What Drives Hepatitis B Virus-Related Hepatic Flares? Virus, T Cells--Or a Bit of Both?

In this issue, Drake and colleagues [1] meticulously describe 2 HIV-infected patients with chronic hepatitis B virus (HBV) infection who develop severe hepatitis flares and hepatic decompensation while receiving lamivudine-containing HAART regimens. They argue that HAART-associated immune restoration played a major role in these clinical events. The first patient recommences treatment with a lamivudine-containing regimen after a 3-month hiatus, with a resultant flare in results of liver function tests. Coincident with this flare is a robust increase in CD4 cell count and percentage. The patient continues to experience hepatic decompensation, but her condition dramatically improves during treatment with lamivudine and tenofovir, with attainment of HBV virological suppression. Although we are not given information about serological test results obtained prior to the initiation of therapy, it appears that an HBV precore mutant has emerged. The second case is not nearly as clear-cut as the first, as the authors themselves concede. The patient is given diagnoses of salmonella sepsis, prior hepatitis D infection, and HBV-related cirrhosis and develops hepatic decompensation after initiation of lamivudine-containing HAART. Tenofovir is added to the regimen, along with steroids to quell the effects of immune reconstitution, but the patient's health continues to deteriorate, and she dies. The role of steroids in immune reconstitution is not clear, because HBV virus load increases with their administration, and hepatic decompensation has occurred with steroid withdrawal [2, 3]. Although the authors argue that HAART-induced immune reconstitution is the direct mechanism responsible for these 2 cases of hepatic decompensation, the second case is too much confounded by infection, poor hepatic reserve, and polypharmacy with other potential hepatotoxic medications (such as dilantin, ritonavir, and indinavir) to clearly support their contention. No dilantin levels are reported, and this information is critical for treatment of a cirrhotic patient. Also, the package insert for indinavir recommends decreased dosing of this medication for patients with advanced liver disease. Severe hepatitis has been reported with indinavir treatment, and pharmacokinetic boosting with ritonavir may have elevated drug serum levels to a toxic range [4]. Furthermore, the increases in absolute CD4 cell count and percentage in the second patient were modest, at best. The case of the first patient is clearly a more compelling example and is supportive of the authors' argument that HAART-induced immune reconstitution plays an important role in the patient's clinical presentation. However, we must also consider the natural history of HBV disease and the flares seen in association with lamivudine therapy in order to fully appreciate this illustrative case. Four main concepts regarding chronic HBV infection should be highlighted: The majority of patients with chronic HBV infection, particularly those who are hepatitis B e antigen (HBeAg) seronegative, experience a recurrence of HBV viremia after discontinuing treatment with lamivudine [5]. 2. Flares of transaminase levels can occur with lamivudine treatment and are associated with more-favorable clinical outcomes [6]. 3. As HBV viremia decreases with effective therapy, CD4+ and CD8+ T cell responses to HBV-specific antigens improve [7]. 4. Spontaneous flares have been attributed to changes in the delicate balance between immunologic responses to HBV and quantitative variations in viremia [2, 8, 9]. Flares of transaminase levels can occur with lamivudine treatment and are associated with more-favorable clinical outcomes [6]. As HBV viremia decreases with effective therapy, CD4+ and CD8+ T cell responses to HBV-specific antigens improve [7]. Spontaneous flares have been attributed to changes in the delicate balance between immunologic responses to HBV and quantitative variations in viremia [2, 8, 9]. Concomitant infection with HIV only complicates the situation further. The risk of HBV chronicity and HBV disease reactivation increases with HIV coinfection [10, 11]. In addition, withdrawal of lamivudine therapy has been temporally correlated with HBV disease reactivation and hepatic decompensation in HIV-infected patients with advanced liver disease [12]. Some authors have suggested that successful treatment of HIV infection may lead to a decrease in TNF levels and disinhibition of HBV replication [13]. In the first clinical case, the qualitative change of HBV-specific T-cell responses secondary to effective HBV infection plus the additive effect of HAART-associated immune reconstitution may have been just the right ingredients for “the perfect storm.” A plausible sequence of events includes: (1) lamivudine-induced virological suppression followed by a rebound when treatment with the drug is withdrawn, (2) subsequent virus load suppression with reintroduction of antiviral therapy, (3) restoration of HBV-specific T cell responsiveness with a subsequent flare of transaminase levels, and (4) exacerbation of a hepatitis flare with HAART-induced immune restoration. Drake et al. [1] also raise awareness of other pressing clinical issues that arise when HIV providers of treatment for HIV disease begin HAART for patients with chronic viral hepatitis. (1) Knowledge of the serologic status of a patient with HBV infection is of paramount importance before prescribing medications, such as lamivudine and tenofovir, that have dual activity against HIV and HBV. (2) Drug holidays from HAART can have devastating consequences if treatment with agents active against HBV is stopped and loss of control of HBV infection ensues. Drake and colleagues [1] suggest strategies for treatment of coinfected patients, including the use of combination therapy with lamivudine and tenofovir in HAART regimens to attain and maintain HBV virological suppression. Although this approach is not yet proven, it is based on the sound concern that the risk of development of lamivudine-resistant virus increases over time in HIV-infected patients [14]. The authors also suggest delaying the initiation of full-dose HAART, to slow the onset of immune reconstitution. This is clearly an option for treatment of patients infected with HIV resistant to lamivudine and tenofovir but is less desirable for those patients who are also depending on these medications for HIV suppression. However, the larger question is, do we want to delay the onset of immune reconstitution? Immune reconstitution can be thought of as a “double-edged sword” in patients infected with HBV. Although an intact immune system is responsible for hepatocyte injury, it is also vital to virus clearance [15]. HBeAg seroconversion represents an important transition from active hepatitis to an inactive carrier state in which there is little evidence of inflammatory activity and no detectable viremia [16]. HBV seroconversion has occurred after profound improvement in immune responses secondary to HAART, even in the absence of effective HBV therapy [17]. However, these severe hepatitis flares can be dangerous in the event of poor hepatic reserve, so the clinician must be knowledgeable about the status of his or her patient's liver disease. Drake et al. [1] should be applauded for highlighting again the fact that providers of treatment for HIV infection can no longer dispense HAART “in a vacuum” for patients with chronic viral hepatitis. We must be aware of the dual purposes of lamivudine, tenofovir, and emtricitabine and know the HBV status of our patients prior to initiation or withdrawal of drug treatment. Finally, infectious diseases training programs need to integrate hands-on clinical experience in subspecialty clinics so fellows can become well-versed in the complicated management of HIV-infected patients with chronic viral hepatitis. Financial support. Support was provided through the General Clinical Research Center funded by the National Center for Research Resources of the National Institutes of Health (grant no. MO1-RR00054). Conflict of interest. No conflict.