role here (I2 = 0%), indicating no major differences among studies. Proportions of somatic BRCA1 and BRCA2 PVs were more or less similar; 3.5% (95% CI 2.7 to 4.5, I2 = 21%) and 2.7% (95% CI 2.0 to 3.6, I2 = 0%), respectively. Fig. 3C presents a pie-chart of the histological subtypes of tumors with a somatic BRCA1/2 PV. Tumors with a somatic BRCA1/2 PV are predominantly HGS and endometrioid carcinomas, and rarely clear cell, mucinous, carcinosarcoma and other carcinomas, more or less similar to the general OC population in our study.

      3.5 Proportion and histology of germline variants in other ovarian carcinoma risk genes

      Our systematic review included nine studies that investigated the prevalence of germline PVs in other OC risk genes (BRIP1, RAD51C, RAD51D, PALB2, ATM, MLH1, MSH2, MSH6, PMS2). Individual study results are presented in supplement S5. Rough estimates of the prevalence of germline PVs in these genes are: BRIP1 0.9% (42/4658), RAD51C 0.8% (44/5257), RAD51D 0.7% (34/5195), PALB2 0.6% (27/4658), ATM 0.3% (14/4658), MSH6 0.3% (14/4658), PMS2 0.2% (7/3538), MLH1 0.2% (7/4658), MSH2 0.2% (7/4658) (as presented in supplement S6). In total, the combined probability of detecting a germline PV in BRIP1, RAD51C, RAD51D, PALB2, or ATM is around 3.3% in OC patients. The probability of detecting a germline PV in a mismatch repair gene (MLH1, MSH2, MSH6 or PMS2) was <1% in total.
      Fig. 3D and E present pie-charts that illustrate the OC histological subtypes of patients with a germline PV in the genes BRIP1, RAD51C, RAD51D, PALB2, ATM and the mismatch repair (MMR) genes, respectively. HGS was the predominant histological subtype for both groups of genes. Furthermore, germline PVs in the genes BRIP1, RAD51C, RAD51D, PALB2 and ATM were detected in all histological subtypes, whereas germline PVs in MMR genes were predominantly detected in HGS and endometrioid OCs.

      4. Discussion

      In this systematic review and meta-analysis, we have shown that germline BRCA1/2 PVs were detected in 16.8% (95% CI 14.8 to 19.2) of the patients with OC. The probability of detecting a germline BRCA1/2 PV varied widely among the various histological subtypes of OC; ranging from 22.2% (95% CI 19.6 to 25.0) in patients with HGSOC to 3.0% (95% CI 1.6 to 5.6) in patients with clear cell OC and 2.5% (95% CI 0.6 to 9.6) in patients with mucinous OC. Unlike the generally accepted assumption that BRCA1/2 PVs are exclusively related to HGS histology [
      • Santandrea G.
      • et al.
      EP973 ‘tumour first’: an institutional experience of reflex tumour BRCA testing in ovarian epithelial carcinomas.
      ,
      • Schrader K.A.
      • et al.
      Germline BRCA1 and BRCA2 mutations in ovarian cancer: utility of a histology-based referral strategy.
      ], our meta-analysis indicated that BRCA1/2 PVs are also found in all other histological subtypes of OC (endometrioid, clear cell, LGS, mucinous, carcinosarcoma and ‘other’). Therefore, limiting genetic predisposition testing to HGSOC patients will likely be insufficient to identify all patients with an underlying germline pathogenic variant.
      The overall estimated probability of finding a germline BRCA1/2 PV in OC patients (16.8%, 95% CI 14.8 to 19.2) seemed somewhat higher than reported in a previous systematic review (12.7%, 95% CI 9.5 to 15.9) [
      • Vos J.R.
      • et al.
      Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
      ]. Importantly, proportions of germline BRCA1/2 PVs differed substantially among included studies. The differences could not be explained by ethnicity, since subgroup analysis revealed no difference between Asian (predominantly Asian ethnicity) and European/American countries (predominantly Caucasian ethnicity). This suggests differences are caused by more specific characteristics of the included populations. Outliers were screened and we identified potential explanations: variation in the presence of founder mutations [
      • Plaskocinska I.
      • et al.
      New paradigms for BRCA1/BRCA2 testing in women with ovarian cancer: results of the genetic testing in epithelial ovarian Cancer (GTEOC) study.
      ] (e.g. presence of Ashkenazi Jews in population), variation in histological subtypes included and classification criteria used [
      • Seo J.H.
      • et al.
      Prevalence and oncologic outcomes of BRCA1/2 mutation and variant of unknown significance in epithelial ovarian carcinoma patients in Korea.
      ,
      • Sugino K.
      • et al.
      Germline and somatic mutations of homologous recombination-associated genes in Japanese ovarian cancer patients.
      ], and variation in the testing and prevention program (i.e. risk reducing surgeries) of various countries [
      • Vos J.R.
      • et al.
      Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
      ,
      • Bu H.
      • et al.
      BRCA mutation frequency and clinical features of ovarian cancer patients: a report from a Chinese study group.
      ,
      • Wu X.
      • et al.
      The first Nationwide multicenter prevalence study of germline BRCA1 and BRCA2 mutations in Chinese ovarian Cancer patients.
      ]. Unfortunately, as there were no non-European/American/Asian studies identified during our search, we are uncertain if our results apply to other populations. Furthermore, germline BRCA1 PVs were more frequently observed than germline BRCA2 PVs, probably related to the overall lifetime risk of developing OC that is more than twice as high for women with a BRCA1 PV than for those with a BRCA2 PV [
      • Kuchenbaecker K.B.
      • et al.
      Risks of breast, ovarian, and contralateral breast Cancer for BRCA1 and BRCA2 mutation carriers.
      ,
      • Kotsopoulos J.
      • et al.
      Age-specific ovarian cancer risks among women with a BRCA1 or BRCA2 mutation.
      ]. Somatic (non-germline) BRCA1 and BRCA2 PVs were present in comparable proportions, detected with a total estimate proportion of approximately 6% in all histological subtypes.
      Our result that BRCA1/2 PVs were detected in patients having endometrioid OC, clear cell OC, LGSOC, or mucinous OCs does not fit within the classical hypothesis of the origin of these carcinomas. Endometrioid and clear cell OCs are thought to develop from endometriosis [
      • Fadare O.
      • Parkash V.
      Pathology of endometrioid and clear cell carcinoma of the ovary.
      ], whereas LGSOCs are thought to develop from cystadenomas or tubal lesions [
      • Goulding E.A.
      • et al.
      Low-grade serous ovarian carcinoma: a comprehensive literature review.
      ], and mucinous OCs are thought to develop from Brenner tumors or teratomas [
      • Simons M.
      • et al.
      Two types of primary mucinous ovarian tumors can be distinguished based on their origin.
      ]. The development of non-HGSOC in women with BRCA1/2 PV suggests another carcinogenic pathway, but precursor lesions of these histological subtypes are not known to be more frequently observed during prophylactic RRSO in women with BRCA1/2 PV. In addition, the identification of germline BRCA1/2 PVs in these patients does not necessarily indicate a causal relationship with tumor development. However because the incidence of BRCA1/2 PV is higher than in the general population, the results of our meta-analysis challenge the hypothesis that germline BRCA1/2 are exclusively related to HGSOC.
      We should however take into account the potential risk of misclassification of histological subtypes and its influence on our results. Seven of our included studies had performed pathological revision, and in these specific studies BRCA1/2 PVs were identified in non-HGS OC as well. However, it is impossible to reliably evaluate quality of pathological assessment in a literature review, as this measure is partly subjective depending on experience of pathologist and histological criteria used which are both seldomly reported. We included studies published after 2015 as the WHO2014 OC histological classification system is demonstrated to be more reproducible compared to previous systems [
      • Peres L.C.
      • et al.
      Histotype classification of ovarian carcinoma: a comparison of approaches.
      ], but patient inclusion and pathological assessment were sometimes performed before 2014. In addition, studies rarely indicated which classification system was used and whether immunohistochemistry was used to support histologic classification. Therefore, when interpreting our results, one should consider that these resemble evidence from most recent literature, but they do not represent the most recent diagnostic criteria. Prospectively investigating the effect of assessing histology by experienced gynecological pathologists on the probability of detecting BRCA1/2 PVs in non-HGSOC would be essential in future research.
      The presence of germline BRCA1/2 PVs in patients having non-HGSOC clearly requires additional investigation to elucidate true presence (and potential carcinogenic pathway) or misclassification. The rarity of some non-HGS histological subtypes will likely complicate this investigation. Our meta-analysis was limited by the rarity of some non-HGSOC as well. We demonstrated that BRCA1/2 PVs were detected the least frequent in mucinous OC. However, conclusions on this data should be interpreted with caution as several studies (8 out of 28) specifically excluded mucinous OC. Because of scarce data of some histological subtypes, we used a GLMM meta-analysis model to estimate all pooled proportions. This model has been recommended as good alternative for conventional two-step methods [
      • Schwarzer G.
      • et al.
      Seriously misleading results using inverse of freeman-Tukey double arcsine transformation in meta-analysis of single proportions.
      ,
      • Lin L.
      • Chu H.
      Meta-analysis of proportions using generalized linear mixed models.
      ]. The sensitivity of our model was assessed by removing studies one-by-one, which demonstrated that the probability of detecting a germline BRCA1/2 PV was more uncertain in patients having non-HGSOC (specifically carcinosarcoma OC, LGSOC and mucinous OC) compared to patients having HGSOC. This is also visible from the wide confidence intervals. Future research on the presence of germline BRCA1/2 PVs in patients with non-HGSOC should consider their power and sample size to detect variants in these histological subtypes.
      Additionally, as wider panels are expected to become increasingly common because of decreasing costs, assessment of the added clinical value of expanding the BRCA1/2 test panel to include moderate risk genes for OC is important. In comparison to BRCA1 and BRCA2, the detected proportions of germline PVs were considerably lower in BRIP1 (0.9%), RAD51C (0.8%), RAD51D (0.7%), PALB2 (0.6%), ATM (0.3%), MLH1 (0.2%), MSH2 (0.2%), MSH6 (0.3%), and PMS2 (0.2%). Noticeably, our systematic search did not focus on identification of these articles specifically. Previous studies have more elaborately investigated presence and cancer risks of germline PVs in these genes [
      • Pavanello M.
      • et al.
      Rare germline genetic variants and the risks of epithelial ovarian cancer.
      ,
      • Suszynska M.
      • et al.
      Large-scale meta-analysis of mutations identified in panels of breast/ovarian cancer-related genes — providing evidence of cancer predisposition genes.
      ]. Besides testing for BRCA1/2, we encourage testing all patients with OC for germline PVs in RAD51C, RAD51D, BRIP1 and PALB2, and where possible also ATM. However, testing all OC patients for MLH1, MSH2, MSH6 and PMS2 is debatable given the rare occurrences and the fact that testing for MLH1, MSH2, MSH6 and PMS2 can be restricted to those with additional reasons to do so, for example a family history with Lynch syndrome associated cancers [
      • Tiwari A.K.
      • Roy H.K.
      • Lynch H.T.
      Lynch syndrome in the 21st century: clinical perspectives.
      ,
      • Koornstra J.J.
      • et al.
      Management of extracolonic tumours in patients with Lynch syndrome.
      ], or endometrioid OC (which has been associated to mismatch repair deficiency and Lynch syndrome [
      • Krämer P.
      • et al.
      Endometrial Cancer molecular risk stratification is equally prognostic for Endometrioid ovarian carcinoma.
      ,
      • Woolderink J.M.
      • et al.
      Characteristics of Lynch syndrome associated ovarian cancer.
      ]).
      In conclusion, germline BRCA1/2 PVs are being detected in all histological subtypes of OC, and most frequently in HGSOC. Limiting genetic predisposition testing to HGSOC will likely be inadequate to identify all patients with an underlying germline pathogenic variant. Future research (e.g. focusing on cost-effectiveness) might shed new light on the issue. However, based on current literature, we strongly encourage to test all OC patients for germline BRCA1/2 PVs, irrespective of their histological subtype. These considerations will contribute to optimize recognition of heredity in ovarian carcinoma patients.

      Funding

      This work was supported by a grant from the Dutch Cancer Society ( KUN2019–12732 ).

      Author contributions

      All authors critically revised the manuscript and provided final approval of the version to be published. VW and MvB performed data collection and wrote the original draft of the manuscript. VW, MvB and JV were involved in analyzing the data. VW, MvB, ML, TB and NH contributed to the study design and data interpretation. NH supervised the study.

      Declaration of Competing Interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Acknowledgements

      We thank Dorien Hermkens, Yvonne Smolders and Jozé Braspenning for useful discussions and the Radboudumc Medical library for assistance in establishing comprehensive search strategies.

      Appendix A. Supplementary data

      The following are the supplementary data related to this article.
      • Supplementary Table S2

        Critical appraisal score list. An adjusted version of the Joanna Briggs critical appraisal list for prevalence studies. Six items (2, 4, 5a, 5b, 5c, 5d) were essential to be answered with ‘yes’ to be included in our meta-analysis. The other items were scored without any consequences.

      • Supplementary Figure S4

        Sub-group analysis based on ethnicity (country where study was conducted). Proportion of germline BRCA1/2 PVs in study countries with predominant Caucasian ethnicity and in study countries with predominant Asian ethnicity.

      • Supplementary Table S5

        Number of pathogenic variants identified per study. Total number of ovarian carcinoma patients included in meta-analysis (N meta) and the number of pathogenic variants in various genes identified per study.

      • Supplementary Table S6

        Proportion of germline PV in other ovarian carcinoma risk genes: genes involved in homologous recombination (HR) and mismatch repair (MMR), and histology of positive cases.

      References

        • Norquist B.M.
        • et al.
        Inherited mutations in women with ovarian carcinoma.
        JAMA Oncol. 2016; 2: 482-490https://doi.org/10.1001/jamaoncol.2015.5495
        • Walsh T.
        • et al.
        Mutations in 12 genes for inherited ovarian, fallopian tube, and peritoneal carcinoma identified by massively parallel sequencing.
        Proc. Natl. Acad. Sci. U. S. A. 2011; 108: 18032-18037https://doi.org/10.1073/pnas.1115052108
        • Lu H.-M.
        • et al.
        Association of breast and ovarian cancers with predisposition genes identified by large-scale sequencing.
        JAMA Oncol. 2019; 5: 51-57https://doi.org/10.1001/jamaoncol.2018.2956
        • Pavanello M.
        • et al.
        Rare germline genetic variants and the risks of epithelial ovarian cancer.
        Cancers. 2020; 12: 3046
        • Suszynska M.
        • et al.
        Large-scale meta-analysis of mutations identified in panels of breast/ovarian cancer-related genes - providing evidence of cancer predisposition genes.
        Gynecol. Oncol. 2019; 153: 452-462https://doi.org/10.1016/j.ygyno.2019.01.027
        • Vos J.R.
        • et al.
        Universal tumor DNA BRCA1/2 testing of ovarian Cancer: prescreening PARPi treatment and genetic predisposition.
        J. Natl. Cancer Inst. 2020; 112: 161-169https://doi.org/10.1093/jnci/djz080
        • de Jonge M.M.
        • et al.
        Linking uterine serous carcinoma to BRCA1/2-associated cancer syndrome: a meta-analysis and case report.
        Eur. J. Cancer. 2017; 72: 215-225https://doi.org/10.1016/j.ejca.2016.11.028
        • Miller R.E.
        • et al.
        ESMO recommendations on predictive biomarker testing for homologous recombination deficiency and PARP inhibitor benefit in ovarian cancer.
        Ann. Oncol. 2020; 31: 1606-1622https://doi.org/10.1016/j.annonc.2020.08.2102
        • Arts-de Jong M.
        • et al.
        Germline BRCA1/2 mutation testing is indicated in every patient with epithelial ovarian cancer: a systematic review.
        Eur. J. Cancer. 2016; 61: 137-145https://doi.org/10.1016/j.ejca.2016.03.009
        • Konstantinopoulos P.A.
        • et al.
        Germline and somatic tumor testing in epithelial ovarian Cancer: ASCO guideline.
        J. Clin. Oncol. 2020; 38: 1222-1245https://doi.org/10.1200/jco.19.02960
        • Eccles D.M.
        • et al.
        Selecting patients with ovarian Cancer for germline BRCA mutation testing: findings from guidelines and a systematic literature review.
        Adv. Ther. 2016; 33: 129-150https://doi.org/10.1007/s12325-016-0281-1
        • Piek J.M.
        • et al.
        Dysplastic changes in prophylactically removed fallopian tubes of women predisposed to developing ovarian cancer.
        J. Pathol. 2001; 195: 451-456https://doi.org/10.1002/path.1000
        • Labidi-Galy S.I.
        • et al.
        High grade serous ovarian carcinomas originate in the fallopian tube.
        Nat. Commun. 2017; 8: 1093https://doi.org/10.1038/s41467-017-00962-1
        • Duska L.R.
        • Kohn E.C.
        The new classifications of ovarian, fallopian tube, and primary peritoneal cancer and their clinical implications.
        Ann. Oncol: Off. J. Eur. Soc. Med. Oncol. 2017; 28: viii8-viii12https://doi.org/10.1093/annonc/mdx445
        • Peres L.C.
        • et al.
        Histotype classification of ovarian carcinoma: a comparison of approaches.
        Gynecol. Oncol. 2018; 151: 53-60https://doi.org/10.1016/j.ygyno.2018.08.016
        • Meinhold-Heerlein I.
        • et al.
        The new WHO classification of ovarian, fallopian tube, and primary peritoneal cancer and its clinical implications.
        Arch. Gynecol. Obstet. 2016; 293: 695-700https://doi.org/10.1007/s00404-016-4035-8
        • Moher D.
        • et al.
        Preferred reporting items for systematic reviews and Meta-analyses: the PRISMA statement.
        PLoS Med. 2009; 6e1000097https://doi.org/10.1371/journal.pmed.1000097
        • Munn Z.
        • et al.
        Methodological guidance for systematic reviews of observational epidemiological studies reporting prevalence and cumulative incidence data.
        JBI Evid. Implement. 2015; 13: 147-153https://doi.org/10.1097/xeb.0000000000000054
        • Stijnen T.
        • Hamza T.H.
        • Özdemir P.
        Random effects meta-analysis of event outcome in the framework of the generalized linear mixed model with applications in sparse data.
        Stat. Med. 2010; 29: 3046-3067
        • Schwarzer G.
        • et al.
        Seriously misleading results using inverse of freeman-Tukey double arcsine transformation in meta-analysis of single proportions.
        Res. Synth. Methods. 2019; 10: 476-483https://doi.org/10.1002/jrsm.1348
        • Higgins J.P.T.
        • et al.
        Measuring inconsistency in meta-analyses.
        BMJ (Clin. Res. Ed.). 2003; 327: 557-560https://doi.org/10.1136/bmj.327.7414.557
        • Viechtbauer W.
        • Cheung M.W.-L.
        Outlier and influence diagnostics for meta-analysis.
        Res. Synth. Methods. 2010; 1: 112-125https://doi.org/10.1002/jrsm.11
        • Harter P.
        • et al.
        Prevalence of deleterious germline variants in risk genes including BRCA1/2 in consecutive ovarian cancer patients (AGO-TR-1).
        PLoS One. 2017; 12e0186043https://doi.org/10.1371/journal.pone.0186043
        • Hauke J.
        • et al.
        Deleterious somatic variants in 473 consecutive individuals with ovarian cancer: results of the observational AGO-TR1 study (NCT02222883).
        J. Med. Genet. 2019; 56: 574-580https://doi.org/10.1136/jmedgenet-2018-105930
        • Ataseven B.
        • et al.
        Prevalence of BRCA1 and BRCA2 mutations in patients with primary ovarian cancer - does the German checklist for detecting the risk of hereditary breast and ovarian cancer adequately depict the need for consultation?.
        Geburtshilfe Frauenheilkd. 2020; 80: 932-940https://doi.org/10.1055/a-1222-0042
        • Bu H.
        • et al.
        BRCA mutation frequency and clinical features of ovarian cancer patients: a report from a Chinese study group.
        J. Obstet. Gynaecol. Res. 2019; 45: 2267-2274https://doi.org/10.1111/jog.14090
        • Choi M.C.
        • et al.
        Germline mutations of <em>BRCA1</em> and <em>BRCA2</em> in Korean ovarian cancer patients: finding founder mutations.
        Int. J. Gynecol. Cancer. 2015; 25: 1386-1391https://doi.org/10.1097/igc.0000000000000529
        • Enomoto T.
        • et al.
        The first Japanese nationwide multicenter study of BRCA mutation testing in ovarian cancer: CHARacterizing the cross-sectionaL approach to ovarian cancer geneTic TEsting of BRCA (CHARLOTTE).
        Int. J. Gynecol. Cancer. 2019; 29: 1043-1049https://doi.org/10.1136/ijgc-2019-000384
        • Flaum N.
        • et al.
        Mainstreaming germline BRCA1/2 testing in non-mucinous epithelial ovarian cancer in the north west of England.
        Eur. J. Hum. Genet. 2020; 28: 1541-1547https://doi.org/10.1038/s41431-020-0692-y
        • George A.
        • et al.
        Implementing rapid, robust, cost-effective, patient-centred, routine genetic testing in ovarian cancer patients.
        Sci. Rep. 2016; 6: 29506https://doi.org/10.1038/srep29506
        • Hirasawa A.
        • et al.
        Prevalence of pathogenic germline variants detected by multigene sequencing in unselected Japanese patients with ovarian cancer.
        Oncotarget. 2017; 8: 112258-112267https://doi.org/10.18632/oncotarget.22733
        • Kim S.I.
        • et al.
        Germline and somatic BRCA1/2 gene mutational status and clinical outcomes in epithelial peritoneal, ovarian, and fallopian tube cancer: over a decade of experience in a single institution in Korea.
        Cancer Res. Treat. 2020; 52: 1229-1241https://doi.org/10.4143/crt.2020.557
        • Kowalik A.
        • et al.
        Somatic mutations in BRCA1&2 in 201 unselected ovarian carcinoma samples – single institution study.
        Pol. J. Pathol. 2019; 70: 115-126https://doi.org/10.5114/pjp.2019.82905
        • Lertkhachonsuk A.-A.
        • et al.
        Prevalence of tissue BRCA gene mutation in ovarian, fallopian tube, and primary peritoneal cancers: a multi-institutional study.
        Asian Pacif. J. Cancer Prevent.: APJCP. 2020; 21: 2381-2388https://doi.org/10.31557/APJCP.2020.21.8.2381
        • Lhotova K.
        • et al.
        Multigene panel germline testing of 1333 Czech patients with ovarian cancer.
        Cancers. 2020; 12: 956https://doi.org/10.3390/cancers12040956
        • Li W.
        • et al.
        Germline and somatic mutations of multi-gene panel in Chinese patients with epithelial ovarian cancer: a prospective cohort study.
        J. Ovarian Res. 2019; 12: 80https://doi.org/10.1186/s13048-019-0560-y
        • Manchana T.
        • Phowthongkum P.
        • Teerapakpinyo C.
        Germline mutations in Thai patients with nonmucinous epithelial ovarian cancer.
        World J. Clin. Oncol. 2019; 10: 358https://doi.org/10.5306/wjco.v10.i11.358
        • Morgan R.D.
        • et al.
        Prevalence of germline pathogenic BRCA1/2 variants in sequential epithelial ovarian cancer cases.
        J. Med. Genet. 2019; 56: 301-307https://doi.org/10.1136/jmedgenet-2018-105792
        • Peixoto A.
        • et al.
        Tumor testing for somatic and germline BRCA1/BRCA2 variants in ovarian cancer patients in the context of strong founder effects.
        Front. Oncol. 2020; 10https://doi.org/10.3389/fonc.2020.01318
        • Plaskocinska I.
        • et al.
        New paradigms for BRCA1/BRCA2 testing in women with ovarian cancer: results of the genetic testing in epithelial ovarian Cancer (GTEOC) study.
        J. Med. Genet. 2016; 53: 655-661https://doi.org/10.1136/jmedgenet-2016-103902
        • Rahman B.
        • et al.
        Mainstreamed genetic testing for women with ovarian cancer: first-year experience.
        J. Med. Genet. 2019; 56: 195-198https://doi.org/10.1136/jmedgenet-2017-105140
        • Rivera D.
        • et al.
        Implementing NGS-based <em>BRCA</em> tumour tissue testing in FFPE ovarian carcinoma specimens: hints from a real-life experience within the framework of expert recommendations.
        J. Clin. Pathol. 2020; (p. jclinpath-2020-206840)https://doi.org/10.1136/jclinpath-2020-206840
        • Rumford M.
        • et al.
        Oncologist-led BRCA ‘mainstreaming’ in the ovarian cancer clinic: a study of 255 patients and its impact on their management.
        Sci. Rep. 2020; 10: 3390https://doi.org/10.1038/s41598-020-60149-5
        • Rust K.
        • et al.
        Routine germline BRCA1 and BRCA2 testing in patients with ovarian carcinoma: analysis of the Scottish real-life experience.
        Bjog. 2018; 125: 1451-1458https://doi.org/10.1111/1471-0528.15171
        • Sakamoto I.
        • et al.
        BRCA1 and BRCA2 mutations in Japanese patients with ovarian, fallopian tube, and primary peritoneal cancer.
        Cancer. 2016; 122: 84-90https://doi.org/10.1002/cncr.29707
        • Seo J.H.
        • et al.
        Prevalence and oncologic outcomes of BRCA1/2 mutation and variant of unknown significance in epithelial ovarian carcinoma patients in Korea.
        Obstet. Gynecol. Sci. 2019; 62: 411-419https://doi.org/10.5468/ogs.2019.62.6.411
        • Shi T.
        • et al.
        BRCA1 and BRCA2 mutations in ovarian cancer patients from China: ethnic-related mutations in BRCA1 associated with an increased risk of ovarian cancer.
        Int. J. Cancer. 2017; 140: 2051-2059https://doi.org/10.1002/ijc.30633
        • Sugino K.
        • et al.
        Germline and somatic mutations of homologous recombination-associated genes in Japanese ovarian cancer patients.
        Sci. Rep. 2019; 9: 17808https://doi.org/10.1038/s41598-019-54116-y
        • Wu X.
        • et al.
        The first Nationwide multicenter prevalence study of germline BRCA1 and BRCA2 mutations in Chinese ovarian Cancer patients.
        Int. J. Gynecol. Cancer. 2017; 27: 1650-1657https://doi.org/10.1097/igc.0000000000001065
        • Santandrea G.
        • et al.
        EP973 ‘tumour first’: an institutional experience of reflex tumour BRCA testing in ovarian epithelial carcinomas.
        Int. J. Gynecol. Cancer. 2019; 29: A516https://doi.org/10.1136/ijgc-2019-ESGO.1019
        • Schrader K.A.
        • et al.
        Germline BRCA1 and BRCA2 mutations in ovarian cancer: utility of a histology-based referral strategy.
        Obstet. Gynecol. 2012; 120: 235-240https://doi.org/10.1097/AOG.0b013e31825f3576
        • Kuchenbaecker K.B.
        • et al.
        Risks of breast, ovarian, and contralateral breast Cancer for BRCA1 and BRCA2 mutation carriers.
        JAMA. 2017; 317: 2402-2416https://doi.org/10.1001/jama.2017.7112
        • Kotsopoulos J.
        • et al.
        Age-specific ovarian cancer risks among women with a BRCA1 or BRCA2 mutation.
        Gynecol. Oncol. 2018; 150: 85-91https://doi.org/10.1016/j.ygyno.2018.05.011
        • Fadare O.
        • Parkash V.
        Pathology of endometrioid and clear cell carcinoma of the ovary.
        Surg. Pathol. Clin. 2019; 12: 529-564https://doi.org/10.1016/j.path.2019.01.009
        • Goulding E.A.
        • et al.
        Low-grade serous ovarian carcinoma: a comprehensive literature review.
        Aust. N. Z. J. Obstet. Gynaecol. 2020; 60: 27-33https://doi.org/10.1111/ajo.13105
        • Simons M.
        • et al.
        Two types of primary mucinous ovarian tumors can be distinguished based on their origin.
        Mod. Pathol. 2020; 33: 722-733https://doi.org/10.1038/s41379-019-0401-y
        • Lin L.
        • Chu H.
        Meta-analysis of proportions using generalized linear mixed models.
        Epidemiology. 2020; 31: 713-717https://doi.org/10.1097/ede.0000000000001232
        • Suszynska M.
        • et al.
        Large-scale meta-analysis of mutations identified in panels of breast/ovarian cancer-related genes — providing evidence of cancer predisposition genes.
        Gynecol. Oncol. 2019; 153: 452-462https://doi.org/10.1016/j.ygyno.2019.01.027
        • Tiwari A.K.
        • Roy H.K.
        • Lynch H.T.
        Lynch syndrome in the 21st century: clinical perspectives.
        QJM: Int. J. Med. 2015; 109: 151-158https://doi.org/10.1093/qjmed/hcv137
        • Koornstra J.J.
        • et al.
        Management of extracolonic tumours in patients with Lynch syndrome.
        Lancet Oncol. 2009; 10: 400-408https://doi.org/10.1016/S1470-2045(09)70041-5
        • Krämer P.
        • et al.
        Endometrial Cancer molecular risk stratification is equally prognostic for Endometrioid ovarian carcinoma.
        Clin. Cancer Res. 2020; 26: 5400-5410https://doi.org/10.1158/1078-0432.Ccr-20-1268
        • Woolderink J.M.
        • et al.
        Characteristics of Lynch syndrome associated ovarian cancer.
        Gynecol. Oncol. 2018; 150: 324-330https://doi.org/10.1016/j.ygyno.2018.03.060