Colposcopic analysis of genital human papillomavirus infections during an 8-year prospective follow-up

Abstract

A long-term prospective follow-up program has been conducted for genital human papillomavirus (HPV) infections in 532 women in Kuopio since 1981. By regular examinations (at 6-month intervals) using colposcopy, cytology and/or punch biopsy, 426 of these women have been followed up for the minimum of 48 months (mean 71 ± 19 months). These 426 women were included in the present report. The colposcopic data of all women at the first and the latest visits were recorded from the colpophotographs, analysed and correlated with cytology, histology, HPV DNA detection and clinical behavior (i.e. prognosis) of the disease. Colposcopic abnormalities were found in 313 (73.5%) women at the first visit and in 138 (32.4%) patients at the latest visit. HPV infections were diagnosed in 368 (86.4%) women at the first visit and in 75 (17.9%) at the latest visit on light microscopy (i.e. biopsy and/or PAP smear). CIN was found in 54.7% of the women at the first visit. HPV DNA was detected by in situ hybridization in 38.2% of the biopsies assayed. Colposcopical multicentric lesions were found in one third of the women upon examination of the entire lower genital tract. The majority of the lesions were situated within the transformation zone. HPV infection and CIN were most frequently encountered in women with multiple lesions. Progression to CIN III was most frequent in large lesions (28.6%). No colposcopic pattern was able to distinguish HPV infections from CIN lesions, or predict the progression of the former. The results implicate that: (1) colposcopy is an essential method to disclose abnormalities due to CIN and HPV infections, but histological changes cannot be predicted on a single colposcopy; (2) HPV infection is a multicentric disease, making the careful examination of the entire lower genital tract mandatory on every colposcopy; and (3) the HPV progressors cannot be predicted on the basis of the colposcopic patterns.