Nearly 25 million of the 34 million people estimated to be living with HIV infection worldwide at the end of 2008 reside in sub-Saharan Africa. Although most national epidemic estimates have stabilized or begun to decrease in the region, the current prevalence is dramatically high, with a range of 2%–25% among the adult population. Swaziland, South Africa, Lesotho, and Bostwana are the African countries with the highest rates of HIV infection. Heterosexual sexual activity is the most frequent mechanism of transmission, and women are more frequently infected than men, which implicates vertical transmission of HIV infection .
Hepatitis B virus (HBV) infection is the most common cause of chronic liver disease worldwide. There are nearly 400 million carriers of HBV, and an estimated 65 million of these carriers reside in Africa . Because access to antiretroviral therapy (ART) has begun to be a reality for a growing number of HIV-infected patients in sub-Saharan Africa, new challenges have emerged; the adequate management of chronic hepatitis B in coinfected patients is one of those challenges. Limited availability of diagnostic tools and antiviral agents to treat hepatitis B, as well as lack of expertise among health care providers, contribute to the mismanagement of hepatitis B in this population. The consequences may surpass progression of liver disease in HBV carriers. Thus, complicated clinical care is a pressing issue for providers of HIV treatment, and the emergence of drug-resistant HBV variants is an undesirable outcome with deleterious effects for patients.
In this issue of Clinical Infectious Diseases, Hoffmann et al.  report the impact of HBV infection on the effectiveness and tolerance of ART in HIV-infected individuals in South Africa. The authors retrospectively examined 537 HIV-infected patients (94% were men; median age, 41 years; median CD4 cell count, 150 cells/mm3; median plasma HIV RNA level, 4.9 log10 copies/mL), 106 (19.7%) of whom were serum hepatitis B surface antigen positive. All of the patients were ART naive and initiated therapy with zidovudine, lamivudine, and efavirenz. The proportion of patients who achieved an undetectable plasma HIV RNA level (<400 copies/mL) and the extent of CD4 cell recovery after up to 18 months of follow-up did not differ significantly between HIV-infected patients with and without chronic hepatitis B.
During the study period, a total of 23 individuals (4.3%) experienced grade 3 or 4 hepatotoxicity (defined as an elevated aspartate aminotransferase and/or alanine aminotransferase level >5 times the upper limit of normal during HAART). Episodes of hepatotoxicity occurred more frequently (incidence rate ratio, 4.4) in the subset of HIV-HBV–coinfected patients with high baseline serum HBV DNA levels (>1×104 copies/mL) than in the other patients. of note, the proportion of HIV-HBV–coinfected patients with a baseline serum HBV DNA level >1×104 copies/mL was 43.4%. Most hepatotoxic events occurred within the first 12–24 weeks after HAART initiation, which may suggest an immune reconstitution inflammatory syndrome (IRIS) . However, 3 of the 7 episodes in the subset of patients with high serum HBV DNA levels occurred after 24 weeks of HAART, which might have resulted from the selection of lamivudine-resistant HBV variants (although other reasons could have also accounted for the elevated transaminase levels).
At least 5 aspects of the study by Hoffmann et al.  merit special attention and may have important implications regarding more widespread use of ART among the HIV-infected population in sub-Saharan Africa (table 1). First, The prevalence of chronic hepatitis B in this group of HIV-infected patients in South Africa (∼20%) is higher than that reported in the European and North American HIV-infected populations and is comparable to that in cohorts of HIV-infected patients from Southeast Asia . Deficient HBV vaccine programs in the area and frequent sexual promiscuity among the patients included in this study likely accounted for the high prevalence of HBV infection. These results highlight that baseline screening of serum HBV markers is crucial in HIV-infected patients, especially in regions where universal HBV vaccination during childhood has not yet been well implemented. With this information, individuals not exposed to HBV will benefit from vaccination, and those found to be infected with HBV may receive adequate care.
Second, antiviral response and CD4 cell recovery during HAART did not seem to be impaired in HIV-infected patients with hepatitis B, compared with HIV-monoinfected individuals. This observation is important because some studies, mainly those involving patients with chronic hepatitis C, have claimed that treatment responses could be lower in coinfected patients because of the multiple interferences between HIV and hepatitis viruses. As recently shown with hepatitis C , the results from Hoffmann et al.  highlight that concomitant hepatitis B does not have a negative impact on the response to ART in HIV-infected patients. Moreover, the most recent guidelines have stressed the convenience of considering early introduction of ART in coinfected patients with use of regimens active against both HIV and HBV infection [7, 8]. This recommendation is based on the evidence of faster liver disease progression in HIV-HBV–coinfected patients, compared with HBV-monoinfected patients, because of the immunological impairment attributable to HIV infection .
Third, the overall incidence of HAART-associated hepatotoxicity was low in the study by Hoffmann et al. , except in the subset of HIV-HBV–coinfected patients with high serum HBV DNA levels. This observation challenges any concern about the use of ART in this population, although avoidance of the most hepatotoxic drugs must be considered . Chronic hepatitis B should not discourage the initiation of HAART in coinfected patients.
Fourth, a high serum HBV DNA level was associated with elevated liver enzyme levels at baseline and greater risk of subsequent elevations in liver enzyme levels during HAART use. This observation is interesting and reinforces that quantitative measurement of serum HBV DNA at baseline and periodically thereafter is warranted in HIV-HBV–coinfected patients. Information on viral load in hepatitis B has prognostic value for the natural history of liver disease [11–13] and influences treatment decisions, including considerations of the risk of selection of drug resistance [7, 14]. Therefore, HBV DNA quantification should be mandatory for all chronic hepatitis B surface antigen carriers, including those coinfected with HIV .
Fifth, the use of lamivudine as part of the HAART regimen did not prevent all episodes of elevations in liver enzyme levels in HIV-HBV–coinfected patients, especially in those with high HBV DNA levels. Both early-onset (within 12–24 weeks and onset due to IRIS) and late-onset episodes of elevations in liver enzyme levels (attributable to selection of lamivudine-resistant HBV) may occur in this population. Although a low baseline CD4 cell count may increase the risk of IRIS events, as was recently shown in another study conducted in South Africa in which 10% of HIV-infected patients who initiated HAART developed IRIS , selection of lamivudine-resistant HBV is, generally, the cause of elevations in liver enzyme levels that occur after 6 months of suboptimal anti-HBV therapy . Figure 1 shows, schematically, the different processes that may account for episodes of elevations in liver enzyme levels in HIV-HBV–coinfected patients after initiation of HAART. Timing may help to distinguish the most likely causes and ensure their adequate management, but measurement of HBV DNA levels is necessary for a correct interpretation.
Selection of lamivudine-resistant HBV must be avoided for at least 4 reasons: (1) the benefit of slowing the progression of liver disease disappears ; (2) because the selection of lamivudine-resistant HBV causes cross-resistance, the response to subsequent anti-HBV agents may be impaired ; (3) selection of HBV vaccine escape mutants may be favored [18–21]; and (4) transmission of drug-resistant HBV may increase . Additional anti-HBV drugs with greater potency and that are less vulnerable to the selection of HBV-resistant variants than is lamivudine (particularly tenofovir) should be made available for the treatment of all HIV-HBV–coinfected individuals. The use of lamivudine as the only anti-HBV agent in an ART regimen is discouraged for patients with high HBV DNA levels, because of its low potency and the high risk of selection of HBV mutations that confer wide cross-resistance to other anti-HBV drugs (e.g., emtricitabine, telbivudine, and entecavir). There is no doubt that, at present, a tenofovir-based regimen must be considered as the treatment of choice, in general, for HIV-HBV–coinfected patients [7, 8, 23, 24]. ART regimens based on tenofovir seem to be the most robust for ensuring sustained suppression of HBV replication in the majority of this patient population, which fortunately, is associated with amelioration or even some reversion of HBV-related liver fibrosis [25, 26].
An interesting and positive observation in the study by Hoffmann et al.  should not be missed. Only 46 (43.4%) of the 106 hepatitis B surface antigen–positive individuals had a baseline serum HBV DNA level >10,000 copies/mL. The risk of liver-related complications in hepatitis B is correlated with viral load; patients with low HBV DNA levels (<10,000 copies/mL) have minimal risk of progression to end-stage liver disease [11–13]. This subset of patients generally represents “inactive” HBV carriers; these patients are serum hepatitis B e antigen negative and have liver enzyme levels within normal limits. Moreover, these patients are less infectious than are those with high viral loads . These results are somewhat unexpected, because most HIV-HBV–coinfected individuals in Western countries have high levels of hepatitis B viremia [7, 14]. Different HBV genotypes and the prevalence of serum hepatitis B e antigen positivity in sub-Saharan Africa, compared with North America and Western Europe, may explain this observation . Although hepatitis B e antigen–negative chronic hepatitis B is the most common type of hepatitis B in Africa and typically presents with low-level viremia, hepatitis B e antigen–positive chronic hepatitis B that presents with high-level viremia is the most frequent type of hepatitis B in Western countries among HIV-infected persons . Achievement of complete HBV suppression is much easier for patients with low-level baseline hepatitis B viremia than for those with high baseline levels of replication, even when lamivudine is the only agent active against HBV that is used. In the absence of ongoing HBV replication, the risk of selection of resistance and treatment failure is very low. In this regard, a large group of HIV-HBV–coinfected patients in the study by Hoffmann et al.  could, in fact, benefit from a HAART regimen containing lamivudine as the single anti-HBV agent. However, serum HBV DNA level monitoring must be performed to determine who may benefit from this strategy and who would need a more effective anti-HBV agent. The latter group is mainly represented by patients with a high baseline serum HBV DNA level who do not achieve complete suppression of viremia within 6 months after the start of lamivudine therapy . As highlighted above, continuation of the same HAART regimen in that situation could be deleterious.
Because of the insufficient data in the study by Hoffmann et al.  with regard to serum HBV DNA level decreases and rebounds of viremia during HAART use, it is not possible to determine the mechanisms that explain elevations in transaminase levels in cases of hepatotoxicity. Although some events, especially those occurring early after the initiation of HAART, could be explained by IRIS, others could be the result of flares commonly seen during the course of HBV disease or of HBV DNA level rebounds after selection of resistance. of note, as the authors acknowledge, concomitant use of hepatotoxic drugs, such as antituberculous agents, was not evaluated. Other toxicities for the liver, such as alcohol toxicity, could have also been a factor explaining elevated transaminase levels. The role of liver toxicity other than that caused by HAART is even more elusive in the absence of baseline transaminase levels for some patients.
Financial support. Fundación Investigación y Educación en SIDA, Red de Investigación en SIDA (RD06/0006), Agencia Lain Entralgo, and the European Network of European AIDS Trials project.
Potential conflicts of interest. All authors: no conflicts.