FAM19A4 methylation analysis in self-samples compared with cervical scrapes for detecting cervical (pre)cancer in HPV-positive women

Background: High-risk human papillomavirus (hrHPV)-positive women require triage to identify those with cervical high-grade intraepithelial neoplasia and cancer (>= CIN3 (cervical intraepithelial neoplasia grade 3)). FAM19A4 methylation analysis, which detects advanced CIN and cancer, is applicable to different sample types. However, studies comparing the performance of FAM19A4 methylation analysis in hrHPV-positive self-samples and paired physician-taken scrapes are lacking. Methods: We compared the performance of FAM19A4 methylation analysis (and/or HPV16/18 genotyping) in self-samples and paired physician-taken scrapes for >= CIN3 detection in hrHPV-positive women (n = 450,18-66 years). Results: Overall FAM19A4 methylation levels between sample types were significantly correlated, with strongest correlation in women with >= CIN3 (Spearman's rho 0.697, P<0.001). The performance of FAM19A4 methylation analysis and/or HPV16/18 genotyping did not differ significantly between sample types. In women >= 30 years, >= CIN3 sensitivity of FAM19A4 methylation analysis was 78.4% in self-samples and 88.2% in scrapes (ratio 0.89; CI: 0.75-1.05). In women <30 years, >= CIN3 sensitivities were 37.5% and 45.8%, respectively (ratio 0.82; CI: 0.55-1.21). In both groups, >= CIN3 specificity of FAM19A4 methylation analysis was significantly higher in self-samples compared with scrapes. Conclusions: FAM19A4 methylation analysis in hrHPV-positive self-samples had a slightly lower sensitivity and a higher specificity for >= CIN3 compared with paired physician-taken scrapes. With a similarly good clinical performance in both sample types, combined FAM19A4 methylation analysis and HPV16/18 genotyping provides a feasible triage strategy for hrHPV-positive women, with direct applicability on self-samples.