123 Molecular analysis for ovarian cancer detection in patient-friendly samples Nevertheless, considering our relatively small sample size, we do not exclude the use of cervicovaginal self-samples for ovarian cancer detection yet. The optimization of preanalytical factors, such as increased input of original sample or improved DNA isolation methods, could enhance the ovarian cancer signal in vaginal samples. Alternatively, a non-tumor DNA driven approach could be useful for ovarian cancer detection in cervicovaginal self-samples, as recently described by Barrett et al (36). Their signature consisted of epigenetic differences in cervical cells and allowed ovarian cancer detection in cervical scrapes with an area under the receiver operating characteristic curve value of 0.76. Larger cohort studies, such as the Screenwide study (37), will provide further insight into the use of cervicovaginal self-samples for ovarian cancer detection. Strengths of this study include the collection of a unique paired sample series of both patients diagnosed with a benign ovarian mass and with a malignant ovarian tumor, covering most histological subtypes. Moreover, urine and cervicovaginal self-samples were collected from home to assess the feasibility and potential of home-based sampling for ovarian cancer. The successful sequencing of urine cfDNA of ovarian cancer patients provides opportunities for future (epi)genome profiling using short- or long-read sequencing technologies. Although we have demonstrated the potential diagnostic value of urine for ovarian cancer, this study is limited by still relatively low sample numbers and the lack of early-stage cancers (≤ FIGO stage 2A). Given the heterogeneous nature of benign and malignant ovarian masses, larger sample series are needed to conclude on the clinical applicability of home-collected cervicovaginal self-samples and urine for ovarian cancer detection. Furthermore, direct comparisons with paired plasma samples using DNA-based and other molecular biomarkers (e.g. HE4) would be informative for future studies. This study supports limited existing data on ovarian cancer detection in cervical scrapes by DNA methylation analysis. Moreover, it provides first proof of concept that urine yields increased methylation levels of ovarian cancer-associated genes and contains ovarian cancer-derived DNA as demonstrated by SCNA analysis. Our findings support continued research into urine biomarkers for ovarian cancer detection and highlight the importance of including benign ovarian masses in future studies. Molecular biomarker testing in patient-friendly samples could facilitate earlier ovarian cancer detection and triage women presenting with an ovarian mass to manage specialist referral. Yet, further studies investigating alternative urine (methylation) biomarkers are warranted to develop a clinically useful test. 5
RkJQdWJsaXNoZXIy MjY0ODMw