Skip to main content
PMC home page
JNCI Journal of the National Cancer Institute logoLink to JNCI Journal of the National Cancer Institute
. 2022 Oct 31;115(2):181–189. doi: 10.1093/jnci/djac196

Somatic inactivation of breast cancer predisposition genes in tumors associated with pathogenic germline variants

Belle W X Lim 1,2,#, Na Li 3,4,5, Sakshi Mahale 6, Simone McInerny 7, Magnus Zethoven 8,9, Simone M Rowley 10, Joanne Huynh 11, Theresa Wang 12,13, Jue Er Amanda Lee 14,15,16, Mia Friedman 17,18, Lisa Devereux 19,20, Rodney J Scott 21,22, Erica K Sloan 23,24, Paul A James 25,26,#, Ian G Campbell 27,28,29,✉,#
PMCID: PMC9905963  PMID: 36315097

Abstract

Background

Breast cancers (BCs) that arise in individuals heterozygous for a germline pathogenic variant in a susceptibility gene, such as BRCA1 and BRCA2, PALB2, and RAD51C, have been shown to exhibit biallelic loss in the respective genes and be associated with triple-negative breast cancer (TNBC) and distinctive somatic mutational signatures. Tumor sequencing thus presents an orthogonal approach to assess the role of candidate genes in BC development.

Methods

Exome sequencing was performed on paired normal-breast tumor DNA from 124 carriers of germline loss-of-function (LoF) or missense variant carriers in 15 known and candidate BC predisposition genes identified in the BEACCON case-control study. Biallelic inactivation and association with tumor genome features including mutational signatures and homologous recombination deficiency (HRD) score were investigated.

Results

BARD1-carrying TNBC (4 of 5) displayed biallelic loss and associated high HRD scores and mutational signature 3, as did a RAD51D-carrying TNBC and ovarian cancer. Biallelic loss was less frequent in BRIP1 BCs (4 of 13) and had low HRD scores. In contrast to other established BC genes, BCs from carriers of CHEK2 LoF (6 of 17) or missense (2 of 20) variant had low rates of biallelic loss. Exploratory analysis of BC from carriers of LoF variants in candidate genes such as BLM, FANCM, PARP2, and RAD50 found little evidence of biallelic inactivation.

Conclusions

BARD1 and RAD51D behave as classic BRCA-like predisposition genes with biallelic inactivation, but this was not observed for any of the candidate genes. However, as demonstrated for CHEK2, the absence of biallelic inactivation does not provide definitive evidence against the gene’s involvement in BC predisposition.

Hereditary breast cancer (HBC) often clusters within families and can be attributed to germline variants in susceptibility genes directly or indirectly involved in DNA repair. The major contributors—BRCA1, BRCA2, and PALB2 (1,2)—collectively explain less than half of the familial aggregation of BC (3). Exploratory case-control studies in the past have found that potentially pathogenic variants in individual candidate genes are rare (3-5), precluding any confident conclusion about their role in HBC based solely on this approach.

An orthogonal approach to assess if a candidate gene is driving tumorigenesis is through genomic analysis of the cancers from carriers of germline mutations. For example, approximately 90% of BRCA1 and 50%-60% of BRCA2 breast tumors from germline mutation carriers have a somatic “second-hit” (6-9), resulting in biallelic inactivation. Most commonly, this occurs through loss of heterozygosity (LOH) or, less frequently, through protein truncating somatic point mutations or promoter hypermethylation. Biallelic inactivation of genes such as BRCA1 and BRCA2 is almost invariably associated with specific somatic mutational signatures (10). The presence or absence of these tumor genomic features can provide strong evidence for or against a gene’s cancer predisposition role, even if based on relatively few cancers as previously demonstrated for PALB2, RAD51C, and ATM (11-13). Recent large case-control studies involving more than 65 000 participants each confirmed the association of moderate risk genes RAD51C, RAD51D, and BARD1 but not BRIP1 with breast cancer (4,5).

In this study, we extend the tumor sequencing approach by performing exome sequencing on 124 BCs from individuals harboring germline variants in proposed and candidate HBC genes identified in the BEACCON case-control study (hereditary BrEAst Case CONtrol study) (3) to look for evidence of biallelic inactivation as a means of validating the role of these genes in BC predisposition.

Methods

Case-control study and tumor sequencing

A total of 124 unique breast tumors were selected from cancers arising in individuals with a germline loss-of-function (LoF) or rare, likely pathogenic missense (MS) variant of interest in a known (BARD1, BRIP1, CHEK2, and RAD51D) or candidate (BLM, CDK9, CTH, ERCC5, FANCM, MUTYH, PARP2, RAD50, RAD51B, WRN, and XRCC2) BC predisposition gene detected in the BEACCON case-control study (3). LoF variants included stop-gained, frame-shift, or essential splice-site variants, and MS variants of interest were identified based on a combination of population frequency, in silico prediction, and location in key functional domains as detailed in Supplementary Tables 1-4 (available online). Two ovarian cancers from carriers of BRIP1 and 1 of RAD51D LoF variants, respectively, were also included as these genes are known to be ovarian cancer predisposing genes. Since last reporting (5), this study has been expanded to include 6689 BRCA-negative female index familial BC patients and 14 381 cancer-free female participants (Supplementary Table 5, available online). Candidate genes were selected for this analysis based on an excess of rare coding variants in the case group. Microdissection, DNA extraction, and exome sequencing are described in the Supplementary Methods (available online). Tumor characteristics and personal and family history of the individuals selected for the current study are summarized in Supplementary Table 6 (available online).

Determination of variant allelic status and potential biallelic inactivation

For each tumor, the somatic sequencing data were assessed for the presence of somatic LoF or MS point mutations in the gene of interest as well as the allelic status of the germline variant as described previously (11,12). In summary, locus-specific LOH was determined by tumor variant allele frequency comparisons as adjusted according to estimated tumor purity. All cases had matched germline sequencing data for the gene of interest. Promoter hypermethylation, using targeted Twist Custom Panel methylation sequencing or Sanger sequencing on bisulfite converted DNA, was also assessed for cases where there was no somatic mutation or LOH across the gene of interest. Homologous recombination deficiency (HRD) scores were calculated for each tumor sample using copy number plots as a sum of the occurrence of telomeric allelic imbalances, large-scale state transitions, and homologous recombination deficiency–loss of heterozygosity from copy number plots as described previously (12), where a threshold of an HRD score of 42 or higher is defined as high-HRD (14,15). Mutational signatures were generated against COSMIC v2 catalogue (https://cancer.sanger.ac.uk/signatures/signatures_v2/) using the DeconstructSig package in R (16) on whole-exome sequenced samples.

Statistical analyses

Odds ratios and Fisher exact test (2-sided) were calculated in case-control analyses, with a 2-tailed P value of .05 or less defined as statistically significant. Confidence intervals (CIs) were calculated using a conditional maximum likelihood estimate. All calculations were carried out using R-in built function in R 3.3.2 (17).

Results

Tumor sequencing in individuals harboring a germline variant in known BC genes

Whole or targeted exome sequencing was performed on 41 tumors from individuals harboring germline LoF variants in genes commonly present in HBC panels: BARD1 (n = 7), BRIP1 (n = 13), RAD51D (n = 4), and CHEK2 (n = 17) (Table 1). BARD1, which in the BEACCON case-control data (Figure 1) and other published data (4) is associated specifically with triple-negative breast cancer (TNBC), showed loss of the wild-type (WT) allele via LOH in 4 of 5 assessable triple negative (TN) tumors. A sixth TN tumor also had LOH, but it was not possible to determine which allele had been lost, and the only BARD1 tumor to show loss of the mutant allele was estrogen receptor (ER) positive. The 6 BARD1 TN tumors including 1 in heterozygous status showed high HRD scores and 4 with strong HRD-related mutational signature 3. Carriers of RAD51D LoF mutations, which are also associated with TNBC, were rare in the BEACCON study. One of the 2 TNBC showed LOH of the WT allele, whereas 2 ER-positive tumors remained heterozygous. An additional high-grade serous ovarian cancer (HGSOC) that was available for analysis (Supplementary Table 5, available online) showed biallelic inactivation through LOH. Nine carriers of rare RAD51D MS variants shortlisted based on likely pathogenicity assessment (Supplementary Table 1, available online) were also analyzed, but only 2 cases showed loss of the WT allele with only 1 of these being a TNBC. This case was a compound heterozygote that showed loss of the p.Ala313Val and retention of the p.Ala52Val allele; it had a high HRD score and a strong mutational signature 3.

Table 1.

Tumor sequencing data for 69 tumors from individuals heterozygous for a germline LoF (n = 41) or MS (n = 29) variant in a known breast cancer predisposition gene (BARD1, BRIP1, CHEK2, and RAD51D)

IDa Gene Germline variant Variant type Variant allelic statusb BC subtype HRD Somatic TP53 Somatic PIK3CA Mutation signature 3c Dominant mutation signature Promoter hypermethylation
3530 BARD1 c.1135A>T, p.Lys379Ter LoF Mutant loss ER+/HER2- na na na na na Failed
3977 BARD1 c.1212C>G, p.Tyr404Ter LoF WT loss TN 50 na na Strong 3, 11 DNT
1531 BARD1 c.1652C>G, p.Ser551Ter LoF WT loss TN 83 LoF — Weak 19, 30 DNT
3828 BARD1 c.1652C>G, p.Ser551Ter LoF WT loss TN 92 MS — Strong 1, 3 DNT
425 BARD1 c.1652C>G, p.Ser551Ter LoF LOH TN 80 na na Strong na
3496 BARD1 c.1905G>A, p.Trp635Ter LoF Het TN 82 MS — Strong 3 Failed
1272 BARD1 c.2078_2079insTAATA, p.Lys693AsnfsTer23 LoF WT loss TN 76 LoF — No 19 DNT
2439 BRIP1 c.93 + 1G>T LoF WT loss TN 59 LoF — Strong 3, 19 DNT
4160 BRIP1 c.103G>T, p.Gly35Ter LoF Mutant loss ER-/HER2+ 30 MS — Strong 1, 3 na
3259 BRIP1 c.1426del, p.Thr476LeufsTer50 LoF WT loss ER+/HER2- 41 — MS No 12, 20 DNT
3597 BRIP1 c.1888dup, p.Thr630AsnfsTer9 LoF Het ER+/HER2- 12 — MS No 30 Negative
3093 BRIP1 c.2298_2301delTGAG, p.Ser766ArgfsTer14 LoF Het ER+/HER2- 5 — MS na na na
227 BRIP1 c.2392C>T, p.Arg798Ter LoF WT loss TN 58 MS — No 20, 21 DNT
1325 BRIP1 c.2392C>T, p.Arg798Ter LoF Mutant loss TN 57 LoF — na na na
786 BRIP1 c.2392C>T, p.Arg798Ter LoF Het ER+/HER2- 1 LoF — na na Failed
1928 BRIP1 c.2392C>T, p.Arg798Ter LoF Het ER+/HER2- 11 MS — na na Negative
3829 BRIP1 c.2400C>G, p.Tyr800Ter LoF Mutant loss TN 50 — — Weak 6 na
3635 BRIP1 c.2400C>G, p.Tyr800Ter LoF Het ER+/HER2- 32 LoF — Weak 1 na
3354 BRIP1 c.2492_2492 + 5delGGTAAG LoF WT loss ER+/HER2- 31 MS MS Weak 1 DNT
3468 BRIP1 c.3715del, p.Ser1239ProfsTer15 LoF Mutant loss ER+/HER2- 37 LoF — Weak 1, 13 na
4152 CHEK2 c.629_732delCAGT, p.Ser210PhefsTer6 LoF Mutant loss ER+/HER2- 47 MS — No 3 Negative
2320 CHEK2 c.630delA, p.Val211PhefsTer6 LoF Het na 7 — — Weak 5, 30 na
290 CHEK2 c.902delT, p.Leu301TrpfsTer3 LoF Mutant loss ER+/HER2- 27 — — Weak 1,6 na
3587 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF WT loss ER+/HER2- 34 na na No 11 DNT
1825 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER+/HER2- 32 LoF — Strong 3 na
3174 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER+/HER2- 29 — — No 6, 30 na
2182 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER+/HER2- 45 — — No 11, 19 Failed
2410 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER+/HER2- 27 — — No 1, 30 Failed
2475 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF WT loss ER+/HER2- 80 — — No 6, 19 DNT
1300 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF WT loss ER+/HER2- 36 — MS No 10 Failed
2326 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF WT loss ER+/HER2- 21 — — No 19, 30 Failed
2711 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF WT loss ER+/HER2+ 10 — — No 1, 11 Failed
3500 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER+/HER2unknown 37 — — Strong 3 na
2351 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER-/HER2+ 4 na na Weak 19, 30 DNT
3076 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF Het ER-/HER2+ 1 — — No 1 na
1732 CHEK2 c.1100delC, p.Thr367MetfsTer15 LoF WT Loss TN 17 MS MS No 6 DNT
1853 CHEK2 c.1696delC, p.Thr533GlnfsTer33 LoF Het ER+/HER2- 28 MS — na na na
2625 CHEK2 c.14C>T, p.Ser5Leu MS Het ER+/HER2- 85 — — Weak 19 Failed
1993 CHEK2 c.190G>A, p.Glu64Lys MS Het ER+/HER2- 91 LoF — Weak 1, 6 na
811 CHEK2 c.349A>G, p.Arg117Gly MS Het ER+/HER2+ 0 — MS Weak 5 Negative
1103 CHEK2 c.349A>G, p.Arg117Gly MS Mutant loss ER+/HER2- 16 — MS Weak 3 Negative
616 CHEK2 c.349A>G, p.Arg117Gly MS Het ER-/HER2+ 17 — MS Strong 1, 3 na
787 CHEK2 c.442A>G, p.Arg148Gly MS Het ER+/HER2- na na na na na Failed
2531 CHEK2 c.470T>C, p.Ile157Thr MS Het ER+/HER2- 34 — — Weak 19 Negative
1420 CHEK2 c.470T>C, p.Ile157Thr MS Het ER+/HER2- 28 MS MS No 6 Negative
3240 CHEK2 c.470T>C, p.Ile157Thr MS Het ER+/HER2- 16 — — Strong 3, 6 na
807 CHEK2 c.1036C>T, p.Arg346Cys; c.499G>A, p. Gly167Arg MS
MS		 WT loss; mutant loss ER+/HER2unknown 52 — — Weak 1 na
2091 CHEK2 c.1067C>T, p.Ser356Leu MS Het ER+/HER2- 47 — — Weak 1, 19 Failed
2689 CHEK2 c.1211A>G, p.Tyr404Cys MS WT loss ER+/HER2+ 28 — MS Strong 3, 30 DNT
2221 CHEK2 c.1270T>C, p.Tyr424His MS WT loss ER+/HER2- 13 — MS No 20 DNT
1830 CHEK2 c.1312G>T, p.Asp438Tyr MS Het ER+/HER2- 17 — — na na na
2257 CHEK2 c.1312G>T, p.Asp438Tyr MS Het ER+/HER2- 0 — MS Weak 6 Negative
200 CHEK2 c.1312G>T, p.Asp438Tyr MS Het na 5 — — Strong 3 na
4164 CHEK2 c.1427C>T, p.Thr476Met MS Het ER+/HER2- 13 — — No 11, 19 na
1410 CHEK2 c.1447C>T, p.His483Tyr MS Mutant loss ER+/HER2- 30 MS — na na DNT
2345 CHEK2 c.1525C>T, p.Pro509Ser MS Mutant loss ER+/HER2+ 7 LoF — na na Negative
1198 CHEK2 c.1604G>A, p.Arg535His MS Het TN 22 MS — No 6, 30 Failed
1897 RAD51D c.616C>T, p.Arg206Ter LoF Het ER+/HER2- 4 — — na na Negative
2734 RAD51D c.754C>T, p.Arg252Ter LoF WT loss TN 42 LoF — na na DNT
2866 RAD51D c.808delC, p.His270ThrfsTer2 LoF Het TN 7 — — na na na
3500 RAD51D c.863G>A, p.Trp288Ter LoF Het ER+/HER2- 8 — — Weak 30 na
506 RAD51D c.26G>C, p.Cys9Ser MS Het ER+/HER2- 0 — — na na na
30 RAD51D c.26G>C, p.Cys9Ser MS Het ER+/HER2- 20 MS — na na na
2936 RAD51D c.26G>C, p.Cys9Ser MS Het ER+/HER2- 8 — — No 5 DNT
1980 RAD51D c.26G>C, p.Cys9Ser MS Het TN 3 LoF — No 1 Negative
2224 RAD51D c.137C>G, p.Ser46Cys MS Het ER+/HER2- 4 — — na na Negative
2219 RAD51D c.155C>T, p.Ala52Val; c.938C>T, p. Ala313Val MS compound homozygous WT loss; mutant loss TN 42 LoF — Strong 3 Negative
1686 RAD51D c.308C>T, p.Ala103Val MS Het TN 17 MS — na na Negative
3095 RAD51D c.472A>C, p.Asn158His MS Mutant Loss ER+/HER2+ 20 — — na na na
2606 RAD51D c.551T>C, p.Leu184Pro MS WT Loss ER+/HER2- 11 MS — na na DNT
Open in a new tab
a

Subject 3093 carried 2 variants of interest in BRIP1; subject 2219 carried 2 variants in RAD51C; subject 3500 carried a variant of interest in both CHEK2 and RAD51D. “—” signifies feature not present. BC = breast cancer; DNT = did not test; na = not available; ER+ = estrogen receptor—positive breast cancer; HER2- = HER2 negative; HER2+ = HER2 positive; HRD = homologous recombination deficiency score; LoF = loss of function; MS = missense; TN = triple-negative; WT = wild type.

b

WT loss, somatic loss of the WT allele; mutant loss, somatic loss of the allele carrying the known germline variant; Het, heterozygous. Case 425 showed loss of heterozygosity across the gene regions but unable to determine which allele had been lost.

c

Proportion of mutational signature 3 (COSMIC v2, assessed on whole-exome sequenced tumors only) above 25% is classified as “strong”, under 25% as “weak”.

Figure 1.

Figure 1.

Open in a new tab

Case-control analysis of rare LoF variants (minor allele frequency [MAF] ≤ 0.005) and MS variants (MAF ≤ 0.001) in known or strongly proposed breast cancer genes, including subcategories of estrogen receptor–positive (ER+), ER-negative (ER-), and triple-negative (TN) breast tumor where diagnosis was available. ER+ and ER- groups were mutually exclusive, and the ER- groups include the TN samples. Participants without sufficient pathological information were only included in the overall LoF group and excluded from the subcategory analysis. CHEK2, BARD1, and BRIP1 were screened in 6689 cases and 14 381 controls; RAD51C and RAD51D were screened in 5726 cases and 13 428 controls. The sample sizes of ER+, ER-, and TN were 2146, 1246, and 862, respectively. CI = confidence interval; LoF = loss of function; MS = missense; OR = odds ratio (3,4,11,12,18,19).

The role of BRIP1 in breast cancer predisposition is debated, and our analysis of BRIP1 LoF variants identified an equal number of tumors showing loss of the WT or mutant alleles (4 cases each) with the remaining 5 remaining heterozygous. By comparison, analysis of 2 HGSOCs showed both had loss of the WT (Supplementary Table 5, available online), consistent with the established role of BRIP1 in ovarian cancer predisposition. Both HGSOCs also showed high HRD scores compared with only 2 of 4 BRIP1-null BCs. Mutational signature 3 was observed only in 1 BRIP1-null BC and not in the HGSOCs.

LoF mutations in CHEK2, predominantly the c.1100delC variant, are well established to confer a two- to threefold increase in BC risk (20), with the association being strongest for ER-positive BC. The current tumor data do not provide evidence that CHEK2 requires biallelic inactivation with the majority (9 of 17) of tumors remaining heterozygous, whereas only 6 showed loss of the WT allele, and 2 ER-positive tumors showed loss of the LoF allele. It appears that the 2 tumors with loss of pathogenic variants were not driven by CHEK2 LoF, and given that CHEK2 is only a moderate risk gene, several of the tumors without LOH could also not be driven by the CHEK2 pathogenic variants. In addition, 20 tumors from individuals with rare germline CHEK2 MS variants of interest (Table 1; Supplementary Table 2, available online) were analyzed. Most of these variants except for p.Ile157Thr and p.Arg117Gly are currently classified as variants of unknown significance, and tumor sequencing showed that most retained heterozygosity with only 2 showing loss of the WT allele and 2 showing loss of the variant allele. In particular, all 3 carriers of the known pathogenic, but reduced penetrance, variant CHEK2 p.Ile157Thr retained heterozygosity, and in the 2 tumors that were tested, neither were found to have promoter methylation.

Tumor sequencing in individuals with germline variants in candidate HBC genes

Tumor sequencing was performed on 57 BCs (Table 2) carrying LoF (n = 45) or rare MS (n = 16) variants in 11 genes that have been the subject of debate in the literature (FANCM, RAD50, RAD51B, and XRCC2) or were shortlisted from the BEACCON study (BLM, CDK9, CTH, ERCC5, MUTYH, PARP2, and WRN) (Supplementary Figure 1, available online). The 4 BCs from RAD50 LoF mutation carriers remained heterozygous, consistent with the accumulating literature that it does not predispose to BC (21). Literature support for a role of XRCC2 in BC predisposition is weak but with a potential association with ER-negative cancer (4,22,23). Of the 2 BCs from XRCC2 LoF variant carriers, only 1 (a TNBC) showed biallelic inactivation with both a high HRD score and mutational signature 3. For RAD51B, only tumors from carriers of rare MS variants were available with 2 remaining heterozygous and 2 showing loss of the variant allele.

Table 2.

Sequencing results of 57 tumors from individuals heterozygous for a germline LoF (n = 43) or MS (n = 16) variant in a candidate breast cancer predisposition gene (BLM, CDK9, CTH, ERCC5, FANCM, MUTYH, PARP2, RAD50, RAD51B, WRN, and PARP2)

IDa Germline gene Variant Variant type Variant statusb Subtype HRD Somatic TP53 Somatic PIK3CA Mutation signature 3c Dominant signature Hypermutated
2660 BLM c.318_319insT, p.Leu107PhefsTer36 LoF Het ER+/HER2- 1 — — na na Negative
1471 c.768_769delCT, p.Leu258GlufsTer7 LoF Mutant Loss ER+/HER2unknown 5 — MS na na Negative
462 c.1624delG, p.Asp542ThrfsTer2 LoF Mutant Loss TN 44 LoF — na na Negative
3093 c.2695C>T, p.Arg899Ter LoF Het ER+/HER2- 5 — MS na na na
2287 c.2875C>T, p.Arg959Ter LoF Het ER+/HER2- 11 — — na na Negative
2083 c.3210 + 2delT LoF Het ER+/HER2- 6 — — na na Failed
35 c.3558 + 1G>T LoF Het ER+/HER2unknown 17 — MS na na Negative
1245 CDK9 c.130delA, p.Lys44ArgfsTer4 LoF Het ER+/HER2- 13 MS — na na Failed
2345 c.274delT, p.Tyr92IlefsTer23 LoF Het ER+/HER2+ 4 LoF — na na Negative
3075 c.620_621insC, p.Ile210HisfsTer2 LoF Het TN 56 LoF — na na Failed
2723 c.689_690insG, p.Asn232GlnfsTer20 LoF Het ER+/HER2+ 0 — — na na na
2045 CTH c.465G>A, p.Trp155Ter LoF Het ER+/HER2- 19 — — na na Failed
1322 c.465G>A, p.Trp155Ter LoF Het ER+/HER2+ 20 MS — na na Failed
3119 c.465G>A, p.Trp155Ter LoF Het ER+/HER2+ 0 — — na na na
1865 c.1064delC, p.Thr355IlefsTer19 LoF Het ER+/HER2- 36 MS — na na Failed
1092 c.230C>T, p.Ala77Val MS Het ER+/HER2- 20 ESS — No 5, 12 Failed
307 c.323T>C, p.Ile108Thr MS Het TN 24 LoF — na na Failed
2593 c.620T>C, p.Met207Thr MS Het na 9 — — na na na
1065 c.718C>G, p.Gln240Glu MS Het ER+/HER2- 0 — — na na na
4142 c.794G>A, p.Arg265Gln MS Het ER+/HER2- 10 LoF — na na Failed
3197 c.1124G>A, p.Arg375Gln MS Het ER+/HER2- 19 — — Weak 1, 12 na
2260 ERCC5 c.589delC, p.Pro198LeufsTer3 LoF Mutant Loss TN 41 — — Weak 15 Failed
901 c.1774_1775insAAGCA, p.Val592GlufsTer8 LoF Mutant Loss TN 83 — MS No 11 na
1367 FANCM c.2267G>A, p.Arg756His LoF Het ER+/HER2- 24 — MS No 6 Failed
1709 c.3589delG, p.Asp1197MetfsTer18 LoF Mutant Loss TN 26 MS — Weak 6 Failed
3147 c.5101C>T, p.Gln1701Ter LoF Het ER+/HER2- 6 — — Weak 3, 6 Failed
691 c.5791C>T, p.Arg1931Ter LoF Het ER+/HER2- 2 MS MS No 6 na
1172 c.5791C>T, p.Arg1931Ter LoF Het ER+/HER2+ 0 — — No 1 Failed
2771 c.5791C>T, p.Arg1931Ter LoF WT Loss TN 18 LoF — No 1, 26 DNT
1127 c.163G>A, p.Asp55Asn MS Mutant Loss TN 51 — — Weak 3 Failed
2094 c.2267G>A, p.Arg756His MS Het TN 40 MS — Strong 1, 3 Failed
1879 c.2267G>A, p.Arg756His MS cpd Het ER+/HER2- 72 LoF — Weak 1, 19 DNT
1222 c.3998A>C, p.Gln1333Pro MS WT Loss TN 48 LoF — Weak 5 DNT
901 c.5108A>G, p.His1703Arg MS Het TN 83 — MS No 5 na
2743 MUTYH c.925-2A>G LoF Het TN 10 MS — na na Failed
2727‡ c.925-2A>G LoF Het TN 36 LoF — na na Failed
1253 c.384G>A, p.Trp128Ter Biallelic LoF Germline homozygous TN na — — na na na
1474 PARP2 c.979_980insTT, p.Ser328CysfsTer8 LoF Mutant Loss ER+/HER2- 67 — MS na na DNT
2294 c.985_986insA, p.Ile331AsnfsTer11 LoF Het ER-/HER2+ 81 — — na na Failed
333 c.1109_1110insT, p.Leu372ProfsTer2 LoF Mutant Loss ER+/HER2- na — — na na na
1185 c.1304delG, p.Val436TrpfsTer4 LoF Het ER-/HER2+ 30 MS MS na na Failed
1327 c.965G>A, p.Arg322Gln MS WT Loss ER+/HER2- 28 — — na na DNT
2883 RAD50 c.1291_1297delGAGATAA, p.Asp434LysfsTer7 LoF Het ER+/HER2- 47 — — No 16 Failed
2193 c.1958C>A, p.Ser653Ter LoF Het ER+/HER2- 23 — — Strong 3 Failed
2251 c.2467C>T, p.Arg823Ter LoF Het ER+/HER2- 24 — MS Strong 3, 5 na
1031 c.3207delA, p.Asn1070IlefsTer6 LoF Het ER+/HER2+ 8 — MS No 25 Failed
2923 RAD51B c.103C>T, p.Pro35Ser MS Het TN 32 MS — na na na
1932 c.277G>A, p.Ala93Thr MS Het TN 2 — — na na Failed
1795 c.436G>A, p.Ala146Thr MS Mutant Loss TN 30 LoF — na na na
3024 c.553T>G, p.Cys185Gly MS Mutant Loss TN 76 — — na na DNT
3054 WRN c.171C>A, p.Tyr57Ter LoF Het ER+/HER2- 24 — MS Weak 5 Failed
2963 c.944_948delTAAAC, p.Leu315PhefsTer5 LoF Het ER+/HER2- 24 — — Weak 5 Failed
1115 c.3961C>T, p.Arg1321Ter LoF Het ER+/HER2- 40 MS MS No 5, 6 Failed
1847 c.4216C>T, p.Arg1406Ter LoF WT Loss ER+/HER2- 53 — MS Strong 3 DNT
2562 c.4216C>T, p.Arg1406Ter LoF Het ER+/HER2- 29 — — Strong 3, 11 Failed
3093 c.4216C>T, p.Arg1406Ter LoF WT Loss ER+/HER2- 5 — MS na na na
1349 c.4216C>T, p.Arg1406Ter LoF Het TN 43 — — Strong 3 Failed
863 XRCC2 c.39 + 1G>A LoF Het ER-/HER2+ 12 LoF — Strong 3 Failed
3062 c.794T>A, p.Leu265Ter LoF WT Loss TN 44 LoF — Strong 3 DNT
Open in a new tab
a

Subject 2727 also carried a germline ATM variant that had biallelic loss in tumor. Subject 3054 carried a germline RAD51C that had biallelic loss in tumor. Subject 3093 carried variants of interest in BLM and WRN. Subject 901 carried variants in both ERCC5 and FANCM. “—” signifies feature not present. BC = breast cancer; DNT = did not test; na = not available; ER+ = estrogen receptor–positive breast cancer; HER2- = HER2 negative; HER2+ = HER2 positive; HRD = homologous recombination deficiency score; LoF = loss of function; MS = missense; TN = triple-negative; WT = wild type.

b

WT loss, somatic loss of the wild-type allele; Mutant loss, somatic loss of the allele carrying the known germline variant, Het, heterozygous. Case 425 showed loss of heterozygosity across the gene regions but unable to determine which allele had been lost.

c

Proportion of mutational signature 3 (COSMIC v2, assessed on whole-exome sequenced tumors only) above 25% is classified as “strong,” below 25% as “weak.”

LoF mutations in FANCM have previously been reported to be associated with a small increase in BC risk (24-26), but 4 of 6 tumors from individuals with FANCM LoF variants remained heterozygous with only 1 having lost the WT allele and did not show a high HRD score. Similarly, there was no consistent loss of the WT allele in tumors associated with germline FANCM MS variants (Supplementary Table 3, available online). BLM has previously been implicated in BC predisposition (27-29) and in the BEACCON data (Supplementary Figure 1, available online), but 5 of the BCs with LoF variants remained heterozygous with no evidence of promoter hypermethylation, whereas 2 BCs lost the LoF allele.

For the candidate genes, loss of the WT allele was not observed for the majority. The BCs carrying LoF or MS variants in the candidate genes CDK9 and CTH (Supplementary Table 4, available online) remained heterozygous although promoter methylation assessment was not successful for these genes. Of the 4 PARP2 BCs with LoF mutations, 2 showed loss of the LoF allele with the other 2 remaining heterozygous. The 2 ERCC5 BCs showed loss of the LoF allele. In contrast, the WRN gene, which was found to have a statistcally significant association with ER-positive BC in the BEACCON study (unadjusted P = .003), showed loss of the WT allele in 2 of 6 ER-positive cases. A WRN-heterozygous BC also carried an ATM germline variant that had experienced biallelic loss, suggesting that the ATM variant was instead responsible for the tumor. Overall, evidence of loss of the WT allele was rare among candidate genes despite evidence in case-control frequencies.

Discussion

The frequent observation of loss of the WT allele in BCs carrying germline pathogenic mutations in BRCA1 and BRCA2 supports the model of biallelic inactivation being required for BC predisposition, at least in some high penetrance genes. Recent studies indicate that biallelic inactivation is also common in BCs carrying pathogenic mutations in PALB2 (12,30,31) and ATM (13) and has been used as a biomarker to support the role of RAD51C (11) as a TNBC predisposition gene. These examples suggest tumor sequencing can provide a useful orthogonal approach to validate new BC genes and rare MS variants in known genes.

In this study, tumor sequencing demonstrated that TN tumors from BARD1 LoF mutation carriers frequently exhibit biallelic inactivation consistent with data from case-control studies (3-5) that indicate that BARD1 pathogenic variants are associated with predisposition to TNBC. Recent case-control studies have also provided support for the role of RAD51D in TNBC predisposition (4). Although 1 of 2 LoF BCs studied here showed loss of the WT allele, the rarity of its variants precluded any confident conclusion to be drawn. RAD51D MS variants as a group showed an excess in the BEACCON case-control analysis, but most of the tumors from rare MS variant carriers showed no evidence of a second hit, and the 2 cases with loss of the WT allele had low HRD scores suggesting they are benign variants.

The role of BRIP1 in BC predisposition is debated with most, but not all, published case-control studies failing to identify a statistically significant excess of LoF mutations in cases (4,5). Tumor sequencing did not find evidence to support a role for BRIP1 in BC predisposition with most BCs remaining heterozygous and, importantly, an equivalent number of cases losing the WT and LoF alleles. Previous tumor sequencing studies on BRIP1 BCs are limited, but our findings are consistent with a previous report on 3 BRIP1 BC where only 1 was found to have biallelic inactivation (10). Overall, our results for BARD1, RAD51D, and BRIP1 are consistent with the findings of 2 recent large case-control studies cited previously (4,5) where BARD1 and RAD51D are associated with BC, specifically TNBC, whereas no causative link was identified for BRIP1.

Based on the data for BARD1 and the other previously studied BC predisposing genes BRCA1, BRCA2, PALB2, and ATM, it might be extrapolated that biallelic inactivation is a typical feature for all BC predisposition genes. However, the data for CHEK2, which has highly robust case-control evidence supporting its role as a moderate penetrance BC gene, suggest that this is not true in all cases. Of the 17 BCs with germline CHEK2 LoF mutations, only 6 showed loss of the WT allele, and the majority (53%) showed no evidence of biallelic loss. The established low penetrance CHEK2 variant p.Ile157Thr was also not detected with WT allele loss. This is consistent with previous studies that found that LOH across CHEK2 in BCs from LoF mutation carriers was infrequent (32 of 93, 34%) (32-37) and occurs at a similar rate in sporadic BCs (40%, n = 560) (38). A recent sequencing-based study reported that 13 of 16 (81%) BC from CHEK2 LoF carriers had biallelic inactivation (32), however, only 5 of the 8 (63%) CHEK2-null tumors were of ER-positive and HER2-negative ductal histological subtype that are known to be associated with CHEK2 predisposition. Our data based on 17 LoF and 20 MS CHEK2 variant-carrying tumors suggested that although CHEK2 displayed a selective predisposition to ER-positive ductal BC, there was no consistent biallelic inactivation, and the BCs had low mutational burden and were not consistently associated with a characteristic mutational signature or somatic driver mutations. These data suggest that the effect of pathogenic variants in CHEK2 is possibly mediated by haploinsufficiency (39), which has implications for the reliability of using biallelic inactivation as an indicator of disease association.

Candidate genes analyzed in this study were identified in the BEACCON case-control and included BLM, PARP2, and WRN, which showed statistically significant association with BC with odds ratios of 2.5, 5.0, and 2.0, respectively, whereas other candidate genes CDK9, CTH, and XRCC2 also showed relatively high odds ratios despite the small number of cases. The BEACCON study included more than 11 500 subjects with enrichment for high-risk familial cases, however, LoF variants in candidate genes were still too rare to confidently assert a genuine association with BC predisposition. Inclusion of tumor sequencing from 57 cases did not provide definitive evidence for their roles in BC predisposition: 1 of 2 XRCC2 and 2 of 7 WRN BCs were among the minority that showed loss of the WT allele. Despite multiple studies reporting an association with BLM, no instance of biallelic inactivation has been found in BLM-carrying tumors to date, including our study of 7 tumors and the previous evaluation of 22 cases across 3 studies (28,34,40). Despite early studies and inclusion of RAD50 in many HBC gene panels (41,42), recent large studies have demonstrated that it is not a BC predisposition gene (4,5), consistent with the findings in the 4 RAD50 tumors analyzed here, which all remained heterozygous.

Findings from this study have demonstrated that tumor sequencing is useful in validating BC predisposing genes that operate in carcinogenesis through a mechanism of biallelic inactivation, such as BARD1. However, based on the evidence from CHEK2-associated tumors, the absence of biallelic inactivation does not appear to preclude a role in BC predisposition. It is interesting to note that most of the established HBC genes that have been shown to undergo frequent biallelic inactivation in breast tumors, such as BRCA1, BRCA2, PALB2, and RAD51C, are highly penetrant and/or predispose selectively to TNBC, currently the sole exception being ATM, which is a moderate penetrance and ER-positive BC-associated gene. Interestingly, similar to CHEK2, the lower penetrance BRCA2 variant p.Lys3326Ter does not appear to require biallelic activation. A recent study of 26 BRCA2 p.Lys3326Ter-associated breast tumors found no instance of LOH (43).

Despite the ready availability of archival formalin-fixed, paraffin-embedded tumor blocks from BEACCON study participants, the main limitation to this study, was the small sample size for rare genes. In addition, the quality of formalin-fixed, paraffin-embedded tumor samples may cause potential errors in the determination of LOH. Tumor purity, especially in tumors that have high levels of infiltrating lymphocytes, may also introduce further complexity to the interpretation of allele frequency and copy number status, as addressed in the methods. Lastly, because of the greater requirements in terms of tumor DNA quality and quantity, methylation sequencing was not able to be carried out for all samples, therefore promoter hypermethylation cannot be ruled out for those samples, although in this study, no such instance was found.

In summary, this study demonstrates the utility of inclusion of tumor sequencing in HBC gene discovery and validation, but the absence of consistent biallelic inactivation in CHEK2 suggests this approach might not be reliable for lower penetrance genes.

Supplementary Material

djac196_Supplementary_Data
Click here for additional data file. (802.2KB, zip)

Contributor Information

Belle W X Lim, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Drug Discovery Biology, Monash Institute of Pharmaceutical Sciences, Monash University, Melbourne, Victoria, Australia.

Na Li, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Sir Peter MacCallum Department of Oncology, The University of Melbourne, Melbourne, Victoria, Australia; Parkville Familial Cancer Centre, Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia.

Sakshi Mahale, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Simone McInerny, Parkville Familial Cancer Centre, Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia.

Magnus Zethoven, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Bioinformatics Core Facility, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Simone M Rowley, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Joanne Huynh, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Theresa Wang, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Parkville Familial Cancer Centre, Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia.

Jue Er Amanda Lee, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Sir Peter MacCallum Department of Oncology, The University of Melbourne, Melbourne, Victoria, Australia; Molecular Genomics Core, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Mia Friedman, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Parkville Familial Cancer Centre, Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia.

Lisa Devereux, Sir Peter MacCallum Department of Oncology, The University of Melbourne, Melbourne, Victoria, Australia; Lifepool, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Rodney J Scott, Discipline of Medical Genetics and The Centre for Cancer Detection and Therapy, The University of Newcastle and Hunter Medical Research Institute, Newcastle, New South Wales, Australia; Division of Molecular Medicine, New South Wales Health Pathology North, Newcastle, New South Wales, Australia.

Erica K Sloan, Drug Discovery Biology, Monash Institute of Pharmaceutical Sciences, Monash University, Melbourne, Victoria, Australia; Division of Cancer Surgery, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Paul A James, Sir Peter MacCallum Department of Oncology, The University of Melbourne, Melbourne, Victoria, Australia; Parkville Familial Cancer Centre, Peter MacCallum Cancer Centre and Royal Melbourne Hospital, Melbourne, Victoria, Australia.

Ian G Campbell, Cancer Genetics Laboratory, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; Sir Peter MacCallum Department of Oncology, The University of Melbourne, Melbourne, Victoria, Australia; Lifepool, Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia.

Funding

This work was supported by the National Breast Cancer Foundation (IF-15-004, IGC and PAJ), Cancer Australia/National Breast Cancer Foundation (PdCCRS_1107870, IGC and PAJ), Cancer Australia/National Breast Cancer Foundation (PdCCRS_1188547, IGC and PAJ), the Victorian Cancer Agency (Tumor Stream Grant, PAJ), the National Health and Medical Research Council of Australia (GNT1023698, PAJ; GNT1041975, IGC) and Cancer Australia (PdCCRS_1188547, IGC, PAJ, and EKS). EKS is supported by the National Health and Medical Research Council GNT114749, the National Breast Cancer Foundation IIRS-20-025 and Cancer Council Victoria Grants-in-Aid. NL is supported by Cancer Council Victoria.

Notes

Role of the funder: The funding sources did not participate in the design and conduct of the study, the collection, analysis, and interpretation of the data, the preparation and writing of the manuscript, and the decision to submit the manuscript for publication.

Disclosures: The authors declare that there are no competing interests.

Author contributions: Conceptualization: IGC, PAJ. Data curation: BWXL, NL, SMR, MZ. Formal analysis: BWXL, IGC, PAJ. Funding acquisition: IGC, PAJ. Investigation: BWXL, NL, SM, SM, MZ, SMR, JH, TW, MF, AJEL, LD, RJS, EKS, PAJ, IGC. Methodology: BWXL, NL, MZ, PAJ, IGC. Resources: IGC, PAJ, RJS. Supervision: IGC, PAJ. Visualization: BWXL, NL, MZ. Writing—original draft: BWXL, IGC, PAJ, EKS. Writing—review & editing: BWXL, NL, SM, MZ, LD, EKS, RJS, PAJ, IGC.

Acknowledgements: The authors thank all the participants of the Variants in Practice and Lifepool studies for donating their DNA samples and clinical information. We also thank Norah Grewal, the Variants in Practice study site principal investigators Geoffrey Lindeman, Marion Harris, Lucinda Salmon, Ingrid Winship, and Yoland Antill and the staff at the Peter MacCallum Cancer Centre, Royal Melbourne Hospital, Monash Health, Cabrini Health and Barwon Health Familial Cancer Centres, and the Austin and Tasmanian Clinical Genetics Services, who enrolled participants and provided clinical data. We thank the following staff from Peter MacCallum Cancer Centre; Kaushalya Amarasinghe, Niko Thio, and Richard Lupat from Bioinformatics core facility for helping with the bioinformatic analysis and Steven Macaskill from Pathology Lab for Sanger sequencing services.

Prior presentations: Part of this study was presented at BRCA Symposium 2021, Montreal, Canada and Familial Aspects of Cancer Conference 2021, Melbourne, Australia.

Data availability

All sequencing data has been deposited to the European Genome-phenome Archive under accession numbers Study: EGAS00001006532 and Dataset: EGAD00001009299. Standard R codes were used. Code requests should be addressed to Prof. Ian Campbell.

References

  • 1. Narod SA, Foulkes WD.. BRCA1 and BRCA2: 1994 and beyond. Nat Rev Cancer. 2004;4(9):665-676. [DOI] [PubMed] [Google Scholar]
  • 2. Antoniou AC, Casadei S, Heikkinen T, et al. Breast-cancer risk in families with mutations in PALB2. N Engl J Med. 2014;371(6):497-506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Li N, Lim BWX, Thompson ER, et al. Investigation of monogenic causes of familial breast cancer: data from the BEACCON case-control study. NPJ Breast Cancer. 2021;7(1):76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Breast Cancer Association Consortium. Breast cancer risk genes—association analysis in more than 113,000 women. N Engl J Med. 2021;384(5):428-439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Hu C, Hart SN, Gnanaolivu R, et al. A population-based study of genes previously implicated in breast cancer. N Engl J Med. 2021;384(5):440-451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Maxwell KN, Wubbenhorst B, Wenz BM, et al. BRCA locus-specific loss of heterozygosity in germline BRCA1 and BRCA2 carriers. Nat Commun. 2017;8(1):319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Sokol ES, Pavlick D, Khiabanian H, et al. Pan-cancer analysis of BRCA1 and BRCA2 genomic alterations and their association with genomic instability as measured by genome-wide loss of heterozygosity. J Clin Oncol Precis Oncol. 2020;4:442-465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Tung N, Miron A, Schnitt SJ, et al. Prevalence and predictors of loss of wild type BRCA1 in estrogen receptor positive and negative BRCA1-associated breast cancers. Breast Cancer Res. 2010;12(6):R95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Osorio A, de la Hoya M, Rodriguez-Lopez R, et al. Loss of heterozygosity analysis at the BRCA loci in tumor samples from patients with familial breast cancer. Int J Cancer. 2002;99(2):305-309. [DOI] [PubMed] [Google Scholar]
  • 10. Polak P, Kim J, Braunstein LZ, et al. A mutational signature reveals alterations underlying deficient homologous recombination repair in breast cancer. Nat Genet. 2017;49(10):1476-1486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Li N, McInerny S, Zethoven M, et al. Combined tumor sequencing and case/control analyses of RAD51C in breast cancer. J Natl Cancer Inst. 2019;111(12):1332-1338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Lee JEA, Li N, Rowley SM, et al. Molecular analysis of PALB2-associated breast cancers. J Pathol. 2018;245(1):53-60. [DOI] [PubMed] [Google Scholar]
  • 13. Weigelt B, Bi R, Kumar R, et al. ; for the kConFab Investigators. The landscape of somatic genetic alterations in breast cancers from ATM germline mutation carriers. J Natl Cancer Inst. 2018;110(9):1030-1034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Telli ML, Timms KM, Reid J, et al. Homologous recombination deficiency (HRD) score predicts response to platinum-containing neoadjuvant chemotherapy in patients with triple-negative breast cancer. Clin Cancer Res. 2016;22(15):3764-3773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Lanchbury J, Timms K, Reid J, et al. 3-biomarker HRD score versus individual biomarker (LOH, TAI, LST) scores in platinum treated serous ovarian cancer (SOC). Ann Oncol. 2016;27(Suppl 6):vi34. [Google Scholar]
  • 16. Rosenthal R, McGranahan N, Herrero J, Taylor BS, Swanton C.. DeconstructSigs: delineating mutational processes in single tumors distinguishes DNA repair deficiencies and patterns of carcinoma evolution. Genome Biol. 2016;17:31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2016. [Google Scholar]
  • 18.Thompson ER, Rowley SM, Li N, et al. Panel Testing for Familial Breast Cancer: Calibrating the Tension Between Research and Clinical Care. J Clin Oncol. 2016;34(13):1455-1459. [DOI] [PubMed] [Google Scholar]
  • 19.Rowley SM, Mascarenhas L, Devereux L, et al. Population-based genetic testing of asymptomatic women for breast and ovarian cancer susceptibility. Genet Med. 2019;21(4):913-922. [DOI] [PubMed] [Google Scholar]
  • 20. Weischer M, Bojesen SE, Ellervik C, Tybjærg-Hansen A, Nordestgaard BG.. CHEK21100delC genotyping for clinical assessment of breast cancer risk: meta-analyses of 26,000 patient cases and 27,000 controls. J Clin Oncol. 2008;26(4):542-548. [DOI] [PubMed] [Google Scholar]
  • 21. Couch FJ, Shimelis H, Hu C, et al. Associations between cancer predisposition testing panel genes and breast cancer. JAMA Oncol. 2017;3(9):1190-1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Hilbers FS, Wijnen JT, Hoogerbrugge N, et al. Rare variants in XRCC2 as breast cancer susceptibility alleles. J Med Genet. 2012;49(10):618-620. [DOI] [PubMed] [Google Scholar]
  • 23. Park DJ, Lesueur F, Nguyen-Dumont T, et al. ; for the Kathleen Cuningham Foundation Consortium for Research into Familial Breast Cancer. Rare mutations in XRCC2 increase the risk of breast cancer. Am J Hum Genet. 2012;90(4):734-739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Figlioli G, Bogliolo M, Catucci I, et al. ; for kConFab. The FANCM:p.Arg658 truncating variant is associated with risk of triple-negative breast cancer. NPJ Breast Cancer. 2019;5(1):38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Kiiski JI, Pelttari LM, Khan S, et al. Exome sequencing identifies FANCM as a susceptibility gene for triple-negative breast cancer. Proc Natl Acad Sci USA. 2014;111(42):15172-15177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Neidhardt G, Hauke J, Ramser J, et al. Association between loss-of-function mutations within the FANCM gene and early-onset familial breast cancer. JAMA Oncol. 2017;3(9):1245-1248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Prokofyeva D, Bogdanova N, Dubrowinskaja N, et al. Nonsense mutation p.Q548X in BLM, the gene mutated in Bloom’s syndrome, is associated with breast cancer in Slavic populations. Breast Cancer Res Treat. 2013;137(2):533-539. [DOI] [PubMed] [Google Scholar]
  • 28. Sokolenko AP, Iyevleva AG, Preobrazhenskaya EV, et al. High prevalence and breast cancer predisposing role of the BLM c.1642 C>T (Q548X) mutation in Russia. Int J Cancer. 2012;130(12):2867-2873. [DOI] [PubMed] [Google Scholar]
  • 29. Thompson ER, Doyle MA, Ryland GL, et al. ; for kConFab. Exome sequencing identifies rare deleterious mutations in DNA repair genes FANCC and BLM as potential breast cancer susceptibility alleles. PLoS Genet. 2012;8(9):e1002894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Li A, Geyer FC, Blecua P, et al. ; for kConFab Investigators. Homologous recombination DNA repair defects in PALB2-associated breast cancers. NPJ Breast Cancer. 2019;5(1):23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Ng PS, Pan JW, Ahmad Zabidi MM, et al. Characterisation of PALB2 tumours through whole-exome and whole-transcriptomic analyses. NPJ Breast Cancer. 2021;7(1):46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Mandelker D, Kumar R, Pei X, et al. The landscape of somatic genetic alterations in breast cancers from CHEK2 germline mutation carriers. JNCI Cancer Spectr. 2019;3(2):pkz027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Massink MPG, Kooi IE, Martens JWM, Waisfisz Q, Meijers-Heijboer H.. Genomic profiling of CHEK2*1100delC-mutated breast carcinomas. BMC Cancer. 2015;15:877- [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Suspitsin EN, Yanus GA, Sokolenko AP, et al. Development of breast tumors in CHEK2, NBN/NBS1 and BLM mutation carriers does not commonly involve somatic inactivation of the wild-type allele. Med Oncol. 2014;31(2):828. [DOI] [PubMed] [Google Scholar]
  • 35. Muranen TA, Greco D, Fagerholm R, et al. Breast tumors from CHEK2 1100delC-mutation carriers: genomic landscape and clinical implications. Breast Cancer Res. 2011;13(5):R90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Oldenburg RA, Kroeze-Jansema K, Kraan J, et al. The CHEK21100delC variant acts as a breast cancer risk modifier in non-BRCA1/BRCA2 multiple-case families. Cancer Res. 2003;63(23):8153-8157. [PubMed] [Google Scholar]
  • 37. Sodha N, Bullock S, Taylor R, et al. CHEK2 variants in susceptibility to breast cancer and evidence of retention of the wild type allele in tumours. Br J Cancer. 2002;87(12):1445-1448. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Nik-Zainal S, Davies H, Staaf J, et al. Landscape of somatic mutations in 560 breast cancer whole-genome sequences. Nature. 2016;534(7605):47-54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Teugels E, De Brakeleer S.. An alternative model for (breast) cancer predisposition. NPJ Breast Cancer. 2017;3(1):13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Kluźniak W, Wokołorczyk D, Rusak B, et al. Inherited variants in BLM and the risk and clinical characteristics of breast cancer. Cancers (Basel). 2019;11(10):1548. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Heikkinen K, Rapakko K, Karppinen S-M, et al. RAD50 and NBS1 are breast cancer susceptibility genes associated with genomic instability. Carcinogenesis. 2006;27(8):1593-1599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Tommiska J, Seal S, Renwick A, et al. Evaluation of RAD50 in familial breast cancer predisposition. Int J Cancer. 2006;118(11):2911-2916. [DOI] [PubMed] [Google Scholar]
  • 43. Butz H, Papp J, Bozsik A, et al. Application of multilayer evidence for annotation of c-terminal BRCA2 variants. Cancers (Basel). 2021;13(4):881. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

djac196_Supplementary_Data
Click here for additional data file. (802.2KB, zip)

Data Availability Statement

All sequencing data has been deposited to the European Genome-phenome Archive under accession numbers Study: EGAS00001006532 and Dataset: EGAD00001009299. Standard R codes were used. Code requests should be addressed to Prof. Ian Campbell.

Articles from JNCI Journal of the National Cancer Institute are provided here courtesy of Oxford University Press

RESOURCES