Some hormonal factors in the etiology of endometrial cancer

Abstract

The main purpose of this dissertation was to study the impact of some hormone-related factors in the etiology of endometrial neoplasms, i.e. hormone replacement therapy, use of oral contraceptives, serum levels of 20 different organochlorine substances, and polymorphisms in the estrogen receptor alpha (ER) gene.

We conducted two population-based case-control studies among post-menopausal women. In the first one, 789 women with a reported diagnosis of primary endometrial cancer and 3368 age-frequency-matched control women were enrolled from all over Sweden. These women answered a questionnaire on use of oral contraceptives and hormone replacement therapy, among other questions. We used unconditional logistic regression to calculate odds ratios (OR) as estimates of relative risks. We found a duration-dependent increase in the relative risk of endometrial cancer both among women who used orally administered estriol, 1-2 mg/day (multivariate OR following 5 or more years of use: 3.0; 95% confidence interval [CI] 2.0-4.4) and medium-potency estrogens without addition of progestins (multivariate odds ratio following 5 or more years of use: 6.7; 95% CI 4.3-10.5), compared to women who never used these substances. Following combined estrogen-progestin use, the association was considerably weaker (multivariate odds ratio following 5 or more years of use: 1.6; 95% CI 1.1-2.4) than for estrogens without progestins, and the increased relative risk was confined to women using cyclic regimens (i.e. with less than 16 days of added progestins, mostly 10 days; multivariate OR for 5 or more years of use: 2.9, 95% CI 1.8-4.6). Continuous addition of progestins reduced the relative risk (OR for 5 or more years of use 0.2, 95% CI 0.1-0.8), compared to women who never used these hormone replacement regimens. Women who used oral contraceptives had a 30% decreased relative risk for endometrial cancer compared to women who never used these compounds. The protective effect of combined oral contraceptive use remained for at least 20 years after cessation of use. Subsequent use of hormone replacement, did not modify these protective effects.

In the second study we enrolled women - 154 with endometrial cancer and 205 age-frequency-matched controls - who never used hormone replacement, and who were residents in 12 selected counties of Sweden. We collected questionnaire information and blood samples, from which we separated serum for analysis of organochlorine compounds (pesticides and polychlorinated biphenyls), and extracted DNA for analyses of ER gene polymorphisms. We found no significant associations between endometrial cancer risk and serum levels of the 10 pesticides and 10 polychlorinated biphenyls studied. We conclude that our data do not support the hypothesis that organochlorine exposures studied increase risk for endometrial cancer. Results from analysis of the ER gene polymorphism showed a multivariate OR for the Xba I XX genotype of 0.51 (95% CI 0.20-1.27) compared to the xx genotype. The PP Pvu II genotype was also associated with a non-significantly decreased risk for endometrial cancer (multivariate OR 0.69, 95% CI 0.34-1.43) compared with the pp genotype. The multivariate OR for two short TA (< 19 repeats) alleles versus two long alleles was 1.50 (95% CI 0.72-3.17). We observed the same pattern of results in an expanded group of subjects, which included women who had used hormone replacement (in total 288 cases and 392 controls). These data suggest that variants of the ER gene may be associated with an altered risk of endometrial cancer.