Cervical Cancer Screening in Postmenopausal Women: Is It Time to Move Toward Primary High-Risk Human Papillomavirus Screening?

Abstract

Background: Cervical cytology in postmenopausal women is challenging due to physiologic changes of the hypoestrogenic state. Misinterpretation of an atrophic smear as atypical squamous cells of uncertain significance (ASCUS) is one of the most common errors. We hypothesize that high-risk human papillomavirus (hrHPV) testing may be more accurate with fewer false positive results than co-testing of hrHPV and cervical cytology for predicting clinically significant cervical dysplasia in postmenopausal women. Materials and Methods: We conducted a retrospective analysis of 924 postmenopausal and 543 premenopausal women with cervical Pap smears and hrHPV testing. Index Pap smear diagnoses (ASCUS or greater vs. negative for intraepithelial lesion) and hrHPV testing results were compared with documented 5-year clinical outcomes to evaluate sensitivity and specificity of hrHPV compared with co-testing. Proportions of demographic factors were compared between postmenopausal women who demonstrated hrHPV clearance versus persistence. Results: The prevalence of hrHPV in premenopausal and postmenopausal women was 41.6% and 11.5%, respectively. The specificity of hrHPV testing (89.6% [87.4-91.5]) was significantly greater compared with co-testing (67.4% [64.2-70.4]) (p < 0.05). A greater proportion of women with persistent hrHPV developed cervical intraepithelial lesion 2 or greater (CIN2+) compared with women who cleared hrHPV (p = 0.012). No risk factors for hrHPV persistence in postmenopausal women were identified. Conclusion: Our data suggest that hrHPV testing may be more accurate than co-testing in postmenopausal women and that cytology does not add clinical value in this population. CIN2+ was more common among women with persistent hrHPV than those who cleared hrHPV, but no risk factors for persistence were identified in this study.

Keywords: Papanicolaou test; cervical intraepithelial neoplasia; human papillomavirus; sensitivity and specificity.

FIG. 1.

Study inclusions and exclusions.

FIG. 2.

Sensitivity and specificity of cytology, hrHPV, and co-testing of cytology and hrHPV with 95% confidence intervals. (A) There was no difference in sensitivity between postmenopausal and premenopausal women for cytology alone, hrHPV, or co-testing. (B) Specificity for cytology alone, hrHPV, and co-testing was significantly different when comparing postmenopausal and premenopausal cohorts. hrHPV was more specific than co-testing among postmenopausal women. *p < 0.05 comparing postmenopausal to premenopausal women. ^p < 0.05 comparing hrHPV to cytology alone and co-testing.

FIG. 3.

Inclusions and exclusions for a subgroup analysis of “clearers” and “persisters” to evaluate for factors related to persistence of hrHPV among postmenopausal women. hrHPV, high-risk HPV.