Human immunodeficiency virus is a driven factor of human papilloma virus among women: evidence from a cross-sectional analysis in Yaoundé, Cameroon

Human papillomavirus (HPV) is the leading cause of cervical cancers, causing 270.000 deaths annually worldwide of which 85% occur in developing countries with an increasing risk associated to HIV infection. This study aimed at comparing HPV’s positivity and genotype distribution in women according to their HIV status and determinants. A comparative study was carried out in 2012 at the Chantal BIYA International Reference Centre (CIRCB) among 278 women enrolled consecutively at the General Hospital and the Gynaeco-Obstetric and Paediatric Hospital of the City of Yaoundé. HPV genotyping was performed by real-time PCR, HIV serological screening by serial algorithm, CD4 T cell phenotyping by flow cytometry and HIV viral load by Abbott m2000RT. Statistical analyses were performed using Microsoft Excel 2016 and Graph Pad version 6.0 software; with P < 0.05 considered statistically significant. Globally, mean age was 37 ± 3 years; median CD4-count for HIV+ was 414 cells/mm3 [IQR: 264.75–588] and median viremia was 50 RNA copies/mL [IQR: < 40–8288]. Overall HPV rate was 38.49% (107/278); 58.88% for single women vs. others (28.97% married, 2.80% divorced, 9.34% for widows), OR: 2.164; p = 0.0319. Following HIV status, HPV rate was 43.48% (80/184) among HIV+ vs. 28.72% (27/94) among HIV- (OR: 1.937; p < 0.0142); HPV genotypes among HIV+ vs. HIV- were respectively distributed as follows: genotype 16 (3.75% vs. 0.00%, p = 0.57), genotype 18 (3.75% vs. 3.70%, p = 1.00), co-infection 16 and others (8.75% vs. 7.40%, p = 1.00), co-infection 18 and others (8.75% vs. 11.11%, p = 0.71), co-infection 16, 18 and others (2.50% vs. 0.00%, p = 1.00) and other genotypes (72.50% vs. 77.78%, p = 0.80). Among HIV+ participants, HPV rate following CD4 was 62.88% (61/97) for CD4 < 500 vs. 35.71% (20/56) for CD4 ≥ 500 (OR: 3.05; p = 0.0012) while HPV rate following HIV viremia was 42.71% (41/96) with < 1000 RNA copies/ml vs. 66.00% (33/50) with > 1000 RNA copies/ml (OR = 0.384; p = 0.009). In Yaoundé, HPV rate appear to be very high, with higher rates of genotypes other than 16 and 18. In the event of HIV infection, the risk of HPV positivity is two times higher, favoured essentially by immunodeficiency. Thus, HIV-infected women should be closely monitored to prevent the emergence of cervical cancer.


Keywords: HPV, HIV, Women, Yaoundé
Background Human papilloma viruses (HPV) are non-enveloped DNA viruses that represent the most widespread sexually transmitted infection worldwide, with more than 200 species. HPV genotypes are responsible for 95% of the occurrence of cervical cancer in women [1]. Notably, HPV causes benign and malignant tumours in humans and viral genotypes that are oncogenic are classified as high-risk (HR-HPV) or low-risk oncogenic HPV (LR-HPV) [2]. Although all HR-HPV genotypes have oncogenic potential, some are more frequently involved in cancers than others. Thus, HPV 16 and 18 represent approximately 70% of all cervical cancers worldwide, followed by HPV 31, 33, 45, 52, 58 and 68 which are among the ten most commonly isolated HR-HPV genotypes from cervical tumor biopsies [3,4].
Indeed, cervical cancer is known to be a viro-induced neoplasm and that HPV is the causal agent in 99% of cases [4]. According to the World Health Organization (WHO), cervical cancer will kill more than 443,000 people worldwide each year by 2030, and 98% of these deaths will occur in developing countries, particularly in sub-Saharan Africa (SSA) where the HIV epidemic and other risk factors further increase the burden of this cancer [5,6]. In SSA, HPV and HIV co-infection is commonly found and contributes to a rapid disease progression toward cervical cancer [4]. Importantly, HIV-positive women, especially those who are severely immunocompromised, are 5 times more likely to develop the cervical cancer than their HIVnegative peers.
Even though little is known on HPV genotypes circulating among HIV-infected women in SSA countries, recommendations for HPV vaccination in the population of people living with HIV (PLHIV) are still unclear [7]. Of note, Generally 66% of infections are cleared at 12 months and this increases to 90% at 24 months [8]. For 10% of HPV-infected individuals, there is viral persistence that remains at latency. In this frame, immunodeficiency may be the cause of viral persistence and reactivation subsequently [7,9].
Factors that contribute to HPV infections include: early sexual activity, multiple sexual partners, smoking, prolonged oral contraception and immuno-compromission (due to HIV essentially) [9]. HIV infection also promotes HPV infection at molecular and cellular levels during different phases of HPV replication cycle such as, penetration into the target cell, HPV multiplication and HPV immune evasion from the host's defences. Direct consequences are clinical manifestations of HPV-associated dysplasia, which has been recognized in early HIV epidemics [10]. As reported by Kevin A. Ault in 2006, High-risk, oncogenic HPV types (including HPV 16 and HPV 18) are associated with 99.7% of all cervical cancers [11]. A recent study conducted in Chad during the year 2018 showed that 11 genotypes of HR-HPV were detected in nearly 70% of the study population, and HR-HPV 58 genotype was the most prevalent (22.9%) [12]. In Cameroon, out of 14,000 new cases of cancer reported in 2016, cervical cancer occupied the second leading position of cancers among women (23.2%) [12,13]. These observations are very concerning and suggest a high burden of HPV in these settings where HIV is also highly prevalent [14,15]. We therefore hypothesised that HPV-infection is significantly higher among Cameroonian HIV-seropositive women as compared to their seronegative peers. Thus, in the light of evidence generated from countries with similar programmatic challenges [14,16], understanding the burden of HPV among women, its circulating genotypes and the effect of HIV-infection on HPV-positivity, might help in shaping HPV preventive measures according to local realities in Cameroon.
Our study objective was to compare the rate of HPVpositivity according to HIV-serostatus, and to determine other predictors of HPV-positivity.

Study design and setting
A cross-sectional study was carried-out in 2012 from women attending the General Hospital and the Gynecoobstetrical Hospital, both located in Yaoundé, the capital city of Cameroon. For all enrolled women, blood and cervical samples were collected. HIV screening tests, CD4 lymphocyte counts, HIV viral load, and HPV genotyping were performed at the Chantal BIYA International Reference Centre for research on HIV/AIDS prevention and management (CIRCB) in Yaoundé, Cameroon (http:// circb.cm/btc_circb/web/). Briefly, CIRCB is a reference centre for HIV/AIDS, performing laboratory analysis with external quality controls and proficiency testing for HIV screening (CDC DTS), HIV early infant diagnosis (CDC PT program), CD4 and viral load (QASI, Canada).

Sampling strategy
Following a consecutive non-probabilistic sampling method, a minimum sample size of 272 was required to meet the study outcomes. Consenting women, sexually active, aged 18 years and above, were eligible for the study. However, pregnant women or those who had undergone a total hysterectomy were not considered for inclusion. After consenting, a standard questionnaire was administered to all participants, covering sociodemographic characteristics, gynaeco-obstetrical and reproductive history. Whole blood and cervical samples were collected; cervico-vaginal smear reading was done; HIV screening was performed following a serial algorithm as per the national guidelines, plasma HIV-1 viral load (solely for HIV+ cases) was done using the Abbott m2000 RT-PCR platform, enumeration of CD4 T lymphocyte was performed by using the FACS Calibur of Becton Dickinson.

Statistical analysis
Data were collected using Excel 2016, and analyses were performed using Epi-info version 7 and Graph pad prism version 6. Odd ratio (OR) was calculated to determine whether the variable was a risk or protective factor. The confidence interval (CI) for the statistical tests was set at 95%, and the null hypothesis rejected at a 5% threshold. Chi square or Fisher-Exact tests were used whenever appropriate. P-values ≤0.2, obtained in uni-or bi-variate analysis, were considered in the multivariate analysis.

Ethics considerations
Ethical clearance was obtained from the CIRCB Ethics Committee (ref N°1810), and administrative authorization was provided by the facilities where the study was conducted. Written informed consent was provided by each participant, and results were delivered free of charge to participants for their clinical benefits.

Immune status of HIV+/HIV-patients
Women with a normal immunity (CD4 ≥ 500) were fewer (35.71%) as compared to those immunocompromised (CD4 < 500) representing 62.88% of the general population. The overall median [IQR] CD4 count was 399 cells/mm 3 . According to HIV-serostatus, median CD4 was lower (354 [231-530 cells] cells/mm 3 ) among HIV-seropositive women as compared to their seronegative peers (414 [IQ 262-918 cells/mm 3 ]). The nonsignificant difference in CD4 distribution according to HIV status reflects the systematic HIV diagnosis performed to all women consenting for participation, most of whom were still asymptomatic.

HPV positivity rate and circulating genotypes
Overall, HPV positivity rate was 38.49% (107/278), indicating a high burden of HPV in the study population

Factors associated with HPV-infection
The age range 20-29 years was the most infected by HPV (51.66%). This latter prevalence was significantly higher when compared with that obtained in those aged 40-49 years (25.00%); p = 0.03. Overall, HPV positivity decreased as participants grow older (Table 2). According to marital status, the highest HPV positivity rate was recorded among divorced women (60%), and this latter prevalence was significantly higher when compared to those that were single (p = 0.04) ( Table 3).
Participants with poor immune status (CD4 ≤ 500) had a significantly higher rate of HPV positivity as compared to those with a normal immunity (62.88% versus 35.71% respectively, p = 0.046). Interestingly, the lower the CD4count, the higher the rate of HPV positivity (Table 4).
After adjusting for CD4, viral load and HIV status, only HIV-infection was 12 times more likely to increase the risk of acquiring HPV-infection (p = 0.0001), as shown in (Table 6).

Discussion
The purpose of our study was to compare the rate of HPV positivity among HIV-positive women with that of HIV-negative women, and to determine predictors of HPV-infection. Though all women participating in this study were living in the centre region of Cameroon, the population living in this region is very cosmoplite, as they come from all regions of the country. Therefore,   [12,17,18].
According to marital status, single women were the most represented (58%), a rate above the national proportion published in 2005 by the Central Bureau of Censuses and Population Studies in Cameroon where single women had a percentage of 41.46% [14]. the difference could be explained by changing overtime, as the social realities in the population could justify the increasing rate of single women [15].
The median CD4 count was lower in the frame of HIV-infection (354 versus 414 cells/mm 3 ] in HIVnegative women. Median CD4 obtained in HIV-free population was surprising; however, this could be justified by the fact that these women attending the health facility (likely due to illed-health for some of them) [14,15,19]. Furthermore, it was previously reported that CD4 levels could vary with different conditions, including but not limited to stress, smoking, menstrual cycle, contraceptive pill, physical activity, etc. [20] The HR-HPV positivity rate was 38.48%, with very low rate of genotypes 16 and 18. 3% for mono-infection with HPV genotype 16 and 4% for mono-infection with HPV genotype 18. This suggests that other genotypes are drivers of HPV-infection in the Cameroonian context and need to be depict among cervical cancer cases, especially in the frame of HIV-infection [14,20]. Our findings are comparable to those found in Central Africa Republic [21], which suggest a predominance of nonvaccine HR-HPV different from those found in Western Europe and North America [22,23]. Interestingly, in a systematic review conducted by Doh et al., the prevalence of pure HPV 16 and 18 was 6.25 and 3.28% respectively, a finding very similar to our report showing evidence of 3% HPV 16 and 4% HPV 18 within the same country [24]. This thererfore stresses the fact that a bivalent vaccine would have very little predictive effectiveness within the population of Cameroonin women. Coinfection with HIV infection is a factor facilitating carcinogenesis associated with HR-HPV infections. Prospective studies have reported a higher incidence of HPV among HIV-positive women compared to HIVnegative women [1,14,16]. Co-infections of several HR-HPV genotypes was also been recorded, similar to previous reports [25]. Multi-infections are often associated with viral persistence, especially if co-present with HPV 16.
HIV is confirmed as a contributing factor to HPVinfection (p < 0.019), also confirmed by multi-variate analysis. As expected, women infected with HIV have a weaker immune status [26], which henceforth increases risk of acquiring HPV [27,28]. This calls for the need to regularly test people with HPV infection for HIV [27].
Younger age are at higher risk of acquiring HPVinfection, likely due to sexual habits [29]. Regarding marital status, divorced women stand at high risk of acquiring HR-HPV. The lack of statiscal power limits the interpretation of this finding. Nonetheless, divorced women might be prone to multiple sexual partners, leading to HPV-infection [3,30,31]. The lack of synergistic effect between the HIV viral load and HPV-infection     [32,33]. After adjusting the different variables based on a multi-variate analysis, HIV status remains the independent factor associated with the risk of infection with this high-risk oncogenic HPV (12 times higher risk), as previously demonstrated [34,35]. Similar to our findings, the study by Tartaglia E. et al. also supports HIV-status as the main independent factor associated with risk of HPV infection among women, likely due to immunodeficiency. In other words, following multivariate analysis, none of the other parameters (except CD4 which is strongly related to the HIV condition) were found to be significantly associated with HPV infection [36].
Our study has some limitations. For instance, the low CD4-count in uninfected women was surpoirsing. This could simply be due to the consecutive sampling strategy that might lead to a certain degree of selection bias. Of note, the number of HIV-uninfected women was low; the presence of other comorbidities or infections such as hepatitis B, hepatitis C, and tuberculosis, could also affect the immune status. Unfortunately, diagnosis of these clinical conditions was not covered within the frame of the study. Further studies could therefore considered these aspects. Also, we have enrolled our patients only in the city of Yaounde and the reference molecular technique used for HPV detection made it possible to identify only genotypes 16 and 18. Given the high rate of other genotypes, it becomes essential to characterize this group. This therefore calls for sequencing for a better molecular characterization toward a rational selection of HPV vaccine candidate for the country.

Conclusion
In this sub-population of Cameroonian women, HPV burden appears to be very high, with higher rates of genotypes other than the commonly known 16 and 18. Interestingly, in the event of HIV infection, the risk of HPV-infection becomes higher, favoured essentially by immunodeficiency. Further HPV gentopic studies, coupled to vaccine trials, would are needed to ensure effective prevention of cervical cancer among HIV-infected women.