The management of chronic hepatitis B poses specific problems in the presence of human immunodeficiency virus (HIV) coinfection, because therapeutic approaches have to address both hepatitis B virus (HBV) and HIV infections. Response to interferon (IFN-α) is lower in HBV-HIV–coinfected than in HIV-negative subjects, especially in patients in advanced stages of immunosuppression. Thus far, there are no data on the performance of the new pegylated forms of IFN-α in HBV- and HIV-coinfected persons. After prolonged use of lamivudine, resistance develops in the majority of HBV-HIV–coinfected patients treated with the drug. The more recently approved tenofovir has shown excellent short-term results, and data from longer follow-up studies are eagerly awaited. Several drugs with combined anti-HIV and anti-HBV activity have recently been approved (emtricitabine) or are currently under development. Preliminary results with some of them are quite promising and probably will widen the therapeutic armamentarium against hepatitis B in patients with HIV infection.
Hepatitis B virus (HBV) and HIV share similar routes of transmission. As a consequence, up to 80% of HIV-infected patients have serological markers of present or past HBV infection . The prevalence of HIV-infected persons carrying hepatitis B surface antigen (HBsAg) varies according to geography and risk category, being higher among men who have sex with men and in developing countries. Global figures of 8%–11% have been given, whereas HBsAg positivity was found in 6% of men who have sex with men in an American cohort [2, 3].
The clinical impact of hepatitis virus infection in HIV-positive persons has progressively grown since the introduction of HAART, given the dramatic increase in survival experienced by these patients . On the other hand, HIV acts as a cofactor and accelerates the progression of liver disease due to hepatitis viruses. Thus, according to an American study, HIV infection increases up to 12.7-fold the risk of liver-related mortality in HBV-HIV–coinfected patients, especially in those with low CD4 cell counts . Similar findings have been reported for hepatitis C virus (HCV)– and HIV-coinfected patients . Not surprisingly, chronic viral hepatitis in recent years has become one of the most frequent causes of hospital admissions and mortality among HIV-infected subjects [5–7]. In this way, chronic hepatitis B and C actually behave as opportunistic infections in the context of HIV infection.
The goals of treatment of chronic hepatitis B may be categorized in several steps, from less to more ambitious. First, treatment should pursue suppression of HBV replication, as reflected by the achievement of significant reductions in and/or clearance of serum HBV DNA. Second, therapy may shift HBV infection from active (hepatitis B e antigen [HBeAg]–positive) to nonreplicative forms of chronic HBV infection, reflected by anti-HBe seroconversion. Third, ideally, any anti-HBV therapy should pursue disappearance of the chronic HBV carrier status (HBsAg positive), reflected by anti-HBsAg seroconversion.
As summarized in table 1, back-and-forth interference between HIV and HBV takes place when both infections coexist [8–29]. As a result, the treatment of chronic hepatitis B poses specific problems in the context of HIV infection (table 2). On the one hand, anti-HBV drugs show poorer performance, with lower response rates and faster selection of HBV-resistant strains. On the other hand, nucleoside analogues active against both HBV and HIV (e.g., lamivudine), if not used appropriately, can induce the selection of resistance mutations in the HIV genome. Therefore, the management of both infections should be carefully coordinated. In this review, we update the current view about the treatment of chronic hepatitis B in HIV-coinfected subjects.
Before the identification of HIV, 55 HIV- and HBV-coinfected patients were included in controlled trials assessing the efficacy of IFN-α for the treatment of chronic hepatitis B [30–35]. Only 3 of these studies considered separately the response rate according to HIV infection [32, 34, 35]. HIV coinfection was an exclusion criterion in all but 4 of the trials done since 1985 [36–39]. Therefore, the results of anti-HBV treatment with IFN-α in HIV-positive persons could be extrapolated from 7 studies, which are summarized in table 3 . Putting together the results of the 5 comparative trials, 7 of the 54 treated patients included in those trials had virological response, compared with none of the 46 untreated controls, with a statistical difference between groups of 0.08 (95% CI, 0.067–0.23). Response was defined as clearance of serum HBV DNA as determined by means of an hybridization method, which had a lower limit of detection of 106 copies/mL. The meta-analysis suggested that IFN-α is more effective than no treatment in achieving anti-HBe seroconversion. However, at least 10 patients needed to be treated to find 1 subject with a response. Long-term follow-up data for the patients included in those trials have not been reported. Neither CD4 T lymphocyte counts nor clinical data are given in these studies; therefore, predictors of response cannot be investigated.
Better responses were achieved in 2 more recent uncontrolled studies performed in France in which IFN-α was given to subjects with CD4 cell counts of >300 cells/mm3. Six (19%) of 31 treated patients responded and maintained sustained response, comparable to that seen in HIV-negative persons [38, 39]. However, in a more recent analysis by the same authors, the response in 76 patients (26 HIV-positive patients and 50 HIV-negative patients) treated for 6 months with IFN-α was higher in HIV-negative subjects (15% vs. 52%) . In HIV-positive patients, higher CD4 cell counts, in addition to higher baseline transaminase values, were identified as predictors of response. IFN-α therapy was associated with long-term anti-HBe seroconversion only in HIV-negative patients. Transient elevations in transaminase levels during treatment were significantly more frequent among HIV-infected persons, and flares correlated with lower CD4 cell counts.
The response to HBV treatment with IFN-α is worse when HCV and/or hepatitis D virus coinfections are present, a common situation for persons infected through injection drug use . In the setting of HIV infection, patients with chronic hepatitis B have significantly lower transaminase levels, an unfavorable predictor of response to IFN-α [32, 34–36, 39, 41]. Therefore, several factors may account for the lower response to IFN-α seen in HBV-HIV–coinfected patients (low CD4 cell counts, low transaminase levels, and frequent HCV coinfection). Because most studies of anti-HBV therapy in HIV-coinfected patients were conducted before HAART was introduced, it would be desirable to see the results of IFN-α for the treatment of HBV infection in the context of HAART.
In southern Europe, most HBV-infected patients harbor replication-competent HBV variants that are unable to produce HBeAg because of a mutation in the precore or core promoter HBV genome region. They retain or develop high levels of HBV DNA and persistent or intermittent elevations in alanine aminotransferase despite anti-HBe seroconversion . Treatment of this HBeAg-negative chronic hepatitis B has different goals (suppression of HBV replication rather than anti-HBe seroconversion) and different schedules and duration . IFN-α is the first-line treatment in HBV-monoinfected subjects with HBeAg-negative chronic hepatitis B . Currently, there are no data on the efficacy of IFN-α in HIV-infected patients with this type of hepatitis B. Finally, the new pegylated forms of IFN-α, which are more potent and convenient, probably would also offer better results in HIV-coinfected subjects. These studies need to be designed.
In summary, data on the effect of IFN-α against HBV in HIV-coinfected patients are scarce. Although they suggest that the response to IFN is superior to the response to no treatment, the response seems to be lower than in HIV-negative persons. The available data are insufficient to determine the optimal candidates for IFN-α treatment among HIV-HBV–coinfected patients. However, subjects with CD4 cell counts of >350 cells/mm3 and transaminase levels elevated to at least twice the upper limit of normal probably might get the greatest benefit from IFN-α therapy.
Lamivudine is a nucleoside analogue with both anti-HIV and anti-HBV activity. Although the inhibitory dosage for HBV is lower (100 mg/day) than that needed for HIV, 300 mg/day should be given when treating HBV-HIV–coinfected patients, and lamivudine should always be combined with ⩾2 other anti-HIV agents. Otherwise, lamivudine monotherapy rapidly selects HIV resistance mutations.
HBV replication has been proven to be inhibited by lamivudine in 86.4% (95% CI, 75.7–93.6) of HIV-HBV–coinfected patients. Table 4 shows results of the main studies that have examined the efficacy of lamivudine as an anti-HBV agent in HIV-coinfected patients [44–46]. Although anti-HBe seroconversion occurs in only a minority of patients taking lamivudine, many patients show improvements in liver histological findings, and reversal of liver decompensations is common in those with advanced cirrhosis [44, 45]. However, HBV mutations associated with lamivudine resistance are more rapidly and frequently selected in HIV-coinfected subjects, with incidences of 50% and 90% after 2 and 4 years of lamivudine treatment, respectively [44–48]. The development of lamivudine resistance mutations correlates with the duration of treatment but not with baseline HBV DNA levels, transaminase levels, or CD4 cell counts.
The emergence of lamivudine resistance mutations has been associated with elevations in transaminase levels in a minority of patients and, occasionally, with fatal liver failure [44–51]. Although resistance mutations may reduce HBV fitness and result in lower plasma HBV DNA levels, the histological and clinical benefit of continuing lamivudine treatment for subjects infected with drug-resistant HBV is unclear . However, in the absence of other anti-HBV drugs, it might be advisable to continue treatment with lamivudine despite the presence of resistance mutations, because hepatitis flares after discontinuation of lamivudine have been described [49, 50, 53, 54].
Adefovir is a nucleotide analogue recently licensed for the treatment of chronic hepatitis B at a dosage of 10 mg/day. It is able to suppress HBV replication and normalize alanine aminotransferase levels in up to 70% of patients with HBeAg-negative chronic hepatitis. It also induces HBeAg seroconversion in 23% of patients with HBeAg-positive chronic hepatitis B . It is active against lamivudine-resistant HBV mutants . Adefovir does not have significant antiretroviral effect at this dosage and could be given to HIV-infected patients who are not taking HAART. There is a theoretical risk of inducing HIV resistance, which deserves careful evaluation. However, mutations in HIV reverse-transcriptase codons 65 and 70, which occur in patients treated with higher doses of adefovir, were not observed in 13 HIV-HBV–coinfected patients treated for >1 year with adefovir plus HAART and showing uncontrolled HIV replication . Data about adefovir resistance from HIV-positive patients without HAART are not yet available. Adefovir could be given to HIV-infected patients receiving antiretroviral therapy without tenofovir, in whom it has been demonstrated to be safe and effective .
Lamivudine has lost favor in the treatment of HBV infection in HIV-coinfected patients since the approval of tenofovir for the treatment of HIV infection. Like adefovir, tenofovir is a nucleotide analogue reverse-transcriptase inhibitor with both anti-HIV and anti-HBV activity. It has been proven to be a very potent inhibitor of HBV replication in vitro, even in the presence of lamivudine resistance mutations . The results of trials conducted thus far assessing the efficacy of tenofovir against HBV in HIV-coinfected patients are shown in table 5. In these studies, HBV DNA levels decreased by 4 log10 on average [57–62], despite the fact that the majority of patients carried lamivudine resistance–associated mutations. No breakthroughs in HBV replication derived from the emergence of resistance mutations to tenofovir have been seen thus far, although the time of follow-up was short (⩽24 weeks) in the majority of these studies.
Of concern is the recent report of development of HBV resistance to adefovir. A mutation at codon 236 (N→T) within the HBV polymerase gene seems to be responsible for this effect, although it was selected only in 2 (1.6%) of 124 patients treated for >124 weeks with adefovir . Although this mutation does not result in cross-resistance to lamivudine, it might impair HBV susceptibility to tenofovir .
The new compounds being tested for the treatment of HBV infection may be grouped in 2 categories (table 6). The first includes drugs active against both HBV and HIV. Another group includes medications with activity against HBV alone. These latest drugs might be indicated preferentially for persons who have not yet met criteria for beginning HIV therapy.
Emtricitabine, a new nucleoside analogue, has been recently approved for the treatment of HIV infection. It is well tolerated and has potent anti-HBV activity. It should not be used after lamivudine failure, because both drugs show cross-resistance . Moreover, emtricitabine resistance mutations in HBV were selected in 19% of patients after 2 years of emtricitabine treatment .
Entecavir, another nucleoside analogue with anti-HBV activity, has been shown to be a potent inhibitor of HBV replication [67, 68]. Patients treated with entecavir at doses of 0.5 mg and 1 mg (as a single daily dose) had mean decreases in plasma HBV DNA levels of 2.8 and 2.5 logs, respectively, at 4 weeks, 3.8 and 4.4 logs at 24 weeks, and 4.5 and 5.1 logs at 48 weeks of treatment [69, 70]. No significant adverse events occurred, and 26% of subjects achieved undetectable HBV DNA levels at 48 weeks with both doses. The results were superior when transaminase levels were elevated at baseline. Entecavir has also been proven to be effective in patients previously treated with IFN-α and against lamivudine-resistant HBV strains .
Within the nucleoside analogue family, telbivudine and clevudine are compounds in earlier stages of development as anti-HBV drugs. Of interest, preliminary data suggest that the inhibition of HBV replication is superior over monotherapy with either drug when lamivudine and telbivudine are given in combination . The development of other nucleoside analogues with anti-HBV activity, such as L-Fd4C (2,3-Dideoxy-2,3-didehydro-β-L-fluorocytidine) and DAPD (amdoxovir), is in earlier phases, and data are scarce [72, 73].
Famciclovir, the prodrug of penciclovir, inhibits HBV replication, but it is less potent than lamivudine and shows cross-resistance . These facts, along with its thrice-daily dosing, have decreased the amount of interest in this drug as an anti-HBV agent. The results of studies of the efficacy of thymosin (thymic-derived peptides) are conflicting .
Finally, a new class of drugs, called heteroaryldihydropyrimidines (HAPs), is currently in preliminary stages of research as anti-HBV agents . They seem to act as inhibitors of the HBV nucleocapsid.
Recommendations for the Treatment of Chronic Hepatitis B in Hiv-Coinfected Patients
All HIV-infected persons with active HBV replication (positive for HBsAg and detectable HBV DNA) and elevated transaminase levels should be considered candidates for HBV treatment, even if criteria for HIV therapy are not met yet. There are 2 arguments supporting this recommendation. Hepatitis B behaves as an opportunistic infection in the context of HIV infection, with a more rapid progression to cirrhosis, and the response to anti-HBV agents diminishes as immunodeficiency progresses.
However, the optimal time for initiating anti-HBV therapy in coinfected patients has not been established yet nor is it clear which drug(s) should be used. An integrated approach against both HIV and HBV is needed. On the basis of the advice of different panels of experts [77, 78], if HBV infection is the only viral infection to be treated, IFN-α (preferably any of the pegylated forms) could be a valid option. However, given the low response to standard IFN-α and the lack of data on the effectiveness of pegylated IFN-α in HIV-coinfected subjects, it is desirable to treat these patients within clinical trials and in health care centers with experience.
Adefovir monotherapy (at dosages of 10 mg/day) may be considered a reasonable alternative option for treating HBV infection in HIV-positive subjects who do not met the criteria for HIV therapy. However, the potential risk for selection of resistance mutations in HIV with long-term use should be of concern , although this option is being supported by recent European guidelines for the treatment of hepatitis B .
The use of lamivudine, emtricitabine, or tenofovir in monotherapy should be avoided in HIV-HCV–coinfected patients, because these drugs favor the selection of HIV resistance mutations. Alternatively, some authors advocate providing a triple-antiretroviral regimen that includes either lamivudine, emtricitabine, and/or tenofovir and discontinuing it after complete HBV response is achieved. To minimize relapses, treatment should be discontinued not earlier than 6 months after response (i.e., anti-HBe seroconversion) has been obtained [77, 78]. However, because cure of hepatitis B is a remote prospect, maintenance of the treatment might be required.
In patients who meet criteria for the treatment of HIV infection, a HAART regimen containing lamivudine, emtricitabine, and/or tenofovir should be administered. Of note, tenofovir seems to show a higher activity against HBV than do emtricitabine or lamivudine and has a great genetic barrier for resistance. However, the experience with tenofovir is scarce, and long-term data are still lacking. Close follow-up assessment of renal function and phosphate plasma levels is required, because several cases of Fanconi syndrome associated with tenofovir use have been recently reported [80–82].
For patients who have already received lamivudine as part of HAART and who have detectable HBV viremia, resistance of HBV to the drug should be suspected. Emtricitabine is not useful for treatment of HBV infection in these patients. However, adding or replacing it with tenofovir is a reasonable approach. If HIV is already suppressed, adefovir might also be considered.
As for the role of combination therapy in HBV treatment, data are scarce. In a recent study among noncarriers of HIV that compared adefovir with adefovir and lamivudine given together, combination treatment was not found to be superior to adefovir alone in the first 12 months of treatment .