Edelstein, Zoe R. PhD*; Schwartz, Stephen M. PhD†‡; Hawes, Stephen PhD‡; Hughes, James P. PhD§; Feng, Qinghua PhD¶; Stern, Michael E. MN, ARNP∥; O’Reilly, Sandra BS‡; Lee, Shu-Kuang MS§; Fu Xi, Long MD, PhD‡¶; Koutsky, Laura A. PhD‡
Human papillomavirus (HPV) is an etiologic agent for oropharyngeal cancers, with the strongest associations observed for cancer of the tonsils.1 Among HPV-positive oropharyngeal cancers, up to 90% are associated with HPV-16.2 Although head and neck cancers are rare in the United States, there has been a significant increase in HPV-related oropharyngeal cancer in the last 30 years.3 This increase has been especially notable in the last 10 years and has occurred predominately in males, among whom oropharyngeal cancer is more common.4 If current trends continue, the annual number of new HPV-positive oropharyngeal cancer cases in the United States could surpass that of cervical cancer by 2020.5
Prevalence estimates for HPV in the oral cavity or oropharynx, known collectively as oral HPV infection, vary substantially from study to study (<1%6 to >50%7). The variability has been attributed to differences in study populations and approaches to specimen collection, processing, and testing. A recent population-based study estimated oral HPV prevalence among ages 14 to 69 years in the United States as 6.9%.8 No such estimate is available for incidence.
Although transmission of HPV to the oral cavity and oropharynx is hypothesized to occur mainly through sexual contact, determinants of acquisition have not been fully established through longitudinal studies. Cross-sectional associations between prevalent oral HPV infection and sexual behaviors, such as higher numbers of vaginal sex partners or oral sex partners, have been observed in some studies,8–11 but not in others.12,13 Prevalent oral HPV infection also has been linked with male sex, increasing age, cigarette smoking, and cervical HPV infection.8–12,14,15
We conducted a longitudinal study to determine the prevalence and incidence of oral α-genus HPV infection (overall and by HPV type) in young men and to examine factors associated with new infection.
MATERIALS AND METHODS
Participants were recruited between September 2008 and April 2010 from an ongoing male HPV cohort study at the University of Washington, Seattle, Washington.16 The study protocol was approved by the University of Washington Human Subjects Division.
Men were eligible for the oral HPV study if they (1) were 18 to 25 years; (2) were permanent residents of Washington state; (3) had in general good health and immune competent; (4) were able to provide informed consent; and (5) were sexually active with women. Of the 214 men eligible from the parent study, none refused to participate in the oral HPV study. Two men however were excluded because they reported receiving the HPV vaccine before September 2008.
Participants were seen every 4 months for up to 1.5 years for the oral HPV study (3 years for the parent study). Clinic visit procedures for the parent study have been described elsewhere.16 Briefly, at each visit, the research nurse practitioner administered a medical and sexual history interview followed by genital examination. The examination included collection of exfoliated cell specimens for HPV genotyping from the penile shaft, glans, and scrotum. The exfoliated cells were placed in site-specific vials that contained 1 mL of Specimen Transport Medium (STM) (Qiagen, formerly Digene, Gaithersburg MD). At each visit, urine was collected for detection of HPV DNA from the internal urethral meatus.16 If clinically indicated, urine was tested for Chlamydia trachomatis and Neisseria gonorrhoeae (ATPIMA, Gen-Probe, San Diego, CA). Hyponychial specimens were self-collected on both hands using a cytology brush to rub the underside of each fingernail tip. Between visits, participants completed biweekly online diaries in which they recorded sexual behaviors for each day and each sex partner.
After completing the parent study visit procedures, men were recruited for the oral HPV study. All interested men underwent the informed consent process. At enrollment and at each follow-up visit, the nurse practitioner administered an additional questionnaire. It included questions on oral sex, open-mouth kissing, respiratory tract infections, tobacco use, marijuana use, tonsillectomy, mononucleosis, gastroesophageal reflux disease (GERD), GERD-like symptoms, nail biting, and dental hygiene.
The nurse practitioner then oversaw collection of 2 types of oral specimens for HPV genotyping: a gargle/rinse and 2 self-collected oropharyngeal swabs. For the gargle/rinse, participants rinsed and gargled 10 mL of Scope mouthwash (Proctor & Gamble, Cincinnati, OH) for 30 seconds. After collection, the mouthwash was poured into a separate vial and centrifuged for 5 minutes. The resulting cell pellet was stored in 1 mL STM. The participant was then instructed and supervised in mirror-assisted self-collection with a polyester fiber-tipped swab, first from the tonsils or palatine arches and then from the back of the tongue and pharynx. The swabs were stored in 1 mL of STM.
α-Genus HPV DNA Detection and Genotyping
The STM samples were digested with 20-µg/mL protease K at 37°C for 1 hour. For each sample, 400 µL was used to isolate DNA by QIAamp DNA blood mini kit, following the manufacturer’s protocol (Qiagen, Cat. no. 51104). Each sample was digested with protease K at 56°C for 10 minutes and then loaded onto 1 QIAamp column. The column was washed and DNA eluted in 50 µL heated TE (Tris, EDTA buffer, 70°C). Human papillomavirus–specific DNA was polymerase chain reaction amplified and detected by dot blot hybridization using the MY-09-MY11-HMB01 primers and probes. Human papillomavirus–positive samples were subsequently genotyped for 37 α-genus HPV types (HPV types, 6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 64, 66, 67, 68, 69, 70, 71, 72, 73, 81, 82, 83, 84, IS39, and CP6108) using a liquid bead microarray assay based on Luminex technology, as described in detail by Feng et al.17 Specimens negative for both β-globin and HPV DNA were considered insufficient for HPV DNA testing. Human papillomavirus types 61 and 71 were excluded from the analyses owing to quality control issues with those specific HPV types.
A positive HPV DNA test result for the swab or gargle/rinse was considered to be a detection of oral HPV infection. Prevalent infection was defined as detection of oral HPV infection at enrollment. Incident oral HPV infection of any HPV, or of a specific type, was defined as the first oral HPV detection after observed negative result(s) for all types, or for the specific type, respectively. The time of incident infection was defined as the visit at which it was first detected. We estimated cumulative probabilities of incident type-specific oral HPV infection using Kaplan-Meier analysis with any HPV infection measured as the first infection of any HPV type. We measured concordance between oral collection methods using an unweighted κ statistic and percentile bootstrap methods with 1000 repetitions to calculate 95% confidence intervals (CIs) for κ while accounting for correlation within participants.
To examine potential determinants of incident oral HPV infection, we estimated hazards ratios (HRs) using marginal Cox proportional hazards models.18 Data on all HPV types were included in the models, with stratification by type and correlation within participants accounted for using generalized estimating equations (GEEs) with robust SEs.19 With the exception of age and self-reported race, we modeled potential determinants as time-varying covariates, which were assessed for the 4 months before visit date (e.g., number of oral sex partners in the last 4 months). For characteristics measured continuously, we created categories by quartile or tertile. Characteristics associated with oral HPV infection in unadjusted analyses (P ≤ 0.05) were considered for the multivariate model.
The 212 men (median age, 20 years) enrolled in the oral HPV study were predominately white and mostly nonsmokers (Table 1). The median lifetime number of vaginal sex partners was 3, and the median number of female partners on whom the participants performed oral sex was 2. Four men reported a history of sex with men. The median follow-up time was 10.7 months (interquartile range [IQR], 4.4–13.2 months) and the total number of study visits was 704. The median time between visits was 3.9 months (IQR, 3.4–4.4 months).
During study visits, 75 type-specific oral HPV infections (prevalent, incident, or persistent infection) were detected in 33 men; 37 (49%) by gargle/rinse only, 29 (39%) by self-collected swab only, and 9 (12%) by both methods (κ = 0.21; 95% CI, 0.07–0.34) (Table, Supplemental Digital Content 1, which enumerates HPV type-specific agreement between methods). Insufficiency (absence of β-globin DNA) for the gargle/rinse was 0.2% (2/702) and 0.1% (1/704) for the self-collected swab. At these same visits, 628 type-specific genital HPV infections were detected in 126 men, and 192 type-specific hyponychial HPV infections were detected in 78 men. Among the 75 type-specific oral HPV infections, the same HPV type was detected concurrently in the genitals of 15 men (40.0% of infections) and the same type detected in a hyponychial site of 22 men (21.3% of infections).
Prevalence of oral HPV infection at enrollment was 7.5%, with just more than half of infected men positive for more than 1 HPV type (Table 2). Among the 34 type-specific prevalent infections, the most common were HPV-16 (2.8%), HPV-18 (2.4%), HPV-33 (1.9%), and HPV-39 (1.9%). Of the 28 men, 8 (28.6%) with prevalent infection and at least 1 follow-up visit had the same HPV type detected at 1 or more follow-up visits. Because we did not sample men at shorter intervals, durations shorter than 4 months could not be measured.
The 12-month cumulative incidence of any oral HPV infection was 12.3% (Fig. 1), and it was 3.3% for multiple concurrent oral HPV infections, 0.8% for HPV-16, 2.7% for HPV-18, and 2.3% for HPV-39 (Table 2). Of the 26 men with incident oral HPV infection, 7 had 1 or more visits after the initial detection and none were repeatedly positive for the same HPV type.
Tables 3 and 4 contain HRs for univariate and multivariate associations with incident oral HPV infection that were estimated using data on all HPV types and GEE modeling techniques.
Age and race/ethnicity were not associated with incident oral HPV infection (Table 3). Although the likelihood of incident oral HPV infection was higher for some forms of smoking, none of these associations were statistically significant (Table 3). No incident infections were detected in men who chewed tobacco, although the practice was uncommon (2% at enrollment). Higher frequency of teeth brushing was associated with a nonstatistically significant lower likelihood of incident oral HPV infection. Number of alcoholic drinks when examined by beverage type (beer, wine, or liquor) or in total, history of GERD, symptoms of GERD, and recent upper respiratory tract infection were not associated with incident oral HPV infection (data not shown). Although none of 20 men who reported having a tonsillectomy at enrollment developed a new oral HPV infection during follow-up, 8 (40%) had an oral HPV infection at enrollment.
Higher frequencies of performing oral sex and open-mouth kissing and report of any anal sex with a man in the last 4 months were statistically significantly associated with incident oral HPV infection, whereas recent number of oral, vaginal, or digital sex partners and frequency of vaginal or digital sex were not (Table 3). No incident oral HPV infections were detected in men who used dental dams while performing oral sex, although the use was rare among participants (<1% of visits). There were no incident oral HPV infections among men with genital HSV-2, C. trachomatis, or N. gonorrhoeae infections, but these were also rare among participants (combined <1% of visits).
Concurrent HPV infection of a genital or hyponychial site was associated with incident oral HPV infection with the same HPV type (Table 3). Genital or hyponychial HPV infection 1 visit previously was also associated with incident oral HPV infection. However, among men with incident oral HPV infection, none within the subset that had had a previous genital or hyponychial HPV infection with the same HPV type had cleared that infection(s) at the time of incident oral HPV infection detection. Other factors potentially related to autoinoculation, including masturbation, fingernail biting, current genital warts, and circumcision status, were not associated with incident oral HPV infection (data not shown).
Independent associations observed in the multivariate model indicated that risk of incident oral HPV infection was elevated for men who (1) performed oral sex once or more per week versus less than once a week in the last 4 months (HR, 3.6; 95% CI, 1.4–9.8; P = 0.009), (2) reported recent anal sex with men versus no anal sex with men (HR, 42.9; 95% CI, 9.0–205.5; P < 0.0001), (3) had current infection with the same HPV type at a genital site versus no/other HPV type genital infection (HR, 6.2; 95% CI, 2.4–16.4; P = 0.0002), and (4) had current infection with the same HPV type at a hyponychial site versus no/other type HPV hyponychial infection (HR, 11.8; 95% CI, 4.1, 34.2; P < 0.0001) (Table 4).
In a cohort of sexually active young men, 7.5% were infected in the oral cavity or oropharynx with α-genus HPV at enrollment and an additional 12.3% became infected during the next 12 months. Human papillomavirus type 16 was one of the more common HPV types, although its prevalence (2.8%) and 12-month cumulative incidence (0.8%) were low. The prevalence of oral HPV infection in this study was similar to the estimate of 6.5% for US males aged 18 to 24 years in 2009 to 2010,8 although greater than an estimated 3% reported for 2 studies with more than 200 men in this age range.11,12 These estimates, however, used only gargle/rinse specimens, which would have reduced our prevalence estimate to 5%. Similarly, the 12-month cumulative incidence observed among Finnish fathers-to-be (mean age, 29 years) was approximately 5% as compared with our estimate of 12.3%, but oral specimens were collected by cytobrush only and tested only for 12 HPV types.13 Collection by both gargle/rinse and directed swab/cytobrush seems to be more sensitive than collection by a single method.20 The poor-to-fair agreement between the gargle/rinse and swabs targeting the oropharynx observed in our study (κ = 0.21) has been reported by others (κ = 0.16).14
The duration of most oral HPV infections appeared to be short with 72% of prevalent infections no longer detected, or “cleared,” within approximately 4 months. This is a higher rate of clearance than observed in the Finnish fathers-to-be study,13 which included older married men, and testing for fewer HPV types (clearance rates vary by type for genital infections21,22). That the prevalence of oral HPV infection observed in this study was lower than the cumulative incidence at 12 months also suggests a short duration of infection given that prevalence can be estimated as product of incidence and duration. In fact, one might expect the prevalence to be lower based on the clearance we observed in our study population. However, because of the small number of infections for which we could observe duration (n = 15) and the length of time between study visits (4 months), further research is needed to assess duration of oral HPV infection and its correlates.
Our findings support the hypothesis that α-genus HPV types exhibit a decreased tropism for epithelial cells of the oral cavity and hyponychium compared with the genitals. For this young male cohort, the published 12-month cumulative incidence of genital HPV infection was approximately 35%,16 which is similar to the estimate (39.3%) for a large multinational study of men (mean age, 32 years)21 and approximately 3-fold higher than was estimated in the present study for oral HPV infection (12.3%). Among men with oral HPV infection, 45% had the same type detected in the genitals concurrently. Concurrency was higher than that which has been observed between the oral cavity and the cervix in females,9,15 which could be caused by anatomic differences. The published 12-month cumulative incidence of hyponychial HPV infection (15%)16 in the parent study was similar to the estimate for oral infection (12.3%) in the present study and among men with an oral HPV infection; two thirds were infected in a hyponychial site.
We found that risk of oral HPV infection was more strongly associated with frequency of sexual contact than with number of new sex partners. For example, frequency of performing oral sex but not number of oral sex partners within the last 4 months was associated with oral HPV infection. Other studies that have examined the relationship between oral sex and oral HPV infection have had mixed results.8,11–14,23 Reasons for not observing an association could include not capturing information on oral sex type (performing vs. receiving), frequency, or relevant time frame.
It was difficult to discern whether sexual behaviors other than performing oral sex were independently associated with oral HPV because concurrency of sexual behaviors was high. More than any other sexual behavior measured, men who recently performed oral sex were most likely to report recent deep kissing, vaginal sex, and digital sex with females (data not shown). Although the association we observed between having anal sex with men and incident oral HPV infection was strong, it lacked precision because of small numbers and the absence of information on oral sex with men. A more focused study of men who have sex with men is warranted, given our results and similar associations reported for prevalent infection.8,11,23
The strong associations between incident oral HPV infection and infection with the same HPV type at other sites indicate that transmission may occur through autoinoculation. However, although these associations were independent of recent frequency of sexual contact (oral sex or open-mouth kissing), it is possible that HPV acquisition at each anatomic site occurred independently through an infected partner at some point between study visits. We did not have enough men without some form of sexual contact between visits to investigate this further. It is also possible that there are other unmeasured genetic or environmental factors that make some men more susceptible to α-genus HPV infection at any site.
In men, current smoking has been associated with prevalent oral HPV infection,11,12 although not in all studies,8 and it was not associated with incident infection in our study. It is possible that smoking is related to oral HPV infection persistence9 and not acquisition. Further study of oral HPV risk in populations with higher smoking rates may be warranted. Marijuana use has been reported as a risk factor for HPV-positive oropharyngeal cancer,24 although not prevalent oral HPV infection.8 We found risk estimates for marijuana use and for current cigar and pipe smoking that were modestly elevated but were not statistically significant.
Our study was not without limitations. There is no criterion standard for oral HPV specimen collection,25 and although the HPV type-specific assay used in this study has performed well in comparative analyses of test results for cervical specimens and purified plasmid samples,17,26 similar analyses have not yet been completed for specimens collected for oral HPV DNA detection. By using a swab that targeted the oropharynx and a gargle at the back of the throat, we aimed to improve sensitivity. Our relatively small sample size affected the precision of our measure of oral sampling agreement and our risk estimates as well as potentially limited our ability to detect more modest associations and the extent to which we could separate out correlated behaviors. Lastly, our findings pertain to young men who have sex with women, and they might not generalize to older men, men with different sexual behaviors, women, or immunocompromised men and women.
In summary, approximately 1 in 5 young men enrolled in this study had evidence of at least 1 oral HPV infection within a 12-month interval. Notably, most infections were transient, and HPV-16, the type most often detected in oropharyngeal cancer, was not common. Frequent performance of oral sex, sex with another man, and infection of the same HPV type at another site were independently associated with incident oral HPV infection, but attribution of transmission to specific behaviors was undoubtedly incomplete. Skin-to-skin contact between multiple anatomic sites is an important, but difficult, to measure aspect of sexual activity and personal hygiene. Looking beyond oral HPV infection to the natural history of HPV-related oropharyngeal cancer is essential but complicated because characteristics of precancerous oral lesions and their relationship to oral HPV infection are still being defined.27,28 The increasing rate of HPV-related oropharyngeal cancer (especially in men), the lack of early detection methods similar to those for cervical cancer, the uncertain benefits of promoting less skin-to-skin contact for prevention of HPV infection, and the high efficacy of HPV vaccines for preventing infections at different sites29–31 all suggest that it is time to investigate and implement cost-effective adolescent HPV vaccination programs to prevent oral HPV-16 infection.
1. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Human papillomaviruses. IARC Monogr Eval Carcinog Risks Hum 2007; 90: 1–636.
2. Gillison ML, Koch WM, Capone RB, et al.. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. J Natl Cancer Inst 2000; 92: 709–720.
3. Chaturvedi AK, Engels EA, Anderson WF, et al.. Incidence trends for human papillomavirus-related and -unrelated oral squamous cell carcinomas in the United States. J Clin Oncol 2008; 26: 612–619.
4. Fast Stats. An interactive tool for access to SEER cancer statistics [online database]. Bethesda, MD. Surveillance Research Program, National Cancer Institute. Available at: http://seer.cancer.gov/faststats/
. Accessed on December 10, 2011.
5. Chaturvedi AK, Engels EA, Pfeiffer RM, et al.. Human papillomavirus and rising oropharyngeal cancer incidence in the United States. J Clin Oncol 2011; 29: 4294–4301.
6. Eike A, Buchwald C, Rolighed J, et al.. Human papillomavirus (HPV) is rarely present in normal oral and nasal mucosa. Clin Otolaryngol Allied Sci 1995; 20: 171–173.
7. Terai M, Hashimoto K, Yoda K, et al.. High prevalence of human papillomaviruses in the normal oral cavity of adults. Oral Microbiol Immunol 1999; 14: 201–205.
8. Gillison ML, Broutian T, Pickard RK, et al.. Prevalence of oral HPV infection in the United States, 2009–2010. JAMA 2012; 307: 693–703.
9. D’Souza G, Fakhry C, Sugar EA, et al.. Six-month natural history of oral versus cervical human papillomavirus infection. Int J Cancer 2007; 121: 143–150.
10. Smith E, Swarnavel S, Ritchie J, et al.. Prevalence of human papilloma-virus in the oral cavity/oropharynx in a large population of children and adolescents. Pediatr Infect Dis J 2007; 26: 836–840.
11. D’Souza G, Agrawal Y, Halpern J, et al.. Oral sexual behaviors associated with prevalent oral human papillomavirus infection. J Infect Dis 2009; 199: 1263–1269.
12. Kreimer AR, Villa A, Nyitray AG, et al.. The epidemiology of oral HPV infection among a multinational sample of healthy men. Cancer Epidemiol Biomarkers Prev 2011; 20: 172–182.
13. Rintala M, Grénman S, Puranen M, et al.. Natural history of oral papillomavirus infections in spouses: A prospective Finnish HPV Family Study. J Clin Virol 2006; 35: 89–94.
14. Kreimer A, Alberg A, Daniel R, et al.. Oral human papillomavirus infection in adults is associated with sexual behavior and HIV serostatus. J Infect Dis 2004; 189: 686–698.
15. Fakhry C, D’souza G, Sugar E, et al.. Relationship between prevalent oral and cervical human papillomavirus infections in human immunodeficiency virus-positive and -negative women. J Clin Microbiol 2006; 44: 4479–4485.
16. Partridge JM, Hughes JP, Feng Q, et al.. Genital human papillomavirus infection in men: Incidence and risk factors in a cohort of university students. J Infect Dis 2007; 196: 1128–1136.
17. Feng Q, Cherne S, Winer R, et al.. Development and evaluation of a liquid bead microarray assay for genotyping genital human papillomaviruses. J Clin Microbiol 2009; 47: 547–553.
18. Wei L, Lin D, Weissfeld L. Regression analysis of multivariate incomplete failure time data by modeling marginal distributions. J Am Stat Assoc 1989; 84: 1065–1073.
19. Lin D. Cox regression analysis of multivariate failure time data: The marginal approach. Stat Med 1994; 13: 2233–2247.
20. Lawton G, Thomas S, Schonrock J, et al.. Human papillomaviruses in normal oral mucosa: A comparison of methods for sample collection 1. J Oral Pathol Med 1992; 21: 265–269.
21. Giuliano AR, Lee JH, Fulp W, et al.. Incidence and clearance of genital human papillomavirus infection in men (HIM): A cohort study. Lancet 2011; 377: 932–940.
22. Trottier H, Franco EL. The epidemiology of genital human papillomavirus infection. Vaccine 2006; 24 (suppl 1): S1–S15.
23. Coutlée F, Trottier AM, Ghattas G, et al.. Risk factors for oral human papillomavirus in adults infected and not infected with human immunodeficiency virus. Sex Transm Dis 1997; 24: 23–31.
24. Gillison ML, D’Souza G, Westra W, et al.. Distinct risk factor profiles for human papillomavirus type 16-positive and human papillomavirus type 16-negative head and neck cancers 1. J Natl Cancer Inst 2008; 100: 407–420.
25. Braakhuis B, Brakenhoff R, Meijer C, et al.. Human papilloma virus in head and neck cancer: The need for a standardised assay to assess the full clinical importance. Eur J Cancer 2009; 45: 2935–2939.
26. Eklund C, Forslund O, Wallin KL, et al.. The 2010 global proficiency study of human papillomavirus genotyping in vaccinology. J Clin Microbiol 2012; 50: 2289–2298.
27. Miller C, White D. Human papillomavirus expression in oral mucosa, premalignant conditions, and squamous cell carcinoma: A retrospective review of the literature. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996; 82: 57–68.
28. Varnai A, Bollmann M, Bankfalvi A, et al.. The prevalence and distribution of human papillomavirus genotypes in oral epithelial hyperplasia: Proposal of a concept. J Oral Pathol Med 2009; 38: 181–187.
29. Giuliano AR, Palefsky JM, Goldstone S, et al.. Efficacy of quadrivalent HPV vaccine against HPV Infection and disease in males. N Engl J Med 2011; 364: 401–411.
30. The FUTURE II Study Group. Quadrivalent vaccine against human papillomavirus to prevent high-grade cervical lesions. N Engl J Med 2007; 356: 1915–1927.
31. Palefsky JM, Giuliano AR, Goldstone S, et al.. HPV vaccine against anal HPV infection and anal intraepithelial neoplasia. N Engl J Med 2011; 365: 1576–1585.