Human papillomavirus prevalence in pregnant women living with human immunodeficiency virus infection: a scoping review of the literature

Introduction

Worldwide there are in the year 2021 nearly 1.3 million pregnant women living with HIV (WLWH) [1]. The World Health Organization (WHO) estimate that globally 81 % of pregnant WLWH are currently under antiretroviral therapy (ART) [1]. Around the world, these women make up a small proportion of the global population. Nevertheless, these women are high-risk group for sexually transmitted diseases, such as the human papillomavirus (HPV) and its consequences of cervical dysplasia up to cervical carcinoma.

According to the German guidelines for the prevention of cervical carcinoma, primary prevention of cervical carcinoma includes HPV vaccination for boys and girls from the age of 9–14 years. Secondary prevention includes annual Pap smears for women up to the age of 35 and HPV testing and cytology every three years from then on. In case of cytological abnormalities or high-risk HPV (HR-HPV) positivity, colposcopy should be performed [2].

However, WHO-, European – or German guidelines have not yet established extended and more intensive screening for pregnant WLWH, although, as some studies have already shown, this is a highly vulnerable group with regard to HPV positive dysplasia and cervical carcinoma [2, 3].

Fortunately, more and more studies are pointing out this gap and the lack of standardised guidelines for pregnant WLWH [4, 5]. Few studies on pregnant WLWH worldwide are the reason for a not yet standardized approach. This scoping review was conducted to highlight the importance of expanded and standardised screening in relation to the prevention of cervical cancer in these high-risk women.

Inclusion and exclusion criteria

For inclusion, articles had to conform to the following criteria: 1) study published in English; 2) original study with clear HPV prevalence among pregnant HIV-coinfected women; 3) detailed description of study population; 4) information about the HPV detection method. Only studies which used the highly sensitive Polymerase Chain reaction (PCR) method for HPV detection, were considered; 5) studies had to present a genotyping profile of the HPV infections. Studies focusing on case reports, reviews, abstracts, commentaries as well as studies without the triple combination of HPV, HIV and pregnancy were excluded.

Literature search

Extracted data are presented in Tables 1 and 2. Furthermore, data will be discussed and analysed in a narrative approach. Two databases were screened for identifying articles, PubMed and EMBASE Ovid. For searching process on PubMed no filters or limitations were used. Searching process on EMBASE Ovid were made under “Multi Fields Search” section and by selecting the search area: EMBASE Classic and EMBASE. Due to high number of results during searching process via EMBASE and the fact that detected and assessed articles via PubMed were published between 2001 and 2021, EMBASE search was limited by only including publications between 2001 and 2021. By using the key words or Medical Subject Headings (MeSH) terms listed in Table 3, 329 titles/articles were found in both databases, PubMed and EMBASE combined.

194 titles/articles were found and verified from PubMed database (Figure 1) [6]. 42 articles were duplicated and excluded. 131 articles were excluded after screening of titles and abstracts by non-fitting to our topic and inclusion criteria. 21 articles were screened by reading the full study text, after which 7 studies were finally included in the review. In the EMBASE Ovid Database 135 titles/articles were found and 108 directly excluded. From 27 potentially relevant articles 12 were duplications. The residual 15 articles from EMBASE Ovid were already screened on the database PubMed. So no further articles found on EMBASE Ovid were included (Figure 2) [6]. Also, in the screen of the reference lists no additional article was included. The final search results were exported into EndNote. The search strategy and inclusion criteria were confirmed by a second reviewer. Finally, full text of potentially relevant articles for the scoping review were reviewed by two separate reviewers and discussed before data extraction. Evaluation of full text was based on inclusion criteria and on availability of necessary data to answer to the leading question of this review. Disagreements were resolved by discussion in an expert guided setting. Due to the particular and specific topic of this review and the hence small existing publication repertoire in the online Databases, we did not explicitly search for additional grey literature.

Figure 1:

Flow diagram; selection of studies for inclusion, searching algorithm PubMed [6].

Figure 2:

Flow diagram; selection of studies for inclusion, searching algorithm EMBASE OVID [6].

Data extraction

Data extraction was performed by one reviewer. A data charting form was developed by three authors to decide which relevant information to extract from the included articles. Data extraction was continuously discussed in an interactive process. We extracted the information on publication characteristics, study characteristics, HPV data, Clinical data, HIV data and sociodemographic factors according to the PRISMA Extension (ScR) 2018 from each included article (Table 4) [7].

For the HPV data, the prevalence and distribution of the LR-HPV genotypes 6,11,32,34,40,42,43,44,54,55,61,62,67,69,70,71,72,74,81,83,84,85 and the HR-HPV genotypes 16,18,31,33,35,39,45,51,52,53,56,58,59,66,68,73,82 were extracted.

Prevalence is shown by pooling the number of positive women divided by the total number of women tested from selected studies. Additionally, percentages were calculated. As part of descriptive statistics, we calculated the weighted arithmetic mean wherever useful by relating the indicated percentages from the studies to the total identified cases of HPV from the studies. The studies were grouped by countries and evidence levels are stated.

Discussion

This big gap between the detected HPV prevalence among the studies could be influenced by several risk factors described in the following.

The different PCR methods used for HPV detection in the studies could had an influence for the variability on the detected HPV prevalence due to different sensitivity and specificity among the methods. The gold standard for HPV detection is the PCR method or the Hybrid Capture 2 (HC2) method or even a combination of both [15]. The Inno-LiPA HPV Genotyping Extra Assay and the combination of amplification methods (PCR with MY09/MY11 primers and HC2 method) are one of the most sensitive methods for HPV detection. Studies included in this review, which used this methods detected the highest HPV prevalence (Table 1) [8, 9, 16, 17].

Single PCR has lower sensitivity than semi-nested PCR [15, 18]. Among the included studies low HPV prevalence was seen when single PCR method was used for HPV detection (Table 1) [9, 14].

Geographical- and Sociodemographic aspects of the populations reasons:

Sociodemographic factors are from one side interesting to evaluate the possibility of comparability among the population and on the other hand to identify potential influencing factors for HPV and HR-HPV prevalence. Host factors which might assume a role in this progress are for example, young age of first sexual intercourse, multiple sexual partners, history of sex work, past use of oral contraception, low use of condom, immunodeficiency syndromes like HIV and positive smoking anamnesis (Table 2) [12, 14, 19], [20], [21], [22], [23].

Additionally, HPV genotype prevalence and frequency can differ between countries and regions [24].

The highest HPV prevalence worldwide is found in Sub-Sahara Africa, Eastern Europe and Latin America [24]. Dividing the included studies by geographical regions, a heterogeneity of the estimated HPV prevalence could be seen. Overall, in this review, three studies from South America, i.e. Brazil, three studies from East Africa, i.e. Uganda, Kenya, Tanzania and one study from Southeast Asia, i.e. Thailand, were included. The fact that the two studies from Brazil reported data from the same study population limits the representation of data for pregnant WLWH in Brazil [12, 13]. One of the lowest HPV prevalence was found in the study conducted in Thailand [14].

Reasons for that could be lower number of sexual life partners among Thai women in comparison to women coming from other countries [25]. Pregnant WLWH included in the study from Thailand by Bollen et al. reported in 38.7 % of cases to had only one sexual partner in life or rather 37.9 % indicated two [9]. Pregnant WLWH from other studies and hence other countries reported on average more sexual life partners, when this parameter was asked (Table 1). In the studies performed in Brazil more than half of pregnant WLWH (precisely 62.2 , 66.2, 72.5 %), declared to had more or equal to three sexual partners) (Table 2) [10], [11], [12].

Although the study from Thailand had one of the lowest HPV prevalence among the included studies, detected HPV prevalence is still higher than in other studies comping from Thailand, where women without HIV and pregnancy had been analysed [14, 25]. History of sex work which was reported in 9.4 % of pregnant WLWH in the study by Bollen et al. could also had an influence of their HPV prevalence (Table 2) [14].

As widely known, HIV and HPV are generally more diffused in low-and middle-income countries, i.e. countries from the African continent or Brazil [3, 8, 10]. A lower population education level, insufficient elucidation, and the lack of adequate screening programmes, are factors which increase additionally the diffusion of HPV and could have boosted detected HPV prevalence [3, 8, 10]. 89 % of people living in Low- and Middle-Income Countries, have not heard about HPV infection, transmission, risks and protection [26]. In the study by Brandao et al. less education level as well as the low-income status of 82.4 % and hence missing elucidation and knowledge about HPV could have boosted the detected HPV prevalence of 96 % (Table 2) [10].

HR-HPV genotypes prevalence could also vary by different geographical regions. As already mentioned, HR-HPV52 was with 24.3 % the most frequent genotype among our studies (111/457 women) (Table 1). HR-HPV52 is considered as the second most common HPV genotype in women from the African continent [24]. HR-HPV52 was the most frequent HPV genotype in the two studies performed in Sub-Sahara Africa region (Table 2) [8, 11]. In comparison to other studies from Germany, the genotype HR-HPV52 was also one of the most frequent genotypes due to large part of included participants from the Sub-Sahara Africa region [5, 27].

Hence, knowledge about the origin of pregnant WLWH is important to estimate the risk for an HPV infection and to precise and adapt the HPV genotyping profile.

Another modulating factor on HPV infection is smoking. Previous studies already showed an association of smoking and an elevated risk for any HPV infection [19, 20]. Potentially, immunological effects such as fewer Langerhans’ cells in the epithelium of smokers could be an explanation for increasing risk for HPV infection [21]. In the study by Meyrelles et al. 2012 pregnant WLWH with positive nicotine anamnesis showed higher HPV prevalence than the women without nicotine anamnesis (Table 2) [12].

Overall, comparing the sociodemographic factors of pregnant WLWH among the studies was difficult, because of missing data and inhomogeneous presentation of the sociodemographic factors in the studies.