Introduction
Human papillomavirus (HPV) could be acquired through skin-to-skin or skin-to-mucosa contacts, and sexual transmission is also a common route of transmission [Reference Petca1]. It is one of the most common sexually transmitted infections (STIs) affecting both men and women [Reference Olusola2]. It was estimated that 65%–100% of the sexually active population experienced HPV infection during their lifetime [Reference Bucchi3]. Persistent HPV infections were associated with diseases such as genital warts, anal cancer, cervical cancer, and penile cancer [Reference Della Fera4]. More than 200 types of HPV have been identified, and some of them are divided into high-risk and low-risk HPV types [Reference Burd5]. High-risk HPV types, such as 16, 18, 31, 33, and 45, are responsible for the majority of HPV-related cancers. It has been reported that 99% of cervical cancers, 90% of anal cancers, and 30% of oropharyngeal cancers are attributed to high-risk HPV types of infection worldwide [Reference Song6, Reference de Martel7]. Low-risk HPV types such as HPV 6 and 11 mainly result in genital warts [Reference Sung8].
The prevalence of HPV infection is high in men who have sex with men (MSM) [Reference Meites, Wilkin and Markowitz9], who bear a higher HPV infection burden than women and heterosexual men [Reference Ma10, Reference Marra, Lin and Clifford11]. A meta-analysis conducted in 2021 reported that the prevalence of anal, penile, and oral HPV infections in MSM without HIV was 63.8%, 28.6%, and 14.5% respectively [Reference Farahmand, Monavari and Tavakoli12]. Previous studies reported that the prevalence of HPV-related anal cancer in MSM was 20-fold greater than that of heterosexual men [Reference Patel13, Reference Bouassa14]. MSM are at an increasing risk of HPV acquisition and subsequent HPV-related disease development [Reference Farahmand, Monavari and Tavakoli12]. However, unlike heterosexual men, MSM cannot benefit from herd immunity provided by HPV vaccination in girls [Reference Sauvageau and Dufour-Turbis15]. Moreover, only a few developed countries have included HPV vaccination for MSM in their national programmes [Reference Markowitz16–Reference McGrath18]. To promote HPV vaccination uptake and guide the development of screening strategies for HPV-related diseases in MSM, it’s critical to investigate the prevalence and genotype distribution of HPV in this population.
Despite there being abundant literature on HPV infection burden in MSM, most of these have focused on people living with HIV, and data on MSM from the community are still limited [Reference Donà19]. Furthermore, the HPV prevalence and genotype distribution vary across different countries and regions worldwide [Reference Bruni20]. Accordingly, regional HPV infection burden and genotype distribution should be determined to support the development and implementation of effective HPV preventive strategies. To our knowledge, only a few studies have assessed HPV infections and genotypes in multi-anatomic sites among MSM in the Asia-Pacific region. Studies from Australia [Reference Chow21, Reference Zou22], Mainland China [Reference Ni23], and Taiwan [Reference Lin24] have estimated the HPV infections in oral, anal/rectal, and genital/penile sites, as well as the distribution of HPV genotypes in MSM. Another study conducted in Thailand investigated the prevalence of HPV in anorectal, oropharyngeal, and urethral samples; however, no specific genotypes were identified [Reference Chuerduangphui25]. Additionally, studies conducted in Taiwan [Reference Strong26–Reference Zhou28] and Mainland China [Reference Qian29] examined the prevalence and genotypes of HPV at the anal and penile/genital sites in MSM, but not the oral site. Against this background, this study aims to estimate the prevalence and genotype distribution of HPV infection across multiple anatomical sites in MSM in Hong Kong, a Special Administrative Region (SAR) of China, and an international city in the Asia-Pacific region.
Methods
Study design and participants
This is a longitudinal study with a baseline (September 2021–October 2022) [Reference Wong30] and a one-year follow-up (September 2022–October 2023) (Figure 1). MSM aged 18 or above were recruited from two non-governmental organizations (NGOs) and through online outreach frequented by MSM in Hong Kong [Reference Wong30]. Chinese MSM were included in this study.
Figure 1. Study flow chart. NGOs represent non-governmental organizations.
In the baseline study, following e-consent, participants were asked to complete a self-administered online survey. Self-sampling packages (penile, pharyngeal, and rectal swabs for HPV test, urine, pharyngeal, and rectal swabs for Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (NG) tests, detailed instructions for self-sampling, post-sampling questionnaire, and consent form) were delivered to MSM. The self-collected samples, written consent form, and post-sampling questionnaire in hard copy were returned to the research centre. The self-collected urine specimen, pharyngeal, and rectal swabs were subjected to CT/NG detection by nucleic acid amplification test using the Aptima Combo 2 Assay (Hologic). All specimens were processed according to the manufacturer’s instructions. For HPV detection, DNA was extracted using the QIAamp DNA Mini Kit (Qiagen), and HPV testing was carried out by nested polymerase chain reaction (PCR) using the PGMY09/11 and GP5+/6+ primers targeting the consensus region of the HPV L1 gene [Reference Fuessel Haws31]. Successful nested PCR amplicons were subjected to HPV genotyping by Sanger sequencing using GP6+ primer as the sequencing primer and sequenced using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems). The sequences obtained were analysed and genotyped using the online BLAST search for comparison with HPV sequences in the GenBank database. One year after the baseline study, all participants were invited for another round of self-sampling regardless of their participation in self-sampling at baseline. CT and NG test results were provided to participants regardless of whether the results were positive or negative, and clinical referrals were provided to those who tested positive for CT and/or NG. HPV test results were only provided upon request, and this arrangement has been included in the study promotion materials. Ethical approval of this study was obtained from the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee (approval number: CREC2020.436).
Data analysis
The main outcome variable, HPV positivity (prevalence) at any site at each timepoint, was defined as the number of participants who tested HPV positive at any site (rectal, pharyngeal, and/or penile) divided by the total number of participants who underwent self-sampling at each timepoint. The prevalence was calculated at the individual level. For example, if one participant tested positive for HPV at multiple anatomic sites, we counted this as one in the prevalence calculation. Among participants who tested HPV negative at baseline, the incidence rate was calculated by the number of participants who newly tested HPV positive at follow-up divided by the total person-years of follow-up. In addition, the uptake of self-sampling for HPV testing was considered as a secondary outcome variable in this study.
Factors (socio-demographics, histories (to-date) of Hepatitis B Virus (HBV) and HPV vaccination, histories (to-date) of STIs/HIV testing and diagnosis, and HIV pre-exposure prophylaxis (PrEP) use, as well as sexual behaviours in the past 6 months, and CT/NG positivity at each timepoint) associated with the outcome variables were examined using bivariable logistic regression. Variables that showed a statistical difference (P < 0.05) in the bivariable logistic regression were further analysed using multivariable binary backward stepwise logistic regression. Adjusted odds ratio (aOR) with 95% confidence interval (95%CI) was calculated, and all statistical analyses were performed in SPSS 28.0. Complete case analysis was performed in this study.
Results
Characteristics of participants
Totally 701 Chinese MSM completed the baseline survey, of which 25% (175/701) had attended the follow-up. The median age was 30 years old (interquartile range [IQR], 26–35) in the baseline survey, and 31 years old (IQR, 28–37) in the follow-up survey. In the baseline, 83% were in employment, 85% had attained post-secondary education or above, and 47% had a monthly income of HKD 15,000–30,000 (USD 1 ~ HKD 7.8) (Supplementary Table S1).
HPV positivity in MSM who had completed self-sampling for HPV testing
Totally 380 and 175 MSM have requested the self-sampling kits at baseline and follow-up, respectively. Overall, 270 and 127 returned the completed self-collected samples for laboratory testing at the two respective timepoints. Throughout the whole study period, 41% (288/701) of MSM completed self-sampled HPV testing (Figure 1). Regarding the factors associated with the uptake of self-sampling for HPV testing (Supplementary Table S2), age (years old) was positively related to the uptake of self-sampling for HPV testing (aOR = 1.03, 95% CI = 1.01–1.06). MSM who had attained post-secondary education or above (aOR = 2.45, 95% CI = 1.36–4.41), histories of HIV testing (aOR = 2.00, 95% CI = 1.09–3.62), STIs testing (aOR = 1.62, 95% CI = 1.04–2.52), and HIV PrEP use (aOR = 2.46, 95% CI = 1.60–3.78) were more likely to complete self-sampling for HPV testing. Compared with MSM without a regular sex partner in the past six months, MSM who had 1–4 (aOR = 1.60, 95% CI = 1.04–2.49) and at least 5 (aOR = 2.49, 95% CI = 1.07–5.81) regular sex partners were more likely to complete self-sampling for HPV testing.
Among MSM who returned the completed self-collected samples, 29% (95%CI: 23%–36%) (78/270) and 33% (95%CI: 24%–45%) (42/127) tested HPV positive at any anatomic site at baseline and follow-up timepoints, respectively. By anatomic site at baseline, 26% (71/270) tested positive at rectal sites, 2% (6/270) tested positive at penile sites, and 1% (4/270) tested positive at pharyngeal sites. By anatomic site at follow-up, 30% (38/127) tested positive at rectal sites, and 4% (5/127) tested positive at penile sites (Table 1). At baseline, 2 MSM had rectal–penile concurrence for HPV, 1 MSM had rectal–pharyngeal concurrence, but 0 MSM had penile–pharyngeal concurrence. At follow-up, 1 MSM had rectal–penile concurrence, while 0 MSM had rectal–pharyngeal nor penile–pharyngeal concurrence. Among 71 MSM who tested positive at rectal sites at baseline, 24% (17/71) remained positive at rectal sites at follow-up. Of the 6 MSM who tested positive at penile sites at baseline, only 17% (1/6) remains positive at penile sites at follow-up. Among 4 MSM who tested positive at pharyngeal sites, none tested positive at the same sites at follow-up.
Table 1. HPV positivity in MSM by different anatomic sites
+ve positive.
Change of HPV positivity at baseline and follow-up
Totally 109 MSM returned their samples at both baseline and follow-up, with 32 who tested HPV positive at baseline, and 38 who tested HPV positive at follow-up. At baseline, 31 samples were successfully typed, with detection of HPV 6 (10%), 11 (23%), 16 (13%), 18 (6%), 33 (3%), 44(3%), 53(6%), 58(6%), 59(3%), 66(6%), 68(3%), 74(3%), 83(3%), 84(6%), and 89(3%). Among the 77 MSM who tested HPV negative at baseline, 18 tested positive at follow-up, with an incidence rate of 21.2/100 (95%CI: 12.6–33.5/100) person-years at any anatomic site. Among the 32 MSM who tested HPV positive at baseline, 12 tested negative while 20 remained positive at follow-up.
By anatomic site, 77 rectal, 76 pharyngeal, and 75 penile swabs were tested HPV negative at baseline. Of these, 16 rectal, 2 penile, and 0 pharyngeal samples were tested positive at follow-up. The incidence rate of HPV was 18.9/100 (95%CI:10.8–30.6/100) person-years at rectal sites, 2.4/100 (95%CI:0.3–8.6/100) person-years at penile sites, and 0 /100 person-years at pharyngeal sites.
Genotype distribution of HPV
A total of 109 HPV-positive samples (three samples for each person were collected and tested) from 100 MSM were successfully genotyped at two timepoints, with 27 HPV genotypes detected. Among these samples, HPV 6 was the most common genotype (17%), followed by HPV 11(16%) and HPV 16 (11%). Sixty percent of the identified HPV subtypes were vaccine-preventable (9-valent). Low-risk HPV types ranked first (39%), followed by high-risk HPV types (36%), probably high-risk HPV types (14%), and unclassified risk (11%). Totally 33 MSM were co-infected with HPV and CT/NG at two timepoints. Of these, 35 samples were tested for HPV subtypes at baseline and/or at follow-up, with 46% being low-risk HPV types, 31% being high-risk, 17% being probable high-risk, and 6% being unclassified.
Characteristics of MSM who tested HPV positive at any anatomic site at baseline or follow-up timepoints
Among 78 MSM who tested HPV positive at baseline, 100% were single/never married, 86% were in employment, and 88% had attained post-secondary education or above. Some 28% self-reported a history of HPV vaccination, all of them received 9-valent vaccine, of whom 60% of the detected HPV subtypes were covered by 9-valent HPV vaccine. A majority had ever tested for HIV (88%) and STIs (69%). Additionally, 3% and 47% self-reported living with HIV and had ever been diagnosed with STIs, respectively (Table 2). Among 42 MSM who tested HPV positive at follow-up, 100% were single/never married, 88% were in employment, and 98% had attained post-secondary education or above (Table 2). At both baseline and follow-up timepoints, no sociodemographic factors were found to be associated with HPV positivity (P > 0.05).
Table 2. Characteristics of MSM by self-sampled HPV testing results at any anatomic site at baseline and follow-up timepoints
USD 1 ~ HKD 7.8, IQR interquartile range, STIs sexually transmitted infections, PrEP pre-exposure prophylaxis, OR odds ratio, CI confidence interval.
Sexual behaviours and STIs of MSM who tested HPV positive at any anatomic site at baseline or follow-up timepoints
At baseline, 97% of MSM who tested HPV positive were sexually active in the past 6 months. Among them, 79% had engaged in group sex and 16% had engaged in chemsex. Some 43% had at least five non-regular sex partners, 59% had 1–4 regular sex partners, and 7% had at least one commercial sex partner. About half (41%) had never used condoms with regular sex partners, and 50% had never used condoms with commercial sex partners. About one-third (31%) of MSM used condoms on less than half the occasions or had never used condoms with casual sex partners. Among MSM who tested HPV positive at baseline, 19% and 6% were concurrently tested positive for CT and/or NG, respectively. History of group sex (OR = 1.91, 95%CI = 1.05–3.50) in the past 6 months was positively associated with HPV positivity at any site. Compared with MSM who always used condoms with regular sex partners in the past 6 months, those who used condoms on less than half (OR = 2.74, 95% CI = 1.11–6.75) and never used condoms (OR = 2.91, 95% CI = 1.26–6.73) were more likely to have samples testing HPV positive at any site. Compared with MSM who always used condoms in the past 6 months, those who used condoms on less than half (OR = 3.23, 95% CI = 1.39–7.49) of the occasions with casual sex partners and never used condoms (OR = 8.75, 95% CI = 1.21–63.43) with commercial sex partners were more likely to have samples testing HPV positive at any site (Table 3). In multivariable logistic regression, inconsistent condom use with casual sex partners was the single variable in the final model.
Table 3. Sexual behaviours and STIs of MSM by self-sampled HPV testing results at any anatomic site at baseline and follow-up timepoints
STIs sexually transmitted infections, CT Chlamydia trachomatis, NG Neisseria gonorrhoeae, OR odds ratio, CI confidence interval.
*P < 0.05.
At follow-up, 95% of MSM who tested HPV positive were sexually active in the past 6 months. Of those, 60% had engaged in group sex and 15% had engaged in chemsex. Among MSM who tested HPV positive at follow-up, 17% and 15% were concurrently tested positive for CT and/or NG, respectively. History of group sex (OR = 4.09, 95%CI = 1.84–9.10) in the past 6 months was positively associated with HPV positivity at any site (Table 3).
Discussion
Compared with other studies in the Asia-Pacific region that utilized self-sampled HPV testing, this study demonstrated a slightly lower prevalence of HPV infection. The HPV prevalence rate in this study was 29%–33%%, with 26%–30%% at rectal sites, 2%–4% at penile sites, and 0%–1% at pharyngeal sites. One study conducted in Mainland China reported a rectal HPV positivity rate of 38% in MSM attending university [Reference Fan32]. Similarly, another study conducted in Southern China estimated that 38% of MSM tested HPV positive from self-collected pharyngeal, penile, and rectal specimens [Reference Ni23]. To our knowledge, no other studies in the Asia-Pacific region had used self-sampled HPV testing, except for the two above-mentioned studies conducted in China. In comparison to other studies in the Asia-Pacific region that did not utilize self-sampled testing, this study revealed a lower prevalence rate. A study in Northern China reported HPV infection rate of 53.7% at any anatomic site in MSM without HIV, with 42% at anal sites, 24.7% at genital sites, and 3.2% at oral sites [Reference Bai33]. A study in Taiwan reported that 44% of MSM had at least one HPV type detected in nurses-collected samples, with 39% and 16.5% at anal and penile sites, respectively [Reference Strong26]. In Thailand, the prevalence of anorectal and oropharyngeal HPV infection in MSM was 43.9% and 13.9%, respectively [Reference Chuerduangphui25]. A study from Australia reported that prevalent anal HPV was detected in over 50% of MSM [Reference Poynten34]. The lower HPV prevalence observed in self-sampled testing may be due to differences in participant characteristics. As this study reported that MSM with higher educational attainment, older age, previous HIV/STIs testing and HIV PrEP usage histories, and recent engagement in multiple regular sexual partnerships were more likely to participate in self-sampled testing. When comparing with studies in Western countries that also utilized HPV self-sampled testing, the HPV prevalence in this study was much lower than that in Canada (62.3%) [Reference Gilbert35] and the United States (70.9%) [Reference Singh36]. This discrepancy may be attributed to the disproportionate distribution of HPV types between North America and Asia [Reference Sung8, Reference Giuliano37]. Furthermore, we provided detailed, step-by-step instructions for HPV self-sampling to participants, making it feasible for them to collect samples independently. Consistent with the previous study [Reference Shah38], this study detected rectal–penile and rectal–pharyngeal concurrence for HPV, while no concurrence was observed between penile–pharyngeal sites. This pattern may be explained by the low prevalence of HPV at pharyngeal sites. The pharyngeal HPV infections may be acquired independently of penile infections, along with different durations of HPV infection at different anatomic sites [Reference King39, Reference Darwich40]. Previous studies [Reference Sikora, Morris and Sturgis41–Reference Ferenczy, Fau and Richart43] have also suggested that the mechanisms of HPV infection transmission between different anatomic sites remain unclear, HPV infection at one site could serve as a source for infection at other sites, and transmission may occur through hand carriage and objects.
Few studies have reported on the incidence rate of multi-anatomic HPV infection in MSM. In this study, we estimated a much lower incidence rate (21.2/100 person-years at any site, 18.9/100 person-years at rectal sites, and 2.4/100 person-years at penile sites) compared to previous studies conducted in Mainland China [Reference Zhou44], Taiwan [Reference Zhou28], and the United States [Reference Glick45]. Previous studies conducted in Mainland China (2017) [Reference Zhou44] and the United States (2014) [Reference Glick45] reported an incidence rate of anal HPV infection in MSM of 31.3/1000 person-months (37.56/100 person-years) and 38.5/1000 person-months (46.2/100 person-years), respectively. One study conducted in Taiwan from 2015 to 2019 reported an incidence rate of HPV infection in MSM of 43.6/1000 person-months (52.32/100 person-years) at anal sites and 26.8/1000 person-months (32.16/100 person-years) at penile sites [Reference Zhou28]. Another study conducted in Australia from 2010 to 2012 estimated the definite incidence rate of HPV infection in MSM as 57% and 12% per 100 person-years at anal and penile sites, respectively [Reference Zou22]. The variations in incidence rates between the above studies may be attributed to the differences in testing methods, study follow-up duration, local HPV epidemiology, and vaccination coverage.
In contrast to a previous global meta-analysis that identified HPV 16 as the most prevalent HPV genotype in the anorectal region of MSM [Reference Farahmand, Monavari and Tavakoli12], our study findings suggested that HPV 6 was the most common HPV genotype. Nevertheless, this study showed similar results to meta-analyses conducted in China; one meta-analysis in 2021 reported that the HPV genotypes in Chinese MSM were predominately HPV 6, 11, and 16, and with a higher prevalence of low-risk HPV in Southern China [Reference Zhou46]. Another meta-analysis in 2025 also reported that the HPV 6 and 11 were the most prevalent HPV genotypes in Chinese MSM [Reference Li47]. Importantly, our study findings revealed that more than half of the HPV genotypes identified in MSM could be prevented by HPV vaccines. However, it is noteworthy that only 28%–29% of the MSM who tested HPV positive had self-reported a history of HPV vaccination, and 40%–68% of the detected HPV subtypes among these MSM were not covered by 9-valent HPV vaccine. Since we did not investigate the time of self-reported vaccination, some MSM may have received the vaccine after acquiring HPV.
Consistent with a previous study [Reference Müller Etienne48], our study findings suggested that MSM who recently engaged in group sex have a higher risk of HPV infection. Since condom use plays an important role in preventing HPV infection [Reference Quinlan49], it is unsurprising that a lower frequency of condom use was associated with an increased likelihood of HPV infection. Younger Chinese MSM were found to have a higher risk of acquiring HPV infection [Reference Zhou44], but no association between age and HPV infection was observed in this study. Further investigation is needed to identify the relationship between age and HPV infection in MSM. Inconsistent with previous studies, some risk factors, including a history of STIs [Reference Donà50], multiple sexual partners [Reference Nyitray51], and chemsex engagement [Reference Zhang52], were not significantly associated with HPV infection of MSM in this study. This inconsistency may be due to the differences in the clinical characteristics of MSM and the definition of HPV infection. One study regarded ≥12-month persistent HPV infection as the outcome indicator [Reference Nyitray51], and another study defined HPV infection as the detection of one or more HPV types in participants who were HPV-DNA negative at their previous visit [Reference Donà50].
There were several limitations in this study. First, selection bias may exist in this study. This survey was limited to online users, and self-sampling for STI testing was not compulsory, resulting in a relatively small sample size and a high rate of loss to follow-up, which decreases the representativeness of the sample for estimating prevalence. Additionally, some of the participants were from another research project regarding HIV PrEP use, but the influence of HIV PrEP use has been controlled by the multivariate regression model in this study. Second, the inconvenience of returning self-collected samples during the COVID-19 epidemic may have reduced the participation rate of self-sampling. Third, the incidence of HPV infection may be underestimated due to the potential occurrence of infection shortly followed by seroclearance between the baseline and follow-up sampling timepoints. Fourthly, high-risk sexual behaviours occurring between survey completion and sample collection (normally in 1 month) may not be captured, and new infections occurring in this period might not be reflected in the collected survey data. Lastly, in the survey, we only asked about the type of HPV vaccination received, not the number of doses taken and the time taken. We cannot confirm if all participants have completed the three-dose HPV vaccination regimen, but that they have received at least one dose.
Conclusions
The HPV prevalence in Chinese MSM in Hong Kong was found to be similar to other Chinese studies that also utilized self-sampled HPV testing methods. However, it was notably lower compared to similar studies conducted in Western countries. Among Chinese MSM in Hong Kong, low-risk HPV infections, specifically HPV 6 and HPV 11, were the most prevalent genotypes. HPV vaccination plays a crucial role in preventing HPV infection in MSM, as the majority of identified HPV subtypes are vaccine-preventable.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S0950268825100472.
Data availability statement
The data presented in this study are available from the corresponding author upon request and approval from the ethics committee. The data are not publicly available due to privacy or ethical restrictions.
Acknowledgements
We thank Mr. CY Poon, Ms. SC Mak, and Ms. KY So for their assistance in the laboratory work and Ms. Priscilla Wong, Ms. Mandy Li, and Mr. Alexander Chiu for project coordination and assistance.
Author contributions
Supervision: S.S.L.; Conceptualization: N.S.W., R.Z.; Data curation: D.P.C., S.L.C.; Methodology: C.K.K., T.H.K.
Funding statement
This study is supported by the Health and Medical Research Fund Commissioned Research on Control of Infectious Diseases (Phase IV), Hong Kong Special Administrative Region Government (Reference number: CID-CUHK-E).
Competing interests
The authors declare none.
Ethical standard
E-consent had been obtained before the online baseline survey started. Ethical approval from the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee was obtained (approval number: CREC2020.436).
Open access
