Objective: To investigate the aberrant promoter hypermethylation as a screening tool for cervical adenocarcinomas (CAs) and endometrial adenocarcinomas (EAs) in cervical scrapings.
Study design: A quantitative multiplex methylation-specific polymerase chain reaction approach was used to examine promoter methylation of 5 genes (APC, HIN-1, RAR-beta, RASSF1A and Twist) in biopsy-confirmed CA (n = 31) and EA (n = 27) residual, liquid-based cytology samples. The data of negative for intraepithelial lesions or malignancy and low-grade squamous intraepithelial lesions were used as controls.
Results: Methylation levels of APC, RAR-beta, RASSF1A and Twist were significantly higher in CA than in control cervical samples. For EA, only the methylation levels of RASSF1A differed significantly from those of control. Receiver-operating characteristic analysis demonstrated that APC, RAR-beta and RASSF1A had the ability to distinguish CA/EA, CA and EA from control samples. In CA/EA and CA samples, the best 3-gene combination was RASSF1A/RAR-beta/APC. This 3-gene panel had a sensitivity of 87.0% for CA/EA and of 80.6% for CA and a specificity of 79.3% for both CA/EA and CA. In EA samples, RASSF1A showed the best performance in distinguishing EA from control. The estimated sensitivity of RASSF1A for detecting EA was 63.0%, and its specificity was 96.3%.
Conclusion: This feasibility study demonstrates that quantitative detection of aberrant DNA methylation in cervical scrapings may be a promising new diagnostic tool for the detection of CA and EA.