Anal human papillomavirus and its associations with abnormal anal cytology among men who have sex with men

Human papillomavirus (HPV) infection contributes to most anal cancers and premalignant intraepithelial lesions. This study investigated anal HPV infections and cytological abnormalities among men who have sex with men (MSM). Sociodemographic characteristics and sexual behaviors were collected by using a structured questionnaire. Anal cytological results were examined, and HPV genotyping was performed by the Linear Array HPV Genotyping test. Logistic regression was used to estimate risk factors and their associations with high-risk HPV infection and cytological abnormalities. Among 163 MSM, 101 were seropositive for human immunodeficiency virus (HIV) and 62 were seronegative for HIV. The overall prevalence of HPV was 66.2%. A total of 61.9% and 48.2% of participants had never acquired any of either the quadrivalent or nonavalent vaccine HPV types, respectively. Cytological findings showed 15.3% atypical squamous cells of undetermined significance, 16.6% low-grade squamous intraepithelial lesion, 4.9% atypical squamous cells that cannot exclude high-grade squamous intraepithelial lesion and 17% high-grade squamous intraepithelial lesion. The number of high-risk HPV types was the predominant risk factor for abnormal anal cytology (OR 2.02, 95% CI 1.27–3.24). Infection with high-risk HPV was a significant predictor for cytological abnormality. MSM should be encouraged to obtain the HPV vaccine.

Anal HPV infection is mainly transmitted through anal intercourse, which is usually considered as a major high-risk sexual behavior in men who have sex with men (MSM) 18 . A previous meta-analysis found that the prevalence of anal HPV infection and epithelial dysplasia among MSM greatly exceeded the prevalence in the cervix among women 3 . Moreover, among the HIV infected population, the prevalence of any anal HPV type was 96% among MSM, 90% among women and 59% among men who have sex with women, suggesting that the acquisition of anal HPV infection is dependent on sexual behaviors 19 . However, most of the previous studies were mainly confined to HIV seropositive MSM. Data comparing HPV prevalence among both HIV seropositive and seronegative MSM is limited.
Therefore, this study aimed to evaluate the associations of HPV infection with the risk of abnormal anal dysplasia among MSM. Moreover, the different types or numbers of HPV infections were estimated by stratifying by HIV seronegatives and seropositives.
As shown in Table 3, HIV seropositivity and number of detected high-risk HPV showed significant positive associations with anal intraepithelial neoplasia (p < 0.001). On the other hand, the absence of receptive anal intercourse within one year showed negative associations with abnormal anal cytology (p = 0.034). In the multivariate model, only the number of detected high-risk HPV types was significantly associated with abnormal anal cytology, with an adjusted OR of 2.02 (1.27-3.24).
The prevalence of vaccine-preventable HPV types among the 139 MSM is shown in Table 5. There were 86 (61.9%) and 67 (48.2%) participants that had never acquired any of the HPV types included in the quadrivalent or nonavalent HPV vaccine. Among our study participants, none had acquired all four quadrivalent vaccine HPV types or more than four nonavalent vaccine HPV types. There were significantly fewer detected quadrivalent vaccine types in HIV seronegatives than in seropositives (15.5% versus 54.3%, p < 0.001). The same pattern was seen for nonavalent vaccine HPV types (25.9% versus 70.4%, p < 0.001).

Discussion
Anal HPV infection is primarily acquired through sexual exposure, especially via anal intercourse. HPV infection contributes substantially to the risk for anal cancer and squamous intraepithelial lesions. However, there is still limited epidemiological studies investigating the association between anal HPV infection and anal intraepithelial lesions among MSM regardless of their HIV infection status, particularly among the Asian population. The study was conducted in a large medical center which provides substantial healthcare for HIV-seropositives as well as pre-or post-exposure prophylaxis for HIV-seronegatives. Because MSM are considered as a high risk group for HIV infection 20,21 , the participants enrolled in this study should be representative of most of the MSM in northern Taiwan with considerable generalizability. This study showed that both high-and low-risk anal HPV infections were significantly more prevalent in HIV seropositives than in HIV seronegatives. Individuals with increased numbers of acquired HPV types had increased odds of abnormal anal cytology. However, not many individuals (15.5% of HIV seronegatives and 54.3% of HIV seropositives) were infected with at one or more of www.nature.com/scientificreports www.nature.com/scientificreports/ the quadrivalent vaccine types (HPV6, 11,16,18), suggesting that the HPV vaccine should be advised in these individuals.
We found that the prevalence of anal HPV infection was 85.2% in HIV seropositives and 39.7% in HIV seronegatives. This HPV prevalence is comparable with other studies conducted in Asian countries 17,22,23 . Among MSM seropositive for HIV, the prevalence of HPV infection was 82.7% in China, 85% in Thailand and 85.3% in Taiwan 17,22,23 . On the other hand, the prevalence was even higher in Western countries, showing 97% in the United States and 96.3% in Italy 24,25 . Compared to MSM that were seropositive for HIV, those who were HIV seronegative had consistently lower prevalence of anal HPV infection, which was reported to be 58.5-73.3% and 70% in Asian and Western countries, respectively 17,[22][23][24][25] . Another recent study among MSM of various ethnicities suggested that the estimated global prevalence of anal HPV infection among HIV seropositives and seronegatives was 79% and 47%, respectively 26 .
In terms of HPV types, the most common types of high-risk genital HPV among HIV seropositive MSM in Taiwan were HPV16 (5.9%), HPV51 (7.9%) and HPV52 (7.2%) 20 . In Western countries, the most common types was HPV16, with a prevalence of 29.0% 27 . However, the prevalence of HPV52 was around 10% 7,27 . Our study found the prevalence of HPV16 and HPV52 to be 23.5% and 17.3%, suggesting that anal HPV52 infection among  MSM is more prominent and increasing in Taiwan. A majority of anal cancer was attributed to HPV infections 6 ; nevertheless, the impact of different high-risk HPV types in precancerous lesions on anal cancer has not been clearly identified. The importance of anal HPV52 infection in Taiwan might be worth further investigation. MSM and increased numbers of acquired high-risk HPV types were found to significantly increase the risk for anal atypical squamous lesions in recent reports 19,21,28,29 . On the other hand, a meta-analysis suggested that HPV16 infection, but not the number of acquired high-risk HPV types was associated with increased risk for anal cancer 7 . This may be due to the fact that the study included different populations, such as HIV seropositive male and female patients regardless of their sexual orientation 7 .
In our study, we found that the prevalence of high-risk HPV infection, including HPV16, HPV31, HPV51, HPV56, HPV58 and HPV68, increased with the severity of abnormal anal cytological findings. The prevalence of high-risk HPV was 35.2% for normal anal cytology, 68.9% for LSIL and ASCUS and 78.3% for ASCH and HISL. These findings are consistent with previous reports 7,27 . Interestingly, among those with high-risk HPV infection, the numbers of detected high-risk HPV types also increased with the severity of abnormal anal cytology. The median numbers of detected high-risk HPV infections was 0 (IQR, 0-1) for individuals with normal anal cytology, 1 (IQR, 0-2.5) for ASCUS, 1.5 (IQR, 1-3) for LSIL, and 2 (IQR, 1-3) for ASCH and HSIL. Notably, HIV serostatus was also significantly positively associated with HPV infection. HIV seropositives who are immunosuppressed may be more prone to persistent HPV infection, therefore increasing their risk for anal cancer 17,20 . However, given the multitude of HPV types, the impact of multiple HPV infections on anal dysplasia needs further investigation in large-scale studies.
Most anal cancers or precancerous HSIL among MSM are attributed to HPV infection, especially infection via anal intercourse 6,18,28,30,31 . Strategies for primary prevention of HPV infection are crucial for the control of anal dysplasia or anal cancer among MSM. Our study found fewer detected quadrivalent or nonavalent vaccine HPV types among HIV seronegatives than among HIV seropositives. Moreover, 61.9% of our study population did not yet acquire any of the HPV types covered in quadrivalent vaccines (HPV6/11/16/18). Thus, this study suggests that there is still a substantial proportion of MSM that may benefit from HPV vaccination. Therefore, increasing   www.nature.com/scientificreports www.nature.com/scientificreports/ awareness of HPV vaccines may be a practical strategy for prevention among MSM, as HPV vaccines have also been suggested for MSM in previous studies 21,32 .
There are limitations to this study that should be considered. Due to its cross-sectional design, it may be difficult to infer causality, since we could not establish causal temporality. Although we found that HIV seropositivity was associated with acquisition of both low-risk and high-risk HPV types, it was not clear whether HIV infection preceded the HPV infection or not. It is possible that individuals with HIV infection were unable to clear HPV because of impaired immunity. On the other hand, there is also the possibility that individuals with HPV infection may be more sexually active, therefore making them more vulnerable to HIV infections. Lastly, due to the heterogeneity in HPV types, larger sample sizes are needed to further elucidate the influence of specific HPV types or number of anal HPV infections on abnormal anal cytological lesions.
In conclusion, we found a high prevalence of anal dysplasia and HPV infection among MSM regardless of their HIV serostatus. Individuals with increased numbers of HPV infections had stronger associations with anal dysplasia. In addition, there were still a substantial number of MSM who had not yet acquired any of the HPV types that are preventable by vaccines. Therefore, MSM should be advised to receive HPV vaccinations.

Methods
Study design and subjects. This cross-sectional study was conducted in a tertiary medical center, Taipei Veterans General Hospital in northern Taiwan. Males at least 20 years of age who reported having regular sex with men in the past 6 months were enrolled from March, 2015 to July, 2016. The MSM who were HIV seropositive were recruited from outpatient clinics during their regular follow-ups. On the other hand, the MSM who were seronegative for HIV were recruited from voluntary HIV testing clinics. These men were confirmed as HIVnegative by Western blot using the recomline HIV-1 & HIV-2 IgG (Mikrogen GmbH, Neuried, Germany) or HIV viral load examination using cobas HIV-1 (Roche Diagnostics, Mannheim, Germany). Participants with a history of anal cancer or abnormal anal cytology were excluded. All methods in this study were carried out in accordance with the principles of the Declaration of Helsinki. This study was approved by the Institution Review Board of Taipei Veterans General Hospital. All of the study subjects provided informed consent. Data collection. Study subjects were interviewed at study entry using a structured questionnaire (Supplementary questionnaire). The information collected included demographic characteristics (age, educational levels and income), smoking status, partnership (single or fixed partner), sexual behavior within 1 year, HPV vaccination status and history of sexually transmitted diseases. Clinical data on HIV infection at study entry was obtained by medical chart review with a standardized form. Data on HIV viral load, CD4 + cell count and nadir CD4 + cell count was abstracted.
Anal specimen collection for cytology and HpV genotyping. At the enrollment visit, well-trained health-care professionals sampled anal specimens from all participants using LIBO Specimen Collection and Transport Swab (LIBO Medical Products Inc., New Taipei City, Taiwan) with 5 cm insertion into the anal canal and a slow 360-degree rotating extraction. The swab was then shaken in SurePath liquid-based cytology fluid (Becton, Dickinson and Company, NJ, USA). The anal cytology slides were screened and labeled by cytotechnologists first then interpreted by two cytopathologists separately blinded for HIV or HPV infection status. Slides with discordant interpretations were reviewed again under a multi-head microscope for agreements. The cytological results were interpreted according to the Bethesda System terminology as negative, ASCUS, LSIL, ASCH and HSIL 33 . The remaining samples were sent for HPV genotyping using the Linear Array HPV Genotyping test (Roche Molecular Systems, Inc., Pleasanton, CA), per the manufacturer's instructions. Briefly, the target DNA was