Theses Doctoral

Breast Cancer Risk Factors and Associations with Breast Cancer Tumor Characteristics in High Risk Populations

Work, Meghan E.

Background: Estrogen receptor (ER)- and progesterone receptor (PR)-negative (ER-PR-) breast cancer is associated with higher grade and poorer prognosis compared with other breast cancer subtypes. High parity, coupled with lack of breastfeeding, has been associated with an increased risk of ER-PR- cancer. The mechanism of this etiology is unclear, and may be obfuscated by ER and PR correlation with each other as well as other prognostic tumor characteristics.
Methods: Using population-based and clinic-based ascertained cases and controls from the Breast Cancer Family Registry, I examined reproductive risk factors, including parity, breastfeeding, and oral contraceptive (OC) use, in relation to ER and PR status, using polytomous logistic regression (for the population-based data) and the method of generalized estimating equations (GEE) (for the clinic-based data) as well as the pseudo-conditional likelihood approach, which accounts for correlated outcome variables.
Results: High parity (≥ 3 live births) combined with lack of breastfeeding, was positively associated with ER-PR- tumors (odds ratio [OR]=1.57, 95% confidence interval [CI] 1.10-2.24, population-based cases vs. controls) relative to nulliparity. There was no association with ER-PR- tumors and parity in women who breastfed (OR=0.93, 95%CI 0.71-1.22) relative to nulliparous women. Associations with ER-PR- cancer were higher across all races/ethnicities among women who did not breastfeed compared with women who did. Population-based and clinic-based data were generally in agreement (OR=2.07, 95% CI 1.09-3.91, clinic-based cases vs. controls, relative to nulliparity). When adjusted for the correlation of PR-status and grade, to ER-status, the association between high parity +lack of breastfeeding and ER- status, was maintained. OC use before year 1975 was associated with an increased risk of ER-PR- tumors (OR=1.32, 95% CI 1.04-1.67, population-based data, cases vs. controls) relative to never use of OCs. For women who began OC use in 1975 or later there was no increased risk. Analysis of OC use in clinic-based data agreed with the findings of the population-based data.
Conclusions: My findings support that there are modifiable factors for ER-PR- breast cancer, and that breastfeeding in particular may mitigate the increased risk of ER-PR-cancers seen from multiparity. The mechanism of both risk and risk mitigation may operate primarily through the estrogen, rather than progesterone, pathway.


  • thumnail for Work_columbia_0054D_14822.pdf Work_columbia_0054D_14822.pdf application/pdf 17.1 MB Download File

More About This Work

Academic Units
Thesis Advisors
Terry, Mary Beth
Ph.D., Columbia University
Published Here
July 21, 2018