A randomized study of adefovir dipivoxil in place of HBIG in combination with lamivudine as post–liver transplantation hepatitis B prophylaxis


  • Peter W. Angus,

    Corresponding author
    1. Victorian Liver Transplant Unit, Austin Health, Victoria, Australia
    • Liver Transplant Unit, Austin Health, Studley Road, Heidelberg, VIC 3084, Australia
    Search for more papers by this author
    • Peter Angus, Geoffrey McCaughan, and Simone Strasser received minor honoraria from Gilead Sciences for participation in Advisory Boards or Speakers' Bureau.

    • fax: (61)-3-94963487

  • Scott J. Patterson,

    1. Victorian Liver Transplant Unit, Austin Health, Victoria, Australia
    Search for more papers by this author
  • Simone I. Strasser,

    1. AW Morrow Gastroenterology and Liver Center, Royal Prince Alfred Hospital, New South Wales, Australia
    Search for more papers by this author
    • Peter Angus, Geoffrey McCaughan, and Simone Strasser received minor honoraria from Gilead Sciences for participation in Advisory Boards or Speakers' Bureau.

  • Geoffrey W. McCaughan,

    1. AW Morrow Gastroenterology and Liver Center, Centenary Research Institute, Australian Liver Transplant Unit, Royal Prince Alfred Hospital, University of Sydney, New South Wales, Australia
    Search for more papers by this author
    • Peter Angus, Geoffrey McCaughan, and Simone Strasser received minor honoraria from Gilead Sciences for participation in Advisory Boards or Speakers' Bureau.

  • Edward Gane

    1. New Zealand Liver Transplant Unit, Auckland City Hospital, Auckland, New Zealand
    Search for more papers by this author

  • Potential conflict of interest: Dr. Angus advises and received grants from Gilead. Dr. Strasser advises and is on the speakers' bureau of Gilead. Dr. McCaughan advises Gilead.


Prior to effective prophylaxis, liver transplantation for hepatitis B virus (HBV)-related disease was frequently complicated by recurrence, which could be severe and rapidly progressive. Combination hepatitis B immunoglobulin (HBIG) and lamivudine prophylaxis reduces this rate of recurrence to <5% at 5 years; however, HBIG administration is costly and inconvenient. We conducted a multicenter randomized study of adefovir dipivoxil substitution for low-dose intramuscular (IM) HBIG in patients without HBV recurrence at least 12 months posttransplantation for HBV-related disease. Thirty-four patients were randomized, 16 to adefovir (1 patient withdrew consent at 3 months and is not considered in the results) and 18 to continue HBIG. All continued lamivudine. Groups were well matched by age, sex, and time since transplantation (median, 4.5 years), and background virological risk for HBV recurrence (30% of patients in the adefovir group, 24% in the HBIG group having detectable HBV DNA at transplantation). All patients were alive at study completion without recurrence. One patient in the adefovir group became hepatitis B surface antigen–positive at 5 months but was persistently HBV DNA undetectable via polymerase chain reaction (sensitivity 14 IU/mL) over the following 20 months. Median creatinine was not significantly changed over the course of the study in either group. One patient in the adefovir group with a background of diabetic and hypertensive nephropathy (baseline creatinine 150 μmol/L) developed increased creatinine leading to dose reduction and ultimately cessation of adefovir at 15 months. Yearly cost of combination adefovir/lamivudine prophylaxis was $8,290 versus $13,718 IM HBIG/lamivudine. Conclusion: Compared with combination HBIG plus lamivudine prophylaxis, combination adefovir plus lamivudine provides equivalent protection against recurrent HBV infection but with better tolerability and less cost. (HEPATOLOGY 2008.)

Prior to the advent of effective posttransplantation antiviral prophylaxis, liver transplantation for hepatitis B virus (HBV)-related disease was usually followed by immediate HBV infection of the allograft.1–3 In the setting of posttransplantation immunosuppression, recurrent HBV infection was associated with severe liver injury, which often progressed rapidly to graft loss and death.2, 4

Over the past 15 years, advances in antiviral prophylaxis have reduced the risk of HBV recurrence significantly. Long-term intravenous (IV) high-dose hepatitis B immunoglobulin (HBIG) administration5, 6 and lamivudine monotherapy7 both reduce recurrence rates from over 90% to less than 50% in patients who are HBV DNA–positive prior to transplantation. More recently, several studies have shown that these two approaches are synergistic and that combination lamivudine plus IV HBIG reduces recurrence rates to less than 5% at 5 years.8 Consequently, this combination is the accepted prophylaxis regimen at most transplantation units. However, a major drawback of this therapy is that high-dose intravenous HBIG is very expensive, and long-term IV administration of HBIG each month is inconvenient and unpleasant for patients.

Several strategies have been used to reduce the cost associated with long-term high-dose IV HBIG administration. These include cessation of HBIG at some interval posttransplantation while continuing lamivudine,9–12 vaccination of patients prior to cessation of HBIG,13–18 and substitution of high-dose IV HBIG preparations with low-dose intramuscular (IM) HBIG.19–21 Another possible approach is to replace HBIG with a direct antiviral drug such as adefovir dipivoxil, because recent studies have demonstrated that the risk of virologic breakthrough during combination adefovir and lamivudine therapy is negligible.22, 23 The combination of both drugs has been shown to be safe and effective in transplant recipients with lamivudine resistance.24, 25 We therefore conducted a randomized study comparing the safety and efficacy of replacing HBIG with adefovir in combination with lamivudine versus continuing HBIG plus lamivudine in patients at least 1 year posttransplantation for HBV-related disease.


ADV/LAM, adefovir dipivoxil and lamivudine; HBIG, hepatitis B immunoglobulin; HBIG/LAM, HBIG and lamivudine; HBsAb, hepatitis B surface antibody; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; IM, intramuscular; IV, intravenous; PCR, polymerase chain reaction.

Patients and Methods

This was a randomized, open-label, multicenter trial conducted at the Victorian (Austin Health, Melbourne), New South Wales (Royal Prince Alfred Hospital, Sydney), and New Zealand (Auckland City Hospital) liver transplantation units between August 2004 and October 2006. The trial was approved by the appropriate institutional ethics review committees, and all patients gave written informed consent prior to entry into the study. No patient had received donor organs obtained from executed prisoners or other institutionalized persons.

All adult patients who were at least 12 months posttransplantation for HBV-related disease and who were receiving protocol low-dose IM HBIG and lamivudine prophylaxis without evidence of graft reinfection (defined below) were eligible. Patients were excluded if they had undergone transplantation for fulminant hepatic failure, had a serum creatinine level of ≥180 μmol/L, were human immunodeficiency virus–seropositive (or other significant comorbidity), or were pregnant or lactating. Patients with hepatitis C or hepatitis delta coinfection were included.

Patients were randomized to receive either adefovir dipivoxil 10 mg/day (Gilead Sciences, East Melbourne, Victoria, Australia) or to continue their current prophylaxis with low-dose (800 IU/month) IM HBIG (CSL Bioplasma, Parkville, Victoria, Australia). Randomization was stratified by center and was accomplished through sequential assignment of treatment numbers from lowest to highest, cross-referenced to a separately administered randomization spreadsheet. All patients continued to receive lamivudine 100 mg/day (GlaxoSmithKline Australia, Boronia, Victoria, Australia). Patients were reviewed (physical examination, physician assessment, urine pregnancy test for females of child-bearing potential) at least monthly for the first 3 months and at least every 3 months thereafter. At each study visit, blood was taken for routine hematology and biochemistry, prothrombin time, hepatitis B surface antigen (HBsAg) and hepatitis B surface antibody (HBsAb). In addition, serum was stored at −70°C for possible later reflex (if HBsAg-positive) HBV DNA testing and resistance mutation sequencing.

The primary endpoint of the study was recurrence of HBV infection, defined as the reappearance of both HBsAg and HBV DNA in serum. The secondary endpoint was to conduct an economic evaluation of the costs and outcomes of combination adefovir and lamivudine (ADV/LAM group) prophylaxis compared with combination HBIG and lamivudine (HBIG/LAM group) prophylaxis. Analysis of data was performed when all patients reached 48 weeks postrandomization.

Baseline Antiviral Prophylaxis.

The low-dose IM HBIG protocol used in all patients included in the study has been described.20 Prior to transplantation, patients with detectable serum HBV DNA (hybridization or polymerase chain reaction [PCR]) received lamivudine 100 mg/day. After transplantation, patients received lamivudine 100 mg/day plus IM HBIG 800 IU/day for 1 week, then monthly thereafter. At one center (New South Wales), 400 IU HBIG was administered unless the trough HBsAb fell below 50 IU/L when the dose was increased to the 800 IU dose.

HBV Assays.

Viral markers (HBsAg, HBsAb, hepatitis B e antigen, and antibody) were measured using standard commercial assays. HBV recurrence at the time of screening for entry into the study was excluded via seronegativity for both HBsAg and HBV DNA using a standardized PCR assay (Roche COBAS AMPLICOR HBV monitor; Roche Molecular Systems, Branchburg, NJ) at a lower level of detection (300 viral copies/mL). HBV DNA at the latest reported follow-up was assessed via various highly sensitive real-time PCR assays (COBAS TaqMan HBV test, Roche Molecular Systems, Branchburg, NJ; Abbott RealTime HBV, Abbott Molecular, Des Plaines, IL; or Artus HBV LC PCR Kit, QIAGEN, Hilden, Germany) according to the local reference laboratory used by the transplantation unit. The lower limit of detection for the three assays is quoted as 12 IU/mL, 15 IU/mL, and 14 IU/mL, respectively.

Prior to study baseline, the three transplantation centers used a variety of HBV DNA assays, depending on the year of transplantation and assay availability. The results of this testing were collected retrospectively in order to characterize the virological background of the groups. Results of HBV DNA testing were sought to determine the viral load at the commencement of antiviral therapy pretransplantation, the time when HBV viral load became undetectable, and the HBV DNA level at transplantation. Assays that reported results in pg/mL or copies/mL were converted to IU/mL using the manufacturer's recommended conversion multiple.

Statistical Methods.

Continuous variables are expressed as the median (range) unless specified otherwise. Categorical data are presented as the number (percentage). Cumulative patient survival and HBV recurrence-free survival were calculated using the Kaplan-Meier method. The Wilcoxon signed-rank test was used to detect change in serum creatinine over the course of the study. P values <0.05 were considered statistically significant.


Between September 2004 and July 2006, there were 147 patients who were potentially eligible for this study. These were post–liver transplantation patients in Australia or New Zealand (without pretransplantation lamivudine resistance) who had received low-dose HBIG and lamivudine prophylaxis. Of these 147 patients, 24 underwent transplantation at sites not involved in this study, 20 had died (one with HBV recurrence), and 4 were alive but had developed HBV recurrence and were started on adefovir as rescue therapy. This left 99 potential candidates who met inclusion criteria and were being treated at investigational sites. All eligible patients were approached unless they had known comorbidity that precluded inclusion (8 patients with significant renal impairment, 2 patients with tumor, 3 patients who had been transplanted for fulminant hepatitis B, and 2 patients with human immunodeficiency virus), or resided distant to the treating unit making the trial impractical (18 patients). This left 66 patients, 34 of whom were recruited (16 in New Zealand, 9 in Victoria, and 9 in New South Wales). The final obstacle to recruitment was failure to give consent, because a significant proportion (32/65 [48%]) of eligible patients preferred to remain on their current therapy. Over the three sites, 16 patients were randomized to ADV/LAM and 18 to HBIG/LAM.

The demographic details of the 34 patients are included in Table 1. The median age (55.0 in the ADV/LAM group, 51.9 in the HBIG/LAM group), proportion of male patients (13/16 [81%] in the ADV/LAM group, 15/18 [83%] in the HBIG/LAM group), and ethnic background of the groups were similar. Median time since transplantation until baseline was 4.4 years (range, 1.1-6.3 years) in the ADV/LAM group and 4.6 years (range, 1.3-7.2 years) in the HBIG/LAM group. Approximately half of all patients (10/16 [63%] in the ADV/LAM group versus 8/18 [44%] in the HBIG/LAM group) had a history of hepatocellular cancer prior to their transplantation. One (6%) patient in the ADV/LAM group had hepatitis C coinfection (none in the HBIG/LAM group) and 2/18 (11%) patients in the HBIG/LAM group had hepatitis delta coinfection (none in the ADV/LAM group).

Table 1. Patient Demographics
 ADV/LAM Group n = 16)HBIG/LAM Group (n = 18)
Median age (range)55 (26-67)51.9 (37-69)
Median years since transplantation (range)4.4 (1.1-6.3)4.6 (1.3-7.2)
Male sex13 (81%)15 (83%)
 Asian6 (38%)7 (39%)
 European5 (31%)5 (28%)
 South Pacific5 (31%)6 (33%)
Comorbid liver disease etiology  
 Hepatocellular cancer10 (63%)8 (44%)
 Hepatitis C1 (6%)0 (0%)
 Hepatitis delta0 (0%)2 (11%)
 Alcohol0 (0%)1 (6%)

Patient histories were reviewed to determine potential risk factors for posttransplantation HBV recurrence and antiviral breakthrough (Table 2). Because these data predate recruitment into the study, a variety of HBV DNA assays were used, some of which were insensitive by modern standards. Nevertheless, the majority of patients had detectable HBV DNA at commencement of lamivudine therapy pretransplantation, and the median HBV load prior to lamivudine therapy was similar in both groups. The duration of lamivudine therapy prior to liver transplantation was also similar in both groups. At transplantation, 7/30 patients (4 unknown) had detectable HBV DNA (3 in the ADV/LAM group, 4 in the HBIG/LAM group).

Table 2. Background Virological Risk Factors for Recurrence
 ADV/LAM Group (n = 16)HBIG/LAM Group (n = 18)
HBV DNA detectable prior to lamivudine (pretransplantation)88%70%
Median detectable HBV DNA load prior to lamivudine (pretransplantation)4.75 log10 IU/mL5.5 log10 IU/mL
Time on lamivudine (pretransplantation)4.1 months (1.2-13.5)4.2 months (0.7-47.9)
Detectable HBV DNA at transplantation3 (2 patients ≤ 4.4 log10 IU/mL, 1 patient 4.4-6.4 log10IU/mL), 10 negative, 3 unknown4 (2 patients ≤ 4.4 log10 IU/mL, 1 patient 4.4-6.4 log10IU/mL, 1 patient ≥ 6.4 log10IU/mL), 13 negative, 1 unknown

Patient Survival and Withdrawal from Trial.

All patients in both groups were alive at the time of manuscript preparation. Median follow-up was 21.1 months (range, 9.4-35.9 months) in the ADV/LAM group and 21.8 months (range, 13.5-35.6 months) in the HBIG/LAM group. At 3 months following randomization, 1 patient in the ADV/LAM group withdrew consent from the study and is not considered in further results. This patient was switched back to HBIG/LAM and has not developed recurrence with a further 28 months of follow-up.

HBsAb Levels.

The median trough HBsAb level in the HBIG/LAM group was approximately 50 IU/L throughout the study (Figs. 1 and 2). As expected, HBsAb levels fell rapidly in the ADV/LAM group, and 6 months after HBIG cessation, no patient was found to have an HBsAb level >10 IU/L.

Figure 1.

HBsAb versus Time: HBIg/LAM group.

Figure 2.

HBsAb versus Time: ADV/LAM group.

Recurrent Hepatitis B.

There were no cases of recurrence (defined as detectable HBsAg and HBV DNA via highly sensitive real-time PCR assay) in either arm of the study (Table 3). One patient in the ADV/LAM group had detectable HBsAg (titer 0.05 IU/mL) at 5 months; however, HBV DNA was undetectable via PCR (Artus RealArt, lower limit of detection 14 IU/mL) at the time and when tested repeatedly over the following 24 months. The patient displayed no clinical or biochemical evidence of hepatitis at any time.

Table 3. HBV Recurrence at Latest Follow-up
 Positive HBsAgDetectable HBV DNA
ADV/LAM group1/15 (6%)0/18 (0%)
HBIG/LAM group0/15 (0%)0/18 (0%)

Safety and Renal Function.

Median creatinine was unchanged over the course of the study in both groups (see Table 4). In 1 patient in the ADV/LAM group, serum creatinine rose significantly following inclusion in the study, necessitating adefovir dose reduction and finally withdrawal from the study. This patient had pre-existing diabetic and hypertensive nephropathy, and in the 2 years prior to baseline serum creatinine had fluctuated between 130 μmol/L and 180 μmol/L. Baseline creatinine at study entry was 150 μmol/L. Over the next 15 months, adefovir was dosed according to the manufacturer's recommendation. At 15 months, serum creatinine was persistently above 230 μmol/L. Adefovir therapy was ceased, and the treatment was switched back to HBIG/LAM therapy. At that time there was no evidence of recurrence, and the patient has remained HBsAg– and HBV DNA–negative. Serum creatinine has not improved since cessation or adefovir therapy, fluctuating between 210 μmol/L and 250 μmol/L.

Table 4. Renal Function
Creatinine (μmol/L)Start MedianEnd MedianChange Over Study
ADV/LAM group (n = 15)107 μmol/L108 μmol/L+1 μmol/L
HBIG/LAM group (n = 18)109 μmol/L108 μmol/L−1 μmol/L

There were no other significant adverse advents considered related to therapy in either arm of the study.

Costs of Therapy.

Cost elements (Table 5) were obtained from the largest single center (the New Zealand Liver Transplant Unit) and converted into U.S. dollars (exchange rate: US $1 = New Zealand $1.31). Total costs include drug (adefovir dipivoxil 10 mg/day and lamivudine 100 mg/day versus HBIG 800 IU/month and lamivudine 100 mg/day), attendance (quarterly versus monthly for IM HBIG injections), and minimum mandatory serological testing (quarterly HBsAg versus monthly HBsAb testing). Societal costs (patient time, travel, and loss of income) are not included. Equal efficacy in prophylaxis was assumed.

Table 5. Cost Elements
Cost ElementCost per YearSource
HBIG$10,410New Zealand Pharmaceutical Schedule
Lamivudine$1,426New Zealand Pharmaceutical Schedule
Adefovir$6,236New Zealand Pharmaceutical Schedule
Visits and serology (HBIG/LAM group)$1,882Auckland Hospital Schedule
Visits and serology (ADV/LAM group)$627Auckland Hospital Schedule

The total cost per year in the ADV/LAM group was $8,290, and the total cost per year in the very low-dose IM HBIG/LAM group was $13,718, yielding a yearly saving of $5,428 in favor of combination ADV/LAM prophylaxis.


Several strategies have been proposed to avoid the costs and inconvenience associated with long-term, high-dose IV HBIG administration for post–liver transplantation hepatitis B prophylaxis. In Australia and New Zealand, the lack of availability of an intravenous HBIG preparation has lead to the adoption of a low-dose IM HBIG and lamivudine regimen for prophylaxis against HBV recurrence.19, 20 In 147 patients receiving transplants over 7 years, the rate of HBV recurrence was 4%,21 similar to rates achieved with high-dose IV HBIG-based regimes, regardless of pretransplantation HBV DNA status. This low-dose protocol reduces the direct expenses of HBIG, but it does not avoid the cost and inconvenience associated with the long-term administration of HBIG. An alternative strategy is to simply cease IV HBIG at some point posttransplantation while continuing lamivudine. However, even in patients who have no detectable HBV DNA at transplantation (and are therefore considered at lower risk of recurrence), HBIG cessation at various time points posttransplantation has been found to be associated with a significant risk of recurrence (up to 13% at 5 years).26 Posttransplantation HBV vaccination has been used in some studies with the aim of achieving a durable endogenous surface antibody response, thereby eliminating the need for continued exogenous HBIG. Although very good results have been reported in one study,16 other investigators have been unable to achieve similar success,17, 18 and this approach, although attractive, remains experimental.

In this current study, long-term prophylaxis with low-dose IM HBIG and lamivudine was replaced with adefovir dipivoxil and lamivudine in patients who were at least 12 months posttransplantation without evidence of HBV recurrence or lamivudine resistance. Adefovir was chosen for this study because it has been shown to be generally safe and well tolerated in liver transplant recipients.24, 25 Adefovir has been shown to be effective both in improving hepatitis and indices of liver function in patients with lamivudine resistance pretransplantation and in the setting of posttransplantation hepatitis B recurrence. Furthermore, lamivudine and adefovir represent two different classes of anti-HBV agents (L-nucleoside and acyclic phosphonate) with different mechanisms of action and without cross-resistance, and this combination greatly reduces the risk of the development of subsequent adefovir resistance in nontransplantation patients.22, 23

The combination of adefovir and lamivudine was effective in preventing recurrence following the cessation of HBIG. Importantly, the majority of patients in both arms of the study had detectable HBV DNA prior to lamivudine therapy pretransplantation with a moderate viral load, and a significant minority in both groups still had detectable HBV DNA at time of transplantation (Table 2). Thus, the majority of patients in both treatment arms were at high risk of the development of HBV recurrence in the absence of effective prophylaxis.5, 7 Although the median follow-up in patients who have been switched is 21 months and a majority have been followed for more than 2 years, we cannot exclude the possibility that with more prolonged follow-up, some patients may experience viral breakthrough due to the development of combined ADV/LAM resistance. We believe this is unlikely, however, because none of our patients had pre-existing lamivudine resistance, and resistance to both adefovir and lamivudine is very rare when the two are used in combination.22, 23

It was of considerable interest that 1 patient in the ADV/LAM group developed detectable HBsAg (at a low titer of 0.05 IU/mL) at 5 months but has retained persistently undetectable HBV DNA via highly sensitive PCR assay (lower limit of detection 14 IU/mL) over a further 20 months of follow-up. The likely explanation for this was that HBsAg was being produced at a low level prior to inclusion in the study and that this was unmasked as HBsAb titers fell following the switch to ADV/LAM (Fig. 1). This finding is in keeping with studies that have shown that the majority of patients who have no evidence of HBV reinfection while receiving long-term HBIG have detectable HBV DNA sequences within the liver.27 Importantly, the combination of adefovir and lamivudine has prevented the development of measurable HBV replication in this patient.

We found that the use of combination ADV/LAM for posttransplantation prophylaxis resulted in significant cost savings compared with our standard low-dose HBIG/LAM therapy. The cost of both treatments was very much less than that of continuing high-dose IV HBIG, which is approximately $50,000 (U.S.) per year.28 Furthermore, this regimen is significantly less burdensome to patients, allowing 3 monthly rather than monthly clinic visits and obviating the need for parenteral administration. Although the regime was well tolerated, the patient who developed progressive renal impairment during the study illustrates the need to monitor renal function. It should be noted that this patient had significant renal impairment pretransplantation that was attributed to diabetic nephropathy and also had difficult-to-control hypertension. It is likely that the deterioration in renal function during the course of the study was not primarily due to adefovir use, because renal function has not improved after the drug was ceased.

It is possible that combination ADV/LAM therapy could also provide adequate HBV prophylaxis in patients with less than 12 months of initial posttransplantation HBIG/LAM therapy.29 Indeed, in a recent prospective pilot study performed by our group, primary prophylaxis with combination lamivudine and adefovir from the time of listing prevented HBV recurrence in patients without documented resistance to either drug.30 It is likely that other antiviral combinations could provide similar protection to ADV/LAM, the underlying principle in choosing which combination to use in future studies being to maximize the virological hurdle for resistance while minimizing therapeutic costs and side effects.

In conclusion, in patients who are at least 12 months posttransplantation, the switch to combination ADV/LAM provides effective prophylaxis against recurrence, equivalent to that provided by ongoing HBIG/LAM therapy. Furthermore, combination ADV/LAM therapy is less expensive and is less burdensome to patients. These findings support the use of combination adefovir and lamivudine in preference to long-term HBIG therapy in patients undergoing transplantation for HBV-related liver disease.