ABSTRACT
OBJECTIVE: We recently demonstrated that a self-collected sample tested with a high-throughput polymerase chain reaction-based high-risk human papillomavirus (HR-HPV) assay is equal in sensitivity to a physician-obtained direct endocervical sample. We now explore some secondary screening options to improve specificity. METHODS: The Shenzhen Cervical Cancer Screening Trial II is a multisite, population-based cross-sectional cervical cancer screening study conducted in Guangdong Province, China. Two HR-HPV assays were used for self-collected specimens, and 3 assays were used for directly collected specimens along with cytology. The polymerase chain reaction-based matrix-assisted laser desorption/ionization time-of-flight mass spectrometry assay reported 14 HR-HPV types. Any patient with a positive lesion on any assay or cytology underwent colposcopy and biopsy. RESULTS: A total of 8,556 women with a mean age of 38.9 years (range = 25-54 years) were included in the analysis. Primary self-collection had a sensitivity of 94.3% and a specificity of 87.5% (for cervical intraepithelial neoplasia grade 3 or cancer). Secondary cervical cytology had a sensitivity and specificity of 83.0% and 95.2%, respectively, which would require a pelvic examination and sacrifice some sensitivity. Secondary genotyping for HPV types 16 or 18 had a sensitivity and specificity of 53.9% and 97.7%, respectively; and HPV types 16, 18, 31, 33, 45, 52, and 58 had a sensitivity and specificity of 92.2% and 90.4%, respectively. CONCLUSIONS: Genotyping is efficient if it is part of the primary test result. It potentially identifies a high percentage of the cancers (types 16/18 = 84.5% in China).
