Prevalence of vaccine and non-vaccine human papillomavirus types among women in Accra and Kumasi, Ghana: a cross-sectional study

Background Human Papillomavirus (HPV) infection is the main etiological factor for pre-invasive and invasive cervical cancer. HPV type-specific vaccination is being widely recommended to control the burden of disease, but the genotype-specific distribution of HPV may vary in different countries. The aim of the study was to determine the prevalence and distribution of HPV genotypes among women attending reproductive health services in Ghana, their associated risk factors, and to assess the potential coverage of identified HPV genotypes by three licensed vaccines among these women. Method Women presenting for reproductive health services in two regional hospitals in Accra and Kumasi from October 2014 to March 2015 were conveniently recruited into the study (n = 317). HPV-DNA detection and genotype identification were carried out by a nested multiplex PCR assay that combines degenerate E6/E7 consensus primers and type-specific primers for the detection and typing of eighteen HPV genotypes. Cytology was performed to screen women for cervical cancer lesions. Risk factors for HPV infection were analyzed by logistic regression. Statistical significance was accepted for p < 0.05. Results The age of study participants ranged from 21 to 76 years. Among women positive for HPV, 35.0% were infected with high-risk HPV, 14.5% with probable high-risk HPV, and 17.0% with low-risk HPV. The prevalence of HPV 16/18 was 8.2%, HPV 6/11/16/18 was 9.1% and HPV 6/11/16/18/31/33/45/52/58 was 28.4%. The most prevalent among HR-HPV were types 52 (18.3%) and 58 (8.8%). HPV positivity may be associated with educational background (p < 0.001), age at first pregnancy (p = 0.028), and age at coitarche (p = 0.016). Conclusions Our study revealed a high prevalence of HR-HPV infection among women. The high prevalence of HR HPV indicates that multivalent vaccines will be useful for controlling HPV burden in general population contexts. The distribution of HPVs in this population suggests that of the three currently available vaccines the nonavalent vaccine, which protects against seven HPV types in addition to HPV 16 and 18, has the highest coverage of HPV infections among Ghanaian women. Healthcare officials planning to reduce the transmission of HPV and cervical cancer must consider the coverage of the nonavalent vaccine as an advantage. Supplementary Information The online version contains supplementary material available at 10.1186/s12905-021-01511-1.


Introduction
Human papillomavirus (HPV) infection is the most common sexually transmitted viral infection worldwide [1]. Over 200 HPV types have been identified and have been completely sequenced [2]. Genital HPV types are classified as high risk (HR), probable high risk (PHR) or low risk (LR) according to the degree or likelihood associated with developing cervical cancer [1,2]. Approximately 50 genotypes are known to be oncogenic or HR and are correlated with invasive cervical cancer [3]. Among the 15 most common types known to be carcinogenic, HPV16 and HPV18 are responsible for approximately 70% of cervical cancer worldwide [4,5]. Other oncogenic HPV types, including HPV 31,33,45,52,and 58, are estimated to be responsible for another 18% of cervical cancers [6].
In Ghana, cervical cancer is ranked as the second most common cancer with an estimated incidence of 27.4 per 100,000 women [7,8]. Around 2797 women are diagnosed yearly and approximately 1699 deaths occur from cervical cancer [7,8]. The World Health Organization (WHO) predicts that there will be 5000 new cases of cervical cancer and 3361 cervical cancer deaths will occur annually in Ghana by 2025 [9].
HPV vaccination for young girls and cervical screening programs for older women can be an effective strategy to prevent cervical cancer [10,11]. Cervical screening programs, such as cervical cytology and visual inspection with acetic acid (VIA), are available but not mandatory in Ghana. The cervical screening programs are only effective on cervical cancer mortality, if a high proportion of women participate [12]. Moreover, it has been difficult to implement screening programs for cervical cancer in Ghana as well as in most sub-Saharan African countries, partly due to competing health needs such as HIV, malaria, tuberculosis and malnutrition [13,14].
In Ghana, the quadrivalent-Gardasil ® vaccine was piloted into the school-based vaccination program for 6000 girls aged years, in four districts in the Northern and Greater Accra Regions with support from the GAVI Alliance in November 2013 [24]. Meanwhile, the vaccine is available to the public for vaccination at a cost of approximately USD 50 per dose, potentially rising to USD 150 for three doses [25]. In spite of their availability for almost a decade, the uptake of HPV vaccines in Ghana has been poor. A number of reasons such as exclusion from services covered by universal health insurance and the low patronage of reproductive health services by adolescent girls may be cited [25][26][27]. However, the cost of HPV vaccines is a well-known barrier to vaccine accessibility in Sub-Saharan Africa and more innovative pricing solutions are constantly sought [25,28,29]. The large-scale deployment of vaccines with a broad coverage of common HPV types in a single-dose regimen could represent a sustainable, cost-effective, preventive strategy and potentially increase uptake of vaccination [30]. However, very few studies on HPV types not targeted by the present HPV-vaccines among Ghanaian women are available. The aim of our study was to determine the prevalence and distribution of HPV genotypes among Ghanaian women, their associated risk factors and to assess the potential coverage of identified HPV genotypes by available licensed vaccines.

Study design, population and sampling technique
Between October, 2014 and March, 2015, women presenting to the Cervicare/Reproductive Health Clinics at the Kumasi South Regional Hospital in Kumasi, Ashanti Region and Greater Accra Regional Hospital in Accra, Greater Accra Region, Ghana, for reproductive counselling, routine family planning, visual screening, Pap smear testing and other support services were invited to participate in this hospital-based study. The minimum sample size was determined using the Cochran formula [31] for estimating single population proportions. In a previous study, the prevalence of HPV 16/18 was 5% [32] and the utilization of reproductive health services was approximately 8% for women of fertility age (WIFA) in study areas [33]. Based on a confidence level of 95% (z = 1.96), 5% margin of error (d), and 10% non-response rate, a minimum sample size of 317 was required. The convenient sampling technique was employed to enroll participants. Criteria for exclusion included: pregnancy; < 20 years-old; actively menstruating on the day of sample collection; history of hysterectomy or conization; history of Pap smear and prior HPV vaccination. Participants were fully informed about the purpose, procedures, risks, and benefits of participating in this study and informed consent was obtained from all subjects.

Data collection
At enrolment, participants completed a questionnaire and provided data on sexual behavior, reproductive history, contraceptive practice, smoking habits, history of sexually transmitted disease, screening history, and various measures of demographics and socioeconomic status (e.g., occupation, education), cervical cancer screening options and HPV vaccination.

Sample collection and processing
Two cervical samples (one each for Pap and HPV testing) were collected by trained nurses. Cervical specimens for HPV DNA test were suspended in a proprietary DNA solution (Biomatrica Co., San Diego, USA) for DNA preservation at room temperature until DNA extraction.
For cytological examination, all samples were stained using the modified Papanicolaou technique [34]. Results from cervical cytology specimens were reported according to the 2014 Bethesda Classification System for reporting cervical cytology [35].

Genomic DNA extraction
Genomic DNA from cervical swabs was extracted using commercial spin-column based QIAamp Mini kit (QIA-GEN, Hilden, Germany), according to the manufacturer's instructions. The DNA was aliquoted in duplicate (25 µl each into separate 2 ml Eppendorf tubes) and stored at − 20 °C until further analysis. The concentration of extracted DNA was determined by spectrophotometry at 260 nm (Nano-Drop 2000c spectrophotometer, Thermo Fisher Scientific, USA). All samples were pre-screened with the human β-globin primers PCO3+/PCO4+ to assess sample integrity [36]. Briefly, for a PCR volume of 25 ml, 1-4 ml of the DNA lysate and 25 pmol of each human beta-globulin consensus primers PCO3+ and PCO4+ (Integrated DNA Technologies, Inc, USA) were used. Purified DNA was used in PCR as templates to amplify target regions.

DNA HPV analysis
HPV-DNA detection and identification of the genotypes was carried out by nested multiplex PCR (NMPCR) as described by Sotlar et al. [37]. Briefly, a single consensus forward primer (GP-E6-3F) and two consensus back primers (GP-E7-5B and GP-E7-6B) were used for the general primer PCR. The PCR reaction mix of 50 µl contained 10X PCR buffer, 2.5 mM MgCl 2 200 µM of each of the four deoxyribonucleoside triphosphates (dNTP), 15 pmols of each E6/E7 consensus primers and 1.25 units of Taq polymerase enzyme (New England Biolabs Inc., UK). Four microlitres (4 µl) of DNA extracts was used as template for the amplification reactions. This was carried out using a thermal cycler (Stratagene Robocycler Gradient 96, Roche Molecular System Inc, USA). The cycling parameters for the first round PCR with E63F/E75B/E76B consensus primers were as follows: 94 °C for 4 min (initial denaturation), followed by 40 cycles of 94 °C for 1 min (denaturation), 40 °C for 2 min (annealing), 72 °C for 2 min (extension) and a single final elongation step of 72 °C for 10 min. In the second round PCR, 2 µl of first round PCR product, 15 pmols of forward and reverse primers for genotyping were used. Primers for the identification of high-risk genotypes 16,18,31,33,35,39,45,51,52,56,58,59; probable high-risk genotypes 66 and 68, and low-risk genotypes 6/11, 42, 43, and 44 were used in four cocktails, each containing four to five different primer pairs. The other parameters that were used in the first round PCR mix were maintained. However, the cycling parameters were as follows: 94 °C for 4 min followed by 35 cycles of 94 °C for 30 s, 56 °C for 30 s, 72 °C for 45 s and a single final elongation step of 72 °C for four minutes.

Detection of PCR products
The amplified products were detected by agarose gel electrophoresis (2%), containing 0.5 µg/ml EZ-Vision ® Bluelight DNA dye (AMRESCO, LLC USA). Ten microlitres of each sample was added to 2 µl of orange G (5X) gel loading dye for the electrophoresis. Hundred base pair DNA molecular weight marker (New England Biolabs Inc., UK) was run alongside the PCR products. The gel was prepared and electrophoresed in 1X TAE buffer using a mini gel system at 100 V for one hour. The gels were viewed in a benchtop UV illuminator (UVP, LLC, Upland, CA, USA) and photographed using Canon camera Sx230 HS.
All methods were carried out in accordance with relevant guidelines and regulations.

Statistical analysis
Data obtained from the questionnaire was checked for accuracy, stored in Microsoft Excel 2010 software (Microsoft Corporation, Redmond Campus, Washington DC, USA) and analyzed using the Statistical Package for Social Scientists (SPSS) version 22. Qualitative variables were described by simple counts and percentages. The confidential interval (95%CI) for the prevalence was determined using the Cochran formula [31]. Quantitative variables were represented as mean ± SD. Any HPV infection/overall HPV infection was defined as any or inclusive of all 18 HPV types. HPV genotypes were grouped according to the licensed vaccines: HPV-16/18, HPV-6/11/16/18 and HPV-6/11/16/18/31/33/45/52/58. Type-specific reporting on HPV genotypes accounted for each infection independently in women with multiple infections. The distribution of HPV genotypes was summarized using frequency distributions. The relation of HPV genotypes with demographic, gynecological and behavioral variables were examined by logistic regression. A p value < 0.05 was considered statistically significant.

Study population characteristics
In total, complete the consent process and questionnaire. 317 women were screened for HPV genotypes after documenting informed consent and questionnaire administration. Analysis of the study population (n = 317) by sociodemographic characteristics revealed that more than half (67.8%: 215/317) of the respondents were between ages 25 and 44 years, married or cohabiting (64.7%: 205/317) ( Table 1). Most were economically active (86.8: 275/317) with an educational level described as senior high school and higher (59.3%: 188/317). Concerning their sexual and reproductive characteristics, most (76.9%: 244/317) of the respondents were multiparous women who had their first pregnancy after age of 18 years (73.2%: 232/317) and had their first sexual contact before age 20 years (60.9%: 193/317).

Associated risk factors
The results were further analyzed for cross-sectional associations between infection with HPV and sociodemographic, obstetric and behavioral characteristics, as well as the development of squamous intraepithelial lesions (Tables 5 and 6). Potentially significant associations were observed between infection with HPV and educational background (p < 0.001), age of first sexual intercourse (p = 0.016) and age of first pregnancy (p = 0.028).

Discussion
Epidemiological data on circulating type-specific HPV prevalence in a population is a very important rationale for introducing HPV vaccines and can provide an early measure of vaccine impact. In particular, it is important in countries where organized screening programs for the prevention of cervical cancer have not yet been established at scale. We conducted a multicenter descriptive study of the prevalence of HPV infection, and the vaccine-related genotype distribution among  [40]. In general, differences in HPV prevalence reports may also be explained by the differences in study methodology: more specifically, the case volumes, the type of the case groups, study population and the method employed for HPV detection [43]. The present study relies on a highly-sensitive nested multiplex PCR (NMPCR) assay that combines degenerate E6/E7 consensus primers and type-specific primers for 16 HPV genotypes. This assay has been tried and proven to give consistent and reliable results when followed [37] and has been used in a couple of studies in Sub-Saharan Africa [34,42,44]. Polymerase chain reaction-based assays showed HPV prevalence of 40% in rural Mozambique [41], 31% in Harare, Zimbabwe [45], and 44% in Nairobi, Kenya. These figures suggest similar endemicity of HPV infection in the region.
Higher HPV prevalence may also be attributed to the nature of the study population and age range. Women of reproductive age may be more sexually active than women of other age-groups such as adolescent girls and elderly women. Available epidemiological data on HPV in high-risk population groups promote the idea that overrepresentation of high-risk women in general population studies may raise the observed prevalence. For instance, among HIV-positive and HIV-negative women recruited at the Cape-Coast Teaching Hospital, prevalence of HPV infection was reported as 75% and 42.6%, respectively [46]. In addition, HIV-positive women have higher rates of persistent HPV infections [40,46].
Infection with more than a single genotype is a common feature of HPV infections [3,47]. Cervical coinfection with multiple HPV types was observed for both HR HPV and LR HPV infections, and was supported by previous studies in Ghana [40,42,48] and sub-Saharan Africa [39,49,50]. Although co-infection of HPV genotypes occurs very frequently, the evidence suggests that the presence of multiple types does not especially influence clearance of HPV infections either by natural mechanism or vaccine-induced immunity [51,52]. Rather, multiple infections occur as independent events sharing common transmission routes and having a similar profile of risk factors [53].
Additionally, the prevalence of HPV-16 and HPV-18 was 4.4% and 4.1% respectively. The higher prevalence of HPV-18 over HPV-16 in this study is similar to previous reports among available local studies in normal cervixes [40,42,55] and cancerous tissue [44,56]. Although the

HPV genotype
High-risk proportions of specific genotypes differ, similar reports confirm that the most common genotypes detected in this study are equally prominent across other African countries both in women with normal cytology and in those with HSIL or worse [38,41,45,49,57]. It is possible that local differences in HPV distributions actually exist and must be continuously investigated in order to better understand the complex interplay of factors that shape HPV distributions in populations. However, it is also plausible that type-specific HPV prevalence may be influenced by the type of assay used and by the preponderance of multiple HPV infection in certain populations [45]. An important question which population HPV studies seek to answer pertains to the frequency of detection of HPV-16 and-18, the two most common high-risk genotypes prevented by available vaccines. The combined prevalence of HPV-16 and 18 in this study is similar to estimates from other local studies from Kumasi (6.2%) [32] and Accra (6.6%) [42]. The result of this study is consistent with other studies, which reported lower prevalence of these two vaccine-preventable genotypes in the general population compared to studies involving histologically confirmed cancer tissue [58][59][60]. These findings support the knowledge that HPV16 and HPV18 are mostly under-represented in women with normal cytology by comparison with their importance in severe cervical lesions and underscore their epidemiological reputation as more aggressive carcinogenic agents and justify the use of preventive vaccination in cancer prevention [5,61,62]. Since 2015, the Advisory Committee on Immunization Practices (ACIP) recommended an FDA approved nonavalent (9-valent) human papillomavirus (HPV) vaccine (9vHPV) (Gardasil 9, Merck and Co., Inc.) containing HPV-31, 33, 45, 52, and 58 VLPs in addition to the quadrivalent vaccine coverage [17,18]. This new vaccine is hoped to expand the range of existing vaccines and protect even more women from HPV infections. Gardasil 9 vaccine includes five more HPV genotypes (HPV-52, 58, 45, and 31) covering the most common HR-HPV types found in our study among Ghanaian women. The high prevalence of nonavalent vaccine-preventable types coincidental with abnormal cytological findings taken together with previous reports of the distribution of vaccine preventable HPV genotypes in malignant cervical tissue [56,58] show that expanded genotype vaccines may be more beneficial than previous vaccines in this population.
Earlier age at sexual debut puts women at greater risk for HPV infection [63]. An association between young age and cervical HPV infections is generally attributed to a higher susceptibility to the infection at the beginning of sexual activity, with peak prevalence in younger women and progressive decline with increasing age [64].
In this study we found that although there was a high burden of HPV infection in young adult women (18-25 years) (Additional file 1), the overall prevalence of infection from vaccine-type HPVs in sexually active women was low in the age group less than 25 years (0% for HPV-16/18, 0% for HPV-6/11/16/18 and 1.3% (4/318) for HPV-6/11/16/18/31/33/45/53/58) and a wide margin of interventions with vaccine primary prophylaxis beyond the preadolescent girls aged 9-12 years, who are actually the target for vaccine, could be expected.
In this work, an important socio-demographical risk factor associated with of HPV DNA infection was education. In general, low educational status is thought to indicate poor knowledge and compliance for safer sexual behavior; including number of sex partners and use of protective condoms [65]. Our results suggest that higher education may not necessarily be indicative of knowledge of HPV or safer sexual behavior. Our data showed that women who were active (have any job, government or private) had 2 times higher risk to be infected with HPV, than those who were not active (do not have any job). But our study did not look at the association of prevalence of HPV infection with socio-economic status of the women. Though this contradiction would be hard to explain, it may be related to difference in a risk behavior correlated with education.

Limitations
The study used a convenient hospital-based sampling. The benefits of this approach include time and costsavings. However, there are many setbacks with this approach to sampling. These include a high chance of bias and the applicability of findings to other population groups such as community-dwelling women or the general population of women. Bias was minimized by excluding previously screened women and skipping every third woman presenting to the clinic. These factors should always be borne in mind when interpreting study results.

Conclusion
Vaccine preventable high-risk HPV types represent 8.2% (HPV-16 and 18) and 9.1% (HPV-6, 11, 16 and 18) of the burden of HPV infections in the population studied. Although the impact of universal vaccination with existing vaccines would be much greater with data from women with pre-cancerous lesions, our data shows that the nonavalent vaccine targeted at HR HPV-16, -18, -31, -33, -45, -52 and -58, would cover 28.4 0% of HPV infections. As the country prepares to achieve 90% universal vaccine coverage, the role of expanded-scope vaccines that confer immunity against region-specific oncogenic HPV types can be essential. It is an urgent need to introduce the nonavalent vaccine to the country at scale.