87 DNA methylation testing for endometrial cancer detection in patient-friendly sample types Correlation coefficients for the individual markers between the different sample types are presented in Table 2. Correlation coefficients between all markers are illustrated in Supplementary Figure 3, showing that the majority of markers correlated highly between urine and self-samples while less correlation was seen when comparing urine and self-samples with cervical scrapes. Table 2: Correlation of DNA methylation markers between sample types. ADCYAP1 BHLHE22 CDH13 CDO1 GALR1 GHSR HAND2 SST ZIC1 Urine vs Self-sample 0.72 0.80 0.76 0.72 0.77 0.72 0.72 0.43 0.64 Urine vs Scrape 0.70 0.67 0.09 0.51 0.14 0.63 0.01 0.44 0.58 Scrape vs Self-sample 0.68 0.77 0.09 0.61 0.07 0.62 0.04 0.60 0.61 The Spearman’s rank correlation coefficients (r) of DNA methylation markers ADCYAP1, BHLHE22, CDH13, CDO1, GALR1, GHSR, HAND2, SST, and ZIC1 between paired samples of women diagnosed with endometrial cancer (n = 103). The Spearman’s rank correlation coefficient was calculated based on 2log-transformed Ct ratios. Ct, cycle threshold. r ≤ .19 poor correlation, r = .20 - .39 fair correlation, r = .40 - .59 moderate correlation, r = .60 - .79 strong correlation, r ≥ .80 very strong correlation. Performance of DNA methylation analysis for endometrial cancer detection The diagnostic performance of each marker was assessed individually by univariable logistic regression analysis and validated by LOOCV. In urine, the non-cross-validated AUCs of the DNA methylation markers to detect endometrial cancer ranged between 0.61 – 0.93, in cervicovaginal self-samples between 0.62 – 0.91, and in clinician-taken cervical scrapes between 0.61 – 0.95 (Figure 2, Table 3). Most markers, seven out of nine, showed the highest performance in urine: ADCYAP1 (AUC 0.83), BHLHE22 (AUC 0.85), CDH13 (AUC 0.90), GALR1 (AUC 0.79), GHSR (AUC 0.93), HAND2 (AUC 0.71), and ZIC1 (AUC 0.78). The remaining markers performed best in clinician-taken cervical scrapes: CDO1 (AUC 0.95) and SST (AUC 0.74). Nonetheless, except for CDH13 in urine, the 95% confidence interval of AUCs were overlapping between paired sample types. DNA methylation marker panels, rather than single genes, may increase the diagnostic accuracy for endometrial cancer detection. Multivariable logistic regression with backward selection was applied to identify the most optimal three-marker combinations for each sample type. This selection procedure created marker panels with increased AUC values of 0.95 (95% CI: 0.92-0.98) for urine by combining CDH13 + GHSR + SST, 0.94 (95% CI: 0.90-0.97) for cervicovaginal self-samples by combining CDO1 + GHSR + ZIC1, and 0.97 (95% CI: 0.960.99) for clinician-taken cervical scrapes by combining CDH13 + CDO1 + ZIC1. Marker panels allowed endometrial cancer detection with increased sensitivity, without a major impact on specificity. This was especially the case in cervicovaginal self-samples and cliniciantaken cervical scrapes. The sensitivity and specificity of single genes in urine ranged 4
RkJQdWJsaXNoZXIy MjY0ODMw