117 Molecular analysis for ovarian cancer detection in patient-friendly samples Table 1: (Continued) n % Age: median (IQR) Healthy controls: 110 Sample type Urine 30 60 (53 - 74) Cervicovaginal self-sample 40 60 (60 - 60) Clinician-taken cervical scrape 40 60 (60 - 60) *Including one mixed clear cell and low-grade endometrioid carcinoma. DNA methylation levels are elevated in cervical scrapes and urine samples of women with ovarian masses The discriminatory power of qMSP assays was verified in tissue, in which all markers showed clear significant differences when comparing methylation levels in normal fallopian tube (n=22) with HGSOC (n=35) tissues (p<0.0001; Supplemental Figure 1, Mann-Whitney U). The feasibility of ovarian cancer detection in urine by methylation analysis was evaluated by testing nine methylation markers in full void (i.e. unfractionated) urine, urine supernatant, and urine sediment of healthy controls and patients diagnosed with a benign or malignant ovarian mass (Figure 1, Supplemental Figure 2-4). When comparing healthy controls with ovarian cancer patients, three markers showed a significant discrimination in full void urine (C2CD4D, p=0.008; CDO1, p=0.022; MAL, p=0.008, Mann-Whitney U), one in urine supernatant (MAL, p=0.001) and one in urine sediment (GHSR, p=0.018, Mann-Whitney U). Benign and malignant masses revealed comparably high methylation levels for most methylation markers, except for GHSR. GHSR showed significantly elevated methylation levels in the urine sediment of ovarian cancer patients (p=0.024, Mann-Whitney U; Figure 1, Supplemental Figure 4). Similarly, the feasibility of ovarian cancer detection in cervicovaginal self-samples and clinician-taken cervical scrapes by methylation analysis was assessed by testing the same methylation markers. While methylation levels of two markers were significantly increased in clinician-taken cervical scrapes of ovarian cancer patients as compared to controls (C2CD4D, p=0.001; CDO1, p=0.004, Mann-Whitney U), benign and malignant ovarian masses could not be distinguished using these markers (Figure 1, Supplemental Figure 5). None of the markers were significantly elevated in cervicovaginal self-samples when comparing these groups (Figure 1, Supplemental Figure 6). Numbers were insufficient to compare methylation levels between different histological subtypes and stages. 5
RkJQdWJsaXNoZXIy MjY0ODMw