[Show abstract][Hide abstract] ABSTRACT: Clin Microbiol Infect 2011; 17: 386–394
The present study aimed to evaluate a novel fluorescence in situ hybridization (FISH) assay for detecting the high-risk human papillomavirus (HR-HPV) DNA and signal pattern in cervical cytology specimens and for identifying cervical intraepithelial neoplasia (CIN) lesions. One hundred and ninety-six liquid-based cytology specimens with CIN were recruited. The signal pattern (punctate, mixed punctate and diffuse, and diffuse) detected by FISH was compared with E6 mRNA and correlated with histological classification. FISH and E6-type specific polymerase chain reaction (PCR) had fair to good agreement for detecting HPV DNA across all grades of CIN (kappa coefficient, 0.37–0.73). Among 44 samples of negative FISH and positive E6 type-specific PCR in HPV 16, 18, 31, 33, 52 and 58, 82% (36/44) of E6 mRNA were not detected, in contrast to 41% (48/118) of positive FISH and positive E6 type-specific PCR (p <0.0001). Among HR-HPV DNA positive cases tested by the FISH assay, the specificity of predicting CIN3 using the punctuate pattern is higher than that using E6 mRNA (96.3% vs. 44.8%). The punctate pattern was 0% in patients with <CIN1 lesions, 8.7% for CIN1 lesions, 6.1% for CIN2 lesions, and 34.0% for CIN3 lesions (p 0.001). The odds ratios were 8.7-fold higher (2.7–27.8, p <0.0001) for the punctate pattern versus the mixed punctate and diffuse pattern, and the diffuse pattern, for predicting CIN3 lesions. The novel FISH assay is comparable to PCR for detecting HPV DNA in cervical cytology with CIN lesions. The punctate signal pattern detected by the FISH assay can be more biologically and clinically relevant for clinically detecting CIN3 lesions.