Anal human papillomavirus infection in HIV-positive men and women at two opportunistic infections clinics in Harare, Zimbabwe

Anal cancer is rare in the general population but is significantly more prevalent in high-risk groups such as men who have sex with men (MSM) and those infected with human immunodeficiency virus (HIV). In up to 90% of cases anal cancer is linked to human papillomavirus (HPV) infection [1]. HIV predisposes an individual to multiple and often persistent anal HPV infections. There has been a downward trend in HIV prevalence in Zimbabwe thanks to aggressive efforts to manage the epidemic. However, the overall burden of HIV in the general population is still high, at 13% [2].

A substantial amount of data are now available from studies on anal HPV infections and consequent disease in MSM. There are numerous HPV genotypes and anal cancer in this population is most strongly linked to persistent HPV-16 and HPV-18 infection [3]. However data from other high-risk groups, in particular, African heterosexual males and females are relatively scant [4].The patterns of HPV infection, HPV clearance or progression to anal squamous intraepithelial lesions and cancer in this latter group are yet to be fully elucidated.

Anecdotal observation and crude cancer registry data suggest an upward trend in the number of cases of anal cancer in Zimbabwe in the past few years. Anal cancer screening is not routinely offered in Zimbabwe. Against this backdrop we sought to establish the prevalence of anal HPV infection in asymptomatic HIV-positive patients presenting at two Opportunistic Infections Clinics (OIC) in two large urban hospitals in Harare.

A cross-sectional study was performed on 152 consecutive patients (88 females; 64 males) attending OIC at Parirenyatwa Group of Hospitals and Harare Central Hospital between November 2014 and June 2015. Eleven patients opted not to be included in this study with the main reason given for refusal being lack of time. A total of 9000 adult patients are enrolled at the two OIC, with a male:female ratio of 1:1.5. The OIC serve a low to middle-income urban population of Harare, Zimbabwe. Patients visit the OIC for HIV testing and counselling services, routine follow up of HIV disease, and fulfilment of prescribed antiretroviral drugs. 85% of patients enrolled at the OIC are on highly active antiretroviral therapy (HAART). Ethical approval for this study was obtained from the University of Zimbabwe, College of Health Sciences Joint Research Ethics Committee and the Harare Central Hospital Ethics Committee. The study was registered with the Medical Research Council of Zimbabwe.

Males and females age ≥ 18 years were included in the study after providing informed signed consent. The source test document was checked to verify that each participant was HIV-positive. Demographic data and behavioural characteristics were collected through a face-to-face structured interview. CD4 counts and viral loads were verified by inspection of the participant’s clinical record.

Data from 122 patients were analysed after excluding 30 whose samples tested negative for beta-globulin; 73 (60%) were female and 49 (40%) were male. The mean age was 39.6 years (Range 18–69; SD 10.7). All patients were Black Africans. A summary of the study population is shown in Table 1.

Whilst we set out to probe for 14 high-risk and 15 low-risk HPV genotypes, we chose not to report results for genotypes 26, 70 and 82 (intermediate risk) and genotypes 32, 54, 55 and 61 (low-risk) due to lack of availability of working standards for those particular genotypes. These genotypes were therefore excluded from analyses and are not reported. Fifty-four of 122 samples tested HPV positive (44%). Of the 54 HPV-positive samples, 49 (91%) had specific HPV genotypes detected. Anal HPV was detected in 10/49 males (20%) and in 44/73 females (60%). High risk HPV types were detected in 25 patients (20%); low-risk HPV types were detected in 24 patients (20%). Nine patients (7%) had multiple high-risk HPV types i.e. more than one specific high-risk HPV type detected per sample. All high-risk HPV genotypes that we sought except HPV-39 were detected in this study population.

Table 2 summarises the prevalence and type distribution of HPV genotypes in the study sample.

The most commonly detected high-risk HPV genotypes were HPV-58 (13%), HPV-31 (11%) and HPV-16 (9%). HPV-18 was detected in 2.5% of samples. HPV-11 (17%) and HPV-53 (11%) were the two most commonly detected low-risk genotypes.

Figures 1 and 2 show the type-specific distribution of anal HPV infections in the study sample.

A positive test for any HPV was significantly associated with female gender (p < 0.0001). It was not significantly associated with marital status, HAART (highly active anti-retroviral therapy) use, cigarette smoking, alcohol use, number of sex partners, condom use, current CD4 count nor nadir CD4 count. For males and females combined, detection of any high-risk HPV was significantly associated with ≥5 lifetime partners (p = 0.021). When data were analysed for males and females separately, there was no change noted in the above associations. A summary of statistical associations between a positive HPV result and selected socio-demographic and immunological factors is shown in Table 3.

In this study of a largely heterosexual, HIV-positive population in Harare, Zimbabwe, the overall prevalence of anal HPV infection (high-risk and low-risk HPV genotypes) was 44%. This is the first paper describing prevalence of anal HPV infection in African heterosexual men, and one of the first in African women.

The anal HPV prevalence in females was three-fold higher than in men (60% vs. 20%). Most studies on females note that prevalence of anal HPV is actually higher than cervical HPV [6]. Putatively the cervix acts as a reservoir for HPV infections, allowing cross-infections with other sites such as the anus [7]. Simpson et al. suggest that in women, post-toilet front-to back wiping is a plausible biological mechanism for introduction of HPV into the anal canal [8]. There is likely under-reporting of anal sex in our study as this is a cultural taboo. Nonetheless, it is notable that the prevalence in males, though less than in females, is substantial, and suggests the importance of screening for anal cancer in both male and female HIV-positive patients regardless of self-reported sexual practices [9, 10].

High-risk HPV prevalence was 21% being found in 10% of males and 27% of females. Overall HPV-16 was found at a prevalence of 4% which is important because HPV-16 is one of the most consistent genotypes associated with anal cancer [11]. To date, however, studies on prevalence of HPV-16 in anal cancer samples have included no or very few samples, from Africa. In a recent meta-analysis by Lin et al. that included 2358 anal cancer samples, only 23 samples were from studies in Africa [12]. As more data become available it will become clearer how significant a role HPV-16 plays in anal cancer within the African population. An interesting finding in our data is that HPV-58 was the most common high-risk type detected. Other local and regional studies have also noted HPV-58 to be amongst the most frequently detected in the genital region [13, 14]. This may reflect different geographical distribution of HPV genotypes than reported elsewhere in the world. In addition, or alternatively, the frequent detection of HPV-58 may be a reflection of the relatively slow rate at which this particular genotype is cleared. Shvetsov et al. found that HPV-59 and 58 had the longest clearance times in a Hawaiian study of healthy women [15].

Overall, low-risk HPV infections were as common as high-risk infections in our study (20% vs. 21%). Though low-grade types are much less oncogenic by definition, they cause a significant burden of genital warts. The majority of these warts are attributable to infection by HPV-6 and 11 [16]. These are bothersome to the patient and difficult to treat [17]. This is an important consideration as Zimbabwe, through the Ministry of Health and Child Care, is in the process of rolling out a nationwide vaccination programme. The currently available vaccine is divalent, providing protection against HPV-16 and HPV-18 [18]. Ideally a vaccine protecting against HPV-6 and HPV-11 would reduce the future burden of genital warts.

There are several limitations to our results. The study was cross-sectional yet many HPV infections are transient [15]. We used the MY09/MY11 primer set for PCR and the dot blot method for HPV typing. These are older methods compared to newer technologies such as next-generation DNA sequencing. They are not able to detect all possible HPV types in the anogenital region but enabled us to describe the prevalence of the most important HPV types known to be associated with cervical and anal cancer and their precursor lesions. In addition we chose not to report on the prevalence of 7 other lower risk HPV genotypes (26, 32, 54, 55, 61, 70 and 82). Our overall HPV positivity remains unaffected because consensus primers were used initially, as do our results for the specific genotypes most commonly associated with anogenital cancer in Africa. Self-reporting of sexual practices may be unreliable especially where such practices are stigmatising. There is likely under-reporting of anal sexual intercourse in our study, especially by men, for this reason. Our study was set in two urban hospital clinics. This limits the generalisation of our findings to the rest of Zimbabwe where behavioural risk for sexually transmitted infections and access to health care may be different.

In conclusion the prevalence of anal HPV was 44% in this group of largely heterosexual HIV-positive males and females. Oncogenic HPV genotypes accounted for almost half of infections, supporting the need for surveillance of anal cancer in this population. These data represent a baseline upon which future studies may be built. Specifically, longitudinal studies can potentially reveal the pattern of persistence of anal HPV and chronicle the development of cytopathological changes over time.