Clustering patterns of human papillomavirus infections among HIV-positive women in Kenya

BACKGROUND: HIV-positive women are at increased risk of human papillomavirus (HPV) infection, and, especially, multiple infections compared to HIV-negative women. Whether certain HPV types have a tendency to cluster in multiple infections beyond or below what would be expected by shared risk factors (e.g., sexual behavior and the degree of immunosuppression) is unclear. We, therefore, investigated clustering patterns of 44 HPV types in HIV-positive women from Kenya. FINDINGS: HPV status was assessed on cervical scrapings from 498 women using GP5+/6+ PCR and reverse line blot. Logistic regression was used to model type-specific HPV positivity, adjusted for age, specific HPV type prevalence, CD4, combination antiretroviral therapy, and, in the Full Model, individual-level random effects that represent unobservable risk factors common to all HPV types. We found a modest excess of women with co-infections with 2 HPV types (1.12; 95% credible intervals: 1.03-1.21) in the Full Model but no significant associations of individual types. No significant deviations of observed/expected counts were observed for any 2-way combination of HPV types at either the chosen level of significance, p = 0.00005, or at p = 0.01. Findings were substantially similar when women with CIN2/3 were excluded and when they were stratified by use of anti-retroviral therapy or CD4 count. CONCLUSIONS: HPV co-infections occurred at random in the cervix of HIV-positive women as previously found in HIV-negative women. The removal of HPV types through vaccination should not result, therefore, in an increase or decrease in the prevalence of HPV types not targeted by vaccination in immunosuppressed women.