Risk factors, diagnosis and prognosis of cervical intraepithelial neoplasia among HIV-infected women

Abstract

The prevalence of cervical intraepithelial neoplasia (CIN) is high among HIV-infected women. Decreased CD4 lymphocytes, high human immunodeficiency viral load (HIVL) and human papillomavirus (HPV) infection are risk factors for CIN. We characterized the prevalence, risk factors and prognosis of histologically-verified CIN among systematically followed HIV-infected women enrolled from a low HIV-prevalence population. The study population comprised 153 HIV-infected women followed between 1989 and 2006. The mean +/- SD duration of follow-up was 5.6 +/- 3.8 years. Demographic as well as treatment-related data were derived from medical reports. During the follow-up, 51 subjects (33%) displayed CIN (16% CIN 1 and 18% CIN 2 +), whereas 102 subjects had Pap smear results of normal cells, atypical squamous cells of uncertain significance, or signs of low-grade squamous intraepithelial lesion (LSIL) but no CIN in histological specimens from the cervix. Only one case of cancer of the uterine cervix was detected. Pap smears were reliable in screening for CIN; 75% of patients with CIN had high-grade squamous intraepithelial lesion (HSIL) or LSIL in Pap smears taken at the time of dysplasia. The incidence of CIN decreased from 12.7 to 3.5 (per 100 subjects) between 2000 and 2005 (P = 0.07). The risk of CIN was not associated with decreased levels of CD4 lymphocytes, duration of HIV infection, use of antiretroviral medication or plasma HIVL. In univariate analysis, bacterial vaginosis (BV) was associated with a significantly increased risk of CIN, whereas parity was associated with lower risk of CIN. Each delivery lowered the risk of CIN by 30% (P = 0.02). The significantly lower risk of CIN among parous women (P = 0.04) persisted in multivariate analysis. CIN was treated by means of loop electrosurgical excision procedure (LEEP), (n = 34). The recurrence rate was low; seven subjects (14%) had a recurrence of CIN during follow-up. The nadir of CD4 lymphocytes was lower (P = 0.04) and the HIVL higher (P = 0.03) among subjects with recurrence of CIN. Duration of HIV infection, use of antiretroviral medication and positive margins in LEEP specimens were indistinguishable among subjects with vs. without recurrence of CIN. The prevalence of CIN is high among systematically managed HIV-infected women. However, the incidence of CIN decreased during the 21st century. BV was associated with an increased risk of CIN whereas parous women had lower risk of CIN. However, the patients with and without CIN could not be distinguished on the basis of previously described risk factors. Regular follow-up by means of Pap smears is warranted in all HIV-infected women.