Prevalence, Incidence, and Clearance of Anogenital Warts in Kenyan Men Reporting High-Risk Sexual Behavior, Including Men Who Have Sex With Men

Background. Human papillomavirus (HPV) causes a spectrum of disease, ranging from warts to cancer. Prevalence, incidence, and factors associated with anogenital warts in East African men are unknown. Methods. Kenyan men reporting high-risk sexual behavior were inspected for anogenital warts at enrollment and follow-up visits. Logistic regression was performed to identify associations with anogenital warts at baseline. Cox regression was performed to analyze predictors of incident anogenital warts, and Kaplan–Meier curves were used to estimate clearance. Results. Baseline anogenital wart prevalence in 1137 men was 2.9% (95% confidence interval [CI], 2.0%–4.0%) overall, 2.0% in human immunodeficiency virus (HIV)-uninfected men, and 9.4% in HIV-1-infected men (adjusted odds ratio, 5.43; 95% CI, 2.03–11.29). Over a median of 1.4 years, anogenital wart incidence among 1104 men was 5.3 (95% CI, 4.3–6.5) per 100 person-years. Having HIV-1 infection at baseline (adjusted hazard ratio [aHR], 1.66; 95% CI, 1.01–2.72) or a genital syndrome during follow-up (aHR, 4.78; 95% CI, 3.03–7.56) was associated with increased wart incidence. Wart clearance was lower in HIV-1-infected men (log-rank P<.001). Conclusions. Anogenital wart prevalence and incidence were increased in HIV-1-infected men, and anogenital warts co-occurred with other genital syndromes. Quadrivalent HPV vaccination should be recommended for young men in settings with high HIV-1 prevalence.

objective was to describe the prevalence, incidence, and factors associated with anogenital warts in a cohort of Kenyan men reporting high-risk sexual behavior, including male sex workers and men who have sex with men (MSM). We included both HIV-1-infected and uninfected men in the study, with the aim to estimate anogenital wart prevalence, incidence, and clearance in these 2 groups.

Study Population
In July 2005, the Kenya Medical Research Institute (KEMRI) Human Immunodeficiency Virus/Sexually Transmitted Disease clinic opened in Mtwapa, Kenya (approximately 13 miles north of the coastal city of Mombasa), as a collaborative research site for an HIV-1 vaccine feasibility study. Since that time, HIV-1uninfected persons reporting high-risk sexual behavior (defined as having sex with men, sex in exchange for money, recent sexually transmitted infections, serodiscordant sex partners, or multiple sex partners) have been prospectively observed and monitored for HIV-1 seroconversion. In February 2006, the clinic added a cohort study to evaluate the risk behavior and health needs of HIV-1-infected persons from the same highrisk groups. Participants were recruited through community outreach, local voluntary counseling and testing sites, and links with local lesbian, gay, bisexual, and transgender groups. Men who reported any anal sex with another man within the last 3 months were classified as MSM. All participating men, regardless of reported sexual activity, orientation, or preference, were included in the present study. Human immunodeficiency virus-1-uninfected men who acquired HIV-1 during follow-up were censored at the estimated date of infection, because data on anogenital wart detection were not available after this time point.

Clinic Procedures
Eligible individuals were invited to receive information about participation in research and alternatives for care, including several nonresearch HIV clinics in the study area. Upon enrollment, sociodemographic and sexual behavior information was obtained with a standardized questionnaire, blood was collected for HIV-1 screening, and appropriate counseling was provided [11].
Men who reported receptive anal intercourse had monthly follow-up visits; all other participants were seen quarterly. At each follow-up visit, seronegative participants underwent HIV counseling and testing and seropositive participants underwent risk reduction counseling. Trained clinical officers obtained a medical history, performed a standardized physical examination and sexually transmitted infection (STI) screening, and provided care as indicated. Syndromic STI treatment and antiretroviral therapy, when indicated, were provided in accordance with World Health Organization and Kenyan guidelines [12]. Podophyllin was the only treatment available for warts. Participants were reimbursed KES 350 (approximately $3.70) per visit.
Trained clinical officers identified anogenital warts by visual inspection of the external genitalia. Visual inspection of the perianal area was also performed if the participant reported receptive anal sex or anorectal symptoms. Genital and perianal warts were recorded separately and included both verrucous and flat lesions. Starting in September 2006, participants who reported receptive anal sex or experienced anal symptoms were offered proctoscopy to screen for rectal lesions or discharge.

Diagnostic Procedures
Two rapid HIV-1 test kits (Determine, Abbott Laboratories; Unigold Recombigen HIV, Trinity Biotech PLC, Ireland) were used for HIV-1 testing. Discrepant results were verified with a third-generation enzyme-linked immunosorbent assay test (Genetic System HIV-1/2 plus O EIA; Bio-Rad Laboratories, Redmond, WA). Genital syndromes diagnosed included urethritis (defined as 5 or more polymorphonuclear cells [PMNs] per high-power field on a Gram-stained urethral smear), proctitis (defined as 5 or more PMNs per high-power field on a Gram-stained rectal smear), or genital ulceration (any ulceration, either penile or perianal, on physical examination). Detection of Gram-negative, intracellular diplococci in either the urethra or rectum was used for diagnosis of gonococcal infection. All patients with laboratory-diagnosed STIs were traced and treated.

Statistical Methods
Descriptive statistics were used to summarize baseline characteristics of the study population at cohort enrollment. The prevalence of anogenital warts in HIV-1-infected and HIV-1-uninfected men was calculated, with 95% confidence intervals (CIs), using exact binomial methods. Logistic regression was used to identify independent associations between prevalent anogenital warts (ie, the primary outcome) and potential predictors, before and after adjustment for potential confounders that were associated with anogenital warts in bivariate analysis at P < .10 [13]. We included HIV status and soap use as a priori predictors in multivariable analysis. Incidence rates with 95% CI were calculated using the quadratic approximation to the Poisson log likelihood for the log-rate parameter, both in the entire population and within predictor categories. Kaplan-Meier estimates were obtained, and the log-rank test was used to test for differences in risk of anogenital warts by HIV-1 status. Cox regression was then performed to analyze predictors of incident warts, with and without adjustment for potential confounders using the same model-building approach as above. Anogenital wart clearance (ie, no visible warts after previous detection) rates with 95% CI were calculated in the subgroup of men who had anogenital warts identified at enrollment or during follow-up. Kaplan-Meier estimates were obtained, and the log-rank test was used to test for differences in anogenital wart clearance by HIV-1 status. Stata/IC 11.2 was used.

Ethics
Written, informed consent was obtained from all study participants, following procedures approved by the ethical review committees of the KEMRI ( protocols 894 and 1224) and University of Washington ( protocol 28292).

Population
The study was conducted between January 2005 and June 2013. Sociodemographic and other characteristics of the 1137 men in the study population are presented in Table 1. The majority of participants (59.4%) reported having sex with both men and women, 25.1% reported having sex with women only, and 15.5% reported having sex with men only. Only 26% reported  b Sex work was defined as having received cash, living expenses, or goods in exchange for sex. c Defined as having a diagnosis of urethritis (defined as 5 or more PMNs per high-power field on a Gram-stained urethral smear), proctitis (defined as 5 or more PMNs per high-power field on a Gram-stained rectal smear), or genital ulceration (any ulceration, either penile or perianal, on physical exam). formal employment, and 58% reported having been paid for sex in the past 3 months. Human immunodeficiency virus-1 prevalence at baseline was 12.1%.

Prevalence and Risk Factors for Anogenital Warts
At enrollment, 33 of the 1137 participants were diagnosed with 1 or more anogenital warts for an estimated wart prevalence of 2.9% (95% CI, 2.0%-4.0%). Twelve men had 1 or more penile warts, of whom 6 reported sex with both men and women and 6 reported sex with women exclusively. Twenty-one men had 1 or more perianal warts, of whom 17 reported sex with both women and men and 4 reported sex with men exclusively. Anogenital wart prevalence was 2.1% for men who reported sex with women exclusively, 2.3% for men who reported sex with men exclusively, and 3.4% for men who reported sex with both men and women. In bivariate analysis, prevalent HIV-1 infection was the only variable associated with an increased prevalence of anogenital warts, with 9.4% of HIV-1-infected versus 2.0% of uninfected men having warts (odds ratio [OR], 5.09; 95% CI, 2.47-10.49; P < .001) ( Table 2). Anogenital wart point prevalence was slightly lower among men who used soap for genital washing than those who did not (2.8% vs 3.8%), and increasing frequency of soap use showed an inverse relationship with anogenital wart prevalence (see Table 2); however, these differences were not statistically significant. In multivariate analysis including HIV-1 status and soap use as a priori predictors, only prevalent HIV-1 infection was associated with an increased anogenital wart prevalence (adjusted OR [aOR], 5.43; 95% CI, 2.03-11.29; P < .001).

Anogenital Wart Incidence and Risk Factors
The 33 men with prevalent anogenital warts on their initial exam were excluded from the incidence analysis, leaving 1104 men at risk for incident anogenital warts. Overall, follow-up time for this group was 1639 person-years (PY), with a median of 1.4 years (range, 0.03-7.59 years) and a median of 6 visits. We identified 87 men with incident anogenital warts, for an incidence rate of 5.3 per 100 PY (95% CI, 4.3-6.5 per 100 PY). Fifty men had 1 or more penile warts, of whom 34 reported sex with both men and women, 8 reported sex with men exclusively, and 8 reported sex with women exclusively. Fifty-men had 1 or more perianal warts, of whom 41 reported sex with both men and women, 12 reported sex with men exclusively, and 1 reported sex with women exclusively. A Kaplan-Meier plot of time to detection of anogenital warts (combining both penile and perianal) by HIV-1 status is shown in Figure 1A. Anogenital wart incidence in HIV-1-infected men was 7.6 per 100 PY vs 4.9 per 100 PY in uninfected men (log-rank P = .009). b Defined as having a diagnosis of urethritis (defined as 5 or more PMNs per high-power field on a Gram-stained urethral smear), proctitis (defined as 5 or more PMNs per high-power field on a Gram-stained rectal smear), or genital ulceration (any ulceration, either penile or perianal, on physical exam). c Sex work was defined as having received cash, living expenses, or goods in exchange for sex.

Anogenital Wart Clearance and Human Immunodeficiency Virus-1 Status
Among 120 men diagnosed with 1 or more anogenital warts, 106 had documented clearance at a subsequent visit. Median time to documented clearance was 2.7 months (interquartile range, 1.5-3.2 months). The estimated rate of clearance was 166.7 per 100 PY (95% CI, 137.7-201.8 per 100 PY). A Kaplan-Meier plot of time to anogenital wart clearance by HIV-1 status is shown in Figure 1B. Anogenital wart clearance in HIV-1-infected men was 96.7 per 100 PY vs 253.4 per 100 PY in HIV-1-uninfected men (log-rank P < .001).

DISCUSSION
To our knowledge, this is the first study to report the prevalence and incidence of anogenital warts in a population of MSM and other high-risk men in Africa. We found a prevalence of 2.9% and incidence of 5.3 per 100 PY; both were substantially higher among HIV-1-infected men than HIV-1-uninfected men. Men diagnosed with a genital syndrome such as urethritis, proctitis, or genital ulceration were also at increased risk of an anogenital wart at the same visit. Soap use was not associated with decreased anogenital prevalence and incidence; however, this may have been due to the relatively small number of events, which could have limited our power to detect an association.
Our data fall within the range of global prevalence of anogenital warts (0.1%-5.1%) estimated in a 2013 systematic review by Patel et al [14]. The prevalence of anogenital warts we found in high-risk men in Kenya is similar to data from urban North American men. For example, the prevalence of genital warts in several urban US cities ranged from 2.9% to 9.2% for MSM and from 2.6% to 7.2% for men who have sex with women [15].
The anogenital wart incidence estimated in our cohort (5.3 per 100 PY) is almost 4 times higher than the anogenital wart incidence found in a cohort of male trucking company employees in Mombasa, Kenya (1.4 per 100 PY) and 23 times higher than the incidence rate of genital warts found in a multinational (United States, Mexico, Brazil) cohort of men in the general population (0.23 per 100 PY) [10,16]. Of note, both studies excluded HIV-1-infected men. The very high overall incidence in our study even among HIV-1-uninfected men could also be due to a relatively young study population, close follow-up intervals, and frequently reported high-risk sexual behavior, including unprotected receptive anal intercourse.  In our cohort, HIV-1-infected men had higher anogenital wart prevalence and incidence compared with uninfected men, with minimal change in the magnitude of risk after adjusting for potential confounders. These results are consistent with the findings of several studies describing higher rates of HPV infection in HIV-1-infected individuals in populations across the globe [16][17][18][19][20], including a systematic review that found a relative risk (RR) similar to ours (1.62 vs 1.66) [21]. In a South African study, HIV-1-infected men had an increased risk of new HPV detection (any type) compared with uninfected men (RR, 2.00; 95% CI, 1.49-2.69) [22]. In addition, HIV-1infected men had a reduced rate of HPV viral clearance relative to HIV-1-uninfected men (RR, 0.71; 95% CI, .55-.93) [22]. Human papillomavirus infection tends to persist and reactivate more frequently in HIV-infected individuals than in uninfected individuals [22]. In the present study, HIV-1 infection was associated with a decreased rate of anogenital wart clearance. Although our study is not the first to report increased anogenital wart prevalence in HIV-1-infected men, we were unable to identify any previous studies reporting anogenital wart incidence or clearance in HIV-1-infected men for comparison.
In agreement with previous studies [23], we found a significant association between incident anogenital warts and having another genital syndrome, including urethritis, proctitis, and genital ulcer disease, even after adjustment for potential confounders. Although we did not detect a significant association between prevalent anogenital warts and being diagnosed with a genital syndrome, this finding could be due to the small number of anogenital wart cases detected at enrollment or to the relatively low sensitivity of the diagnostics used for STI testing in the present study.
Because poor hygiene has been associated with increased HPV prevalence, we analyzed associations between the use of soap for genital washing and both anogenital wart prevalence and incidence [7,24,25]. Although we found that point estimates of the OR for prevalent anogenital warts and the HR for incident anogenital warts were less than 1, no statistically significant association was found. This is in contrast to our finding, in the same Kenyan cohort of men reporting high-risk sexual behavior, that the use of soap for genital washing was associated with a decreased risk of herpes simplex virus type 2 acquisition (adjusted incidence rate ratio, 0.3; 95% CI, .1-0.8) [26]. Additional study of the impact of hygiene on HPV Defined as having a diagnosis of urethritis (defined as 5 or more PMNs per high-power field on a Gram-stained urethral smear), proctitis (defined as 5 or more PMNs per high-power field on a Gram-stained rectal smear), or genital ulceration (any ulceration, either penile or perianal, on physical exam).
c Sex work was defined as having received cash, living expenses, or goods in exchange for sex. viral persistence and clearance could help determine whether hygiene practices could potentially reduce the incidence of anogenital warts in this study population. Anogenital warts are not a mere nuisance, but they can cause considerable psychological distress and morbidity [1][2][3]. Warts disrupt the epithelial surface and can bleed, occasionally becoming superinfected. Moreover, anogenital warts have been associated with infection by multiple HPV types (including high-risk types) and, as a result, may be associated with a long-term increased risk of anogenital cancer [27,28]. Although it is unclear whether anogenital warts increase the risk for HIV-1 acquisition per se, 1 Kenyan study reported an increased risk of risk of HIV-1 acquisition among Kenyan men infected with HPV [29]. Other studies are currently examining the impact of genital warts on HIV incidence [30]. The quadrivalent HPV vaccine (HPV6/11/16/18) has demonstrated efficacy for the prevention of anogenital warts and anal cancer in men [9]. As evidence of the burden of HPV-related disease in African men accrues, the case for vaccination of both men and women in this region will become more and more compelling.
Strengths of this study include the longitudinal study design, the long duration of follow-up, close monitoring with monthly or quarterly visits, and a standardized physical exam including inspection for both genital and perianal warts. Our study had several limitations. First, the relatively low number of events limited our power to detect potentially important associations. In order to increase power, we created a combined variable, anogenital warts, including both genital and perianal warts. In doing this, our ability to focus on a single anatomic site was reduced. Second, we used visual inspection as our method for anogenital wart detection. Without histologic testing, it is possible that other conditions (eg, penile intraepithelial neoplasia) were misclassified as warts. However, visual inspection alone is the most frequently used method for genital wart diagnosis in several parts of the world. Finally, we were unable to test for HPV DNA and subtypes, including HPV 6/11, an important predictor of the development of warts.

CONCLUSIONS
This population of men reporting high-risk sexual behaviors had a moderate prevalence of anogenital warts and one of the highest rates of anogenital wart acquisition ever described. We have confirmed the association between anogenital wart prevalence and HIV-1 infection and estimated a 1.7-fold increase in the risk of anogenital wart acquisition among HIV-1-infected men. We have also confirmed an association between incident anogenital wart and genital syndromes other than warts. Genital wart disease negatively impacts the quality of life of men worldwide. Treatment of persistent genital warts can be challenging, and treatment is unavailable in many resource-limited settings. The quadrivalent HPV vaccination, which has demonstrated efficacy for the prevention of anogenital warts and anal cancer in men, should be considered for young men in East Africa and in other settings with high HIV-1 prevalence.