Professor Montse Garcia-Closas
Group Leader: Integrative Cancer Epidemiology
Biography
Professor Montserrat García-Closas is a Professor of Epidemiology and Group Leader for the Integrative Epidemiology Group. She co-leads the Cancer Epidemiology and Prevention Unit, a joint initiative from the ICR and Imperial College London, and holds an honorary appointment at The Royal Marsden Hospital.
Professor Montserrat García-Closas received her M.D. from the University of Barcelona, Spain, a Master of Public Health in quantitative methods, and a Doctorate of Public Health in epidemiology from the Harvard School of Public Health.
She joined the Division of Cancer Epidemiology and Genetics (DCEG) of the National Cancer Institute in the USA in 1996 as a postdoctoral fellow, became a tenure-track investigator in 1999 and a tenured senior investigator in 2007. From 2008-2010, she was a visiting scientist at the Department of Oncology and Strangeways Laboratory, Cambridge University, U.K.
In 2010, she became a Professor of Epidemiology at the Division of Genetic and Epidemiology at the ICR. In 2015, she returned to DCEG as Deputy Director and senior investigator. She served as Interim Branch Chief for the Integrative Tumor Epidemiology Branch (ITEB) from 2016-2020, and as Director of the Trans-Divisional Research Program (TDRP) from 2020-2023. In 2023 she returned to the ICR in London as Professor of Epidemiology.
Professor García-Closas investigates the causes of cancer with the aim of understanding carcinogenic processes and improving risk assessment for precision prevention. She carries out a multidisciplinary research program on the genetic susceptibility, etiologic heterogeneity, and risk prediction for breast cancer.
She is principal investigator (PI) of the Breast Cancer Now Generations Study, a prospective cohort study of 110,000 women in the UK to study cancer aetiology and its outcomes, and the Male Breast Cancer Study, a case-control study of over 1,000 men with breast cancer and 1,000 controls to study the genetic and environmental causes of male breast cancer. Professor García-Closas also contributes to molecular epidemiology studies of bladder and other cancers. She has published over 400 peer-reviewed papers, as well as book chapters and review articles.
Professor García-Closas is a founder member of several large consortia. She served as the Chair of the Coordinating Committee for the International Consortium of Case-Control Studies of Bladder Cancer from 2005-2011, and since 2006 she co-Chairs the Pathology and Survival Working Group of the Breast Cancer Association Consortium (BCAC).
Professor Garcia-Closas co-led the European Commission Horizon 2020-funded Breast CAncer STratification (BCAST) Project within BCAC. She initiated and is the Chair of the Scientific Steering Committee for the Confluence Project which brings together multiple consortia for trans-ancestry genetic susceptibility studies of female and male breast cancer.
She is an active member of the NCI Cohort Consortium, where she co-leads the U01 Multi-factorial Breast Cancer Risk Prediction Accounting for Ethnic and Tumor Diversity to develop a comprehensive tool to predict breast cancer risk, overall and by sub-types, across major racial/ethnic groups in the U.S.
She served as the founding PI of Connect for Cancer Prevention (2016-2023), a new prospective study of 100,000 adults in the USA, and she currently serves in the International Advisory Committee.
Dr.P.H. Epidemiology, Harvard School of Public Health, Boston.
M.P.M. Quantitative Methods, Harvard School of Public Health, Boston.
MD, University of Barcelona, Spain.
Merit Award, NIH, 2001.
Intramural Research Award, National Cancer Institute, 2004.
Mentor of Merit Award, National Cancer Institute, 2007.
Merit Award, NIH, 2008.
Editorial BoardsCancer Epidemiology, Biomarkers & Prevention, 2005.
AACR Cancer Epidemiology and Prevention Awards Committee, Member, AACR, 2009-2011.
Scientific Review Committee for the 2010 Landon Foundation-AACR INNOVATOR Award for Cancer Prevention Research, Member, AACR, 2010-2010.
Member of the 2011 AACR Annual Meeting Education Committee, Member, AACR, 2010-2010.
Program Committee for the 2010 AACR Annual Meeting, Member, AACR, 2009-2010.
External Scientific Committee for the Multicenter Case-Control Study in Spain, Centre for Research in Environmental Epidemiology (CREAL), Member, CREAL, 2009.
Scientific Advisory Board for the Sister Study, Member, National Institute of Environmental Health Sciences, Research, 2007.
Steering Committee of the Molecular Epidemiology Working Group, Member, AACR, 2007-2009.
Search Committee for Tenure-track/Tenure Investigator, Laboratory of Translational Genetics, Member, National Cancer Institute, 2007.
Estrogen Receptor Negative Breast Cancer Health Disparities Work Group, Member, National Cancer Institute, 2007-2008.
Search Committee for Branch Chief, Member, National Cancer Institute, 2007.
Search Committee for Tenure-track Investigator, Biostatics Branch, Member, National Cancer Institute, 2007.
Etiology and Early Marker Studies (EEMS) Review Panel for PLCO Studies, Member, National Cancer Institute, 2006-2008.
Tenure-track Search Committee, Occupational & Environmental Epidemiology Branch, Member, National Cancer Institute, 2005.
Tenure-track Search Committee, Biostatistics Branch, Member, National Cancer Institute, 2005.
Epidemiology & Biostatistics Subsection, Planning Committee of the AACR 96th Annual Meeting, Chair, American Association for Cancer Research, 2004.
Technical Evaluation of Protocols Committee, Member, National Cancer Institute, 2003.
Technical Evaluation of Questionnaires Committee, Member, National Cancer Institute, 1998.
Related pages
Types of Publications
Journal articles
Power and sample size considerations are critical for the design of epidemiologic studies of gene-environment interactions. Hwang et al. (Am J Epidemiol 1994;140:1029-37) and Foppa and Spiegelman (Am J Epidemiol 1997;146:596-604) have presented power and sample size calculations for case-control studies of gene-environment interactions. Comparisons of calculations using these approaches and an approach for general multivariate regression models for the odds ratio previously published by Lubin and Gail (Am J Epidemiol 1990; 131:552-66) have revealed substantial differences under some scenarios. These differences are the result of a highly restrictive characterization of the null hypothesis in Hwang et al. and Foppa and Spiegelman, which results in an underestimation of sample size and overestimation of power for the test of a gene-environment interaction. A computer program to perform sample size and power calculations to detect additive or multiplicative models of gene-environment interactions using the Lubin and Gail approach will be available free of charge in the near future from the National Cancer Institute.
OBJECTIVES: This study was undertaken to evaluate the relationship between vaginal pH and factors related to cervical cancer. STUDY DESIGN: In a population-based sample of 9161 women from Guanacaste Province in Costa Rica women were categorized into 2 groups, those with vaginal pH in the reference range (4.0-4.5) and those with elevated vaginal pH (5.0-5.5). Odds ratios were used to estimate the relationship between elevated pH and its potential determinants. RESULTS: Aging was strongly associated with increasing vaginal pH, starting at around 45 years of age and continuing into old age. Menopause was responsible for an additional 1.7-fold increase in the odds of having an elevated pH (odds ratio 1.7, 95% confidence interval 1.4-2.0). Human papillomavirus infection and cervical intraepithelial neoplasia were not associated with changes in pH. CONCLUSIONS: Our data indicate that vaginal pH is strongly related to age and to menopausal status and thus could be a marker of age-related hormonal changes. Elevated pH does not appear to be associated with risk of high-grade intraepithelial neoplasia among women infected with human papillomavirus.
Red meat or meat-cooking methods such as frying and doneness level have been associated with an increased risk of colorectal and other cancers. It is unclear whether it is red meat intake or the way it is cooked that is involved in the etiology of colorectal cancer. To address this issue, we developed an extensive food frequency questionnaire module that collects information on meat-cooking techniques as well as the level of doneness for individual meat items and used it in a study of colorectal adenomas, known precursors of colorectal cancer. A case-control study of colorectal adenomas was conducted at the National Naval Medical Center (Bethesda, MD) between April 1994 and September 1996. All cases (n = 146) were diagnosed with colorectal adenomas at sigmoidoscopy or colonoscopy and histologically confirmed. Controls (n = 228) were screened with sigmoidoscopy and found not to have colorectal adenomas. The subjects completed a food frequency questionnaire and answered detailed questions on meat-cooking practices. We used frequency and portion size to estimate grams of meat consumed per day for total meat as well as for meat subgroups defined by cooking methods and doneness levels. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using logistic regression, adjusted for age, gender, total caloric intake, reason for screening (routine or other), and several established risk factors for colorectal adenomas or cancer, including the use of nonsteroidal anti-inflammatory drugs, physical activity, and pack-years of cigarette smoking. There was an increased risk of 11% per 10 g/day (or 2.5 oz/week) of reported red meat consumption (OR, 1.11; CI, 1.03-1.19). The increased risk was mainly associated with well-done/very well-done red meat, with an excess risk of 29% per 10 g/day (OR, 1.29; CI, 1.08-1.54) versus an excess of 10% per 10 g/day (OR, 1.10; CI, 0.96-1.26) for consumption of rare/medium red meat. High-temperature cooking methods were also associated with increased risk; 26% per 10 g/day (OR, 1.26; CI, 1.06-1.50) of grilled red meat and 15% per 10 g/day (OR, 1.15; CI, 0.97-1.36) of pan-fried red meat consumption. There was an increased risk of colorectal adenomas associated with higher intake of red meat, most of which was due to the subgroup of red meat that was cooked until well done/very well done and/or by high-temperature cooking techniques, such as grilling. These results are consistent with the hypothesis that carcinogenic compounds formed by high-temperature cooking techniques, such as heterocyclic amines and polycyclic aromatic hydrocarbons, may contribute to the risk of developing colorectal tumors.
In this chapter we describe the impact of risk factor misclassification in case-control studies designed to estimate gene-environment interactions. We show that under certain scenarios even small amounts of exposure or genotype misclassification can substantially attenuate the interaction effect and, as a consequence, dramatically increase the sample size required to study these interactions. A consideration of how sample size is affected by exposure and genotype misclassification in the study design phase should help to identify situations where obtaining better risk factor information is crucial for the feasibility of studies.
BACKGROUND: The enzymes encoded by the glutathione S-transferase mu 1 (GSTM1) and theta 1 (GSTT1) genes are involved in the metabolism (mainly inactivation, but activation is possible) of a wide range of carcinogens that are ubiquitous in the environment; the enzyme encoded by the GSTT1 gene may also be active in endogenous mutagenic processes. Homozygous deletions of the GSTM1 and GSTT1 genes are commonly found in the population and result in a lack of enzyme activity. This study was undertaken to evaluate the association between GSTM1 and GSTT1 gene polymorphisms and breast cancer risk. METHODS: Our study included 466 women with incident cases of breast cancer occurring from May 1989 through May 1994 and 466 matched control subjects. These individuals were part of a prospective cohort of U.S. women (i.e., the Nurses' Health Study). Odds ratios (ORs) and 95% confidence intervals (CIs) from conditional logistic regression models were used to estimate the association between genetic polymorphisms and breast cancer risk. RESULTS: The GSTM1 and GSTT1 null genotypes were not associated with an increased risk of breast cancer (OR = 1.05 [95% CI = 0.80-1.37] for GSTM1 null; OR = 0. 86 [95% CI = 0.61-1.21] for GSTT1 null). On the contrary, a suggestion of a decreased risk of breast cancer associated with the GSTT1 null genotype was observed among premenopausal women. When considered together, no combination of the GSTM1 and GSTT1 genotypes was associated with an increased risk of breast cancer. The relationship between GSTM1 and GSTT1 gene deletions and breast cancer risk was not substantially modified by cigarette smoking. CONCLUSIONS: Our data provide evidence against a substantially increased risk of breast cancer associated with GSTM1 and/or GSTT1 homozygous gene deletions.
In studies of gene-environment interactions, exposure misclassification can lead to bias in the estimation of an interaction effect and increased sample size. The magnitude of the bias and the consequent increase in sample size for fixed misclassification probabilities are highly dependent on the prevalence of the misclassified factor and on the interaction model. This paper describes a relatively simple approach to assess the impact of misclassification on bias in the estimation of multiplicative or additive interactions and on sample size requirements. Applications of this method illustrate that even small errors in the assessment of environmental or genetic factors can result in biased interaction parameters and substantially increased sample size requirements that can compromise the feasibility of the study. Also, an example is provided where nondifferential misclassification biases an additive interaction parameter away from the null value, even under conditions where a multiplicative interaction parameter will always be biased toward the null value. Efforts to improve the accuracy in measuring both genetic and environmental factors are critical for the valid assessment of gene-environment interactions in case-control studies.
Tobacco use is an established cause of bladder cancer. The ability to detoxify aromatic amines, which are present in tobacco and are potent bladder carcinogens, is compromised in persons with the N-acetyltransferase 2 slow acetylation polymorphism. The relationship of cigarette smoking with bladder cancer risk therefore has been hypothesized to be stronger among slow acetylators. The few studies to formally explore such a possibility have produced inconsistent results, however. To assess this potential gene-environment interaction in as many bladder cancer studies as possible and to summarize results, we conducted a meta-analysis using data from 16 bladder cancer studies conducted in the general population (n = 1999 cases), Most had been conducted in European countries. Because control subjects were unavailable for a number of these studies, we used a case-series design, which can be used to assess multiplicative gene-environment interaction without inclusion of control subjects. A case-series interaction odds ratio (OR) > 1.0 indicates that the relationship of cigarette smoking and bladder cancer risk is stronger among slow acetylators as compared with rapid acetylators. We observed an interaction between smoking and N-acetyltransferase 2 slow acetylation (OR, 1.3; 95% confidence interval, 1.0-1.6) that was somewhat stronger when analyses were restricted to studies conducted in Europe (OR, 1.5; confidence interval, 1.1-1.9), a pooling that included nearly 80% of the collected data. Using the predominantly male European study population and assuming a 2.5-fold elevation in bladder cancer risk from smoking, we estimated that the population attributable risk percent was 35% for slow acetylators who had ever smoked and 13% for rapid acetylators who had ever smoked. These results suggest that the relationship of smoking and bladder cancer is stronger among slow acetylators than among rapid acetylators.
We devised a simple, noninvasive, cost-efficient technique for collecting buccal cell DNA for molecular epidemiology studies. Subjects (n = 52) brushed their oral mucosa and expectorated the fluid in their mouths, which was applied to "Guthrie" cards pretreated to retard bacterial growth and inhibit nuclease activity (IsoCode, Schleicher and Schuell, Keene, NH). The cards are well-suited for transport and storage because they dry quickly, need no processing, and are compact and lightweight. We stored the samples at room temperature for 5 days to mimic a field situation and then divided them into portions from which DNA was extracted either immediately or after storage for 9 months at room temperature, -20 degrees C, or -70 degrees C. The fresh samples had a median yield of 2.3 microg of human DNA (range, 0.2-53.8 microg), which was adequate for at least 550 PCR reactions. More than 90% of the samples were amplified in all three beta-globin gene fragment assays attempted. DNA extract frozen for 1 week at -20 degrees C also performed well. Stored samples had reduced DNA yields, which achieved statistical significance for room temperature and -70 degrees C, but not -20 degrees C, storage. However, because all of the stored samples tested were successfully amplified, the observed reduction may represent tighter DNA fixation to the card over time rather than loss of genetic material. We conclude that treated cards are an alternative to brushes/swabs and mouth rinses for the collection of buccal cell DNA and offer some advantages over these methods, particularly for large-scale or large-scale or long-term studies involving stored samples and studies in which samples are collected off-site and transported. Future studies that enable direct comparisons of the various buccal cell collection methods are needed.
The results obtained from experimental studies of estrogen carcinogenesis need validation in epidemiologic studies. Such studies present additional challenges, however, because variations in human populations are much greater than those in experimental systems and in animal models. Because epidemiologic studies are often used to evaluate modest differences in risk factors, it is essential to minimize sources of errors and to maximize sensitivity, reproducibility, and specificity. In the first part of this chapter, critical factors in designing and executing epidemiologic studies, as well as the influence of sample collection, processing, and storage on data reliability, are discussed. One of the most important requirements is attaining sufficient statistical power to assess small genetic effects and to evaluate interactions between genetic and environmental factors. The second part of this chapter describes innovative technology, namely, Fourier transform-infrared (FT-IR) spectra of DNA that reveal major structural differences at various stages of the progression from normal to cancer tissue. The structural differences become evident from wavenumber-by-wavenumber statistical comparisons of the mean FT-IR spectra of DNA from normal to cancer tissues. This analysis has allowed distinguishing benign tissues from cancer and metastatic tissues in human breast, prostate, and ovarian cancers. This analysis, which requires less than 1 microg of DNA, is predicted to be used for detecting early cancer-related changes at the level of DNA, rather than at the cellular level.
Overwhelming evidence indicates that environmental exposures, broadly defined, are responsible for most cancer. There is reason to believe, however, that relatively common polymorphisms in a wide spectrum of genes may modify the effect of these exposures. We discuss the rationale for using common polymorphisms to enhance our understanding of how environmental exposures cause cancer and comment on epidemiologic strategies to assess these effects, including study design, genetic and statistical analysis, and sample size requirements. Special attention is given to sources of potential bias in population studies of gene--environment interactions, including exposure and genotype misclassification and population stratification (i.e., confounding by ethnicity). Nevertheless, by merging epidemiologic and molecular approaches in the twenty-first century, there will be enormous opportunities for unraveling the environmental determinants of cancer. In particular, studies of genetically susceptible subgroups may enable the detection of low levels of risk due to certain common exposures that have eluded traditional epidemiologic methods. Further, by identifying susceptibility genes and their pathways of action, it may be possible to identify previously unsuspected carcinogens. Finally, by gaining a more comprehensive understanding of environmental and genetic risk factors, there should emerge new clinical and public health strategies aimed at preventing and controlling cancer.
Blood samples are an excellent source of large amounts of genomic DNA. However, alternative sources are often needed in epidemiological studies because of difficulties in obtaining blood samples. This report evaluates the buccal cytobrush and alcohol-containing mouthwash protocols for collecting DNA by mail. Several DNA extraction techniques are also evaluated. The study was conducted in two phases. In phase 1, we compared cytobrush and mouthwash samples collected by mail in two different epidemiological studies: (a) cytobrush samples (n = 120) from a United States case-control study of breast cancer; and (b) mouthwash samples (n = 40) from a prospective cohort of male United States farmers. Findings from phase 1 were confirmed in phase 2, where we randomized cytobrush (n = 28) and mouthwash (n = 25) samples among participants in the breast cancer study to directly compare both collection methods. The median human DNA yield determined by hybridization with a human DNA probe from phenol-chloroform extracts was 1.0 and 1.6 microg/2 brushes for phases 1 and 2, respectively, and 27.5 and 16.6 microg/mouthwash sample for phases 1 and 2, respectively. Most (94-100%) mouthwash extracts contained high molecular weight DNA (>23 kb), in contrast to 55-61% of the brush extracts. PCR success rates for amplification of beta-globin gene fragments (268, 536, and 989 bp) were similar for cytobrush and mouthwash phenol-chloroform extracts (range, 94.4-100%). Also, we obtained high success rates in determining the number of CAG repeats in the androgen receptor gene, characterizing tetranucleotide microsatellites in six gene loci, and screening for mutations in the BRCA1/2 genes in a subset of phenol-chloroform DNA extracts. Relative to DNA extracted by phenol-chloroform from cytobrush samples, DNA extracted by NaOH had lower molecular weight, decreased PCR success rates for most assays performed, and unreliably high spectrophotometer readings for DNA yields. In conclusion, although DNA isolated from either mouthwash or cytobrush samples collected by mail from adults is adequate for a wide range of PCR-based assays, a single mouthwash sample provides substantially larger amounts and higher molecular weight DNA than two cytobrush samples.
Experimental studies suggest that vitamin D and calcium protect against cancer by reducing proliferation and inducing differentiation. The effects of vitamin D and calcium may be mediated by the vitamin D receptor (VDR), which is encoded by the VDR gene. The present study investigated whether calcium intake and serum vitamin D, as an integrated measure of intake and endogenous production, were associated with risk of colorectal adenoma, known precursors of invasive colorectal cancer. In addition, the interrelation among vitamin D, calcium, and FokI polymorphism of the VDR gene was investigated. Persons (239) with histologically confirmed colorectal adenomas and 228 control individuals without colorectal adenomas confirmed by sigmoidoscopy were enrolled in this case control study conducted at the National Naval Medical Center, Bethesda, MD. We observed an inverse association of serum 25-OH vitamin D [25-(OH)D] with colorectal adenoma. With each 10 ng/ml increase of serum 25-(OH)D, the risk of colorectal adenoma decreased by 26% (odds ratio 0.74, 95% confidence interval 0.60-0.92). The results provided limited evidence for a weak association between calcium intake and colorectal adenoma (odds ratio 0.97, 95% confidence interval 0.93-1.01 per each 100-mg calcium intake). However, the inverse association of serum 25-(OH)D with colorectal adenoma is suggested to be stronger in subjects with calcium intake above the median (P for multiplicative interaction 0.13). The VDR FokI polymorphism was not significantly associated with colorectal adenoma and did not modify the effect of vitamin D or calcium. In conclusion, the study results suggested a protective effect for vitamin D on colorectal adenoma.
To study genetic risk factors for common diseases, researchers have begun collecting DNA specimens in large epidemiologic studies and surveys. However, little information is available to guide researchers in selecting the most appropriate specimens. In an effort to gather the best information for the selection of specimens for these studies, we convened a meeting of scientists engaged in DNA banking for large epidemiologic studies. In this discussion, we review the information presented at that meeting in the context of recent published information. Factors to be considered in choosing the appropriate specimens for epidemiologic studies include quality and quantity of DNA, convenience of collection and storage, cost, and ability to accommodate future needs for genotyping. We focus on four types of specimens that are stored in these banks: (1) whole blood preserved as dried blood spots; (2) whole blood from which genomic DNA is isolated, (3) immortalized lymphocytes from whole blood or separated lymphocytes, prepared immediately or subsequent to cryopreservation; and (4) buccal epithelial cells. Each of the specimens discussed is useful for epidemiologic studies according to specific needs, which we enumerate in our conclusions.
Smoking is a known risk factor for bladder cancer. The product of the GSTM1 gene, glutathione S-transferase M1 (GSTM1), is involved in the detoxification of polycyclic aromatic hydrocarbons found in tobacco smoke; a homozygous deletion of this gene in approximately 50% of Caucasians and Asians results in a lack of GSTM1 enzyme activity. Most studies examining the relation between bladder cancer and GSTM1 have reported an increased risk associated with a lack of GSTM1 activity. The authors performed meta- and pooled analyses of published and unpublished, case-control, genotype-based studies that examined this association (17 studies, 2,149 cases, 3,646 controls) and excluded studies conducted in populations with a high prevalence of exposure to known bladder cancer risk factors other than tobacco smoke. Using random effects models in the meta-analysis, the authors obtained a summary odds ratio of 1.44 (95% confidence interval (CI): 1.23, 1.68) for GSTM1 null status with all studies included. Results from studies with at least 100 cases and 100 controls produced a summary odds ratio of 1.42 (95% CI: 1.26, 1.60). Pooled analyses using original data sets from 10 studies (1,496 cases and 1,444 controls) and adjusting for age, sex, and race produced similar results. There was no evidence of multiplicative interaction between the GSTM1 null genotype and ever smoking in relation to bladder cancer, although there was a suggestion of additive interaction (additive interaction = 0.45, 95% CI: -0.03, 0.93). These results indicate that, among populations studied to date, GSTM1 null status is associated with a modest increase in the risk of bladder cancer.
Reproductive characteristics, alcohol intake and polymorphisms in genes encoding sex-steroid metabolizing enzymes might influence the risk of hormone-related cancers by changing circulating concentrations of sex hormones. The relationship between these factors and serum concentrations of estradiol, progesterone, androstenedione, testosterone and DHEA was evaluated in a cross-sectional study of 218 pre-menopausal women from Kaiser Permanente Health Plan in Portland, Oregon. Risk factor information was obtained from questionnaires and hormone serum concentrations were determined by radioimmunoassays. Genotypes for CYP11A 5'UTR(tttta)n, CYP17 5'-UTR -34 T>C, CYP19 IVS4(ttta)n, CYP1B1 (L432V and S453N) and COMT (V158M) were determined from genomic DNA samples. Increasing number of full-term pregnancies was associated with a significant decrease in late-follicular progesterone levels (p-trend = 0.03). Increasing alcohol consumption was associated with higher estradiol levels averaged through the menstrual cycle (p-trend = 0.009) and higher progesterone levels during luteal phase (p-trend = 0.04). Androstenedione and testosterone levels were higher among light to moderate drinkers compared to non-drinkers, although we only observe a significant trend with increasing levels of alcohol consumption for androstenedione. Women heterozygous or homozygous for the CYP1B1 L432V or the S453N polymorphisms had increased luteal estradiol levels (p-value = 0.04 for L432V and 0.04 for S453N). None of the other factors evaluated was significantly associated with serum concentration of hormones. In conclusion, results from this cross-sectional study of pre-menopausal women provide support for an association between light to moderate alcohol intake and elevated levels of estrogen and androgen levels. Our data suggest that circulating levels of progesterone might be related to parity and alcohol consumption, however the biological plausibility of the observed associations is unclear. We found little support for an influence of the evaluated genetic polymorphisms in the steroid synthesis and metabolism pathway on serum hormone levels, except for a possible effect of the CYP1B1 L432V and S453N polymorphisms on serum estradiol levels.
Buccal cells were collected from 29 participants, by use of mouthwash rinses, and were split into equal aliquots, with one aliquot irradiated by electron-beam (E-beam) irradiation equivalent to the sterilizing dosage used by the U.S. Postal Service and the other left untreated. Aliquots were extracted and tested for DNA yields (e.g., TaqMan assay for quantifying human genomic DNA), genomic integrity, and amplification-based analysis of genetic variants (e.g., single-nucleotide polymorphisms [SNPs] and single tandem repeats [STRs]). Irradiated aliquots had lower median DNA yields (3.7 microg/aliquot) than untreated aliquots (7.6 microg/aliquot) (P<.0005) and were more likely to have smaller maximum DNA fragment size, on the basis of genomic integrity gels, than untreated aliquots (P<.0005). Irradiated aliquots showed poorer PCR amplification of a 989-bp beta-globin target (97% for weak amplification and 3% for no amplification) than untreated aliquots (7% for weak amplification and 0% for no amplification) (P<.0005), but 536-bp and 268-bp beta-globin targets were amplified from all aliquots. There was no detectable irradiation effect on SNP assays, but there was a significant trend for decreased detection of longer STRs (P=.01) in irradiated versus untreated aliquots. We conclude that E-beam irradiation reduced the yield and quality of buccal-cell specimens, and, although irradiated buccal-cell specimens may retain sufficient DNA integrity for some amplified analyses of many common genomic targets, assays that target longer DNA fragments (>989 bp) or require whole-genome amplification may be compromised.
Too many reports of associations between genetic variants and common cancer sites and other complex diseases are false positives. A major reason for this unfortunate situation is the strategy of declaring statistical significance based on a P value alone, particularly, any P value below.05. The false positive report probability (FPRP), the probability of no true association between a genetic variant and disease given a statistically significant finding, depends not only on the observed P value but also on both the prior probability that the association between the genetic variant and the disease is real and the statistical power of the test. In this commentary, we show how to assess the FPRP and how to use it to decide whether a finding is deserving of attention or "noteworthy." We show how this approach can lead to improvements in the design, analysis, and interpretation of molecular epidemiology studies. Our proposal can help investigators, editors, and readers of research articles to protect themselves from overinterpreting statistically significant findings that are not likely to signify a true association. An FPRP-based criterion for deciding whether to call a finding noteworthy formalizes the process already used informally by investigators--that is, tempering enthusiasm for remarkable study findings with considerations of plausibility.
Immunohistochemical characterization of tumor tissues in epidemiological studies is a promising approach to identify breast cancer subtypes with distinct etiology. The recent development of the tissue microarray (TMA) technique allows for standardized, rapid, and cost-effective immunohistochemical characterization of many cases, which is critical in epidemiological studies. Sectioning paraffin blocks at different times results in loss of material, which can be reduced by preparing many sections each time a block is cut. However, data suggest that staining intensity declines in whole sections prepared from conventional paraffin blocks with storage time, resulting in false-negative results. This problem would be accentuated in TMAs because of the limited tissue representation of each case. To evaluate this concern, we prepared a single TMA block from 125 invasive breast carcinomas collected in a population-based case-control study conducted in Poland and compared estrogen receptor (ER-alpha), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression in sections cut and stored for 6 months at room temperature with sections cut from the same TMA block and stained on the same day. Percentage of positive cases for stored versus fresh sections was similar for ER (59.0%) but significantly higher in fresh sections for PR (56.3% versus 64.1%, P = 0.01) and HER2 (45.5% versus 64.4%, P < 0.001). Among cases positive in both stored and fresh sections, the median percentage of immunoreactive cells was significantly reduced and the staining intensity was consistently lower in stored compared with fresh sections. We conclude that loss of immunoreactivity is an important problem in TMAs of breast cancer. Improved methods for sectioning TMAs and storing tissue sections aimed at reducing loss of immunoreactivity are critical for the use of TMAs in epidemiological studies.
Errors in genotype determination can lead to bias in the estimation of genotype effects and gene-environment interactions and increases in the sample size required for molecular epidemiologic studies. We evaluated the effect of genotype misclassification on odds ratio estimates and sample size requirements for a study of NAT2 acetylation status, smoking, and bladder cancer risk. Errors in the assignment of NAT2 acetylation status by a commonly used 3-single nucleotide polymorphism (SNP) genotyping assay, compared with an 11-SNP assay, were relatively small (sensitivity of 94% and specificity of 100%) and resulted in only slight biases of the interaction parameters. However, use of the 11-SNP assay resulted in a substantial decrease in sample size needs to detect a previously reported NAT2-smoking interaction for bladder cancer: 1,121 cases instead of 1,444 cases, assuming a 1:1 case-control ratio. This example illustrates how reducing genotype misclassification can result in substantial decreases in sample size requirements and possibly substantial decreases in the cost of studies to evaluate interactions.
Breast cancers classified by estrogen receptor (ER) and/or progesterone receptor (PR) expression have different clinical, pathologic, and molecular features. We examined existing evidence from the epidemiologic literature as to whether breast cancers stratified by hormone receptor status are also etiologically distinct diseases. Despite limited statistical power and nonstandardized receptor assays, in aggregate, the critically evaluated studies (n = 31) suggest that the etiology of hormone receptor-defined breast cancers may be heterogeneous. Reproduction-related exposures tended to be associated with increased risk of ER-positive but not ER-negative tumors. Nulliparity and delayed childbearing were more consistently associated with increased cancer risk for ER-positive than ER-negative tumors, and early menarche was more consistently associated with ER-positive/PR-positive than ER-negative/PR-negative tumors. Postmenopausal obesity was also more consistently associated with increased risk of hormone receptor-positive than hormone receptor-negative tumors, possibly reflecting increased estrogen synthesis in adipose stores and greater bioavailability. Published data are insufficient to suggest that exogenous estrogen use (oral contraceptives or hormone replacement therapy) increase risk of hormone-sensitive tumors. Risks associated with breast-feeding, alcohol consumption, cigarette smoking, family history of breast cancer, or premenopausal obesity did not differ by receptor status. Large population-based studies of determinants of hormone receptor-defined breast cancers defined using state-of-the-art quantitative immunostaining methods are needed to clarify the role of ER/PR expression in breast cancer etiology.
Electron-beam (E-beam) irradiation, currently being used to sterilize mail addressed to selected ZIP codes in the United States, has significant negative effects on the genomic integrity of DNA extracted from buccal-cell washes. We investigated the yield, composition, and genotyping performance of whole genome amplified DNA (wgaDNA) derived from 24 matched samples of E-beam-irradiated and nonirradiated genomic DNA (gDNA) as a model for the effects of degraded gDNA on the performance of whole genome amplification. gDNA was amplified using the Multiple Displacement Amplification method. Three methods of DNA quantification analysis were used to estimate the yield and composition of wgaDNA, and 65 short tandem repeat and single nucleotide polymorphism genotyping assays were used to evaluate the genotyping performance of irradiated and nonirradiated gDNA and wgaDNA. Compared with wgaDNA derived from nonirradiated gDNA, wgaDNA derived from irradiated gDNA exhibited a significantly reduced yield of wgaDNA and significantly reduced short tandem repeat and single nucleotide polymorphism genotyping completion and concordance rates (P < 0.0001). Increasing the amount of irradiated gDNA input into whole genome amplification improved genotyping performance of wgaDNA but not to the level of wgaDNA derived from nonirradiated gDNA. Multiple Displacement Amplification wgaDNA derived from E-beam-irradiated gDNA is not suitable for genotyping analysis.
BACKGROUND: Skewed X chromosome inactivation may be more common in women with epithelial ovarian cancer and early-onset breast cancer. We tested this hypothesis in a group of 235 breast cancer patients and 253 controls (mean age 45.8 years) from a larger population based case control study. METHODS: We measured X chromosome inactivation with the AR gene assay in lymphocyte DNA digested with the methylation specific enzyme HpaII. We judged skewness using an adjusted measure (relative to the undigested sample) with a cut point of 75%, and an unadjusted measure where skewed was defined as > 90% of the signal from one allele in the HpaII digested sample. RESULTS: There were no significant differences in any of the skewing measures between cases and controls. Using the adjusted skewing measure among pre-menopausal subjects under the age of 50, 14% of cases versus 11% of controls were skewed, OR = 1.2, 95% CI 0.6 to 2.3; using the unadjusted measure, OR = 0.9, 95% CI 0.4 to 2.0. CONCLUSIONS: While we cannot rule out a subtle difference of approximately twofold or less, we have failed to find a significant difference in the prevalence of skewed X chromosome inactivation in younger women with breast cancer compared to controls.
The promise of whole genome amplification (WGA) is that genomic DNA (gDNA) quantity will not limit molecular genetic analyses. Multiple displacement amplification (MDA) and the OmniPlex PCR-based WGA protocols were evaluated using 4 and 5 ng of input gDNA from 60 gDNA samples from three tissue sources (mouthwash, buffy coat, and lymphoblast). WGA DNA (wgaDNA) yield and genotyping performance were evaluated using genotypes determined from gDNA and wgaDNA using the AmpFlSTR Identifiler assay and N = 49 TaqMan SNP assays. Short tandem repeat (STR) and SNP genotyping completion and concordance rates were significantly reduced with wgaDNA from all WGA methods compared with gDNA. OmniPlex wgaDNA exhibited a greater reduction in genotyping performance than MDA wgaDNA. Reduced wgaDNA genotyping performance was due to allelic (all protocols) and locus (OmniPlex) amplification bias leading to heterozygote and locus dropout, respectively, and %GC sequence content (%GC) was significantly correlated with TaqMan assay performance. Lymphoblast wgaDNA exhibited higher yield (OmniPlex), buffy coat wgaDNA exhibited higher STR genotyping completion (MDA), whereas mouthwash wgaDNA exhibited higher SNP genotyping discordance (MDA). Genotyping of wgaDNA generated from < or = 5 ng gDNA, e.g., from archaeological, forensic, prenatal diagnostic, or pathology samples, may require additional genotyping validation with gDNA and/or more sophisticated analysis of genotypes incorporating observed reductions in genotyping performance.
BACKGROUND: Many reported associations between common genetic polymorphisms and complex diseases have not been confirmed in subsequent studies. An exception could be the association between NAT2 slow acetylation, GSTM1 null genotype, and bladder-cancer risk. However, current evidence is based on meta-analyses of relatively small studies (range 23-374 cases) with some evidence of publication bias and study heterogeneity. Associations between polymorphisms in other NAT and GST genes and bladder-cancer risk have been inconsistent. METHODS: We investigated polymorphisms in NAT2, GSTM1, NAT1, GSTT1, GSTM3, and GSTP1 in 1150 patients with transitional-cell carcinoma of the urinary bladder and 1149 controls in Spain; all the participants were white. We also carried out meta-analyses of NAT2, GSTM1, and bladder cancer that included more than twice as many cases as in previous reports. FINDINGS: In our study, the odds ratios for bladder cancer for individuals with deletion of one or two copies of the GSTM1 gene were 1.2 (95% CI 0.8-1.7) and 1.9 (1.4-2.7) respectively (p for trend <0.0001). Compared with NAT2 rapid or intermediate acetylators, NAT2 slow acetylators had an increased overall risk of bladder cancer (1.4 [1.2-1.7]) that was stronger for cigarette smokers than for never smokers (p for interaction 0.008). No significant associations were found with the other polymorphisms. Meta-analyses showed that the overall association for NAT2 was robust (p<0.0001), and case-only meta-analyses provided support for an interaction between NAT2 and smoking (p for interaction 0.009). The overall association for GSTM1 was also robust (p<0.0001) and was not modified by smoking status (p=0.86). INTERPRETATION: The GSTM1 null genotype increases the overall risk of bladder cancer, and the NAT2 slow-acetylator genotype increases risk particularly among cigarette smokers. These findings provide compelling evidence for the role of common polymorphisms in the aetiology of cancer. RELEVANCE TO PRACTICE: Although the relative risks are modest, these polymorphisms could account for up to 31% of bladder cancers because of their high prevalence.
The double-strand break DNA repair pathway has been implicated in breast carcinogenesis. We evaluated the association between 19 polymorphisms in seven genes in this pathway (XRCC2, XRCC3, BRCA2, ZNF350, BRIP1, XRCC4, LIG4) and breast cancer risk in two population-based studies in USA (3,368 cases and 2,880 controls) and Poland (1,995 cases and 2,296 controls). These data suggested weak associations with breast cancer risk for XRCC3 T241M and IVS7-14A>G (pooled odds ratio (95% confidence interval): 1.18 (1.04-1.34) and 0.85 (0.73-0.98) for homozygous variant vs wild-type genotypes, respectively), and for an uncommon variant in ZNF350 S472P (1.24 (1.05-1.48)), with no evidence for study heterogeneity. The remaining variants examined had no significant relationships to breast cancer risk. Meta-analyses of studies in Caucasian populations, including ours, provided some support for a weak association for homozygous variants for XRCC3 T241M (1.16 (1.04-1.30); total of 10,979 cases and 10,423 controls) and BRCA2 N372H (1.13 (1.10-1.28); total of 13,032 cases and 13,314 controls), and no support for XRCC2 R188H (1.06 (0.59-1.91); total of 8,394 cases and 8,404 controls). In conclusion, the genetic variants evaluated are unlikely to have a substantial overall association with breast cancer risk; however, weak associations are possible for XRCC3 (T241M and IVS7-14A>G), BRCA2 N372H, and ZNF350 S472P. Evaluation of potential underlying gene-gene interactions or associations in population subgroups will require even larger sample sizes.
Impaired base-excision repair (BER) function can give rise to the accumulation of DNA damage and initiation of cancer. We evaluated whether genetic variation in six BER pathway genes (XRCC1, ADPRT, APEX1, OGG1, LIG3, and MUTYH) is associated with breast cancer risk in two large population-based case-control studies in the United States (3,368 cases and 2,880 controls) and Poland (1,995 cases and 2,296 controls). A detailed evaluation was first done in a subset of 1,898 cases and 1,514 controls with mouthwash DNA samples in the U.S. study. Significant findings were followed up in the remainder of the U.S. study population that provided cytobrush DNA samples and in the Polish study. Using data from U.S. study participants with mouthwash DNA, we found no significant overall association between breast cancer risk and XRCC1 R280H and R194W, ADPRT V726W, APEX1 D148E, OGG1 S326C, LIG3 R780H, or MUTYH 5' untranslated region. These data suggested a decreased risk for XRCC1Q399R homozygous variants compared with homozygous wild-type in premenopausal women, but these findings were not confirmed when data from cytobrush DNA samples were added [combined odds ratio (OR), 0.8; 95% confidence interval (95% CI), 0.6-1.1] or in the Polish study (OR, 1.0; 95% CI, 0.7-1.5). Meta-analyses based on our data and published data from studies of two single nucleotide polymorphisms in XRCC1 showed no evidence of an overall association between breast cancer risk and homozygous variants versus wild-type for Q399R (OR, 1.1; 95% CI, 1.0-1.2) or R194W (OR, 1.0; 95% CI, 0.7-1.8), although there was a suggestion for an association in Asian populations for Q399R (OR, 1.6; 95% CI, 1.1-2.4; P = 0.02). In conclusion, our results do not support that the polymorphisms evaluated in six BER pathway genes play a major role in breast carcinogenesis, particularly in Caucasian populations.
Nucleotide excision repair (NER) is critical for protecting against damage from carcinogens in tobacco smoke. We evaluated the influence of common genetic variation in the NER pathway on bladder cancer risk by analyzing 22 single nucleotide polymorphisms (SNP) in seven NER genes (XPC, RAD23B, ERCC1, ERCC2, ERCC4, ERCC5, and ERCC6). Our study population included 1,150 patients with transitional cell carcinoma of the urinary bladder and 1,149 control subjects from Spain. Odds ratios (OR) and 95% confidence intervals (95% CI) were adjusted for age, gender, region, and smoking status. Subjects with the variant genotypes for SNPs in four of the seven genes evaluated had small increases in bladder cancer risk compared to subjects with the homozygous wild-type genotypes: RAD23B IVS5-15A>G (OR, 1.3; 95% CI, 1.1-1.5; P = 0.01), ERCC2 R156R (OR, 1.3; 95% CI, 1.1-1.6; P = 0.006), ERCC1 IVS5+33A>C (OR, 1.2; 95% CI, 1.0-1.5; P = 0.06; P(trend) = 0.04), and ERCC5 M254V (OR, 1.4; 95% CI, 1.0-2.0; P = 0.04). A global test for pathway effects indicated that genetic variation in NER characterized by the 22 SNPs analyzed in this study significantly predicts bladder cancer risk (P = 0.04). Pairwise comparisons suggested that carrying variants in two genes could result in substantial increases in risk. Classification tree analyses suggested the presence of subgroups of individuals defined by smoking and NER genotypes that could have substantial increases in risk. In conclusion, these findings provide support for the influence of genetic variation in NER on bladder cancer risk. A detailed characterization of genetic variation in key NER genes is warranted and might ultimately help identify multiple susceptibility variants that could be responsible for substantial joint increases in risk.
Homozygous mutation in the ATM gene causes ataxia telangiectasia and heterozygous mutation carriers may be at increased risk of breast cancer. We studied a total of 22 ATM variants; 18 variants were analyzed in one of two large population-based studies from the U.S. and Poland, and four variants were analyzed in all 2,856 breast cancer cases and 3,344 controls from the two studies. The missense mutation Ser49Cys (c.146C>G, p.S49C), carried by approximately 2% of subjects, was more common in cases than controls in both study populations, combined odds ratio (OR) 1.69 (95% CI, 1.19-2.40; P=0.004). Another missense mutation at approximately 2% frequency, Phe858Leu (c.2572T>C, p.F858L), was associated with a significant increased risk in the U.S. study but not in Poland, and had a combined OR of 1.44 (95% CI, 0.98-2.11; P=0.06). These analyses provide the most convincing evidence thus far that missense mutations in ATM, particularly p.S49C, may be breast cancer susceptibility alleles. Because of their low frequency, even larger sample sizes are required to more firmly establish these associations.
The role of active and passive cigarette smoking in breast cancer etiology remains controversial. Using data from a large population-based case-control study in Poland (2386 cases, 2502 controls) conducted during 2000-2003, we examined the associations between active and passive smoking overall and for different age categories. We also evaluated differences in risk by estrogen receptor (ER) and progesterone receptor (PR) status in tumors, and the potential modification of the smoking association by N-acetyl transferase 2 (NAT2) genotype. Women ever exposed to passive smoking at home or at work had a risk of breast cancer similar to those never exposed to active or passive smoking (OR (95%CI) = 1.11 (0.85-1.46), and no trends were observed with increasing hours/day-years of passive smoking exposure. Active smoking was associated with a significant increase in risk only among women younger than 45 years of age (OR (95%CI) = 1.95 (1.38-2.76); 1.15 (0.93-1.40); 0.91 (0.77-1.09) for < 45, 45-55 and > 55 years of age, respectively; p-heterogeneity < 0.001 for < 45 vs. > 55 years) and prevailed for both ER+ and ER- tumors. The smoking association among women < 45 years was stronger for current than former smokers, and a significant trend was observed with duration of smoking (p = 0.04). NAT2 slow vs. rapid/intermediate acetylation genotype was not related to breast cancer risk (0.99 (0.87-1.13)), and did not significantly modify the smoking relationships. In conclusion, our data indicate that passive smoking is not associated with breast cancer risk; however, active smoking might be associated with an increased risk for early onset breast cancers.
An increased bladder cancer risk has been suggested among users of hair dyes. We evaluated this association among females in a hospital-based case-control study in Spain (152 female incident cases, 166 female controls). The effect of hair dye use was also evaluated among potentially susceptible subgroups defined by NAT1, NAT2, CYP1A2, GSTM1, GSTT1 and GSTP1 genotypes. Use of any hair dye (OR=0.8, CI 0.5-1.4) or of permanent hair dyes (OR=0.8, CI 0.5-1.5) was not associated with increased risk. Small non-significant increases in risks were observed in a lagged analysis that ignores exposures within ten years of diagnosis (OR=1.3, CI 0.8-2.2). No trend in risk with increasing exposure was seen for duration of use, average use or cumulative use. None of the polymorphisms examined significantly modified the hair dye associated risk. Overall, this study does not support an association between hair dye use and bladder cancer.
Breast cancer is a morphologically and clinically heterogeneous disease; however, it is less clear how risk factors relate to tumour features. We evaluated risk factors by tumour characteristics (histopathologic type, grade, size, and nodal status) in a population-based case-control of 2386 breast cancers and 2502 controls in Poland. Use of a novel extension of the polytomous logistic regression permitted simultaneous modelling of multiple tumour characteristics. Late age at first full-term birth was associated with increased risk of large (> 2 cm) tumours (odds ratios (95% confidence intervals) 1.19 (1.07-1.33) for a 5-year increase in age), but not smaller tumours (P for heterogeneity adjusting for other tumour features (Phet) = 0.007). On the other hand, multiparity was associated with reduced risk for small tumours (0.76 (0.68-0.86) per additional birth; Phet = 0.004). Consideration of all tumour characteristics simultaneously revealed that current or recent use of combined hormone replacement therapy was associated with risk of small (2.29 (1.66-3.15)) and grade 1 (3.36 (2.22-5.08)) tumours (Phet = 0.05 for size and 0.0008 for grade 1 vs 3), rather than specific histopathologic types (Phet = 0.63 for ductal vs lobular). Finally, elevated body mass index was associated with larger tumour size among both pre- and postmenopausal women (Phet = 0.05 and 0.0001, respectively). None of these relationships were explained by hormone receptor status of the tumours. In conclusion, these data support distinctive risk factor relationships by tumour characteristics of prognostic relevance. These findings might be useful in developing targeted prevention efforts.
High estrogen exposure in utero may increase breast cancer risk later in life. However, studies of the associations between perinatal factors presumed to affect the fetal hormonal environment and breast cancer risk are inconsistent. We used data from a population-based case-control study of 2,386 incident breast cancers and 2,502 controls in Poland to evaluate risks associated with various perinatal characteristics. After adjusting for confounders, we found a significant trend (p = 0.01) of breast cancer risk with birth weight (OR = 1.54, 95% CI 1.08-2.19 for birth weights >4,000 g vs. <2,500 g). Subjects with a high birth order (> or =6) were at reduced risk (OR = 0.81, 0.61-1.06) when compared with first born subjects. Birth weight was somewhat a stronger risk predictor among subjects whose cancers were diagnosed at 50 years of age or older (OR = 1.84, 1.19-2.85) than among those with cancers diagnosed at younger ages (OR = 1.14, 0.61-2.12). Subjects whose mothers smoked during their pregnancies were at slightly higher risk than those who never smoked (OR = 1.21, 0.99-1.47), but the risk was similar to mothers who only smoked at other times (OR = 1.22, 0.81-1.84). Breast cancer risk was not related to paternal smoking, maternal age, gestational age or twin status. Our results add support to the growing evidence that some perinatal exposures may relate to breast cancer risk. Additional studies are needed to confirm associations and clarify the biologic mechanisms underlying these associations.
We examined the effects of dose, type of tobacco, cessation, inhalation, and environmental tobacco smoke exposure on bladder cancer risk among 1,219 patients with newly diagnosed bladder cancer and 1,271 controls recruited from 18 hospitals in Spain. We used unconditional logistic regression to estimate odds ratios (OR) and 95% confidence intervals (95% CI) for the association between bladder cancer risk and various characteristics of cigarette smoking. Current smokers (men: OR, 7.4; 95% CI, 5.3-10.4; women: OR, 5.1; 95% CI, 1.6-16.4) and former smokers (men: OR, 3.8; 95% CI, 2.8-5.3; women: OR, 1.8; 95% CI, 0.5-7.2) had significantly increased risks of bladder cancer compared with nonsmokers. We observed a significant positive trend in risk with increasing duration and amount smoked. After adjustment for duration, risk was only 40% higher in smokers of black tobacco than that in smokers of blond tobacco (OR, 1.4; 95% CI, 0.98-2.0). Compared with risk in current smokers, a significant inverse trend in risk with increasing time since quitting smoking blond tobacco was observed (> or =20 years cessation: OR, 0.2; 95% CI, 0.1-0.9). No trend in risk with cessation of smoking black tobacco was apparent. Compared with men who inhaled into the mouth, risk increased for men who inhaled into the throat (OR, 1.7; 95% CI, 1.1-2.6) and chest (OR, 1.5; 95% CI, 1.1-2.1). Cumulative occupational exposure to environmental tobacco smoke seemed to confer increased risk among female nonsmokers but not among male nonsmokers. After eliminating the effect of cigarette smoking on bladder cancer risk in our study population, the male-to-female incidence ratio decreased from 8.2 to 1.7, suggesting that nearly the entire male excess of bladder cancer observed in Spain is explained by cigarette smoking rather than occupational/environmental exposures to other bladder carcinogens.
Four single nucleotide polymorphisms (SNPs) in CYP1B1 (Ex2 + 143 C > G, Ex2 + 356 G > T, Ex3 + 251 G > C, Ex3 + 315 A > G) cause amino acid changes (R48G, A119S, L432V and N453S, respectively) and are associated with increased formation of catechol estrogens; however, epidemiologic evidence only weakly supports an association between these variants and breast cancer risk. Because genetic variability conferring increased susceptibility could exist beyond these putative functional variants, we comprehensively examined the common genetic variability within CYP1B1. A total of eight haplotype-tagging (ht)SNPs (including Ex3 + 315 A > G), in addition to two putatively functional SNPs (Ex2 + 143 C > G and Ex3 + 251 G > C), were selected and genotyped in a large case-control study of Polish women (1995 cases and 2296 controls). Haplotypes were estimated using the expectation-maximization algorithm, and overall differences in the haplotype distribution between cases and controls were assessed using a global score test. We also evaluated levels of tumor CYP1B1 protein expression in a subset of 841 cases by immunohistochemistry, and their association with genetic variants. In the Polish population, we observed two linkage disequilibrium (LD)-defined blocks. Neither haplotypes (global P-value of 0.99 and 0.67 for each block of LD, respectively), nor individual SNPs (including three putatively functional SNPs) were associated with breast cancer risk. CYP1B1 was expressed in most tumor tissues (98%), and the level of expression was not related to the studied genetic variants. We found little evidence for modification of the estimated effect of haplotypes or individual SNPs by age, family history of breast cancer, or tumor hormone receptor status. The present study provides strong evidence against the existence of a substantial overall association between common genetic variation in CYP1B1 and breast cancer risk.
Construction of tissue microarrays (TMAs) to efficiently characterize large sets of noninvasive epithelial lesions in the breast by immunohistochemistry is an appealing investigative approach, but presents technical challenges. We report methodologic studies performed to optimize methods for building TMAs from noninvasive breast tissues collected in a large case-control study of breast cancer. Using a manual arraying technique with 2.0-mm diameter needles, we constructed TMAs from specimens obtained from 32 women with breast cancer containing the following targets: (1) 28 terminal duct lobular units (TDLUs); (2) 28 ductal carcinomas in situ, and (3) 23 invasive carcinomas. Using careful target selection, we achieved representation of approximately 80% of noninvasive targets with sustained preservation through section 30 of the TMAs. Immunohistochemical staining of TDLU targets demonstrated positive staining for estrogen receptor (ER) in 30.8% of tubules and for progesterone receptor (PR) in 50.0%. To establish an efficient method to evaluate staining results in TDLUs, we created a categorical scoring system to approximate the percentage of tubules containing positive stained cells (<10%, 10% to 50%, >or=50%), and compared the results with those obtained by tubule counting. Comparison between the two methods demonstrated exact agreement for 70.8% of ER and 79.2% of PR stains without two-category discrepancies. ER/PR expression levels in multiple (up to 4) noninvasive targets of the same tissue type (TDLU or DCIS) from a single block showed good correlation. These data suggest that it is feasible to produce TMAs of noninvasive breast structures, albeit with careful selection of targets, and that immunostains of such cores may permit efficient immunohistochemical characterization of peritumoral tissues. Additional exploration of this approach is needed.
Buccal cell samples are increasingly used in epidemiological studies as a source of genomic DNA. The accurate and precise quantitation of human DNA is critical for the optimal use of these samples. However, it is complicated by the presence of bacterial DNA and wide inter-individual variation in DNA concentration from buccal cell collections. The paper evaluated the use of ultraviolet light (UV) spectroscopy, Höechst (H33258) and PicoGreen as measures of total DNA, and real-time quantitative polymerase chain reaction (PCR) as a measure of human amplifiable DNA in buccal samples. Using serially diluted white blood cell DNA samples (at a concentration range of 300 to 0.5 ng microl-1), UV spectroscopy showed the largest bias, followed by Höechst, especially for low concentrations. PicoGreen and real-time PCR provided the most accurate and precise estimates across the range of concentrations evaluated, although an increase in bias with decreasing concentrations was observed. The ratio of real-time PCR to PicoGreen provided a reasonable estimate of the percentage of human DNA in samples containing known mixtures of human and bacterial DNA. Quantification of buccal DNA from samples collected in a breast cancer case-control study by PicoGreen and real-time PCR indicated that cytobrush and mouthwash DNA samples contain similar percentages of human amplifiable DNA. Real-time PCR is recommended for the quantification of buccal cell DNA in epidemiological studies since it provides precise estimates of human amplifiable DNA across the wide range of DNA concentrations commonly observed in buccal cell DNA samples.
BACKGROUND: We assessed use of nonaspirin nonsteroidal anti-inflammatory drugs (NSAID), aspirin, paracetamol (acetaminophen), phenacetin, and metamizol (dipyrone) and risk of bladder cancer and their interaction with polymorphisms in drug-metabolizing genes. METHODS: We analyzed personal interview data from 958 incident bladder cancer cases and 1,029 hospital controls from a multicenter case-control study in Spain. A drug matrix was developed to estimate cumulative lifetime dose of active ingredients. Polymorphisms in GSTP1, SULT1A1, CYP2E1, CYP2C9, and NAT2 were examined. RESULTS: A significant reduction in bladder cancer risk [adjusted odds ratio (OR), 0.4; 95% confidence interval (95% CI), 0.2-0.9] was observed for regular users of nonaspirin NSAIDs compared with never users. Regular users of aspirin experienced no reduction in risk (OR, 1.0; 95% CI, 0.7-1.5). Regular users of paracetamol had no overall increased risk of bladder cancer (OR, 0.8; 95% CI, 0.4-1.3), but our data suggested a qualitative interaction with the GSTP1 I105V genotype. Subjects with at least one copy of the 359L or 144C variant alleles in the NSAID-metabolizing gene CYP2C9 had a slightly decreased risk of bladder cancer (OR, 0.8; 95% CI, 0.7-1.0; P = 0.037); however, having at least one copy of the 359L or 144C variant alleles did not significantly modify the protective effect of nonaspirin NSAID use. CONCLUSION: Regular use of nonaspirin NSAIDs was associated with a reduced risk of bladder cancer, which was not modified by polymorphisms in the NSAID-metabolizing gene CYP2C9. We found no evidence of an overall effect for paracetamol or aspirin use.
Because catechol-O-methyltransferase (COMT) catalyzes the addition of methyl groups to stabilize catechol estrogens that may induce DNA damage, genetic variants could influence breast cancer risk. To comprehensively characterize genetic variation in this gene, we selected haplotype-tagging single nucleotide polymorphisms (htSNP) in COMT. A total of 11 htSNPs (including COMT Val(158)Met) were selected based on the resequencing and dense genotyping approach of the Breast and Prostate Cancer Cohort Consortium. htSNPs were genotyped in a population-based, case-control study in Poland (1,995 cases and 2,296 controls). Individual SNPs were not significantly associated with risk. Haplotypes were estimated using the expectation-maximization algorithm. Overall differences in the haplotype distribution between cases and controls were assessed using a global score test. The TGAG haplotype (frequent in 4.3% of controls), in a linkage disequilibrium (LD) block that included the 3' untranslated region (UTR) of COMT, was associated with breast cancer risk (odds ratio, 1.29; 95% confidence interval, 1.06-1.58) compared with the most common haplotype TGAA; however, the global test for haplotype associations was not significant (P = 0.09). Haplotypes in another LD block, which included COMT Val(158)Met, were not associated with breast cancer risk (global P = 0.76). Haplotype-breast cancer risk associations were not significantly modified by hormonally related risk factors, family history of breast cancer, or tumor characteristics. In summary, our data does not support a substantial overall association between COMT haplotypes and breast cancer. The suggestion of increased risk associated with a haplotype in the 3' UTR of COMT needs to be confirmed in independent study populations.
Genetic polymorphisms in DNA repair genes may impact individual variation in DNA repair capacity and alter cancer risk. In order to examine the association of common genetic variation in the base-excision repair (BER) pathway with bladder cancer risk, we analyzed 43 single nucleotide polymorphisms (SNPs) in 12 BER genes (OGG1, MUTYH, APEX1, PARP1, PARP3, PARP4, XRCC1, POLB, POLD1, PCNA, LIG1, and LIG3). Using genotype data from 1,150 cases of urinary bladder transitional cell carcinomas and 1,149 controls from the Spanish Bladder Cancer Study we estimated odds ratios (ORs) and 95% confidence intervals (CIs) adjusting for age, gender, region and smoking status. SNPs in three genes showed significant associations with bladder cancer risk: the 8-oxoG DNA glycosylase gene (OGG1), the Poly (ADP-ribose) polymerase family member 1 (PARP1) and the major gap filling polymerase-beta (POLB). Subjects who were heterozygous or homozygous variant for an OGG1 SNP in the promoter region (rs125701) had significantly decreased bladder cancer risk compared to common homozygous: OR (95%CI) 0.78 (0.63-0.96). Heterozygous or homozygous individuals for the functional SNP PARP1 rs1136410 (V762A) or for the intronic SNP POLB rs3136717 were at increased risk compared to those homozygous for the common alleles: 1.24 (1.02-1.51) and 1.30 (1.04-1.62), respectively. In summary, data from this large case-control study suggested bladder cancer risk associations with selected BER SNPs, which need to be confirmed in other study populations.
OBJECTIVE: Clarifying age-specific female breast cancer risks and interactions may provide important etiologic clues. METHOD: Using a population-based case-control study in Poland (2000-2003) of 2,386 incident breast cancer cases and 2,502 control subjects aged 25-74 years, we estimated age-specific breast cancer incidence rates according to risk factors. RESULTS: Breast cancer risks were elevated among women with positive family history (FH), younger age at menarche, older age at first full-term birth, nulliparity, exogenous hormonal usage, and reduced physical activity (PA). Notwithstanding overall risks, we observed statistically significant quantitative (non-crossover) and qualitative (crossover) age interactions for all risk factors except for FH and PA. For example, nulliparity compared to parity reduced breast cancer risk among women ages 25-39 years then rates crossed or reversed, after which nulliparity increased relative risks among women ages 40-74 years. CONCLUSION: Though quantitative age interactions could be expected, qualitative interactions were somewhat counterintuitive. If confirmed in other populations, qualitative interactions for a continuous covariate such as age will be difficult to reconcile in a sequential (multistep or monolithic) 'stochastic' breast cancer model. Alternatively, the reversal of relative risks among younger and older women suggests subgroup heterogeneity with different etiologic mechanisms for early-onset and late-onset breast cancer types.
Tumor necrosis factor (TNF) is critical to regulation of inflammation. Genetic variation in the promoter region of TNF has been associated with expression differences, and a range of auto-immune, infectious, and oncologic diseases. We analyzed eight common single nucleotide polymorphisms (SNPs) (rs746868, rs909253, rs1799964, rs1800630, rs1800750, rs1800629, rs361525, and rs1800610) to capture most of the genetic variation in TNF in addition to SNPs in lymphotoxin-alpha (LTA), a pro-inflammatory cytokine in linkage disequilibrium with the TNF promoter region. SNPs were genotyped in a USA population-based case-control study (3,318 cases, 2,841 controls). Promising results were followed-up in an independent population-based case-control study in Poland (2,228 cases, 2,378 controls). In both studies, women carrying the variant allele of rs361525 were at elevated breast cancer risk compared to the GG genotype (per allele OR = 1.18, 95% CI 1.04-1.35; P for trend = 0.008). Other SNPs were not significantly associated with breast cancer risk. Haplotype analyses did not reveal any additional associations between TNF and breast cancer risk. Data from 5,269 cases and 4,982 controls suggested that the rs361525 A allele, located in the TNF promoter region, was associated with a modest increase in breast cancer risk. Additional studies are required to replicate these findings and to determine whether rs361525 is a causative SNP or is a marker of a causative SNP.
BACKGROUND: The etiology of breast cancer is not well understood and the role of occupational exposures in breast carcinogenesis is still uncertain. METHODS: The population-based case-control study included 2,386 incident breast cancer cases diagnosed in 2000-2003, and 2,502 controls. Lifetime occupational histories and information on other potential breast cancer risk factors were obtained through personal interviews. Conditional logistic regression analyses calculated odds ratios (ORs) associated with various occupations and industries after control for potential confounders. RESULTS: We found statistically significant excesses of breast cancer among engineers (OR=2.0; 95% CI: 1.0-3.8), economists (2.1; 1.1-3.8), sales occupations-retail (1.2; 1.0-1.5), and other sales occupations (1.2; 1.0-1.5). Industries showing significantly elevated risks included special trade contractors (2.2; 1.2-4.3), electronic and electric equipment manufacturers (1.7; 1.1-2.7); and public administration/general government n.e.c. (2.7; 1.3-5.7). Each of these findings was supported by a statistically significant positive trend for duration of employment (P<0.05). A decreased breast cancer risk was observed in janitors and cleaners (0.7; 0.5-0.8). CONCLUSIONS: In this study, we found few associations for breast cancer and occupations or industries. The suggestive findings for the electronic and electric equipment manufacturing industry and for the occupations with potential exposure to magnetic fields deserve further evaluation.