Self-collection of saliva has the potential to provide molecular epidemiologic studies with DNA in a user-friendly way. We evaluated the new Oragene saliva collection method and requested saliva samples by mail from 611 men (ages 53-87 years). We obtained a response rate of, on average, 80% [varying from 89% (ages 67-71 years) to 71% (ages 77-87 years)]. DNA was extracted from 90 randomly selected samples, and its usefulness was evaluated with respect to quality, quantity, and whole-genome amplification (WGA). Visual inspection of DNA on agarose gels showed high molecular weight DNA (>23 kb) and no degradation. Total DNA yield measured with PicoGreen ranged from 1.2 to 169.7 Mg, with a mean of 40.3 Mg (SD, 36.5 Mg) and a median of 29.4 Mg. Human DNA yield was estimated by real-time PCR of the human prothrombin gene to account for 68% (SD, 20%) of total DNA. We did WGA on 81 saliva DNA samples by using the GenomiPhi DNA kit and genotyped both saliva DNA and WGA DNA for 10 single-nucleotide polymorphisms randomly selected from the human genome. Overall genotyping success rate was 96% for saliva DNA and 95% for WGA DNA; 79% of saliva DNA samples and 79% of WGA DNA samples were successfully genotyped for all 10 single-nucleotide polymorphisms. For the 10 specific assays, the success rates ranged between 88% and 100%. Almost complete genotypic concordance (99.7%) was observed between saliva DNA and WGA DNA. In conclusion, Oragene saliva DNA in this study collected from men is of high quality and can be used as an alternative to blood DNA in molecular epidemiologic studies. (Cancer Epidemiol Biomarkers Prev 2006;15(9):1742 -5)

The observed association between risk of developing postmenopausal ER+ breast cancer and alcohol drinking, especially among those women who use postmenopausal hormones, may be important, because the majority of breast tumors among postmenopausal women overexpress ER.

Although obesity is one of the established risk factors for postmenopausal breast cancer, it is not clear whether this positive association differs across estrogen receptor (ER) and progesterone receptor (PR) status of breast tumors. We evaluated the association between body weight and ER/PR defined breast cancer risk stratified by postmenopausal hormone (PMH) use and a family history of breast cancer in the population-based Swedish Mammography Screening Cohort comprising 51,823 postmenopausal women. Relative body weight was measured by body mass index (kg/m 2 ) based on self-reported weight and height collected in 1987 and 1997. Relative risks (RRs) were estimated by hazard ratios derived from Cox proportional hazards regression models. During an average of 8.3-year follow-up, 1,188 invasive breast cancer cases with known ER and PR status were diagnosed. When comparing to normal weight group, we observed a positive association between obesity and risk for the development of ER1 PR1 tumors (RR 5 1.67, 95% CI 5 1.34-2.07) and an inverse association for the development of all PR2 tumors (RR 5 0.68, 95% CI 5 0.47-0.98). Statistically significant heterogeneity was observed in the RRs between ER1 PR1 tumors and all PR2 tumors (p heterogeneity < 0.0001). The positive association of obesity with the development of ER1 PR1 tumors was confined to never-users of PMHs (RR 5 1.90 (CI 95%:1.38-2.61)) and to those without a family history of breast cancer (RR 5 1.82 (CI 95%:1.45-2.29)). Our results support the hypothesis that excess endogenous estrogen due to obesity contributes to an increased risk of ER1 PR1 postmenopausal breast cancer. ' 2006 Wiley-Liss, Inc.Key words: breast cancer; body mass index; estrogen receptor; progesterone receptor Obesity is one of the established risk factors for postmenopausal breast cancer.1 A pooling study of 8 prospective cohorts including the Swedish Mammography Cohort (SMC), and a recent metaanalysis confirmed a positive association between relative body weight and the risk of postmenopausal breast cancer. 2,3

The association between the cytochrome P-450 1B1 (CYP1B1) Val432Leu polymorphism and breast cancer was assessed through a meta-analysis of all published case-control studies and a pooled analysis of both published and unpublished case-control studies from the Genetic Susceptibility to Environmental Carcinogens (GSEC) database ( www.upci.upmc.edu/research/ccps/ccontrol/g_intro.html ). GSEC is a collaborative project that gathers information on studies of metabolic gene polymorphisms and cancer. Thirteen articles were included in the meta-analysis (14,331 subjects; 7,514 cases, 6,817 controls); nine data sets were included in the pooled analysis (6,842 subjects; 3,391 cases, 3,451 controls). A summary meta- or pooled estimate of the association between the CYP1B1 Val432Leu polymorphism and breast cancer could not be calculated because of statistically significant heterogeneity in the point estimates among studies. No association between the CYP1B1 Val432Leu polymorphism and breast cancer was observed in Asians (for Val/Val and Val/Leu combined, odds ratio (OR) = 1.0, 95% confidence interval (CI): 0.8, 1.2). An inverse association was observed in populations of mixed/African origin (OR = 0.8, 95% CI: 0.7, 0.9). The pooled analysis suggested a possible association in Caucasians (for Val/Val and Val/Leu combined, OR = 1.5, 95% CI: 1.1, 2.1), with effect modification across age categories. The observed effect of age on the association in Caucasians indicates that further studies are needed on the role of CYP1B1 Val432Leu in estrogen metabolism according to age, ethnicity, and menopausal status.

Cytochrome P450 1B1 (CYP1B1) is active in the metabolism of estrogens to reactive catechols and of different procarcinogens. Several studies have investigated the relationship between genetic polymorphisms of CYP1B1 and breast cancer risk, however, with inconsistent results. We investigated such an association in postmenopausal Swedish women, with special emphasis on long-term menopausal hormone users, in a large population-based case-control study. We genotyped 1521 cases and 1498 controls for the CYP1B1 single nucleotide polymorphisms (SNPs) m2, m3 and m4 and reconstructed haplotypes. The frequencies of CYP1B1*1, CYP1B1*2, CYP1B1*3 and CYP1B1*4 alleles among controls were estimated to be 0.087, 0.293, 0.444 and 0.175, respectively. It thus appeared that very few haplotypes contained combinations of SNPs at two or three loci and that single SNP genotype data effectively represented haplotypes. Odds ratios (OR) and 95% confidence intervals (CI) were calculated from logistic regression models. We found no overall association between any CYP1B1 genotype and breast cancer risk. The data indicated, however, that women who had used menopausal hormones for 4 years or longer, and carried the CYP1B1*3/*3 genotype may be at increased risk of breast cancer, OR 2.0 (95% CI 1.1-3.5), compared with long-term users without this genotype. We explored the effect of CYP1B1 genotype on breast cancer risk in subgroups defined by body mass index, family history, smoking and catechol-O-methyl transferase genotype, but found no convincing evidence for interaction. In summary, our results strongly indicate that the studied CYP1B1 gene polymorphisms do not influence breast cancer risk overall but may modify the risk after long-term menopausal hormone use.