Article Text


Human papillomavirus (HPV) infection in heterosexual South African men attending sexual health services: associations between HPV and HIV serostatus
  1. Etienne E Müller1,
  2. Tobias F Chirwa2,
  3. David A Lewis1,3,4
  1. 1Sexually Transmitted Infections Reference Centre, National Institute for Communicable Diseases of the National Health Laboratory Service, Johannesburg, South Africa
  2. 2Epidemiology and Biostatistics Division, School of Public Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
  3. 3Department of Internal Medicine, School of Clinical Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
  4. 4Division of Medical Microbiology, University of Cape Town, Cape Town, South Africa
  1. Correspondence to Dr Etienne E Müller, Sexually Transmitted Infections Reference Centre, National Institute for Communicable Diseases of the National Health Laboratory Service, Private Bag X4, Sandringham, Johannesburg 2131, South Africa; etiennem{at}


Objectives To investigate the distribution of human papillomavirus (HPV) genotypes and determine the associations between HPV infection and HIV coinfection in sexually active heterosexual men with anogenital warts (GW), male urethral discharge or asymptomatic men.

Methods Valid specimens for HPV genotyping were obtained from three patient groups consisting of 108 men with GW, 56 men with urethral discharge syndrome and 50 asymptomatic men attending for HIV voluntary counselling and testing. The Linear Array HPV Genotyping Test was used to determine the HPV genotype distribution among study participants. Sera were tested for HIV antibodies using two commercial rapid tests.

Results The prevalence of anogenital HPV among study participants was 78% (166). HPV DNA was detected in 100% (108) of GW, 48% (27) of men with urethral discharge syndrome and 62% (31) of voluntary counselling and testing participants. HPV types 6, 11, 16 and 18 were prevalent as either single or combined infections in 81% (134) of all HPV-positive study participants. HPV types 6 and/or 11 were significantly higher among GW patients (p<0.001). After adjusting for patient groups, HIV seropositivity was significantly associated with multiple HPV infections (OR=3.98, 95% CI 1.58 to 10.03) but not with the presence of a foreskin (OR=0.67, 95% CI 0.32 to 1.40).

Conclusions Infections with HPV were prevalent among sexually active heterosexual men attending the men's sexual health clinic. Associations were observed between HIV coinfection and multiple HPV infections. Further population-based studies on the prevalence of HPV genotypes are required to determine if men should be included in any future national HPV vaccination programme in South Africa.

  • HPV
  • HIV
  • anogenital warts
  • men
  • genital warts
  • heterosexuals

Statistics from


Men are key to the transmission of human papillomaviruses (HPVs), and treatment of clinically detectable lesions as well as subclinical disease has been advisable to prevent the spread of HPV infection to new sexual partners.1 Limited data exist on the incidence and transmission efficiency of HPV in heterosexual men, and few studies have focused on risk factors for genital HPV DNA detection in men.2–4 A number of researchers have focused their attention and research into the development and utilisation of an HPV vaccine for men to eliminate cervical cancer in women, and to reduce the rate of anal cancer and genital warts in men.5–9 Within South Africa, HPV-16 is the most dominant HPV type in women with cervical cancer (57%), followed by HPV-18 (11%) and HPV-33 (9%).10 In men, HPV types 16 and 18 are responsible for at least 70% of anal cancers and precancerous lesions of the penis, while HPV types 6 and 11 cause more than 90% of genital warts.11 The quadrivalent HPV vaccine will target HPV types 6, 11, 16 and 18.12

HIV infection is a risk factor for HPV infection and subsequent development of invasive cancers in both men and women. Infection with HIV causes immune dysfunction, leading to increased susceptibility to HPV infection, HPV persistence and both development and progression of squamous intraepithelial lesions.13 14 Although the mechanism by which HIV changes the pathogenesis of HPV-associated neoplasia is unclear, the interaction may possibly be through HPV genes, effects on immune functions or altered transcription of the HPV L1 and E1 proteins.15 In HIV-negative individuals, most HPV infections are self-limited, such that only small percentages of patients (2–3%) develop dysplasia, despite the much greater prevalence of asymptomatic HPV infection.16

At present, there are few HPV type specific distribution data available for South African men. This study aims at determining the genotypic distribution of HPV and the associations with HIV coinfection, the presence of multiple HPV types and circumcision status among men presenting with anogenital lesions, urethritis and among asymptomatic men attending a men's sexual health clinic in Johannesburg, South Africa.


Study population

Sexually active heterosexual men (18 years or older), with anogenital warts (GW) (n=112) or urethral discharge (MUS) (n=64), or asymptomatic men attending for HIV voluntary counselling and testing (VCT) (n=63), who were willing to provide specimens for delinked HPV screening and typing, were consecutively recruited between December 2006 and July 2008 at a weekly men's sexual health clinic in Alexandra, Johannesburg. Unpublished data on clinic attendance for the 2-year period 2007–2008 show that GW and MUS accounted for 37% and 17% of 366 new consultations, respectively (DA Lewis, personal communication). The men's STI clinic is staffed by an experienced and well-trained clinical team with cryotherapy facilities for genital wart treatment (no treatment at all exists for genital warts in the PHC setting in Johannesburg). The clinic was situated in Alexandra, which is known to have a high STI/HIV burden based on annual STI microbiological surveillance data (DA. Lewis, personal communication). Men who used condoms regularly (>75% of time) were excluded from the MUS and VCT groups. Only data from participants with valid HPV genotyping results were analysed. A nurse-administered questionnaire was completed by all study participants. The questionnaire collected basic demographic data, clinical data and examination findings. With an estimated HPV prevalence of 95%, 45% and 60% for the GW, MUS and VCT groups, an interim 90% power calculation with 95% confidence required recruitment of a minimum of 44 participants in the MUS and VCT groups.

Sample collection

Dry cotton swabs were collected from participants. Each swab sampled the glans penis, the coronal sulcus, the penile shaft and, if present, anogenital warts. The skin of the perianal area was not sampled in the absence of visible perianal warts. Swab specimens were transported to the laboratory, kept at 4°C and processed within 24 h after sample collection at the clinic. In addition, 5 ml of blood was obtained from all consenting participants for storage and subsequent delinked serological testing. Swabs and sera were labelled with a unique study number which was linked to questionnaire data but delinked from participant details. A separate written informed consent to store sera and swab samples for future research work was obtained from participants.

Patient management

STI syndromes were treated free of charge according to national syndromic management algorithms. Genital warts were treated by cryotherapy and topical application of 20% podophyllin. All participants were offered on-site VCT for HIV at the time of clinical presentation.

Laboratory analyses

HPV detection and genotyping

The cotton swabs were transferred into individual tubes containing 500 μl of nuclease-free water and vortexed for 1 min. Aliquots of these suspensions were used for nucleic acid extraction using the MagNA Pure Compact nucleic acid extractor (Roche Diagnostics, Mannheim, Germany). An extraction negative control was included with each run. After DNA extraction, the Linear Array (LA) HPV Genotyping Test (Roche Molecular Systems, Branchburg, New Jersey) was used to determine the HPV genotype distribution among all study participants. The LA test amplifies the target HPV DNA for 37 anogenital HPV genotypes and include 24 low-risk HPV types (6, 11, 26, 40, 42, 53, 54, 55, 61, 62, 64, 66, 67, 69, 70, 71, 72, 73, 81, 82, 83, 84, IS39 and CP6108) and 13 high-risk HPV types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68).17 HPV-52 was only recorded positive in the absence of HPV types 33, 35 and 58 due to the combined probe used for these four types in the LA assay. The β-globin gene was amplified as a control for cell adequacy, extraction and amplification. Taking into account the manufacturers' instructions, 10 samples with a negative β-globin result and a positive HPV DNA result were considered valid and adequate for analyses, although additional genotypes might have been missed.

HIV antibody testing

After all participants had been recruited, stored sera were screened in a delinked manner for HIV antibodies using two qualitative immunochromatographic tests (Determine HIV-1/2 and Uni-Gold HIV) according to the manufacturers' instructions.

Statistical analyses

Data were analysed using the Epi-Info V.3.3.2 and STATA V.10.0 in order to study the differences in the proportions and types of HPV and HIV antibodies detected in the three patient groups (GW, MUS and VCT). A χ2 test of association was used to explore the association between HPV and HIV coinfection, and logistic regression was used to adjust for potential confounders. Only factors associated with both HPV and HIV were considered as potential confounders. Statistical tests were carried out at the 5% significance level.



A total of 239 participants were enrolled, including 112 men with anogenital warts, 64 men with MUS and 63 asymptomatic men presenting for HIV VCT. Of 239 participants recruited, samples from 214 men (89.5%) produced valid HPV genotyping results. A total of 25 (10.5%) samples were both HPV DNA and β-globin negative, and were excluded from data analysis.

Participant demographics and clinical features

Participants had a mean age of 29.8 years (SD 7.50) with no variation in age by patient group (table 1). All were of Black African ethnicity, and 156 patients (72.9%) were single with a girlfriend. The median number of days since the last episode of sexual intercourse for all participants was 12 days (IQR 5 to 28) with 16.5 days (IQR 7 to 38) for GW; 7 days (IQR 5 to 13) for MUS and 8.5 days (IQR 4 to 24) for VCT. Only 81 (37.9%) of the 214 participants were aware of their HIV status. Of these, 35 (43.2%) self-reported a positive HIV serostatus, 40 (49.4%) indicated a negative HIV serostatus, and six participants (7.4%) declined to give any information. The actual HIV prevalence was 49.5% (104) among all study participants (71.7% for GW, 35.7% for MUS and 16.7% for VCT). On physical examination, 54 (26.0%) of all participants were circumcised (GW 20.0%; MUS 23.2%; VCT 42.6%). Anatomically, most warts (single and/or multiple warts) were present on the prepuce (57.4%), the shaft (37.0%) and the glans penis (30.6%) (data not shown).

Table 1

Human papillomavirus (HPV) DNA and HIV positivity in relation to characteristics of all study participants (n=214)

HPV genotype distribution

Results for HPV genotype distribution among GW, MUS and VCT participants are presented in table 2. The overall prevalence of anogenital HPV among participants was 77.6% (166). All participants presenting with GW were positive for HPV (108/108; 100%). Approximately half of MUS participants (27/56; 48.2%) and almost two-thirds of VCT participants (31/50; 62.0%) were HPV-positive. On univariate analysis, any HPV type, and HPV types 6 and/or 11, were significantly associated with the GW patient group (p<0.001). There was no significant association between patient group and the detection of HPV types 16 and/or 18.

Table 2

High- and low-risk human papillomavirus (HPV) type profiles in HPV-DNA-positive participants

The 166 HPV positive participants were mostly infected with a combination of high- and low-risk HPV genotypes (68.5% for GW; 29.6% for MUS and 54.8% for VCT). Infection with high-risk HPV genotypes alone was uncommon in all groups (GW 0%; MUS 18.5%; VCT 9.7%). Multiple HPV infections were observed in 81.3% of all HPV-positive participants (88.0% for GW; 55.6% for MUS; 80.7% for VCT).

As expected, HPV types 6 and/or 11 were the most prevalent genotypes among HPV-positive GW participants (104/108; 96.3%) and occurred less frequently in HPV-positive MUS (5/27; 18.5%) and VCT (11/31; 35.5%) participants. High-risk HPV types 16 and/or 18 were observed in about a quarter of all HPV-positive individuals (46/166; 27.7%); the prevalence was similar in all three participant groups (table 2). Either one infection or a combination of infections with HPV types 6, 11, 16 and 18 was present in 106/108 (98.1%) of GW, 11/27 (40.7%) of MUS and 17/31 (54.8%) of VCT participants. HPV genotype 64 was the only genotype not detected in any of the study participants.

HIV coinfections

HIV coinfection was detected in 104 (49.5%) of 210 study participants (table 1); four participants declined permission to store their sera, and accordingly their HIV status could not be determined. The GW group had the highest HIV-positivity rate (71.7% (76)), followed by MUS (35.7% (20)) and VCT (16.7% (8)). On univariate analysis, HPV was significantly associated with HIV, with 90.4% among HIV-positive participants and 65.1% in the HIV-negative group being HPV-positive (OR=5.04, 95% CI 2.35 to 10.83). Multiple HPV infections were significantly associated with HIV coinfection in that 90.4% of HIV-positive men and 68.1% of HIV-negative men coinfected with HPV had more than one HPV type (OR=4.42, 95% CI 1.88 to 10.38) (table 3).

Table 3

Unadjusted and adjusted OR estimates (with 95% CI) from logistic regression for associations between human papillomavirus (HPV), HIV and multiple HPV and exposure risk factors adjusting for confounders

After adjusting for patient groups in a logistic regression, the association between multiple HPV and HIV remained significant with the odds of an HIV-positive individual having multiple HPV types being about four times (OR=3.98, 95% CI 1.58 to 10.03) higher than an HIV-negative person. However, the adjusted association between HPV and HIV was not significant (OR=2.07, 95% CI 0.81 to 5.28) (table 3).

Association of circumcision with HPV detection and HIV serostatus

Clinician-observed data on circumcision were recorded for 208/214 participants (table 1). Fifty-four men (26.0%) were circumcised. The presence of a foreskin was not associated with HR-HPV positivity (OR=0.69, 95% CI 0.37 to 1.28) but was associated with LR-HPV positivity (OR=0.41, 95% CI 0.21 to 0.80) and overall HPV positivity with 81.2% (125) in uncircumcised men and 64.8% (35) among circumcised men (OR=0.43, 95% CI 0.21 to 0.85). The results were statistically insignificant after adjusting for HIV status and patient group, (OR=0.51, 95% CI 0.21 to 1.25).

Among all participants, the unadjusted results show that there was an association between being HIV-seropositive and having an intact foreskin (OR=0.50, 95% CI 0.26 to 0.94). After adjusting for patient group and HPV, the relationship was not statistically significant with an OR of 0.67 (95% CI 0.32 to 1.40) showing that those who were circumcised were 0.67 times less likely to be HIV-positive than those who were not.


In this paper, we report the HPV prevalence and genotype distribution among a group of sexually active heterosexual men with clinically apparent anogenital warts, urethritis and asymptomatic individuals attending a weekly men's sexual health clinic in Alexandra, Johannesburg. At present, no specific ‘gold standard’ exists for sampling HPV DNA at a specific genital site in men.18 A wide range of sampling methods is described in the literature. Most sampling methods involved rubbing/rotating of a dry or wet swab on the genital epithelium. Some studies increased their HPV detection rates by rubbing the sampling site with a small piece of emery paper before sampling with a swab.19 In previous studies, demonstrating HPV distribution in different anogenital sites in asymptomatic men, it was shown that the external genital samples, especially the penile shaft, prepuce, glans penis, coronal sulcus and scrotum were more likely than anal, urethral or meatal samples to contain HPV DNA.4 20 21 Approximately 10% of swab specimens were excluded from the current study due to failure of both HPV and β-globin gene amplification, suggesting that insufficient cellular material was collected on the swab.

Dunne et al conducted a systematic review of the literature regarding the prevalence of HPV infection in men, and they reported HPV prevalence rates between 1.3% and 72.9% in studies in which multiple anatomical sites were evaluated.4 An overall HPV prevalence of 77.6% was observed in our study population. As expected, all participants presenting with GW were HPV DNA-positive, and approximately half of the MUS and VCT control groups were HPV DNA-positive. These results are in agreement with data from previous studies suggesting very high HPV positivity rates in GW patients and prevalence rates between 44% and 51% in patients who had no history of genital warts.22 23

The high prevalence of HPV types 6/11/16/18 (80.7%) among the study participants emphasises the importance of conducting more HPV studies among men, especially those in the general population, prior to introducing national HPV vaccination programmes. In addition to impacting on cervical dysplasia and cancer, the currently available quadrivalent HPV vaccine could also potentially decrease the socio-economic burden of anogenital warts in South Africa. Vaccination of men with the quadrivalent HPV vaccine is not currently recommended. The strongest case for vaccinating men is the establishment of herd immunity; however, if all women were immunised with the HPV vaccine, some argue that there may be a limited benefit in vaccinating men.24 Arguments against HPV vaccination of boys or young men include the absence of efficacy data of the vaccine in preventing anal, penile and oral/pharyngeal cancers among men, the difficulties in attracting men to health services and the cost of implementing such an intervention.24 25

HIV-positive men are more likely to be infected with HPV and develop intraepithelial lesions than their HIV-negative equivalent.13 Linked AIDS and cancer registry data from the USA have highlighted the fact that HIV-positive men have a 37.9-fold higher risk of anal cancer and a 3.7-fold higher risk of penile cancer than the general population.13 Several mechanisms may be involved in the increased prevalence of HPV and HPV-associated disease among HIV-positive individuals, including HIV-induced immunodeficiency, modulation of the immune response to HPV, the local immune response at tissue level, genetic instability and the interaction between the two viruses at molecular level.13 16 26 27

A number of cross-sectional studies have demonstrated that male circumcision is associated with a lower risk of HPV infection.3 21 22 Among young men enrolled in a circumcision trial in South Africa, circumcision has been shown to have a partial protective effect on the prevalence of HR-HPV.28 The present study showed an association between HPV detection and circumcision status, although this relationship was not significant after adjusting for HIV serostatus and patient group. The adjusted analyses also revealed no association between multiple HPV infections and the presence of a foreskin. Significant results from the unadjusted analysis are in line with published literature. However, the non-significant results after adjusted analyses should be treated with caution, as this could partly be attributed to colinearity of factors, especially with patient group, and probably small disaggregated numbers.

In conclusion, this study within a men's sexual health clinic has demonstrated a high prevalence of HPV types 6, 11, 16 and 18, as well as HIV infection, among men presenting with genital warts, urethritis or HIV VCT (asymptomatic men). Multiple HPV infection was associated with HIV coinfection but not with the presence of a foreskin.

Key messages

  • The burden of HPV infection was high among men attending a primary healthcare-based men's sexual health clinic within Alexandra, Johannesburg.

  • Non-oncogenic HPV types 6 and 11 were associated with genital warts.

  • Oncogenic HPV types 16 and 18 were detected in about a quarter of men attending the men's sexual health clinic, regardless of the presence or absence of genital warts.

  • Multiple HPV types were more prevalent in HIV seropositive men.


The authors would like to thank O Mohlamonyane, R Ilunga and A Vezi for their assistance with data collection and sampling at the men's sexual health clinic, as well as the staff of the 8th Avenue clinic in Alexandra and N Mehana and her colleagues in the City of Johannesburg Health Department for their support of this study.


View Abstract


  • Funding This research was internally funded at the National Institute for Communicable Diseases, National Health Laboratory Service, South Africa.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (Protocol nos M060434 and M070912).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Whistlestop tour
    Jackie A Cassell