Advertisement

Hereditary cancer variants and homologous recombination deficiency in biliary tract cancer

Open AccessPublished:October 12, 2022DOI:https://doi.org/10.1016/j.jhep.2022.09.025

      Highlights

      • Hereditary cancer-predisposing gene alterations were identified in 5.5% of BTCs.
      • Germline variants in BRCA1/2, APC, MSH6 and PALB2 contributed significantly to BTC.
      • BTCs with germline variants in BRCA2 and PALB2 accompanied by LOH showed an HRD phenotype.

      Background & Aims

      The heritability and actionability of variants in homologous recombination-related genes in biliary tract cancers (BTCs) are uncertain. Although associations between BTC and BRCA germline variants have been reported, homologous recombination deficiency has not been investigated in BTCs.

      Methods

      We sequenced germline variants in 27 cancer-predisposing genes in 1,292 BTC cases and 37,583 controls without a personal nor family history of cancer. We compared pathogenic germline variant frequencies between cases and controls and documented the demographic and clinical characteristics of carriers. In addition, whole-genome sequencing of 45 BTC tissues was performed to evaluate homologous recombination deficiency status.

      Results

      Targeted sequencing identified 5,018 germline variants, which were classified into 317 pathogenic, 3,611 variants of uncertain significance, and 1,090 benign variants. Seventy-one BTC cases (5.5%) had at least one pathogenic variant among 27 cancer-predisposing genes. Pathogenic germline variants enriched in BTCs were present in BRCA1, BRCA2, APC, and MSH6 (p <0.00185). PALB2 variants were marginally associated with BTC (p = 0.01). APC variants were predominantly found in ampulla of Vater carcinomas. Whole-genome sequencing demonstrated that three BTCs with pathogenic germline variants in BRCA2 and PALB2, accompanied by loss of heterozygosity, displayed homologous recombination deficiency. Conversely, pathogenic germline variants in without a second hit or other homologous recombination-realted genes such as ATM and BRIP1 showed homologous recombination-proficient phenotypes.

      Conclusions

      In this study, we describe the heritability and actionability of variants in homologous recombination-related genes, which could be used to guide screening and therapeutic strategies for BTCs.

      Impact and implications

      We found that 5.5% of biliary tract cancers (BTCs) in a Japanese population possessed hereditary cancer-predisposing gene alterations, including in BRCA and genes associated with colorectal cancer. Two hits in homologous recombination-related genes were required to confer a homologous recombination-deficient phenotype. PARP inhibitors and DNA-damaging regimens may be effective strategies against BTCs exhibiting homologous recombination deficiency. Hence, in this study, genome-wide sequencing has revealed a potential new therapeutic strategy that could be applied to a subset of BTCs.

      Graphical abstract

      Keywords

      Abbreviations:

      BTC (biliary tract cancer), HRD (homologous recombination deficiency)

      Introduction

      Biliary tract cancer (BTC) is a rare malignant tumor that is anatomically classified into intrahepatic cholangiocarcinoma (ICC), extrahepatic cholangiocarcinoma (ECC), gallbladder carcinoma (GBC), and ampulla of Vater carcinoma (AVC). The incidence of BTC is high in East Asia and South America and increasing worldwide.
      • Bertuccio P.
      • Malvezzi M.
      • Carioli G.
      • Hashim D.
      • Boffetta P.
      • El-Serag H.B.
      • et al.
      Global trends in mortality from intrahepatic and extrahepatic cholangiocarcinoma.
      BTC is diagnosed at an advanced stage and has a worse prognosis than other gastrointestinal cancers because early BTC lacks clinical symptoms and shows more aggressive behavior. Even in patients who have undergone aggressive surgery, the 5-year survival rate is 5–10% for GBC and 10–40% for cholangiocarcinoma.
      • Eckel F.
      • Brunner T.
      • Jelic S.
      Biliary cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up.
      Furthermore, the median overall survival of individuals with unresectable BTC treated with gemcitabine-based chemotherapy is less than 1 year.
      • Valle J.
      • Wasan H.
      • Palmer D.H.
      • Cunningham D.
      • Anthoney A.
      • Maraveyas A.
      • et al.
      Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer.
      Risk factors for BTC include inflammation of the biliary tract caused by gallstones, viral hepatitis, primary sclerosing cholangitis, metabolic conditions, and exposure to chemicals such as aflatoxin.
      • Palmer W.C.
      • Patel T.
      Are common factors involved in the pathogenesis of primary liver cancers? A meta-analysis of risk factors for intrahepatic cholangiocarcinoma.
      ,
      • Koshiol J.
      • Gao Y.T.
      • Dean M.
      • Egner P.
      • Nepal C.
      • Jones K.
      • et al.
      Association of aflatoxin and gallbladder cancer.
      However, these risk factors are not essential for BTC development. Germline variants related to hereditary cancers have been reported in BTC. Lynch syndrome-causing variants (MLH1, MSH2, MSH6, and PMS2) have been shown to confer an increased risk of BTC,
      • Win A.K.
      • Lindor N.M.
      • Young J.P.
      • Macrae F.A.
      • Young G.P.
      • Williamson E.
      • et al.
      Risks of primary extracolonic cancers following colorectal cancer in lynch syndrome.
      and BRCA2 germline variants have also been associated with BTC.
      Breast Cancer Linkage Consortium
      Cancer risks in BRCA2 mutation carriers.
      Our previous report identified BRCA1, BRCA2, MLH1, and RAD51D germline variants in multiple BTCs.
      • Wardell C.P.
      • Fujita M.
      • Yamada T.
      • Simbolo M.
      • Fassan M.
      • Karlic R.
      • et al.
      Genomic characterization of biliary tract cancers identifies driver genes and predisposing mutations.
      However, the National Comprehensive Cancer Network (NCCN) guideline for hepatobiliary cancer only considers germline testing in patients with mismatch repair deficient or microsatellite instability (MSI)-high tumors or family history suggestive of BRCA1/2 variants, but does not actively recommend it.
      • Benson A.B.
      • D'Angelica M.I.
      • Abbott D.E.
      • Anaya D.A.
      • Anders R.
      • Are C.
      • et al.
      Hepatobiliary cancers, version 2.2021, NCCN clinical practice guidelines in oncology.
      This may be because BTC is a rare disease in European populations, and reports on the prevalence of germline variants in BTC are limited. To determine the feasibility of germline testing for BTC, many cases would need to be analyzed and the contribution of cancer-predisposing germline variants to BTC further clarified.
      BRCA1/2 play a vital role in the DNA repair processes via the homologous recombination (HR) repair pathway.
      • Lord C.J.
      • Ashworth A.
      BRCAness revisited.
      Numerous other genes such as ATM, PALB2, and RAD51 are also involved in this process.
      • Pearl L.H.
      • Schierz A.C.
      • Ward S.E.
      • Al-Lazikani B.
      • Pearl F.M.
      Therapeutic opportunities within the DNA damage response.
      ,
      • Casolino R.
      • Paiella S.
      • Azzolina D.
      • Beer P.A.
      • Corbo V.
      • Lorenzoni G.
      • et al.
      Homologous recombination deficiency in pancreatic cancer: a systematic review and prevalence meta-analysis.
      Alterations in HR-related genes result in HR deficiency (HRD), promoting carcinogenesis in breast, ovarian, prostate, and pancreatic cancers. Screening for HRD is clinically crucial because tumor cells with HRD are sensitive to HRD-targeted therapy, such as platinum drugs and poly-ADP ribose polymerase inhibitors (PARPi), and clinical evidence suggests that germline BRCA variants are the best clinical biomarkers for response to PARPi.
      • Hoppe M.M.
      • Sundar R.
      • Tan D.S.P.
      • Jeyasekharan A.D.
      Biomarkers for homologous recombination deficiency in cancer.
      However, there are several drawbacks to using germline tests to determine HRD status. Many germline variants are registered in databases such as ClinVar as variants of uncertain significance (VUS), and their pathogenicity is still controversial. In addition, not all tumors with alterations in HR-related genes have HRD phenotypes. For example, BRCA-mutated breast and ovarian cancers without BRCA locus-specific loss of heterozygosity (LOH) have been reported to respond poorly to platinum drugs.
      • Maxwell K.N.
      • Wubbenhorst B.
      • Wenz B.M.
      • De Sloover D.
      • Pluta J.
      • Emery L.
      • et al.
      BRCA locus-specific loss of heterozygosity in germline BRCA1 and BRCA2 carriers.
      Thus, to expand therapeutic strategies for BTC and evaluate the pathogenicity of variants, HRD status should be evaluated.
      Recently, our pan-cancer analysis found that pathogenic germline variants of BRCA1 were significantly associated with BTC in the Japanese population.
      • Momozawa Y.
      • Sasai R.
      • Usui Y.
      • Shiraishi K.
      • Iwasaki Y.
      • Taniyama Y.
      • et al.
      Expansion of cancer risk profile for BRCA1 and BRCA2 pathogenic variants.
      In this study we analyzed additional BTC cases and performed a large-scale association study of Japanese individuals with BTC and non-cancer controls to identify high-risk germline variants in BTC. Furthermore, the HRD status of BTCs with and without HR-related germline variants was evaluated by whole-genome sequencing (WGS) and somatic mutation profiles.

      Materials and methods

      The study population for germline variant analysis

      We obtained DNA samples of BTCs from BioBank Japan (BBJ) and Hokkaido University Hospital (Fig. 1). BTCs included ICC, ECC, GBC, and AVC, but not duodenal carcinoma in this study. BBJ is a multi-institutional, hospital-based registry that collected blood and clinical information from patients with various common diseases from 2003 to 2018;
      • Hirata M.
      • Kamatani Y.
      • Nagai A.
      • Kiyohara Y.
      • Ninomiya T.
      • Tamakoshi A.
      • et al.
      Cross-sectional analysis of BioBank Japan clinical data: a large cohort of 200,000 patients with 47 common diseases.
      ,
      • Nagai A.
      • Hirata M.
      • Kamatani Y.
      • Muto K.
      • Matsuda K.
      • Kiyohara Y.
      • et al.
      Overview of the BioBank Japan project: study design and profile.
      976 BTCs from the BBJ were analyzed in a recent pan-cancer study.
      • Momozawa Y.
      • Sasai R.
      • Usui Y.
      • Shiraishi K.
      • Iwasaki Y.
      • Taniyama Y.
      • et al.
      Expansion of cancer risk profile for BRCA1 and BRCA2 pathogenic variants.
      For the 316 BTC cases from Hokkaido University Hospital, DNA samples were extracted from blood or resection specimens from normal tissue, e.g. the liver and bile duct mucosa, collected between 2003 and 2020. We selected controls from BBJ who were ≥60-years-old and did not have a personal nor family history of cancer. Clinical characteristics from BBJ were collected by interview or medical record survey using standardized questionnaires at the time of entry into the BBJ. Clinical characteristics from Hokkaido University Hospital were collected from electronic medical records. Information on family cancer history was collected out to third-degree relatives. Written informed consent was obtained from all participants. This study was approved by the Institutional Review Boards of the Institute of Medical Sciences, the University of Tokyo, RIKEN, and Hokkaido University Hospital.
      Figure thumbnail gr1
      Fig. 1Study schema.
      (Left) Germline variant analysis with targeted sequencing. (Right) HRD status analysis with whole-genome sequencing. AVC, ampulla of Vater carcinoma; ECC, extrahepatic cholangiocarcinoma; GBC, gallbladder carcinoma; HRD, homologous recombination deficiency.; ICC, intrahepatic cholangiocarcinoma; NA, not available; VUS, variants of uncertain significance.

      Sequencing and data analysis for germline variants

      We analyzed 27 cancer-predisposing genes as in our previous studies (Fig. 1).
      • Fujita M.
      • Liu X.
      • Iwasaki Y.
      • Terao C.
      • Mizukami K.
      • Kawakami E.
      • et al.
      Population-based screening for hereditary colorectal cancer variants in Japan.
      ,
      • Mizukami K.
      • Iwasaki Y.
      • Kawakami E.
      • Hirata M.
      • Kamatani Y.
      • Matsuda K.
      • et al.
      Genetic characterization of pancreatic cancer patients and prediction of carrier status of germline pathogenic variants in cancer-predisposing genes.
      We selected these genes based on the Myriad hereditary cancer gene panel.
      • Buys S.S.
      • Sandbach J.F.
      • Gammon A.
      • Patel G.
      • Kidd J.
      • Brown K.L.
      • et al.
      A study of over 35,000 women with breast cancer tested with a 25-gene panel of hereditary cancer genes.
      These genes included Lynch syndrome-causing genes and BRCA1/2, for which germline testing is considered in the NCCN guidelines on hepatobiliary cancer. The germline mutation calling methods were described in our previous studies.
      • Momozawa Y.
      • Iwasaki Y.
      • Parsons M.T.
      • Kamatani Y.
      • Takahashi A.
      • Tamura C.
      • et al.
      Germline pathogenic variants of 11 breast cancer genes in 7,051 Japanese patients and 11,241 controls.
      ,
      • Momozawa Y.
      • Iwasaki Y.
      • Hirata M.
      • Liu X.
      • Kamatani Y.
      • Takahashi A.
      • et al.
      Germline pathogenic variants in 7636 Japanese patients with prostate cancer and 12 366 controls.

      Annotation of germline variants

      As in our previous studies,
      • Momozawa Y.
      • Sasai R.
      • Usui Y.
      • Shiraishi K.
      • Iwasaki Y.
      • Taniyama Y.
      • et al.
      Expansion of cancer risk profile for BRCA1 and BRCA2 pathogenic variants.
      we determined the clinical significance (pathogenic, benign, or VUS) of all variants using the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines and pathogenicity interpretation registered in ClinVar (2020/September/19 version).
      • Momozawa Y.
      • Iwasaki Y.
      • Parsons M.T.
      • Kamatani Y.
      • Takahashi A.
      • Tamura C.
      • et al.
      Germline pathogenic variants of 11 breast cancer genes in 7,051 Japanese patients and 11,241 controls.
      ,
      • Momozawa Y.
      • Iwasaki Y.
      • Hirata M.
      • Liu X.
      • Kamatani Y.
      • Takahashi A.
      • et al.
      Germline pathogenic variants in 7636 Japanese patients with prostate cancer and 12 366 controls.
      Variants not registered in ClinVar on September 19, 2020, were considered novel.

      WGS and RNA sequencing of BTC tissues

      To evaluate HRD status, WGS was performed for BTCs (Fig. 1). First, we obtained WGS data from 36 tumor and normal tissue samples, which were analyzed in our previous studies.
      • Wardell C.P.
      • Fujita M.
      • Yamada T.
      • Simbolo M.
      • Fassan M.
      • Karlic R.
      • et al.
      Genomic characterization of biliary tract cancers identifies driver genes and predisposing mutations.
      ,
      • Fujimoto A.
      • Furuta M.
      • Shiraishi Y.
      • Gotoh K.
      • Kawakami Y.
      • Arihiro K.
      • et al.
      Whole-genome mutational landscape of liver cancers displaying biliary phenotype reveals hepatitis impact and molecular diversity.
      • Fujimoto A.
      • Furuta M.
      • Totoki Y.
      • Tsunoda T.
      • Kato M.
      • Shiraishi Y.
      • et al.
      Whole-genome mutational landscape and characterization of noncoding and structural mutations in liver cancer.
      • Ebata N.
      • Fujita M.
      • Sasagawa S.
      • Maejima K.
      • Okawa Y.
      • Hatanaka Y.
      • et al.
      Molecular classification and tumor microenvironment characterization of gallbladder cancer by comprehensive genomic and transcriptomic analysis.
      All 36 BTCs were included in the germline variant analysis cohort. In addition, we selected 16 BTCs with HR-related pathogenic germline variants or VUS identified in the Hokkaido cohort. DNA was extracted from frozen tissue using the QIAamp DNA mini kit (QIAGEN) and the FlexiGene DNA Kit (QIAGEN) for blood, following the manufacturer's protocol. WGS libraries were prepared using the TruSeq Nano DNA High Throughput Library Prep Kit (Illumina) with 200 ng DNA input. 500–600 base pair (bp) insert libraries were prepared following the manufacturer's protocol. Thirty-five WGS libraries were sequenced on HiSeq 2500/2000, and 17 WGS libraries were sequenced on NovaSeq 6000, yielding paired reads of 100–150 bp. The mean sequence depths were 52x for the tumor WGS and 36x for the normal tissue WGS. We called germline and somatic variants from WGS data by using the integrated pipeline with HRD.
      • Priestley P.
      • Baber J.
      • Lolkema M.P.
      • Steeghs N.
      • de Bruijn E.
      • Shale C.
      • et al.
      Pan-cancer whole-genome analyses of metastatic solid tumours.
      We obtained RNA sequencing (RNA-seq) data, which were analyzed in our previous study.
      • Okawa Y.
      • Ebata N.
      • Kim N.K.D.
      • Fujita M.
      • Maejima K.
      • Sasagawa S.
      • et al.
      Actionability evaluation of biliary tract cancer by genome transcriptome analysis and Asian cancer knowledgebase.
      RNA was extracted from frozen tumor tissues, and RNA-seq libraries were prepared using the TruSeq Stranded mRNA Sample Prep kit (Illumina) following the manufacturer’s protocol. We described further details regarding methods of WGS and RNA-seq analysis in the supplementary materials and methods.

      HRD status analysis

      We predicted HRD status using the R package Classifier of HOmologous Recombination v2.0 (CHORD).
      • Nguyen L.
      • Martens J.W.M.
      • Van Hoeck A.
      • Cuppen E.
      Pan-cancer landscape of homologous recombination deficiency.
      Briefly, the somatic single nucleotide variants, indels, and structural variants (SVs) were passed to the CHORD function extractSigsChord to extract mutation contexts and their function chordPredict. HRD was determined when the calculated HRD probability was 0.5 or higher. To validate the CHORD results, the mutational signature of each sample was calculated based on all the detected single nucleotide variants and the trinucleotide sequence context, including the bases immediately 5' and 3' to the mutated base (96 possible mutation types). Signatures were extracted using the Bayesian non-negative matrix factorization approach of the R package Sigminer.
      • Wang S.
      • Li H.
      • Song M.
      • Tao Z.
      • Wu T.
      • He Z.
      • et al.
      Copy number signature analysis tool and its application in prostate cancer reveals distinct mutational processes and clinical outcomes.
      These signatures were compared to the COSMIC v3 database using cosine similarity.
      • Alexandrov L.B.
      • Kim J.
      • Haradhvala N.J.
      • Huang M.N.
      • Tian Ng A.W.
      • Wu Y.
      • et al.
      The repertoire of mutational signatures in human cancer.

      Statistical analysis

      All statistical analyses were performed using R 4.0.2. Case-control association analysis was performed using Fisher's exact test with a recessive model for MUTYH and with dominant models for the other genes. Bonferroni correction was applied for the association analysis between 27 genes with pathogenic variants and BTC (p <0.00185 = 0.05/27). The Wilcoxon rank-sum test was used for continuous variables, and Fisher's exact test was used for nominal variables to compare pathogenic germline variants and clinical information. The trend test was analyzed using the Cochran-Armitage trend test. Statistical significance was set at p <0.05.

      Results

      Clinicopathological characteristics in germline variant analysis

      We analyzed germline variants in 1,292 BTCs and 37,583 controls aged ≥60 years without a personal nor family history of cancer (Table S1).
      • Fujimoto A.
      • Furuta M.
      • Shiraishi Y.
      • Gotoh K.
      • Kawakami Y.
      • Arihiro K.
      • et al.
      Whole-genome mutational landscape of liver cancers displaying biliary phenotype reveals hepatitis impact and molecular diversity.
      • Fujimoto A.
      • Furuta M.
      • Totoki Y.
      • Tsunoda T.
      • Kato M.
      • Shiraishi Y.
      • et al.
      Whole-genome mutational landscape and characterization of noncoding and structural mutations in liver cancer.
      • Ebata N.
      • Fujita M.
      • Sasagawa S.
      • Maejima K.
      • Okawa Y.
      • Hatanaka Y.
      • et al.
      Molecular classification and tumor microenvironment characterization of gallbladder cancer by comprehensive genomic and transcriptomic analysis.
      The mean age of individuals with BTC was 69.2 years, and 66.8% were male. Of the individuals with BTC, 36.9% had at least one event in their previous personal cancer history, and 62.8% had at least one event in their family cancer history. In terms of BTC location, ECC, including perihilar and distal cholangiocarcinoma, was the most common, followed by GBC, ICC, and AVC (Fig. 1).

      Pathogenic germline variants and VUS

      Targeted sequencing for 27 cancer-predisposing genes identified 5,018 germline variants among 1,292 cases and 37,583 controls (Table S2). We classified germline variants into 317 pathogenic, 3,611 VUS, and 1,090 benign variants using ACMG/AMP guidelines and ClinVar. Of 317 pathogenic variants, 127 variants (40.1%) were novel and not registered in ClinVar. Among the 27 cancer-predisposing genes, BRCA2 had the highest number of germline variants (574 variants), and BRCA2 and ATM had the highest number of pathogenic variants (41 pathogenic variants) (Fig. S1).
      We compared the frequency of pathogenic variants in BTCs with that in controls (Table S3). Seventy-one cases (5.5%) and 520 controls (1.38%) had at least one pathogenic variant among the 27 cancer-predisposing genes (p = 1.017x10-20, odds ratio [OR] 4.1, 95% CI 3.2–5.4). Ten genes had p values of less than 0.05 (BRCA1, BRCA2, APC, MSH6, PALB2, CDK4, ATM, BARD1, MLH1, MSH2). After Bonferroni correction, BRCA1 (p = 4.213x10-10, OR 13.6, 95% CI 6.5–27.3), BRCA2 (p = 2.225x10-7, OR 6.5, 95% CI 3.4–11.8), APC (p = 4.143 ×10-5, OR 18.2, 95% CI 4.7–63.3), and MSH6 (p = 7.591x10-4, OR 5.2, 95% CI 2–11.9) were significantly enriched in BTCs. Next, to investigate whether VUS contribute to the carcinogenesis of BTC, we extracted 2,917 non-synonymous VUS with minor allele frequency <0.001 from 3,611 total VUS and compared the VUS frequency in BTCs with that in controls (Table S4). As a result, no significant difference was observed (p = 0.075, OR 0.9, 95% CI 0.8–1), suggesting that pathogenic variants and not VUS in the four genes (BRCA1, BRCA2, APC, and MSH6) are associated with BTC. Pathogenic variants in these four significant genes and PALB2, whose p value was lowest in six marginally significant genes (p = 0.01), are shown in Fig. 2. Most pathogenic variants in BTCs were singletons, but p.Arg2318∗ of BRCA2, p.Leu63∗ of BRCA1, p.Glu546Gln of MSH6, and p.Ilu558fr of PALB2 were recurrently observed in BTCs. We have previously reported germline variants in breast, prostate, and pancreatic cancers.
      • Mizukami K.
      • Iwasaki Y.
      • Kawakami E.
      • Hirata M.
      • Kamatani Y.
      • Matsuda K.
      • et al.
      Genetic characterization of pancreatic cancer patients and prediction of carrier status of germline pathogenic variants in cancer-predisposing genes.
      ,
      • Momozawa Y.
      • Iwasaki Y.
      • Parsons M.T.
      • Kamatani Y.
      • Takahashi A.
      • Tamura C.
      • et al.
      Germline pathogenic variants of 11 breast cancer genes in 7,051 Japanese patients and 11,241 controls.
      ,
      • Momozawa Y.
      • Iwasaki Y.
      • Hirata M.
      • Liu X.
      • Kamatani Y.
      • Takahashi A.
      • et al.
      Germline pathogenic variants in 7636 Japanese patients with prostate cancer and 12 366 controls.
      Across those three reports, p.Arg2318∗ of BRCA2 was a significant risk factor for all three cancers, p.Leu63∗ of BRCA1 was identified as a risk factor for breast cancer, and p.Glu546Gln of MSH6 was identified only in the pancreatic cancer study but did not pass significance cut-offs. In contrast, p.Ilu558fr of PALB2 was not found in those other cancer studies.
      Figure thumbnail gr2
      Fig. 2Type and number of pathogenic variants in BTCs.
      (A) BRCA1, (B) BRCA2, (C) APC, (D) MSH6, (E) PALB2. The x-axis shows the position number of amino acid residues, and the y-axis shows the total number of cases with each pathogenic variant. Red circles represent truncating variants, and blue circles represent missense variants. Amino acid substitutions observed in multiple cases are labeled. BTC, biliary tract cancer.
      Next, we compared this BTC study with our pancreatic cancer study,
      • Mizukami K.
      • Iwasaki Y.
      • Kawakami E.
      • Hirata M.
      • Kamatani Y.
      • Matsuda K.
      • et al.
      Genetic characterization of pancreatic cancer patients and prediction of carrier status of germline pathogenic variants in cancer-predisposing genes.
      which reported that pathogenic germline variants in BRCA1/2 and ATM were associated with pancreatic cancer. Similar to the current study, the prevalence of pathological germline variants in pancreatic cancer was 6.67%. In addition, the prevalence of BRCA1 variants in pancreatic cancer was 0.9%, identical to that of BTC (1.01%). However, in pancreatic cancer, the prevalence of BRCA2 variants was more than twice that of BTC (1.08% in BTC and 2.49% in pancreatic cancer), and the total prevalence of Lynch syndrome-causing genes was approximately half that observed in BTC (0.93% in BTC, 0.4% in pancreatic cancer).

      Clinicopathological characteristics of pathogenic germline variant carriers

      To evaluate the clinicopathological characteristics of BTCs with pathogenic germline variants (carriers), we compared 71 carriers with 1,221 non-carriers (Table S5). The mean age at diagnosis was 66.9 years in carriers, which was significantly younger than in non-carriers (mean: 69.3 years) (Wilcoxon rank-sum test: p = 0.038). The percentage of carriers increased with decreasing age (Cochran-Armitage trend test: p = 0.019), and carriers accounted for 12.8% of all BTCs under 50 years old (Fig. 3A). The carriers had increased personal and family histories of breast cancer (personal: p = 0.026, OR 5.8, 95% CI 1.3–22.1; family: p = 0.047, OR 2.5, 95% CI 1–5.5). There were no significant differences in sex, lymph node, or distant metastasis. Regarding the anatomical location, ICC had the highest proportion of carriers (17/199 = 8.5%) (Fig. 3B). Variants in BRCA1 were observed with similar frequency in all locations (ICC: 1%, GBC: 0.9%, ECC: 1.1%, APC: 1.4%). In contrast, variants in APC were observed most frequently in AVC (2.7%) but not in ICC or GBC. BTCs with APC variants had no personal or family history of familial adenomatous polyposis (FAP). In addition, there was no significant difference for this aggressive cancer prognoses between carriers and non-carriers in these two BTC cohorts (available data: carrier/non-carrier= 18/259 in Hokkaido and 10/159 in BBJ).
      Figure thumbnail gr3
      Fig. 3Association between germline variants and clinicopathological characteristics in BTCs.
      (A) The proportion of BTCs with germline variants by age group. The trend test was performed using the Cochran-Armitage trend test. (B) The proportion of pathogenic germline variants by location of BTC. AVC, ampulla of Vater carcinoma; BTC, biliary tract cancer; ECC, extrahepatic cholangiocarcinoma; GBC, gallbladder carcinoma; ICC, intrahepatic cholangiocarcinoma; n.a., not available.

      HRD analysis of BTC tissues with or without HR-related variants

      We focused on the point that germline pathogenic variants of BRCA1/2 and other HR-related genes such as PALB2, ATM, BARD1, and BRIP1 were observed in BTCs and investigated whether BTCs with HR-related germline variants showed HRD. We evaluated HRD status from the whole-genome somatic mutation profiles of 52 BTCs by CHORD, which classifies HRD status by analyzing specific indel and SV types using machine learning methods.
      • Priestley P.
      • Baber J.
      • Lolkema M.P.
      • Steeghs N.
      • de Bruijn E.
      • Shale C.
      • et al.
      Pan-cancer whole-genome analyses of metastatic solid tumours.
      ,
      • Nguyen L.
      • Martens J.W.M.
      • Van Hoeck A.
      • Cuppen E.
      Pan-cancer landscape of homologous recombination deficiency.
      We excluded six BTCs because of extremely low tumor purity and one BTC due to MSI-high status. Forty-five BTCs remained for further analysis, which included seven BTCs with HR-related pathogenic germline variants (one ATM, one BRCA1, three BRCA2, one BRIP1, one PALB2), seven BTCs with HR-related VUS (two BRCA1, one BRCA2, one CHEK2, two PALB2, one RAD51D), and 31 BTCs without HR-related germline variants as negative controls (Fig. 1). The clinical characteristics and variant information of the 45 BTCs are shown in Table S6. Candidate driver genes, candidate copy number variant drivers, and all fusion genes are shown in Tables S7–9.
      HRD probabilities were calculated for 45 BTCs by CHORD (Fig. 4, Table S6), which used short deletions with flanking microhomology to predict HRD status. CHORD focuses on deletions with ≥2 bp flanking homology (del.mh.bimh.2.5) to predict HRD, excluding deletions with 1 bp flanking microhomology (del.mh.bimh.1) associated with radiation. CHORD predicts tumors with more del.mh.bimh.2.5 to be HRD. Furthermore, to distinguish between BRCA1-type HRD and BRCA2-type HRD, 1–100 kb structural duplications are used; tumors with more 1–100 kb structural duplications are predicted to be BRCA1-type HRD. Of seven BTCs with HR-related pathogenic germline variants, three BTCs (HK20, BHK072, BHK59) were predicted as BRCA2-type HRD (probability >0.5), all of which had truncating variants (BRCA2 p.Thr3033fs, BRCA2 p.Gln3026∗, PALB2 p.Ile558fs) accompanied by LOH at the variant locus. Furthermore, all three of the affected individuals had family histories of cancer. However, one BTC (HK110), with a germline frameshift variant in ATM (p.Ile2629fs), was predicted to be HR proficient (HRP) despite LOH. Another case (RK567) with a germline missense variant in BRIP1 (p.Ala349Ser), was also predicted to be HRP despite LOH. This missense variant had been registered as VUS in ClinVar but was classified as a pathogenic variant based on the ACMG/AMP guidelines in this study. The two other BTC cases (CHK005, HK09) were predicted to be HRP and did not have LOH at either variant locus. These results indicated that alterations in BRCA2 and PALB2 along with locus-specific LOH are required to elicit HRD in BTCs, consistent with the classical two-hit model of tumor suppressor genes. In addition, this study did not find pathogenic variants in ATM and BRIP1 to influence HRD status in BTCs.
      Figure thumbnail gr4
      Fig. 4HRD status of 45 BTCs by WGS analysis.
      The oncoplot shows clinical information and HRD prediction results by CHORD. Cases are sorted in descending order of the proportion of del.mh.bimh.2.5. BTC, biliary tract cancer; del.mh.bimh.2.5, deletion with ≥2 bp flanking microhomology; del.mh.bimh.x, deletions with flanking microhomology split by bins of x-bp in microhomology; del.none, deletion with none of the above; del.rep, deletion in repeat regions; HRD, homologous recombination deficiency; ins.mh, insertion with flanking microhomology; ins.none, insertion with none of the above; ins.rep, insertion in repeat regions; WGS, whole-genome sequencing.
      In contrast, all seven BTCs with HR-related VUS were predicted as HRP. Of those seven BTCs, only CHK008, with a missense variant in PALB2 (c.2560A>C p.Asn854His), was accompanied by LOH at the variant locus. Of the 31 negative controls, three BTCs (HK122, HK82, RK142) were predicted as BRCA2-type HRD. HK122 had a somatic frameshift mutation in BRCA2 (c.6373dupA p.Thr2125fs) with LOH, which represents two somatic hits. In the latter two BTCs, no obvious genomic events or variants were observed as the cause of HRD. RK272 had a somatic variant in ATM with LOH but was predicted as HRP. No BTC was predicted as BRCA1-type HRD in this study. We examined the gene expression of BRCA1/2 in 31 BTCs for which RNA-seq data were available and found that HK82 and RK142 tended to have low expression of BRCA1/2 (Fig. S2).
      To validate the HRD status results from CHORD, which predicts HRD status mainly by focusing on short indels and SVs, we performed a single base substitution (SBS) signature analysis. The Bayesian non-negative matrix factorization approach extracted 13 SBS signatures (Fig. S3). Sig5 showed the highest similarity to SBS3 (cosine similarity: 0.766), which is associated with HRD. All three BTCs with pathogenic germline variants predicted HRD had high burdens of Sig5 (Fig. 4, Fig. S4), whereas BTCs with pathogenic germline variants or VUS predictive of HRP had low burdens of Sig5, and Sig5 burden levels in the negative controls were variable.
      Finally, circos plots of the genomic structures of 45 BTCs are shown in Fig. 5 and Fig. S5. Three BTCs with pathogenic germline variants predictive of HRD had high levels of SVs. In contrast, four BTCs with pathogenic germline variants predictive of HRP had fewer SVs. These results suggest that both germline testing and HRD status assessment may be essential for developing therapeutic strategies targeting HRD in BTC.
      Figure thumbnail gr5
      Fig. 5Circos plots of BTCs with pathogenic variants in homologous recombination-related genes.
      (A) Circos plots of homologous recombination deficient phenotypes. (B) Homologous recombination proficient phenotypes. The outer first circle displays chromosomes, with darker shaded areas representing large human reference genome gaps. The second circle shows somatic variants divided into an outer ring of single nucleotide polymorphism allele frequencies and an inner ring of indels. The third circle shows copy number changes. The fourth circle represents minor allele copy numbers. The normal minor allele copy number is two, and anything below one represents LOH events. Minor allele copy numbers above one indicate amplification events. The innermost circle displays structural variants. LOH, loss of heterozygosity.

      Discussion

      We revealed that 5.5% of Japanese BTCs had pathogenic germline variants among 27 cancer-predisposing genes. Furthermore, comparing BTCs with non-cancer controls showed that carriers with variants in BRCA1/2, APC, and MSH6 were at high risk of BTC carcinogenesis. WGS showed that BTCs with pathogenic germline variants in BRCA2 and PALB2 with LOH had HRD phenotypes.
      Maynard et al.
      • Maynard H.
      • Stadler Z.K.
      • Berger M.F.
      • Solit D.B.
      • Ly M.
      • Lowery M.A.
      • et al.
      Germline alterations in patients with biliary tract cancers: a spectrum of significant and previously underappreciated findings.
      analyzed germline variants in 131 BTCs in a multi-ethnic population and showed that the prevalence of carriers was 16%. Lin et al.
      • Lin J.
      • Cao Y.
      • Yang X.
      • Li G.
      • Shi Y.
      • Wang D.
      • et al.
      Mutational spectrum and precision oncology for biliary tract carcinoma.
      reviewed germline variants in 803 BTCs in a Chinese population, including cases from archived data, and showed that the prevalence of carriers was 12%. All three studies, including our study, showed different carrier prevalence rates because of differences in BTC subtype proportions and the sequenced gene panels.
      In our study, DNA damage repair genes such as BRCA1, BRCA2, and MSH6 were enriched in BTCs. BTCs with germline variants in DNA damage repair genes demonstrated a higher sensitivity to platinum and PARPi in a previous clinical analysis,
      • Golan T.
      • Raitses-Gurevich M.
      • Kelley R.K.
      • Bocobo A.G.
      • Borgida A.
      • Shroff R.T.
      • et al.
      Overall survival and clinical characteristics of BRCA-associated cholangiocarcinoma: a multicenter retrospective study.
      and immune checkpoint inhibitors have also been shown to be effective among individuals with Lynch syndrome and those with MSI-high tumors.
      • Therkildsen C.
      • Jensen L.H.
      • Rasmussen M.
      • Bernstein I.
      An update on immune checkpoint therapy for the treatment of lynch syndrome.
      These drugs may represent potential therapeutic interventions for BTCs. APC germline variants were identified in this study as a risk factor for BTC and were enriched in AVC. AVC are heterogeneous tumors with an intestinal or biliopancreatic phenotype,
      • Wong W.
      • Lowery M.A.
      • Berger M.F.
      • Kemel Y.
      • Taylor B.
      • Zehir A.
      • et al.
      Ampullary cancer: evaluation of somatic and germline genetic alterations and association with clinical outcomes.
      but it might be that APC alterations in AVC are more prevalent in intestinal type than biliopancreatic type. In addition, all BTCs with APC variants in this study are likely to be a manifestation of attenuated FAP based on the variant location
      • Wong W.
      • Lowery M.A.
      • Berger M.F.
      • Kemel Y.
      • Taylor B.
      • Zehir A.
      • et al.
      Ampullary cancer: evaluation of somatic and germline genetic alterations and association with clinical outcomes.
      ,
      • Knudsen A.L.
      • Bisgaard M.L.
      • Bülow S.
      Attenuated familial adenomatous polyposis (AFAP). A review of the literature.
      – in line with previous reports on BTCs with germline APC variants.
      • Knudsen A.L.
      • Bisgaard M.L.
      • Bülow S.
      Attenuated familial adenomatous polyposis (AFAP). A review of the literature.
      ,
      • Dinarvand P.
      • Davaro E.P.
      • Doan J.V.
      • Ising M.E.
      • Evans N.R.
      • Phillips N.J.
      • et al.
      Familial adenomatous polyposis syndrome: an update and review of extraintestinal manifestations.
      However, more cases will be required to elucidate the association between APC variants and BTC.
      HRD is caused by aberrations in HR-related genes and promotes breast, ovarian, prostate, and pancreatic cancers. Clinical trials for PARPi have used BRCA1/2 variants and HRD scores from the commercial my Choice assay to determine HRD status.
      • Telli M.L.
      • Jensen K.C.
      • Vinayak S.
      • Kurian A.W.
      • Lipson J.A.
      • Flaherty P.J.
      • et al.
      Phase II study of gemcitabine, carboplatin, and iniparib as neoadjuvant therapy for triple-negative and BRCA1/2 mutation-associated breast cancer with assessment of a tumor-based measure of genomic instability: PrECOG 0105.
      ,
      • Coleman R.L.
      • Oza A.M.
      • Lorusso D.
      • Aghajanian C.
      • Oaknin A.
      • Dean A.
      • et al.
      Rucaparib maintenance treatment for recurrent ovarian carcinoma after response to platinum therapy (ARIEL3): a randomised, double-blind, placebo-controlled, phase 3 trial.
      Several clinical trials have shown that breast and ovarian cancers with HRD-high have good sensitivity to PARPi and platinum drugs. Our WGS analysis found that three BTCs with HR-related pathogenic germline variants accompanied by LOH at the variant loci were predicted as HRD. In contrast, BTCs with BRCA1/2 pathogenic germline variants were predicted HRP when LOH was absent. These results suggest that HRD requires the coexistence of HR-related pathogenic variants and LOH-induced loss of function of tumor suppressor genes. Alternatively, a somatic double hit by mutation or epigenomic alteration is required, which was observed in our study in at least one case. BRIP1 and ATM are also involved in HR repair pathways, but BTCs with these gene variants were predicted to be HRP in this study. A missense variant in BRIP1 classified as pathogenic in this study may not have led to a deleterious loss of function in BTC. Furthermore, a clinical study of prostate cancer found that PARPi was not effective based on ATM variant status alone,
      • de Bono J.
      • Mateo J.
      • Fizazi K.
      • Saad F.
      • Shore N.
      • Sandhu S.
      • et al.
      Olaparib for metastatic castration-resistant prostate cancer.
      suggesting that, at least in BTC, ATM variants may have little effect on the HR repair pathway. All cases with HR-related VUS were predicted to be HRP. In particular, a case with a missense variant in PALB2 (c.2560A>C p.Asn854His) was predicted to be HRP even though it was accompanied by LOH at the variant locus, suggesting that this variant may be of low pathogenicity. HRD status analysis may improve the interpretation of VUS.
      • Osorio A.
      • de la Hoya M.
      • Rodríguez-López R.
      • Martínez-Ramírez A.
      • Cazorla A.
      • Granizo J.J.
      • et al.
      Loss of heterozygosity analysis at the BRCA loci in tumor samples from patients with familial breast cancer.
      Additionally, driver genes were not identified in some cases with predicted HRD, and the pan-cancer CHORD paper reported that the causes of HRD in some patients were unknown.
      • Nguyen L.
      • Martens J.W.M.
      • Van Hoeck A.
      • Cuppen E.
      Pan-cancer landscape of homologous recombination deficiency.
      These cases may involve epigenetic changes or unknown HR-related genes.
      Golan et al.
      • Golan T.
      • Raitses-Gurevich M.
      • Kelley R.K.
      • Bocobo A.G.
      • Borgida A.
      • Shroff R.T.
      • et al.
      Overall survival and clinical characteristics of BRCA-associated cholangiocarcinoma: a multicenter retrospective study.
      retrospectively reported that in BRCA-mutated BTCs, treatment with platinum drugs or PARPi was associated with better overall survival than previously described for BTCs. In this study, no BTCs were treated by PARPi, but we identified one case (BHK59) with a BRCA2 germline variant and HRD phenotype that responded well to platinum drugs and radiation therapy (Fig. S6). Taken with previous reports that DNA damaging agents, such as platinum drugs and radiation, are favorable therapeutic interventions for tumors with HDR,
      • Abbotts R.
      • Topper M.J.
      • Biondi C.
      • Fontaine D.
      • Goswami R.
      • Stojanovic L.
      • et al.
      DNA methyltransferase inhibitors induce a BRCAness phenotype that sensitizes NSCLC to PARP inhibitor and ionizing radiation.
      ,
      • Park W.
      • Chen J.
      • Chou J.F.
      • Varghese A.M.
      • Yu K.H.
      • Wong W.
      • et al.
      Genomic methods identify homologous recombination deficiency in pancreas adenocarcinoma and optimize treatment selection.
      these results support the use of DNA damaging agents and PARPi for HRD-positive BTCs. Furthermore, this study showed that pathogenic variants of PALB2 were marginally significantly enriched in BTCs, and one BTC with a PALB2 frameshift variant with LOH had an HRD phenotype. Recently, the ACMG guideline has been updated and states that PALB2 variants carry a risk of breast and ovarian cancer comparable to BRCA1/2 and recommend their return as secondary findings from clinical exome and genome sequencing.
      • Miller D.T.
      • Lee K.
      • Chung W.K.
      • Gordon A.S.
      • Herman G.E.
      • Klein T.E.
      • et al.
      ACMG SF v3.0 list for reporting of secondary findings in clinical exome and genome sequencing: a policy statement of the American College of Medical Genetics and Genomics (ACMG).
      Thus, the evaluation of HR-related genes, including BRCA1/2 and PALB2, may be helpful in BTC, for which only limited therapeutic options are available.
      There are several limitations to this study. First, the cohort was limited to Japanese individuals, and germline analysis for BTC should be expanded to include all of Asia, where BTC rates are high. Second, the number of cases for HRD status analysis was small. To investigate the potential of HRD-targeted therapy, we need to accumulate a larger number of cases for analysis.
      We have described the prevalence of germline variants in Japanese BTCs for 27 cancer-predisposing genes. Furthermore, the WGS revealed that HR-related pathogenic germline variants had HRD phenotypes in BTC. These results suggest that HRD-targeted therapies could be appropriate for certain BTCs.

      Abbreviations

      ACMG/AMP, American College of Medical Genetics and Genomics/Association for Molecular Pathology; AVC, ampulla of Vater carcinoma; BBJ, BioBank Japan; bp, base pair; BTC, biliary tract cancer; ECC, extrahepatic cholangiocarcinoma; FAP, familial adenomatous polyposis; GBC, gallbladder carcinoma; HR, homologous recombination; HRD, homologous recombination deficiency; HRP, homologous recombination proficient; ICC, intrahepatic cholangiocarcinoma; LOH, loss of heterozygosity; MSI, microsatellite instability; NCCN, National Comprehensive Cancer Network; PARPi, poly-ADP ribose polymerase inhibitors; RNA-seq; RNA sequencing; SBS, single base substitution; SV, structural variant; VUS, variants of uncertain significance; WGS, whole-genome sequencing.

      Financial support

      This work was partly supported by AMED under Grant Number JP19kk0305010 awarded to Yu.M. and JSPS KAKENHI Grant Numbers 18H04049 and 22H03063 awarded to H.N.

      Conflict of interest

      The authors have no conflicts of interest to declare.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Authors’ contributions

      Conceived the study design: YuM, and HN. Performed the experiments and analyzed the data: YO, YI, TAJ, NE, CI, ME, KaM, SS, MF, YuM, and HN. Collected clinical samples and clinical information: YO, YI, NE, CI, ME, KoM, YoM, TN, SH, and YuM. Wrote the manuscript: YO, TAJ, and HN. Approved the final version of the manuscript: all authors. Contributed to the funding for this study: YuM, and HN.

      Data availability statement

      We deposited raw whole-genome sequencing and RNA-seq data in the National Bioscience Database Center (NBDC) study number JGAS000109, JGAS000151, and JGAS000389.

      Acknowledgements

      We acknowledge the staff of the RIKEN-IMS, BBJ, and Hokkaido University for their technical assistance.

      Supplementary data

      The following are the supplementary data to this article:

      References

        • Bertuccio P.
        • Malvezzi M.
        • Carioli G.
        • Hashim D.
        • Boffetta P.
        • El-Serag H.B.
        • et al.
        Global trends in mortality from intrahepatic and extrahepatic cholangiocarcinoma.
        J Hepatol. 2019; 71: 104-114
        • Eckel F.
        • Brunner T.
        • Jelic S.
        Biliary cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up.
        Ann Oncol. 2011; 22: vi40-vi44
        • Valle J.
        • Wasan H.
        • Palmer D.H.
        • Cunningham D.
        • Anthoney A.
        • Maraveyas A.
        • et al.
        Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer.
        N Engl J Med. 2010; 362: 1273-1281
        • Palmer W.C.
        • Patel T.
        Are common factors involved in the pathogenesis of primary liver cancers? A meta-analysis of risk factors for intrahepatic cholangiocarcinoma.
        J Hepatol. 2012; 57: 69-76
        • Koshiol J.
        • Gao Y.T.
        • Dean M.
        • Egner P.
        • Nepal C.
        • Jones K.
        • et al.
        Association of aflatoxin and gallbladder cancer.
        Gastroenterology. 2017; 153: 488-494.e481
        • Win A.K.
        • Lindor N.M.
        • Young J.P.
        • Macrae F.A.
        • Young G.P.
        • Williamson E.
        • et al.
        Risks of primary extracolonic cancers following colorectal cancer in lynch syndrome.
        J Natl Cancer Inst. 2012; 104: 1363-1372
        • Breast Cancer Linkage Consortium
        Cancer risks in BRCA2 mutation carriers.
        J Natl Cancer Inst. 1999; 91: 1310-1316
        • Wardell C.P.
        • Fujita M.
        • Yamada T.
        • Simbolo M.
        • Fassan M.
        • Karlic R.
        • et al.
        Genomic characterization of biliary tract cancers identifies driver genes and predisposing mutations.
        J Hepatol. 2018; 68: 959-969
        • Benson A.B.
        • D'Angelica M.I.
        • Abbott D.E.
        • Anaya D.A.
        • Anders R.
        • Are C.
        • et al.
        Hepatobiliary cancers, version 2.2021, NCCN clinical practice guidelines in oncology.
        J Natl Compr Canc Netw. 2021; 19: 541-565
        • Lord C.J.
        • Ashworth A.
        BRCAness revisited.
        Nat Rev Cancer. 2016; 16: 110-120
        • Pearl L.H.
        • Schierz A.C.
        • Ward S.E.
        • Al-Lazikani B.
        • Pearl F.M.
        Therapeutic opportunities within the DNA damage response.
        Nat Rev Cancer. 2015; 15: 166-180
        • Casolino R.
        • Paiella S.
        • Azzolina D.
        • Beer P.A.
        • Corbo V.
        • Lorenzoni G.
        • et al.
        Homologous recombination deficiency in pancreatic cancer: a systematic review and prevalence meta-analysis.
        J Clin Oncol. 2021; 39: 2617-2631
        • Hoppe M.M.
        • Sundar R.
        • Tan D.S.P.
        • Jeyasekharan A.D.
        Biomarkers for homologous recombination deficiency in cancer.
        J Natl Cancer Inst. 2018; 110: 704-713
        • Maxwell K.N.
        • Wubbenhorst B.
        • Wenz B.M.
        • De Sloover D.
        • Pluta J.
        • Emery L.
        • et al.
        BRCA locus-specific loss of heterozygosity in germline BRCA1 and BRCA2 carriers.
        Nat Commun. 2017; 8: 319
        • Momozawa Y.
        • Sasai R.
        • Usui Y.
        • Shiraishi K.
        • Iwasaki Y.
        • Taniyama Y.
        • et al.
        Expansion of cancer risk profile for BRCA1 and BRCA2 pathogenic variants.
        JAMA Oncol. 2022 1; 8: 871-878
        • Hirata M.
        • Kamatani Y.
        • Nagai A.
        • Kiyohara Y.
        • Ninomiya T.
        • Tamakoshi A.
        • et al.
        Cross-sectional analysis of BioBank Japan clinical data: a large cohort of 200,000 patients with 47 common diseases.
        J Epidemiol. 2017; 27: S9-S21
        • Nagai A.
        • Hirata M.
        • Kamatani Y.
        • Muto K.
        • Matsuda K.
        • Kiyohara Y.
        • et al.
        Overview of the BioBank Japan project: study design and profile.
        J Epidemiol. 2017; 27: S2-S8
        • Fujita M.
        • Liu X.
        • Iwasaki Y.
        • Terao C.
        • Mizukami K.
        • Kawakami E.
        • et al.
        Population-based screening for hereditary colorectal cancer variants in Japan.
        Clin Gastroenterol Hepatol. 2022; 20: 2132-2141
        • Mizukami K.
        • Iwasaki Y.
        • Kawakami E.
        • Hirata M.
        • Kamatani Y.
        • Matsuda K.
        • et al.
        Genetic characterization of pancreatic cancer patients and prediction of carrier status of germline pathogenic variants in cancer-predisposing genes.
        EBioMedicine. 2020; 60103033
        • Buys S.S.
        • Sandbach J.F.
        • Gammon A.
        • Patel G.
        • Kidd J.
        • Brown K.L.
        • et al.
        A study of over 35,000 women with breast cancer tested with a 25-gene panel of hereditary cancer genes.
        Cancer. 2017; 123: 1721-1730
        • Momozawa Y.
        • Iwasaki Y.
        • Parsons M.T.
        • Kamatani Y.
        • Takahashi A.
        • Tamura C.
        • et al.
        Germline pathogenic variants of 11 breast cancer genes in 7,051 Japanese patients and 11,241 controls.
        Nat Commun. 2018; 9: 4083
        • Momozawa Y.
        • Iwasaki Y.
        • Hirata M.
        • Liu X.
        • Kamatani Y.
        • Takahashi A.
        • et al.
        Germline pathogenic variants in 7636 Japanese patients with prostate cancer and 12 366 controls.
        J Natl Cancer Inst. 2020; 112: 369-376
        • Fujimoto A.
        • Furuta M.
        • Shiraishi Y.
        • Gotoh K.
        • Kawakami Y.
        • Arihiro K.
        • et al.
        Whole-genome mutational landscape of liver cancers displaying biliary phenotype reveals hepatitis impact and molecular diversity.
        Nat Commun. 2015; 6: 6120
        • Fujimoto A.
        • Furuta M.
        • Totoki Y.
        • Tsunoda T.
        • Kato M.
        • Shiraishi Y.
        • et al.
        Whole-genome mutational landscape and characterization of noncoding and structural mutations in liver cancer.
        Nat Genet. 2016; 48: 500-509
        • Ebata N.
        • Fujita M.
        • Sasagawa S.
        • Maejima K.
        • Okawa Y.
        • Hatanaka Y.
        • et al.
        Molecular classification and tumor microenvironment characterization of gallbladder cancer by comprehensive genomic and transcriptomic analysis.
        Cancers (Basel). 2021; 13
        • Priestley P.
        • Baber J.
        • Lolkema M.P.
        • Steeghs N.
        • de Bruijn E.
        • Shale C.
        • et al.
        Pan-cancer whole-genome analyses of metastatic solid tumours.
        Nature. 2019; 575: 210-216
        • Okawa Y.
        • Ebata N.
        • Kim N.K.D.
        • Fujita M.
        • Maejima K.
        • Sasagawa S.
        • et al.
        Actionability evaluation of biliary tract cancer by genome transcriptome analysis and Asian cancer knowledgebase.
        Oncotarget. 2021; 12: 1540-1552
        • Nguyen L.
        • Martens J.W.M.
        • Van Hoeck A.
        • Cuppen E.
        Pan-cancer landscape of homologous recombination deficiency.
        Nat Commun. 2020; 11: 5584
        • Wang S.
        • Li H.
        • Song M.
        • Tao Z.
        • Wu T.
        • He Z.
        • et al.
        Copy number signature analysis tool and its application in prostate cancer reveals distinct mutational processes and clinical outcomes.
        PLoS Genet. 2021; 17e1009557
        • Alexandrov L.B.
        • Kim J.
        • Haradhvala N.J.
        • Huang M.N.
        • Tian Ng A.W.
        • Wu Y.
        • et al.
        The repertoire of mutational signatures in human cancer.
        Nature. 2020; 578: 94-101
        • Maynard H.
        • Stadler Z.K.
        • Berger M.F.
        • Solit D.B.
        • Ly M.
        • Lowery M.A.
        • et al.
        Germline alterations in patients with biliary tract cancers: a spectrum of significant and previously underappreciated findings.
        Cancer. 2020; 126: 1995-2002
        • Lin J.
        • Cao Y.
        • Yang X.
        • Li G.
        • Shi Y.
        • Wang D.
        • et al.
        Mutational spectrum and precision oncology for biliary tract carcinoma.
        Theranostics. 2021; 11: 4585-4598
        • Golan T.
        • Raitses-Gurevich M.
        • Kelley R.K.
        • Bocobo A.G.
        • Borgida A.
        • Shroff R.T.
        • et al.
        Overall survival and clinical characteristics of BRCA-associated cholangiocarcinoma: a multicenter retrospective study.
        Oncologist. 2017; 22: 804-810
        • Therkildsen C.
        • Jensen L.H.
        • Rasmussen M.
        • Bernstein I.
        An update on immune checkpoint therapy for the treatment of lynch syndrome.
        Clin Exp Gastroenterol. 2021; 14: 181-197
        • Wong W.
        • Lowery M.A.
        • Berger M.F.
        • Kemel Y.
        • Taylor B.
        • Zehir A.
        • et al.
        Ampullary cancer: evaluation of somatic and germline genetic alterations and association with clinical outcomes.
        Cancer. 2019; 125: 1441-1448
        • Knudsen A.L.
        • Bisgaard M.L.
        • Bülow S.
        Attenuated familial adenomatous polyposis (AFAP). A review of the literature.
        Fam Cancer. 2003; 2: 43-55
        • Dinarvand P.
        • Davaro E.P.
        • Doan J.V.
        • Ising M.E.
        • Evans N.R.
        • Phillips N.J.
        • et al.
        Familial adenomatous polyposis syndrome: an update and review of extraintestinal manifestations.
        Arch Pathol Lab Med. 2019; 143: 1382-1398
        • Telli M.L.
        • Jensen K.C.
        • Vinayak S.
        • Kurian A.W.
        • Lipson J.A.
        • Flaherty P.J.
        • et al.
        Phase II study of gemcitabine, carboplatin, and iniparib as neoadjuvant therapy for triple-negative and BRCA1/2 mutation-associated breast cancer with assessment of a tumor-based measure of genomic instability: PrECOG 0105.
        J Clin Oncol. 2015; 33: 1895-1901
        • Coleman R.L.
        • Oza A.M.
        • Lorusso D.
        • Aghajanian C.
        • Oaknin A.
        • Dean A.
        • et al.
        Rucaparib maintenance treatment for recurrent ovarian carcinoma after response to platinum therapy (ARIEL3): a randomised, double-blind, placebo-controlled, phase 3 trial.
        Lancet. 2017; 390: 1949-1961
        • de Bono J.
        • Mateo J.
        • Fizazi K.
        • Saad F.
        • Shore N.
        • Sandhu S.
        • et al.
        Olaparib for metastatic castration-resistant prostate cancer.
        N Engl J Med. 2020; 382: 2091-2102
        • Osorio A.
        • de la Hoya M.
        • Rodríguez-López R.
        • Martínez-Ramírez A.
        • Cazorla A.
        • Granizo J.J.
        • et al.
        Loss of heterozygosity analysis at the BRCA loci in tumor samples from patients with familial breast cancer.
        Int J Cancer. 2002; 99: 305-309
        • Abbotts R.
        • Topper M.J.
        • Biondi C.
        • Fontaine D.
        • Goswami R.
        • Stojanovic L.
        • et al.
        DNA methyltransferase inhibitors induce a BRCAness phenotype that sensitizes NSCLC to PARP inhibitor and ionizing radiation.
        Proc Natl Acad Sci USA. 2019; 116: 22609-22618
        • Park W.
        • Chen J.
        • Chou J.F.
        • Varghese A.M.
        • Yu K.H.
        • Wong W.
        • et al.
        Genomic methods identify homologous recombination deficiency in pancreas adenocarcinoma and optimize treatment selection.
        Clin Cancer Res. 2020; 26: 3239-3247
        • Miller D.T.
        • Lee K.
        • Chung W.K.
        • Gordon A.S.
        • Herman G.E.
        • Klein T.E.
        • et al.
        ACMG SF v3.0 list for reporting of secondary findings in clinical exome and genome sequencing: a policy statement of the American College of Medical Genetics and Genomics (ACMG).
        Genet Med. 2021; 23: 1381-1390