Determinants of Type-Specific HPV Concordance Across Anatomic Sites in Young Men Who Have Sex with Men

Abstract

Men who have sex with men (MSM) are at high risk for HPV infection and HPV associated cancers. The dynamics of HPV infections at different anatomical sites and whether type-specific HPV infections are detected simultaneously at these sites in this population is not well understood. Information on HPV type concordance can be useful in determining the biological relationship between anatomic sites, the extent of auto-inoculation between sites, and susceptibility of different anatomic areas to specific HPV types. The aim of this study is to describe the concordance of HPV genotypes across anal, oral, and genital samples and to assess factors that predict concordance. We enrolled MSM participants aged between 18 and 26 years attending sexual health clinics or community centers in three US cities. Samples were obtained from oral, genital, and/or anal sites of 1876 study participants for type-specific HPV DNA testing. Clinical record details and data from a patient filled questionnaire were used to assess for risk factors associated with concordant infections. Concordance of type-specific HPV detection across anatomic sites was described with kappa statistics based on percent positive agreement (kappa +). Generalized estimating equations were used to measure the univariate and multivariate associations for the correlates of type-specific concordant HPV detection at oral-anal and anal-genital sample pairs. The median age of the participants was 23 years, 8% were HIV positive and 39.6% received HPV vaccine. The prevalence of any HPV type was highest at the anal site (69%) followed by the penile site (49%) and the oral sites (7.4%). Concurrent detection of any HPV type was most prevalent in anal-genital sample pairs (40%) and was uncommon in pairings involving oral samples. There was little to no genotypic agreement across all sites with poor concordance noted across all sample pairings (kappa + <0.20). Nevertheless, participants who were younger and those who reported an older age of first sex were significantly more likely to have type concordant anal-genital infections while sexual behavior characteristics were not significant correlates of concordance. Lack of oral and anogenital HPV type concordance suggests that direct transmission of HPV infection from anogenital sites to oral sites is rare. Nonetheless, there is some degree of concurrence and concordance observed between anal and genital sites that were not found to be associated with sexual characteristics, suggesting that bidirectional transmission could be occurring between these two sites. However, large longitudinal studies are necessary in order to better demonstrate this mechanism of transmission.