High prevalence of and factors associated with human papillomavirus infection among women attending a tertiary hospital in Gauteng Province, South Africa

Persistent high-risk (hr) human papillomavirus (HPV) infection is a necessary cause of cervical cancer. Cervical cancer is a major public health problem in Sub-Saharan Africa including South Africa. This study investigated the prevalence of and factors associated with hr-HPV infection among women attending a tertiary hospital in Gauteng Province, South Africa. Cervical samples were collected from 526 participants aged ≥ 18 years using a Cervex Brush® Combi and tested for hr-HPV types on the Abbott m2000 analyzer using the Abbott RealTime HR HPV assay. Samples that tested hr-HPV deoxyribonucleic acid (DNA)-positive were further tested for hr-HPV E6/E7 messenger ribonucleic acid (mRNA) using the APTIMA® HPV assay on the Panther system (Hologic, Inc.). Sociodemographic data were collected using a self-administered questionnaire. Binomial regression analysis was used to assess factors associated with hr-HPV infection. Overall hr-HPV DNA prevalence was 48.1% (95%CI: 43.8–52.4%). Of the hr-HPV DNA-positives, 24.5% (95%CI: 19.3–30.1) had HPV-16; 12.3% (95%CI: 8.5–16.9) had HPV-18 and 87.4% (95%CI: 82.6–91.2) had other 12 h-HPVs. Of the samples positive for hr-HPV DNA, 84.2% (95%CI: 79.1–88.5) (213/253) were positive for hr-HPV E6/E7 mRNA. Advanced age was an important factor linked to hr-HPV E6/E7 mRNA positivity. Based on multivariate binomial regression analysis, unemployment (PR: 1.50; 95%CI: 1.23–1.83) and being married (PR: 0.61; 95%CI: 0.47–0.81) were identified as statistically significant (p < 0.0001) predictive and protective factors, respectively, for hr-HPV infection. The prevalence of hr-HPV infection was high. Furthermore, hr-HPV DNA-positive samples had a high hr-HPV E6/E7 mRNA prevalence. The presence of hr-HPV E6/E7mRNA indicates active infection and thus a greater risk of developing the cervical disease. Therefore, HPV mRNA testing could be a better test to monitor women who are positive with Pap smear before colposcopy is performed to reduce the burden of referrals.

. Furthermore, cervical cancer cases are most common among women between the ages of 15 to 44 years) [2]. High-risk human papillomavirus (hr-HPV) is the most common sexually transmitted infection (STI) [3]; and an etiologic agent for the development of cervical cancer [4]. Although 70% of sexually active individuals are infected with HPV, the infection usually clears up within a few months, with 90% of cases clearing within two years after HPV acquisition [5]. However, 10 to 20% of HPV infections persist latently, leading to disease progression and eventually invasive cancer [6]. HPV messenger ribonucleic acid (mRNA) tests determine whether the virus is actively expressing the proteins responsible for oncogenesis in HPV infection [7] and thus allow the identification of transient and potentially transforming HPV infections [8]. E6 and E7 oncogene expression cause cellular changes [9]. Cellular changes are achieved when genes hijack the cellular ubiquitin-proteasome system (UPS) to degrade or inactivate the cellular tumor suppressor protein p53 and the retinoblastoma protein (pRb), forcing the host cell into the S phase [10].
To curb cervical cancer incidence and mortality rates in South Africa, two HPV vaccines; Cervarix ® (GlaxoS-mithKline Biologicals, Rixensart, Belgium) and Gardisil ® (Merck & Co., Whitehouse Station, NJ USA) were approved in 2008 [11]. In 2014, the first-dose phase of the South African National HPV campaign was launched through the Integrated School Health Programme implemented by the National Department of Health (NDoH), in collaboration with the Department of Basic Education targeting female learners aged 9 years and older in Grade 4 [12]. Cervarix ® protects against two hr-HPV types, HPV-16, and HPV-18, which together account for 70% or more of all cervical cancers globally [13]. Gardisil ® confers protection against two low-risk (lr) HPV types causing about 90% of warts (HPV-6 and HPV-11) and two hr-HPV types (HPV- [16][17][18]. Although the approved vaccines do not prevent diseases caused by other hr-HPV types, it is anticipated that they will contribute to significant reductions in cervical cancer incidence in the country. Therefore, effective cervical cancer screening will remain the most important secondarylevel prevention intervention. Advances in cervical cancer screening technologies have introduced HPV testing due to the higher sensitivity for precancer and cancer detection than cytology [14]. In primary HPV screening programs, two types of hr-HPV assays are used; HPV tests that are very sensitive for HPV deoxyribonucleic acid (DNA) detection and HPVspecific tests that target HPV E6/E7 oncogenes with high specificity but lower sensitivity [15]. Although HPV testing is replacing cytology, particularly in HICs, the test is associated with lower specificity, which hence requires a second test to identify women who are at risk of developing cervical cancer [16]. The low specificity associated with HPV DNA testing may lead to unnecessary colposcopy referral and overtreatment, hence some countries tailor the triage strategy and/or follow-up [17].
HICs routinely use cervical cytology and HPV testing to detect high-grade intraepithelial neoplasia to prevent invasive cervical cancer. However, in low and middle-income countries (LMICs) there is a limitation of resources such as health infrastructure, and trained personnel, which therefore hinder the feasibility and effectiveness of Papanicolaou (Pap) smear tests [18]. The lack of adequate infrastructure may also pose a challenge to the introduction of liquid-based cytology (LBC). Nevertheless, some LMICs have introduced HPV testing in their national cervical cancer screening programs [17]. In South Africa, in addition to LBC-based cervical cancer screening, the NDoH Cervical Cancer Prevention and Control Policy has recommended HPV DNA-based testing in the public sector [19].
The acquisition and persistence of HPV infection and development of HPV-associated cancers are associated with high-risk sexual behavior, such as the age of first vaginal sex, age of first oral sex, and the number of oral and vaginal sexual partners [20]. Other risk factors of HPV infection include STIs such as HIV (Swaziland) [21]; being single and cohabiting with a sexual partner (Ghana) [22]; practicing intravaginal substance use (South Africa) [23]; and an increased number of lifetime sexual partners (South Africa) [24].
Hr-HPV testing is relatively new to sub-Saharan Africa (SSA), and most studies have utilized hr-HPV DNA assays compared with hr-HPV E6/E7 mRNA assays in the region. Although data on HPV DNA prevalence and associated factors are available in SSA regions including South Africa, there is a lack of data on the prevalence and associated factors of hr-HPV E6/E7 mRNA in the country. Furthermore, despite the high burden of cervical cancer in the country, there is limited research on factors associated with hr-HPV E6/E7 mRNA expression. Therefore, this study aimed to investigate the prevalence of and factors associated with hr-HPV infection among women attending tertiary hospital gynecology clinics in Gauteng Province, South Africa.

Study design and population
This was a cross-sectional study conducted from March 2016 to September 2018 at a tertiary hospital, situated in the north of Pretoria in the Gauteng Province, South Africa. The study participants were women aged 18 years and older, attending gyne-oncology and termination of pregnancy (TOP) clinics within the tertiary hospital. All women who had provided written informed consent were included. Those who had undergone a complete hysterectomy or were menstruating at the time of enrolment were excluded. Services offered at the gyne-oncology clinic included large loop excision of the transformation zone (LLETZ) procedure, a review with Pap test after treatment with LLETZ, and colposcopy examination after a positive Pap test. The termination of pregnancy (TOP) clinic under the gynecology department offered family planning, cervical cancer screening using the Pap test, and other gynecology-related services.

Ethical approval and consent to participate
The study was conducted with the approval of the Research Ethics Committee of Sefako Makgatho Health Sciences University, (SMUREC/P/75/2016: IR) and (SMUREC/M/279/2019: PG). Written informed consent was signed by each participant.

Data and specimen collection
Women were provided with written information sheets that had a brief description and purpose of the study. Consenting participants completed a self-administered questionnaire covering demographic characteristics, as well as information related to sexual practices and reproductive history, knowledge of HPV, cervical cancer, and participants' experiences and preferences using the selfsampling device before clinical examination for cervical sample collection. The researchers distributed the questionnaire after the study had been explained to the participants.
An endo/ectocervical sample was collected using a Cervex Brush ® Combi (Rovers ® Medical Devices B.V., the Netherlands). The speculum was inserted and opened to widen the vagina to assist the pelvic examination, and thereafter, the brush was delicately inserted into the endocervical canal until the lateral bristles were touching the ectocervix and rotated two full turns clockwise. The collection brushes were immediately rinsed into Thin-Prep PreservCyt ® Solution vials (Hologic Incorporated, Bedford, Massachusetts) for the preservation of collected cells, before being discarded. For laboratory testing, all samples were transported at room temperature to the HPV and STIs Training Centre for Africa, Department of Virology at the Sefako Makgatho Health Sciences University.

Detection of hr-HPV E6/E7 mRNA
Samples that tested positive for hr-HPV DNA with the Abbott m2000 system were tested for hr-HPV E6/E7 messenger ribonucleic acid (mRNA) using the APTIMA ® HPV assay (Hologic Gen-Probe, Inc., San Diego, Canada). This assay is an in vitro nucleic acid amplification test for the qualitative detection of hr-HPV E6/E7 viral mRNA of the same 14 h-HPV types detected by the Abbott RealTime HR HPV assay. An aliquot of 1 mL of each Thin Prep PreservCyt vial was diluted into 2.9 mL of Aptima specimen transport media and tested according to the manufacturer's instructions. The assay incorporated an internal control (IC) to monitor nucleic acid capture, amplification, and detection, as well as operator or instrument errors. The results were interpreted as negative, positive, or invalid if the run or IC failed. The results did not discriminate between the 14 h-HPV types, with the detection of any of the 14 h-HPV types showing a positive result.

Statistical analysis
Statistical analysis was performed using SAS version 9.4 for Windows. Descriptive data analyses included the mean, standard deviation (SD), the range for age, and frequency distributions for categorical data. Prevalence estimates were calculated by dividing the number of infected individuals by the total number of the sample. Exact binomial 95% Confidence intervals (CIs) were calculated for proportions. The Prevalence Ratios (PRs) of each risk factor for each outcome variable (hr-HPV infection and hr-HPV E6/E7 mRNA expression) were determined, together with its 95% confidence interval, using binomial regression. Categories with n < 15 overall were not included in analyses (no reliable inference can be made based on such small groups). Study variables significant at p < 0.20 were combined into a multivariable model, after examining each pair of variables for possible confounding using the chi-squared test (or Fisher's exact test for 2 × 2 tables). All variables (with p < 0.20) were at first included in the multivariate model, but after sequentially removing those which were not significant. A value of Cramer's V (or the phi coefficient for Fisher's exact test) > 0.60 was regarded as too strong an association to include both variables in a multivariable model. In the multivariable model, non-significant variables were removed one at a time until only significant variables remained. A p-value of ≤ 0.05 was considered statistically significant.

Demographic characteristics
The mean age of participants was 36.8 (SD: ± 11.04; range: 18 to 68 years. The majority (61.6% [322/523]) of the participants were between the ages of 18 to 39 years. All 526 enrolled women were black Africans (99.6%), except for two women (0.4%). The majority (67.3%) of the participants were single. The majority (85.6%) of participants were from semi-urban (township) areas compared to only 2.3% from rural areas. More than fifty percent (57.4%) of the participants had 1 or 2 children and most (79.8%) of the participants were sexually active with 74.5% having at least one sexual partner at the time of the study. Of the 526 participants, few (28.9%) reported the use of contraceptives, only 17.1% were pregnant, and 16.0% of pregnancies were in their first trimester Table1.

Prevalence of hr-HPV types and E6/E7 mRNA in different groups
Hr-HPV DNA and hr-HPV E6/E7 mRNA among participants attending gynecology clinics were stratified by reasons for clinic visits. There was a significant moderate association between hr-HPV infection and reason for visit (p < 0.0001; Cramer's V = 0.33). The prevalence of hr-HPV DNA and hr-HPV E6/E7 mRNA was significantly higher for those who visited the clinic for the LLETZ procedure and significantly lower for those who came for a routine Pap smear and termination of pregnancy (TOP) Table 3. There was a significant but weak association between hr-HPV E6/E7 mRNA expression and reasons for a visit (p = 0.0001; Cramer's V = 0.29).

Age-specific prevalence of hr-HPV DNA and hr-HPV E6/E7 mRNA
There was a high hr-HPV DNA prevalence (57.0%) in women younger than 30 years of age. Hr-HPV DNA prevalence decreased (45.6%) in women aged 30 to 39 years and increased (48.8%) aged 40 to 49 years. Hr-HPV prevalence then decreased with increasing age Table 4. For hr-HPV E6/E7 mRNA, the prevalence was high in women over 40 years. In contrast to hr-HPV DNA, hr-HPV E6/E7 mRNA was lower in women younger than 30 years Table 5.      being married were statistically significant predictive and protective factors, respectively, for hr-HPV infection.

Factors associated with hr-HPV E6/E7 HPV expression
Based on the bivariate analysis, age was significantly associated with hr-HPV E6/E7 mRNA expression. Being between the ages of 40 to 49 years was significantly associated with an increased risk (PR: 1.38; 95% CI: 1.20-1.60; p < 0.001) of hr-HPV E6/E7 mRNA expression. Furthermore, contraceptive use for more than 24 months was a (PR: 1.16; 95% CI: 1.06-1.28; p = 0.0022) was a predictive factor for hr-HPV E6/E7 mRNA. Multivariate binomial regression identified age as a statistically significant predictive factor for hr-HPV E6/E7 mRNA Table 5.

Factors associated with hr-HPV infection in women attending gynecology clinics for different reasons
Bivariate analysis revealed a significant association between hr-HPV infection with age, marital, and employment status in women attending routine Pap smear, In a multivariate binomial regression, only marital status and employment status remained significantly associated with hr-HPV infection Table S1.

Discussion
The current study used the Abbott RealTime HR HPV assay and the APTIMA ® HPV assay to detect 14 h-HPV types and the expression of hr-HPV E6/E7 mRNA of these HPV types among women attending gynecology clinics at a tertiary hospital in Gauteng Province, South Africa from 2016 and 2018. The overall hr-HPV DNA prevalence was 48.1%. Comparable hr-HPV DNA prevalence was reported in previous South African studies; 54.5% among HIV-negative women aged 14 to 30 years [23] and 54.3% among women attending public sector primary healthcare clinics for routine gynecological and non-gynecological primary healthcare-related reasons [25], respectively. Similar HPV prevalence was reported by Ginindza et al. [21] in Swaziland (46.2%) and Mbulawa et al. [24] in South Africa (45.7%) using the GeneXpert HPV assay among women aged 15 to 49 years, and the Roche Linear Array HPV genotyping test among HIVnegative women aged 16 to 22 years, respectively. Low HPV DNA prevalence (28.5%) compared to our study was found among 30 to 98 years HIV-positive and -negative women recruited from a community health clinic in South Africa [26]. The relatively low HPV DNA prevalence reported by Taku et al. [26] could be influenced by the exclusion of the younger population that has been shown to harbor higher HPV infection. Although HPV DNA prevalence is higher among young women, most infections are transient and asymptomatic [27]; therefore, clearing in the majority and only persisting in the minority.
High hr-HPV DNA prevalence (57.0%) among the younger population aged 18-29 years in the current study supports the continuation of HPV vaccination for South African girls. Consistent with previous studies, a high prevalence of HPV DNA among young women has previously been reported in Swaziland, South Africa, and the Democratic Republic of Congo [21,23,[28][29][30]. The second peak of HPV DNA prevalence in the group of women aged 40 to 49 years old observed in the current study was previously reported in the same age group in Korea [31]. Previously, other studies have observed a second peak of prevalence in the group of women aged 46 to 66 and 60 years and older, respectively [28,29]. The second peak of prevalence in the current study might be explained by the low cervical cancer screening coverage observed among the study population.
The current study also demonstrates a substantially higher hr-HPV E6/E7 mRNA prevalence than what was found by other studies within the African continent. Hr-HPV E6/E7 mRNA prevalence in our study was higher compared to the overall hr-HPV E6/E7 mRNA of 36.7% reported among HIV-infected women in South Africa and 30.0% among sex workers in Kenya [32,33]. The abovementioned studies tested all samples as opposed to the current study that only tested hr-HPV DNA-positive samples; this could explain the high prevalence found in our study. Comparable to our study, a previous study reported a slightly lower hr-HPV E6/E7 mRNA prevalence (67%) among HPV DNA-positive women [9]. The high hr-HPV DNA and hr-HPV E6/E7 mRNA prevalence in the current study is likely due to the high-risk nature of the study population, particularly women who were scheduled for the LLETZ procedure, colposcopy, and those who had already undergone the LLETZ procedure. In the current study, not all women who tested positive for hr-HPV DNA expressed hr-HPV E6/E7 mRNA. The detection of HPV E6/E7 mRNA is suggestive of an increased risk of cervical cancer development as it shows active or progressing HPV infection [34]. Additionally, HPV E6/E7 mRNA detection is more specific than HPV DNA in identifying women at risk of cervical disease [27]. In the current study, hr-HPV E6/E7 mRNA prevalence increased with age, where women aged ≥ 30 years had the highest prevalence and those < 30 years had the lowest prevalence, contrary to hr-HPV DNA. In the current study, hr-HPV E6/E7 mRNA prevalence increased with age, where women aged ≥ 30 years had the highest prevalence and those < 30 years had the lowest prevalence, contrary to hr-HPV DNA. Although HPV infection is very high in younger women; the infections are transient and asymptomatic [27]. However, advanced age is one of the factors for the persistence of HPV infection [35]; which results from increased E6/E7 mRNA expression [36]. In advanced age groups, a high prevalence of hr-HPV mRNA has been previously reported [37].
As expected, a high prevalence of hr-HPV infection and expression of hr-HPV E6/E7 mRNA was found among women undergoing the LLETZ procedure and colposcopy. Factors including smoking, hr-HPV type, and persistence of hr-HPV posttreatment amongst others may contribute to treatment failure [40]; thus, patients are followed up following the LLETZ procedure. Hr-HPV DNA prevalence was high in women who had the LLETZ procedure. This is expected as HPV infection is still present in up to one-third of women who had undergone the procedure [41]. An observational study among patients with a high-grade squamous intraepithelial lesion (HSIL), who had LLETZ reported that HPV infection was not completely eradicated [35].
Two risk factors remained significantly associated with HPV infection after multivariate binomial regression analysis: marital and employment status. In a multivariate model, marital status was associated with reduced odds of hr-HPV infection. Single women may have a higher risk of HPV infection compared to married women due to risky sexual behavior such as not having a consistent sexual partner [42]. The data is consistent with previous studies that have reported a higher risk of HPV infection in single women [21,22,43]. Consistent with a previous study, unemployed women were significantly associated with hr-HPV infection [44]. This could be explained by the possibility that these women are low-income earners which could increase their level of poverty and high-risk sexual practices [44]. Contrasting findings were reported by Ginindza et al. [21] whereby there was no significant association between employment status and hr-HPV infection.
Furthermore, in a multivariate binomial regression model, only marital status (married vs. single) and employment status remained significantly associated with hr-HPV infection in women who attended routine Pap smear.
To our knowledge, no studies have investigated potential factors associated with hr-HPV E6/E7 mRNA expression in our setting.. Multivariate binomial regression analysis identified age as the most important risk factor for hr-HPV E6/E7 mRNA. Advanced age is an important factor linked to hr-HPV E6/E7 mRNA positivity. In a multivariate model, only age remained significantly associated with hr-HPV E6/E7 mRNA in women who attended routine Pap smear.

Limitations
Hr-HPV E6/E7 mRNA could have been overestimated as only hr-HPV DNA-positive samples were tested. Hr-HPV DNA-positive samples were more likely to be hr-HPV E6/E7 mRNA positive than hr-HPV DNA-negative samples. Furthermore, the Abbott HR-HPV assay targets the L1 region which is often deleted when integration occurs during disease progression. Therefore, we may have missed samples that could be hr-HPV DNA-positive when excluding hr-HPV DNA-negative samples for hr-HPV E6/E7 mRNA testing. This could have caused underestimation. Technical aspects such as storage of the samples could have degraded the genetic material (i.e., mRNA is less stable than DNA). HIV status was not included in the study, given that South Africa has a high prevalence of HIV, it would have been interesting to determine if HIV infection had any significant effect on hr-HPV infection and hr-HPV E6/E7 mRNA expression in this population.

Conclusions
The study found a high prevalence of hr-HPV infection and hr-HPV E6/E7 mRNA among women attending a tertiary hospital in Gauteng Province, South Africa. The high prevalence of hr-HPV E6/E7 mRNA indicates active HPV infection and thus a greater risk of developing cervical disease. The use of hr-HPV E6/E7 mRNA testing could identify women at greater risk and reduce unnecessary referrals, and subsequently the burden on healthcare facilities. Further research should be conducted to evaluate hr-HPV E6/E7 mRNA testing as a primary screening test, or co-testing together with liquid-based cytology, and/or follow up test for women already positive using cytology in the South African cervical cancer screening context.