HUMAN PAPILLOMAVIRUSES (HPVs) infect cutaneous and mucosal epithelial cells.1 The HPV DNA also has been detected in a variety of oral lesions.2–4 The presence of HPV infection in the oral cavity inducing oral papillomas can affect up to 10% of individuals seropositive for the human immunodeficiency virus (HIV).5 Periodontal HPV infection was reported in 20% of HIV-infected adults with acute periodontitis.4 Oral HPV infection inducing koilocytic changes have been described in the course of HIV infection.3,6 Human papillomavirus also may sometimes generate atypical oral lesions.3,7–9 Unusual HPV types can infect the oral cavity of HIV-infected individuals.8
Thus far, the estimates of prevalence of HPV infection in the oral cavity were based on studies that recruited patients with oral lesions or on studies with small sample sizes of healthy patients.4,10,11 Several studies have reported an association between genital HPV infection and sexual activity.12–14 Much less is known about risk factors involved in HPV infection of the oral cavity. A recent case-control study showed that oral-genital sexual activity predisposed to oral carcinoma.15 Oral squamous carcinomas also contain HPV DNA sequences from oncogenic types that usually are detected in the genital epithelium.2,3 Nevertheless, the role of HPV infection in oral carcinoma remains controversial.16,17 In patients infected with HIV, the presence of the same HPV types in the oral and anal epithelia suggests that HPV could be transmitted by oral-genital sexual contact.18 Moreover, HPVs isolated in the mouth often are genital types.4,10,19 Multivariate analyses need to be done to further define the relation between oral HPV infection and sexual activity.
We conducted this cross-sectional study to estimate the prevalence of HPV oral infection in HIV-infected and HIV-negative sexually active individuals, in the absence of proliferative oral lesion. We also investigated the correlation between oral HPV infection assessed with a consensus polymerase chain reaction (PCR) assay and sexual activity.
Materials and Methods
Study protocols were approved by the Ethics Committees of participating institutions. Written informed consent was obtained from all participants at enrollment. The HIV-seropositive and HIV-negative individuals were enrolled through the following clinical sources. First, HIV-seropositive and HIV-negative men and women 16 to 60 years of age were recruited consecutively between January 1994 and April 1994 from four outpatient sexually transmitted disease (STD)-HIV clinics. Individuals were considered HIV-negative if they tested negative for anti-HIV antibodies less than 2 months before inclusion. Participants also were recruited consecutively from September 1992 to April 1994, as part of a study on the role of HPV in esophageal diseases (Coutlée F, et al, manuscript submitted). The HIV-seropositive adults 16 to 60 years of age who were scheduled for gastrointestinal endoscopy were recruited consecutively by two endoscopists. The HIV-negative adults matched for age (10-year age groups) with HIV-seropositive individuals were recruited from the same endoscopy clinics. Subjects were eligible whether they consulted the endoscopy clinic as outpatients or while they were admitted to the hospital. The presence of risk factors for HIV infection was evaluated for all individuals by a standardized questionnaire. Individuals from the endoscopy clinics were considered HIV-negative if they tested negative for anti-HIV antibodies less than 2 months before inclusion or if they had no risk factor for HIV infection. In the presence of risk factors, HIV-1 antibodies were measured by serologic analysis. Serologic testing for HIV infection was done by commercial enzyme-linked immunosorbent assay tests with confirmation by Western blot or radioimmunoprecipitation assay by the Laboratoire de Santé Publique du Québec.
Two trained study personnel administered a structured and standard questionnaire to obtain information on sociodemographic variables, current and past sexual behavior, history of STD, and tobacco and alcohol consumption. During oral sampling (see below), the presence of proliferative lesions of the oral mucosa was recorded. Medical records of all participants were reviewed to confirm the HIV infection status, CD4 T-lymphocyte counts, previously diagnosed STDs, and opportunistic infections. The CD4 T-lymphocyte counts had been measured less than 2 months' time from the time of inclusion.
Before the interview, oral epithelial cells were collected with a cytobrush by scraping the oral mucosa at both cheeks, hard palate, and dorsum of the tongue. The cytobrush was agitated in 10 ml of phosphate-buffered saline. Cell suspensions were kept at 4 °C for at most 4 days before processing. Cells then were pelleted after centrifugation at 2,500 rpm for 10 minutes at 4 °C and resuspended in 10-mM Tris, pH 8.3. Cell suspensions were lysed with Tween 20 and NP-40 and digested with proteinase K as described previously.20 Cell lysates were stored at −70 °C until amplified.
Human Papillomavirus DNA Amplification
Cell lysates were amplified in duplicates with the L1 consensus HPV primers MY09 and MY11 under standard conditions,21,22 without adding the HPV-51-specific primer HMB01. Negative controls, weakly positive controls of 10 DNA copies of HPV-18, and strongly positive controls of HPV types 6/11, 16, 31, 33, 35, and 45, were included. A 268-pb β-globin DNA fragment was amplified with PC04/GH2021 to control for DNA integrity, the presence of an adequate number of epithelial cells for PCR analysis, and the absence of inhibitors. The DNA from β-globin-negative lysates was extracted with phenol chloroform, precipitated with ethanol, and amplified for β-globin and HPV DNA. If negative for β-globin, the specimen was rejected as inadequate for HPV testing. Measures to avoid false-positive reactions due to contamination were followed strictly.20
Human Papillomavirus Amplified Products Detection
The PCR products spotted on nylon membranes were first reacted under low stringency conditions with an HPV generic probe mixture labeled with 32P-deoxynucleotides.23 This generic probe efficiently detects common genital types.23,24 All membranes also were reacted under stringent conditions with type-specific oligonucleotide probes end-labeled with 32P-ATP,21 for HPV-6/11, -16, -18, -31, -33, -35, -39, -45, -51, -52, -53, -56, and -58.22,25,26 Untyped HPVs were those reactive with the generic probe but not with type-specific probes.
In univariate analyses, odds ratio and 95% confidence interval were calculated to estimate the relative risk of HPV oral infection with several demographic and sexual factors. The P values were calculated with chi-square tests and Mantel-Haensel tests for ordinal variables. The student's t test was applied for comparisons of continuous variables. Continuous variable are expressed as the mean ± one standard deviation. Multivariate analyses were performed using stepwise logistic regression to examine the independent and joint effects of risk factors. To evaluate the role of variables that could be masked by HIV status, a second model was constructed for logistic regression without considering the HIV status. The most significant variables in univariate analysis were included in the models. All analyses were performed using the software SPSS.
Source of Patients
We identified a total of 323 eligible individuals (Figure 1), and oral samples were obtained from 317 (98.1%). Data collected from questionnaires were complete for 287 (88.9%) eligible participants. The rate of refusal to participate was 2.6% for the endoscopy clinic and 1.5% for STD-HIV outpatient clinic settings (P = 0.372).
Characteristics of the Study Population
No proliferative lesion was noticed in the oral cavity of the participants. Considering only the 287 individuals for whom questionnaires were completed, 178 (62%) were HIV-seropositive (158 men and 20 women) and 109 (38%) were HIV-negative (73 men and 36 women). Overall, 148 (51.6%) individuals were homosexual men (134 were HIV-seropositive), 29 (10.1%) were bisexual men (24 were HIV-seropositive), 54 (18.8%) were heterosexual men (2 were HIV-seropositive), and 56 (19.5%) were heterosexual women (20 were HIV-seropositive). The mean age of participants reached 36.7 ± 9.3 years (median, 36.8 years; range, 16.1–60.3 years). Participants had a median of eight sex partners (range, 1–7,000) in the past 10 years. Median age at first sexual intercourse was 17.2 years old (range, 6–31 years old).
Participants infected with HIV differed from HIV-negative individuals. The proportion of men recruited was greater in the HIV-seropositive group (88.5%) than in the HIV-negative group (68.7%; P < 0.0001). A higher percentage of homosexual men was found in the HIV-infected group (79.3% versus 20.3%; P < 0.0001). The HIV-infected individuals were more often single (77%) than were the HIV-negative individuals (51.8%; P < 0.0001) and also were older (38.7 ± 8.4 years compared to 33.4 ± 9.9 years; P < 0.001). The HIV-infected individuals had more sexual partners (202 ± 727 in the past 10 years) than did the HIV-negative participants (66 ± 275; P = 0.056). Also, 15.3% of HIV-seropositive and 2.6% of HIV-negative participants had made use of illicit intravenous drugs (P = 0.005).
Human Papillomavirus DNA Detection in Oral Brushings
Of the 317 individuals for whom oral samples were tested, 202 were HIV-seropositive and 115 were HIV-negative. Three hundred fifteen (99.4%) samples yielded a positive signal for β-globin amplification, showing the absence of PCR inhibitor and sufficient quantity of cellular DNA for analysis. The HPV DNA sequences were detected in 32 (10.2%) of 315 oral brushings. Thirteen (11.4%) of 113 subjects enrolled from the endoscopy clinics and 19 (9.4%) of 202 individuals from the HIV-STD clinics provided samples containing HPV sequences (P = 0.691). Of those who tested positive for HPV, 29 were HIV-seropositive, including 12 (23.1%) of the 52 patients from the endoscopy clinics and 17 (11.4%) of the 149 from the HIV outpatient clinic (P = 0.067). Three HIV-negative individuals also were infected with HPV, including 1 (1.6%) of the 61 patients from the endoscopy clinics and 2 (3.7%) of the 54 patients from STD-HIV clinics (P = 0.155).
Distribution of Human Papillomavirus Types
The distribution of HPV types is listed in Table 1. Only one sample yielded multiple HPV types (HPV-33 and HPV-35). The most frequent types were HPV-16, HPV-35, and unidentified HPV types. Most oral HPV infections (71.9%) were caused by intermediate or high-risk oncogenic HPV types. Unadjusted rates for HPV DNA detection were significantly greater in HIV-seropositive (29 [14.4%] of 201) than in HIV-negative (3 [2.6%] of 114) individuals (P = 0.001).
Univariate Analysis of Risk Factors for Oral Human Papillomavirus Infection
Of the 287 interviewed individuals who completed the questionnaire, 29 (12.6%) of 231 men and 3 (5.4%) of 56 women were infected with HPV in the oral cavity (P = 0.13). To evaluate the role of sexual risk factors for oral HPV infection, we examined the associations between HPV DNA detection with several markers of sexual activity (Table 2). Most of the demographic variables failed to exhibit important correlations with the likelihood of finding a specimen containing HPV DNA. In univariate analyses, there were statistically significant increases in risk of HPV infection with HIV infection, homosexuality, unprotected oral intercourse, and previous STDs. There was a significant linear trend between the number of sexual partners in the past 10 years and oral HPV infection. Although a trend was found between age and oral HPV infection, the chisquare test, including all the age categories, was not significant and the 95% confidence interval of each odds ratio included 1.0. The small number of HPV-positive individuals precluded us to analyze risk factors for each HPV genotype. Of the 32 individuals with oral HPV infection, 26 (81%) admitted to homosexual activity, whereas 3 were heterosexual men and 3 were heterosexual women.
A greater proportion of HPV-infected (90.6%) individuals than uninfected (63.9%) individuals reported that they had practiced oral sex on their partners without the protection of a condom (odds ratio, 5.5; 95% confidence interval, 1.6–18.4). Oral contact with female genital organs (data not shown) and oral intercourse with a condom were not risk factors. There was a trend for a higher number of sexual partners for HPV-infected participants (mean, 327 ± 1232 partners in the past 10 years) as opposed to those (mean, 131 ± 479 partners in the past 10 years) without HPV infection (P = 0.08). The number of sexual partners in the past 10 years was correlated with age at first intercourse (P = 0.026) when participants were classified as in Table 2. The mean age at first sexual intercourse was similar for patients infected (16.8 ± 4.2 years) and not infected (17.2 ± 3.8 years) with HPV (P = 0.58).
Association of Oral Human Papillomavirus Infection With Other Sexually Transmitted Diseases
A history of one or more non-HPV-related STD (Chlamydia trachomatis, genital herpes, syphilis, or gonorrhea) was reported by 64.8% of participants. Among those with oral HPV infection, 27 (84.4%) of 32 individuals reported at least 1 prior STD, as compared with 159 (62.4%) of 255 individuals among those without HPV infection (P = 0.024). The number of reported previous episodes of STD also was highly associated with oral HPV infection (data not shown). In particular, HPV infection was significantly associated with past gonorrhea infection. The other STDs also were associated with HPV infection to a lesser extent (Table 2). A history of genital warts was not associated with an increase in risk. An HPV infection was significantly correlated with HIV infection status, but it was not correlated with the level of immunosuppression. Considering only HIV-seropositive individuals, the mean CD4+ T-lymphocyte counts were similar between HPV-negative (180.9 ± 216.5/mm3) and HPV-infected (139.6 ± 190.9/mm3) participants (P = 0.37). The proportion of subjects with acquired immune deficiency syndrome in the HPV-negative group (82 [52.2%] of 157) and in the HPV-infected group (15 [51.7%] of 29) was not statistically different (P = 0.10).
Independent Risk Factors for Oral Human Papillomavirus Infection
Because of the wide difference in sexual activity between patients infected with HIV and those not infected with HIV, variables significantly related to HPV infection in the univariate analyses (sexual orientation, number of sexual partners, HIV status, C. trachomatis, genital herpes, syphilis, or gonorrhea, unprotected oral intercourse), gender, age, genital condyloma, and protected oral intercourse, were included in a stepwise multivariate model analysis with and without knowledge of HIV status, to determine independent risk factors for HPV infection in the oral cavity. In the initial model, the strongest independent predictor of HPV infection remained the HIV infection status (Table 3). A history of gonorrhea and C. trachomatis (Table 3) also persisted as risk factors for HPV infection. Redundant variables associated with the significant variables included homosexuality, a variable strongly associated with the HIV status, syphilis, genital herpes, and unprotected intercourse. Because the presence of previous STDs was assessed by means of a questionnaire, the infectious agents associated with HPV should be considered as illustrating the degree of exposure of participants to STDs.
When the HIV status was not considered in the logistic regression model (Table 3), the presence of previous gonorrhea remained a strong independent predictor for HPV infection. However, practice of unprotected orogenital intercourse emerged as independently associated with HPV infection.
In this study, we have evaluated the rate of detection of HPV in oral epithelial cells collected from sexually active individuals, 62% of whom were infected with HIV. The overall prevalence of HPV DNA of 10.2% in our study is in agreement with rates from 3.8% to 15.0% of HPV infection in oral scrapings reported by previous groups.15,18 Some investigators have, however, described higher detection rates of HPV infection from 31% to 60% on a limited number of individuals.10,27,28
In some studies, full-depth biopsies of lesions were performed and may have increased the yield of HPV as compared to exfoliated cells.2,4,5,8,11,17,29,29–32 Oral brushings are, however, adequate to detect HPV in the oral cavity of individuals without lesion and to collect cells from various areas of the mouth.10 Some studies did not provide their protocol to control PCR contamination. The L1 consensus PCR assay we used was sensitive, as the 10 HPV-18 DNA copy control consistently scored positive. All samples included in the analysis tested positive for β-globin, indicating the absence of PCR inhibitors and showing the presence of a sufficient amount of cellular DNA for PCR analysis.
The selection of individuals with oral disease or with different risk factors also could explain the higher prevalence rates reported by some investigators.2,4,11,17,30,32 Unfortunately, the characteristics of the populations selected as well as the strategy for recruitment in some of these studies were not provided in detail. Individuals recruited here were at risk for STDs, as suggested by the important proportion of HIV-seropositive individuals and the fact that 53% of HIV-negative individuals had had at least six sexual partners in the past 10 years. Patient selection is critical, as shown by a trend for a higher rate of HPV infection in HIV-seropositive patients recruited from endoscopy clinics compared with those from outpatient STD-HIV clinics. Some studies cited above also included a limited number of participants.2,4,10,17,27,28,30
The prevalence of HPV infection in sexually active adults was lower than in the genital epithelium, where HPV DNA can be detected in approximately 50% of adults.14 We have shown that several HPV types can infect the oral epithelium without resulting in clinical disease, although the prevalence of oral HPV infection in HIV-negative individuals was low in our study. These results suggest that HPV frequently establishes a subclinical or latent infection in the oral cavity, which could serve as a reservoir for HPV transmission and future disease. Most of the HPV types detected in the mouth were intermediate- and high-risk oncogenic genital types. As found in other studies,2,28,29 HPV-16 was the most frequent type detected. In agreement with another study,10 infections with more than one HPV type was not as frequently encountered in the mouth as in genital infections.21,22 An important proportion of HPVs was of an unidentified type, which may contain HPV genotypes not transmitted sexually. Unfortunately, reagents to identify nongenital types HPV-13, HPV-32, and HPV-7 with the L1 PCR assay have not been developed yet. These types can be detected in oral lesions from individuals who are HIV infected.8
This study shows that individuals who are HIV-seropositive are infected more frequently in the oral cavity by HPV than are sexually active adults who are HIV-negative, even if recruited from STD clinics. The HIV status remained the strongest predictor for HPV infection in multivariate analysis. Homosexuality, strongly associated with HIV status, was a redundant variable in our analysis. It did not persist as an independent risk factor in the multivariate model in which HIV status was excluded. Most of the differences between HIV-seropositive and -negative individuals were related to HIV infection risk factors, including sexual preference, use of illicit intravenous drugs, and number of sexual partners. Individuals who are HIV infected also were older. Because the prevalence of genital HPV infection usually decreases with age,14,33,34 this difference could have introduced a bias against the association of HIV serologic status and oral HPV infection. Age did not persist as a significant factor for oral HPV infection after adjustment for the other variables. The immunosuppression induced by HIV could be a predisposing factor for clinical expression of latent HPV infection or could facilitate new infections. As opposed to anal HPV infection,18 the level of immunity was not a predisposing factor for oral infection.
Transmission of HPV by oral sex already had been proposed.8,18,35 Cytologic evidence of HPV infection in the uterine cervix was found in a minority of women with oral HPV-related lesions, thus not supporting concurrent transmission of HPV in the genital and oral tracts.35 In one study, nearly one fourth of mouth biopsy specimens from women with concurrent HPV genital infection contained HPV DNA, but HPV types isolated at both sites were different.29 Results obtained in our cross-sectional study showed a strong correlation between markers of sexual activity and oral HPV infection, suggesting that HPV can be transmitted in the oral cavity through sexual contact.
The present study showed that non-HPV-related STDs were strong independent predictors of HPV infection in the mouth. These STDs could, however, be markers for other forms of high-risk behaviors not identified here. A history of genital warts was not associated with an increase in risk. The presence of genital warts is a crude measure of HPV infection because most infections are subclinical and only a few HPV types cause these lesions. In fact, HPV types 6/11 rarely were encountered in our participants.
The number of sexual partners was significantly related to the presence of oral HPV infection in univariate analysis, but did not persist as an independent risk factor in multivariate analysis. A better examination of magnitude of risk would have resulted from the knowledge of total number of partners with whom unprotected oral sex occurred instead of the total number of partners irrespective of type of sexual contact. The presence of sexually transmissible agents in partners, indirectly measured by the history of previous STD in participants, was more important than was the actual number of partners. Similarly, prior STD also correlates with genital HPV infection in women.12,14 Our findings also could be explained by unmeasured sexual behavior acting as confounding, which would not refute the hypothesis of sexual transmission of HPV to the oral cavity. As an example, we have not considered oral-anal intercourse, which could increase the risk of oral HPV infection significantly.36 Oral infection with HPV was not evaluated in studies on risk factors for genital infections, but could represent an important variable that should be measured. The analysis of association between HPV infection and sexual behavior of HIV-positive and -negative homosexual men was limited by the small sample of HIV-seronegative homosexual men.
In contrast to our results, genital HPV infection has been correlated with the lifetime number of partners, younger age at first intercourse, and years of sexual activity.12–14 Several studies had found no increase in risk for genital HPV infection with sexual activity,37–39 but could have been affected by misclassification of HPV infection status.40 The presence of HPV DNA was assessed here with a standardized and sensitive PCR assay.21–23 Epithelial cells from different organs may exhibit various susceptibilities to HPV infection,41 thus affecting the role of sexual practices in transmission.
We found evidence to suggest that barrier contraception was associated with a decreased risk of HPV infection in the oral cavity. In our study, individuals who practiced unprotected oral sex on their partners nearly were four times more likely to be infected with HPV. Protection by condoms against genital HPV infection has not been shown clearly.12,14,42–44 This lack of protection could be explained by multifocal genital infections,21 infection before condom use, and passage of virus through pores of condoms.43 However, our results need confirmation because of the limited number of participants.
The results of the current study suggest that oral HPV infection without visible lesion is more frequently encountered in HIV-seropositive individuals and that the main determinants of oral HPV infection were correlates of sexual activity. Results obtained in our study cannot be generalized because of the selected population recruited. More studies are now needed to better define the natural history of HPV infection in the oral cavity in various populations. 13 Studies recruiting homosexual individuals who are HIV negative also would allow to isolate further risk factors for this infection.
1. Koutsky LA, Galloway DA, Holmes KK. Epidemiology of genital human papillomavirus infection. Epidemiol Rev 1988; 10:122–163.
2. Chang F, Syrjanen S, Kellokoski J, Syrjanen K. Human papillomavirus (HPV) infections and their associations with oral disease. J Oral Pathol Med 1991; 20:305–317.
3. Garlick JA, Taichman LB. Human papillomavirus infection of the oral mucosa. Am J Dermatopathol 1991; 13:386–395.
4. Madinier I, Doglio A, Cagnon L, Lefebvre JC, Monteil RA. Southern blot detection of human papillomaviruses (HPVs) DNA sequences in gingival tissues. J Periodontal Res 1992; 63:667–673.
5. Snidjers PJ, Engelbert AJM, Schulten, et al. Detection of human papillomavirus and Epstein-Barr virus DNA sequences in oral mucosa of HIV-infected patients by the polymerase chain reaction. Am J Pathol 1990; 137:659–666.
6. Silverman S, Migliorati CS, Lozada-Nur F, Greenspan D, Conant MA. Oral findings in people with or at high risk for AIDS: a study of 375 homosexual males. J Am Dent Assoc 1986; 112:187–192.
7. Eversole LR. Viral infections of the head and neck among HIV-seropositive patients. Oral Surg Oral Med Oral Pathol 1992; 73:155–163.
8. Greenspan D, de Villiers EM, Greenspan JS, De Souza YG, zur Hausen H. Unusual HPV types in oral warts in association with HIV infection. J Oral Pathol Med 1988; 17:482–487.
9. Syrjanen S, Von Krogh G, Kelloski J, Syrjanen K. Two different human papillomavirus (HPV) types associated with oral mucosal lesions in an HIV-seropositive man. J Oral Pathol Med 1989; 18:366–370.
10. Lawton G, Thomas S, Schonrock J, Monsour F, Frazer I. Human papillomaviruses in normal oral mucosa: a comparison of methods for sample collection. J Oral Pathol Med 1992; 21:265–269.
11. Yeudall WA, Campo MS. Human papillomavirus DNA in biopsies of oral tissues. J Gen Virol 1991; 72:173–176.
12. Davidson M, Schnitzer PG, Bulkow LR, et al. The prevalence of cervical infection with human papillomaviruses and cervical dysplasia in Alaska native women. J Infect Dis 1994; 169:792–800.
13. Reeves WC, Gary HE, Johnson PR, et al. Risk factors for genital papillomavirus infection in populations at high and low risk for cervical cancer. J Infect Dis 1994; 170:753–758.
14. Ley C, Bauer HM, Reingold A, et al. Determinants of genital human papillomavirus infection in young woman. J Natl Cancer Inst 1991; 83:997–1003.
15. Maden C, Beckman AM, Thomas DB, et al. Human papillomaviruses, Herpes simplex viruses, and the risk of oral cancer in men. Am J Epidemiol 1992; 135:1093–1102.
16. Abdelsayed RA. Study of human papillomavirus in oral epithelial dysplasia and epidermoid carcinoma in the absence of tobacco and alcohol use. Oral Surg Oral Med Oral Pathol 1991; 71:730–732.
17. Holladay EB, Gerald WL. Viral gene detection in oral neoplasms using the polymerase chain reaction. Am J Clin Pathol 1993; 100:36–40.
18. Melbye M, Palefsky J, Gonzales J, et al. Immune status as a determinant of human papillomavirus detection and its association with anal epithelial abnormalities. Int J Cancer 1990; 46:203–206.
19. Miller CS, White DK, Royse DD. In situ hybridization analysis of human papillomavirus in orofacial lesions using a consensus biotinylated probe. Am J Dermatopathol 1993; 15:256–259.
20. Coutlée F, Bobo L, Hawwari A, et al. Detection of HPV-16 in cell lines and cervical lavage specimens by a polymerase chain reaction-enzyme immunoassay assay. J Med Virol 1992; 37:22–29.
21. Bauer HM, Ting Y, Greer CE, et al. Genital human papillomavirus infection in female university students as determined by a PCR-based method. JAMA 1991; 265:472–477.
22. Hildesheim A, Schiffman MH, Gravitt PE, et al. Persistence of type-specific human papillomavirus infection among cytologically normal women. J Infect Dis 1994; 169:235–240.
23. Guerrero E, Daniel RW, Bosch FX, et al. Comparison of Virapap, Southern blot hybridization, and polymerase chain reaction methods for human papillomavirus identification in an epidemiological investigation of cervical cancer. J Clin Microbiol 1992; 30:2951–2959.
24. Coutlée F, Provencher D, Voyer H. Detection of human papillomavirus DNA in cervical lavage specimens by a non-isotopic consensus PCR assay. J Clin Microbiol 1995; 33:1973–1978.
25. Bauer HM, Greer CE, Manos MM. Determination of genital human papillomavirus infection by consensus polymerase chain reaction amplification. In: Herrington CS and McGee JOD, eds. Diagnostic Molecular Pathology, A Practical Approach. Oxford: IRL Press, 1992:131–152.
26. Schiffman MH, Bauer HM, Hoover RN, et al. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993; 85:958–963.
27. Jenison SA, Yu XP, Valentine JM, Koutsky LA, Christiansen AE, Galloway DA. Evidence of prevalent genital-type human papillomavirus infections in adults and children. J Infect Dis 1990; 162:60–69.
28. Jahal H, Sanders JH, Prime SS, Scully C, Maitland NJ. detection of human papillomavirus type 16 DNA in oral squames from normal young adults. J Oral Pathol Med 1992; 21:465–470.
29. Kellokoski JK, Syrjanen SM, Chang F, Yliskoski M, Syrjanen KJ. Southern blot hybridization and PCR in detection of oral human papillomavirus (HPV) infections in women with genital HPV infections. J Oral Pathol Med 1992; 21:459–464.
30. Maitland NJ, Cox MF, Lynas C, Prime SS, Meanwell CA, Scully C. Detection of human papillomavirus DNA in biopsies of human oral tissue. Br J Cancer 1987; 56:245–250.
31. Kashima HK, Kutcher M, Kessis T, Levin LS, de Villiers EM, Shah K. Human papillomavirus in squamous cell carcinoma, leukoplakia, lichen planus, and clinically normal epithelium of the oral cavity. Ann Otol Rhinol Laryngol 1990; 99:55–61.
32. Chang K-W, Chang C-S, Lai K-S, Chou M-J, Choo K-B. High prevalence of human papillomavirus infection and possible association with Betel Quid chewing and smoking in oral epidermoid carcinomas in Taiwan. J Med Virol 1994;28:57–61.
33. de Villiers EM, Wagner D, Schneider A, et al. Human papillomavirus DNA in women without and with cytological abnormalities: results of a 5-year follow-up study. Gynecol Oncol 1992; 44:33–39.
34. Meanwell CA, Cox M, Blackledge G, Maitland NJ. HPV 16 DNA in normal and malignant cervical epithelium: implications for the etiology and behaviour of cervical neoplasia. Lancet 1987; 1:703–707.
35. Kellokoski J, Syrjanen S, Syrjanen K, Yliskoski M. Oral mucosal changes in women with genital HPV infection. J Oral Pathol Med 1990; 19:142–148.
36. IARC (WHO). The Epidemiology of Human Papillomavirus and Cervical Cancer. 2nd ed. Munoz N, Bosch FX, Shah KV, Meheus A, eds. Lyon: IARC Publications #119, 1992:1–288.
37. Reeves WC, Brinton LA, Garcia M, et al. Human papillomavirus infection and cervical cancer in Latin America. New Engl J Med 1989; 320:1437–1441.
38. Kjaer SK, DeVilliers EM, Hauugaard BJ, et al. Human papillomavirus, herpes simplex virus and cervical cancer incidence in Greenland and Denmark. A population based cross-sectional study. Int J Cancer 1988; 41:518–524.
39. Villa LL, Franco ELF. Epidemiologic correlates of cervical neoplasia and risk of human papillomavirus infection in asymptomatic women in Brazil. J Natl Cancer Inst 1989; 81:332–340.
40. Franco E. Viral etiology of cervical cancer: a critique of the evidence. Rev Inf Dis 1991; 13:1195–1206.
41. Kreider JW, Howett MK, Stoler MH, Zaino RJ, Welsh P. Susceptibility of various human tissues to transformation in vivo with human papillomavirus type 11. Int J Cancer 1987; 39:459–465.
42. Burkett BJ, Peterson CM, Birch LM. The relationship between contraceptives, sexual practices, and cervical human papillomavirus infection among a college population. J Clin Epidemiol 1992; 45:1295–1302.
43. Zhang J, Rose BR, Thompson CH, et al. Associations between oncogenic human papillomaviruses and local invasive patterns in cervical cancer. Gynecol Oncol 1995; 57:170–177.
© Copyright 1997 American Sexually Transmitted Diseases Association
44. Syrjanen K, Vayrynen M, Castrnen O, et al. Sexual behavior of women with human papillomavirus (HPV) lesions of the uterine cervix. Br J Vener Dis 1984; 60:243–248.