Prevention Strategies for Gynecologic Cancers

Abstract

This dissertation focused on prevention strategies for ovarian and cervical cancers. Primary prevention of ovarian cancer was explored in Aim 1 using data from the Ovarian Cancer Association Consortium (OCAC; 7,743 cases; 11,882 controls) with the goal of developing a risk stratification model for ovarian cancer. This model was the most comprehensive to date and included body mass index (BMI), height, age at menarche, parity, breastfeeding, incomplete pregnancy, age at last pregnancy, tubal ligation, age at menopause, combined oral contraceptive use, depot-medroxyprogesterone acetate use, menopausal hormone therapy use, first-degree family history of ovarian cancer, endometriosis, and a 36-variant polygenic risk score. We found that associations between ovarian cancer and family history of the disease and endometriosis were modified by menopausal status, but no pairwise interactions between the 15 risk factors themselves were identified. Hence, we developed an ovarian cancer 15-factor multiplicative risk stratification model separately for pre- and post-menopausal women (based on age). Our model stratifies women into finer risk profiles than prior models thereby allowing us to identify people who are candidates for ovarian cancer precision prevention strategies.

Tertiary prevention for ovarian cancer was examined in Aim 2 by studying the associations between lifestyle and reproductive factors and residual disease after ovarian cancer primary cytoreductive surgery (PCS). We included 2,169 OCAC participants with advanced stage high-grade serous ovarian cancer who underwent PCS. We found that parous compared to nulliparous women and menopausal estrogen users compared to never users were statistically significantly less likely to have macroscopic residual disease after PCS after adjusting for relevant clinical factors (OR=0.65, 95% CI 0.45-0.93, p=0.018 and OR=0.69, 95% CI 0.48-1.00, p=0.048, respectively). Conversely, women who had ever breastfed compared to those who had not were more likely to have residual disease (adjusted OR=1.41, 95% CI 1.03-1.92, p=0.032). If these findings are replicated, this scope of work has tremendous potential to assist with ovarian cancer precision treatment. These factors could be included in treatment decision tools to help determine whether a patient should undergo PCS or have neoadjuvant chemotherapy followed by interval debulking surgery.

Lastly, cervical cancer secondary prevention was studied in Aim 3 where a cross-sectional survey on awareness of, experience with, and attitudes toward cervical cancer screening was carried out among urban (n=202) and rural women (n=196) in Southern Vietnam. Women in both areas reported similarly low awareness of Human papillomavirus (HPV; 37.6% in urban and 34.2% in rural areas had ever heard of it) and low cervical cancer screening (51.8% in urban; 49.1% in rural). Urban participants were statistically significantly more willing to try HPV self-sampling for cervical cancer screening than rural participants (56.2% in urban; 42.2% in rural; OR=2.02, 95% CI 1.26-3.23) adjusting for demographic and socioeconomic factors. Rural women were more likely to have the concern of self-sampling incorrectly (73.4% in urban; 82.5% in rural; adjusted OR=0.49, 95% CI 0.28-0.88) while urban women were more likely to fear HPV testing revealing cancer (59.9% in urban; 47.1% in rural; adjusted OR=1.78, 95% CI 1.11-2.86). Improving health literacy and healthcare access and developing rural-urban tailored health education programs are warranted to reduce the cervical cancer burden in Southern Vietnam.

This dissertation has tremendous translational potential to decrease the burden of gynecologic cancers by assisting with prevention, screening, and treatment for ovarian and cervical cancers.