Incidence of Cervical Squamous Intraepithelial Lesions in HIV-Infected Women

Abstract

This study was conducted to investigate the elevated incidence of squamous intraepithelial lesions (SIL) of the cervix among women infected with human immunodeficiency virus (HIV). Three hundred twenty-eight women with HIV, recruited from HIV/AIDS clinics, were followed over a 2.5-year period and compared with 325 similar but noninfected women, recruited from a methadone treatment program or sexually transmitted disease clinic. All subjects underwent an initial clinical examination that included an interview, cervical smear, cervicovaginal lavage for human papillomavirus (HPV) DNA testing, cervicography, and colposcopy with biopsy of suspicious areas. The examination was repeated every 6 months; cervical biopsies were performed if indicated. The two study groups had similar clinical and demographic characteristics, except that the HIV-infected women were significantly more likely to have a history of prostitution, intravenous drug use, genital herpes, or genital warts than noninfected women (P < .05). The 328 HIV-infected patients had a mean CD4 cell count of 429 × 106/liter at the initial clinical examination. More than half of them (178 of 328, 54 percent) took one or more antiretroviral medications for at least 6 consecutive months during the study. At the initial clinical examination, HPV DNA was found in 143 of the 264 women (54 percent) with HIV whose HPV testing was adequate for analysis and in 85 of the 265 noninfected women (32 percent) with adequate HPV testing. HPV DNA was detected at two or more visits in 61 percent of HIV-infected women (155 of 256) and in 23 percent of noninfected women (58 of 257) who returned for at least two follow-up visits over a 3- to 12-month period. This difference in the percentages of women with persistent HPV infection is significant (P < .001). Over the course of the study, 67 of the 328 women (20 percent) with HIV infection developed SILs, compared with 16 of the 325 non-HIV infected women (5 percent). This difference is significant (P < .001). More than 90 percent (61 of 67, 91 percent) of the lesions among the women in the HIV group and 75 percent (12 of 16) of the lesions in the non-HIV group were low-grade SIL. No women developed invasive carcinoma of the cervix during the study. For the subjects who had 54 months of follow-up, the presence of HPV DNA infection on the cervix at the initial examination was found to be a significant predictor of the development of SILs (P < .001). After 54 months, 48 percent of the women with HPV type 16 or 18 infection and 24 percent of women with HPV infection other than type 16 or 18 developed SIL, compared with only 10 percent of the women who were negative for all types of HPV. SIL developed in 54 percent of the women who were infected with both HIV and HPV type 16 or 18 and in 12 percent of those who were HIV negative but positive for HPV type 16 or 18. Seven percent of women negative for both HIV and HPV and 15 percent of those infected with HIV but negative for HPV developed SIL within 54 months. A univariate analysis of factors associated with SILs found HIV, HPV DNA status, younger age, and young age at first intercourse to be significant predictors of SIL development. In a multivariate analysis, women with persistent HPV type 16 or 18 or HPV other than type 16 or 18 were significantly more likely to develop SILs than women with transient HPV of any type (P < .001 for type 16 or 18, P = .004 for other types). In women who were not infected with HIV, a multivariate analysis showed a continued association of HPV infection with SIL development. Transient HPV infection, persistent HPV infection with types other than 16 or 18, and persistent infection with HPV type 16 or 18 were all significantly associated with the development of SILs of the cervix (P < .06, .03, and .02, respectively). In women infected with HIV, antiretroviral therapy did not have a significant effect on the development of SIL. JAMA 2000;283:1031–1037