High prevalence of human papillomavirus type 58 in HIV infected men who have sex with men: A preliminary report in Central Italy

INTRODUCTION

Human papillomavirus (HPV) infection is a leading cause of genital malignancies, and there are evidences showing a strong association between HPV infection and risk of HIV acquisition [Rositch et al., 2013; Videla et al., 2013]. Moreover, the incidence of anal cancer increased in men who have sex with men (MSM) compared to heterosexual men and in HIV infected individuals despite highly active antiretroviral treatment [Houlihan et al., 2012; Lawton et al., 2013; Hernandez et al., 2014]. Among these risk population groups, where life expectancy is reduced by non-HIV related morbidities [Ucciferri et al., 2012, 2013], HPV infections are generally caused by high-risk (HR) genotypes [De Vuyst et al., 2009; Ripabelli et al., 2010], particularly HPV-16. Therefore, in order to design and develop successful preventive protocols for HIV infected MSM, it is necessary to assess the prevalence of HR HPV genotypes, including those targeted by available vaccines, in order to estimate the potential impact of the vaccination strategies on types circulation and to monitor their relative frequencies [Sammarco et al., 2013].

To date, there are two licensed HPV vaccines showing high efficacy to prevent HPV infections or precancerous lesions caused by two HR genotypes. Particularly, the bivalent vaccine offers active protection against HPV-16 and 18 associated to vaginal, vulvar, and anal cancers, as well as the quadrivalent vaccine which is also used to prevent low-risk (LR) HPV-6 and 11 genotypes, associated with anogenital warts. Schiller et al. [2012] reviewed the results of large scale clinical trials and reported that both vaccines provide significant protection against HR genotypes similar to HPV-16, such as 39, 45, 59, and 68, and also against HPV-31 in females; in addition, the bivalent vaccine significantly reduce HPV-33 and 52 rates.

Currently, second generation vaccines are under development to extend protection against additional HR HPV genotypes [Schiller and Müller, 2015]. On December 2014, the U.S. Food and Drug Administration (FDA) approved for use a nonavalent vaccine offering immunization against five additional HPV types, HPV-31, 33, 45, 52, and 58. The nonavalent vaccine has proved to offer effective prevention of 90% of cervical, vulvar, vaginal, and anal cancers caused by HPV-16, 18, 31, 33, 45, 52, and 58, as well as genital warts related to HPV-6 and 11 genotypes [Chatterjee, 2014]. Recently (March 26, 2015), the Committee for Medicinal Products for Human Use (CHMP) adopted a positive opinion, recommending the granting of a marketing authorization for nonavalent vaccine in Europe (EMA/CHMP/76588/2015).

Here, we describe the HPV prevalence and genotype-specific distribution at four anatomical sites, including anal canal, oral cavity, urethral mucosa, and coronal sulcus in HIV positive MSM population recruited from a non-metropolitan area of Central Italy.

MATERIALS AND METHODS

In this study, approved by Bioethics Committee at University of Molise, Italy (Protocol number 28536_II/5/2012), fifty Caucasian HIV positive MSM, aged ≥18 years, outpatients attending the Clinic of Infectious Diseases of University of Chieti (Italy) were recruited. All participants were HPV unvaccinated, had no history of acute hepatitis or advanced liver disease and no virological failure in 12 months preceding the study. Clinical data and information on socio-demographic characteristics, drug use and sexual behaviors were obtained for each participant by questionnaire administration. A total of 200 specimens were collected using sterile flocked swabs (UTM-RT viral transport media Flocked Polyester Swabs, Copan Diagnostics Inc., Corona, CA) and analyzed. HPV DNA was isolated by Maxwell® 16 viral total nucleic acid purification kit (Promega, Milan, Italy) in Maxwell® 16 System instrument. A Multiplex-PCR assay, including MY09/MY11 degenerate consensus primers, located within L1 highly conserved polymorphic region, and GH20/PC04 primers pair (Biosense srl, Milan, Italy) targeting β globin gene, was used for viral detection. HPV DNA negative and positive controls were obtained from previous studies [Ripabelli et al., 2010; Sammarco et al., 2013]. Restriction Fragment Length Polymorphism (RFLP) analysis using BamHI, DdeI, HaeIII, HinfI, PstI, RsaI, Sau3AI (Promega) enzymes was used for genotyping, and results were confirmed by sequencing.

RESULTS

All enrolled HIV positive MSM (mean age 38 ± 9 years, interquartile range [IQR] 20–53 years) were on active antiretroviral therapy, 98.7% had undetected HIV viral loads, 91.3% T CD4 ≥350 cells/mm³, 49.3% and 42.0% had Nadir CD4 of 500–200 cells/mm³ and <200 cells/mm³, respectively. A high proportion of patients was current smoker (n = 32, 64.0%) and used drugs (n = 28, 56.0%), particularly hashish/cannabis (91.8%), popper (43.4%), and cocaine (33.8%). Participants stated to have 9 ± 13 partners per year, and 69.1% had a steady partner at time of the study; moreover, 85.7% reported to use condom in sexual intercourses, especially in the anal intercourses, both insertive (88.9%) and receptive (69.8%).

We detected HPV infection in 60.0% (n = 30; 95%CI 46.4–73.5) of HIV positive MSM, with a high prevalence among individuals aged 32–42 years (n = 16, 53.3%), followed by 21–31 years (n = 7, 23.3%) and 43–53 years (n = 7, 23.3%). HPV infection was also common among who smoked (n = 17, 61.8%) and used drugs (n = 19, 63.3%). Among patients, the highest (28/50, 56.0%) HPV detection rate was found in anal specimens compared to the other anatomic sites (χ² test P < 0.01) (Table I). HPV was also frequently detected at one single site (n = 15, 50.0%), and only one patient had detectable viral DNA at all sites (Table I).

Prevalence of vaccine-preventable HPV types was 20.0% (6/30), 63.3% (19/30), and 80.0% (24/30) for the bivalent, quadrivalent, and nonavalent vaccine, respectively. We found 63.3% (n = 19) of HPV positive MSM to carry at least one HR genotype, irrespectively of anatomical sites, and oncogenic types proportion per body sites is reported in Table I.

Among patients harboring at least one HR genotype, HPV-58 was the most prevalent (n = 9, 47.4%), followed by HPV-16 (n = 6, 31.6%) (Table II). We also observed that HPV-58 was isolated from all anatomical sites, mainly in coronal sulcus and oral specimens, while HPV-16 was only found at anal and oral levels; the other HR types were all detected at a single site (Table II).

In our study population, co-infections at the same anatomical site (anal and coronal sulcus, respectively) were found in two patients who harbored types 31/70/84, and 59/72/CP4173. Multiple HR-HPV infections at same anatomical site were not detected; however, among patients with at least one HR type, the 31.6% (n = 6) presented concurrent HR-HPV infections at multiple sites. Particularly, 21.0% (n = 4) of MSM harbored an oncogenic type at least at two different sites, and 10.5% (n = 2) at three sites.

Among HPV positive MSM, LR types 6 and 11 were identified in 36.7% (n = 11) and 20.0% (n = 6), respectively. Furthermore, LR types 81 and 84 were both detected in two individuals, and HPV-42, 61, 70, 72, 145, CP4173 were identified in a single patient, respectively.

Generally, among MSM with infection at multiple sites, we found a lack of concordance of HPV type distribution, except for one patient who had LR HPV-6 infection at both anal canal and coronal sulcus. However, the 40.0% (6/15) harbored the same HPV type at two anatomical sites.

DISCUSSION

To date, few epidemiological studies have assessed the simultaneous presence of HPV infection at different body sites involved in sexual practices in HIV infected MSM. To the best of our knowledge, we describe the first study on HPV detected at four anatomical sites in HIV positive MSM in a small non-metropolitan area of Central Italy. In our study population, we found a high prevalence of HPV infection, in agreement with previous studies indicating that HIV positive MSM are at increased risk of infection compared with HIV negative MSM and men in the general population [Hernandez et al., 2014; Méndez-Martínez et al., 2014]. These findings may be associated to an increased persistence of infection as consequence of compromised immunity and/or to a high incidence of infection as an effect of sexual behaviors [Houlihan et al., 2012]. We also found a higher prevalence of anal vs oral HPV infections, in agreement with previous reports indicating that more than 80% of HIV positive MSM have anal HPV infections [Goldstone et al., 2009; Houlihan et al., 2012; Machalek et al., 2012], probably associated to higher HPV incidence and persistence rates compared to other body sites [Machalek et al., 2012]. Furthermore, anal infections were more associated with HR types compared to the other anatomical sites, suggesting that HR genotypes prevalence differs by the infected epithelium type, and implying that mucosal infection may provide a stronger signal to the immune system.

Our analysis also indicated that infections were more frequent at coronal sulcus level than urethral site, suggesting that it should be preferred for HPV detection. The different site-specific prevalence supports the hypothesis of a significant influence of the anatomical site on the natural history of HPV [Van Rijn et al., 2014], most likely related to differences in the frequency of sexual practices associated to HPV transmission, local mucosa immunity and predisposition to epithelial microtrauma.

We found a high proportion of HIV positive MSM with detectable HR genotypes, in agreement with Goldstone et al. [2009] who have shown that about 40% of MSM carry more than one HR HPV type, increasing also the likelihood of acquiring HIV infection. Surprisingly, we found that HPV-58 was the most frequent HR genotype, followed by HPV-16. Previous observations in HIV positive MSM have shown the highest detection rate of HPV-16 compared to other HR genotypes [Houlihan et al., 2012; Latini et al., 2014; Méndez-Martínez et al., 2014].

Hence, to the best of our knowledge, this is the first report showing an exceeding prevalence, which was not expected, of HR type HPV-58 respect to HPV-16 among HIV-infected MSM. The decreased detection of HPV-16 could be partially considered as the result of vaccination programs whose advantages may be counterbalanced by an increased circulation of non-vaccine oncogenic genotypes.

Since a cross-protection against HPV-31, 33, 45, 52 genotypes included in the nonavalent vaccine was provided by previous vaccines, the highest prevalence of HPV-58 suggests the emergence of this oncogenic genotype, and significant changes in HR type distribution. Interestingly, HPV-58 was detected at all anatomical sites compared to HPV-16 and other HR types, suggesting a potential broader tropism that should be further investigated.

A recent phylogenetic analysis suggested that host-adaptations have created new ecological niches for papillomaviruses and, through this route, HPV types have developed a great diversity of epithelial tropisms. Egawa et al. [2015] hypothesized that papillomaviruses have epithelial sites where they have evolved to complete their productive life-cycle, and that differences of host cells (the site of infection) affect viral pathogenicity. Different papillomaviruses may have become well adapted to epithelial niches and developed preferential tissue distributions within anatomical sites, as well as new opportunities for infection. Hence, both individual status of the host and the infected anatomical sites may have contributed to a different type-specific prevalence.

Genotyping further revealed the lack of genotype concordance between oral and ano-genital specimens collected from HIV-positive MSM, and our finding was similarly reported in previous studies [Parisi et al., 2011; Videla et al., 2013; King et al., 2015]. Indeed, differences in sampling methods may contribute to some disproportion in HPV type detection within the anatomical sites, but it is unlikely to account for the extent of the differences observed. Other variables could be related to sexual practices. Anyway, very limited information is available on the natural history of HPV infection in sites other than cervix, and on viral molecular mechanisms occurring during carcinogenesis of the other anatomic sites [Tommasino, 2014]. Hence, HPV infection and pathogenicity depend on multiple factors, including the genotype, the nature of the cell infected (tropism), and the status of host immunity. A better understanding of site-specific HPV infection could provide new insights into epithelial biology and immunology, and facilitate the development of strategies for disease management and treatment.

In conclusion, our findings suggest that HPV infection should be detected at anus, penis, and mouth levels in HIV infected MSM, and further strengthen the need of public health campaigns promoting positive attitudes towards HPV vaccine and awareness about HPV risk, HPV-associated cancer risks and HPV vaccine efficacy. Although the study was carried out in a limited sample from a non-metropolitan area, these preliminary results underline a potential modification in the circulation of HR genotypes among MSM, which indeed needs further investigation to be valuable for improving prevention strategies. Hence, according to our results, the next-generation nonavalent vaccine which includes the successful HPV-58 genotype should be suitable to prevent a larger number of infections caused by potentially emerging high-risk genotypes.