Association of BRCA2 K3326* With Small Cell Lung Cancer and Squamous Cell Cancer of the Skin

Association of BRCA2 K3326* With Small Cell Lung Cancer and Squamous Cell Cancer of the Skin Abstract Background Most pathogenic mutations in the BRCA2 gene carry a high risk of hereditary breast and ovarian cancer (HBOC). However, a stop-gain mutation, K3326* (rs11571833), confers risk of lung cancer and cancers of the upper-aero-digestive tract but only a modest risk of breast or ovarian cancer. The Icelandic population provides an opportunity for comprehensive characterization of the cancer risk profiles of K3326* and HBOC mutations because a single mutation, BRCA2 999del5, is responsible for almost all BRCA2-related HBOC in the population. Methods Genotype information on 43 641 cancer patients and 370 971 control subjects from Iceland, the Netherlands, and the United States was used to assess the cancer risk profiles of K3326* and BRCA2 999del5. BRCA2 expression was assessed using RNAseq data from blood (n = 2233), as well as 52 tissues reported in the GTEx database. Results The cancer risks associated with K3326* are fundamentally different from those associated with 999del5. We report for the first time an association between K3326* and small cell lung cancer (odds ratio [OR] = 2.06, 95% confidence interval [CI] = 1.35 to 3.16) and squamous cell carcinoma of the skin (OR = 1.69, 95% CI = 1.26 to 2.26). Individuals homozygous for K3326* reach old age and have children. Unlike BRCA2 999del5, the K3326* allele does not affect the level of BRCA2 transcripts, and the allele is expressed to the same extent as the wild-type allele. Conclusions K3326* associates primarily with cancers that have strong environmental genotoxic risk factors. Expression of the K3326* allele suggests that a variant protein may be made that retains the DNA repair capabilities important to hormone-responsive tissues but may be less efficient in responding to genotoxic stress. Pathogenic mutations in BRCA2 predispose to hereditary breast and ovarian cancer (HBOC) syndrome, characterized by greatly increased risk of cancers of the breast and ovary as well as other cancers (1). However, HBOC-associated mutations do not increase risk of lung cancer, suggesting that lung epithelium may be less dependent on BRCA2 function than the tissues involved in HBOC. It was therefore unexpected when a stop-gain variant close to the 3’ end of the BRCA2 gene, rs11571833 (NM_000059.3:c.9976A>T; NP_000050.2:p.Lys3326Ter, hereafter referred to as K3326*), was reported to confer risk of lung cancer (2). Analysis of 21 594 lung cancer patients and 54 156 control subjects of European origin found that carriers of the variant had an odds ratio (OR) of 1.83 (95% confidence interval [CI] = 1.61 to 2.09) of developing the disease (2). The same study also showed that the association was stronger with squamous cell lung carcinoma (SQLC) than adenocarcinoma of the lung (OR = 2.47, 95% CI = 2.03 to 3.00; OR = 1.47, 95% CI = 1.19 to 1.82, respectively). K3326* is located in the last of the 27 exons of the BRCA2 gene and is predicted to result in the loss of the 93 C-terminal amino acids of the protein product. In addition to its association with lung cancer risk, K3326* associates with substantial risk of cancers of the upper aero-digestive tract (UADT; OR = 2.53, 95% CI = 1.89 to 3.38) (3) and esophageal squamous cell carcinoma (OR = 6.0, 95% CI = 1.3 to 28) (4). Unlike variants in BRCA2 that associate with HBOC, the K3326* variant has a small effect on hormone-related cancers (5,6). A recent study including 76 637 cancer patients and 83 796 control subjects showed a modest increase in risk of breast cancer (OR = 1.28, 95% CI = 1.17 to 1.40) and invasive ovarian cancer (OR = 1.26, 95% CI = 1.10 to 1.43) (6). No association with prostate cancer was observed. Finally, K3326* was reported to be more common in familial pancreatic cancer patients than in control subjects (144 cancer patients and 250 control subjects, OR = 4.84, 95% CI = 1.27 to 18.55) (7). The difference between the cancer risk profiles of K3326* and variants in BRCA2 that associate with HBOC could provide insights into how the roles of BRCA2 differ between tissues. However, dissection of this phenomenon is complicated because rare pathogenic mutations may reside on the background of K3326*, affecting the risk estimates for the variant. The Icelandic population is ideally suited to the characterization of cancer risk and clinical presentation of K3326*. This is because a single founder mutation in BRCA2, commonly referred to as 999del5 (rs80359671, NM_000059.3:c.767_771delCAAAT, NP_000050.2:p.Asn257Lysfs), is responsible for virtually all BRCA2-related HBOC in the population, and this mutation is never on the same chromosome as K3326* (2). Here, we combined extensive information on genetic variation in the Icelandic population with data from the nation-wide cancer registry in order to compare the cancer risk profiles of K3326* and HBOC mutations. Methods This study used genotype information on a total of 43 641 cancer patients and 370 971 control subjects from Iceland, the Netherlands, and the United States. An in-depth description of the study populations and methods is presented in the Supplementary Methods (available online). Icelandic Study Population This study is based on extensive genetic information on the Icelandic population, which has been previously described (8). The whole genomes of 15 220 Icelanders were sequenced, unveiling 40 780 213 single nucleotide polymorphisms (SNP) and short indels. These variants were imputed into 151 677 Icelanders whose DNA had been genotyped with various Illumina SNP chips and phased using long-range phasing (9,10). Genealogical deduction of carrier status of 282 894 untyped relatives of chip-typed individuals further increased the sample size for association analysis. Information on cancer in the genotyped individuals is from the population-based Icelandic Cancer Registry (ICR) (11). A total of 42 331 Icelandic cancer patients and 354 488 control subjects were used in the analysis. With the aid of hospital charts, 53 genotyped lung cancer K3326* carriers diagnosed in the years 2000–2015 were staged according to the International Association for the Study of Lung Cancer 7th edition of TNM staging. These were compared with the stages of all 155 lung cancer patients diagnosed in 2009 in Iceland. The study was approved by the National Bioethics Committee of Iceland (ref. 12–122-V7). Written informed consent was obtained from all genotyped subjects. Statistical Analysis Association results for the BRCA2 variants tested in the Icelandic cancer patients come from genome-wide association studies (GWAS) on the 20 cancers. The methods used for association testing in the Icelandic population have been described in detail (12). To test for association between SNPs and cancer in the Icelandic study, logistic regression was used, treating disease status as the response and genotype counts as covariates. Other relevant covariates that might correlate with disease status were also included in the model as nuisance, for example, sex, county of birth, current age or age at death (first- and second-order terms included), blood sample availability for the individual, and an indicator function for the overlap of the lifetime of the individual with the time span of phenotype collection. To account for inflation in test statistics due to cryptic relatedness and stratification in the Icelandic population, we applied the method of linkage disequilibrium (LD) score regression (13). With a set of 1.1 M variants, we regressed the χ2 statistics from our GWAS scan against LD score and used the intercept as a correction factor. The LD scores were downloaded from an LD score database (see the URL in the “Notes”), and the estimated correction factors are listed in Supplementary Table 1 (available online). All statistical tests were two-sided, and a P value of less than .05 was considered statistically significant unless otherwise noted. UADT Cancer Patients and Control Subjects In total, 696 head and neck cancers and 119 squamous cell esophageal carcinomas were collected in two studies (14,15). Information on the ICD codes for the UADT populations used and referred to can be found in Supplementary Table 2 (available online). The 4139 Dutch control subjects were from the Nijmegen Biomedical Study (16). The protocols of all the studies were approved by the respective institutional review boards, and all study subjects gave written informed consent. In the UADT cancer patients and control subjects, the rs11571833 (K3326*) variant was genotyped using the Centaurus (Nanogen) platform (17). The c.6275delTT and c.4889C>G mutations were genotyped by Sanger sequencing. Primer sequences are listed in Supplementary Table 3 (available online). Squamous Cell Skin Cancer Patients and Control Subjects Association results for squamous cell skin cancer (SQCSC) and the BRCA2 variants were looked up in two GWAS studies, one from Rotterdam and the other from Ohio. The Rotterdam Study The Rotterdam Study (RS) is a prospective population-based follow-up study of the determinants and prognosis of chronic diseases in the middle age and elderly participants living in the Ommoord district (Rotterdam, the Netherlands) (18). The Medical Ethics Committee of the Erasmus Medical Center and the review board of the Dutch Ministry of Health, Welfare and Sports have ratified the RS. Written informed consent was obtained from all participants. Details on the ascertainment of skin cancer in the RS is presented elsewhere (19). Briefly, all RS participants with informed consent (n = 14 628) were linked to the nationwide registry of histo- and cytopathology in the Netherlands (PALGA; up to December 31, 2013). All prevalent SQCSCs were labeled as patients. Participants from the RS without PALGA reports were considered control subjects. Cohorts RS-I and RS-II were genotyped with the Infinium II HumanHap550K Genotyping BeadChip version 3 (Illumina, San Diego, CA), and the cohort RS-III was genotyped using the Illumina Human 610 Quad Arrays. The three data sets were aligned to the same strand as the GoNL data set (20) and restricted to a common set of 452 673 autosomal variants. Variants with a minor allele frequency of less than 0.5% (nine variants) that fail (P < 1e-7) the Hardy-Weinberg test and that have statistically significant differences (P < 1e-6) in frequency between data sets were excluded. The resulting genotypes for 450 395 variants were phased using SHAPEIT (v2.790) (21). Imputation was done with IMPUTE2 (v2.3.2) (22), using the GoNL data set as reference (https://molgenis26.target.rug.nl/downloads/gonl_public/variants/release5/) (20). A logistic case–control study GWAS was carried out on 398 SQCSC patients and 10 629 control subjects of Dutch origin, using SNPTEST (v2.5.2) (23) with a frequentist (additive) and expected dosage as main parameters. The model was adjusted for age at the start of the study, sex, and four principal components. Further quality control of the GWAS results included the removal of markers with low imputation quality (R2 < 0.3), duplicated markers, and markers with high betas or standard errors (larger than 10), using EasyQC (24). The rs11571833 (K3326*) variant is not present in GoNL. Therefore, we used rs11571815, which is fully correlated with rs11571833 (R2 = 1) in Northern Europeans from Utah (CEU) and Iceland as a proxy. The Ohio Study This study was approved by the Ohio State University Institutional Review Board. All study participants signed informed consent. Genomic DNA was collected from 103 SQCSC patients ascertained from dermatology clinics. Control subjects (n = 1715) were obtained from the Columbus area control study and were individuals without a self-reported SQCSC diagnosis and were matched to the population and banked patients by sex and age of diagnosis. Samples were genotyped on Illumina chips as described above, phased using SHAPEIT (v2.790), and imputed using 1000 Genomes Phase 3 (October 2014) as a reference set. Test for Heterogeneity Between the Different Populations The results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors (Supplementary Methods, available online) (25). RNA Analysis Preparation of poly(A)+ cDNA sequencing libraries and RNA sequencing were carried out as described previously (26). A detailed protocol for the RNA analysis is presented in the Supplementary Methods (available online). Results To compare the cancer risks associated with K3326* and 999del5, we tested the associations between the variants and 20 cancer types (Table 1). As expected, very strong associations were observed between 999del5 and cancers of the breast, ovary, prostate, and pancreas. In addition, squamous UADT cancers showed statistically significant association with 999del5 (20 tests, statistical significance threshold of P < 0.05/20 = 2.5 × 10-3). Several cancer types show suggestive association with 999del5, that is, SQCSC and cancers of the stomach, urinary bladder, and cervix. When all cancers are combined, the risk of 999del5 carriers being diagnosed with some form of cancer is more than fourfold that of the general population. We observed no statistically significant association between lung cancer and 999del5 (OR = 1.02, 95% CI = 0.69 to 1.52, P = .09). In contrast, K3326* associates with lung cancer (OR = 1.54, 95% CI = 1.23 to 1.91, P = 1.2 × 10−4) and squamous cell skin cancer (OR = 1.69, 95% CI = 1.26 to 2.26, P = 4.2 × 10−4). Within lung cancer, the cancer risk was confined to SQLC (OR = 1.71, 95% CI = 1.10 to 2.67, P = .02) and small cell lung cancer (SCLC; OR = 2.06, 95% CI = 1.35 to 3.16, P = 9.0 × 10−4), whereas no association to adenocarcinoma was detected in our sample set (OR = 0.98, 95% CI = 0.66 to 1.45, P = .92). The associations of K3326* with SCLC and SQCSC have not been reported before. Table 1. Association between 20 cancer types and two BRCA2 mutations in the Icelandic population; K3326* and the Icelandic founder mutation 999del5† K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 † Association results are shown for the 19 most common solid tumors in the Icelandic cancer registry (41 976 diagnoses) and all cancers combined (42 331 individuals, including some rare cancers not listed in the table), as well as for the three most common histological subtypes of lung cancer. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; MGUS = monoclonal gammopathy of undetermined significance; OR = odds ratio. ‡ The numbers of cancer patients (Naff) and control subjects (Ncontrols) correspond to the number of individuals used in the association tests. § The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. ‖ The odds ratios correspond to minor alleles. ¶ Results that remain statistically significant after correction for number of tests. Table 1. Association between 20 cancer types and two BRCA2 mutations in the Icelandic population; K3326* and the Icelandic founder mutation 999del5† K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 † Association results are shown for the 19 most common solid tumors in the Icelandic cancer registry (41 976 diagnoses) and all cancers combined (42 331 individuals, including some rare cancers not listed in the table), as well as for the three most common histological subtypes of lung cancer. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; MGUS = monoclonal gammopathy of undetermined significance; OR = odds ratio. ‡ The numbers of cancer patients (Naff) and control subjects (Ncontrols) correspond to the number of individuals used in the association tests. § The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. ‖ The odds ratios correspond to minor alleles. ¶ Results that remain statistically significant after correction for number of tests. To refine our estimate of the association of K3326* with SQCSC, we genotyped the variant in two additional SQCSC sample sets from the Netherlands and the United States (Table 2). Combining the results from the three sample sets gives a Pcombined value of 1.7 × 10−4 (OR = 1.66, 95% CI = 1.27 to 2.16, Phet = .91). Table 2. Association between K3326* and SQCSK and UADT cancer in Iceland, the Netherlands, and the United States Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 † The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; OR = odds ratio; SQCSC = squamous cell skin cancer; UADT = upper aero-digestive tract. ‡ Imputation information for marker estimated by the ratio of the variance of imputed expected allele counts and the variance of the actual allele counts. § The variant rs11571815 was used as a surrogate for rs1157833 (K3326* variant) in the Dutch SQCSC population because the GoNL data set does not contain rs1157833. The two variants are fully correlated in CEU (R2 = 1). ‖ rs1157833 was directly genotyped in the Dutch UADT patients and control subjects. Table 2. Association between K3326* and SQCSK and UADT cancer in Iceland, the Netherlands, and the United States Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 † The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; OR = odds ratio; SQCSC = squamous cell skin cancer; UADT = upper aero-digestive tract. ‡ Imputation information for marker estimated by the ratio of the variance of imputed expected allele counts and the variance of the actual allele counts. § The variant rs11571815 was used as a surrogate for rs1157833 (K3326* variant) in the Dutch SQCSC population because the GoNL data set does not contain rs1157833. The two variants are fully correlated in CEU (R2 = 1). ‖ rs1157833 was directly genotyped in the Dutch UADT patients and control subjects. A strong association between the 999del5 variant and UADT cancers was observed in Iceland (OR = 3.16, 95% CI = 1.82 to 5.89, P = 2.1 × 10-5). This association was driven primarily by squamous esophageal cancer (OR = 5.71, 95% CI = 2.63 to 12.41, P = 1.1 × 10-5, 366 patients, 186 714 control subjects). However, contrary to previous reports (3,4), there was no statistically significant association observed between UADT cancers and K3326* in Iceland. To exclude systematic errors in the registration of this cancer group, a pathology review was performed on a random set of 30 samples from different periods. The pathology review confirmed the diagnoses of all 30 samples. In line with published results, we observed a strong association of K3326* with UADT cancers in samples from the Netherlands (OR = 2.68, 95% CI = 1.67 to 4.31, P = 5.0 × 10-5) (Table 2). Comparing the distribution of cancer sites between the Icelandic and Dutch UATD cancer patients, we note a higher fraction of oral cavity cancer in the Icelandic patients and a higher fraction of cancers of the oropharynx and hypopharynx in the Dutch patients (Supplementary Table 4, available online). However, this difference does not explain the difference in cancer risk between the populations because K3326* is associated with risk of all subtypes of UADT cancers in our Dutch data set (Supplementary Table 5, available online) as well as in the published report (3). Previous studies of families with breast cancer of northern European ancestry show that two rare mutations, BRCA2 c.6275delTT and c.4889C>G, which are highly penetrant for breast and ovarian cancer, occur on the K3326* haplotype (27). These mutations have not been found in Iceland. However, the BRCA2 c.6275delTT mutation has been found in several HBOC families in the Netherlands (28), prompting speculations that these mutations could be responsible for the UADT signal, rather than the K3326* variant in the Dutch population. We genotyped the c.6275delTT and c.4889C>G mutations in the Dutch UADT cancer sample set using Sanger sequencing (Supplementary Table 6, available online). Neither mutation was found, suggesting that the K3326* variant alone explains the association with UADT cancer in the Netherlands. Carriers of HBOC mutations in BRCA2 are on average diagnosed with breast, ovarian, and prostate cancers at a younger age than noncarriers (29–31). The K3326* variant did not affect the age at diagnosis for lung cancer (Supplementary Table 7, available online). When considering the whole Icelandic population, carriers of K3326* have a slightly shorter lifespan than noncarriers among deceased individuals who lived to be age 50 years and were born after 1890 (P = .01, β = –0.063, n = 118 626). Considering lung cancer patients only, the ratio of males to females among carriers of K3326* was not statistically significantly different from the ratio of males to females with lung cancer in the total cancer registry (55.5% vs 51.9%, P = .54). Bi-allelic mutations in BRCA2 result in Fanconi anemia (FA) complementation group D1 (FA-D1) (32), a chromosome instability disorder that causes childhood-onset aplastic anemia and cellular hypersensitivity to DNA crosslinking agents, as well as a strong cancer predisposition. Although the cancer risk profile of the K3326* mutation is more modest than that of the HBOC mutations, it is possible that homozygosity for the variant could have consequences over and beyond a multiplicative risk for the cancers observed in heterozygotes. In total, 17 homozygous individuals were observed in our population of 151 677 chip-typed Icelanders; this is in agreement with the expectation under the Hardy-Weinberg equilibrium (17 expected), given the allelic frequency of 1.1% for K3326*. Homozygous individuals reach adulthood (oldest 92 years) and have children (Table 3). Information on cancer in the 17 K3326* homozygotes, as well as the seven compound K3326*/999del5 heterozygous carriers observed, is listed in Table 3. No homozygotes for the 999del5 mutation were observed in 151 677 chip-typed individuals (expected to be two under the Hardy-Weinberg equilibrium and allelic frequency of 0.36%). Table 3. Information on homozygous carriers of K3326* and compound heterozygous carriers of K3326* and 999del5† Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 † Age at dx = age at diagnosis; BCC = basal cell carcinoma; SCLC = small cell lung cancer; SQCSC = squamous cell skin cancer. Table 3. Information on homozygous carriers of K3326* and compound heterozygous carriers of K3326* and 999del5† Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 † Age at dx = age at diagnosis; BCC = basal cell carcinoma; SCLC = small cell lung cancer; SQCSC = squamous cell skin cancer. We tested if the stage distribution of carriers of K3326* was different from noncarriers. Because disease stage is not registered in the ICR, we retrieved the medical records and registered the stage of disease of all 53 carriers of K3326* diagnosed with LC in the years 2000–2015 as well as all 155 LC patients diagnosed in the year 2009 for comparison. We found no difference in the stage distribution between carriers and noncarriers with NSCLC (Supplementary Table 8, available online). The effect of K3326* on the function of the BRCA2 protein in different tissues remains to be determined. As the variant introduces a stop codon in the last exon of the gene, the mRNA will likely escape nonsense-mediated decay (NMD), allowing translation of a truncated protein product (33). To assess the effects of the K3326* and 999del5 variants on transcript levels, BRCA2 mRNA was assessed in RNA-Seq data from the blood of individuals with wild-type BRCA2 (n = 2181), individuals heterozygous for K3326* (n = 40), and individuals homozygous for K3326* (n = 2), as well as individuals heterozygous for 999del5 (n = 10). As previously reported (34), heterozygous carriers of the 999del5 mutation have fewer BRCA2 transcripts than noncarriers, in line with 999del5 inducing NMD (Figure 1). In contrast, carriers of K3326* express normal levels of BRCA2 mRNA, and no allelic imbalance in expression was detected for heterozygous individuals (87 reference vs 90 alternative total number of fragments). This indicates that transcripts carrying the K3326* are not subjected to NMD and that a truncated BRCA2 protein may be generated. Similarly, using data from GTEx, K3326* was not found to affect BRCA2 transcript levels in other tissues, such as lung (Supplementary Figure 1, available online), sun-exposed skin, subcutaneous adipose tissue, thyroid, or whole blood (Supplementary Table 9, available online). Figure 1. View largeDownload slide Expression of BRCA2 in whole blood by genotypes of K3326* and 999del5. Expression values are quantile-normalized and mapped to a standard normal distribution. The bottom and top of the boxes correspond to the 25th (Q1) and 75th (Q3) percentiles, the line inside the box corresponds to the median, and the whiskers are located at max(min(Expression), Q1 – 1.5 IQR) and min(max(Expression), Q3 + 1.5 IQR), respectively (where IQR is the interquartile range = Q3 – Q1). Linear regression, assuming an additive model for genotype effects and including both variants in the model, estimated a β of –0.71 (95% confidence interval [CI] =–1.30 to –0.11, P=.02) for 999del5 and a β of 0.22 (95% CI = –0.06 to 0.49, P= .13) for K3326*. Figure 1. View largeDownload slide Expression of BRCA2 in whole blood by genotypes of K3326* and 999del5. Expression values are quantile-normalized and mapped to a standard normal distribution. The bottom and top of the boxes correspond to the 25th (Q1) and 75th (Q3) percentiles, the line inside the box corresponds to the median, and the whiskers are located at max(min(Expression), Q1 – 1.5 IQR) and min(max(Expression), Q3 + 1.5 IQR), respectively (where IQR is the interquartile range = Q3 – Q1). Linear regression, assuming an additive model for genotype effects and including both variants in the model, estimated a β of –0.71 (95% confidence interval [CI] =–1.30 to –0.11, P=.02) for 999del5 and a β of 0.22 (95% CI = –0.06 to 0.49, P= .13) for K3326*. Discussion Here, we used the comprehensive information on genetic variation in the Icelandic population, along with information on 49 051 diagnoses of 20 different cancers to dissect and compare the cancer risks of the K3326* variant and the 999del5 founder mutation that behaves like a typical HBOC mutation. We confirm the previously reported associations with total LC and SQLC. In addition, we demonstrate associations with SCLC and SQCSC not previously reported. The reported association between K3326* and UADT cancer was not replicated in the Icelandic population, although a strong effect of the 999del variant on this cancer group was observed. Possible explanations of the difference in UADT cancer risk between Iceland and the other populations include different etiological factors, such as infection with human papilloma virus (HPV). HPV associates most strongly with oropharyngeal cancers, and this cancer subtype is less common in the Icelandic population than in the Dutch data set used here. However, K3326* was previously reported to associate with all subtypes of UADT cancers (oral cavity, oropharynx, larynx/hypopharynx, and esophagus) (3), and these associations are replicated in our Dutch sample set. In the Icelandic UADT cancer group, the strongest effect of K3326* (albeit statistically nonsignificant) was with oropharyngeal cancer. It is of interest that K3326* associates with cancer types that have strong environmental genotoxic risk factors, that is, smoking for SQLC, SCLC and UADT cancer, and ultraviolet (UV) radiation for SQCSC. This suggests that the DNA repair capability of the truncated BRCA2 protein may be impaired, rendering affected individuals more vulnerable to genotoxic stress. BRCA2 has two well-characterized roles in the maintenance of genomic integrity, homology-directed repair (HDR) of double-strand breaks (35) and a role in preventing degradation of stalled replication forks (36). In both cases, the function of BRCA2 is dependent on the interaction with RAD51. The RAD51 interaction domain needed to stabilize stalled replication forks is located between amino acids 3270 and 3305 in exon 27 (37), in close proximity to K3326*, prompting speculations that the truncation might interfere with the interaction between RAD51 and the C-terminus of BRCA2. This would affect the ability of BRCA2 to prevent the MRE11-mediated degradation of stalled replication forks, resulting in genomic instability and increased cancer risk. Such a defect would be more likely to have a biological impact under stressful conditions where replication fork progression is constantly being challenged, for instance, in the presence of bulky DNA adducts caused by frequent exposure to carcinogenic agents such as tobacco smoke or UV light. On the contrary, studies have shown that the BRCA2 C-terminus RAD51 interaction domain is dispensable for HDR repair (36), which may explain why the K3326* variant does not have a strong effect on breast or ovarian cancer risks. Our population-based assessment of the 999del5 founder mutation highlights the complexity of cancer risks conferred by loss-of-function mutations in BRCA2. Current evidence suggests that other cancer susceptibility genes, many of which have a role in DNA repair, also have complex risk patterns that still need to be characterized in full (38). Dissection of the risk patterns of these genes is likely to have important implications for the genetic counseling of mutation carriers, which has historically been strongly focused on a limited number of cancer types. The limitation of this study lies in the lack of power for rare cancer types. Additional studies with more patients will provide more robust estimates of the risk for these cancer types. Also, due to the low frequency of the K3326* variant, it was not possible to test if the variant affects survival of lung cancer patients. In conclusion, our study shows that the K3326* variant in BRCA2 associates primarily with cancer types that have strong environmental genotoxic risk factors. We report for the first time an association between the variant and risk of SCLC and SQCSC. The risk profile of K3326* suggests that while the variant protein retains the ability to maintain genomic stability in hormone-related tissues, it may be less efficient in repairing damages caused by carcinogenic agents such as tobacco smoke or UV light. Funding The study was funded by deCODE genetics/Amgen. The Ohio Study on SQCSC was supported in part by the American Cancer Society (RSG-07-083 MGO to AET), the National Institutes of Health (R03 CA173788), and by the National Cancer Institute (grant P30 CA16058 to the OSU Comprehensive Cancer Center). Notes Affiliations of authors: deCODE Genetics/Amgen, Reykjavik, Iceland (TR, SNS, GH, PS, HH, AS, JG, JS, JKS, HJ, DFG, GM, UT, GT, KS); Department of Oncology (GRS, KA, OTJ, FS, GBR), Department of Pathology (JGJ, HI), Department of Dermatology (BS, KT, JHO), Department of Medicine (HH, SJ), and Department of Surgery (TG), Landspitali-University Hospital, Reykjavik, Iceland; Department of Dermatology, Erasmus MC Cancer Institute, Rotterdam, the Netherlands (LMP, TN); Faculty of Electrical and Computer Engineering (HH), Faculty of Medicine (STS, TG, LT, JGJ, BS, KT, TG, JHO, UT, KS), and School of Engineering and Natural Sciences (DFG), University of Iceland, Reykjavik, Iceland; Icelandic Cancer Registry, Reykjavik, Iceland (LT, GHO, KA); Department of Internal Medicine, Erasmus MC University Medical Center, Rotterdam, the Netherlands (AU); Radboud University Medical Center, Radboud Institute for Health Sciences, Nijmegen, the Netherlands (SHV, TEG, LAMLK); Department of Internal Medicine (DCA), Department of Cancer Biology (AET), and Department of Genetics and Internal Medicine (AET), Division of Human Genetics, and The Comprehensive Cancer Center (AET), The Ohio State University, Columbus, OH (DCA); Department of Otorhinolaryngology, Head and Neck Surgery, Maastricht University Medical Center, Maastricht, the Netherlands (ML); Department of Gastroenterology, Radboud University Nijmegen Medical Center, Nijmegen, the Netherlands (WHMP). The funders had no role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the manuscript for publication. All authors from deCODE are employees of deCODE genetics/Amgen. From Ohio: The Human Genetics Sample bank was essential for processing of samples for DNA. We thank Victoria Klee and Madison Bernhardt for ascertainment of medical records and database management and Elizabeth Solinger for accrual of control individuals for this study. We thank Dr. Albert de la Chapelle for provision of some of the control samples from the Ohio set. From the Rotterdam Study: The authors are grateful to the study participants, the staff from the Rotterdam Study, and the participating general practitioners and pharmacists. We thank the staff from Internal Medicine for generation and maintenance of GWAS data (P. Arp, M. Jhamai, M. Verkerk, L. Herrera, M. Peters). We also thank E. den Broek, L. Overbeek, and S. Koljenovic for help in the linkage with the pathology data from PALGA. Data can be obtained upon request. Requests should be directed to the management team of the Rotterdam Study (secretariat.epi@erasmusmc.nl), which has a protocol for approving data requests. Because of restrictions based on privacy regulations and informed consent of the participants, data cannot be made freely available in a public repository. From NBS: The Nijmegen Biomedical Study is a population-based survey conducted at the Department for Health Evidence and the Department of Laboratory Medicine of the Radboud University Medical Center. The principal investigators of the Nijmegen Biomedical Study are L. A. L. M. Kiemeney, A. L. M. Verbeek, D. W. Swinkels, and B. Franke. LD score database (accessed June 23, 2015): ftp://atguftp.mgh.harvard.edu/brendan/1k_eur_r2_hm3snps_se_weights.RDS. GTEx Portal (accessed June 12, 2016): http://www.gtexportal.org/home/. References 1 Lynch HT , Snyder C , Casey MJ. Hereditary ovarian and breast cancer: What have we learned? Ann Oncol. 2013 ; 24(suppl 8) : viii83 – viii95 . Google Scholar PubMed 2 Wang Y , McKay JD , Rafnar T , et al. . Rare variants of large effect in BRCA2 and CHEK2 affect risk of lung cancer . Nat Genet. 2014 ; 46 7 : 736 – 741 . http://dx.doi.org/10.1038/ng.3002 Google Scholar Crossref Search ADS PubMed 3 Delahaye-Sourdeix M , Anantharaman D , Timofeeva MN , et al. . A rare truncating BRCA2 variant and genetic susceptibility to upper aerodigestive tract cancer . J Natl Cancer Inst. 2015 ; 107 5 :djv037. 4 Akbari MR , Malekzadeh R , Nasrollahzadeh D , et al. . Germline BRCA2 mutations and the risk of esophageal squamous cell carcinoma . Oncogene. 2008 ; 27 9 : 1290 – 1296 . http://dx.doi.org/10.1038/sj.onc.1210739 Google Scholar Crossref Search ADS PubMed 5 Michailidou K , Hall P , Gonzalez-Neira A , et al. . Large-scale genotyping identifies 41 new loci associated with breast cancer risk . Nat Genet. 2013 ; 45 4 : 353 – 361 , 361–362. Google Scholar Crossref Search ADS PubMed 6 Meeks HD , Song H , Michailidou K , et al. . BRCA2 polymorphic stop codon K3326X and the risk of breast, prostate, and ovarian cancers . J Natl Cancer Inst. 2015 ; 108 2 :djv315. 7 Martin ST , Matsubayashi H , Rogers CD , et al. . Increased prevalence of the BRCA2 polymorphic stop codon K3326X among individuals with familial pancreatic cancer . Oncogene 2005 ; 24 22 : 3652 – 3656 . http://dx.doi.org/10.1038/sj.onc.1208411 Google Scholar Crossref Search ADS PubMed 8 Gudbjartsson DF , Helgason H , Gudjonsson SA , et al. . Large-scale whole-genome sequencing of the Icelandic population . Nat Genet. 2015 ; 47 5 : 435 – 444 . http://dx.doi.org/10.1038/ng.3247 Google Scholar Crossref Search ADS PubMed 9 Kong A , Masson G , Frigge ML , et al. . Detection of sharing by descent, long-range phasing and haplotype imputation . Nat Genet. 2008 ; 40 9 : 1068 – 1075 . http://dx.doi.org/10.1038/ng.216 Google Scholar Crossref Search ADS PubMed 10 Kong A , Steinthorsdottir V , Masson G , et al. . Parental origin of sequence variants associated with complex diseases . Nature. 2009 ; 462 7275 : 868 – 874 . http://dx.doi.org/10.1038/nature08625 Google Scholar Crossref Search ADS PubMed 11 Sigurdardottir LG , Jonasson JG , Stefansdottir S , et al. . Data quality at the Icelandic Cancer Registry: Comparability, validity, timeliness and completeness . Acta Oncol. 2012 ; 51 7 : 880 – 889 . Google Scholar Crossref Search ADS PubMed 12 Helgason H , Rafnar T , Olafsdottir HS , et al. . Loss-of-function variants in ATM confer risk of gastric cancer . Nat Genet. 2015 ; 47 8 : 906 – 910 . http://dx.doi.org/10.1038/ng.3342 Google Scholar Crossref Search ADS PubMed 13 Bulik-Sullivan BK , Loh P-R , Finucane HK , et al. . LD Score regression distinguishes confounding from polygenicity in genome-wide association studies . Nat Genet. 2015 ; 47 3 : 291 – 295 . http://dx.doi.org/10.1038/ng.3211 Google Scholar Crossref Search ADS PubMed 14 Kristinsson JO , van Westerveld P , te Morsche RHM , et al. . Cyclooxygenase-2 polymorphisms and the risk of esophageal adeno- or squamous cell carcinoma . World J Gastroenterol. 2009 ; 15 28 : 3493 – 3497 . http://dx.doi.org/10.3748/wjg.15.3493 Google Scholar Crossref Search ADS PubMed 15 Lacko M , Roelofs HMJ , Te Morsche RHM , et al. . Genetic polymorphism in the conjugating enzyme UGT1A1 and the risk of head and neck cancer . Int J Cancer. 2010 ; 127 12 : 2815 – 2821 . http://dx.doi.org/10.1002/ijc.25296 Google Scholar Crossref Search ADS PubMed 16 Galesloot TE , Vermeulen SH , Swinkels DW , et al. . Cohort Profile: The Nijmegen Biomedical Study (NBS) . Int J Epidemiol. 2017 ; 46 4 : 1099 – 1100 . Google Scholar PubMed 17 Kutyavin IV , Milesi D , Belousov Y , et al. . A novel endonuclease IV post-PCR genotyping system . Nucleic Acids Res. 2006 ; 34 19 : e128 . Google Scholar Crossref Search ADS PubMed 18 Hofman A , Brusselle GGO , Darwish Murad S , et al. . The Rotterdam Study: 2016 objectives and design update . Eur J Epidemiol. 2015 ; 30 8 : 661 – 708 . http://dx.doi.org/10.1007/s10654-015-0082-x Google Scholar Crossref Search ADS PubMed 19 Verkouteren JAC , Pardo LM , Uitterlinden AG , Hofman A , Nijsten T. Common variants affecting susceptibility to develop multiple basal cell carcinomas . J Invest Dermatol. 2015 ; 135 8 : 2135 – 2138 . Google Scholar Crossref Search ADS PubMed 20 Genome of the Netherlands Consortium . Whole-genome sequence variation, population structure and demographic history of the Dutch population . Nat Genet. 2014 ; 46 8 : 818 – 825 . http://dx.doi.org/10.1038/ng.3021 Crossref Search ADS PubMed 21 Delaneau O , Howie B , Cox AJ , Zagury J-F , Marchini J. Haplotype estimation using sequencing reads . Am J Hum Genet. 2013 ; 93 4 : 687 – 696 . Google Scholar Crossref Search ADS PubMed 22 Howie BN , Donnelly P , Marchini J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies . PLoS Genet. 2009 ; 5 6 : e1000529 . Google Scholar Crossref Search ADS PubMed 23 Marchini J , Howie B , Myers S , McVean G , Donnelly P. A new multipoint method for genome-wide association studies by imputation of genotypes . Nat Genet. 2007 ; 39 7 : 906 – 913 . Google Scholar Crossref Search ADS PubMed 24 Winkler TW , Day FR , Croteau-Chonka DC , et al. . Quality control and conduct of genome-wide association meta-analyses . Nat Protoc. 2014 ; 9 5 : 1192 – 1212 . http://dx.doi.org/10.1038/nprot.2014.071 Google Scholar Crossref Search ADS PubMed 25 Mantel N , Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease . J Natl Cancer Inst. 1959 ; 22 4 : 719 – 748 . Google Scholar PubMed 26 Gretarsdottir S , Helgason H , Helgadottir A , et al. . A splice region variant in LDLR lowers non-high density lipoprotein cholesterol and protects against coronary artery disease . PLoS Genet. 2015 ; 11 9 : e1005379 . Google Scholar Crossref Search ADS PubMed 27 Mazoyer S , Dunning AM , Serova O , et al. . A polymorphic stop codon in BRCA2 . Nat Genet. 1996 ; 14 3 : 253 – 254 . http://dx.doi.org/10.1038/ng1196-253 Google Scholar Crossref Search ADS PubMed 28 Verhoog LC , van den Ouweland AM , Berns E , et al. . Large regional differences in the frequency of distinct BRCA1/BRCA2 mutations in 517 Dutch breast and/or ovarian cancer families . Eur J Cancer. 2001 ; 37 16 : 2082 – 2090 . http://dx.doi.org/10.1016/S0959-8049(01)00244-1 Google Scholar Crossref Search ADS PubMed 29 Thorlacius S , Olafsdottir G , Tryggvadottir L , et al. . A single BRCA2 mutation in male and female breast cancer families from Iceland with varied cancer phenotypes . Nat Genet. 1996 ; 13 1 : 117 – 119 . Google Scholar Crossref Search ADS PubMed 30 Rafnar T , Benediktsdottir KR , Eldon BJ , et al. . BRCA2, but not BRCA1, mutations account for familial ovarian cancer in Iceland: A population-based study . Eur J Cancer. 2004 ; 40 18 : 2788 – 2793 . http://dx.doi.org/10.1016/j.ejca.2004.09.008 Google Scholar Crossref Search ADS PubMed 31 Tryggvadóttir L , Vidarsdóttir L , Thorgeirsson T , et al. . Prostate cancer progression and survival in BRCA2 mutation carriers . J Natl Cancer Inst. 2007 ; 99 12 : 929 – 935 . Google Scholar Crossref Search ADS PubMed 32 Howlett NG , Taniguchi T , Olson S , et al. . Biallelic inactivation of BRCA2 in Fanconi anemia . Science. 2002 ; 297 5581 : 606 – 609 . http://dx.doi.org/10.1126/science.1073834 Google Scholar Crossref Search ADS PubMed 33 Singh G , Lykke-Andersen J. New insights into the formation of active nonsense-mediated decay complexes . Trends Biochem Sci. 2003 ; 28 9 : 464 – 466 . http://dx.doi.org/10.1016/S0968-0004(03)00176-2 Google Scholar Crossref Search ADS PubMed 34 Mikaelsdottir EK , Valgeirsdottir S , Eyfjord JE , Rafnar T. The Icelandic founder mutation BRCA2 999del5: Analysis of expression . Breast Cancer Res. 2004 ; 6 4 : R284 – R290 . Google Scholar Crossref Search ADS PubMed 35 Moynahan ME , Pierce AJ , Jasin M. BRCA2 is required for homology-directed repair of chromosomal breaks . Mol Cell. 2001 ; 7 2 : 263 – 272 . http://dx.doi.org/10.1016/S1097-2765(01)00174-5 Google Scholar Crossref Search ADS PubMed 36 Schlacher K , Christ N , Siaud N , Egashira A , Wu H , Jasin M. Double-strand break repair-independent role for BRCA2 in blocking stalled replication fork degradation by MRE11 . Cell. 2011 ; 145 4 : 529 – 542 . Google Scholar Crossref Search ADS PubMed 37 Davies OR , Pellegrini L. Interaction with the BRCA2 C terminus protects RAD51-DNA filaments from disassembly by BRC repeats . Nat Struct Mol Biol. 2007 ; 14 6 : 475 – 483 . http://dx.doi.org/10.1038/nsmb1251 Google Scholar Crossref Search ADS PubMed 38 Näslund-Koch C , Nordestgaard BG , Bojesen SE. Increased risk for other cancers in addition to breast cancer for CHEK2*1100delC heterozygotes estimated from the Copenhagen General Population Study . J Clin Oncol. 2016 ; 34 11 : 1208 – 1216 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI: Journal of the National Cancer Institute Oxford University Press

Loading next page...
 
/lp/ou_press/association-of-brca2-k3326-with-small-cell-lung-cancer-and-squamous-KJ4q3v01oC
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
0027-8874
eISSN
1460-2105
D.O.I.
10.1093/jnci/djy002
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Most pathogenic mutations in the BRCA2 gene carry a high risk of hereditary breast and ovarian cancer (HBOC). However, a stop-gain mutation, K3326* (rs11571833), confers risk of lung cancer and cancers of the upper-aero-digestive tract but only a modest risk of breast or ovarian cancer. The Icelandic population provides an opportunity for comprehensive characterization of the cancer risk profiles of K3326* and HBOC mutations because a single mutation, BRCA2 999del5, is responsible for almost all BRCA2-related HBOC in the population. Methods Genotype information on 43 641 cancer patients and 370 971 control subjects from Iceland, the Netherlands, and the United States was used to assess the cancer risk profiles of K3326* and BRCA2 999del5. BRCA2 expression was assessed using RNAseq data from blood (n = 2233), as well as 52 tissues reported in the GTEx database. Results The cancer risks associated with K3326* are fundamentally different from those associated with 999del5. We report for the first time an association between K3326* and small cell lung cancer (odds ratio [OR] = 2.06, 95% confidence interval [CI] = 1.35 to 3.16) and squamous cell carcinoma of the skin (OR = 1.69, 95% CI = 1.26 to 2.26). Individuals homozygous for K3326* reach old age and have children. Unlike BRCA2 999del5, the K3326* allele does not affect the level of BRCA2 transcripts, and the allele is expressed to the same extent as the wild-type allele. Conclusions K3326* associates primarily with cancers that have strong environmental genotoxic risk factors. Expression of the K3326* allele suggests that a variant protein may be made that retains the DNA repair capabilities important to hormone-responsive tissues but may be less efficient in responding to genotoxic stress. Pathogenic mutations in BRCA2 predispose to hereditary breast and ovarian cancer (HBOC) syndrome, characterized by greatly increased risk of cancers of the breast and ovary as well as other cancers (1). However, HBOC-associated mutations do not increase risk of lung cancer, suggesting that lung epithelium may be less dependent on BRCA2 function than the tissues involved in HBOC. It was therefore unexpected when a stop-gain variant close to the 3’ end of the BRCA2 gene, rs11571833 (NM_000059.3:c.9976A>T; NP_000050.2:p.Lys3326Ter, hereafter referred to as K3326*), was reported to confer risk of lung cancer (2). Analysis of 21 594 lung cancer patients and 54 156 control subjects of European origin found that carriers of the variant had an odds ratio (OR) of 1.83 (95% confidence interval [CI] = 1.61 to 2.09) of developing the disease (2). The same study also showed that the association was stronger with squamous cell lung carcinoma (SQLC) than adenocarcinoma of the lung (OR = 2.47, 95% CI = 2.03 to 3.00; OR = 1.47, 95% CI = 1.19 to 1.82, respectively). K3326* is located in the last of the 27 exons of the BRCA2 gene and is predicted to result in the loss of the 93 C-terminal amino acids of the protein product. In addition to its association with lung cancer risk, K3326* associates with substantial risk of cancers of the upper aero-digestive tract (UADT; OR = 2.53, 95% CI = 1.89 to 3.38) (3) and esophageal squamous cell carcinoma (OR = 6.0, 95% CI = 1.3 to 28) (4). Unlike variants in BRCA2 that associate with HBOC, the K3326* variant has a small effect on hormone-related cancers (5,6). A recent study including 76 637 cancer patients and 83 796 control subjects showed a modest increase in risk of breast cancer (OR = 1.28, 95% CI = 1.17 to 1.40) and invasive ovarian cancer (OR = 1.26, 95% CI = 1.10 to 1.43) (6). No association with prostate cancer was observed. Finally, K3326* was reported to be more common in familial pancreatic cancer patients than in control subjects (144 cancer patients and 250 control subjects, OR = 4.84, 95% CI = 1.27 to 18.55) (7). The difference between the cancer risk profiles of K3326* and variants in BRCA2 that associate with HBOC could provide insights into how the roles of BRCA2 differ between tissues. However, dissection of this phenomenon is complicated because rare pathogenic mutations may reside on the background of K3326*, affecting the risk estimates for the variant. The Icelandic population is ideally suited to the characterization of cancer risk and clinical presentation of K3326*. This is because a single founder mutation in BRCA2, commonly referred to as 999del5 (rs80359671, NM_000059.3:c.767_771delCAAAT, NP_000050.2:p.Asn257Lysfs), is responsible for virtually all BRCA2-related HBOC in the population, and this mutation is never on the same chromosome as K3326* (2). Here, we combined extensive information on genetic variation in the Icelandic population with data from the nation-wide cancer registry in order to compare the cancer risk profiles of K3326* and HBOC mutations. Methods This study used genotype information on a total of 43 641 cancer patients and 370 971 control subjects from Iceland, the Netherlands, and the United States. An in-depth description of the study populations and methods is presented in the Supplementary Methods (available online). Icelandic Study Population This study is based on extensive genetic information on the Icelandic population, which has been previously described (8). The whole genomes of 15 220 Icelanders were sequenced, unveiling 40 780 213 single nucleotide polymorphisms (SNP) and short indels. These variants were imputed into 151 677 Icelanders whose DNA had been genotyped with various Illumina SNP chips and phased using long-range phasing (9,10). Genealogical deduction of carrier status of 282 894 untyped relatives of chip-typed individuals further increased the sample size for association analysis. Information on cancer in the genotyped individuals is from the population-based Icelandic Cancer Registry (ICR) (11). A total of 42 331 Icelandic cancer patients and 354 488 control subjects were used in the analysis. With the aid of hospital charts, 53 genotyped lung cancer K3326* carriers diagnosed in the years 2000–2015 were staged according to the International Association for the Study of Lung Cancer 7th edition of TNM staging. These were compared with the stages of all 155 lung cancer patients diagnosed in 2009 in Iceland. The study was approved by the National Bioethics Committee of Iceland (ref. 12–122-V7). Written informed consent was obtained from all genotyped subjects. Statistical Analysis Association results for the BRCA2 variants tested in the Icelandic cancer patients come from genome-wide association studies (GWAS) on the 20 cancers. The methods used for association testing in the Icelandic population have been described in detail (12). To test for association between SNPs and cancer in the Icelandic study, logistic regression was used, treating disease status as the response and genotype counts as covariates. Other relevant covariates that might correlate with disease status were also included in the model as nuisance, for example, sex, county of birth, current age or age at death (first- and second-order terms included), blood sample availability for the individual, and an indicator function for the overlap of the lifetime of the individual with the time span of phenotype collection. To account for inflation in test statistics due to cryptic relatedness and stratification in the Icelandic population, we applied the method of linkage disequilibrium (LD) score regression (13). With a set of 1.1 M variants, we regressed the χ2 statistics from our GWAS scan against LD score and used the intercept as a correction factor. The LD scores were downloaded from an LD score database (see the URL in the “Notes”), and the estimated correction factors are listed in Supplementary Table 1 (available online). All statistical tests were two-sided, and a P value of less than .05 was considered statistically significant unless otherwise noted. UADT Cancer Patients and Control Subjects In total, 696 head and neck cancers and 119 squamous cell esophageal carcinomas were collected in two studies (14,15). Information on the ICD codes for the UADT populations used and referred to can be found in Supplementary Table 2 (available online). The 4139 Dutch control subjects were from the Nijmegen Biomedical Study (16). The protocols of all the studies were approved by the respective institutional review boards, and all study subjects gave written informed consent. In the UADT cancer patients and control subjects, the rs11571833 (K3326*) variant was genotyped using the Centaurus (Nanogen) platform (17). The c.6275delTT and c.4889C>G mutations were genotyped by Sanger sequencing. Primer sequences are listed in Supplementary Table 3 (available online). Squamous Cell Skin Cancer Patients and Control Subjects Association results for squamous cell skin cancer (SQCSC) and the BRCA2 variants were looked up in two GWAS studies, one from Rotterdam and the other from Ohio. The Rotterdam Study The Rotterdam Study (RS) is a prospective population-based follow-up study of the determinants and prognosis of chronic diseases in the middle age and elderly participants living in the Ommoord district (Rotterdam, the Netherlands) (18). The Medical Ethics Committee of the Erasmus Medical Center and the review board of the Dutch Ministry of Health, Welfare and Sports have ratified the RS. Written informed consent was obtained from all participants. Details on the ascertainment of skin cancer in the RS is presented elsewhere (19). Briefly, all RS participants with informed consent (n = 14 628) were linked to the nationwide registry of histo- and cytopathology in the Netherlands (PALGA; up to December 31, 2013). All prevalent SQCSCs were labeled as patients. Participants from the RS without PALGA reports were considered control subjects. Cohorts RS-I and RS-II were genotyped with the Infinium II HumanHap550K Genotyping BeadChip version 3 (Illumina, San Diego, CA), and the cohort RS-III was genotyped using the Illumina Human 610 Quad Arrays. The three data sets were aligned to the same strand as the GoNL data set (20) and restricted to a common set of 452 673 autosomal variants. Variants with a minor allele frequency of less than 0.5% (nine variants) that fail (P < 1e-7) the Hardy-Weinberg test and that have statistically significant differences (P < 1e-6) in frequency between data sets were excluded. The resulting genotypes for 450 395 variants were phased using SHAPEIT (v2.790) (21). Imputation was done with IMPUTE2 (v2.3.2) (22), using the GoNL data set as reference (https://molgenis26.target.rug.nl/downloads/gonl_public/variants/release5/) (20). A logistic case–control study GWAS was carried out on 398 SQCSC patients and 10 629 control subjects of Dutch origin, using SNPTEST (v2.5.2) (23) with a frequentist (additive) and expected dosage as main parameters. The model was adjusted for age at the start of the study, sex, and four principal components. Further quality control of the GWAS results included the removal of markers with low imputation quality (R2 < 0.3), duplicated markers, and markers with high betas or standard errors (larger than 10), using EasyQC (24). The rs11571833 (K3326*) variant is not present in GoNL. Therefore, we used rs11571815, which is fully correlated with rs11571833 (R2 = 1) in Northern Europeans from Utah (CEU) and Iceland as a proxy. The Ohio Study This study was approved by the Ohio State University Institutional Review Board. All study participants signed informed consent. Genomic DNA was collected from 103 SQCSC patients ascertained from dermatology clinics. Control subjects (n = 1715) were obtained from the Columbus area control study and were individuals without a self-reported SQCSC diagnosis and were matched to the population and banked patients by sex and age of diagnosis. Samples were genotyped on Illumina chips as described above, phased using SHAPEIT (v2.790), and imputed using 1000 Genomes Phase 3 (October 2014) as a reference set. Test for Heterogeneity Between the Different Populations The results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors (Supplementary Methods, available online) (25). RNA Analysis Preparation of poly(A)+ cDNA sequencing libraries and RNA sequencing were carried out as described previously (26). A detailed protocol for the RNA analysis is presented in the Supplementary Methods (available online). Results To compare the cancer risks associated with K3326* and 999del5, we tested the associations between the variants and 20 cancer types (Table 1). As expected, very strong associations were observed between 999del5 and cancers of the breast, ovary, prostate, and pancreas. In addition, squamous UADT cancers showed statistically significant association with 999del5 (20 tests, statistical significance threshold of P < 0.05/20 = 2.5 × 10-3). Several cancer types show suggestive association with 999del5, that is, SQCSC and cancers of the stomach, urinary bladder, and cervix. When all cancers are combined, the risk of 999del5 carriers being diagnosed with some form of cancer is more than fourfold that of the general population. We observed no statistically significant association between lung cancer and 999del5 (OR = 1.02, 95% CI = 0.69 to 1.52, P = .09). In contrast, K3326* associates with lung cancer (OR = 1.54, 95% CI = 1.23 to 1.91, P = 1.2 × 10−4) and squamous cell skin cancer (OR = 1.69, 95% CI = 1.26 to 2.26, P = 4.2 × 10−4). Within lung cancer, the cancer risk was confined to SQLC (OR = 1.71, 95% CI = 1.10 to 2.67, P = .02) and small cell lung cancer (SCLC; OR = 2.06, 95% CI = 1.35 to 3.16, P = 9.0 × 10−4), whereas no association to adenocarcinoma was detected in our sample set (OR = 0.98, 95% CI = 0.66 to 1.45, P = .92). The associations of K3326* with SCLC and SQCSC have not been reported before. Table 1. Association between 20 cancer types and two BRCA2 mutations in the Icelandic population; K3326* and the Icelandic founder mutation 999del5† K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 † Association results are shown for the 19 most common solid tumors in the Icelandic cancer registry (41 976 diagnoses) and all cancers combined (42 331 individuals, including some rare cancers not listed in the table), as well as for the three most common histological subtypes of lung cancer. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; MGUS = monoclonal gammopathy of undetermined significance; OR = odds ratio. ‡ The numbers of cancer patients (Naff) and control subjects (Ncontrols) correspond to the number of individuals used in the association tests. § The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. ‖ The odds ratios correspond to minor alleles. ¶ Results that remain statistically significant after correction for number of tests. Table 1. Association between 20 cancer types and two BRCA2 mutations in the Icelandic population; K3326* and the Icelandic founder mutation 999del5† K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 K3326* (MAF = 1.1%) 999del5 (MAF = 0.36%) Heterogeneity Phenotype Naff‡ Ncontr‡ P§ OR‖ (95% CI) P§ OR‖ (95% CI) P§ I2 Breast 6013 324 022 .42 1.09 (0.86 to 1.40) 1.1 × 10 to 162 18.3 (14.6 to 22.2) 1.1 × 10−77¶ 99.7 Lung 4461 244 693 1.2 × 10−4¶ 1.54 (1.23 to 1.91) .91 1.02 (0.69 to 1.52) .048 74.5  Adenocarcinoma 1713 246 959 .92 0.98 (0.66 to 1.45) .95 1.02 (0.55 to 1.90) .91 0  Squam. cell carcinoma 901 159 694 .02 1.71 (1.10 to 2.67) .20 1.60 (0.77 to 3.44) .88 0  Small cell carcinoma 800 186 762 9.0 × 10−4¶ 2.06 (1.35 to 3.16) .43 0.69 (0.27 to 1.76) .04 77.6 Gastric adenocarcinoma 2651 226 257 .96 0.99 (0.71 to 1.40 .005 1.90 (1.21 to 2.97) .04 78.1 Ovary epithelial 822 151 199 .73 0.91 (0.52 to 1.57) 2.8 × 10 to 27¶ 9.7 (6.44 to 14.68) 6.6 × 10−12¶ 97.9 Pancreas 1262 282 298 .12 1.38 (0.93 to 2.10) 2.5 × 10 to 12¶ 4.7 (3.06 to 7.38) 5.2 × 10−5¶ 93.9 Prostate 5689 102 259 .62 1.06 (0.84 to 1.34) 3.5 × 10 to 17¶ 3.8 (2.78 to 5.16) 9.6 × 10−11¶ 97.6 Urinary bladder 2010 300 813 .70 1.07 (0.76 to 1.51) .005 2.1 (1.24 to 3.40) .03 77.8 Colorectal 3987 270 271 .42 1.10 (0.86 to 1.44) .81 0.94 (0.59 to 1.52) .58 0 Upper aero-digestive tract 946 264 566 .62 1.14 (0.50 to 1.50) 2.1 × 10 to 5¶ 3.16 (1.82 to 5.89) .007 86.2 Skin squamous 2081 296 015 4.2 × 10−4¶ 1.69 (1.26 to 2.26) .004 2.2 (1.28 to 3.81) .40 0 Skin basal cell 5559 301 407 .03 1.27 (1.03 to 1.56) .95 1.01 (0.67 to 1.53) .23 29.5 Melanoma 1786 309 646 .96 1.01 (0.71 to 1.43) .33 1.3 (0.75 to 2.34) .45 0 Thyroid 1209 287 439 .62 0.86 (0.54 to 1.37) .09 1.7 (0.91 to 3.23) .12 58.5 Kidney 1673 396 768 .04 1.42 (1.01 to 2.00) .70 0.88 (0.45 to 1.71) .20 38.9 Endometrium 970 118 340 .66 1.11 (0.69 to 1.78) .75 1.2 (0.48 to 2.76) .90 0 Cervix 849 144 382 .15 1.42 (0.88 to 2.27) .007 2.4 (1.27 to 4.47) .19 40.6 Brain (glioma) 723 374 314 .88 0.95 (0.51 to 1.77) .63 0.8 (0.32 to 2.00) .76 0 Non-Hodgkin lymphoma 1176 363 693 .33 1.23 (0.81 to 1.88) .13 0.45 (0.16 to 1.27) .08 68 MGUS 1770 305 079 .75 0.94 (0.64 to 1.38) .69 1.1 (0.60 to 2.17) .61 0 All cancer types 42 331 354 488 1.6 × 10−5¶ 1.23 (1.12 to 1.35) 3.8 × 10 to 88¶ 4.2 (3.66 to 4.86) 1.2 × 10−45¶ 99.5 † Association results are shown for the 19 most common solid tumors in the Icelandic cancer registry (41 976 diagnoses) and all cancers combined (42 331 individuals, including some rare cancers not listed in the table), as well as for the three most common histological subtypes of lung cancer. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; MGUS = monoclonal gammopathy of undetermined significance; OR = odds ratio. ‡ The numbers of cancer patients (Naff) and control subjects (Ncontrols) correspond to the number of individuals used in the association tests. § The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. ‖ The odds ratios correspond to minor alleles. ¶ Results that remain statistically significant after correction for number of tests. To refine our estimate of the association of K3326* with SQCSC, we genotyped the variant in two additional SQCSC sample sets from the Netherlands and the United States (Table 2). Combining the results from the three sample sets gives a Pcombined value of 1.7 × 10−4 (OR = 1.66, 95% CI = 1.27 to 2.16, Phet = .91). Table 2. Association between K3326* and SQCSK and UADT cancer in Iceland, the Netherlands, and the United States Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 † The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; OR = odds ratio; SQCSC = squamous cell skin cancer; UADT = upper aero-digestive tract. ‡ Imputation information for marker estimated by the ratio of the variance of imputed expected allele counts and the variance of the actual allele counts. § The variant rs11571815 was used as a surrogate for rs1157833 (K3326* variant) in the Dutch SQCSC population because the GoNL data set does not contain rs1157833. The two variants are fully correlated in CEU (R2 = 1). ‖ rs1157833 was directly genotyped in the Dutch UADT patients and control subjects. Table 2. Association between K3326* and SQCSK and UADT cancer in Iceland, the Netherlands, and the United States Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 Sample set No. of cases No. of controls MAF cases MAF controls P† OR (95% CI) Info‡ Phet I2 SQCSC  Iceland 2081 296 015 0.019 0.011 4.2 × 10−4 1.69 (1.26 to 2.26) 1.00  The Netherlands* 398 10 629 0.013 0.008 .19 1.62 (0.79 to 3.3) 0.92  Ohio, USA 103 1715 0.021 0.016 .68 1.29 (0.38 to 4.35) 0.97  Comb. Iceland, Netherlands, USA 2582 308 359 1.7 × 10−4 1.66 (1.27 to 2.16) .91 UADT cancer  The Netherlands‖ 809 4139 0.019 0.007 5.0 × 10−5 2.68 (1.67 to 4.31)  Iceland 946 264 566 0.013 0.011 .62 1.14 (0.50 to 1.50) 1.00  Comb. Iceland Netherlands 1755 268 705 9.0 × 10−4 1.81 (1.27 to 2.57) .02 82.4 † The association was tested using logistic regression, and results for the multiple case–control groups were combined using a fixed-effects inverse variance method based on effect estimates and standard errors. Heterogeneity in the effect estimate was tested using a Cochran’s Q test statistic. All P values reported are two-sided. Odds ratios are based on population control subjects. CI = confidence interval; MAF = minor allele frequency; OR = odds ratio; SQCSC = squamous cell skin cancer; UADT = upper aero-digestive tract. ‡ Imputation information for marker estimated by the ratio of the variance of imputed expected allele counts and the variance of the actual allele counts. § The variant rs11571815 was used as a surrogate for rs1157833 (K3326* variant) in the Dutch SQCSC population because the GoNL data set does not contain rs1157833. The two variants are fully correlated in CEU (R2 = 1). ‖ rs1157833 was directly genotyped in the Dutch UADT patients and control subjects. A strong association between the 999del5 variant and UADT cancers was observed in Iceland (OR = 3.16, 95% CI = 1.82 to 5.89, P = 2.1 × 10-5). This association was driven primarily by squamous esophageal cancer (OR = 5.71, 95% CI = 2.63 to 12.41, P = 1.1 × 10-5, 366 patients, 186 714 control subjects). However, contrary to previous reports (3,4), there was no statistically significant association observed between UADT cancers and K3326* in Iceland. To exclude systematic errors in the registration of this cancer group, a pathology review was performed on a random set of 30 samples from different periods. The pathology review confirmed the diagnoses of all 30 samples. In line with published results, we observed a strong association of K3326* with UADT cancers in samples from the Netherlands (OR = 2.68, 95% CI = 1.67 to 4.31, P = 5.0 × 10-5) (Table 2). Comparing the distribution of cancer sites between the Icelandic and Dutch UATD cancer patients, we note a higher fraction of oral cavity cancer in the Icelandic patients and a higher fraction of cancers of the oropharynx and hypopharynx in the Dutch patients (Supplementary Table 4, available online). However, this difference does not explain the difference in cancer risk between the populations because K3326* is associated with risk of all subtypes of UADT cancers in our Dutch data set (Supplementary Table 5, available online) as well as in the published report (3). Previous studies of families with breast cancer of northern European ancestry show that two rare mutations, BRCA2 c.6275delTT and c.4889C>G, which are highly penetrant for breast and ovarian cancer, occur on the K3326* haplotype (27). These mutations have not been found in Iceland. However, the BRCA2 c.6275delTT mutation has been found in several HBOC families in the Netherlands (28), prompting speculations that these mutations could be responsible for the UADT signal, rather than the K3326* variant in the Dutch population. We genotyped the c.6275delTT and c.4889C>G mutations in the Dutch UADT cancer sample set using Sanger sequencing (Supplementary Table 6, available online). Neither mutation was found, suggesting that the K3326* variant alone explains the association with UADT cancer in the Netherlands. Carriers of HBOC mutations in BRCA2 are on average diagnosed with breast, ovarian, and prostate cancers at a younger age than noncarriers (29–31). The K3326* variant did not affect the age at diagnosis for lung cancer (Supplementary Table 7, available online). When considering the whole Icelandic population, carriers of K3326* have a slightly shorter lifespan than noncarriers among deceased individuals who lived to be age 50 years and were born after 1890 (P = .01, β = –0.063, n = 118 626). Considering lung cancer patients only, the ratio of males to females among carriers of K3326* was not statistically significantly different from the ratio of males to females with lung cancer in the total cancer registry (55.5% vs 51.9%, P = .54). Bi-allelic mutations in BRCA2 result in Fanconi anemia (FA) complementation group D1 (FA-D1) (32), a chromosome instability disorder that causes childhood-onset aplastic anemia and cellular hypersensitivity to DNA crosslinking agents, as well as a strong cancer predisposition. Although the cancer risk profile of the K3326* mutation is more modest than that of the HBOC mutations, it is possible that homozygosity for the variant could have consequences over and beyond a multiplicative risk for the cancers observed in heterozygotes. In total, 17 homozygous individuals were observed in our population of 151 677 chip-typed Icelanders; this is in agreement with the expectation under the Hardy-Weinberg equilibrium (17 expected), given the allelic frequency of 1.1% for K3326*. Homozygous individuals reach adulthood (oldest 92 years) and have children (Table 3). Information on cancer in the 17 K3326* homozygotes, as well as the seven compound K3326*/999del5 heterozygous carriers observed, is listed in Table 3. No homozygotes for the 999del5 mutation were observed in 151 677 chip-typed individuals (expected to be two under the Hardy-Weinberg equilibrium and allelic frequency of 0.36%). Table 3. Information on homozygous carriers of K3326* and compound heterozygous carriers of K3326* and 999del5† Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 † Age at dx = age at diagnosis; BCC = basal cell carcinoma; SCLC = small cell lung cancer; SQCSC = squamous cell skin cancer. Table 3. Information on homozygous carriers of K3326* and compound heterozygous carriers of K3326* and 999del5† Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 Individual No. Year of birth Year of death Age at diagnosis, y Cancer type No. of children Homozygous carriers of K3326* 1 1924 2002 — — 5 2 1928 — 82 Skin (SQCSC) 5 3 1937 2002 64 Lung (SCLC) 5 4 1939 — 60/68 Colon/rectum 4 5 1942 — — — 3 6 1945 2012 66 Lung (SQCLC) 3 7 1948 — 58 Skin (BCC) 3 8 1956 — 57 Pancreas carcinoid tumor 0 9 1957 — — — 0 10 1959 — — — 2 11 1967 — — — 4 12 1971 — 42 Skin (BCC) 1 13 1983 — — — 3 14 1990 — — — 0 15 1992 — — — 0 16 1994 — — — 0 17 2000 — — — 0 Compound heterozygous carriers of K3326 and 999del5 1 1914 2011 80/83 Kidney/skin (SQCSC) 2 2 1936 2010 55/74 Breast/pancreas 4 3 1950 — — — 2 4 1959 — 44 Breast 4 5 1978 — — — 3 6 1982 — — — 2 7 1993 — — — 0 † Age at dx = age at diagnosis; BCC = basal cell carcinoma; SCLC = small cell lung cancer; SQCSC = squamous cell skin cancer. We tested if the stage distribution of carriers of K3326* was different from noncarriers. Because disease stage is not registered in the ICR, we retrieved the medical records and registered the stage of disease of all 53 carriers of K3326* diagnosed with LC in the years 2000–2015 as well as all 155 LC patients diagnosed in the year 2009 for comparison. We found no difference in the stage distribution between carriers and noncarriers with NSCLC (Supplementary Table 8, available online). The effect of K3326* on the function of the BRCA2 protein in different tissues remains to be determined. As the variant introduces a stop codon in the last exon of the gene, the mRNA will likely escape nonsense-mediated decay (NMD), allowing translation of a truncated protein product (33). To assess the effects of the K3326* and 999del5 variants on transcript levels, BRCA2 mRNA was assessed in RNA-Seq data from the blood of individuals with wild-type BRCA2 (n = 2181), individuals heterozygous for K3326* (n = 40), and individuals homozygous for K3326* (n = 2), as well as individuals heterozygous for 999del5 (n = 10). As previously reported (34), heterozygous carriers of the 999del5 mutation have fewer BRCA2 transcripts than noncarriers, in line with 999del5 inducing NMD (Figure 1). In contrast, carriers of K3326* express normal levels of BRCA2 mRNA, and no allelic imbalance in expression was detected for heterozygous individuals (87 reference vs 90 alternative total number of fragments). This indicates that transcripts carrying the K3326* are not subjected to NMD and that a truncated BRCA2 protein may be generated. Similarly, using data from GTEx, K3326* was not found to affect BRCA2 transcript levels in other tissues, such as lung (Supplementary Figure 1, available online), sun-exposed skin, subcutaneous adipose tissue, thyroid, or whole blood (Supplementary Table 9, available online). Figure 1. View largeDownload slide Expression of BRCA2 in whole blood by genotypes of K3326* and 999del5. Expression values are quantile-normalized and mapped to a standard normal distribution. The bottom and top of the boxes correspond to the 25th (Q1) and 75th (Q3) percentiles, the line inside the box corresponds to the median, and the whiskers are located at max(min(Expression), Q1 – 1.5 IQR) and min(max(Expression), Q3 + 1.5 IQR), respectively (where IQR is the interquartile range = Q3 – Q1). Linear regression, assuming an additive model for genotype effects and including both variants in the model, estimated a β of –0.71 (95% confidence interval [CI] =–1.30 to –0.11, P=.02) for 999del5 and a β of 0.22 (95% CI = –0.06 to 0.49, P= .13) for K3326*. Figure 1. View largeDownload slide Expression of BRCA2 in whole blood by genotypes of K3326* and 999del5. Expression values are quantile-normalized and mapped to a standard normal distribution. The bottom and top of the boxes correspond to the 25th (Q1) and 75th (Q3) percentiles, the line inside the box corresponds to the median, and the whiskers are located at max(min(Expression), Q1 – 1.5 IQR) and min(max(Expression), Q3 + 1.5 IQR), respectively (where IQR is the interquartile range = Q3 – Q1). Linear regression, assuming an additive model for genotype effects and including both variants in the model, estimated a β of –0.71 (95% confidence interval [CI] =–1.30 to –0.11, P=.02) for 999del5 and a β of 0.22 (95% CI = –0.06 to 0.49, P= .13) for K3326*. Discussion Here, we used the comprehensive information on genetic variation in the Icelandic population, along with information on 49 051 diagnoses of 20 different cancers to dissect and compare the cancer risks of the K3326* variant and the 999del5 founder mutation that behaves like a typical HBOC mutation. We confirm the previously reported associations with total LC and SQLC. In addition, we demonstrate associations with SCLC and SQCSC not previously reported. The reported association between K3326* and UADT cancer was not replicated in the Icelandic population, although a strong effect of the 999del variant on this cancer group was observed. Possible explanations of the difference in UADT cancer risk between Iceland and the other populations include different etiological factors, such as infection with human papilloma virus (HPV). HPV associates most strongly with oropharyngeal cancers, and this cancer subtype is less common in the Icelandic population than in the Dutch data set used here. However, K3326* was previously reported to associate with all subtypes of UADT cancers (oral cavity, oropharynx, larynx/hypopharynx, and esophagus) (3), and these associations are replicated in our Dutch sample set. In the Icelandic UADT cancer group, the strongest effect of K3326* (albeit statistically nonsignificant) was with oropharyngeal cancer. It is of interest that K3326* associates with cancer types that have strong environmental genotoxic risk factors, that is, smoking for SQLC, SCLC and UADT cancer, and ultraviolet (UV) radiation for SQCSC. This suggests that the DNA repair capability of the truncated BRCA2 protein may be impaired, rendering affected individuals more vulnerable to genotoxic stress. BRCA2 has two well-characterized roles in the maintenance of genomic integrity, homology-directed repair (HDR) of double-strand breaks (35) and a role in preventing degradation of stalled replication forks (36). In both cases, the function of BRCA2 is dependent on the interaction with RAD51. The RAD51 interaction domain needed to stabilize stalled replication forks is located between amino acids 3270 and 3305 in exon 27 (37), in close proximity to K3326*, prompting speculations that the truncation might interfere with the interaction between RAD51 and the C-terminus of BRCA2. This would affect the ability of BRCA2 to prevent the MRE11-mediated degradation of stalled replication forks, resulting in genomic instability and increased cancer risk. Such a defect would be more likely to have a biological impact under stressful conditions where replication fork progression is constantly being challenged, for instance, in the presence of bulky DNA adducts caused by frequent exposure to carcinogenic agents such as tobacco smoke or UV light. On the contrary, studies have shown that the BRCA2 C-terminus RAD51 interaction domain is dispensable for HDR repair (36), which may explain why the K3326* variant does not have a strong effect on breast or ovarian cancer risks. Our population-based assessment of the 999del5 founder mutation highlights the complexity of cancer risks conferred by loss-of-function mutations in BRCA2. Current evidence suggests that other cancer susceptibility genes, many of which have a role in DNA repair, also have complex risk patterns that still need to be characterized in full (38). Dissection of the risk patterns of these genes is likely to have important implications for the genetic counseling of mutation carriers, which has historically been strongly focused on a limited number of cancer types. The limitation of this study lies in the lack of power for rare cancer types. Additional studies with more patients will provide more robust estimates of the risk for these cancer types. Also, due to the low frequency of the K3326* variant, it was not possible to test if the variant affects survival of lung cancer patients. In conclusion, our study shows that the K3326* variant in BRCA2 associates primarily with cancer types that have strong environmental genotoxic risk factors. We report for the first time an association between the variant and risk of SCLC and SQCSC. The risk profile of K3326* suggests that while the variant protein retains the ability to maintain genomic stability in hormone-related tissues, it may be less efficient in repairing damages caused by carcinogenic agents such as tobacco smoke or UV light. Funding The study was funded by deCODE genetics/Amgen. The Ohio Study on SQCSC was supported in part by the American Cancer Society (RSG-07-083 MGO to AET), the National Institutes of Health (R03 CA173788), and by the National Cancer Institute (grant P30 CA16058 to the OSU Comprehensive Cancer Center). Notes Affiliations of authors: deCODE Genetics/Amgen, Reykjavik, Iceland (TR, SNS, GH, PS, HH, AS, JG, JS, JKS, HJ, DFG, GM, UT, GT, KS); Department of Oncology (GRS, KA, OTJ, FS, GBR), Department of Pathology (JGJ, HI), Department of Dermatology (BS, KT, JHO), Department of Medicine (HH, SJ), and Department of Surgery (TG), Landspitali-University Hospital, Reykjavik, Iceland; Department of Dermatology, Erasmus MC Cancer Institute, Rotterdam, the Netherlands (LMP, TN); Faculty of Electrical and Computer Engineering (HH), Faculty of Medicine (STS, TG, LT, JGJ, BS, KT, TG, JHO, UT, KS), and School of Engineering and Natural Sciences (DFG), University of Iceland, Reykjavik, Iceland; Icelandic Cancer Registry, Reykjavik, Iceland (LT, GHO, KA); Department of Internal Medicine, Erasmus MC University Medical Center, Rotterdam, the Netherlands (AU); Radboud University Medical Center, Radboud Institute for Health Sciences, Nijmegen, the Netherlands (SHV, TEG, LAMLK); Department of Internal Medicine (DCA), Department of Cancer Biology (AET), and Department of Genetics and Internal Medicine (AET), Division of Human Genetics, and The Comprehensive Cancer Center (AET), The Ohio State University, Columbus, OH (DCA); Department of Otorhinolaryngology, Head and Neck Surgery, Maastricht University Medical Center, Maastricht, the Netherlands (ML); Department of Gastroenterology, Radboud University Nijmegen Medical Center, Nijmegen, the Netherlands (WHMP). The funders had no role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the manuscript for publication. All authors from deCODE are employees of deCODE genetics/Amgen. From Ohio: The Human Genetics Sample bank was essential for processing of samples for DNA. We thank Victoria Klee and Madison Bernhardt for ascertainment of medical records and database management and Elizabeth Solinger for accrual of control individuals for this study. We thank Dr. Albert de la Chapelle for provision of some of the control samples from the Ohio set. From the Rotterdam Study: The authors are grateful to the study participants, the staff from the Rotterdam Study, and the participating general practitioners and pharmacists. We thank the staff from Internal Medicine for generation and maintenance of GWAS data (P. Arp, M. Jhamai, M. Verkerk, L. Herrera, M. Peters). We also thank E. den Broek, L. Overbeek, and S. Koljenovic for help in the linkage with the pathology data from PALGA. Data can be obtained upon request. Requests should be directed to the management team of the Rotterdam Study (secretariat.epi@erasmusmc.nl), which has a protocol for approving data requests. Because of restrictions based on privacy regulations and informed consent of the participants, data cannot be made freely available in a public repository. From NBS: The Nijmegen Biomedical Study is a population-based survey conducted at the Department for Health Evidence and the Department of Laboratory Medicine of the Radboud University Medical Center. The principal investigators of the Nijmegen Biomedical Study are L. A. L. M. Kiemeney, A. L. M. Verbeek, D. W. Swinkels, and B. Franke. LD score database (accessed June 23, 2015): ftp://atguftp.mgh.harvard.edu/brendan/1k_eur_r2_hm3snps_se_weights.RDS. GTEx Portal (accessed June 12, 2016): http://www.gtexportal.org/home/. References 1 Lynch HT , Snyder C , Casey MJ. Hereditary ovarian and breast cancer: What have we learned? Ann Oncol. 2013 ; 24(suppl 8) : viii83 – viii95 . Google Scholar PubMed 2 Wang Y , McKay JD , Rafnar T , et al. . Rare variants of large effect in BRCA2 and CHEK2 affect risk of lung cancer . Nat Genet. 2014 ; 46 7 : 736 – 741 . http://dx.doi.org/10.1038/ng.3002 Google Scholar Crossref Search ADS PubMed 3 Delahaye-Sourdeix M , Anantharaman D , Timofeeva MN , et al. . A rare truncating BRCA2 variant and genetic susceptibility to upper aerodigestive tract cancer . J Natl Cancer Inst. 2015 ; 107 5 :djv037. 4 Akbari MR , Malekzadeh R , Nasrollahzadeh D , et al. . Germline BRCA2 mutations and the risk of esophageal squamous cell carcinoma . Oncogene. 2008 ; 27 9 : 1290 – 1296 . http://dx.doi.org/10.1038/sj.onc.1210739 Google Scholar Crossref Search ADS PubMed 5 Michailidou K , Hall P , Gonzalez-Neira A , et al. . Large-scale genotyping identifies 41 new loci associated with breast cancer risk . Nat Genet. 2013 ; 45 4 : 353 – 361 , 361–362. Google Scholar Crossref Search ADS PubMed 6 Meeks HD , Song H , Michailidou K , et al. . BRCA2 polymorphic stop codon K3326X and the risk of breast, prostate, and ovarian cancers . J Natl Cancer Inst. 2015 ; 108 2 :djv315. 7 Martin ST , Matsubayashi H , Rogers CD , et al. . Increased prevalence of the BRCA2 polymorphic stop codon K3326X among individuals with familial pancreatic cancer . Oncogene 2005 ; 24 22 : 3652 – 3656 . http://dx.doi.org/10.1038/sj.onc.1208411 Google Scholar Crossref Search ADS PubMed 8 Gudbjartsson DF , Helgason H , Gudjonsson SA , et al. . Large-scale whole-genome sequencing of the Icelandic population . Nat Genet. 2015 ; 47 5 : 435 – 444 . http://dx.doi.org/10.1038/ng.3247 Google Scholar Crossref Search ADS PubMed 9 Kong A , Masson G , Frigge ML , et al. . Detection of sharing by descent, long-range phasing and haplotype imputation . Nat Genet. 2008 ; 40 9 : 1068 – 1075 . http://dx.doi.org/10.1038/ng.216 Google Scholar Crossref Search ADS PubMed 10 Kong A , Steinthorsdottir V , Masson G , et al. . Parental origin of sequence variants associated with complex diseases . Nature. 2009 ; 462 7275 : 868 – 874 . http://dx.doi.org/10.1038/nature08625 Google Scholar Crossref Search ADS PubMed 11 Sigurdardottir LG , Jonasson JG , Stefansdottir S , et al. . Data quality at the Icelandic Cancer Registry: Comparability, validity, timeliness and completeness . Acta Oncol. 2012 ; 51 7 : 880 – 889 . Google Scholar Crossref Search ADS PubMed 12 Helgason H , Rafnar T , Olafsdottir HS , et al. . Loss-of-function variants in ATM confer risk of gastric cancer . Nat Genet. 2015 ; 47 8 : 906 – 910 . http://dx.doi.org/10.1038/ng.3342 Google Scholar Crossref Search ADS PubMed 13 Bulik-Sullivan BK , Loh P-R , Finucane HK , et al. . LD Score regression distinguishes confounding from polygenicity in genome-wide association studies . Nat Genet. 2015 ; 47 3 : 291 – 295 . http://dx.doi.org/10.1038/ng.3211 Google Scholar Crossref Search ADS PubMed 14 Kristinsson JO , van Westerveld P , te Morsche RHM , et al. . Cyclooxygenase-2 polymorphisms and the risk of esophageal adeno- or squamous cell carcinoma . World J Gastroenterol. 2009 ; 15 28 : 3493 – 3497 . http://dx.doi.org/10.3748/wjg.15.3493 Google Scholar Crossref Search ADS PubMed 15 Lacko M , Roelofs HMJ , Te Morsche RHM , et al. . Genetic polymorphism in the conjugating enzyme UGT1A1 and the risk of head and neck cancer . Int J Cancer. 2010 ; 127 12 : 2815 – 2821 . http://dx.doi.org/10.1002/ijc.25296 Google Scholar Crossref Search ADS PubMed 16 Galesloot TE , Vermeulen SH , Swinkels DW , et al. . Cohort Profile: The Nijmegen Biomedical Study (NBS) . Int J Epidemiol. 2017 ; 46 4 : 1099 – 1100 . Google Scholar PubMed 17 Kutyavin IV , Milesi D , Belousov Y , et al. . A novel endonuclease IV post-PCR genotyping system . Nucleic Acids Res. 2006 ; 34 19 : e128 . Google Scholar Crossref Search ADS PubMed 18 Hofman A , Brusselle GGO , Darwish Murad S , et al. . The Rotterdam Study: 2016 objectives and design update . Eur J Epidemiol. 2015 ; 30 8 : 661 – 708 . http://dx.doi.org/10.1007/s10654-015-0082-x Google Scholar Crossref Search ADS PubMed 19 Verkouteren JAC , Pardo LM , Uitterlinden AG , Hofman A , Nijsten T. Common variants affecting susceptibility to develop multiple basal cell carcinomas . J Invest Dermatol. 2015 ; 135 8 : 2135 – 2138 . Google Scholar Crossref Search ADS PubMed 20 Genome of the Netherlands Consortium . Whole-genome sequence variation, population structure and demographic history of the Dutch population . Nat Genet. 2014 ; 46 8 : 818 – 825 . http://dx.doi.org/10.1038/ng.3021 Crossref Search ADS PubMed 21 Delaneau O , Howie B , Cox AJ , Zagury J-F , Marchini J. Haplotype estimation using sequencing reads . Am J Hum Genet. 2013 ; 93 4 : 687 – 696 . Google Scholar Crossref Search ADS PubMed 22 Howie BN , Donnelly P , Marchini J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies . PLoS Genet. 2009 ; 5 6 : e1000529 . Google Scholar Crossref Search ADS PubMed 23 Marchini J , Howie B , Myers S , McVean G , Donnelly P. A new multipoint method for genome-wide association studies by imputation of genotypes . Nat Genet. 2007 ; 39 7 : 906 – 913 . Google Scholar Crossref Search ADS PubMed 24 Winkler TW , Day FR , Croteau-Chonka DC , et al. . Quality control and conduct of genome-wide association meta-analyses . Nat Protoc. 2014 ; 9 5 : 1192 – 1212 . http://dx.doi.org/10.1038/nprot.2014.071 Google Scholar Crossref Search ADS PubMed 25 Mantel N , Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease . J Natl Cancer Inst. 1959 ; 22 4 : 719 – 748 . Google Scholar PubMed 26 Gretarsdottir S , Helgason H , Helgadottir A , et al. . A splice region variant in LDLR lowers non-high density lipoprotein cholesterol and protects against coronary artery disease . PLoS Genet. 2015 ; 11 9 : e1005379 . Google Scholar Crossref Search ADS PubMed 27 Mazoyer S , Dunning AM , Serova O , et al. . A polymorphic stop codon in BRCA2 . Nat Genet. 1996 ; 14 3 : 253 – 254 . http://dx.doi.org/10.1038/ng1196-253 Google Scholar Crossref Search ADS PubMed 28 Verhoog LC , van den Ouweland AM , Berns E , et al. . Large regional differences in the frequency of distinct BRCA1/BRCA2 mutations in 517 Dutch breast and/or ovarian cancer families . Eur J Cancer. 2001 ; 37 16 : 2082 – 2090 . http://dx.doi.org/10.1016/S0959-8049(01)00244-1 Google Scholar Crossref Search ADS PubMed 29 Thorlacius S , Olafsdottir G , Tryggvadottir L , et al. . A single BRCA2 mutation in male and female breast cancer families from Iceland with varied cancer phenotypes . Nat Genet. 1996 ; 13 1 : 117 – 119 . Google Scholar Crossref Search ADS PubMed 30 Rafnar T , Benediktsdottir KR , Eldon BJ , et al. . BRCA2, but not BRCA1, mutations account for familial ovarian cancer in Iceland: A population-based study . Eur J Cancer. 2004 ; 40 18 : 2788 – 2793 . http://dx.doi.org/10.1016/j.ejca.2004.09.008 Google Scholar Crossref Search ADS PubMed 31 Tryggvadóttir L , Vidarsdóttir L , Thorgeirsson T , et al. . Prostate cancer progression and survival in BRCA2 mutation carriers . J Natl Cancer Inst. 2007 ; 99 12 : 929 – 935 . Google Scholar Crossref Search ADS PubMed 32 Howlett NG , Taniguchi T , Olson S , et al. . Biallelic inactivation of BRCA2 in Fanconi anemia . Science. 2002 ; 297 5581 : 606 – 609 . http://dx.doi.org/10.1126/science.1073834 Google Scholar Crossref Search ADS PubMed 33 Singh G , Lykke-Andersen J. New insights into the formation of active nonsense-mediated decay complexes . Trends Biochem Sci. 2003 ; 28 9 : 464 – 466 . http://dx.doi.org/10.1016/S0968-0004(03)00176-2 Google Scholar Crossref Search ADS PubMed 34 Mikaelsdottir EK , Valgeirsdottir S , Eyfjord JE , Rafnar T. The Icelandic founder mutation BRCA2 999del5: Analysis of expression . Breast Cancer Res. 2004 ; 6 4 : R284 – R290 . Google Scholar Crossref Search ADS PubMed 35 Moynahan ME , Pierce AJ , Jasin M. BRCA2 is required for homology-directed repair of chromosomal breaks . Mol Cell. 2001 ; 7 2 : 263 – 272 . http://dx.doi.org/10.1016/S1097-2765(01)00174-5 Google Scholar Crossref Search ADS PubMed 36 Schlacher K , Christ N , Siaud N , Egashira A , Wu H , Jasin M. Double-strand break repair-independent role for BRCA2 in blocking stalled replication fork degradation by MRE11 . Cell. 2011 ; 145 4 : 529 – 542 . Google Scholar Crossref Search ADS PubMed 37 Davies OR , Pellegrini L. Interaction with the BRCA2 C terminus protects RAD51-DNA filaments from disassembly by BRC repeats . Nat Struct Mol Biol. 2007 ; 14 6 : 475 – 483 . http://dx.doi.org/10.1038/nsmb1251 Google Scholar Crossref Search ADS PubMed 38 Näslund-Koch C , Nordestgaard BG , Bojesen SE. Increased risk for other cancers in addition to breast cancer for CHEK2*1100delC heterozygotes estimated from the Copenhagen General Population Study . J Clin Oncol. 2016 ; 34 11 : 1208 – 1216 . Google Scholar Crossref Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

JNCI: Journal of the National Cancer InstituteOxford University Press

Published: Sep 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off