Treatment of Chronic Hepatitis B in HIV-Infected Persons: Thinking Outside the Black Box

Worldwide, there are >300 million persons with chronic hepatitis B, 40 million with HIV-1 infection, and a substantial (but unknown) number with both. Available data indicate that ∼8% of HIV-infected persons in the United States and Europe have chronic hepatitis B (i.e., HIV—hepatitis B virus [HBV] coinfection) [1, 2]. The prevalence is higher among HIV-infected persons in some areas in sub-Saharan Africa and Asia, although data are sparse [3, 4]. In several ways, HIV-HBV coinfection adversely affects the chronic viral illnesses due to monoinfection with either virus. HIV-HBV—coinfected persons have more severe liver disease than do persons with either chronic infection alone. In the Multicenter AIDS Cohort Study, the incidence of liver-related mortality for persons with HIV-HBV coinfection was 17 times that for HIV-uninfected persons with chronic hepatitis B [5]. In addition, HBV infection can affect the tolerability of antiretroviral medications. Wit et al. [6] found that the incidence of grade 3–4 hepatotoxicity for HIV-HBV coinfected persons was 9 times the incidence for those without chronic hepatitis B, a result corroborated by findings of other studies [7–9]. There are a number of pathways through which antiretroviral medications can lead to hepatotoxicity in HIV-HBV—coinfected persons. In addition to the direct effects of the medications on the liver, which also occur in persons without chronic viral hepatitis [10], antiretroviral medications administered to HIV-HBV—coinfected persons may cause toxicity by restoring the immune responses to HBV [11]. Furthermore, with use of antiretroviral drugs that effect the replication of both HIV and HBV, liver injury can occur both when particular HBV variants that are less fully suppressed are selected (resulting in breakthrough infection) or when treatment with those medications is stopped (resulting in relapse of infection) [12]. Relapse is common, because most medications active against HBV do not eliminate intrahepatic covalently closed circular DNA, which is the replication template. Thus, discontinuation of medications that constrained replication without elimination of the covalently closed circular DNA reservoir is typically associated with infection relapse. Because of the increased risks for liver disease progression and for hepatotoxicity from antiretroviral agents, treatment of chronic hepatitis B is important in the HIV-infected adult. However, the optimal treatment regimen is not clear, because of the paucity of data and the overlapping drug activity against both viruses. Antiretroviral agents with activity against HBV include lamivudine, emtricitabine, and tenofovir disoproxil fumarate (DF), with only lamivudine being approved by the US Food and Drug Administration (FDA) for hepatitis B treatment in HIV-uninfected individuals (table 1). The other agents with anti-HBV activity are adefovir dipivoxil, entecavir, standard IFN-α, and pegylated IFN-α (peg—IFN-α), all of which have FDA approval for the treatment of chronic hepatitis B in HIV-uninfected individuals. Entecavir has an additional indication for persons infected with HIV who have previously received lamivudine therapy. Medications available for treatment of chronic hepatitis B (CHB). Recently, HBV-related concerns have been (or will be) added to the black box warnings for the antiretroviral drugs active against both HIV and HBV, which appropriately highlight the possibility of clinically significant hepatitis B relapse with their discontinuation. We applaud the FDA for bringing this important issue to the attention of clinicians. However, the black box warning for one of these agents, tenofovir DF, further states that this drug is “not indicated” for the treatment of chronic hepatitis B (figure 1). In FDA parlance, this means that the drug has not been registered for this use. However, it is important that clinicians do not misinterpret this warning, because the warning does not mean that tenofovir DF is not active against HBV, nor should it be interpreted as a contraindication for use in persons with chronic hepatitis B. Tenofovir DF has strong anti-HBV activity, and there is mounting clinical evidence that the drug plays an important role in treating chronic hepatitis B in HIV-infected persons. Accordingly, tenofovir DF use is supported in a US HIV-HBV treatment guideline reported by Benson et al. [16], as well as a European guideline reported by Alberti et al. [17] on the treatment of HBV-HIV coinfection. Black box warning for tenofovir, from [15] Currently, there are several options for treatment of chronic hepatitis B in HIV-infected persons. IFN-α2b, which is the first medication approved for the treatment of chronic hepatitis B in HIV-uninfected persons, has not been well-studied in HIV-HBV—coinfected persons, but its efficacy appears to be low [18]. We are aware of a cumulative total published experience of 98 HIV-HBV—coinfected persons (all prior to effective antiretroviral therapy) treated with IFN-α with an overall response rate of 14.3% [19–25]. Peg—IFN-α appears to be more effective than standard IFN-α for treatment of chronic hepatitis B in HIV-uninfected persons [26], but there are no published data on the effectiveness of peg—IFN-α treatment in persons coinfected with HBV and HIV. Lamivudine is a nucleoside analogue that, in its active triphosphate form, inhibits HBV DNA polymerase and HIV reverse transcriptase. Although HBV DNA levels decrease a mean of 3 log copies/mL in HIV-HBV—coinfected persons after initiation of lamivudine treatment [27], the annual incidence of lamivudine resistance among coinfecting HBV strains is ∼25% in HIV-infected persons [28]. When lamivudine-resistant variants are selected (or emerge), the HBV DNA level increases; concentrations of liver enzymes, such as alanine aminotransferase, may increase; and in some but not most persons, the resulting hepatitis can be fatal. In addition, the bulk of available data suggest that the benefit of preventing progression of liver disease is substantially diminished in the presence of lamivudine-resistant HBV [29]. Similarly, emtricitabine is a nucleoside analogue that, after intracellular phosphorylation, exerts potent inhibition of both HIV and HBV replication. Although emtricitabine has not been approved by the FDA for the treatment of chronic hepatitis B, HBV DNA levels decreased by ∼3 log copies/mL in HIV-infected and HIV-uninfected patients treated for 48 weeks [30]. However, mutations conferring HBV resistance to both emtricitabine and lamivudine were observed in HBV isolates from ∼13% of treated patients, potentially limiting the utility of this agent in HIV-HBV—coinfected patients. Recently, emtricitabine in combination with tenofovir DF was approved for the treatment of HIV infection, pairing 2 agents with potent anti-HBV activity. Entecavir is a nucleoside analogue that is licensed for the treatment of chronic hepatitis B in persons with and persons without HIV infection. It inhibits all 3 functions of HBV polymerase, including base priming, reverse transcription of the negative strand, and synthesis of the positive strand of HBV DNA, but it does not have activity against HIV reverse transcriptase. Because lamivudine-resistant mutants are less susceptible to entecavir, a 1.0-mg dose is required, compared with a 0.5-mg dose of entecavir in lamivudine-naive patients. In a randomized, controlled trial involving 68 HIV-HBV—coinfected persons receiving lamivudine, a 24-week course of entecavir resulted in a decrease in the HBV DNA load of 3.65 log copies/mL, which is similar to findings for HIV-uninfected persons [31,32]. In clinical trials, entecavir resistance mutations were detected after 48 weeks in 7% of HIV-uninfected persons with lamivudine-resistant HBV infection, but to date, no entecavir resistance has been identified in lamivudine-naive patients. The long-term rate of developing entecavir-resistant HBV in HIV-HBV—coinfected persons, many of whom have been treated with lamivudine, is not known. Adefovir dipivoxil is a nucleotide analogue that, in its active diphosphate form, inhibits DNA polymerase and reduces HBV DNA levels an average of 3.5 logs at 48 weeks [33]. Adefovir dipivoxil is licensed for treatment of chronic hepatitis B in persons without HIV infection and is active against lamivudine-resistant HBV. In addition, a total of 35 HBV-HIV coinfected persons received treatment with adefovir for 192 weeks that led to substantial reduction in the HBV DNA load [34]. It is clear from this study and the experience with HIV-uninfected persons that the incidence of clinically evident HBV resistance to adefovir is substantially lower than that for lamivudine [35]. However, use of adefovir in HIV-HBV—coinfected persons poses a theoretical risk of HIV developing cross-resistance to tenofovir, because adefovir is active against HIV at higher doses. Such cross-resistance has not been demonstrated in the small study by Benhamou et al. [34], but more data are needed to insure that this does not occur. Tenofovir DF is a nucleotide analogue approved by the FDA for treatment of HIV infection and differs from adefovir dipivoxil by 1 methyl group. In vitro, tenofovir DF has activity that is at least equivalent to that of adefovir dipivoxil, with a similar IC50 [36]. The efficacy of tenofovir DF appeared to be superior to that of adefovir dipivoxil against a triple lamivudine-resistant mutant HBV strain [37], which is found in ∼30% of HIV-HBV—coinfected individuals [38]. Clinical studies and 1 AIDS Clinical Trial Group (ACTG)—sponsored randomized, controlled trial (ACTG A5127) have clearly demonstrated that tenofovir DF is not inferior to other drugs, including adefovir dipivoxil, approved for the treatment of chronic hepatitis B in HIV-HBV—coinfected persons [39, 40]. Furthermore, in ACTG A5127, in which 27 and 25 persons received tenofovir DF and adefovir dipivoxil, respectively, there was a trend toward the superiority of tenofovir versus adefovir dipivoxil, owing to a larger mean decrease in the HBV DNA load (5.74 vs. 4.03 log copies/mL) [39]. In addition to the report by Peters et al. [39], to date, 115 HIV-HBV—coinfected individuals treated with tenofovir have been described, with a minimum decrease in the HBV DNA load of 4 log copies/mL during 24–48 weeks of treatment (table 2). One study prospectively followed up 53 individuals infected with lamivudine-resistant HBV. Tenofovir DF was given to 35 of these individuals, and 18 received adefovir dipivoxil [40]. The tenofovir DF—treated group had a more rapid decrease in the HBV DNA load and greater decrease in the hepatitis B e antigen concentration, suggesting a stronger anti-HBV effect of tenofovir. In addition, multiple retrospective studies have been presented at meetings, with the largest following up 107 HIV-HBV—coinfected persons who received tenofovir for a median duration of 10 months (range, 2–24 months) [51]. Ninety of these individuals had a detectable HBV DNA level at the start of tenofovir therapy, of whom one-third reached an HBV DNA level of <200 copies/mL. To date, no significant side effects have been reported in any of the studies involving tenofovir DF, and the number is expected to be low, as observed for adefovir dipivoxil. Data from studies that examined the efficacy of tenofovir DF (TDF) for treatment of chronic hepatitis B. There are no randomized, controlled trials comparing the possible options, and therefore, strong treatment recommendations cannot be made. Given the widespread use of lamivudine for treatment of HIV infection and the 25% annual incidence of resistance among HBV isolates, the majority of HIV-HBV—coinfected persons taking antiretroviral medications already carry lamivudine-resistant HBV. Because of this situation, coupled with the fact that resistance breakthroughs can be clinically significant, many experts use alternative anti-HBV medications in HIV-infected persons. Tenofovir DF and entecavir are both active against lamivudine-resistant HBV, but they have not been compared in HIV-HBV—coinfected persons. In our opinion, tenofovir DF should be strongly considered as treatment for chronic hepatitis B when an adult patient also needs HIV treatment, according to published guidelines for the latter. Clearly, the strength of our recommendation for the use of tenofovir DF would be much greater if the drug had been licensed for the treatment of chronic hepatitis B, but the use of tenofovir DF is easily justified in light of available data. In vitro, tenofovir DF is at least as potent as adefovir and any other available drug. Entecavir is the only other drug approved for treatment of chronic hepatitis B in HIV-infected persons, but it is not active against HIV. Thus, unlike tenofovir DF, the use of entecavir is an additional pill in persons requiring treatment of HIV infection. Only lamivudine has been involved in more published studies than tenofovir DF, but these data point to a high rate of virologic resistance and clinical failure. Adefovir dipivoxil appeared to durably suppress HBV replication in one study of 35 HIV-HBV—coinfected persons [34]. However, adefovir dipivoxil at a dosage of 10 mg/day raises concerns regarding development of HIV cross-resistance to tenofovir DF, but it does not contribute to an antiretroviral regimen for which tenofovir DF is a leading candidate. Thus, although tenofovir DF has not been licensed for use for treatment of chronic hepatitis B, we believe that the drug is clinically indicated for treatment of chronic hepatitis B for HIV-infected persons in whom tenofovir DF could also be a component of the antiretroviral regimen. Furthermore, with the commercial availability of tenofovir DF—emtricitabine, a single pill that is part of an HIV regimen can be used for hepatitis B treatment, providing 2 agents with dual activity against HIV and HBV. It is regrettable that better data are not available to support this recommendation, and we strongly support the conduct of clinical trials to guide treatment guidelines. In fact, because of the high prevalence of hepatitis B (and hepatitis C) among HIV-infected persons, we believe that the FDA should consider requiring that such studies be done in these special populations. In the meantime, the best approach to the treatment of chronic hepatitis B in HIV-infected persons will require thinking “outside the black box,” and in particular, coupling the available data on the treatment of chronic hepatitis B in persons without HIV infection with sound clinical judgment. Financial support.National Institutes of Health (DA00441, DA13086, DA16065, and AI060454), which had no role in the preparation, review, or approval of the article in manuscript form. Potential conflict of interest.D.L.T. is a member of a scientific advisory board for Bristol-Myers Squibb, which is bringing a drug for treatment of hepatitis B virus infection to market; was a member of a scientific advisory board for Roche Pharmaceuticals, which is studying pegylated IFN-α for treatment of chronic hepatitis B, in 2003; and has received payment during the past 3 years from Gilead, the manufacturer of tenofovir disoproxil fumarate, for giving a talk about chronic hepatitis B. C.L.T. and M.S.S.: no conflicts.