TGFβ<Subscript>1</Subscript> signaling via α<Subscript>V</Subscript>β<Subscript>6</Subscript> integrin

Background

Transforming growth factor β₁ (TGFβ₁) is a potent inhibitor of epithelial cell growth, thus playing an important role in tissue homeostasis. Most carcinoma cells exhibit a reduced sensitivity for TGFβ₁ mediated growth inhibition, suggesting TGFβ₁ participation in the development of these cancers. The tumor suppresor gene DPC4/SMAD4, which is frequently inactivated in carcinoma cells, has been described as a key player in TGFβ₁ mediated growth inhibition. However, some carcinoma cells lacking functional SMAD4 are sensitive to TGFβ₁ induced growth inhibition, thus requiring a SMAD4 independent TGFβ₁ pathway.

Results

Here we report that mature TGFβ₁ is a ligand for the integrin α_Vβ₆, independent of the common integrin binding sequence motif RGD. After TGFβ₁ binds to α_Vβ₆ integrin, different signaling proteins are activated in TGFβ₁-sensitive carcinoma cells, but not in cells that are insensitive to TGFβ₁. Among others, interaction of TGFβ₁ with the α_Vβ₆ integrin resulted in an upregulation of the cell cycle inhibitors p21/WAF1 and p27 leading to growth inhibition in SMAD4 deleted as well as in SMAD4 wildtype carcinoma cells.

Conclusions

Our data provide support for the existence of an alternate TGFβ₁ signaling pathway that is independent of the known SMAD pathway. This alternate pathway involves α_Vβ₆ integrin and the Ras/MAP kinase pathway and does not employ an RGD motif in TGFβ₁-sensitive tumor cells. The combined action of these two pathways seems to be necessary to elicit a complete TGFβ₁ signal.

     

TGFβ<Subscript>1</Subscript> activates c-Jun and Erk1 via α<Subscript>V</Subscript>β<Subscript>6</Subscript> integrin

Transforming growth factor β (TGFβ) plays an important role in animal development and many cellular processes. A variety of cellular functions that are required for tumor metastasis are controlled by integrins, a family of cell adhesion receptors. Overexpression of α_Vβ₆ integrin is associated with lymph node metastasis of gastric carcinomas. It has been demonstrated that a full TGFβ₁ signal requires both α_Vβ₆ integrin and SMAD pathway. TGFβ₁ binds to α_Vβ₆ via the DLXXL motif, a freely accessible amino acid sequence in the mature form of TGFβ₁. Binding of mature TGFβ₁ to α_Vβ₆ leads to immobilization and tyrosine phosphorylation of proteins, which are associated with focal adhesions, a hallmark of integrin-mediated signal transduction. Here, we show that binding of mature TGFβ₁ recruits the mitogen-activated protein kinase kinase kinase 1 (MEKK1), a mediator of c-Jun activation, and the extracellular signaling-regulated kinase-1 (Erk1) to focal adhesions. In addition, the p21-activated kinase 1 (PAK1) is associated with focal adhesions and differentially phosphorylated upon TGFβ₁ stimulation. We conclude that TGFβ₁ activates c-Jun via the MEKK1/p38 MAP kinase pathway and influences cytoskeletal organization. These finding may provide a link between TGFβ₁ and the metastatic behavior of cancers.     

A study on the CD<Subscript>4</Subscript><Superscript>+</Superscript> CD<Subscript>25</Subscript><Superscript>+</Superscript> regulatory T cells in patients with gastrointestinal malignancies

Objectives Regulatory T cells play an active role in the maintenance of the immune system’s tolerance of both foreign and self antigens. Particularly, CD₄ ⁺ CD₂₅ ⁺ regulatory T cells participate in tumor immunity. The study provided further evidence on the involvement of CD₄ ⁺ CD₂₅ ⁺ regulatory T cells in immune system impairment in patients with gastrointestinal malignancies. Methods Using flow cytometry, CD₄ ⁺ CD₂₅ ⁺ regulatory T cells were analyzed in peripheral blood from 114 patients with gastrointestinal malignancies and 15 healthy controls. Results The prevalence of the CD₂₅ ⁺ subset in CD₄ ⁺ T cells was increased in patients with colorectal carcinoma compared with healthy controls. The phenotic characteristics of the CD₄ ⁺ CD₂₅ ⁺ T cells in patient with malignancies were low expression of CD₄₅ RA and no expression of CD₆₉. Our results indicated that when compared with healthy control, the proportions of CD₄ ⁺ CD₂₅ ⁺ T cells in the peripheral blood of patients with colorectal, gastric, and esophageal carcinoma were significantly higher (P < 0.05) in colorectal carcinoma (22.11 ± 9.65%), gastric carcinoma (17.74 ± 4.24%), and esophageal carcinoma (24.37 ± 4.82)%, respectively. Further analysis on the proportion of CD₄ ⁺ CD₂₅ ⁺ T cells revealed that those patients with gastrointestinal malignancies in stages IV were higher than those of in stage I–III, though no significant difference was observed (P > 0.05). However, the proportion of CD₄ ⁺ CD₂₅ ⁺ T cells in the patients with relapse gastric carcinoma (23.32 ± 4.98%) was significantly higher than that of patients with primary gastric carcinoma (P < 0.01). Conclusions The increased CD₄ ⁺ CD₂₅ ⁺ T cells in patients with gastrointestinal malignancies may be related to immunosuppression and tumor progression. This suggests that elimination or reduction of CD₄ ⁺ CD₂₅ ⁺ regulatory T cells can improve effective tumor immunity for immunotherapy.     

Genetic polymorphisms of <Emphasis Type="Italic">TGF-β</Emphasis>1 &; <Emphasis Type="Italic">TNF-β</Emphasis> and breast cancer risk

Objective. The proliferation of malignant breast epithelial cells is regulated by various stimuli including cytokines and growth factors, thus the variants of those genes may modify the breast cancer risk. To evaluate the potential influences of TGF-1 T²⁹C and TNF- A²⁵²G gene polymorphisms on breast cancer risk, a case–control study was conducted in Korea.Methods. Histologically confirmed breast cancer cases (n = 560) and controls (n=509) with no previous history of cancer were recruited from three teaching hospitals in Seoul, Korea. Genotypes were determined by PCR-CTPP (polymerase chain reaction with confronting two-pair primers) method. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated by unconditional logistic regression model adjusting for age, body mass index, education, parity, age at first full-term pregnancy, and family history of breast cancer. Results. The TGF-1 ²⁹C-allele containing genotypes posed an increased risk of breast cancer (OR=1.3, 95% CI=1.02–1.79), especially in postmenopausal women (OR=1.6, 95% CI=1.01–2.44). Similarly, the TNF- ²⁵²G-allele containing genotypes posed an increased risk of postmenopausal breast cancer (OR=1.7, 95% CI=1.09–2.55). The risk of postmenopausal breast cancer increased in parallel with the number of the risk genotypes (p for trend <0.01). When data were stratified by the presumed non-genetic risk factors, TGF-1 C-allele containing genotypes were found to increase breast cancer risk almost two-fold in postmenopausal women with greater than median body mass index (>22.8 kg/m²) (OR=1.9, 95% CI=1.04–3.37).Conclusion. The results of this study therefore suggest that polymorphisms of TGF-1 and TNF- genes may modify individual susceptibility to breast cancer in Korean women.     

Association of genetic polymorphisms of <Emphasis Type="Italic">ER-α</Emphasis> and the estradiol-synthesizing enzyme genes <Emphasis Type="Italic">CYP17</Emphasis> and <Emphasis Type="Italic">CYP19</Emphasis> with breast cancer risk in Chinese women

Estrogen plays a role in breast cancer development, and genetic polymorphisms in estrogen receptor gene ER-α and genes regulating estrogen biosynthesis and metabolisms are associated with the risk of breast cancer in women in western countries. Therefore, we hypothesized that SNPs in ER-α and other estrogen-metabolizing genes contribute to breast cancer risk in Chinese women. In this study, we genotyped polymorphisms in the regulatory regions of ER-α (rs3798577) and other two estrogen-metabolizing enzyme genes CYP17 (rs743572) and CYP19 (rs10046) among 300 breast cancer cases and 390 controls in a Chinese population. Crude and adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were calculated by unconditional logistic regression analyses to estimate breast cancer risk associated with these polymorphisms. We found that the T allele frequency of ER-α was significantly higher in cases (59.8%) than controls (54.5%) (P = 0.047), but no significant difference was found in the genotype distribution. However, postmenopausal breast cancer risk was associated with the CYP17 TC genotype (aOR = 1.77, 95% CI = 1.11–2.83) compared with the TT genotype. The CYP19 variant TC + TT genotypes were associated with both overall cancer risk (TT + TC vs. TT aOR = 1.73, 95% CI = 1.13–2.65) and premenopausal cancer risk (TT + TC vs. TT aOR = 1.78, 95% CI = 1.03–3.09), particularly for ER +/PR + tumors. Furthermore, there were joint effects between CYP19 T and ER-α T varint genotypes (aOR = 1.67, 95% CI = 1.03–2.69 for CYP19 TC + TT vs. CC among ER-α T variant carriers) and between CYP19 T and CYP17 C variant genotypes (aOR = 1.77, 95% CI = 1.11–2.83 for CYP19 TC + TT vs. CC among CYP17 variant C carriers). This study provides evidence that polymorphisms CYP17 rs743572, CYP19 rs10046 and ER-α rs3798577 are associated with breast cancer risk among Chinese women.     

A new <Emphasis Type="Italic">POT1</Emphasis> germline mutation—expanding the spectrum of <Emphasis Type="Italic">POT1</Emphasis>-associated cancers

Melanomas are associated with several hereditary conditions. We present a large family with several family members affected with primary melanomas and dysplastic nevi as well as thyroid cancer and other malignant tumors. Clinical work-up did not reveal a mutation in any of the genes usually considered with evaluation for predisposition to melanoma (BRCA1/2, CDKN2A, CDK4, PTEN, TP53). Whole exome sequencing of five affected family members showed a new variant in POT1. POT1 is associated with the telomere shelterin complex that regulates telomere protection and telomerase access. Germline mutations in POT1 were recently shown to be associated with hereditary predisposition to melanoma. Our findings support a role of POT1 germline mutations in cancer predisposition beyond melanoma development, suggesting a broader phenotype of the POT1-associated tumor predisposition syndrome that might also include thyroid cancer as well as possibly other malignant tumors.     

Are <Emphasis Type="Italic">BRCA1</Emphasis>- and <Emphasis Type="Italic">BRCA2</Emphasis>-related breast cancers associated with increased mortality?

There has been contradictory evidence as to whether BRCA1 associated breast cancers have a poorer prognosis than non-BRCA1 cancers. In this issue of Breast Cancer Research Robson and colleagues provide further evidence for poorer survival in BRCA1 carriers and show that it could be attributed to failure to treat small node-negative grade 3 breast cancers with chemotherapy. There still remains little evidence for a survival difference for BRCA2 related breast cancers. Although the high contralateral breast cancer risk is confirmed by this study there is no real evidence for an increase in ipsilateral recurrence or new primary breast cancers in mutation carriers up to the 10-year point.     

15-Deoxy-Δ<Superscript>12,14</Superscript>-prostaglandin J<Subscript>2</Subscript> inhibits G<Subscript>2</Subscript>-M phase progression in human breast cancer cells via the down-regulation of cyclin B1 and survivin expression

The cyclopentenone prostaglandin 15-deoxy-Δ^12,14-prostaglandin J₂ (15d-PGJ₂) exerts a growth inhibitory effect on cancer cells, and this effect is linked to the induction of apoptosis or cell cycle arrest. Induction of apoptosis by 15d-PGJ₂ is associated with the down-regulation of anti-apoptotic proteins. G₀-G₁→S phase progression is inhibited by 15d-PGJ₂ via the degradation of cyclin D1. In this study, we further investigated the mechanism by which 15d-PGJ₂ inhibits cancer cell growth by using the breast cancer cell lines MCF-7 and T-47D. Treatment with 20 μM 15d-PGJ₂ for 72 h completely blocked the growth in both cell lines. However, the proportions of apoptotic MCF-7 and T-47D cells were 21.1% and 40.9%, respectively, indicating that the induction of apoptosis did not appear to fully account for growth inhibition by 15d-PGJ₂. Cell cycle analysis using cells synchronized at the G₀-G₁ or S phase revealed that 15d-PGJ₂ blocked not only G₀-G₁→S phase progression but also G₂-M phase progression. The expression of both cyclins D1 and B1 was decreased by 15d-PGJ₂. Furthermore, 15d-PGJ₂ inhibited aurora-B kinase activity, which coincided with the down-regulation of survivin. Thus, 15d-PGJ₂ induced cell cycle arrest at the G₂-M phase via inhibition of cyclin B1 expression and aurora-B kinase activity. We conclude that survivin may be an important target for 15d-PGJ₂, and its down-regulation may lead to a decrease in aurora-B kinase activity. Naoki Tsuji substantially contributed to this work and should also be considered a first author.     

Prevalence of <Emphasis Type="Italic">BRCA1</Emphasis> and <Emphasis Type="Italic">BRCA2</Emphasis> mutations in unselected breast cancer patients from Medellín,Colombia

Background

Approximately 5% of all breast cancers can be attributed to a mutation in the BRCA1 or BRCA2 gene. The genetic component of breast cancer in Colombia has been, for the most part, studied on cases from the Bogota region. Five different founder mutations have been identified in two studies of breast cancer patients in the Bogota region. It is important that the frequency of mutations be established among unselected cases of breast cancer of other regions of Colombia in order to estimate the genetic burden of this cancer in Colombia and to plan genetic services. The aim of this study was to establish the mutation frequencies of the BRCA genes in breast cancer patients unselected for family history or age, from Medellin, Colombia.

Methods

We enrolled 280 unselected women with breast cancer from a large public hospital in Medellin, Colombia. A detailed family history from each patient and a blood sample was obtained and processed for DNA analysis. Mutations in BRCA1 and BRCA2 were sought using a combination of techniques including a panel of recurrent Hispanic BRCA mutations which consists of fifty BRCA1 mutations and forty-six BRCA2 mutations, including the five recurrent Colombian BRCA mutations. All mutations were confirmed by direct sequencing.

Results

Genetic testing was successfully completed for 244 of the 280 cases (87%). Among the 244 cases, three deleterious mutations were identified (two in BRCA1 and one in BRCA2), representing 1.2% of the total. The average age of breast cancer in the mutation-positive cases was 34 years. The two BRCA1 mutations were known founder mutations (3450del4 in exon 11 and A1708E in exon 18). The BRCA2 mutation was in exon 11 (5844del5) and has not been previously reported in individuals of Colombian descent. Among the three mutation-positive families was a breast cancer family and two families with no history of breast or ovarian cancer.

Conclusion

The frequency of BRCA mutations in unselected breast cancer cases from the Medellin region of Colombia is low and is approximately 1.2%.

     

Relative contributions of <Emphasis Type="Italic">BRCA1</Emphasis> and <Emphasis Type="Italic">BRCA2</Emphasis> mutations to “triple-negative” breast cancer in Ashkenazi Women

Approximately 10% of Ashkenazi Jewish (AJ) women with breast cancer (BC) carry a founder mutation in BRCA1 or BRCA2. There is an association between BRCA1 mutations and “triple-negative” breast cancer (TNBC) [estrogen receptor (ER) and progesterone receptor (PR) negative, HER2 negative]. We sought to determine the predictive value of the TNBC phenotype for the presence of a BRCA mutation in AJ women ascertained without respect to family history. DNA samples were collected between 8/2000 and 6/2004 from a prevalent cohort of unselected AJ women with breast cancer (median age at diagnosis 56 years). Samples (n = 451) were genotyped for AJ founder mutations. 352 (78.0%) cancers were ER positive, 254 (56.3%) PR positive, and 91 (20.2%) ER negative/PR negative. 63 (14.0%) cancers were HER2 positive (immunohistochemistry 3+ or FISH >2.2). TNBC was observed in 64 patients (14.2%). Founder mutations were detected in 48 samples (10.6%) including 25/64 TNBC (39.1%; 19 BRCA1, 6 BRCA2). Among TNBC patients with family history (FH) information, 6/15 (40%) mutations were found in women without breast or ovarian cancer in a close relative. The positive predictive value of TNBC for a BRCA1 mutation was 30% overall, 50% in women diagnosed<50 years, and 14% in women diagnosed ≥50. TNBC was significantly associated with detecting a mutation in either BRCA1 or BRCA2, but only 25/52 (48%) mutation-associated cancers were TNBC. The prevalence of BRCA founder mutations exceeds 50% in subsets of AJ women with TNBC. FH is an imperfect predictor of mutation status in this group. A significant number of mutation-associated TNBC are due to BRCA2.     

Sex Ratio Distortion in Offspring of Families with <Emphasis Type="BoldItalic">BRCA1</Emphasis> or <Emphasis Type="BoldItalic">BRCA2</Emphasis> Mutant Alleles: An Ascertainment Bias Phenomenon?

Summary Background. There has been controversy regarding whether BRCA1 germline mutations favor female births or whether the sex imbalances observed are attributable to ascertainment bias. Our aims were to compare the sex ratios among offspring of BRCA1-positive, BRCA2-positive, and BRCA-negative families undergoing genetic testing in clinical programs, and to determine whether ascertainment bias is responsible for the observed preponderance of female offspring.Patients and methods. A total of 145 breast and/or ovarian cancer families with mutations in BRCA1 (n = 83) or BRCA2 (n = 62), and 90 families without identifiable mutation were collected for the study from familial cancer clinics in Barcelona, Spain, and Boston, US. Sex ratio was analyzed among all births in the families and offspring of all (tested and obligate) carriers. In order to minimize the effect of family history of cancer, the analysis was also performed among offspring of the most recent generation of mutation-positive carriers who did not have affected children and compared with a control group comprised of the offspring of the most recent adult generation of non-carriers from families with a known mutation.Results. There was a statistically higher proportion of female births in all groups (BRCA1 59% (95% CI = 57–61%), BRCA2 58% (56–61%), and BRCA-negative 59% (56–61%), respectively). The female preponderance persisted in analyses limited to offspring of BRCA1 and BRCA2 carriers (61% (57–65%), and 62% (58–66%), respectively), with no differences between the two mutation groups. In contrast, the excess of female offspring disappeared when ascertainment or recall biases were minimized, 44% (37–52%), and 39% (26–53%) for BRCA1; 51% (44–58%), and 46% (33–60%) for BRCA2.Conclusions. Our findings suggest that there is no asymmetry in birth outcomes among BRCA1 or BRCA2 mutations carriers. Rather ascertainment bias in families participating in genetic testing, or in the family history information they provide is likely to account for excess of female offspring previously reported.     

Phenocopy breast cancer rates in Israeli <Emphasis Type="Italic">BRCA1</Emphasis><Emphasis Type="Italic">BRCA2</Emphasis> mutation carrier families: is the risk increased in non-carriers?

BRCA1 and BRCA2 mutation carriers have an increased risk for developing breast (and ovarian) cancer. Non-carriers from within such families (=true negatives) are counseled that their risk for developing breast cancer is similar to that of the average-risk population. Breast cancer diagnosed in a non-carrier from a family with a known mutation is coined phenocopy. The rate of breast cancer phenocopy and the risk for breast cancer in true negatives are unsettled. The rate of phenocopy breast cancer was assessed in non-carriers from Jewish families with a BRCA1 or BRCA2 mutation, identified at the Sheba medical center. Analysis was performed by t test for comparison of mean age at counseling or breast cancer diagnosis, and by calculating a standardized incidence ratio (SIR). Overall, 1318 females from 884 mutation carrying families (620 with BRCA1 264 with BRCA2 mutations) were genotyped, of whom 307 women from 245 families were assigned a true negative status (mean age at counseling 43.01 ± 13.03 years (range 19.7–92.8 years). Of these true negatives, 20 women (6.51–2.26% of families) developed breast cancer at a mean age of 54.1 ± 12.9 years (range 48.1 –60.1 years). The SIR for breast cancer in true negatives was not significantly different than the expected in the average-risk Israeli population [observed 20-expected 23.8 cases SIR = 0.84, 95% CI (0.51, 1.30)]. The rate of phenocopy breast cancer in non-carriers from Israeli BRCA1 BRCA2 mutation carrier families is 2.26% with no increased breast cancer risk over the average-risk population.     

<Emphasis Type="Italic">NRAS</Emphasis><Superscript><Emphasis Type="Italic">Q61K</Emphasis></Superscript> mutated diffuse leptomeningeal melanomatosis in an adult patient with a brief review of the so-called “forme fruste” of neurocutaneous melanosis

Primary melanocytic tumors of central nervous system represent rare tumors arising from melanocytes of the leptomeninges. These neoplasms include focal forms like melanocytoma and primary malignant melanoma and diffuse forms like leptomeningeal melanocytosis and primary leptomeningeal melanomatosis. The clinical diagnosis remains challenging, with clinical and radiologic features overlapping with other more common diseases. Here we present a case of a 38 years old male with primary diffuse leptomeningeal melanomatosis with presence of a NRAS^Q61K mutation without features of neurocutaneous melanosis.     

Association between the <Emphasis Type="Italic">T27C</Emphasis> polymorphism in the <Emphasis Type="Italic">cytochrome P450 c17</Emphasis>α (<Emphasis Type="Italic">CYP17</Emphasis>) gene and risk factors for breast cancer

Mammographic density is associated with increased breast cancer risk and is influenced by sex hormones. A T27C polymorphism (alleles A1 and A2, respectively) in the 5 promoter region of CYP17 may be associated with elevated sex hormone levels. In a cross-sectional study of 181 pre- and 173 postmenopausal women, we examined the relationship of this polymorphism with mammographic density and other risk factors for breast cancer. Subjects were recruited across five categories of density. Risk factor and dietary information, anthropometric measures, and blood samples were obtained. Sex hormone, lipid, growth factor levels, and CYP17 genotypes were determined. CYP17 genotype was not associated with mammographic density levels before or after adjusting for risk factors for breast cancer. In premenopausal women, the A2 allele was associated with higher levels of dehydroepiandrosterone sulfate, and in postmenopausal women, with higher levels of total estradiol and lower levels of follicle stimulating hormone. Among premenopausal women, interactions were observed between CYP17 genotype and endogenous insulin levels as well as dietary variables associated with mammographic density. Our findings suggest that the CYP17 A2 allele is associated with hormone levels, and interacts with insulin levels and diet to affect breast density levels and potentially breast cancer risk.     

<Emphasis Type="Italic">Estrogen receptor</Emphasis> α, <Emphasis Type="Italic">BRCA1</Emphasis>, and <Emphasis Type="Italic">FANCF</Emphasis> promoter methylation occur in distinct subsets of sporadic breast cancers

Introduction Alcohol intake has been consistently associated with breast cancer risk, but the importance of timing of intake and the impact of beverage type are unclear. Methods We evaluated whether early, lifetime or recent alcohol intake was associated with breast cancer risk, and whether risk varied by type of alcoholic drinks in 1,728 newly diagnosed population-based breast cancer patients and 435 control subjects aged 20–49 years. We used multivariable logistic regression models to estimate odds ratios (OR) and 95% confidence intervals (CI) as measures of the relative risk of breast cancer associated with intake of alcoholic drinks. Results Intake of alcoholic drinks during the recent five year period before the breast cancer diagnosis was associated with increased breast cancer risk (P _trend = 0.04). Intake of two or more alcoholic drinks per day during this five year period was associated with an 82% increase in breast cancer risk relative to never drinkers (OR = 1.82, 95% CI = 1.01–3.28). No risk increase was observed for alcohol intake at ages 15–20 years or for lifetime alcohol intake. Risk did not vary by type of alcohol consumed. Conclusions Our results suggest that recent alcohol consumption may be associated with increased breast cancer risk in young women.     

Smoking and risk of breast cancer in carriers of mutations in <Emphasis Type="Italic">BRCA1</Emphasis> or <Emphasis Type="Italic">BRCA2</Emphasis> aged less than 50 years

Background Cigarette smoke contains compounds that may damage DNA, and the repair of damage may be impaired in women with germline mutations in BRCA1 or BRCA2. However, the effect of cigarette smoking on breast cancer risk in mutation carriers is the subject of conflicting reports. We have examined the relation between smoking and breast cancer risk in non-Hispanic white women under the age of 50 years who carry a deleterious mutation in BRCA1 or BRCA2. Methods We conducted a case-control study using data from carriers of mutations in BRCA1 (195 cases and 302 controls) and BRCA2 (128 cases and 179 controls). Personal information, including smoking history, was collected using a common structured questionnaire by eight recruitment sites in four countries. Odds-ratios (OR) for breast cancer risk according to smoking were adjusted for age, family history, parity, alcohol use, and recruitment site. Results Compared to non-smokers, the OR for risk of breast cancer for women with five or more pack-years of smoking was 2.3 (95% confidence interval 1.6–3.5) for BRCA1 carriers and 2.6 (1.8–3.9) for BRCA2 carriers. Risk increased 7% per pack-year (p < 0.001) in both groups. Conclusions These results indicate that smoking is associated with increased risk of breast cancer before age 50 years in BRCA1 and BRCA2 mutation carriers. If confirmed, they provide a practical way for carriers to reduce their risks. Previous studies in prevalent mutation carriers have not shown smoking to increase risk of breast cancer, but are subject to bias, because smoking decreases survival after breast cancer. Northern California Family Registry for Breast Cancer: AS Whittemore, Stanford University School of Medicine, EM John, Northern California Cancer Center, A Felberg, Stanford University School of Medicine, V McGuire, Stanford University School of Medicine, DW West, Northern California Cancer Center, A Miron, Dana-Farber Cancer Institute, Harvard Medical School, DC Thomas, USC Keck School of Medicine, R Haile, USC Keck School of Medicine and Norris Comprehensive Cancer. Fox Chase Familial Breast Cancer Registry: M Daly, Fox Chase Cancer Center, A Godwin, Fox Chase Cancer Center, E Ross, Fox Chase Cancer Center. Coriell Institute: J Beck. New York Familial Breast Cancer Registry: MB Terry, Joseph L. Mailman School of Public Health, Columbia University. Utah Breast Cancer Family Registry: SS Buys, Huntsman Cancer Institute, V Venne, Huntsman Cancer Institute. Australian Breast Cancer Family Study: JL Hopper, The University of Melbourne, GG Giles, The Cancer Council Victoria, MRE McCredie, University of Otago, New Zealand, RL Milne, Spanish National Cancer Centre, MC Southey, The University of Melbourne, MA Jenkins, The University of Melbourne, C Apicella, The University of Melbourne. Kathleen Cuningham Consortium for Research into Familial Breast Cancer (kConFab), Peter MacCallum Cancer Centre. Ontario Familial Breast Cancer Registry: I Andrulis, Cancer Care Ontario, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, NF Boyd, Ontario Cancer Institute, J Knight, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, H Ozcelik, Samuel Lunenfeld Research Institute, Mount Sinai Hospital. Correspondence to: Dr NF Boyd, Campbell Family Institute for Breast Cancer Research, Room 10-415, Ontario Cancer Institute, 610 University Ave., Toronto, Ontario, Canada M5G 2M9. Boyd@uhnres.utoronto.ca