Beta and gamma human papillomaviruses in anal and genital sites among men: prevalence and determinants

Data regarding the anogenital distribution of and type-specific concordance for cutaneous β- and γ-HPV types in men who have sex with women is limited and geographically narrow. Knowledge of determinants of anogenital detection of cutaneous HPV types in different regions is needed for better understanding of the natural history and transmission dynamics of HPV, and its potential role in the development of anogenital diseases. Genital and anal canal samples obtained from 554 Russian men were screened for 43 β-HPVs and 29 γ-HPVs, using a multiplex PCR combined with Luminex technology. Both β- and γ-HPVs were more prevalent in the anal (22.8% and 14.1%) samples than in the genital (16.8% and 12.3%) samples. Low overall and type-specific concordance for β-HPVs (3.5% and 1.1%) and γ-HPVs (1.3% and 0.6%) were observed between genital and anal samples. HIV-positive men had higher anal β- (crude OR = 12.2, 95% CI: 5.3–28.1) and γ-HPV (crude OR = 7.2, 95% CI: 3.3–15.4) prevalence than HIV-negative men. Due to the lack of genital samples from the HIV-positive men, no comparison was possible for HIV status in genital samples. The lack of type-specific positive concordance between genital and anal sites for cutaneous β- and γ-HPV types in heterosexual men posits the needs for further studies on transmission routes to discriminate between contamination and true HPV infection. HIV-positive status may favor the anal acquisition or modify the natural history of cutaneous HPV types.

cutaneous HPVs in some cases of neoplasia development 25 . However, most studies on genital HPV infection have examined only α-HPV types 20 .
The presence of βand γ-HPVs in the anal canal of men has been established 21,[26][27][28][29] but the determinants of the anogenital presence of HPVs in males is largely based on geographically-concentrated populations of HIV-positive and HIV-negative MSM [26][27][28] . It is not clear how cutaneous HPVs are transmitted into the anal canal, i.e., if it can be due to sexual behavior or passing through the entire digestive system. A possible commensal colonization of the anogenital skin with cutaneous HPVs in HIV-positive MSM was posited 21,25 . Moreover, the risk of β-HPV infection in the anogenital region in men who have sex with women (MSW) may occur through direct skin contact 20,30 . Supportively, the first report on sequencing previously unclassified βand γ-HPVs in the anal canal of men who have sex with women (MSW) from the HIM Study has suggested that other forms of transmission apart from penile-anal intercourse may exist 29 . However, these transmission hypotheses have yet to be fully established.
Studies on genital-anal type-specific positive concordance of cutaneous HPVs involving MSW are nearly absent and represented with the recent report from the HIM Study on the diversity of β-HPV types 21 . To further this etiological knowledge and better understand transmission dynamics of HPV and its role in the development of anogenital diseases, more evidence must be gathered to establish the spectrum of cutaneous HPVs in the anogenital area among diverse population groups worldwide. The aim of the current study is to analyse the prevalence, type-specific positive concordance and determinants for the presence of genital and anal βand γ-HPVs in a large cohort of MSW from the Eastern hemisphere.

Results
The characteristics of the remaining 554 (97.7%) men are presented in Table 1. Of those analysed, 31 (5.6%) men were HIV-positive. Genital C. trachomatis infections was detected in 31 (5.7%) MSW. Up to one third of men (n = 153) reported having Chlamydia infections in the past. Over half of the men (56.3%) had their sexual debut when they were younger than 18 years old. Less than half of the men (47.2%) reported having more than 20 life-time sex partners.
Having a sexual debut at the age of 18 years and above resulted in higher but nonsignificant β-HPV-positivity in both genital (ORs = 1.4, 95% CI: 0.8-2.4) and anal (ORs = 1.3 (95% CI: 0.8-2.2)) samples compared to those starting sexual activities below 18 years. The prevalence of anal β-HPVs increased with increasing number of lifetime sexual partners (p for trend: 0.006 for anal β-HPVs and 0.865 for genital β-HPVs); this association was strongest among the men with more than 20 lifetime sexual partners (OR 2.6, 95% CI: 1.3-5.1). For the γ-HPV types, no linear trend was observed in the genial or anal samples.
Due to the lack of genital samples from the HIV-positive men, no comparison was possible for HIV status in genital samples. Among the anal sites, both cutaneous HPV genera were more commonly detected in   (Table 6). Forty-seven (7.9%) recently diagnosed STIs were observed in the study, including HIV in one (0.2%), N. gonorrhea in one (0.2%), C. trachomatis in 36 (6.1%), herpes simplex virus in 3 (0.5%) and M. genitalium in 7 (1.2%) men, respectively (data not shown). The evaluation of patients with the National Institutes of Health Chronic   Table 5. Adjusted analysis of effect of participant characteristic on cutaneous gamma HPV types in genital and anal samples among men who have sex with women (MSW). Footnote: HPV: human papilloma virus; CI: confidence interval; *Only individuals whose samples were tested for both sites were included in the regression model and HPV positivity is defined as positive on either genital or anal sites; **All variables included in the regression. Prostatitis Symptom Index (NIH-CPSI) 31 observed no association between the clinical symptoms and the prevalence of cutaneous HPVs in the anal and genital area (data not shown).

Discussion
This is the first report of the determinants for anal and genital presence of a broad range of both βand γ-HPV types in a cohort of 554 Russian heterosexual men investigated and compared. Importantly, despite the abundance of both HPV genera observed in the anogenital area in the current study, the type-concordant association between genital and anal sites was uncommon. Although some association was observed between the number of lifetime sex partners and genital and anal prevalence for either genera (in particular for β-HPVs), poor type-specific positive concordance between the two sites and higher but nonsignificant prevalence of both HPV genera in anal sites compared to genital samples provides more evidence to the existence of such transmission route as autoinoculation, which needs to be explored further. Importantly, a substantially higher prevalence of βand γ-HPVs was detected in HIV-positive than HIV-negative men. No significant association between the anal prevalence of the two cutaneous genera with HIV status or sexual behavioral factors were observed in a study among Italian MSM, applying the same diagnostic test 27 . The current study supports the evidence that impairment of the host's immune surveillance may impact βand γ-HPV infections differently 28 . Further studies should be extended with genital and non-anogential sites, to better understand the association between HIV infection status and subsequent HPV acquisition and vice versa.
To our knowledge, the only study that investigated the prevalence of two cutaneous genera in anal and genital specimens obtained among MSW was the HIM study. First, the study reported the high prevalence of βand γ-HPVs in the anogenital skin by sequencing 25 βand 3 γ-HPVs in the anal canal in 164 MSW 29 from the USA, Brazil and Mexico. As the epithelium of male genitals is known to be rich with a broad range of HPVs 20 , the current study employed the proficient 32 Luminex assay, which allowed for the detection of the broadest range of βand γ-HPV types and compared genital and anal prevalence within the same individuals. Applying the Luminex assay, the study from the Western hemisphere assessed diversity of β-HPV types at oral gargles, anal canal and genital sites specimens obtained from 717 men 21 . The current study adds to this new evidence by including 554 MSW, the first such known sample in the Eastern hemisphere.
Another strength of the current study was the use of combined penile and urethral samples in the detection of cutaneous HPVs. Penile and urethral swabs have been found more likely to have the highest prevalence with α-HPV infections 33 . The addition of urine samples to penile swabs in the detection of HPV in men has illustrated how useful the combination of genital samples could be for epidemiological or clearance studies 34 . The combination of the external genital and scrotum swabs was also employed in the detection of β-HPV types in the HIM study 21 .
The obtained data also do not represent the general population. The study limitations also include a low number of MSM observations, as research on MSM populations in Russia remains delicate 35 and the lack of retrospective genital samples obtained from the HIV-positive men, as only anal and serum samples were primary collected 36 during the survey on Chlamydia LGV infection. Larger, preferably prospective, studies to further explore these issues would ideally include both penile and urethral specimens to elucidate the role of immunosuppression on the prevalence of genital βand γ-HPVs among HIV-positive and HIV-negative MSW and MSM. To help in understanding the potential transmission routes of cutaneous HPVs, further behavioral studies should also include more detailed surveys on sexual behaviors.
The routes for the transmission of HPVs into the anal canal in MSW are not clear. Minimizing the risk of potential contamination from the anatomical sites close to the anal verge at the time of anal sampling, the procedure should be performed by a single trained physician in a standard manner, as in the present study. The presence of cutaneous HPV DNA in the anogenital area may reflect deposition of virions released from other body sites with productive infections 20 . In this respect, self-inoculation or transmission from skin, mucosa or secretions from the partners may result in the detection of βand γ-HPV DNAs in the anal canal of men. Identical cutaneous β-HPV types in both penile and anal intraepithelial neoplasia were found in a study on HIV-positive MSM, assuming the dissemination between different anogenital regions within a given patient 25 . In a study among 25 heterosexual couples, the man's hand either βor γ-HPV types were found in the anogenitals of the female  partner in 25% of the visits, while the woman's hand βand γ-HPV types were found in the males' anogenital area in approximately 50% and 25% of the visits, respectively 30 .
In the HIM Study, multiple β-HPV types were more likely to be detected at the genital than at the anal canal 21 . In the current study, next-to-nothing type-specific concordance between genital and anal sites was observed for both HPV genera in HIV-negative MSW. It could also be envisioned that different transmission routes may result in the discrepancy on HPV prevalence between different sites, regardless their sexual practices. In the studies of heterosexual couples, transmission of α-HPVs between hands and genitals or apparent self-inoculation (primarily in men) events 37 resulted in modest concordance rates between genital and non-genital sampled sites 38 . The further studies on extended number of anatomical sites may provide the answer.
Similar to others 39 , the current study could not distinguish new or persistent HPV infection. Detecting HPV transcripts could be seen as a useful options, since the detection of HPV DNA may be the result of transient deposition 39 . The presence of HPV DNA does not necessarily indicate presence of infectious virus 18,[40][41][42] , although the relatively high rate of concordance of βand γ-HPVs between sex partners found in the recent study of 25 couples has suggested that the anogenitals represents a more common area for infectivity than previous thought and that sexual transmission is possible 30 . β-HPV does seem to exhibit activities that can promote oncogenesis, although using distinct from α-HPV mechanisms such as cell transformation promotion and deregulation of pathways linked to the host immune response [43][44][45] . In addition, recently reported associations with the risk of incident head and neck cancer others than α-HPV types, which included γ11and γ12-HPV species and β1-HPV-5 type 46 , advocates the needs for further, preferably prospective, studies on etiological role of cutaneous HPVs in carcinogenesis.
In summary, the current study extends our knowledge regarding βand γ-HPV types and their distribution in diverse male populations, which is essential for developing a better understanding of the natural history, transmission dynamics, and the potential role of different HPV types as co-factors in the development of anogenital malignancies.

Methods
Study population. The study settings and methods have been described in detail elsewhere 22  For both studies, men were eligible for enrolment if they were at least 18 years old and reported no anorectal disorders. A medical history and a standard physical examination were completed. Participants completed a questionnaire concerning sexual behaviour (age at enrolment, age at the time of sexual debut, number of lifetime sex partners, and sexual preferences). Data on age and sexual preferences were only available from the infectious disease hospital participants.
All men self-reported their sexual preferences. If a man reported having had sex (anal or oral) with at least one other man during his lifetime, he was categorized as MSM. If only sex with women was reported, a man was classified as MSW. The questionnaire was administered during a face-to-face interview, before anal sampling. Importantly for the local context, all participants were informed that this information would be unavailable for the third part. In addition, the responses were coded 22 .
After excluding samples from 40 MSM (6.6%) and β-globin-negative samples from 13 (out of 569) individuals (2.3%), in total the genital and anal samples obtained from 554 self-reported as MSW were included in the analysis (Fig. 1).
All participants from the urology units provided blood samples for HIV testing. All previously-tested HIV-positive patients were receiving HAART therapy at the time of enrolment.
All DNA samples and questionnaire data were anonymized, and no study personnel besides the principal investigator had access to participants' identifying information.
Collection of samples and DNA extraction. Samples from genital sites and anal canal were collected.
To minimize the risk of potential contamination from the anatomical sites close to the anal verge at the time of anal sampling, the procedure was performed by a single trained physician in a standard manner. Before sampling, men were instructed to abstain from any form of sex for 3-5 days and from urination for 3-4 hours. Urethral sampling was performed as described elsewhere 47 . The study clinician first sampled the distal urethra (inserted up to about 2 cm inside) followed by the penis (including the coronal sulcus, glans penis and the penile shaft) using two separate brushes. The swabs were rinsed in 1000 µl of phosphate buffer in two separate tubes. The anal canal was sampled with a third brush wetted with phosphate buffered saline, inserted about 2 cm into the anal canal and rotated 360 degree clockwise and anticlockwise directions. The swab was then rinsed in 1000 µl of phosphate buffered saline in a separate tube. The tubes were placed in a refrigerator at 4 °C. The samples were then transferred to a −20 °C freezer and stored until HPV testing.
Using a 100 µl aliquot of the original anal sample, DNA was extracted in the laboratory of the Department of Laboratory Medicine at Karolinska Institutet (Stockholm, Sweden), using a freeze-thaw-boil procedure, as previously described 48  . HPV type species were classified according to de Villiers 3 . Two primers for the amplification of the β-globin gene were included to provide a positive control for the quality of the DNA in the sample 55 .
In the current study, 10 μl of the anal sample and 10 μl of the genital samples were analysed. A 10 μl volume of genital sample was obtained by combining into one sample 5 μl of penile and 5 μl of urethral aliquots. Following multiplex PCR amplification, 10 μl of each reaction mixtures were analysed by multiplex genotyping using the Luminex technology 49,52 . Statistical analysis. In the current analyses we included only individuals, who had both genital and anal samples, whose samples were tested as beta-globin-positive. Paired genital and anal samples were collected in urology participants in parallel. For HIV-positive patients, only anal samples were obtained.
The prevalence (overall, type-and species-specific) of βand γ-HPVs by anatomical sites was presented as proportions. "Type-specific positive concordance" was defined as having the same HPV genotype in both genital and anal samples of a participant, with the Kappa-based extent of agreement between the two anatomical sites measured (Kappa values: <0.0 poor, 0.00-0.20 slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial and 0.81-1.00 almost perfect) 56 . Odds Ratios (OR) and their 95% confidence intervals (CI) were calculated using logistic regression to assess associations between βand γ-HPV positivity and age, age at sexual debut, number of lifetime sexual partners, past and present C. trachomatis infection (all categorical). Except HIV status (when crude analysis was performed only), all variables were included in the adjusted regression model. All analyses were completed using Stata versions 14 (StataCorp, College Station, TX, USA).