Reproductive factors with respect to breast cancer risk and breast cancer survival

Abstract

Aims: The primary aim of this thesis was to examine the potential relationship between indirect markers of exposure to hormones during pregnancy and the risk of and survival from breast cancer, with special emphasis on young patients. Our specific objectives were as follows: to determine whether the association between placental weight and offspring size, on the one hand, and maternal mortality from breast cancer, on the other, are influenced by tumor characteristics; to examine the association between birth weight and risk of breast cancer in the female member of opposite-sexed twins; and to investigate whether familial factors influence previously reported association between reproductive factors and risk of breast cancer.

Methods: Based on the Swedish Quality Register of Breast Cancer, two different cohort studies were designed in the Stockholm-Gotland and Uppsala-Örebro regions, where records on characteristics of breast cancer have been collected since 1992. The first cohort was restricted to women who had a pregnancy between 1982 and 1989, and subsequently developed breast cancer. The cohort included 1,067 subjects and 180 deaths, and was conducted to investigate if placental weight is associated with maternal risk of dying from breast cancer, taking tumor characteristics into account. In the second study, we studied the possible association between birth weight and maternal risk of death from breast cancer, also taking tumor characteristics into account. We included 6,019 women who had a pregnancy between 1973 and 2008 and subsequently developed breast cancer, of whom 1017 died from the disease. Two case-control studies were also performed. In a nested case-control study, involving the female members of opposite sexed twin pairs, 543 cases and 2715 controls were included to investigate the potential association between offspring birth weight and risk of breast cancer, as well as a possible modifying effect of birth weight of the male twin sibling. Information on the twins (including birth weight, birth height, head circumference and gestational age of the females, and birth weight of the male co-twin) was extracted from the Swedish Twins Register and data on women diagnosed with breast cancer from the Swedish Cancer Register. A second case-control study examined the potential modifying effect of familial factors on the association between reproductive factors and the risk of breast cancer. All women who delivered between 1973 and 2010 and had a full sister were selected as the study population, using the Swedish Medical Birth Register. Information on breast cancer was obtained from the Swedish Cancer Register and sisters were identified using the Swedish Multi-Generation Register. The cases examined included all parous women diagnosed with breast cancer between 1973 and 2010 who were 50 years old or younger and had at least one sister who also gave birth during this same period. The two control groups were sister controls (including the sister without breast cancer and closest in age to the case) and population controls (all parous women without breast cancer with at least a full sister except those in the sisters control group). In total, 8,349 cases, 8,349 sister controls, and 1,053,688 population controls were used.

Results: Our findings indicate that the association between higher placental weight in connection with the most recent pregnancy and maternal risk of mortality from premenopausal breast cancer is dependent on the receptor status of the tumor. A positive association was more pronounced in the case of ER-/PR- tumors, but we did not find a dose– response association. Birth weight demonstrated no association with maternal mortality from premenopausal breast cancer, even in analyses stratified by the time that elapsed between pregnancy and cancer diagnosis, tumor stage, and receptor status. There was an inverse association between birth-weight-for-gestational age and mortality from premenopausal breast cancer among uniparous women. The nested case-control study of opposite-sexed twins did not reveal any statistically significant association between birth weight and risk of breast cancer. Furthermore, we observed no associations between other birth characteristics, including co-twin birth weight, and the risk of developing pre- or postmenopausal breast cancer. Our last study provided some evidence that the association between reproductive factors and maternal risk of breast cancer or between maternal factors and maternal risk of breast cancer may differ when using population or sister controls. We found that parity exhibited an inverse association to premenopausal breast cancer using population controls and was a risk factor using sister controls, suggesting a gene-environment interaction. Very preterm delivery (<31 weeks) was associated with a higher breast cancer risk using sister controls than when population controls were used, also suggesting a gene-environment interaction. environment interaction, as was the association between gestational age and the risk of breast cancer.

Conclusions: We found some, but no strong evidence in support of the hypothesis that higher hormone levels during pregnancy are associated with mortality from premenopausal breast cancer. The hypothesis was supported when placental weight was employed as indirect indicator of estrogen levels during pregnancy, although birth weight showed no such association. The more pronounced effect of placental weight among ER-/PR- tumors suggests that premenopausal hormonal exposure might exert a greater impact on such tumors. The association between parity and risk of premenopausal breast cancer was modified by a gene-environment interaction, as was the association between gestational age and the risk of breast cancer.