Objective: The purpose of this study was to evaluate the effect of different combination models of high-risk human papilloma viruses (HPV) genotyping in triaging Chinese women with atypical squamous cells of undetermined significance (ASCUS). Methods: We established a screening cohort of 3,997 Chinese women who underwent cervical cytology and HPV genotyping test. Women with ASCUS cytology underwent punch biopsy under colposcopy/endocervical curettage. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of different combination models of HR-HPV genotyping calculated that cervical intraepithelial neoplasia 2 or higher (CIN2+) on histology were endpoints. Results: Of the full sample, 393 women had ASCUS. Among ASCUS women with a CIN2 lesion, the prevalence for HPV were 40.0% (type 16), 10.0% (type 18), 0.0% (type 33), 30.0% (type 52), 40.0% (type 58), and 30.0% (other nine types). For ASCUS women with a CIN3 lesion, the prevalence for HPV were 68.4% (type 16), 15.8% (type 18), 10.5% (type 33), 31.6% (type 52), 15.8% (type 58), and 36.8% (other nine types). Combination model including HPV16/18/33/52/58 for predicting CIN2+ lesion in women with ASCUS had relatively higher sensitivity [93.1% (78.0, 98.1)], specificity [75.8% (71.2, 79.9)], PPV [23.5% (16.7, 32.0)], and NPV [99.3% (97.4, 99.8)] than other combination models. Moreover, the referral rate of HPV16/18/33/52/58 (29.3%) was lower than HR-HPV (36.1%). Conclusions: The study demonstrates that specific HR-HPV types HPV16/18/33/52/58 may be an effective strategy in ASCUS triage. This improves the subsequent selection of ASCUS patients.
Keywords: atypical squamous cells of undetermined significance (ASCUS); cervical cancer; genotyping; human papilloma virus (HPV); triage.