Hepatitis B virus infection among men who have sex with men and transgender women living with or at risk for HIV: a cross sectional study in Abuja and Lagos, Nigeria

Background Despite the development of a safe and efficacious hepatitis B vaccine in 1982, the hepatitis B virus (HBV) remains a public health burden in sub-Saharan Africa. Due to shared risk factors for virus acquisition, men who have sex with men (MSM) and transgender women (TGW) living with HIV are at increased risk of HBV. We estimated the prevalence of HBV and associated factors for MSM and TGW living with or without HIV in Nigeria. Methods Since March 2013, TRUST/RV368 has recruited MSM and TGW in Abuja and Lagos, Nigeria using respondent driven sampling. Participants with HIV diagnosis, enrollment as of June 2015, and available plasma were selected for a cross-sectional study and retrospectively tested for hepatitis B surface antigen and HBV DNA. Logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for factors associated with prevalent HBV infection. Results A total of 717 MSM and TGW had a median age of 25 years (interquartile range [IQR]: 21–27), 5% self-reported HBV vaccination, 61% were living with HIV, 10% had prevalent HBV infection and 6% were HIV-HBV co-infected. HIV mono-infected as compared to HIV-HBV co-infected had a higher median CD4 T cell count [425 (IQR: 284–541) vs. 345 (IQR: 164–363) cells/mm3, p = 0.03] and a lower median HIV RNA viral load [4.2 (IQR: 2.3–4.9) vs. 4.7 (IQR: 3.9–5.4) log10copies/mL, p < 0.01]. The only factor independently associated with HBV was self-report of condomless sex at last anal intercourse (OR: 2.2, 95% CI: 1.3, 3.6). HIV infection was not independently associated with HBV (OR: 1.0, 95% CI: 0.7–1.6). Conclusion HBV prevalence was moderately high but did not differ by HIV in this cohort of MSM and TGW. Recent condomless sex was associated with elevated HBV risk, reinforcing the need to increase communication and education on condom use among key populations in Nigeria. Evaluating use of concurrent HIV antiretroviral therapy with anti-HBV activity may confirm the attenuated HBV prevalence for those living with HIV.

Conclusion: HBV prevalence was moderately high but did not differ by HIV in this cohort of MSM and TGW. Recent condomless sex was associated with elevated HBV risk, reinforcing the need to increase communication and education on condom use among key populations in Nigeria. Evaluating use of concurrent HIV antiretroviral therapy with anti-HBV activity may confirm the attenuated HBV prevalence for those living with HIV.
Keywords: HBV DNA, Sexual and gender minorities, Antiretroviral therapy, Sub-Saharan Africa

Background
Despite the development of a safe and efficacious hepatitis B vaccine in 1982 [1], the hepatitis B virus (HBV) remains a public health burden in sub-Saharan Africa with a reported prevalence of about 6% [2,3]. While mortality in sub-Saharan Africa from the World Health Organization (WHO)-targeted infectious diseases (i.e. HIV, tuberculosis and malaria) is now declining, morbidity and mortality due to viral hepatitis is increasing [2]. Globally, HBV is the leading cause of end-stage liver disease (ESLD) and hepatocellular carcinoma (HCC) [4]. In 2015, HBV infections were responsible for 887,000 deaths, in addition to 257 million individuals living with chronic HBV infections worldwide [5].
In resource-limited settings, enzyme immunoassays detecting HBV surface antigen (HBsAg) are commonly used as a standard of care diagnostic test for active HBV infection. However, concerns have been raised about the use of this test for those co-infected with HBV and HIV, in whom the targeted HBV immune markers might be suppressed [6,7]. A more sensitive diagnostic assay that quantifies HBV-DNA rather than circulating antibodies is the polymerase chain reaction (PCR)-based molecular diagnostic assay that has served as a confirmatory and treatment progress-monitoring test for active HBV [8,9].
Due to shared risk factors for acquisition, HBV is common in persons living with HIV (PLWHs) [10]. Furthermore, men who have sex with men (MSM) are 2 to 4 times [2] and 3 to 6 times [11] more likely to be infected with HBV and HIV, respectively, as compared to reproductive aged adults [12]. HBV-HIV co-infection is associated with a 14 fold increased risk in progressing to ESLD compared to mono-infection with either virus [13] and is also associated with increased risk of hospitalization for non-liver-related causes [14,15]. As the advent of ubiquitous combination antiretroviral therapy has improved life expectancies of MSM and transgender women (TGW) living with HIV, these individuals are now at higher risk of developing chronic complications from comorbid infections such as HBV [16,17].
In Nigeria, prevalence studies of HBV range between 7 to 14% among reproductive aged adults [18,19], and in Calabar, Southeast Nigeria, 69% of patients with ESLD tested positive for HBsAg [20]. HBV prevalence among PLWH range from 4 to 8% [21][22][23] and as high as 10 and 18% for MSM in Ibadan and Lagos, Nigeria, respectively [24]. The objective of this study was to investigate the prevalence and factors associated with PCRdiagnosed HBV among MSM and TGW living with or at risk for HIV in Nigeria. We hypothesized that HIV infections would be associated with increased odds of HBV infection in Nigerian MSM and TGW.

Study design and population
The TRUST/RV368 cohort study used respondentdriven sampling (RDS) to recruit MSM and TGW between March 2013 and February 2020 in Abuja and between April 2014 and May 2018 in Lagos as previously described [25]. Eligibility criteria included the following: (1) assigned male sex at birth, (2) history of insertive or receptive anal intercourse with another man in the past 12 months, (3) adulthood (aged 16 years or older in Abuja and 18 years or older in Lagos), (4) possession of a valid RDS coupon. Age inclusion criteria differed between sites because of differences in the recommendations from the instutional review boards. Exclusion critiera included those who declined HIV counseling and testing. In addition to these criteria, participants included in this cross-sectional study had an enrollment visit at the TRUST/RV368 clinics before June 2015, an HIV diagnostic result, and plasma for retrospective HBV screening. During enrollment evaluations, spaced approximately 2 weeks apart, consented participants underwent HIV counseling and testing and provided demographic, behavioral and clinical data through inperson interviews with trained staff using a standardized questionnaire. In addition, participants received physical examinations and sexually transmitted infection (STI) diagnostics [26].

Laboratory procedures
HBsAg screening was completed using the Genescreen HBsAg 3.0 EIA assay (Bio-Rad Laboratories, Redmond, WA). To confirm the presence of HBV and to quantify DNA, we performed in vitro PCR using the Abbott Real-Time HBV assay (Abbott Molecular, Des Plaines, IL). Participants who tested positive for HBsAg with detectable HBV DNA were considered HBV-infected.
Whole blood was tested for HIV using a parallel series of rapid tests (Alere Determine, Waltham, MA and Trinity biotech Uni-Gold HIV, Wicklow, Ireland). In cases of discordance, a 3rd rapid test (Chembio Diagnostics HIV-1/2 Stat Pak test, Medford, NY) was used to confirm or exclude HIV infection as outlined by the parallel testing algorithm for at-risk individuals in Nigeria [27]. For participants living with HIV, HIV RNA was quantified using the Ampliprep/COBAS TaqMan HIV-1 Test (Roche Molecular Diagnostics, Pleasanton, CA) and CD4 counts using the Partec CyFlow Counter (Sysmex, Kobe, Japan). Voided urine and rectal swabs were tested for Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) using the Aptima Combo 2 CT/ NG Assay (Hologic, San Diego, CA). All participants testing positive for HIV or other STIs were offered appropriate treatment regardless of CD4 count.

Statistical analyses
The primary outcome of this investigation was prevalent HBV infection, and the main explanatory variable was HIV infection status, categorized as infected or uninfected. Potential cofactors evaluated included age (16-20, 20-25, 25-60 years), site (Abuja, Lagos), education (primary, secondary, tertiary), marital status (ever, never, other), willingness to access health care services (yes,no), HBV vaccination status (yes, no), sexual orientation (homosexual or bisexual), sexual positioning (insertive only or any receptive), ever had anal sex without a condom (yes, no), condom use during last anal sex (yes, no), concurrent sexual partners (yes, no), prevalent STIs (any urethral or rectal NG or CT), and vaginal sex in the last year (yes, no).
Distributions of demographic characteristics were compared by HBV status using Pearson's Chi-square and Fisher's exact tests for categorical variables. Wilcoxon rank sum test was used to compare continuous variables. Bivariate and multivariable logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association between HIV and other factors with HBV. Confounders were identified based on biological plausibility, statistical associations with exposure and outcome (p < 0.10), and at least a 10% change in the crude beta estimates of the main independent variable. Potential confounders were entered into the multivariable model via forward stepwise selection and retained based on signficance after likelihood ratio testing [28]. Missing observations accounted for less than 4% of the total sample and were retained in the models with indicators. A p-value less than 0.05 was considered statistically significant and a pvalue less than 0.10 was suggestive of a trend. Characteristics of individuals who were included in the sample versus excluded because of incomplete data were compared using Pearson's Chi-square test. Data analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC USA).

Results
Of the 771 participants enrolled from March 2013 to June 2015, 717 had real-time HIV and retrospective HBV testing results available and were included in our study. Out of these, 438 (61%) were HIV-infected and 69 (10%) were HBV-infected, including 41 (6%) with HIV-HBV co-infection. HIV-HBV co-infected participants had a lower median CD4 T cell counts of 345 (interquartile range [IQR]: 163-363) cells/mm 3 than HIV mono-infected participants, 425 (IQR: 284-541) cells/ mm 3 (p = 0.03). Similarly, HIV-HBV co-infected participants had a higher median HIV RNA viral load of 4.7 (IQR: 3.9-5.4) log 10 copies/mL than HIV mono-infected participants, 4.2 (IQR: 2.3-4.9) log 10 copies/mL (p < 0.01). Among participants living with HIV, 51% (223/ 438) had been told by a physician or healthcare provider before enrolling in TRUST/RV368 that they needed to begin treatment for HIV. Of these participants, 24% (107/438) self-reported receiving medication to treat their HIV from a healthcare facility or pharmacy, but not all were able to recall the specific medication regimen they received prior to enrollment.
The median age of participants included in these analyses was 25 years with an IQR of 22 to 27 years. Sixtytwo percent of participants identified as bisexual and 75% engaged in any receptive sex (Table 1). Eighty-seven percent had never married, and 93% had at least a secondary school (high school equivalent) education. Regarding health services, 63% were willing to access care but only 5% self-reported HBV vaccination. Compared to HIV-uninfected persons, HIV-infected persons were more likely to be older, married, and engaged in any receptive sex. However, HIV-infected MSM and TGW were less willing to access health care services. The prevalence of HBV in the study population did not differ between those living with or without HIV (9% [95% CI 6.6, 12.1] vs. 10% [95% CI 6.5, 13.6], respectively, p = 0.8). After adjusting for potential confounders, HIV was not associated with HBV prevalence (adjusted OR: 1.0, 95% CI: 0.7, 1.6) ( Table 2).
Those who had not used a condom during last anal sex were twice as likely to be infected with HBV (OR: 2.2, 95% CI: 1.3, 3.6) as compared to those who had used a condom in the unadjusted analysis (Table 2). In the multivariable analysis, condomless sex at last anal sex remained an independent factor and was associated with a two-fold increased odds of HBV infection as compared to those who used condoms (adjusted OR: 2.2, 95% CI: 1.3, 3.6). Although not significant, self-reported homosexual orientation independently trended towards an association with HBV infection (adjusted OR: 1.5, 95% CI: 0.9-2.6). Age, marital status, willingness to access healthcare services, HBV vaccination, sexual position, ever condomless sex, concurrent sexual partners, prevalent bacterial STIs, and vaginal sex in the last year were not associated with HBV. After comparing those included and excluded from the analyses, the retained individuals were less willing to access health care services and more likely to self-identify as homosexual as compared to bisexual (all p < 0.05).

Discussion
The HBV prevalence of 10% observed among MSM and TGW in our study was comparable to those reported by cross-sectional studies of MSM in Lagos, Nigeria (10%) [24] and Brazil (11%) [29] but higher than Tanzania (5%) [30]. Our estimates were slightly higher than previous estimates of PLWH (4 to 8%) in Nigeria with unknown sexual orientation [21][22][23]. However, our results were lower than those from a study conducted in Ibadan (Southwest Nigeria) where the reported HBV prevalence among MSM was 18% [24]. The higher estimates from the Ibadan study [24] may be attributed to an older age, smaller sample size, and overestimation from enzymelinked immunosorbent assay based methods commonly used in resource-limited settings [6]. In contrast to our hypothesis, the prevalence of HBV was not higher among MSM and TGW living with HIV. This finding is consistent with a prior systematic review that reported an 8% pooled prevalence of HBsAg that did not differ by HIV among a population of Malawian adults (OR:1.2, 95% CI: 0.8, 1.6) [31]. A small proportion of our participants self-reported taking doctor or pharmacy prescribed medicine for their HIV infection before study entry. If our participants were prescribed antiretroviral therapy (ART) with anti-HBV activity, this may have reduced HBV infectivity [32,33]. Multiple studies have shown that tenofovir disoproxil fumarate (TDF) containing ART regimen, a first-line regimen in Nigeria [34], have significant anti-HBV activity [35][36][37]. A systematic review and network meta-analysis of randomized controlled trials [36] confirmed that TDF and tenofovir alafenamide fumarate (TAF) are the most effective ART regimens for HBV virologic suppression. In the Multicenter AIDS Cohort Study, MSM receiving effective ART (HIV RNA level < 400 copies/mL) had the same risk for incident HBV infection as MSM not living with HIV [32]. This pre-exposure prophylaxis effect of TDF containing regimens for HBV co-infections was further confirmed in the Swiss HIV Cohort where there was a reduction of HBV incidence among those on effective ART for their HIV infections [33]. TDF containing regimens may have virally suppressed HBV and prevented transmission for those living with HIV in our cohort.
Although our study did not find a significant association between HIV and HBV prevalence, it offers some insight into how the use of condoms and sexual activities may affect HBV infection. In our TRUST/RV368 cohort, condom usage has been suboptimal and this is a population known to have a high burden of human immunodeficiency virus [38], anorectal STIs [26], and sexual risk practices [39]. Consistent with our findings, the Swiss HIV Cohort study reported condomless sex increased the hazard of HBV acquisition 2-fold among those living with HIV independent of TDF-containing regimens [33]. In Nigeria, Adebajo et al. found an increased odds of HBV infection in Nigeria among MSM engaging in condomless sex that was predominantly for the insertive as opposed to receptive sexual position [24]. In our study, sexual positioning was not associated with HBV infection. Also in the study by Adebajo et al., participants who identified as bisexual were significantly more likely to have HBV (OR:5.2, 95 CI: 3.04, 9.0) [24], whereas our findings suggested the opposite, demonstrating a 2-fold increased prevalence of HBV for those who self-identified as homosexual. Socio-cultural differences in reporting of sexual practices between the North-central (Abuja) and Southwestern regions (Ibadan and Lagos) of Nigeria may affect how participants were categorized [40]. Despite sexual position or sexual orientation, culturally acceptable methods to communicate the benefits of condoms during anal sex for unvaccinated individuals could reduce the risk of HBV and associated morbidities such as ESLD and HCC.
Self-reported vaccination coverage among MSM and TGW was 5% and insufficient in preventing active HBV infection [5]. Routine childhood HBV vaccination in Nigeria was only enacted 15 years ago as part of the expanded immunization program and possibly explains the low vaccination coverage [41]. Most high-income countries with effective vaccination programs have much lower estimates of active HBV. A nationwide, multicenter prospective German study of HIV-infected MSM reported active HBV prevalence as 2%, cleared HBV as 27% and occult HBV as 8% [42]. Similarly, a Danish study from a large tertiary hospital evaluating HBV infection and vaccine coverage among MSM reported active HBV infection as 1%, cleared infection as 7%, and HBV vaccinated as 14% [43]. These studies highlight the reduction in active HBV infection that is possible if a catch-up vaccination strategy were implemented for atrisk individuals in low and middle-income countries. Our study had some limitations. With relatively small prevalence estimates, the multivariable model was not sufficiently powered to detect independent associations between HBV infection and known demographic and behavioral factors. However, most associations between sexual behaviors such as homosexual orientation and exclusively male partners with HBV infection aligned with the direction of the study hypothesis. Although, because more MSM who self-identified as homosexual were retained in our analyses, our prevalence estimates may have been an overestimate because of enrichment of those engaging in high risk sexual practices [38]. HBV vaccination was self-reported as ever receiving and did not specify the number of doses. Additionally, it was unclear who had been exposed to HBV from vaccination versus a prior infection, as testing for HBV surface antibody was unavailable due to limited sample volume. Due to uncertainty regarding participants' past exposure to specific ART regimens, we were unable to assess whether HBV prevalence was lower among those living with HIV and exposed to TDF-containing regimens. The temporality of associations between HIV and HBV could not be determined because of the cross-sectional study design. Lastly, MSM and TGW originated from two cities in Nigeria and results from this study may not be generalizable to other regions of Nigeria.