Our study assessed the factors associated with cervical squamous intra-epithelial lesions (SILs) and invasive cervical cancer, with special attention to human immunodeficiency virus (HIV) and human papillomavirus (HPV) infections. Women from 3 outpatient gynecology clinics of Abidjan, Côte d'Ivoire, were screened for cervical abnormalities: 151 women with low-grade SILs and 151 controls, 60 with high-grade SILs and 240 controls, and 13 with invasive cancer and 65 controls were enrolled in 3 case-control studies. Controls were chosen at random among the women without lesions, with a frequency matching for age and center. We used the PCR method for the detection of cervical HPV DNA and the restriction fragment length polymorphism analysis for HPV typing. HIV antibody testing and CD4 cell count were performed. In multivariate analyses, factors associated with cervical lesions were: for low-grade SILs, HPV positivity, HIV-1 seropositivity and parity >3; for high-grade SILs, HPV positivity, chewing tobacco, HIV-1 seropositivity and illiteracy, and for invasive cancer, HPV positivity only. We found a diversity of HPV types associated with SILs. In HIV-1-infected women, SILs occurred at an early stage of HIV disease. Women infected with both HIV-1 and HPV were at much higher risk of SILs than women infected with each of these 2 viruses separately. Invasive cancer was linked to HIV-2 infection in univariate analysis only. Our results suggest that the relation of SILs with HIV-1 infection is mainly explained by HPV infection and that HIV-1-infected African women may not often reach the invasive stage of cervical cancer.
PIP: The factors associated with cervical squamous intraepithelial lesions (SILs) and invasive cervical cancer were assessed in case-control studies of women from 3 outpatient gynecology clinics in Abidjan, Ivory Coast. Enrolled were 151 women with low-grade SILs and 151 controls, 60 women with high-grade SILs and 240 controls, and 13 women with invasive cervical cancer and 65 controls. Human papillomavirus (HPV) was detected in 75% of the high-grade SILs and cancers compared to 20% in the age-matched control groups. In the multivariate analysis, low-grade SILs were associated with HPV positivity, HIV-1 seropositivity, and parity above 3; for high-grade SIL, these risk factors were HPV positivity, chewing tobacco, HIV-1 seropositivity, and illiteracy. Cervical cancer was associated only with HPV positivity. Women infected with both HPV and HIV-1 were at a significantly higher risk of SILs than women infected with only one of these two viruses. These findings suggest that the association of SILs with HIV-1 infection is primarily explained by HPV infection and that HIV-infected African women may not live to reach the invasive stage of cervical cancer. The feasibility of cervical screening directed preferentially to African women with a low educational level or multipara merits assessment.