Human Papillomavirus Prevalence and Type Distribution in 3603 HIV-Positive and HIV-Negative Women in the General Population of Tanzania

The PROTECT Study

Abstract

Objective: 

The aim of the Prevention of Cervical Cancer in Tanzania (PROTECT) study is to assess the prevalence of oncogenic human papillomavirus (HPV) and to determine the type distribution among women in the general population according to human immunodeficiency virus (HIV) status, in preparation for a potential HPV immunization program.

Methods: 

We included 3603 women from the general population in urban and rural areas of Tanzania. All women underwent a gynecological examination where a Pap smear was obtained and cervical cells were collected to assess the presence of high-risk (HR) HPV DNA by hybrid capture 2 test. Genotyping was performed by the LiPaExtra method. These women were also tested for HIV.

Results: 

The prevalence of HR HPV types was 20.1%, ranging from 14.8% in women with normal cytology to 94.2% in women with high grade squamous intraepithelial lesion (HSIL) or worse (100% in 5 cancers). In women with normal cytology or low-grade lesions, the most common type was HPV52 (3.2%), followed by HPV16 (2.1%). In contrast, HPV16 was the dominating type in HSIL or worse (32.8%). No cancers contained HPV52. The HR HPV prevalence was higher in HIV-positive women (46.7%) than in HIV-negative women (17.2%). No specific HR HPV types were significantly more common in HIV-positive women.

Conclusion: 

The HPV type distribution is similar in HIV-positive and HIV-negative women. These results suggest that the HPV vaccines that are currently available could protect women from HPV infection independently of their HIV status.

In Brief

This study of human papillomavirus (HPV) among women in Tanzania showed that the type distribution is similar, regardless of human immunodeficiency virus status, suggesting that HPV vaccination could protect all women from HPV infection.

Human papillomavirus (HPV) infection is one of the most common sexually transmitted infections; about 80% of all sexually active persons will be infected at some time during their life. Persistent high-risk (HR) HPV infections are known to be a necessary first step in the development of cervical cancer.¹ The prevalence of HPV in different populations varies according to a variety of factors, notably age and sexual behavior.

In Tanzania, cervical cancer is the most common type of cancer in women, and this combined with the highest mortality rate and an estimate from 2007 showed that approximately 7500 new cases are diagnosed every year and about 6000 women die because of this cancer yearly.² Screening for precursor lesions has decreased the incidence of cervical cancer in many countries. In Tanzania, a screening program was initiated at the Ocean Road Cancer Institute (Dar es Salaam) in 2002 and is slowly being scaled up to cover several regions; however, screening is still at best rare and in most places not available at all.

Persistent genital HPV infection has been established as the main etiological factor in the development of cervical cancer. HPV types 16 and 18 infection alone account for up to 70% of all cervical cancers. The HPV prevalence and HPV type distribution according to human immunodeficiency virus (HIV) status have been addressed in some studies in sub-Saharan Africa; however, only few have been population based and more information is needed in order to fully understand the burden of HPV infection according to HIV status.^3,4 The aim of the current study, based on women from the general population coming for screening, was to assess both the overall and type-specific HPV prevalence in women of Tanzania and to explore the distribution of HPV types according to HIV status and to Pap smear results. This information will be valuable for assessing the potential value of HPV vaccination in this population.

MATERIALS AND METHODS

Enrollment and Data Collection

Between February 2008 and March 2009, we conducted a cross-sectional study of women in both urban and rural areas in Tanzania. The study was approved by the ethical committee of the Tanzanian National Institute of Medical Research, and all the women who participated in the study were informed orally about the goal of the study and gave informed consent.

The women from urban areas comprised a group living in urban and semiurban areas in Dar es Salaam Region (Kinondoni, Temeke, and Ilala districts), who spontaneously came for screening at the Ocean Road Cancer Institute screening clinic, and a second, community-based group of women who had never spontaneously responded to the National Cervical Cancer Screening program. In the community-based part of the study, multistage cluster sampling was performed, where 3 wards from each of the 3 municipals in Dar es Salaam were randomly selected and 5 streets then randomly selected from each of the 9 wards.

The group from rural Tanzania consisted of women from Pwani (northeast and southeast of Dar es Salaam Region) and Mwanza (northwestern Tanzania) regions. For the study in rural area, cervical cancer screening clinics were set up in small district hospitals and health centers, at which at least 2 health care personnel per site had undergone the cervical cancer training program organized by the Ocean Road Cancer Institute within the previous 12 months.

In all locations, the women were informed by general public announcement that cervical screening was taking place and that they were invited to meet at their nearest hospital or health centre for an interview and a gynaecological examination.

All women were interviewed by trained female nurses using a structured questionnaire to collect information on socioeconomic factors, lifestyle, and reproductive history. The women also underwent gynecological examination, including visual inspection with acetic acid, a conventional Pap smear, and collection of cervical cells in sample transport medium (Digene Corporation, Gaithersburg, MD) for later HPV testing. Subsequently, the suspended specimens were stored at −20°C and shipped at the end of the study to the Department of Experimental Virology for analysis (Tuebingen, Germany). In addition, a blood sample was obtained for HIV testing from all women.

HPV results were not used in the diagnostic work up as the results were not available until several months after the end of the study. Screening against cervical cancer in the current study was based on the current standard of care in Tanzania (visual inspection with acetic acid) and in addition to Pap smear results.

Conventional Pap Smear

Cervical cells were collected by mean of Aylesbury wooden spatula (CellPath, Wales, United Kingdom) and endocervical brush (CellPath, Wales, United Kingdom), smeared on a glass slide and fixed in ethanol. After fixation slides were air dried and stored in glass slides plastic containers.⁵ The slides were subsequently sent to Department of Pathology, Hvidovre University Hospital, Denmark, where all the cervical cytology samples were analyzed. Laboratory personnel were unaware of any information concerning the study participants, and Pap smears results were categorized according to Bethesda classification system 2001.⁶

HPV DNA Testing

The method used to test for the presence of HR HPV DNA in cervical cell swabs has been described previously.⁷ Briefly, HPV detection was performed without knowledge about the study participants. The hybrid capture 2 (HC2) test (Qiagen, Hildesheim, Germany) with the HR probe cocktail was used for all women included in the study. A result was considered positive when a woman was found to have one or more of the 13 HR HPV types—16,18,31,33,35, 39,45,51,52,56,58,59,68—included and attained or exceeded the United States Food and Drug Administration-approved threshold of 1.0 pg HPV DNA/mL, which corresponds to 1.0 relative light unit coefficient.

HPV-positive samples were further analyzed to determine the HPV genotype (LiPaExtra; Innogenetics, Gent, Belgium). DNA was isolated in a MagnaPure device (Roche Systems, Indianapolis, IN) from 200 μL of the remaining denatured product from the HC2 test. The INNO-LiPA HPV Genotyping Extra test (Innogenetics) is a line blot assay based on the principle of reverse hybridization. It covers all currently known HR and probably HR HPV types (16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, and 82) as well as 7 low-risk types (6, 11, 40, 43, 44, 54, and 70) and 3 additional types (69, 71, and 74).⁸ Specific sequences of the L1 region of the HPV genome are amplified with short-fragment polymerase chain reaction assay (SPF10) primers, and the resulting biotinylated amplicons are then denatured and hybridized with specific oligonucleotide probes. A set of primers for the amplification of the human HLA-DPB1 gene is added to monitor sample quality and extraction. The strips are processed automatically in the Auto-LiPA 48 device. Objective, automatic interpretation of the strips is provided with the LiRAS line probe assay (LiPA) HPV.

HIV Testing

Blood specimens were tested according to an algorithm that included 1 rapid immunoassay (SD Bioline HIV-1/2 3.0 rapid test, Standard Diagnostics Inc., Republic of Korea), and, if positive, an additional confirmatory immunologic test (Determine HIV-1/2, Abbott Laboratories SA, South Africa) was performed. If there was a discrepancy between these 2 tests, a third test was used (UNIGold method/Recombigen HIV, Trinity Biotech, US). This algorithm is in accordance with Tanzanian Ministry of Health guidelines.

The results were given to each woman individually in a separate room in order to ensure confidentiality. The women were given pre- and post-HIV test counseling by a trained nurse according to national and international guidelines. Women who were HIV positive were referred to a care and treatment clinic and were offered a thorough follow-up, with determination of primary CD4 cell level, advice, information, and psychological support. They were offered immediate treatment when relevant. This follow-up was voluntary and free of charge.

Statistical Analysis

The overall HR HPV prevalence and type-specific HPV prevalence were assessed for all women. In addition, the HR HPV prevalence and 95% confidence interval (CI) were estimated according to age group (<25, 25–29, 30–34, 35–39, 40–49, 50–59, and ≤60 years) and cytology (normal, atypical squamous cells of undetermined significance [ASCUS], low grade squamous intraepithelial lesion [LSIL], and high grade squamous intraepithelial lesion [HSIL]). Finally, according to HIV status (positive, negative), we estimated the type-specific HPV prevalence with 95% CI. HPV types were grouped in species based on the analysis of the HPV genome. An HPV phylogenetic tree has previously been developed and described, showing that the genital HPV types belong to the genus α papillomavirus and that lower-order clusters are described as species.⁹

Statistical analyses were carried out using the Social Analysis Statistical software (SAS version 9.1).

RESULTS

A total of 3767 women were willing to participate in the PROTECT study. We excluded 37 women with clinical conditions who prevented a proper examination (e.g., total hysterectomy, menstruation, and pregnancy) and 31 women because of their incomplete samples could not be tested for HPV (empty tubes or too little cell material), leaving 3699 women enrolled. An additional 96 women for whom Pap smear results were not adequate or were missing were also excluded, leaving 3603 women for analysis.

The women were 15 to 82 years old (mean, 38 years); most (82.8%) were married; 43.3% were Christian, 41.1% were Muslim, and 15.7% were of other denominations. Two-thirds (63.5%) had completed only primary school education, which is the minimal compulsory level in Tanzania. About half the women (49.3%) lived in rural areas and 50.7% lived in urban areas. The mean age at first intercourse was 18.2 years, the average lifetime number of partners reported was 3, and they had an average of 3.8 pregnancies (data not shown).

Among the 3603 women included in the study, a total of 3235 (89.8%) had a normal Pap smear result, 169 (4.7%) had ASCUS, 62 LSIL (1.7%), 132 HSIL (3.7%), and 5 women had cervical cancer (0.1%). Finally, a total of 334 women (9.3%) were HIV positive, 3005 (83.4%) were HIV negative, and 264 women (7.3%) did not want to be tested for HIV.

Overall and Type-Specific HPV Prevalence

HR HPV was found in 725 women (20.1%) by means of HC2 test (Table 1). The prevalence was high among younger women, peaking in those aged 25 to 29 years (27.6%; 95% CI, 24.1–31.1), decreased thereafter until age 50 to 59 years (11.7%; 95% CI, 8.6–14.9), and then showed a second peak among women aged ≥60 years (28.6%; 95% CI, 21.1–36.1). When we restricted the analysis to include only women with normal cytology, the HR HPV prevalence pattern was the same with prevalences of respectively 20.7% (95% CI, 15.6–25.8) among women 25 to 29 years old, 5.7% (95% CI, 3.3–8.1) among 50- to 59-year-old women, and 18.4% (95% CI, 11.3–25.5) among ≥60-year-old women.

Table 1: Type-Specific Prevalence of High-Risk (HR) HPV Infection Among 3603 Tanzanian Women

Overall, the prevalence of HR HPV was significantly higher in women from rural area (23.4%; 95% CI, 21.5–25.4) than in women from urban area (16.9%; 95% CI, 15.2–18.6), and HR HPV prevalence among women younger than 25 years old was significantly higher in rural area (32.5%; 95% CI, 25.0–39.9) compared with urban area (13.8%; 95% CI, 7.3–20.2) but was similar in all other age groups in relation to the place of site of residence (data not shown). The same pattern was observed when we included only women with normal cytology (data not shown).

Genotyping showed that among the 725 HC2-positive HR HPV samples, 21 samples were LiPA negative, no specific HPV type could be identified in 85 samples, and in 23 samples, only low-risk HPV types were detected by LiPA. Table 1 shows the type-specific HPV prevalence and the relationship between HPV types and age. The most common HR HPV type was HPV52 (4.4% of the overall population; 21.9% of HPV-positive women), followed by HPV16 (3.8% of overall population; 18.8% of HPV-positive women). Other common types in HPV positive women were HPV51 (13.4%), HPV35 (13.2%), HPV18 (10.5%), and HPV66 (10.3%); 11.7% of the HPV-positive women had uncharacterized HPV types, and 2.9% of HC2-positive women tested negative by LiPA.

For all HPV types, the prevalence was highest in the youngest age groups and then decreased with age, with a second peak in the oldest age group (women aged ≥60 years). The infection with multiple HR HPV types showed a decreasing tendency with age (Table 1).

Overall and Type-Specific HPV Prevalence According to Pap Smear Results

The HR HPV prevalence among women with normal Pap smear was 14.8%; the corresponding figures among women with ASCUS, LSIL, and HSIL or worse were 41.4%, 77.4%, and 94.2%, respectively (Table 2). The category with HSIL or worse lesions contained 5 cervical cancers that were all HR HPV positive. In women with normal Pap smears, single HR HPV infections were identified in 10.4%, whereas HR HPV single infections were present in 28.4%, 46.8%, and 60.6% women with ASCUS, LSIL, and HSIL or worse, respectively.

Table 2: High-Risk (HR) HPV-Type Distribution According to Cervical Cytology Results Among 3603 Tanzanian Women

Table 2 shows also that among women with HSIL and worse, HPV16 (32.8%) was the most common type, HPV52 was the second most frequent type (24.1%) followed by HPV18 (18.2%) and HPV35 (13.1%). Among the 5 cancers contained in this group, 4 cancers contained HPV16 and 1 had HPV18, whereas none were positive for HPV52 (data not shown). In women with LSIL, HPV52 (14.5%) was the most prevalent type followed by HPV51 (12.9%), HPV35 (12.9%), HPV16 (9.7%), and HPV56 (9.7%).

When looking at type-specific single infection distribution in HSIL or worse, HPV16 (19.7%) was the most prevalent followed by HPV52 (13.9%), HPV18 (7.3%), and HPV35 (5.1%) (Table 2).

HR HPV Prevalence According to HIV Status

HIV-positive women had a significantly higher prevalence of HR HPV infection (46.7%; 95% CI, 41.4–52.1) than HIV-negative women (17.2%; 95% CI, 15.8–18.5) and than women who refused to be HIV tested (20.1%, 95% CI, 15.2–24.9). In all 3 groups, the prevalence of HR HPV was high among younger women, peaking in those aged 25 to 29 years, subsequently decreasing with increasing age and showing a second peak among women aged ≥60 years (Fig. 1). Multiple HR HPV infections were significantly (P = 0.016) more prevalent among HIV-positive women (41.7%; 95% CI, 33.9–49.4) than HIV-negative women (28.9%; 95% CI, 25.0–32.8), but none of the phylogenetic groups or specific HR HPV types were significantly more common among HIV-positive than HIV-negative women, e.g., HPV16_{HIV pos.} = 19.9%; 95% CI, 13.6 to 26.1; and HPV16_HIVneg. = 17.8, 95% CI, 14.4 to 21.1 (Table 3).

Figure 1: Age-specific prevalence of high-risk (HR) HPV infection among HIV-positive (n = 334), HIV-negative (n = 3005), and women refusing HIV testing (n = 264) from Tanzania.

Table 3: Prevalence of HR HPV Types Among HC2-Positive Women According to HIV Status

DISCUSSION

We found a high prevalence of HR HPV among Tanzanian women, 20.1% of all women carrying one or more HR HPV types, ranging from 15% among women with normal cytology to 100% among women with cervical cancer. This study is, to the best of our knowledge, the largest study among general population investigating both HR HPV prevalence and type distribution in women in the general population in sub-Saharan Africa. In other studies in the region, HPV infection showed large regional variation, ranging from 13% to 40%.^3,4 In accordance with studies all over the world, we observed that HR HPV prevalence decreased significantly with increasing age; however, we saw a second peak among women aged ≥60 years, as also observed in some studies in Central and South America,^10,11 Africa,¹² and Asia.¹³ In the current study, this could not be explained by a higher frequency of abnormal cervical lesions among the older women as the same significant second peak was observed among women with normal cytology.

Several factors have been suggested as contributing to the peak in HPV prevalence in older women: recent nonmonogamous sexual activity of women or their current partners leading to new infections; decreasing immunity with age implying a subsequently lower rate of HPV clearance and increased duration of HPV infection (persistence)¹⁴; changes in hormonal or immunologic factors leading to reactivation of a latent HPV infection¹⁵; or a cohort effect in HPV prevalence.¹⁶ Gonzalez et al. similarly showed a second peak in HR HPV prevalence in study conducted among older women in Guanacaste (Costa Rica).¹⁶ The study further assessed both recent and past sexual behaviors, as well as cellular immune response, but was not able to point out single important factor and concluded that the HPV infections observed among older women (≥45 years) could both be explained by recent sexual behavior of the woman or her partner and by past sexual behavior (i.e., reactivation of a past HPV infection) with the reappearance of the past HPV infection possibly associated with a reduced immune response.¹⁶ Longitudinal studies with thorough follow-up of large groups of women in the general population in different regions of the world, including information on cervical cytology, HPV prevalence, sexual behavior, hormonal levels, and immunologic factors, will be necessary to elucidate this age-related aspect of the natural history of HPV.

In the current study, HPV52 was the most prevalent HR HPV type (4.3% among all women), closely followed by HPV16 (3.8%); HPV Types 51, 35, 18, and 66 were also common with prevalences ranging from 2.1% to 2.7%. Our finding of a high prevalence of HPV52 is in line with those of several studies in other African countries^17,18 but not all.¹⁹ A recent meta-analysis of studies of HPV prevalence among women with normal cytology on 5 continents also showed that HPV52 is particularly frequent in Africa.²⁰ The differences between the studies might be because of differences in the sensitivity of the methods used or might reflect true differences in the distribution of HPV types in different populations. It has been reported that the concomitant presence of HPV52 with other HPV types, particularly 33, 35, and 58, compromised its detection, and it was therefore underestimated.²¹ The use of more sensitive polymerase chain reaction methods, like the LiPaExtra test, to detect HPV52 in samples containing other HPV types might be one explanation for the higher prevalence of HPV52 detected in recent studies.²²

In the current study, we find that in women with normal cytology or low-grade cervical lesions, the prevalences of HPV52 and HPV16 are high and largely of the same magnitude with a slight predominance of HPV52, especially in lesions with a single HPV infection. In contrast, HPV16 is the dominating HPV type in HSIL or worse, but HPV52 and HPV18 also play an important role. This result was even clearer in the 5 cancers, where HPV16 was detected in 4 of them and HPV 18 was detected in 1 cancer, whereas HPV52 was not detected in any of them. When considering single-type HPV infections, it is the general hypothesis that HPV types belonging to the α 9 species group as well as HPV18 from the α 7 group showed an increasing prevalence with increasing severity of the cervical lesion, whereas this was not the case for the other HPV types. These finding are in concordance with a large meta-analysis study showing that HPV16 and HPV18 are the 2 most common types among women with invasive cervical cancer in different regions of the world and that HPV52 is also common in women with HSIL or invasive cervical cancer in Africa.²³ Nevertheless, data on women infected with HPV52 alone and diagnosed with HSIL or worse are scare, and larger longitudinal studies are needed to investigate further the role of HPV52.

We found that the prevalence of HR HPV was substantially higher among HIV-positive women (46.7%) than HIV-negative women (17.2%). The general pattern for HR HPV prevalence in relation to age showed a decrease with increasing age, independently of HIV status. We observed a second peak in both HIV-positive and HIV-negative women aged ≥60 years and also in the group of women who were not tested for HIV; however, because of the limited number of HIV-positive/HR HPV-positive older women, this pattern was not statistically significant. Several studies have shown that immunocompromised individuals are at increased risk for carrying HPV,²⁴ and women with the highest HIV RNA viral load and the lowest median CD4 lymphocyte count have a higher HPV prevalence.²⁵ Moreover, it has been shown that HIV-positive individuals are more prone to cervical cancer and other HPV-related cancers, and this may also indicate a high probability of HPV persistence.²⁶ In our study, we did not have access to the women's CD4 counts or any proxy information on HIV-infection stage; therefore, we could not investigate these factors in relation to HPV infection. A study in Zimbabwe showed that women infected with cervical HPV were 2.4 times more likely to become HIV positive,²⁷ suggesting that not only being HIV positive increases the odds of being HPV positive but being HPV positive also increases the risk for becoming HIV positive. This has been attributed to the fact that HPV uses several mechanisms to downregulate innate and cell-mediated immunity.²⁸ As our study was cross-sectional, we could not determine the timing of acquisition of HPV and HIV.

Our results show that none of the carcinogenic HPV types are significantly more prevalent among HIV-positive women than HIV-negative women, in line with the findings of a previous study.²⁹ This suggests that HIV status does not influence the distribution of HPV types. Therefore, the currently available HPV vaccines could prevent HPV infection independently of HIV status. In our study, we were unable to investigate whether women with different CD4+ T-cell counts showed different susceptibility to different HPV types; however, significantly more HIV-positive women than HIV-negative women had multiple HPV infections, indicating that immune deficiency leads to greater susceptibility to concurrent infection with >1 type of HR HPV and to a lower or impaired ability to clear HPV infection.³⁰ HIV-positive women have a naturally poor response to HPV infection, as judged by an increased occurrence of genital warts and an increased risk for cervical cancer.³¹ The quadrivalent HPV vaccine (Merck) has been shown to be safe and highly immunogenic in HIV-positive men,³² and several studies are under way to investigate the effect of both bivalent (Glaxo Smith Kline) and quadrivalent HPV vaccines in HIV-infected women, although, to our knowledge, no results have been published yet. The guidelines for studies of vaccination of HIV-positive individuals against HPV are very strict, and most of the clinical trials address specific subgroups. Two important inclusion factors in these trials are good compliance with antiretroviral therapy in the past 6 months for people under treatment before recruitment into the study, and a CD4+ T-cell count exceeding 350 cells/μL. Future studies should investigate the efficiency of HPV immunization in people who are HIV positive but not necessarily under antiretroviral therapy.

In conclusion, we found a high HPV prevalence among women in Tanzania, the most frequent HPV type among all women being HPV52, followed by HPV16, HPV51, HPV35, and HPV18, but with HPV16 as the dominating HPV type in high-grade cervical lesions. Our study confirms that HIV-positive women in all age groups have a higher HR HPV prevalence than HIV-negative women. In contrast, we found no significant difference in HPV type distribution between HIV-positive and HIV-negative women. These results suggest that existing HPV vaccines might benefit women regardless of their HIV status.