This study will construct a longitudinal risk model of VaIN according to the HPVs distribution of cervix and vaginal for those had CIN2+. The study will include three arms to complete the follow-up data for the previous cohort constructed, and prospectively recruit new subjects with the appropriate inclusion/excluding criteria in order to increase sample size of this study.

The long-term outcomes of patients with antecedent cervical intraepithelial neoplasia (CIN) were lacking. A systematic review published in 2008 identified no studies that met the review eligibility criteria for the prevalence of HPV types among vaginal precursors and cancers. Two sporadic studies containing 16 and 81 cases were reported in that review. Given that vaginal intraepithelial neoplasia (VAIN) post-hysterectomy for CIN is an uncommon disease which has a prevalence of being 1 to 6 %; it has the potential to progress to malignancy. Although the etiology of VAIN has not been as thoroughly investigated as that of CIN, evidence implicates that the human papillomavirus (HPV) as the probable carcinogenic agent for the onset and evolution of some VAIN and vaginal cancers.

This study will construct a longitudinal risk model of VaIN according to the HPVs distribution of cervix and vaginal for those had CIN2+.

Inclusion Criteria:

1. Age >= 20 years
2. Those with a previous history of CIN+

Those with a previous history of cervical cancer or CIN before incident VaIN and those with VaIN detected concomitantly with cervical cancer or CIN for whom both vaginal biopsy and cervical specimens