Prevalence of germline TP53 variants among early-onset breast cancer patients from Polish population.

BACKGROUND: The objective of this study was to determine spectrum and prevalence of germline mutations in TP53 gene among Polish women with early-onset breast cancer (BC), which has not been determined until now. METHODS: A cohort of 100 females with BC diagnosed ≤ 30 years of age and with a positive family history of cancer was used as a discovery cohort. 1880 women with BC ≤ 45 years old and a control group of 2000 healthy women were genotyped as a replication phase of this study. RESULTS: Four heterozygous pathogenic missense mutations were detected in a group of 100 patients with early-onset breast cancer. On the basis of software prediction and available literature data, all these variants were defined as pathogenic. None of these TP53 variants were detected among 1880 breast cancer patients and 2000 healthy controls. No large mutations were found among early-onset cases using MLPA reaction. CONCLUSION: Germline pathogenic TP53 variants were found in 4% early-onset Polish BC patients. No founder mutations were identified in Polish population. To improve the treatment and surveillance screening, the search for germline TP53 pathogenic variants is recommended for all female BC cases diagnosed ≤ 30 years old.

Constitutional variants in POT1, TERF2IP, and ACD genes in patients with melanoma in the Polish population.

Evaluation of the prevalence of POT1, ACD, and TERF2IP mutations among Polish melanoma patients. A cohort of 60 patients from melanoma-prone families, 1500 unselected cases and 1500 controls were genotyped. Methodology included Sanger sequencing, in-silico software predilection, and TaqMan assays. We identified three nonsynonymous variants: POT1 c.903 G>T; TERF2IP c.970 A>G; and ACD c.1544 T>C and a splice site variant ACD c.645 G>A. The c.903 G>T was predicted to be pathogenic according to PolyPhen-2, benign according to Mutation Taster, PROVEAN, AGVGD, and SIFT. The c.645 G>A was defined as disease caused by Mutation Taster and Human Splicing Finder and as variant of unknown significance by ClinVar. The other detected variants were described as benign. The c.903 G>T variant was present in two unselected cases and one control [P = 0.57, odds ratio (OR) = 2.00]; the c.645 G>A variant was not detected among the unselected cases and the controls; the c.970 A>G variant was present in 110 cases and 133 controls (P = 0.14, OR = 0.81); the c.1544 T>C variant was present in 687 cases and 642 controls (P = 0.11, OR = 1.07). We found no loss of heterozygosity of the c.903 G>T, c.970 A>G, and c.645 G>A variants. C.645 G>A variant had no effect on splicing or expression. The changes in POT1 c.903 G>T and ACD c.645 G>A can be classified as rare variants of unknown significance, the other variants appear to be polymorphisms. Germline mutations in POT1, ACD, and TERF2IP are infrequent among Polish melanoma patients.

Prevalence of Recurrent Mutations Predisposing to Breast Cancer in Early-Onset Breast Cancer Patients from Poland.

There are twenty recurrent mutations in six breast-cancer-predisposing genes in Poland (BRCA1, BRCA2, CHEK2, PALB2, NBN, and RECQL). The frequencies of the twenty alleles have not been measured in a large series of early-onset breast cancer patients from Poland unselected for family history. We genotyped 2464 women with breast cancer diagnosed below age 41 years for twenty recurrent germline mutations in six genes, including BRCA1, BRCA2 CHEK2, PALB2, NBN, and RECQL. A mutation in one of the six genes was identified in 419 of the 2464 early-onset breast cancer cases (17%), including 22.4% of those cases diagnosed below age 31. The mutation frequency was 18.8% for familial breast cancer cases and 6% for non-familial cases. Among women with breast cancer below age 31, the mutation frequency was 23.6% for familial cases and 17.4% in non-familial cases. The majority of mutations (76.2%) were seen in BRCA1 and BRCA2. In Poland, a panel of twenty recurrent mutations in six genes can identify a genetic basis for a high percentage of early-onset cases and testing is recommended for all women with breast cancer at age 40 or below.

Founder Mutations for Early Onset Melanoma as Revealed by Whole Exome Sequencing Suggests That This is Not Associated with the Increasing Incidence of Melanoma in Poland.

PURPOSE: Germline mutations within melanoma susceptibility genes are present only in minority of melanoma patients and it is expected that additional genes will be discovered with next generation sequence technology and whole-exome sequencing (WES). MATERIALS AND METHODS: Herein we performed WES on a cohort of 96 unrelated Polish patients with melanoma diagnosed under the age of 40 years who all screened negative for the presence of CDKN2Avariants. A replication study using a set of 1,200 melanoma patient DNA samples and similarly large series of healthy controls was undertaken. RESULTS: We selected 21 potentially deleterious variants in 20 genes (VRK1, MYCT1, DNAH14, CASC3, MS4A12, PRC1, WWOX, CARD6, EXO5, CASC3, CASP8AP2, STK33, SAMD11, CNDP2, CPNE1, EFCAB6, CABLES1, LEKR1, NUDT17, and RRP15), which were identified by WES and confirmed by Sanger sequencing for an association study. Evaluation of the allele distribution among carriers and their relatives in available family trios revealed that these variants were unlikely to account for many familial cases of melanoma. Replication study revealed no statistically significant differences between cases and controls. CONCLUSION: Although most of the changes seemed to be neutral we could not exclude an association between variants in VRK1, CREB3L3, EXO5, and STK33 with melanoma risk.

BRCA1/2 mutations are not a common cause of malignant melanoma in the Polish population.

The association of BRCA1/2 mutations with melanoma is not completely determined; the interpretation of variants of unknown significance is also problematic. To evaluate these issues we explored the molecular basis of melanoma risk by performing whole-exome sequencing on a cohort of 96 unrelated Polish early-onset melanoma patients and targeted sequencing of BRCA1/2 genes on additional 30 melanoma patients with familial aggregation of breast and other cancers. Sequencing was performed on peripheral blood. We evaluated MutationTaster, Polyphen2, SIFT, PROVEAN algorithms, analyzed segregation with cancer disease (in both families with identified BRCA2 variants) and in one family performed LOH (based on 2 primary tumors). We found neither pathogenic mutations nor variants of unknown significance within BRCA1. We identified two BRCA2 variants of unknown significance: c.9334G>A and c.4534 C>T. Disease allele frequency was evaluated by genotyping of 1230 consecutive melanoma cases, 5000 breast cancer patients, 3500 prostate cancers and 9900 controls. Both variants were found to be absent among unselected cancer patients and healthy controls. The MutationTaster, Polyphen2 and SIFT algorithms indicate that c.9334G>A is a damaging variant. Due to lack of tumour tissue LOH analysis could not be performed for this variant. The variant segregated with the disease. The c.4534 C>T variant did not segregate with disease, there was no LOH of the variant. The c.9334G>A variant, classified as a rare variant of unknown significance, on current evidence may predisposes to cancers of the breast, prostate and melanoma. Functional studies to describe how the DNA change affects the protein function and a large multi-center study to evaluate its penetrance are required.

Serum 25(OH)D concentration, common variants of the VDR gene and lung cancer occurrence.

The first aim of our study was to examine the association between common variants in VDR [rs2228570 (FokI), rs1544410 (BsmI), rs7975232 (ApaI), rs731236 (TaqI) and rs11568820 (Cdx2)] and lung cancer risk in the Polish population. Genotyping and statistical analysis which included Chi-square test with Yates correction and haplotype frequency analysis were performed on a series of 840 consecutively collected lung cancer patients and 920 healthy controls. The second aim was to evaluate the link between serum 25(OH)D concentration and the number of lung cancers in a subgroup of 200 patients. A separate control group that consisted of 400 matched (by age, sex, smoking habits and the season of blood collection) healthy individuals was used to avoid posterior adjustment on the matched variables. Statistical analysis with the use of Chi-square test with Yates was performed. We found no statistically significant difference in the distribution of the allels of studied VDR variants among cases and controls. A statistically significant over-representation of VDR haplotypes: rs731236_A + rs1544410_T [odds ratio (OR) = 2.43, 95% confidence interval (CI) = 1.11-5.32, p < 0.001], rs731236_G + rs1544410_T (OR = 1.54, 95% CI = 1.31-1.81, p < 0.001) and rs731236_G + rs1544410_C (OR = 0.04, 95% CI = 0.03-0.07, p < 0.001) was detected. We found a tendency toward an increased number of lung cancers among individuals with low serum levels of 25(OH)D. To answer the question, whether VDR can be regarded as lung cancer susceptibility gene and low 25(OH)D serum levels is associated with lung cancer occurrences, additional, multicenter study needs to be performed.

Clinical features and outcomes of germline mutation BRCA1-linked versus sporadic ovarian cancer patients.

BACKGROUND: The role of germline mutations in BRCA1 and BRCA2 genes in the risk of the development of ovarian cancer is clinically well established. BRCA1/2 testing seems to have increasing role in clinical management in patients with advanced ovarian cancer who require treatment with poly(ADP-ribose) polymerase inhibitors. METHODS: Between 2002 - 2008, 125 consecutive patients with ovarian cancer were categorized as having three founder mutations in the BRCA1 gene in Poland as: 5382insC [exon 20], 4153delA [exon 11.17], and 300 T > G [exon 5]. PFS (progression free survival) and OS (overall survival) were determined by Kaplan-Meier analysis with log rank test, univariate comparisons, and multivariate regression analysis using Cox proportional hazards model. RESULTS: Of the 125 patients, the founder mutations of BRCA1 were reported in 17 patients (13.6 %). The median OS was longer for BRCA mutated patients (not reached vs 35.6 months, p = 0.041). PFS was similar for both kinds of ovarian cancer. In multivariate analysis, age ≥70 years, suboptimal surgery, and BRCA1 wild type were poor prognostic factors. The BRCA1 mutation reduced the likelihood of death in ovarian cancer by 86 % (HR 0.14; CI: 0.032-0.650, p = 0.012). CONCLUSION: In conclusion, we found better overall survival for ovarian cancer patients with BRCA1 germline mutations in comparison with patients without these mutations (sporadic) ovarian cancer. Thus, BRCA1 germline mutations appear to be an independent prognostic factor for ovarian cancer.

Clinical outcomes in women with breast cancer and a PALB2 mutation: a prospective cohort analysis.

BACKGROUND: Mutations in PALB2 predispose to breast cancer, but the effect on prognosis of carrying a PALB2 mutation has not been ascertained. We aimed to estimate the odds ratio for breast cancer in women with an inherited mutation in PALB2 and 10-year survival after breast cancer in patients who carry a PALB2 mutation. METHODS: Between 1996 and 2012, patients with invasive breast cancer were recruited prospectively from 18 hospitals in Poland and genotyped for two deleterious mutations in PALB2 (509_510delGA and 172_175delTTGT). A control group of 4702 women without cancer was recruited for comparison. The primary endpoint was death from any cause, as determined by medical records from the Polish Ministry of the Interior and Administration. In patients with breast cancer, 10-year survival of carriers of a PALB2 mutation was calculated and compared with that of non-carriers. FINDINGS: 17 900 women with breast cancer were invited to participate, of whom 12 529 were genotyped successfully. A PALB2 mutation was present in 116 (0·93%, 95% CI 0·76-1·09) of 12 529 patients and in ten (0·21%, 0·08-0·34) of 4702 controls (odds ratio 4·39, 95% CI 2·30-8·37; p<0·0001). 10-year survival for women with breast cancer and a PALB2 mutation was 48·0% (95% CI 36·5-63·2), compared with 74·7% (73·5-75·8) for patients with breast cancer without a mutation (adjusted hazard ratio for death 2·27, 95% CI 1·64-3·15; p<0·0001). INTERPRETATION: Women with a PALB2 mutation face an increased risk of breast cancer and might be at a higher risk of death from breast cancer compared with non-carriers. Increased surveillance should be offered to unaffected women who carry a PALB2 mutation. FUNDING: Polish National Science Centre.

Germline RECQL mutations are associated with breast cancer susceptibility.

Several moderate- and high-risk breast cancer susceptibility genes have been discovered, but more are likely to exist. To discover new breast cancer susceptibility genes, we used 2 populations (from Poland and Quebec, Canada) and applied whole-exome sequencing in a discovery phase (n = 195), followed by validation. We identified rare recurrent RECQL mutations in each population. In Quebec, 7 of 1,013 higher-risk breast cancer cases and 1 of 7,136 newborns carried the c.634C>T (p.Arg215*) variant (P = 0.00004). In Poland, 30 of 13,136 unselected breast cancer cases and 2 of 4,702 controls carried the c.1667_1667+3delAGTA (p.K555delinsMYKLIHYSFR) variant (P = 0.008). RECQL is implicated in resolving stalled DNA replication forks to prevent double-stranded DNA (dsDNA) breaks. This function is related to that of other known breast cancer susceptibility genes, many of which are involved in repairing dsDNA breaks. We conclude that RECQL is a breast cancer susceptibility gene.

Management of ovarian and endometrial cancers in women belonging to HNPCC carrier families: review of the literature and results of cancer risk assessment in Polish HNPCC families.

BACKGROUND: Over half the cancer deaths in HNPCC families are due to extra-colonic malignancies that include endometrial and ovarian cancers. The benefits of surveillance for gynecological cancers are not yet proven and there is no consensus on the optimal surveillance recommendations for women with MMR mutations. METHODS: We performed a systematic review of the literature and evaluated gynecological cancer risk in a series of 631 Polish HNPCC families classified into either Lynch Syndrome (LS, MMR mutations detected) or HNPCC (fulfillment of the Amsterdam or modified Amsterdam criteria). RESULTS: Published data clearly indicates no benefit for ovarian cancer screening in contrast to risk reducing surgery. We confirmed a significantly increased risk of OC in Polish LS families (OR = 4,6, p < 0.001) and an especially high risk of OC was found for women under 50 years of age: OR = 32,6, p < 0.0001 (95% CI 12,96-81,87). The cumulative OC risk to 50 year of life was calculated to be 10%. Six out of 19 (32%) early-onset patients from LS families died from OC within 2 years of diagnosis. We confirmed a significantly increased risk of EC (OR = 26, 95% CI 11,36-58,8; p < 0,001). The cumulative risk for EC in Polish LS families was calculated to be 67%. CONCLUSIONS: Due to the increased risk of OC and absence of any benefit from gynecological screening reported in the literature it is recommended that prophylactic oophorectomy for female carriers of MMR mutations after 35 year of age should be considered as a risk reducing option. Annual transvaginal ultrasound supported by CA125 or HE4 marker testing should be performed after prophylactic surgery in these women. Due to the high risk of EC it is reasonable to offer, after the age of 35 years, annual clinical gynecologic examinations with transvaginal ultrasound supported by routine aspiration sampling of the endometrium for women from either LS or HNPCC families. An alternative option, which could be taken into consideration for women preferring surgical prevention, is risk reducing total hysterectomy (with bilateral salpingo-oophorectomy) for carriers after childbearing is complete.

The variant allele of the rs188140481 polymorphism confers a moderate increase in the risk of prostate cancer in Polish men.

A number of single nucleotide polymorphisms (SNPs) in the human genome have been associated with increased risk of prostate cancer. Recently, a single SNP in the region of chromosome 8q24 (rs188140481) has been associated with a three-fold increased risk of prostate cancer in Europe and North America. To establish whether rs188140481 is associated with the risk of prostate cancer in Poland, we genotyped 3467 men with prostate cancer and 1958 controls. The A allele of rs188140481 was detected in 44 of 3467 (1.3%) men with prostate cancer and in seven of 1958 (0.4%) controls (odds ratio=3.6; 95% confidence interval 1.6-7.9; P=0.0006). The allele was present in eight of 390 (2.1%) men with familial prostate cancer (odds ratio=5.8; 95% confidence interval 2.1-16.2; P=0.001). A positive family history of cancers at sites other than the prostate was observed in 27% of men who carried the rs188140481 risk allele and in 44% of noncarriers (P=0.04). No cancer at a site other than the prostate was more common in first-degree or second-degree relatives of carriers of the rs188140481 risk allele than relatives of noncarriers. The rs188140481 polymorphism in the 8q24 region confers a moderate increase in the risk of prostate cancer in Polish men. The SNP does not appear to be associated with susceptibility to cancers of other types.

Polymorphisms in nucleotide excision repair genes and susceptibility to colorectal cancer in the Polish population.

Xeroderma pigmentosum (XP) is a rare autosomal recessive disease that is associated with a severe deficiency in nucleotide excision repair. Genetic polymorphisms in XP genes may be associated with a change in DNA repair capacity, which could be associated with colorectal cancer development. We assessed the association between 94 single nucleotide polymorphisms (SNPs) within seven XP genes (XPA-XPG) and the colorectal cancer risk in the Polish population. We genotyped 758 unselected patients with colorectal cancer and 1,841 healthy adults. We found that a significantly decreased risk of colorectal cancer was associated with XPC polymorphism rs2228000_CT genotype (OR 0.59; p < 0.0001) and the rs2228000_TT genotype (OR 0.29; p < 0.0001) compared to the reference genotype (CC). And an increased disease risk was associated with the XPD SNP, rs1799793_AG genotype (OR 1.44, p = 0.018) and rs1799793_AA genotype (OR 3.31, p < 0.0001) compared to the reference genotype. Haplotype analysis within XPC, XPD and XPG revealed haplotypes associated with an altered colorectal cancer risk. Stratified analysis by gender showed differences between the association of three SNPs: XPC rs2228000, XPD rs1799793 and XPD rs238406 in females and males. Association analysis between age of disease onset and polymorphisms in XPD (rs1799793) and XPC (rs2228000) revealed differences in the prevalence of these variants in patients under and over 50 years of age. Our results confirmed that polymorphisms in XPC and XPD may be associated with the risk of colorectal cancer.

First recurrent large genomic rearrangement in the BRCA1 gene found in Poland.

Mutation in the BRCA1 gene increases the risk of the person developing breast and/or ovarian cancer. The prevalence and spectrum of large genomic rearrangements (LGRs) varies considerably among different tested populations. In our previous study we described three LGRs in BRCA1 (exons 13-19, exon 17 and exon 22) in Polish families at high risk of breast and ovarian cancer. In this study we analyzed a group of 550 unselected women with ovarian cancer for the three previously identified LGRs. We used a rapid, single-step and closed-tube method: high-resolution melting analysis (HRMA). In this group of unrelated patients diagnosed with ovarian cancer we found three cases with the same deletions of exon 22. This is the first recurrent large deletion in BRCA1 found in Poland. We conclude that screening for the exon 22 deletion in BRCA1 should be included in the Polish BRCA1 genetic testing panel and possibly extended into other Slavic populations.

Common variants of xeroderma pigmentosum genes and prostate cancer risk.

The genetic basis of prostate cancer (PC) is complex and appears to involve multiple susceptibility genes. A number of studies have evaluated a possible correlation between several NER gene polymorphisms and PC risk, but most of them evaluated only single SNPs among XP genes and the results remain inconsistent. Out of 94 SNPs located in seven XP genes (XPA-XPG) a total of 15 SNPs were assayed in 720 unselected patients with PC and compared to 1121 healthy adults. An increased risk of disease was associated with the XPD SNP, rs1799793 (Asp312Asn) AG genotype (OR=2.60; p<0.001) and with the AA genotype (OR=531; p<0.0001) compared to the control population. Haplotype analysis of XPD revealed one protective haplotype and four associated with an increased disease risk, which showed that the A allele (XPD rs1799793) appeared to drive the main effect on promoting prostate cancer risk. Polymorphism in XPD gene appears to be associated with the risk of prostate cancer.

Prevalence of the E318K and V320I MITF germline mutations in Polish cancer patients and multiorgan cancer risk-a population-based study.

The E318K mutation in the MITF gene has been associated with a high risk of melanoma, renal cell carcinoma, and pancreatic cancer; the risk of other cancers has not been evaluated so far. Herein, we examined the possible association of E318K and a novel variant of the MITF gene, V320I, with the risk of cancers of different sites of origin in a Polish population. We assayed for the presence of the E318K and V320I missense mutations in 4,226 patients with one of six various cancers (melanoma or cancer of the kidney, lung, prostate, colon, or breast) and 2,114 controls from Poland. The E318K mutation was detected in 4 of 2,114 participants (0.19%) in the Polish control population, the V320I in 3 of 2,114 participants (0.14%) in the control group. We found no statistically significant differences in the prevalence of the E318K and V320I variants among cases and controls. We found two carriers of the E318K variant among melanoma patients (P = 0.95), one carrier among breast cancer patients (P = 0.77), one carrier among colorectal cancer patients (P = 0.82), and one carrier among kidney cancer patients (P = 0.64). Our study demonstrates a lack of strong association of E318K and V320I with increased risk of melanoma or cancers of the kidney, breast, prostate, lung, or colon.

The presence of prostate cancer at biopsy is predicted by a number of genetic variants.

Several single nucleotide polymorphisms (SNPs) have been associated with an elevated risk of prostate cancer risk. It is not established if they are useful in predicting the presence of prostate cancer at biopsy or if they can be used to define a low-risk group of men. In this study, 4,548 men underwent a prostate biopsy because of an elevated prostate specific antigen (PSA; ≥4 ng/mL) or an abnormal digital rectal examination (DRE). All men were genotyped for 11 selected SNPs. The effect of each SNP, alone and in combination, on prostate cancer prevalence was studied. Of 4,548 men: 1,834 (40.3%) were found to have cancer. A positive association with prostate cancer was seen for 5 of 11 SNPs studied (rs1800629, rs1859962, rs1447295, rs4430796, rs11228565). The cancer detection rate rose with the number of SNP risk alleles from 29% for men with no variant to 63% for men who carried seven or more risk alleles (OR = 4.2; p = 0.002). The SNP data did not improve the predictive power of clinical factors (age, PSA and DRE) for detecting prostate cancer (AUC: 0.726 vs. 0.735; p = 0.4). We were unable to define a group of men with a sufficiently low prevalence of prostate cancer that a biopsy might have been avoided. In conclusion, our data do not support the routine use of SNP polymorphisms as an adjunct test to be used on the context of prostate biopsy for Polish men with an abnormal screening test.

A common nonsense mutation of the BLM gene and prostate cancer risk and survival.

BACKGROUND: Germline mutations of BRCA2 and NBS1 genes cause inherited recessive chromosomal instability syndromes and predispose to prostate cancer of poor prognosis. Mutations of the BLM gene cause another chromosomal instability clinical syndrome, called Bloom syndrome. Recently, a recurrent truncating mutation of BLM (Q548X) has been associated with a 6-fold increased risk of breast cancer in Russia, Belarus and Ukraine, but its role in prostate cancer etiology and survival has not been investigated yet. METHODS: To establish whether the Q548X allele of the BLM gene is present in Poland, and whether this allele predisposes to poor prognosis prostate cancer, we genotyped 3337 men with prostate cancer and 2604 controls. RESULTS: Q548X was detected in 13 of 3337 (0.4%) men with prostate cancer compared to 15 of 2604 (0.6%) controls (OR=0.7; 95% CI 0.3-1.4). A positive family history of any cancer in a first- or second-degree relative was seen only in 4 of the 13 (30%) mutation positive families, compared to 49% (1485/3001) of the non-carrier families (p=0.3). The mean follow-up was 49months. Survival was similar among carriers of Q548X and non-carriers (HR=1.1; p=0.9). The 5-year survival for men with a BLM mutation was 83%, compared to 72% for mutation-negative cases. CONCLUSIONS: BLM Q548X is a common founder mutation in Poland. We found no evidence that this mutation predisposes one to prostate cancer or affect prostate cancer survival. However, based on the observed 0.6% population frequency of the Q548X allele, we estimate that one in 100,000 children should be affected by Bloom syndrome in Poland.

Large BRCA1 and BRCA2 genomic rearrangements in Polish high-risk breast and ovarian cancer families.

BRCA1 and BRCA2 are two major genes associated with familial breast and ovarian cancer susceptibility. In Poland standard BRCA gene test is usually limited to Polish founder BRCA1 mutations: 5382insC, C61G and 4153delA. To date, just a few single large genomic rearrangements (LGRs) of BRCA1 gene have been reported in Poland. Here we report the first comprehensive analysis of large mutations in BRCA1 and BRCA2 genes in this country. We screened LGRs in BRCA1 and BRCA2 genes by multiplex ligation-dependent probe amplification in 200 unrelated patients with strong family history of breast/ovarian cancers and negative for BRCA1 Polish founder mutations. We identified three different LGRs in BRCA1 gene: exons 13-19 deletion, exon 17 deletion and exon 22 deletion. No LGR was detected in BRCA2 genes. Overall, large rearrangements accounted for 3.7 % of all BRCA1 mutation positive families in our population and 1.5 % in high-risk families negative for Polish founder mutation.

Ten-year survival in patients with BRCA1-negative and BRCA1-positive breast cancer.

PURPOSE: To estimate 10-year overall survival (OS) rates for patients with early-onset breast cancer, with and without a BRCA1 mutation, and to identify prognostic factors among those with BRCA1-positive breast cancer. PATIENTS AND METHODS: A total of 3,345 women with stage I to III breast cancer, age ≤ 50 years, were tested for three founder mutations in BRCA1. Information on tumor characteristics and treatments received was retrieved from medical records. Dates of death were obtained from the vital statistics registry. Survival curves for the mutation-positive and -negative subcohorts were compared. Predictors of OS were determined using the Cox proportional hazards model. RESULTS: Of the 3,345 patients enrolled onto the study, 233 (7.0%) carried a BRCA1 mutation. The 10-year survival rate for mutation carriers was 80.9% (95% CI, 75.4% to 86.4%); for noncarriers, it was 82.2% (95% CI, 80.5% to 83.7%). The adjusted hazard ratio (HR) associated with carrying a BRCA1 mutation was 1.81 (95% CI, 1.26 to 2.61). Among BRCA1 carriers with a small (< 2 cm) node-negative tumor, the 10-year survival rate was 89.9%. Among BRCA1 mutation carriers, positive lymph node status was a strong predictor of mortality (adjusted HR, 4.1; 95% CI, 1.8 to 8.9). Oophorectomy was associated with improved survival in BRCA1 carriers (adjusted HR, 0.30; 95% CI, 0.12 to 0.75). CONCLUSION: The 10-year survival rate among women with breast cancer and a BRCA1 mutation is similar to that of patients without a BRCA1 mutation. Among women with a BRCA1 mutation, survival was much improved after oophorectomy.