The use of host cell DNA methylation analysis in the detection and management of women with advanced cervical intraepithelial neoplasia: A review

Premalignant lesions such as cervical intraepithelial neoplasia (CIN) can develop into cervical cancer through a slow process taking up to 30 years from initial high-risk human papillomavirus (hrHPV) infection. Primary hrHPV DNA testing has superior sensitivity compared with cytology and has a high negative predictive value for high-grade CIN or worse (CIN2+). Because most hrHPV infections do not progress to cervical cancer, a triage test is required to identify hrHPV-positive women with clinically relevant lesions compared those with transient infections. DNA methylation of host cell genes is being explored as a new biomarker and triage method to improve risk stratification in hrHPV-positive women. DNA methylation is the most well-studied host cell modification in cervical cancer. The overall loss of methylation during carcinogenesis, resulting in chromosomal instability, and the silencing of tumor suppressor genes by local hypermethylation of CpG-rich promoter regions contribute to cancer development. Methylated host cell DNA can be detected through quantitative methylation-specific polymerase chain reaction or pyrosequencing on different sample types including cervical tissue, cervical scrapes, self-collected cervicovaginal cells, and urine. Methylation level of promoter regions in host cell genes increases with increasing CIN grade, and high methylation levels in these regions are likely to be associated with advanced cervical lesions with a longer duration of lesion existence. Triage strategies for hrHPV-positive women require a balance between safety (high sensitivity for cervical cancer and CIN3+) and screening-related burden. Methylation analysis has been shown to have high reproducibility, objectivity, and applicability on both clinician-collected and self-collected cervical specimens. Marker panel FAM19A4/miRI24-2 has been evaluated in population-based cervical screening studies, and methylation analysis of FAM19A4 and/or miRI24-2 has shown good sensitivity and specificity for the detection of CIN3+ (sensitivity 68.2%-86.7%, specificity 60.6%-79.3% for clinician-collected samples; sensitivity 65.3%-70.5%, specificity 67.8%-81.3% for self-collected samples). Depending on the setting and inclusion of viral methylation markers, sensitivities from 67.0% to 93.2% and specificities from 41.8% to 85.0% have been reported for the detection of CIN or CIN3+. Multiple studies in the Netherlands have shown that cervical cancer risk among hrHPV-positive women after 14 years is lower following a negative FAM19A4/miRI24-2 methylation result compared with a negative cytology result. Current HPV-based screening programs often use cytology for triage of hrHPV-positive women and refer all women with abnormal cytology at baseline or after 6 months for colposcopy. There is evidence that this strategy may lead to over-referral of women without clinically relevant lesions. Recent data show that the referral rate could be lowered by approximately 34% while still detection all cervical carcinomas and at least 70% of CIN3 lesions by performing secondary triage of hrHPV-positive women with atypical cytology results using FAM19A4/miRI24-2 methylation. Methylation analysis may lead to earlier detection of cervical cancer and advanced CIN lesions if used in combination with hrHPV testing as an exit test for women aging out of screening programs. Recent data justify the further validation of methylation triage testing on cervical scrapes and self-collected specimens of hrHPV-positive women in prospective implementation studies within cervical screening programs. Given robust sensitivity and specificity as well as reproducibility, methylation analysis may eventually be considered a primary cervical screening tool.