BREAST CANCER RISK IN HIV: SOCIODEMOGRAPHIC FACTORS AND VIRAL ETIOLOGY

Background: Epidemiologic studies suggest that the risk of breast cancer is lower comparing women with versus without HIV; however, it remains unclear why. Evidence indicates that sociodemographic differences between women with and without HIV do not completely explain this phenomenon. The overarching goal of this dissertation was to assess breast cancer risk factors and incidence among women with HIV to explore potential mechanisms driving this association. Specifically, this dissertation sought to: (aim 1) characterize breast cancer trends over time accounting for the competing risk of death, (aim 2) quantify differences in estrogen comparing women with versus without HIV, and (aim 3) assess the association between HIV viremia and breast cancer risk. Methods: For aims 1 and 3 of this dissertation, the large size and representativeness of the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) was leveraged to conduct two prospective, longitudinal cohort studies. Both aims were assessed from 1996-2016 among women with HIV who initiated antiretroviral therapy (ART) and had no history of any cancer. For aim 2, differences in estrogen comparing women with versus without HIV were examined using a cross-sectional analysis within a nested substudy of the Women’s Interagency HIV Study, a longitudinal prospective cohort of women with and without HIV across the United States. Results: Breast cancer risk did not change over time after accounting for the competing risk of death, though there were significant declines in all-cause mortality. There were no clinically impactful differences in total or free estradiol comparing women with versus without HIV. There was also no association between HIV viremia (measured using cumulative viral load on ART) and breast cancer risk among women with HIV who initiated ART. When the exposure (HIV viremia) was lagged, increasing viremia signaled an inverse association with breast cancer, which got stronger with longer exposure lag. Conclusions: Declining mortality in women with HIV did not influence trends in breast cancer risk over time and may not contribute to the reduced risk of breast cancer observed among women with HIV. There were no differences in total or free estradiol among premenopausal women with and without HIV; however, this should continue to be investigated in postmenopausal women, where breast cancer risk is highest. Finally, although no significant association was observed between HIV viremia and breast cancer, estimates trending in a protective direction should be further explored. As women with HIV continue to age, becoming increasingly at risk for breast cancer, understanding potential differences in etiology is needed and continuing appropriate breast cancer screening remains an important aspect of clinical care