Detection of recurrent endometrial cancer via DNA methylation analysis of cervicovaginal self-samples and urine

DNA methylation analysis of self-collected samples has shown potential for primary endometrial cancer detection. Here, we aimed to explore DNA methylation testing as a non-invasive alternative for post-treatment surveillance. Endometrial cancer patients (n = 43) without recurrence collected pre- and post-treatment cervicovaginal self-samples and urine at home. Additionally, 17 patients with recurrence provided these samples at the time of recurrence. Healthy controls were used as reference. In total, 207 cervicovaginal self-samples and 203 urine samples were tested for nine markers (ADCYAP1, BHLHE22, CDH13, CDO1, GALR1, GHSR, HAND2, SST and ZIC1) using quantitative methylation-specific PCR. Diagnostic performance was assessed using previously established logistic regression models. In patients without recurrence, most marker levels decreased post-treatment compared to pre-treatment, in both sample types. DNA methylation positivity in cervicovaginal self-samples reduced from 90.7% pre-treatment to 19.5% post-treatment (p < .0001). In urine, a decrease from 80.5% to 20.9% was observed (p < .0001). In patients with recurrence, DNA methylation levels of CDH13, CDO1, GALR1, GHSR and HAND2 in cervicovaginal self-samples, and CDO1, GHSR and HAND2 in urine, were higher than in patients without recurrence. DNA methylation positivity was 62.5% in cervicovaginal self-samples and 58.8% in urine among patients with any recurrence, and 100% in cervicovaginal self-samples and 90.0% in urine of patients with a local recurrence specifically. These findings support methylation-based testing as a promising post-treatment surveillance method for endometrial cancer patients. This non-invasive approach could reduce the follow-up burden on patients and healthcare, alleviating challenges related to resource constraints while improving patient comfort.