Longitudinal patterns of penile human papillomavirus detection among Ugandan men: A prospective sampling study.

Abstract

Human papillomavirus (HPV) is a highly prevalent sexually transmitted infection, yet data on HPV dynamics in African male populations remain limited. We conducted a prospective study of 300 human immunodeficiency virus-negative, sexually active Ugandan men enrolled at circumcision and followed every 6 weeks for 6 months. Penile swabs were collected by clinical staff, and HPV DNA was detected using the Roche polymerase chain reaction-based linear array assay at six time points. We estimated both point and cumulative prevalence of any and high-risk HPV, and assessed empirical and bootstrapped absolute and relative differences to quantify bias from relying solely on baseline testing. The baseline prevalence of any HPV was 49.7%, increasing to 66.0% when cumulative results were considered. Among 27 men with swabs with amplifiable viral or cellular DNA at all six visits, 100% tested positive at least once. HPV detection was often transient: only 6.1% of men with any HPV were consistently positive at all visits. Men who were ever HPV positive reported a younger age at first sex and more sexual partners. Single-timepoint testing significantly underestimated HPV prevalence, with relative biases of 24.7% for any HPV and 40.3% for high-risk HPV. Bootstrap analyses showed that reducing absolute bias to below 5% required two visits for any HPV and three for high-risk HPV, whereas achieving relative bias below 5% required three and five visits, respectively. The results remained consistent in sensitivity analyses that excluded pre-circumcision visits. These findings highlight the limitations of cross-sectional HPV testing and emphasize the importance of repeated sampling for accurate surveillance and vaccine impact assessments in male populations.