Genetic diversity of human papillomavirus (HPV) as specified by the detection method, gender, and year of sampling: a retrospective cross-sectional study

This study assesses HPV prevalence and genotype distribution in Lebanon, and identifies differentials in HPV infection, infection with multiple genotypes, and with high-risk genotypes, by sex, age, and year of data collection. Study participants comprised 1042 female and 160 male participants between 2006 and 2018. HPV genotyping was done by PCR and hybridization (2006–2013) or real-time PCR (2013 onwards). Diversity of HPV genotypes across gender, age groups, and years of data collection was tested by applying Shannon Diversity Index. The overall HPV prevalence was 44.8% among study participants, and threefold higher in women than men. Single HPV infection was seen in two-third of HPV-positive participants. Women were less likely to be infected with multiple HPV strains, but more likely to be infected with high-risk or mixed-risk HPV genotypes. HPV-16 (11.0%, 9.8%) and HPV-53 (8.5%, 4.9%) were the most prevalent high-risk HPV genotypes in women and men, respectively, while HPV-18 prevalence was 4.9% in men and 3.1% in women, while HPV-59 prevalence was 6.6% in men and 2.1% in women. Samples collected post-2011 from women showed twice higher odds of HPV infection than those collected earlier and were threefold more likely to be infected with multiple HPV strains, and twice more likely to be infected with high-risk genotypes compared to those tested earlier. Women scored higher on Shannon index indicating high diversity in HPV types and frequency, with trend of increased diversity over time. While the odds of HPV infection remained associated with sex and temporal trend in multivariable analysis, odds of having high-risk genotypes was mainly associated with infection with multiple HPV strains. Our study showed high diversity in HPV genotypes and an increasing trend of infection with multiple and high-risk genotypes in recent years. Findings underscore the need for effective screening/surveillance and HPV vaccination programs.

New categorization based on complete genome analysis, divided HPV types into lineages (1-10%) and sublineages (0.5-1%) linked with differential infection prognosis [22,23]. Disruption of HPV proteins is key for emergence of new HPV types related to papillomaviruses phylogenetic assembly [24], while sublineages possess biological characteristics consistent with persistent infection and cervical cancer [23]. Despite the increased HPV infection prevalence among women and men in Lebanon [5,25], and the heightened risk of progression to cervical cancer, comprehensive data on the prevalence and distribution of HPV genotypes in Lebanon are sparse. Here, we assess HPV prevalence and genotype diversity, and examine associations with HPV infection, infection with multiple genotypes, and with highrisk genotypes in asymptomatic men and women between 2006 and 2018.

Study design and data collection
Between 2006 and 2018, HPV testing information were collected for 1202 participants, of whom 1042 (86.7%) were females seen at OB/GYN and Urology outpatient clinics. Patients' socio-demographic information were retrieved through review of medical charts. HPV genotyping was warranted in case of recent abnormal Papanicolaou test, after immunization with HPV vaccine (Cervarix Gardasil-4, Gardasil-9), known HPV carrier, self-declared multiple sexual partners, or partner with multiple sexual partners, before starting new relationship, and after breaking highrisk relationship. HPV DNA testing was also indicated in non-protected sex, vulvar warts, extra conjugal affair, or concern about potential sexually transmitted infections (STI). Inclusion criteria for females were ≥ 21 years of age, presence of vaginal portion of uterine cervix, sexually active (vaginal, anal), willing to undergo primary HPV high-risk (PCR) screening individually or as co-test, testing concern of potentially having STI. Exclusion criteria were young age (< 20 years), pregnancy, history of cervical lesions, and current/recent history of radiotherapy, chemotherapy, or surgical treatment. This study was performed in line with the principles of the Declaration of Helsinki

HPV genotyping
Cervical and vaginal scrapes were collected from females using Ayre spatula, and genital scraps (urethral epithelium, penile epidermis) were collected from males using nylon flocked swabs and placed in sterile saline. Between 2006 and 2010, total genomic DNA was extracted from isolated scrapes by the mini-spin column method using DNeasy Blood and Tissue kits, per manufacturer's instructions (Qiagen, Courtaboeuf Cedex, France). From 2011 onwards, genomic DNA extraction was done using MagPurix automated extractor (Zinexts Life Science Corp., Taiwan). Between 2006 and 2013, HPV genotyping was done by PCR followed by hybridization, using AID HPV TYPING PCR kit (AID AUTOIMMUN DIAGNOSTIKA GMBH, Strasberg, Germany). This involved PCR, line probe assay, and manual reading using AID Scanner (AID AUTOIM-MUN DIAGNOSTIKA). Control bands were present on all strips for validation of DNA extraction, amplification, and hybridization. From 2013 onwards, real-time (RT)-PCR was used for HPV genotyping, using automated DNA extraction (PREP-NA kits; DNA Technology, Moscow, Russia), followed by RT-PCR using cycling conditions provided by the manufacturer (DNA Technology).

Statistical analysis
Analyses were performed in Stata/SE 15.1 (College Station, TX). Frequency distributions were used to describe participants' baseline characteristics stratified by sex. Bivariable analyses applying χ 2 test and univariable logistic regression were conducted to associations with HPV positivity. Additional bivariable analyses were conducted on subsamples, including only HPV-positive patients, to examine differences in the type and risk classification of HPV genotypes. Odds ratios (ORs) and relative risk ratios (RRRs) with corresponding 95% confidence intervals (CIs) were reported Frequency distributions were used to describe HPV genotype distribution (frequency of specific genotype relative to all identified genotypes), stratified by sex. Study participants testing positive for multiple genotypes, contributed separately to quantitative estimation of each identified genotype. HPV genotype diversity by sex, age group, and year of data collection was assessed by Shannon Diversity Index [26] to resulting distributions, and using data on the global HPV genotype diversity [27]. Univariable and multivariable regression analyses were performed to identify associations with HPV infection. Factors with P value ≤ 0.1 in univariable analysis were included in the multivariable model, which was conducted to assess the contribution of sex, age, and temporal trend to the observed variability in HPV infection. The strength of association with HPV infection was determined using adjusted ORs (AORs), 95% CIs, and P value ≤ 0.05. Similar analyses applying multinomial logistic regression were performed to investigate the association of sex, age, temporal trend, and typology of HPV genotype (single/multiple) with high-risk HPV genotypes in HPV-positive patients.
Sensitivity analysis was performed to assess the robustness of study results to changes in HPV testing post-2011. Data were restricted to study participants undergoing testing post-2011 before re-conducting the bivariable analyses examining differences in type and risk classification of HPV genotypes, and regression analyses examining associations with HPV infection and with high-risk HPV genotypes in HPV-positive study participants.

HPV prevalence among study participants
Data were available for 160 men and 1042 women (Table 1). Among men, over two-thirds were 30-49 years, 43.1% of whom in 30-39 years age group, compared to women who were slightly older (32.3% aged 30-39 years, 24.7% aged 40-49 years, and 21.8% aged 60+ years). HPV prevalence was 44.8%, with the odds of infection being threefold higher in women than in men [P < 0.001; OR (95% CI), 3.2 (2.2-4.8)]. No significant differences in HPV positivity were noted among men across age groups, or years of data

Status of HPV infection
Most HPV-positive men (61.1%) and women (66.2%) were infected by single HPV strain (

Distribution of HPV genotypes
HPV-16 (11.0% and 9.8%) and HPV-53 (8.5% and 4.9%) were the most prevalent high-risk HPV genotypes in women and men, respectively (Fig. 1). By comparison, the prevalence of HPV-18 was 4.9% in men and 3.1% in women, while HPV-59 prevalence was 6.6% in men and 2.1% in women. Assuming equal distribution of HPV genotypes [27], the largest Shannon Diversity Index value was 5.28 [26]. Women scored higher than men on this index (56.3%, 47.8%) indicating high diversity in the types and frequency of HPV genotypes (Table 3). Higher diversity was observed in younger (19-29 years) than older (60+ years) participants (57.4%, 51.1%), and trend of increased diversity over time was noted, as samples from recent years had over 55% diversity compared to samples from earlier years.  (Table 4). Older age, recent years of data collection, and infection with multiple HPV genotypes were associated with high-risk genotypes in univariable analyses (Table 5). Of these, infection with multiple HPV genotypes was the only factor that remained associated with high-risk genotypes in the multivariable model [P < 0.001; ARRR (95% CI), 7.6 (3.8-15.0)], indicating that it was predictor of high-risk genotypes.

Sensitivity analysis
Sensitivity analyses conducted using post-2011 data affirmed the robustness of results, specifically significant decrease in the probability of multiple HPV strains, and of high-risk or mixed-risk genotypes with increasing age (Table S1).
Further analysis showed increased frequency of high-risk and mixed-risk genotypes per year post-2011 (Table S1). Whereas the odds of HPV infection remained strongly associated with sex in multivariable analysis, no evidence for variability in prevalence was found in 2011-2018 (Table S2). There was evidence for increased probability of having high-risk genotypes in participants infected with multiple HPV strains, and over year-by-year, starting from 2011 (Table S3).

Discussion
Few studies reported on HPV genotype diversity in MENA region, and fewer examined HPV genotype distribution among men in Lebanon [28]. This study contributed to filling this gap through assessment of HPV prevalence and genotype diversity and association with HPV infection, infection with multiple and high-risk genotypes in asymptomatic subjects. Study showed high diversity in HPV genotypes and increasing trend of infection with multiple and high-risk genotypes in recent years, highlighting the need for sexual health, screening/surveillance, and targeted HPV vaccination. Cervical and penile samples provided to accredited laboratories documented HPV prevalence of 44.8%, with threefold higher odds of infection in women. This high diversity is attributed to liberalization in sexual norms in Lebanon and other MENA countries [29,30]. This was partially comparable to Iranian study, which documented 53% HPV positivity in 5176 cases [31], with notable difference that HPV prevalence was higher among men, and among younger (20-40 years) women [31]. In contrast, a recent smaller Chinese study reported lower incidence of HPV among men, with no effect of age groups, or data collection years on its prevalence [32]. This varied prevalence of HPV among men and women are attributed to factors including high-risk group screening [33], method used [33,34], and availability of screening facilities.
Earlier HPV genotyping relied on the universal MY09/11 primers [18,35,36], and detection by DNA Enzyme Immu-noAssays, RFLP, RT-PCR, or sequencing [18,[35][36][37]. We compared HPV incidence and genotype profile of specimens tested by PCR-hybridization (2006-2013) with those tested by RT-PCR (post-2011), as reliance on one method may underestimate the HPV prevalence in clinical samples [38,39]. Significant increased likelihood of infection with multiple HPV strains and high-risk genotypes were noted, in agreement with recent studies which documented varied prevalence of HPV, and novel HPV genotypes based on detection method [37,40]. This recommends selection of the most appropriate HPV detection based on high specificity and sensitivity, and cost for developing/low-income countries [34,38,39,41]. Effective HPV screening depends on the genotyping method used [41,42]. Women tested post-2011 were threefold more likely to be infected with multiple HPV strains, and twice more likely to be infected with high-risk genotypes, compared to women tested pre-2011. These differences are attributed to increased sensitivity, and broader detection capacity of RT-PCR compared to conventional (strip-based) hybridization [31,41,42]. This was in agreement with Chinese [43] and Turkish [38] studies, which documented superiority of RT-PCR in detecting variants not captured by hybridization assays, and in determining HPV viral loads, thus serving as quantitative marker for diagnosing and treating single and multiple HPV infections.
Women were less likely to be infected with multiple HPV strains than men, suggesting high-risk behavior among men. This was consistent with evidence implicating positive spouse as the main source of STI among women in the region. HPV vaccination among men remains limited. Our study suggests that curtailing HPV infections and sequelae entail designing sexual health and vaccination programs, specifically tailored for men.
We evaluated HPV genotypic diversity among men and women from the large and diverse collection of HPV-genotyped samples. While HPV-16 and HPV-53 were the most prevalent HR genotypes among women, higher prevalence of HPV-18 and HPV-59 was noted in men. As women, particularly younger subjects, scored higher than men on Shannon Diversity index [26], this confirms high diversity and frequency of HPV genotypes, in agreement with recent studies on gender-dependent distribution of HPV genotypes [31,44]. We demonstrated gender-dependent distribution of select HPV genotype, rather than functional genotypic distribution, in contrast to Iranian study documenting high prevalence of HR genotypes in females, and high prevalence of LR genotypes in males [31], and Czech study that suggested higher prevalence of HR and LR HPV genotypes in males [44], These are attributed to anatomical factors (anal, genital, oral) [44,45], method of detection [34,38], and sample size [31,44].
Significant increase in HPV positivity was noted in 40-49-year-old and elder (≥ 60 years) women. This was in disagreement with German [46] and Chinese [47,48] studies, which suggested that peak HPV positivity was highest in younger women (20-22 years) linked with earlier sexual activity with (unprotected) risky behavior, and with Iranian study suggesting 30-32-year peak HPV infection in males and females [49]. It is noteworthy that the German study was limited to women up to 30 years of age [46], while the study of Wang was based on uneven age group distribution [48]. Later peak of HPV positivity seen in 55-59-year category attributed to reduced immunity at menopause [47,50] was also reported. Lebanon was considered conservative regarding sexual norms, and low incidence of sexually transmitted infections (STI) was reported. Recently, the incidence of HPV infections has significantly increased, prompting nationwide campaigns about STI, and protective measures and tools, including HPV vaccines. Gardasil ® 4/9 (Merck) and Cervarix ® (GlaxoSmithKline) are the current HPV vaccines available in Lebanon [51]. As there is no nation-wide adult HPV vaccination program, and with HPV vaccination not covered by private insurance companies, HPV vaccines are prescribed based on the requirements of the young woman and/or her family [6,52]. This has resulted in a very low rate (2.5%) given that the vaccine is available only on-demand in private pharmacies and some doctors' clinics [52].
In summary, this population-based cross-sectional study provides needed data on the distribution of HPV genotypes, along with prevalence of infection with multiple genotypes and high-risk genotypes, and factors associated with HPV infectivity namely sex, age, and year of data collection. The diversity of HPV types shown here is consistent with other nationalities and ethnic groups, and underscores the need to for routine screening and vaccination, and development of newer HPV vaccines covering a larger panel of HPV genotypes. There are some limitations to this study. The number of male participants was low, hindering conducting further stratified analyses. The study took place in Greater Beirut, where participants had easier access to HPV genotyping, with almost no subject from rural areas which lack the testing facilities. This has limited the generalizability of our findings. Furthermore, its retrospective nature limited assessment of the outcome of HPV infection. Future studies involving larger number of subjects and from different regions of the country, in particular HPV-vaccinated individuals, are of paramount importance towards assessment of HPV infection scope, which is needed for effective national vaccination.