An Examination of HPV16 Natural Immunity in Men Who Have Sex with Men (MSM) in the HPV in Men (HIM) Study

Human papillomavirus (HPV) can infect and lead to cancer at several anatomic sites, including the cervix, anus, vulva, penis, and oropharynx (1). Although HPV is a common infection (2), most individuals are able to clear or control their infection (3–5). However, there is no evidence of lifelong immunity by HPV infection, as type-specific infections can often reappear in individuals previously infected with HPV (6).

While a majority of unvaccinated females who acquire an HPV infection develop a measurable type-specific serum antibody response against epitopes on the HPV L1 capsid protein (7, 8), a minority of HPV-infected men develop these responses (9–11). While a few studies have suggested a lack of natural immunity in females, a recent systematic review and meta-analysis found evidence that HPV16 IgG L1 seropositivity induced through natural HPV16 infection provides some protection against subsequent acquisition of genital HPV16 infection in females, but not in males (12). Another study suggested that HPV natural immunity may be restricted to females with the highest antibody levels (13), while no study has directly evaluated whether seropositive men reach these higher antibody levels potentially necessary for protection. Several studies in the infectious disease field have observed a more robust immune response among adult females compared with adult males (14). If “HPV natural immunity” only occurs in females, it would suggest that men are particularly prone to HPV reacquisition later in life.

Few studies have examined natural immunity among men who have sex with men (MSM). MSM may be more likely to acquire HPV natural immunity given that they are more likely to be exposed to HPV at an anatomic site with access to the mucosal immune system (the anus) and, as has been shown (10), are more likely to seroconvert after infection than men who have sex with women (MSW). Additionally, HPV16-seropositive MSM have higher antibody titers than HPV16-seropositive MSW (previously unpublished, Supplementary Fig. S1). However, prior studies have not suggested evidence of HPV natural immunity in MSM, but that may be because they were restricted to a limited number of MSM, often only included HIV-infected MSM, and utilized different methods of HPV16 antibody measurement compared with many of the female studies demonstrating natural immunity (15–18). In addition, the previous natural immunity studies in MSW and MSM have not examined the duration of antibody persistence in men, as a rapid seroreversion rate could also explain a potential lack of natural immunity.

Understanding the natural history and immunity of HPV in MSM is particularly important given the relatively low HPV vaccine uptake in men, MSM's lack of protection from herd immunity through female vaccination, and their high incidence of HPV16-associated anal cancer (19). Therefore, we conducted a study examining whether antibodies generated in response to natural HPV 16 infection provides protection against subsequent genital or anal HPV16 infection, and whether HPV16 seropositivity persisted over time among the MSM in the prospective Human Papillomavirus Infection in Men (HIM) study.

The HIM study is a prospective multinational study of HPV infection in 4,074 men conducted in three locations: Tampa (Florida), Sao Paulo (Brazil), and Cuernavaca (Mexico; refs. 4, 20). Enrollment for the HIM cohort study occurred between June 2005 and September 2009, and participants contributed semiannual follow-up visits for a median of 4.2 years. Eligibility requirements for the HIM study included being a male between the ages 18 and 70 years, no prior diagnosis of penile or anal cancer, and no history of HIV or of HPV vaccination (15). For this analysis, we included all 494 men who ever reported having sex with a man except for 19 men who were anal and genital HPV16 DNA positive at HIM baseline, leaving 475 MSM available for this study. There were 29 MSM who were only genital HPV16 DNA positive at baseline that were excluded from the genital HPV16 analysis, and 127 MSM who were either anal HPV16 DNA positive at baseline or were not tested for anal HPV16 who were excluded from the anal HPV16 analysis.

A previous study examined HPV natural immunity against genital HPV16 among both MSW and 234 MSM from the HIM study (15). This follow-up study further investigates HPV16 natural immunity in MSM in the HIM study by increasing the power to examine the association (includes 475 MSM with longer follow-up), examines natural immunity against anal HPV16, using a standardized L1 VLP ELISA that has been extensively used to characterize antibody responses in women (13, 21), and examines the persistence of HPV16 seropositivity over time among those HPV16 seropositive at baseline.

Anal and genital samples were collected semiannually in the HIM study. Briefly, a clinician utilized three saline-wetted swabs to collect exfoliated cells from the penis (coronal sulcus, glans, and ventral and dorsal areas of the shaft) and scrotum (4). Then, utilizing a separate swab, 360 degrees of the anal epithelium was swabbed between the anal os and the anal canal dentate line (22). HPV DNA testing was performed utilizing a polymerase chain reaction consensus primer system (PGMY 09/11), as described elsewhere (23).

Baseline serum samples from 234 MSM were previously tested with HPV16 L1 VLP-ELISA, a standard measure of immunogenicity measuring polyclonal antibodies, by a laboratory at Johns Hopkins University (JHU; ref. 15). VLP-ELISAs have been widely used in natural immunity studies in females, and are considered a more sensitive measure of cumulative HPV16 exposure than other neutralizing assays such as the Competitive Luminex Immunoassay (cLIA; ref. 24). However, studies utilizing VLP-ELISA often have different procedures, such as virus-like particle production, that could theoretically impact results. This study utilized a validated VLP-ELISA from the National Cancer Institute (NCI)'s HPV Immunology lab, which uses the same laboratory procedures (such as virus-like particle production) as a natural HPV16 immunity study performed in females who demonstrated immunity (13). Baseline serum was tested for all 475 MSM, and subsequently serum from the MSM who were HPV16 seropositive at baseline was tested at the 1-, 2-, 3-, and 4-year follow-up visits.

We also evaluated whether HPV16 antibody titers differed between HIM MSM in the highest tertile of HPV16 seropositivity and a random sample of 42 women ages 18 to 25 from the control arm of the Costa Rica vaccine trial (CVT) who were in the highest tertile of HPV16 seropositivity (≥60 EU/mL). A previous study in the CVT suggested that having a high level of natural antibodies (>60 EU/mL) may protect against HPV infection (13). We reran the HIM and CVT samples together using the same assay and VLP batch, to avoid any laboratory related bias.

Utilizing the JHU HPV16 VLP-ELISA results, we compared the HPV16 optical density of HPV16-seropositive MSM with HPV16-seropositive MSW using box plots, t test, and Mann–Whitney tests. The NCI and JHU HPV16 VLP-ELISA results were compared using percent agreement and kappa statistic. Characteristics between HPV16-seropositive and HPV16-seronegative individuals (by the NCI assay) were compared using χ² and Fisher exact tests for categorical variables and t tests for continuous variables.

Cumulative incidence curves and Cox proportional hazards regression were utilized to compare incident genital HPV16 in HPV16-seropositive versus -seronegative individuals—adjusting for age, country, sexual orientation [MSM or men who have sex with men and women (MSMW)], circumcision status, and number of new (past six months) and lifetime sex partners (any type of sex and anal sex particularly). This analysis was repeated comparing incident anal HPV16 in HPV16-seropositive versus -seronegative individuals. Seropositivity was also described by antibody titer level—continuously and categorically. Antibody titers were categorized a priori into tertiles to match the approach of previous studies (13, 18).

We conducted stratified analyses by whether an individual had a new sexual partner during follow-up as previously described (25). Incident HPV infections detected among those without a new sexual partner may be more likely to represent reactivation of a latent HPV infection and may be less likely to be protected against through natural immunity (23). Considering that sexual behavior measures were self-reported and may be prone to misclassification, we examined the effect of adjusting for incident nononcogenic HPV infection in the models as separate marker of sexual behavior.

We conducted sensitivity analyses restricting to HPV16 infections that persisted for 6 or more months. For this analysis, we excluded the 19 individuals who had an incident HPV16 infection at their last visit or only contributed two visits of follow-up. We also described the time to clearance for HPV16 seropositivity utilizing Kaplan–Meier curves and by utilizing Cox proportional hazard regression. All statistical tests were two sided and were considered significant at the alpha = 0.05 level. Analyses were performed utilizing STATA v13.1 and SAS.

All 475 men who ever reported having sex with a man in the HIM study and who were not positive for anal and genital HPV16 DNA at baseline were included in this analysis. Among the 475 remaining MSM, the mean age was 33.1 years with over half of the participants being white, never smokers, and residing in Brazil (Table 1). Approximately a quarter of participants reported having two or more new sexual partners in the last six months and over 20 lifetime sexual partners. A majority (73%) of the MSM had some report of also having sex with a woman in their lifetime and will be termed MSMW henceforth.

The baseline HPV16 seroprevalence in this population was 27% (n = 128/475). Among the 128 HPV16-seropositive MSM, the median and mean antibody titer were 29.0 EU/mL and 63.3 EU/mL (SD = 83.0), respectively. Baseline HPV16 seropositivity was considerably more common among those with more recent sex partners, more lifetime sex partners, and in MSM compared with MSMW (all P ≤ 0.01). Additionally, HPV16 seropositivity was modestly more common in men who were uncircumcised (30% vs. 20%; P = 0.04) and those from Brazil (31% vs. 20%; P = 0.03). HPV16 seropositivity was not associated with age, race, smoking status, or HIV status (Table 1).

Results from the VLP-ELISAs were compared across the Hopkins and NCI laboratories and between duplicates of the NCI lab. All samples were tested twice by the NCI laboratory with strong agreement (Kappa = 0.95, 95% confidence interval (CI) = 0.92–0.98; Supplementary Fig. S2). The intra-assay CV between duplicates from the NCI laboratory with detectable antibody was 6.2% and the interassay CV between batches was 4.1%. Results among the 234 MSM tested by the JHU and NCI laboratories demonstrated substantial but lower agreement (Kappa = 0.68, 95% CI = 0.60–0.75; Supplementary Fig. S2). The percent agreement between two HPV16 VLP assays was 87.6%, as 95 of the 120 (79%) individuals classified as HPV16 seropositive by the Hopkins lab were classified as HPV16 seropositive by the NCI labs and 95 of the 129 (74%) individuals classified as HPV16 by the NCI lab were classified as HPV16 seropositive by the Hopkins lab. All subsequently described serologic analyses were tested via the NCI lab.

Among the 446 eligible MSM, the incidence of genital HPV16 infection was similar in HPV16-seropositive MSM and HPV16-seronegative MSM (7.18 vs. 5.92 per 100 person-years; HR = 1.24, 95% CI = 0.80–1.93; Table 2 and Fig. 1). After adjustment for age, country, sexual orientation (MSM vs. MSMW), lifetime number of sex partners, number of recent sex partners, and circumcision status, the risk of HPV16 infection remained similar between HPV16-seropositive and -seronegative MSM (aHR = 1.05, 95% CI = 0.66–1.68; Table 2).

Risk of genital HPV16 infection was also similar among HPV16-seropositive men in the highest tertile of HPV16-seropositive with levels above 54.4 EU/mL (aHR = 1.12, 95% CI = 0.56–2.23) compared with HPV16-seronegative MSM; Table 2). Similarly, HPV16 seropositivity was not associated with risk of genital HPV16 infection when restricting to those with new sex partners, when further adjusting for incident nononcogenic HPV infection, or when excluding individuals who were anal HPV16 positive at baseline (each P > 0.20; Table 2). Results were similar by age groups, and there was no significant evidence of HPV16 natural immunity protection against subsequent infection among MSM in their 20s (aHR = 1.09, 95% CI = 0.55–2.20), 30s (aHR = 0.98, 95% CI = 0.43–2.27), or among MSM 40 years or older (aHR = 0.62, 95% CI = 0.17–2.19).

There were only 25 incident genital HPV16 infections that persisted ≥6 months, with 19 detected among HPV16-seronegative MSM (1.85 per 100 person-years) and 6 detected among HPV16-seropositive MSM (1.54 per 100 person-years). No differences in risk of subsequent incident persistent infection was present by sero-status (aHR = 0.73, 95% CI = 0.28–1.89) or by antibody titer (P = 0.36); however, none of the six persistent genital HPV16 infections detected among the seropositive MSM were among those with the highest HPV16 antibody levels (≥54.4 EU/mL).

Among the 348 eligible MSM, the incidence of anal HPV16 was modestly lower than genital HPV16 (4.38 vs. 6.27 per 100 person-years; P < 0.01). There were 22 incident anal HPV16 infections detected. The incidence of anal HPV16 was nonsignificantly higher in MSM who were HPV16 seropositive at baseline (6.45 per 100 person-years) than HPV16 seronegative (6.45 vs. 3.46 per 100 person-years; HR = 1.81, 95% CI = 0.78–4.21; Table 3). Similarly, after adjustment the risk of anal HPV16 infection was nonsignificantly higher in HPV16-seropositive MSM (aHR = 2.34, 95% CI = 0.92–5.98; Table 3). Results remained nonsignificantly elevated when restricting to those with the highest tertile of seropositivity (>54.5 EU/mL), when also adjusting for incident nononcogenic HPV type during follow-up, and when excluding those who had a genital HPV16 infection at baseline (P > 0.20).

One-hundred and eighteen MSM who were HPV16 seropositive at baseline were tested for HPV16 seropositivity at HIM follow-up visits. At the 1-year follow-up visit, 94% (111/118) remained HPV16 seropositive, and at the 4-year follow-up visit, 90% remained HPV16 seropositive. Among the eight MSM who “HPV16 seroreverted” during the study and had a serology test at a visit after the one that defined seroreversion, there were five (62.5%) who tested seropositive at a subsequent annual visit.

We evaluated whether HPV16 antibody titers differed between MSM in highest tertile of HPV16 seropositivity (>54 EU/mL) compared with a random sample of 42 women from the control arm of the Costa Rica vaccine trial (CVT) who were in the highest tertile of HPV16 seropositivity (≥60 EU/mL; ref. 20). We retested the high HPV16 antibody CVT and HIM samples in the same run with the NCI's HPV16 VLP-ELISA, and found that the mean HPV16 antibody titer level between women from the CVT and MSM from the HIM to be similar (CVT: mean = 134.5 EU/mL, HIM: mean = 130.3 EU/mL, P = 0.84). The median HPV16 antibody titer level was 98.9 EU/mL in women from the CVT and 81.1 EU/mL in MSM from the HIM (P = 0.23).

This study, one of the largest evaluations of natural immunity in MSM, observed that HPV16 seropositivity is common among MSM and appears to persist for multiple years. However, there is no evidence that HPV16 natural antibodies protect against subsequent genital or anal HPV16 infection in MSM despite the fact that antibody levels observed were comparable with those seen in women for whom protection from reinfection have previously been documented (13). If confirmed, this may partially explain the high incidence of genital and anal HPV16 infection and related anal cancer seen in MSM throughout the lifespan (19).

The lack of natural HPV16 immunity seen in MSM in this study provides further support to previous findings that found natural immunity did not provide protection among MSW or MSM (15–18). Similarly, a meta-analysis found no evidence that natural immunity provided protection in men (MSW and MSM; pooled RR = 1.22, 95% CI = 0.67–1.77) against genital HPV16, despite evidence for moderate natural immunity in women (pooled RR = 0.65, 95% CI = 0.50–0.80, P heterogeneity = 0.05) against cervical HPV (12). We originally questioned whether the lack of natural immunity in MSW would apply to MSM for several reasons. First, MSM are more likely to acquire HPV natural immunity given that they are more likely to be exposed to HPV at an anatomic site with further access to the mucosal immune system (the anus), and are more likely to seroconvert after infection than MSW (10). Second, it appears, based on the preliminary results of this study, that antibodies often persist for years in HPV16-seropositive MSM and that the antibody levels are higher than those in HPV16-seropositive MSW. Third, previous studies of natural immunity in MSM had a limited number of MSM, often only included HIV-infected MSM, and utilized different methods of HPV16 antibody measurement compared with many of the female studies demonstrating natural immunity.

No protection against subsequent genital or anal HPV16 was observed in the current study despite the fact that the study included a larger number of HIV-uninfected MSM and utilized the same methods of antibody methods as a previous natural immunity study that observed protection in females (13). Although we did observe a fair bit of disagreement between the two labs with different methods of VLP production, suggesting that differences in the serologic assay methods could have an impact on the results. Results also remained similar in a number of sensitivity analyses including no evident protection among individuals with the highest level of antibody titers (≥55 EU/mL) at baseline, a level that has been suggested to be most protective in women (13).

There are a number of potential hypotheses as to why natural immunity against genital or anal HPV16 was not observed in this study. First, it is possible that MSM (and other men) do not obtain natural immunity against HPV infection, while a subset of women may obtain natural immunity. Our initial examination of those with higher antibody titers suggested a relatively similar antibody titer level among HIM MSM and women from the CVT (although there are several cohort differences between the two cohorts such age, sexual behavior, and geographic location that preclude a direct comparison by sex in this study). However, the mechanism of natural immunity is not well understood and antibody measures may only be a marker of other immune functions that may correlate with antibody titers in females but not males. Second, it is possible that the HPV DNA endpoints that were detected at follow-up visits may represent deposition rather true infections, particularly for genital HPV infection, which has a very low likelihood of having clinical significance in men. While results remained similar when restricting those with new sex partners and when adjusting for nononcogenic HPV infection (another measure of sexual activity), it is still possible that results are biased to the null in men due to the higher proportion of HPV deposition. Indeed, while there were a limited number of persistent (6 month+) genital and anal HPV16 infections identified, future studies should consider utilizing this endpoint given its lower likelihood to represent deposition and our nonsignificant point estimate (HR = 0.73), suggesting natural immunity may be evident in larger studies with persistent endpoints.

This study has several strengths. It was one of the largest evaluations of natural immunity in MSM, a group at high risk of HPV16-associated cancer. It also included robust serologic assays to measure natural HPV16 antibodies utilizing processes that have demonstrated HPV16 natural immunity in women (13). However, there are also limitations. While one of the larger studies in MSM, this study is still underpowered to observe endpoints such as persistent genital or anal HPV infection or HPV-associated disease. The primary endpoint used in this study, one time detection of genital and anal HPV16 is prone to deposition (a potential measurement bias). While there were fewer anal HPV16 infections detected during follow-up, the lower 95% CI of 0.92 excludes substantial protection against anal HPV16. In fact, other natural immunity studies in men have also suggested a potential higher risk of anal or oral HPV infection in HPV-seropositive men, potentially due to residual confounding by sexual risk behaviors (17, 26). Additionally, this study only examined natural immunity against HPV16, and it is possible that the level of antibody needed to confer protection may differ by HPV type (27).

Understanding the natural history and immunity of HPV in MSM is particularly important given the relatively low HPV vaccine uptake in men, MSM's lack of protection from herd immunity through female vaccination, and their high incidence of HPV16-associated anal cancer (19). Further research into vaccination and screening for anal cancer screening in middle-aged MSM is warranted.

A.G. Nyitray reports receiving a commercial research grant from Merck & Co. and has received renumeration from EUROGIN. No potential conflicts of interest were disclosed by the other authors.

Conception and design: D.C. Beachler, L.A. Pinto, A. Hildesheim, R. Viscidi, A.R. Kreimer, A.R. Giuliano

Development of methodology: D.C. Beachler, R. Viscidi, A.R. Kreimer

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L.A. Pinto, T.J. Kemp, A.G. Nyitray, R. Viscidi, A.R. Kreimer, A.R. Giuliano

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D.C. Beachler, L.A. Pinto, A. Hildesheim, J. Schussler, A.R. Kreimer, A.R. Giuliano

Writing, review, and/or revision of the manuscript: D.C. Beachler, L.A. Pinto, A.G. Nyitray, A. Hildesheim, R. Viscidi, J. Schussler, A.R. Kreimer, A.R. Giuliano

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D.C. Beachler, L.A. Pinto, T.J. Kemp, J. Schussler

Study supervision: A. Hildesheim, A.R. Kreimer, A.R. Giuliano

We would like to thank the participants of the HIM and CVT studies. Genotyping of the anal canal samples in the HIM Study—PI: A.G. Nyitray: NIAID-NIH (1R21AI101417-01), NCI-NIH (1R03 CA134204-01), Merck & Co., Inc. (IISP 33707). HIM study infrastructure was supported through an NCI grant to ARG (CA R01CA098803). D.C. Beachler was supported by NCI's cancer prevention fellowship. The Costa Rica HPV Vaccine Trial is a long-standing collaboration between investigators in Costa Rica and the NCI. The trial is sponsored and funded by the NCI (contract N01-CP-11005).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.