Molecular biomarkers for gynecological cancer detection in cervicovaginal samples and urine

Cervical cancer screening programs have resulted in a major decrease in cancer diagnoses and deaths, because cervical (pre)cancer is more often detected and treated at an earlier stage. However, cervical cancer screening is hampered by low participation rates. With the recent transition from cytology-based to high-risk human papillomavirus (hrHPV)-based screening in The Netherlands, accidentally detected cases of endometrial cancer remain undetected. Endometrial cancer is nowadays diagnosed using invasive biopsy procedures. Hence, there is a clinical need for non- or minimally-invasive methods to detect cervical and endometrial cancer, for which testing of tumor-associated molecular markers in urine offers a promising liquid biopsy. PART I: CERVICAL INTRAEPITHELIAL NEOPLASIA AND CERVICAL CANCER In Chapter 2 we demonstrated that tumor DNA in the urine of urogenital cancer patients originates not only from shedding of the local tumor, but also from transrenal excretion of circulating tumor DNA. Urine samples can be separated into three fractions using centrifugation: full void urine, urine sediment and urine supernatant. In Chapter 3, we assessed which urine fraction is most suitable for cervical intraepithelial neoplasia (CIN3) and cervical cancer detection by DNA methylation analysis. All urine fractions performed excellent for cervical cancer detection. We also demonstrated the potential of CIN3 detection by urinary DNA methylation analysis, and showed best performance of urine sediment to detect CIN3. In Chapter 4 we evaluated the performance of DNA methylation analysis and of high-risk human papillomavirus (hrHPV) DNA testing for cervical cancer and high grade CIN lesion (CIN2 and CIN3) detection in urine. For comparison, paired cervicovaginal self-samples and cervical scrapes were analyzed. This prospective study indicates that DNA methylation and hrHPV DNA testing in urine can be used to detect cervical cancer and high grade CIN lesions. In Chapter 5 we used questionnaires to evaluate the experiences and preferences on urine samples, cervicovaginal self-samples and clinician-taken cervical scrapes in women referred for colposcopy and scheduled for a large loop excision of the transformation zone (LLETZ) procedure. Most women considered self-collection of a urine sample and a cervicovaginal self-sample as acceptable, easy and trustworthy. Compared to cervicovaginal self-samples, urine samples were considered to be even more acceptable and reliable. Moreover, the majority of the women preferred urine collection for future cervical cancer screening. From PART I it can be concluded that DNA methylation analysis and hrHPV DNA testing in urine is a promising solution to lower the barrier of women to participate in cervical cancer screening. This could consequently greatly improve cervical cancer prevention strategies for women currently unreached by conventional screening methods. PART II: ENDOMETRIAL CANCER In Chapter 6 we investigated the feasibility of endometrial cancer detection in urine using DNA methylation analysis. A systematic comparison of different urine fractions demonstrated that full void urine is most optimal for endometrial cancer detection. DNA methylation analysis in full void urine showed an excellent discriminatory power for endometrial cancer detection. In Chapter 7 we systematically reviewed the available literature to assess the potential of this approach and define methylation markers deserving further development. A systematic search was conducted in which fifteen methylation markers from nine studies were selected with high potential for endometrial cancer detection in cytological specimens. Chapter 8 - EMBARGO The results obtained in PART II indicate the high potential of DNA methylation marker testing in cervicovaginal samples and urine, as a promising alternative diagnostic strategy to detect endometrial cancer, including the early stages. Potential clinical applications include triaging women with postmenopausal bleeding symptoms, population screening of asymptomatic women, and monitoring women with increased endometrial cancer ris