BACKGROUND: Human papillomavirus (HPV) is the leading cause of cervical cancer (CC) in women. Abnormal CC screening is followed up with colposcopy, a procedure that utilizes a magnified image of the cervix. Criteria have been developed based on lesion characteristics to aid the clinician in targeting areas at high risk for pre-cancer and cancer for biopsy. HPV types 16 and 18 are the most common genotypes associated with CC and have also been associated with more severe colposcopic findings. However, HPV 16 and 18 have become less prevalent with the introduction of the HPV vaccine thus leaving HPV infections with potentially less obvious sites for biopsy. Our study examines whether HPV vaccination status impacts the severity of colposcopic findings in women with high-risk HPV (hrHPV) infections, which may have implications on the utility of colposcopy in HPV vaccinated women.

METHODS: Women were enrolled in the UCSF HPV natural history study between 1990 and 2004, which included healthy women ages 13-21 years who were routinely examined with colposcopy, colpophotography, and HPV DNA testing every 4-6 months. Colposcopic photographs for 60 vaccinated and 56 unvaccinated women were reviewed. Colpophotographs were included in final analysis only if hrHPV testing was positive at that visit date. Outcomes included 1) colposcopic impression (normal vs. abnormal) and 2) colposcopy severity score (derived from 0 – 7 point scale for each lesion based on vascular changes, size, acetowhitening, and demarcation). Chi-squared analysis and Fischer’s Exact Test were used to test for associations between vaccination status, hrHPV genotype, and colposcopic findings for all colpophotographs and separately for individual women using the colpophotograph with highest colposcopy score.

RESULTS: Of 1427 colpophotographs reviewed, 260 met criteria for inclusion in final analysis. Compared to colpophotographs from women with non-16/18 hrHPV, colpophotographs from women with 16/18 hrHPV had significantly higher rates of abnormal colposcopic impressions (44% vs. 76%, respectively, p < 0.001) and higher mean colposcopy scores (2.9 vs. 4.7, respectively, p < 0.001). Colpophotographs from vaccinated women with non-16/18 hrHPV had significantly higher colposcopy scores compared to those from unvaccinated women with non-16/18 hrHPV infections (3.8 vs. 2.9, respectively, p < 0.05). There were no significant differences between vaccinated and unvaccinated women with non-16/18 hrHPV in analysis of individual women. Post-hoc examination of cervical lesions in vaccinated women revealed that 63% of cervical lesions in vaccinated women originated prior to date of first vaccination; the mean colposcopy score for lesions originating prior to vaccination was 5.39, compared to 1.00 for lesions arising after date of first vaccination.

CONCLUSIONS: Our study supports prior findings that HPV types 16 and 18 are associated with more severe colposcopic findings compared to other hrHPV types. The differences observed between vaccinated and unvaccinated women with non-16/18 hrHPV are explained by the high number of cervical lesions with severe colposcopic features pre-dating first vaccination. The disappearance of these differences in sub-analysis of individual women also suggests that a few vaccinated women with severe cervical disease were primarily driving these differences. Overall, our study suggests that colposcopy will be influenced by vaccination and the near eradication of HPV 16/18 lesions; however, considering vaccination does not have therapeutic benefit for pre-existing lesions, colposcopy is likely to remain a useful part of cervical cancer screening for vaccinated women— especially those vaccinated after onset of sexual activity.