Prevalence of hepatitis B and delta according to HIV-type: a multi-country cross-sectional survey in West Africa

In West Africa, approximately 8% of HIV-infected individuals have a chronic hepatitis B virus (HBV) infection [1]. In this part of the world, HBV infection is one of the main causes of end-stage liver disease, cirrhosis and hepatocellular carcinoma (HCC) [2].

West Africa is also characterized by the circulation of both HIV type 1 and 2, which can to lead to co-infections with HIV-1 and HIV-2 (HIV-1&2) [3‐5]. Compared to HIV-1 infection, HIV-2 infection is characterized by a longer asymptomatic phase and a slower disease progression [6]. The management of HIV-2 infection is more challenging than the one of HIV-1, due to the intrinsic resistance of HIV-2 to non-nucleoside reverse transcriptase inhibitors (NNRTIs) and reduced sensitivity to several protease inhibitors (PIs) [7, 8].

Due to shared routes of transmission, HIV and HBV co-infection is common and 5–20% of HIV-infected people have also chronic HBV infection worldwide [9]. Compared to HBV or HIV-mono-infected individuals, HIV/HBV-co-infected patients have a higher risk of impaired immunological recovery and hepatotoxicity during antiretroviral treatment (ART) [10‐12] and a faster rate of progression to cirrhosis and HCC [9]. Thus, knowledge of HBV status among HIV-infected patients is important for clinical monitoring and selection of ART, as tenofovir disoproxil fumarate (TDF) and lamivudine (3TC) or emtricitabine (FTC) should be part of the treatment [13].

In West Africa, little is known in about the epidemiology of hepatitis D virus (HDV) infection, a defective RNA virus that requires the presence of HBV to infect the hepatocytes [14]. Its impact on HIV/HBV co-infected patients is not documented since HDV is not routinely diagnosed. It has been estimated that approximately 5% of HBV carriers are co-infected with HDV worldwide [15, 16] and between 3% and 25% in West Africa [17‐20]. HDV co-infection increases the risk for hepatitis flares and chronic hepatic complications [15, 21] and patients with HBV/HDV co-infection have a significantly increased risk for HCC compared with patients with HBV mono-infection and the general population [22]. Therefore, HDV screening is very important for the monitoring of HIV/HBV co-infected patients.

To our knowledge, only one small-scale study from Guinea Bissau has estimated the prevalence of HDV/HBV co-infection according to HIV types and did not find any difference between HIV-1 and HIV-2 [19]. However, this needs to be confirmed with a large and multicenter study. The objectives of this study were to estimate the prevalence of HBV and HBV/HDV co-infection according to HIV types among a large series of HIV-infected patients in the WADA (West Africa Database on Antiretroviral Therapy) cohort in three West African countries and, to identify risk factors for HBV seropositivity.

In this cross-sectional survey, the prevalence of HBV infection and HBV/HDV co-infection were relatively high, with no statistical difference by country and HIV types. To our knowledge, this is the largest report up to now on HBV/HDV co-infection among HIV-infected individuals in West Africa and the first in the three participating countries (Burkina Faso, Côte d’Ivoire and Mali). There are indeed few published studies in this part of the world [17‐19], and none of them, except in Guinea Bissau, has estimated the prevalence by HIV type [19].

Our study has some limitations. First, about 19% of the patients were considered as HIV-1 & HIV-2 dually reactive after the retesting. We were not able to differenciate between HIV-1 or type 2 since we did not perform PCR DNA for HIV-1 or HIV-2. Second, we did not document occult HBV infection since HBV DNA was only tested in HBsAg-positive patients. This has probably underestimated the prevalence of HBV infection. Indeed, in one study conducted in Côte d’Ivoire, 10% of HBsAg-negative patients had detectable HBV DNA [29]. Third, the lack of molecular data concerning the HDV as HDV RNA testing has become a more compelling tool for recognizing active replication. This may have overestimated the prevalence of HDV co-infection. Fourth, due to low amounts of plasma available, HBV DNA was not performed in all HBsAg-positive patients, and resistance and genotypic tests were not performed. In West Africa region, Genotype E is the predominant genotype [30]. Finally, our study population was selected based on the local identification of HIV-2 and dually reactive patients. The HIV-1 subgroup may thus not be fully representative of all HIV-1 patients in care in the same clinics. Despite these limitations, this study has also two major strengths. First it took place in three different West African countries and second it included a large sample of HIV-infected patients, giving therefore for the first time a global view of the distribution of HBV and HBV/HDV co-infection by HIV type in West Africa.

In our study, the prevalence of HBV was 8.5% (95% CI 6.6–10.6) among HIV-infected patients. This result is consistent with previous studies conducted in West Africa that have estimated the prevalence of HBV between 7.9% and 16.8% [17, 19, 31‐33]. The variation of this prevalence according to studies is possibly due to the HBV vaccination coverage and the specific distribution of some risk factors for HBV infection. Indeed we found that male patients and those <50 years old were more likely to be infected with HBV, which is consistent with previous reports [20, 34‐36]. The high prevalence of HBV in these two groups could be explained by a greater exposure to HBV through sexual behavior and body mutilations [20, 34, 35], but also by the early death of older patients infected with HBV during childhood. Indeed, in sub-Saharan Africa, 25% of adults who have been infected during childhood die from cirrhosis or liver cancer [37]. Other risk factors for HBV infection, such as WHO stage 3 or 4 and CD4 cell count <200 cells/mm³, were also found in other studies [36].

Only one-third of the patients co-infected with HIV and HBV received a standard-of-care ART regimen including TDF and 3TC (or FTC), two drugs active against both HIV and HBV. The remaining two-thirds were only exposed to lamivudine for about four years, a situation which is known to lead to the development of lamivudine-resistant HBV, from 43% in year 1 to >80% in year 4 [38‐40]. Those patients with long-standing lamivudine-resistant mutations may experience worsening liver disease. Since 2013, WHO recommends TDF and 3TC (or FTC) as the preferred NRTI option in first-line ART regimen in adolescents and adults infected with HIV [13, 41], and this recommendation is gradually endorsed by countries. This should allow an optimal management of HIV and HBV co-infection, including occult hepatitis. Indeed, according to a recent meta-analysis, up to 90% of patients receiving such a treatment had a suppressed HBV DNA after three years [42] and resistance to TDF has not yet been described in vivo [43, 44]. However, even with WHO guidelines for ART use in HIV-infected individuals, including two molecules active on both HIV and HBV, the benefits of HBV screening is still important, including referral for HBV vaccination in susceptible individuals and a better management of those already infected with HBV disease.

Regarding HDV infection, previous studies in sub-Saharan Africa have reported a prevalence of HBV/HDV co-infection ranging between 0.0% and 44.4% [17, 18, 20, 45, 46]. More recently, one study conducted in Guinea Bissau found a high prevalence of HDV/HBV co-infection (25.0%), with no statistical difference according to HIV type [19]. In our study, the HBV/HDV prevalence was 14.9%, not different by HIV type. This large variation of HDV prevalence could be explained by the different settings and populations, but also by the tests used for HDV diagnosis. Indeed, in one cohort of over 200 HIV/HBV co-infected individuals in rural Tanzania, no confirmed case of active HDV infection (second serology and nucleic acid amplification) was found among the 11 patients who had a positive anti-HDV antibody screening test [46]. Thus, all anti-HDV antibody positive samples should be confirmed with an additional antibody test or an HDV RNA test if possible. The management of this triple infection is complicated. Indeed, the effect of TDF alone or in association with 3TC (or FTC) on HDV replication is controversial according to studies [47, 48]. More research and tools are needed to improve our knowledge on the distribution and management of HBV/HDV co-infection among HIV-infected patients.

Finally, this large and multicenter study showed that like in Guinea Bissau, there was no statistical difference of the prevalence of HBV or HBV/HDV according to HIV types [19]. Thus, the specificity of HIV-2 infection seems to have no impact on the transmission and the rate of HBV and HBV/HDV infection.

HBV and HBV/HDV co-infection seem to be common in West Africa, irrespective of HIV type and the country. Therefore, the screening of both viruses should be systematically performed in HIV-infected patients to allow a better management of HIV. Follow-up studies are needed to improve our knowledge on HDV infection and to determine the impact of these two viruses on the course of HIV infection, especially now with the universal test and treat policy.