3TC Reduces Hepatitis B Viral Load

Dual Efficacy of Lamivudine Treatment in HIV/Hepatitis B Virus Coinfected Persons in a Randomized, Controlled Study (CAESAR)

Gregory J. Dore,1 David A. Cooper,1 Catherine Barrett,2 Li-Ean Goh,2 Bharat Thakrar,2 and Mark Atkins2 for the CAESAR Coordinating Committee

1National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Sydney, Australia; 2Glaxo Wellcome Clinical Research, Greenford, Middlesex, United Kingdom

Received 28 December 1998; revised 6 May 1999; electronically published 3 August 1999.

The efficacy and safety of lamivudine in persons coinfected with human immunodeficiency virus (HIV) type 1 and hepatitis B virus (HBV) were examined in the CAESAR study, a randomized placebo-controlled trial assessing the addition of lamivudine (150 mg 2x/day) or lamivudine (150 mg 2x/day) plus loviride (100 mg 3x/day) to zidovudine-containing background antiretroviral treatment. Baseline hepatitis B surface antigen (HBsAg) results were available for 1790 study subjects, of whom 122 (6.8%) tested positive. Retrospective analyses for serial HBV DNA, HBsAg, and hepatitis B e antigen (HBeAg) were performed on stored sera from 118 HBsAg-positive subjects. HBV DNA and HBeAg were present in 83% and 63%, respectively. At weeks 12 and 52, median log10 HBV DNA change was -2.0 and -2.7, respectively, in the lamivudine arms, compared with no reduction among placebo recipients (P < .001). A trend to lower alanine transferase level, and delayed progression of HIV-1 disease (relative hazard, 0.26; 95% confidence interval, 0.080.80) were also seen in the lamivudine arms, compared with the placebo group.
------------------------------------------------------------------------
The natural history of hepatitis B virus (HBV) infection is modified by coinfection with human immunodeficiency virus (HIV) type 1. After initial HBV infection, both development and persistence of chronic HBV infection are greater among people with prior HIV-1 infection [1, 2]. Furthermore, higher HBV load, as measured by serum HBV DNA polymerase activity [2], and reactivation of HBV infection in chronic carriers who have lost detectable serum hepatitis B e antigen (HBeAg) [35] are well described for people with HIV-1 coinfection. The altered natural history of HBV infection and recent improvements in HIV-1 survival through combination antiretroviral therapy [6, 7] emphasize the need to address HBV-related morbidity in people with coinfection.

Among persons with chronic HBV infection, in phase II trials daily doses of 100600 mg of lamivudine resulted in HBV DNA suppression below the lower limit of detection by the assay (106 copies/mL; Abbott, Abbott Park, IL) in nearly 100% of patients within 14 weeks [1214]. A recently completed phase III placebo-controlled trial among 358 Asian patients with chronic HBV infection demonstrated reduced progression of liver fibrosis and significantly enhanced HBeAg clearance after 12 months of therapy with 100 mg/day lamivudine [15].

Although there have been isolated case reports and a small series of people with HIV-1/HBV coinfection treated with lamivudine [16, 17], a randomized controlled trial has not been conducted in this patient population. The CAESAR study was a randomized, double-blind, placebo-controlled trial of addition of lamivudine or lamivudine plus loviride to zidovudine-containing regimens for persons with advanced HIV-1 infection [11]. In order to assess the efficacy and safety of lamivudine in an HIV-1/HBVcoinfected population, a retrospective analysis of coinfected patients was undertaken over the 52-week CAESAR study.

Methods

The CAESAR trial was conducted in Canada, Australia, Europe, and South Africa among HIV-1positive patients with 25250 CD4 cells/L. A 1:2:1 randomization to placebo, lamivudine (150 mg 2x/day), or lamivudine (150 mg 2x/day) plus loviride (100 mg 3x/day) in addition to a current treatment regimen (either zidovudine monotherapy or zidovudine plus didanosine or zalcitabine combination therapy) was done. Major inclusion criteria included age 18 years, Karnofsky score 70, and no previous use of lamivudine.

A total of 1895 patients were randomized: 483 to placebo, 937 to lamivudine, and 475 to lamivudine plus loviride. Patients with positive baseline hepatitis B surface antigen (HBsAg) were included in this substudy. Hepatitis B serology was done retrospectively on stored sera (baseline; weeks 4, 8, 12, and 8-week HBsAg and HBeAg/antibody determinations). At the same time points, HBV DNA was measured by Amplicor HBV polymerase chain reaction (PCR) monitor assay (quantitation range, 40040,000,000 copies/mL; Roche, Nutley, NJ) as was liver function. Data on HIV-1 progression were available for analysis. The primary end points of the CAESAR trial were either progression to a new protocol-defined AIDS event, according to the 1993 CDC case definition [18], or death.

Results

Of the 118 coinfected patients with assessment of HBV DNA and HBeAg status, HBV DNA was detectable by PCR in 98 (83%) and HBeAg was present in 74 (63%). Baseline HBV DNA detectability, median log10 HBV DNA, and HBeAg positivity were similar in the lamivudine and placebo arms (table 2).

At week 12, median log10 HBV DNA change was -2.0 and 0.0 in the lamivudine and placebo arms of the study, respectively (P < .001). Significant HBV DNA suppression in the lamivudine arms was maintained throughout the study period, with a median log10 HBV DNA change of -2.7 and 0.0 at 52 weeks in the lamivudine and placebo arms, respectively (P< .001; figure 1). By Kaplan-Meier analysis, HBV DNA loss reached 30% by week 12 and 40% at week 52 in the lamivudine arms, compared with 5% and 10% at the same time points in the placebo arm (log-rank test, P = .018; figure 2). Among patients tested for HBeAg at baseline and week 52, there was a trend to greater loss of HBeAg in the lamivudine arms (7/32) than in the placebo arm (0/7; P = .31).

Over the study period, a trend to lower mean ALT level (74.754.7 U/L) and greater normalization of ALT was also seen in the lamivudine arms, compared with the placebo arm (68.274.9 U/L; figure 3). Although there was a higher prevalence of hepatic flare (increase in ALT to >5x upper limit of normal) in the HIV-1/HBVcoinfected population (11.0%) than in the HIV-1only CAESAR study population (2.2%; P < .0001), there were no significant differences between the lamivudine (11.6%) and placebo (8.7%; P = .69) arms within the HIV-1/HBVcoinfected population.

Progression of HIV disease was delayed in the lamivudine arms, as compared with the placebo groups (relative hazard, 0.26; 95% confidence interval, 0.080.80). This finding is similar to what was seen in the HIV-infectedonly population from the CAESAR study (figure 4). The incidence of serious adverse events was lower in the combined lamivudine arms (20%) than in the placebo arm (32%).

Discussion

In this study, lamivudine displayed dual efficacy in the HIV-1/HBVcoinfected population through both delayed HIV-1 disease progression and a favorable biochemical and HBV virologic response. Twelve months of lamivudine therapy produced HBV DNA and HBeAg losses of 40% and 20%, respectively, and more ALT normalization than in the placebo group. Furthermore, the 75% reduction in HIV-1 disease progression achieved by lamivudine in the HIV-1/HBVcoinfected population was similar to the 60% reduction seen in the total CAESAR study population [11].

The 40% HBV DNA loss (<400 copies/mL) during the study period contrasts with earlier studies of lamivudine therapy among HIV-negative chronic HBV carriers, in which nearly 100% of people had loss of HBV DNA within 14 weeks of initiation of therapy at doses of 100 mg [1215]. However, these studies used considerably less-sensitive HBV DNA detection methods. Other factors, such as higher HBV DNA levels in association with HIV-related immunodeficiency [2] or the relative absence of immune-mediated support for the anti-HBV action of lamivudine in people with HIV-1 infection, may partly explain the apparently lower response rate. The CAESAR study population had advanced HIV-1 disease with 25250 CD4 cells/L; therefore, a more pronounced suppression of HBV DNA may be seen in a less immunodeficient HIV-1infected population.

The use of lamivudine monotherapy for people with chronic HBV infection has produced favorable biochemical, virologic, and histologic outcomes [1214, 19]. In the largest study to date, conducted in Asia (Hong Kong, Singapore, Taiwan), lamivudine (at 100 mg/day) produced significant histologic improvement in most patients; retarded the progression of hepatic fibrosis, compared with placebo; and produced HBeAg seroconversion in 16% of patients, versus 4% for placebo [15]. Our study of patients coinfected with HIV-1 produced a similar rate of HBeAg seroconversion and substantial HBV DNA suppression and ALT normalization.

Although this study demonstrates lamivudine efficacy in the HIV-1/HBV population, there are some limitations. The study was based on a retrospective analysis within a prospective randomized trial assessing the efficacy of lamivudine in HIV-1 disease. The HIV-1/HBVcoinfected study population analyzed was only 7% of the total CAESAR study population, and the study randomization was not done on the basis of HBV infection. However, the similar baseline characteristics for the placebo and lamivudine arms in the HIV-1/HBVcoinfected study population suggest that the latter limitation is not a major concern. The primary end point in the CAESAR study was HIV-1 clinical disease progression, whereas the end points in this study were virologic and biochemical. Liver histologic changes were not assessed in this study. Suppression of HBV DNA by lamivudine in randomized studies of persons with chronic HBV infection has been accompanied by significant histologic and biochemical improvement and increased HBeAg seroconversion [15, 19], with histologic responses even in persons without seroconversion. The combination of a favorable virologic and biochemical response in this study, as measured by HBV DNA suppression and ALT normalization, suggests a favorable histologic response but requires validation in a larger histologically assessed study of HIV-1/HBVcoinfected persons. Another limitation is the nonavailability of virologic and biochemical data beyond week 52 of the CAESAR study period. Because clinically significant recurrent hepatitis has been reported after cessation of lamivudine therapy in a small number of patients with chronic HBV infection [20] it is important to follow-up on coinfected patients after cessation of therapy. However, the vast majority of CAESAR study subjects in the lamivudine-containing arms chose to continue therapy after the 52-week point (CAESAR Coordinating Committee, unpublished data).

The lamivudine response rate is more than comparable to that achieved with interferon (IFN)- in a recent study of HIV-1/HBVcoinfected persons [21]. In a considerably less immunodeficient study population (mean CD4 cells, 480/mm3), 6 months of IFN- produced HBV DNA and HBeAg loss rates of 36% and 8%, respectively [21]. Of note, the HBV DNA loss was 17% in the control population compared with 10% in our study. The response rates are particularly encouraging considering the relatively advanced stage of HIV disease in our study population and previous evidence of poorer HBV therapeutic response with IFN- among HIV-coinfected patients [22].

An important area of further therapeutic research among the HIV-1/HBVinfected population is assessment of antiretroviral therapy combinations that include lamivudine and a protease inhibitor, which should provide significantly greater antiHIV-1 effect and immune restoration than the antiretroviral therapy used in the CAESAR study. A recent report of an HIV-1/HBVcoinfected person with an exacerbation of hepatitis but subsequent suppression of HBV viremia, ALT normalization, and HBeAg seroconversion shortly after initiation of ritonavir therapy, suggests that an improved immune response associated with reductions in HIV-1 viremia may alone have a significant anti-HBV effect [23].

There have been reports of both HBV and HIV-1 with reduced sensitivity to lamivudine in vitro, due to mutations at the YMDD domain of the HBV DNA polymerase [24] and HIV-1 reverse transcriptase [25, 26] ; therefore another crucial area of research is the clinical importance of lamivudine resistance in the HIV-1/HBVcoinfected population. Indeed, the incidence of YMDD mutants and clinical consequences is currently being analyzed in this study population.