Co-existence of Herpes simplex virus type 2 and two other oncoviruses is associated with cervical lesions in women living with HIV in South-Western Nigeria

Background The prevalence of Herpes simplex virus type 2 (HSV-2) in cervical lesions is under-reported, especially in Human immunodeficiency virus (HIV), Epstein-Barr virus (EBV) and Human Papillomavirus (HPV) infected persons. Objectives This study determined the prevalence of viral mono-infections, co-infections and squamous cell intraepithelial lesions (SIL) in HIV seropositive (HIV+) and HIV seronegative (HIV-) women. Methods This study included HIV+ and HIV- women (105 each). Cervical smears and viral antibodies were evaluated by Papanicolaou's technique and ELISA method, respectively. Results The prevalence of HSV-2, HPV and EBV infections, and SIL were higher in HIV+ women (75.2, 41.9, 41 and 32.4%) than in HIV- women (45.7, 26.7, 26.7 and 13.3%) at p< 0.0001, p= 0.029, 0.041 and 0.002, respectively. Higher prevalence of viral mono-infection and tri-infection was observed in HIV+ women (43.8 and 24.8%) than in HIV- women (27.6 and 8.6%) at p= 0.021, and 0.003, respectively. The prevalence of SIL was also higher in HIV+ women with viral mono-infection, bi-infection and tri-infection (15.2, 42.9, and 53.8%) than in HIV- women (6.9, 12.5, and 44.4%) at p= 0.468, 0.041, and 0.711, respectively. Conclusion This study suggests that the high prevalence of SIL in HIV+ women could be associated with viral co-infections.


Introduction
About half a million new cases of cervical cancer (Ca) are reported each year while approximately 49% of these cases result in death worldwide 1 . The reason for the higher prevalence of the disease in developing countries than developed countries is yet to be fully explained. There are varied risk factors associated with Ca but Human Papillomavirus infection remains the major risk factor. However, not all infected women develop the Ca 2 . In West Africa, the prevalence of HPV DNA positive women with normal cytology have significant higher risk for the disease than HPV DNA positive women 13 . Some HPV+ Ca have been found to be positive for HSV-2 antibodies 14 . Interestingly, Bashyai et al. observed increasing antibody titre and prevalence of HSV-2 from LSIL (11%), HSIL (33%) to Ca (40%) 15 . It is believed that HSV-2 associated chronic cervicitis may facilitate EBV entry 16 . The pooled prevalence of EBV-DNA is 3-29, 21-49, and 44-70% in LSIL, HSIL, and Ca, respectively [17][18][19][20][21] . This also implicates EBV in cervical carcinogenesis. This study determined the prevalence of viral mono-infection through tri-infection as correlates of higher SIL in HIV+ women in a developing country.
Interviewer based questionnaire was used to collect socio-economic and clinical demographics: age, marital status, family type, tribe, educational level, residency, religion, occupation, economic status (Low < 18,000 minimum wage, middle= 18,000 to 53000 and High ≥ 54,000) smoking status and alcohol consumption, age at sex debut, parity, sexual behaviour, medical history, number of sex partners, oral sex, type of contraceptives used, vaginal bleeding after sexual intercourse, genital ulcer, itching and burning sensation around the vulva, vaginal discharge, pelvic pain, duration of antiretroviral therapy and uptake of cervical screening.

Data analysis
The sociodemographic data obtained from the questionnaire and the test results were coded as 0 (reference) and 1 (depending on the number of sub-variable) in excel, exported into SPSS (version 23) and analyzed in descending sorting order of categorical targets. Binary logistic regression analysis was used to assess the relationship between some sub-variables. Chi-square/Fisher exact test was used to compare viral infections and Pap smear result (SIL) between HIV+ and HIV-participants in relation to some demographic characteristics. Pearson's correlation was used to assess the relationship between viral infection and cervical lesions. Significant levels were set at p≤ 0.05.

Ethical approvals
Ethical clearances were obtained from State Hospital Abeokuta Research Ethics Services (SHA/RES/ VOL.2/147) and Babcock University Health Research Ethics Committee (BUHREC 353/16) and written informed consents were obtained from participants. All protocols were carried out in line with the guidelines of the ethics committees.

Result
This study determined the prevalence of viral mono-infections and co-infections, and cervical lesions in sexually active HIV+ (mean age= 41.55 ± 11.71 years) and HIV-(mean age= 39.45 ± 11.16 years) participants (p= 0.08). It included HIV+ participants who were receiving highly active anti-retroviral therapy at HIV Testing and Counseling Clinic, State Hospital Ijaiye, Abeokuta. The HIV-participants were apparently healthy women with no history of cervical lesions. The prevalence of pre-cancerous lesions (ASCUS, LSIL and HSIL; figure 1) was significantly higher in HIV+ women (51.3%) than in HIV-women (24.8%) (p= 0.0001). Overall, the prevalence of viral co-infections was higher in HIV+ women (88.6%) than in HIV-women (33.4%) at p< 0.0001. The prevalence of HSV-2, HPV and EBV infections, and SIL were higher in HIV+ women (75.2, 41.9, 41 and 32.4%) than in 26.7,26.7 and 13.3%) at p< 0.0001, p= 0.029, p= 0.041 and p= 0.002). Result showed that there was a significant correlation between viral infection and cervical lesions both in HIV+ women (r=0.363, p= 0.000) and HIVwomen (p= 0.000). However, significant positive relationship between extent of viral infection (mono-, biand tri-infection) and cervical lesions was only seen in HIV+ women (p= 0.040) while insignificant positive relationship was seen in HIV-women (p= 0.326). Statistics revealed that the prevalence of viral mono-infection was higher in HIV+ women (43.8%) than in HIVwomen (27.6%; p= 0.021). However, the prevalence of SIL in viral mono-infection was insignificantly higher in HIV+ women (15.2%) than in HIV-women (6.9%; p= 0.468). No significant difference was observed in the prevalence of viral bi-infection between HIV+ (21%) and HIV-women (22.9%; p= 0.737). Interestingly the prevalnce of SIL in viral bi-infection was significantly higher in HIV+ women (42.9%) than in HIV-women (12.5%; p= 0.041). Although, the prevalence of viral tri-infection was higher in HIV+ women (24.8%) than in HIV-women (8.6%; p= 0.003), no significant difference was observed when the prevalence of SIL was compare between HIV+ (53.8%) and HIV-women (44.4%) with viral tri-infections (p= 0711).

Influence of demographics on prevalence of SIL and viral infections
Descriptive statistics showed that the prevalence of SIL relatively increased with age both in HIV+ and HIVparticipants with the peak prevalence in the age group of 50-59 years. The prevalence of EBV infection increased with age in HIV+ participants while the prevalence of EBV infection peaked in the age group of 30-39 years in HIV-participants and decreased afterwards. A significant difference in EBV infection was observed between HIV+ and HIV-participants in the age range of 50-59 years. The prevalence HPV infection relatively decreased with age in HIV+ and HIV-participants. A significant difference in HPV infection was also observed between HIV+ and HIV-participants in the age group of 30-39 years. A significant correlation was also observed between age and acute HSV-2 infection (r= 0.12, p= 0.03) in both groups. The prevalence of HSV-2 infection relatively decreased with age in HIV+ participants while the prevalence of the virus changes across age groups in HIV-women but peaked in the age group of 40-49 years. Significant differences in HSV-2 infection were also observed between HIV+ and HIV-participants in the age group of 20-39 years. Among the married women and women in polygamous marriages, the prevalence of SIL, HPV and HSV-2 infections were higher in HIV+ participants than their HIV-counterparts. Bivariate analysis revealed that the risk of EBV and HPV infection among HIV+ women in polygamous marriages were 3.26 and 4.67 (95% Cl: 0.28-38.48 and 0.39-55.48) at p= 0.35 and 0.22, respectively when compared with women who have never been married. The odd ratio for both infections were lower among HIV+ women in monogamous marriage  0.79 and 2.00 (95% Cl: 0.35-1.79 and 0.89-4.51) at p= 0.58 and 0.95, respectively when compared with women who have never been married. Among women of Yoruba tribe, HIV+ participants had a significantly higher prevalence of HPV and HSV-2 infections than their HIV-counterparts. The number of participants with post basic (secondary school) education was lower among HIV+ participants than their HIV-counterparts (p= 0.003). The HIV+ participants with only basic education and those living in urban areas had a significantly higher prevalence of SIL, HPV and HSV-2 infections than their HIV-counterparts. The HIV+ women who were Christians had a significantly higher prevalence of SIL, EBV and HSV-2 infections than HIV-women.
Significantly higher prevalence of HPV infection was observed in HIV+ Muslims than HIV-Muslims. The prevalence of participants with low income was higher in HIV+ participants compared with HIV-participants (p < 0.0001). The HIV+ participants who with low income had significantly higher prevalence of SIL, EBV and HSV-2 infections than HIV-participants. The prevalence of SIL and HSV-2 infection were higher in HIV+ multiparous women than their HIV-counterparts. In HIV+ women, although the prevalence of EBV and HPV were higher among those who had first sexual intercourse at ≤ 18 years, these participants surprisingly had lower prevalence of SIL than those who had their sex debut at ≥18years. The HIV+ women with sex debut at ≤ 21 years had higher prevalencef HSV-2 than their HIV-counterparts. The HIV+ debutants at ≤18 years and 19-21 years had higher prevalence of EBV and HPV infections, respectively when compared with HIV-debutants. The odd ratio of HPV infection in HIV+ debutants at 19-21 years is 44.97 (95% Cl: 1.80-1121) at p= 0.02. Participants living with HIV who use only hormonal contraceptives had a significantly higher HPV infection when compared with their HIV-counterparts while HIV+ participants who intermittently use condom had significantly higher prevalence of EBV and HSV-2 infections than their HIV-participants. The HIV+ participants who had multiple sexual partners had significantly higher prevalence of SIL and HSV-2 infection than their HIV-participants. Significantly higher prevalence of HSV-2 infection was observed in HIV+ participants with vulval itching, vaginal discharge and pelvic pain than HIV-participants. The prevalence of EBV and HPV infections in HIV+ women who had HSIL were significantly higher than that of HIVwomen. The prevalence of HPV and HSV-2 is higher in HIV+ women who had cervicitis and ASCUS, respectively than in HIV-women (table 1)

Discussion
The prevalence of cervical lesions in HIV+ and HIVwomen in this study is similar to the 56% and 12.6% reported among HIV+ and HIV-women in North Central Nigeria by Lawal et al., respectively 23 but lower than the 61.1% and 76% reported in HIV+ women at Tanzania and Zambia, respectively 11, 24 . In this study, te prevalence of SIL is higher than the prevalence reported in Etiopian HIV+ women (13.6%) and HIV-women (5.2%), respectively 25  Interestingly, the prevalence of EBV infection was associated with older age in HIV+ women while the prevalence of the virus was associated with younger age in HIV-women. The divergence in age related prevalence of the virus may account for the difference in the prevalence of SIL between the two groups. This is underscored by the fact that the prevalence of SIL in HIV+ women was associated with older age as well.
The relationship between older age and prevalence of EBV has early been reported 20 . Unlike in HIV-women, increasing prevalence of EBV infection was associated with disease state in HIV+ women, with the highest prevalence in HSIL. Similar reports of have been made in earlier studies [20][21] . However, these studies were not explicit on the HIV status of their entire participants.
Though the prevalence of HPV infection was associated with younger age both in HIV+ and HIV-women, its prevalence was associated with higher disease state in HIV+ women. This again demonstrates the importance of EBV and HPV persistence in cervical carcinogenesis, especially in immunocompromised individuals. Studies have shown that the absence or reduced antiviral T-cell response in HIV infection favours superinfections, co-infections, persistence, replication and reactivation of latent oncoviruses 32,33 . This might explains why HIV+ women had higher EBV/HPV associated SIL than HIV-women. Studies have shown that the presence of EBV latent membrane protein 1 and HPV is associated with aggressive and poorly differentiated squamous cell carcinomas phenotype 20,34,35 . This may be associated with the higher prevalence of SIL observed in HIV+ women. The high prevalence of SIL in HIV+ women with tri-infection could be due to immune exhaustion, since these women had a lower expression of Ki67 in HSIL than their HIV-counterparts 18 . Although the prevalence of HSV-2 is lower in HIV-women than their HIV+ counterparts, such HIV-women are at risk of acquiring HIV infections, especially those with acute HSV-2 infections 36 .

Conclusion
This study revealed that the prevalence of viral mono-infections and co-infections, and SIL were higher in HIV+ women than in HIV-women. It also revealed that higher prevalence of SIL in viral tri-infection than in bi-infection and mono-infection both in HIV+ and HIV-women. It suggests that the high prevalence of SIL in HIV+ women could be associated with viral co-infections.