Combination of lamivudine and adefovir without hepatitis B immune globulin is safe and effective prophylaxis against hepatitis B virus recurrence in hepatitis B surface antigen–positive liver transplant candidates
Edward J. Gane,
New Zealand Liver Transplant Unit, Auckland Hospital, Auckland, New Zealand
Chronic hepatitis B (CHB) is the leading cause of liver-related mortality in Asia-Pacific and accounts for more than 1 million deaths per annum. Currently, more than 5000 liver transplants are performed annually for CHB in Asia-Pacific. The most important stimulus has been the improvement in the survival outcomes and cost-effectiveness of transplantation for CHB through the adoption of more effective and less expensive antiviral prophylaxis.
Before the availability of antiviral prophylaxis, hepatitis B virus (HBV) recurrence was universal in patients who were HBV DNA–positive before transplantation. The natural history of recurrent HBV was accelerated by immunosuppression, and it progressed rapidly to graft failure and death.[1, 2] Although passive immunoprophylaxis with intravenous (IV) hepatitis B immune globulin (HBIG) reduced the overall recurrence rate by 60%, this strategy was less effective in HBV DNA–positive transplant candidates. A very-high-dose IV HBIG regimen [on-demand dose=10,000 IU, targeting trough antibody to hepatitis B surface antigen (anti-HBs) titers>500 IU/mL] reduced recurrence rates below 20%. However, lifelong maintenance therapy was required to prevent late recurrence. The cost of high-dose IV HBIG has restricted its use in many countries.
The addition of lamivudine (LAM) to HBIG reduces posttransplant recurrence rates to less than 10%. Pretransplant LAM reduces the viral load at transplant and allows the substitution of high-dose IV HBIG with low-dose intramuscular (IM) HBIG. In an Australasian multicenter study of LAM and low-dose IM HBIG in combination, the actuarial risk of HBV recurrence was only 4% after 5 years, and the cost was approximately 10% of the cost of the high-dose IV HBIG regimen.
However, the high cost and inconvenience of lifelong HBIG have driven the development of HBIG-withdrawal strategies. In a randomized controlled study of patients maintained on LAM plus HBIG for at least 1 year after transplantation (including high-risk patients who were HBV DNA–positive at the time of transplantation), switching to LAM plus adefovir dipivoxil (ADV) provided equivalent protection against HBV recurrence but with improved tolerability and reduced cost. In a recent study, the replacement of LAM plus HBIG by tenofovir plus emtricitabine (a nucleoside inhibitor similar to LAM) was well tolerated with no recurrence after 24 months.
However, all these regimens include the short-term administration of HBIG for 6 to 12 months after liver transplantation with its associated costs and tolerability issues. This study is the first to evaluate the safety and efficacy of an HBIG-free regimen: a combination of LAM and ADV was initiated at the time of listing and was continued after transplantation as long-term prophylaxis against HBV recurrence.
PATIENTS AND METHODS
This was a prospective, multicenter, open-label study conducted at liver transplant units in New Zealand (Auckland City Hospital, Auckland, New Zealand), Victoria (Austin Health, Melbourne, Australia), and New South Wales (Royal Prince Alfred Hospital, Sydney, Australia) between August 2003 and March 2007. The appropriate institutional ethics review committees approved the trial before commencement, and all subjects gave written informed consent before their entry into the study. Eligible subjects included all hepatitis B surface antigen (HBsAg)–positive patients assessed and wait-listed for liver transplantation, including patients with acute HBV infections listed for acute liver failure and patients with chronic HBV infections listed for acute-on-chronic liver failure, decompensated cirrhosis, or hepatocellular carcinoma. Patients with hepatitis C virus and hepatitis delta virus coinfections were included, but patients with human immunodeficiency virus infections were not. All patients either were treatment-naive or were maintained on LAM monotherapy at the time of assessment. Patients with current or previously documented genotypic LAM resistance (eg, rtL180M, rtM204I/V, rtV173L, and rt181V/T) were excluded. Patients with moderate or severe renal dysfunction (serum creatinine level≥180 μmol/L) at the time of wait listing were also excluded.
All patients were commenced on a combination of LAM (100 mg daily; GlaxoSmithKline Australia, Boronia, Australia) and ADV (10 mg daily; Gilead Sciences, East Melbourne, Australia) at the baseline, which was defined as the time of assessment and placement on the waiting list for transplantation when patients were recruited into the study. Those patients already established on LAM monotherapy before listing added ADV at the baseline.
The dosage was adjusted to the renal function according to the manufacturer's recommendation. Patients were reviewed (physical examination, physician assessment, and urine pregnancy test for females of childbearing potential) at least monthly before transplantation and at least every 3 months after transplantation. At each study visit, blood was taken for routine hematology and biochemistry, HBsAg, and antibody to HBsAg testing [hepatitis B surface antibody (HBsAb) or anti-HBs]. At pretransplant study visits, blood was additionally tested for HBV DNA by polymerase chain reaction (PCR; Roche Cobas Amplicor HBV monitor assay or sensitive real-time PCR assay) and the prothrombin time [which enabled the calculation of the Model for End-Stage Liver Disease (MELD) score]. At posttransplant study visits, additional serum (2 mL) was stored at −70°C to allow for later reflex (if HBsAg-positive) HBV DNA testing and resistance mutation sequencing. At study completion, real-time HBV PCR assays were performed for all patients, regardless of the HBsAg status or the time after transplantation.
The renal function (serum creatinine, serum phosphate, and estimated glomerular filtration rate) was monitored throughout the study. LAM and ADV doses were adjusted according to the renal dosing algorithms recommended by the manufacturers.
At transplant, in addition to continuing LAM and ADV therapy, patients received 800 IU of IM HBIG (CSL Bioplasma, Parkville, Victoria, Australia) during the anhepatic phase and daily for 7 days after transplantation (a total of 6400 IU of HBIG administered over 7 days) according to our previous protocol. The patients did not receive any additional HBIG, regardless of the HBsAg status or anti-HBs titers.
The primary endpoint of the study was the recurrence of HBV infection after transplantation, which was defined as the reappearance of both HBsAg and HBV DNA in serum. The secondary endpoints were the outcomes of patients before transplantation (the degree and durability of HBV suppression and death or delisting before transplantation), adverse events, and changes in renal function.
Viral serological markers [HBsAg, HBsAb, hepatitis B e antigen (HBeAg), and hepatitis B e antibody] were measured with standard commercial assays. The measurement of serum HBV DNA at the time of screening for entry into the study for all patients was performed with the Roche Cobas Amplicor HBV monitor assay [Roche Molecular Systems, Branchburg, NJ; lower limit of detection (LLOD) ∼ 55 IU/mL]. Since 2007, serum HBV DNA was measured with a highly sensitive real-time PCR assay: the Cobas TaqMan HBV test (Roche Molecular Systems; LLOD=12 IU/mL) in Auckland, the Abbott real-time HBV test (Abbott Molecular, Des Plaines, IL; LLOD=15 IU/mL) in Melbourne, and the Artus HBV LC PCR kit (Qiagen, Hilden, Germany; LLOD=14 IU/mL) in Sydney.
Posttransplant monitoring included monthly testing for liver function tests and serum HBsAg and HBsAb for 3 months and then testing every 3 months for the long term. Testing for serum HBV DNA was performed annually and at the end of follow-up.
Continuous variables are expressed as medians and ranges unless specified otherwise. Categorical data are presented as numbers and percentages. Cumulative patient survival and HBV recurrence–free survival were calculated with the Kaplan-Meier method. The Wilcoxon signed-rank test was used to detect changes in serum creatinine over the course of the study. P values<0.05 were considered statistically significant.
HBV recurrence was defined as the reappearance of both HBsAg and HBV DNA in serum.
Between August 2003 and March 2007, 29 consecutive patients listed for liver transplantation with an HBV infection were screened for this study. Three had LAM resistance and were excluded. The remaining 26 patients were recruited into this study. While they were waiting for liver transplantation, 3 patients were delisted for hepatocellular carcinoma progression beyond transplant criteria (all within 3 months of the baseline), 2 patients were delisted for improvement (the MELD score improved from 23 to 13 for one patient and was unchanged at 8 for the other), and 1 patient died of cryptococcal sepsis 2 months after study entry.
The remaining 20 patients, including 15 in New Zealand, 3 in Victoria, and 2 in New South Wales, underwent successful transplantation. The demographic details of the 20 transplant patients are included in Table 1. The median age was 50 years (range=19–60 years), and 14 (70%) were male. Seven patients (35%) were Asian, 8 (40%) were Polynesian, 3 (15%) were Caucasian, 1 (5%) was Indian, and 1 (5%) was African. More than half of all patients (12/20) were listed for hepatocellular carcinoma, and all met the expanded University of California San Francisco criteria. One patient had a hepatitis delta coinfection, but none had a hepatitis C virus coinfection. The median MELD score unadjusted for hepatoma was 14 (range=6–34) at the baseline (ie, when the patients were placed on the waiting list for liver transplantation) and 17 (range=6–34) at the time of review immediately before transplantation.
Table 1. Baseline Characteristics of Transplant Patients
The virological background of the 20 transplant patients is detailed in Table 2. Before the baseline, 12 patients (60%) were receiving LAM, although the duration of therapy was short in most (median=58 days, range=0–2815 days). All 12 patients had detectable HBV DNA before LAM, and the median level was 4.0 log10 IU/mL (range=2.3–7.5 log10 IU/mL). At the baseline, 9 patients (45%) were HbeAg-positive. Seventeen (85%) had HBV DNA detectable by PCR (LLOD=55 IU/mL), and 10 (50%) had a level greater than 4 log10 IU/mL. The median HBV DNA level at the baseline was 3.9 log10 IU/mL (range=LLOD to >8).
Twenty patients in the open-label study underwent liver transplantation (a living donor for 1 patient and deceased donors for 19 patients). The median waiting time was 114 days (range=1–505 days). Among the 19 patients for whom HBV DNA was measured at transplant, 13 (68%) had HBV DNA detectable by PCR (>55 IU/mL), and 5 (26%) had a level greater than 4 log10 IU/mL. The median HBV DNA level at transplant was 1.9 log10 IU/mL (range=LLOD to 5.0). For the 6 patients who became HBV DNA–negative before transplantation, the mean time to undetectable HBV DNA was 3.9 months (range=0.9–7.8 months). For the remaining patient who did not have an HBV viral load measured at the time of transplantation, the baseline viral load 3 months before transplantation was low (416 IU/mL), and it remained undetectable from 2 months after transplantation.
All transplant patients were alive and continued to be monitored for the study at the time of this writing. The median follow-up after transplantation was 57 months (range=27–83 months).
One patient was diagnosed with recurrent hepatocellular carcinoma 41 months after transplantation when he presented with a subcutaneous lump at the site of a targeted biopsy of the original tumor before transplantation. Biopsy confirmed metastatic hepatocellular carcinoma, which was presumed to be secondary to needle track seeding. Management included local excision and then conversion from tacrolimus to sirolimus. This patient remained well with no evidence of intrahepatic or extrahepatic recurrence 36 months later.
After the completion of the open-label study, LAM/ADV was adopted as the standard prophylaxis for all HBsAg-positive candidates from the time of wait listing. The perioperative regimen was modified to be completely HBIG-free in patients who had documented suppression of serum HBV DNA below 3 log10 IU/mL before transplantation. All patients who had either HBV DNA levels greater than 3 log10 IU/mL or unknown HBV DNA levels (including patients urgently listed with acute liver failure from either an acute HBV infection or an acute-on-chronic HBV infection) received daily IM HBIG (800 IU) for the first 7 postoperative days. Between September 2007 and September 2011, another 39 HBsAg-positive patients were wait-listed, and 28 underwent transplantation (all deceased donors); 18 did not receive HBIG perioperatively. All 28 patients were alive with normal graft function after a median posttransplant follow-up of 22 months (range=10–58 months).
There were no cases of HBV recurrence (defined as both detectable HBsAg and HBV DNA in serum) in the 20 HBsAg-positive patients undergoing transplantation in the open-label study or in the subsequent 28 HBsAg-positive patients undergoing transplantation with the HBIG-free regimen.
Serum HBsAg was assayed at each posttransplant visit. Forty-seven of the 48 patients had lost HBsAg within 8 weeks after transplantation. In the remaining patient, HBsAg first became undetectable 6 months after transplantation (Fig. 1). The median time to HBsAg nondetectability was 7 days. HBsAg became detectable again in a single patient 41 months after transplantation. This reappearance of HBsAg coincided with the diagnosis of recurrent hepatocellular carcinoma (as discussed previously). The serum HBsAg titer was very low, and serum HBV DNA and HBeAg remained negative throughout (he was HBeAg-positive before transplantation). Serum HBsAg became undetectable again 1 month after the excision of this metastasis, and it remained undetectable 24 months later. No other patients reverted to HBsAg positivity.
For all patients with a quantifiable HBV viral load at the time of transplantation, this became undetectable with a real-time PCR assay during follow-up; this occurred at a median of 6 weeks (range=3–24 weeks) after transplantation. A single patient had a detectable but low level of serum HBV DNA 24 weeks after transplantation (16 IU/mL), but this was undetectable when the test was repeated at 9 and 12 months. HBsAg became undetectable 2 weeks after transplantation and never reappeared.
On posttransplant day 7 [after 7 daily doses (800 IU) of IM HBIG], serum anti-HBs was measurable in 17 patients (35 to >1000 IU/L), but it remained detectable in only 1 patient more than 3 months after transplantation.
At the time of the preparation of this article, all patients were tested for HBsAg and HBV DNA (measured with a real-time HBV PCR assay; LLOD=15 IU/mL). All remained seronegative for both HBsAg and HBV DNA.
Safety and Renal Function
In the open-label study, renal function was stable before transplantation. The median serum creatinine level was 81 μmol/L (range=55–107 μmol/L) at the baseline (when the combination of LAM and ADV was commenced) and 82 μmol/L (range=53–100 μmol/L) at the time of transplantation. Renal function deteriorated after transplantation. At the last follow-up (median=75 months after transplantation), the median serum creatinine level was 119 μmol/L (range=88–170 μmol/L). Serum phosphate levels remained normal in all patients: the median level at the last follow-up was 1.0 mmol/L (range=0.7–1.3 mmol/L). Before transplantation, no patients required alterations of ADV or LAM doses, but after transplantation, 3 patients required a reduction to alternate-day dosing because of decreases in the estimated glomerular filtration rate below 50 mL/minute. There were no patients for whom ADV or LAM therapy had to be stopped.
No serious adverse events were reported that were considered to be related to the study therapy.
This current study demonstrates that the combination of ADV and LAM is highly effective in preventing post–liver transplant HBV recurrence and obviates the need for ongoing prolonged HBIG therapy. After a median follow-up of almost 5 years, no patient developed recurrence. Importantly, more than half of the patients had detectable HBV DNA at the time of transplantation and were at high risk for HBV recurrence. In some patients whose indication for transplantation was liver failure, treatment with this combination resulted in clinical improvements that enabled them to be removed from the waiting list.
At the time of the design of the open-label study, ADV was chosen as the second antiviral because it had been shown to be generally safe and well tolerated in liver transplant recipients.[11, 12] We have previously shown that ADV is safe and effective when it is substituted for HBIG more than 12 months after transplantation in patients receiving long-term HBIG plus LAM. All patients established on LAM at the time of their referral for a transplant evaluation were tested for LAM resistance because of the concern that the LAM/ADV combination might not provide sufficient suppression of viral replication before transplantation.
The limitations of this study include the relatively small number of patients and the lack of a randomized control group. However, the standardized approaches across the 3 study centers and the length of follow-up are particular strengths. The median posttransplant follow-up of the 20 patients enrolled in the open-label study was almost 5 years. Another 28 patients underwent transplantation with LAM/ADV with a median follow-up of almost 2 years. Although it is possible that longer follow-up may identify virological breakthrough from the emergence of dual ADV and LAM resistance, we believe that this is unlikely. Not only does the LAM/ADV combination reduce the emergence of LAM resistance (15% versus 43%) in treatment-naive patients, but no cases of combination genotypic resistance to both ADV and LAM have occurred with 5 years of continuous therapy.
The observations from 2 previous studies of LAM/ADV therapy for post–liver transplant HBV prophylaxis support the results of our current prospective study. In a retrospective study of 16 HBsAg-positive transplant candidates at Queen Mary Hospital (Hong Kong), ADV was added to LAM before transplantation at the onset of LAM resistance. Eight of these patients later underwent transplantation without HBIG. At a median follow-up of 11 months, all remained HBV DNA–negative, although 2 remained persistently HBsAg-positive. In a larger international study of ADV in liver transplant candidates with LAM resistance, 57 patients were commenced on ADV before transplantation; 34 received long-term posttransplant prophylaxis with ADV plus HBIG, whereas 23 did not receive HBIG and were maintained on ADV with or without LAM. After a median follow-up of 36 months, the rates of HBsAg were similar in the 2 groups (12% with HBIG versus 13% without HBIG), and none became HBV DNA–positive. In this study, the efficacy of oral antiviral prophylaxis was not influenced by the addition of HBIG.
It is likely that other potent antiviral agents or antiviral combinations could also provide effective protection without the use of HBIG. In comparison with LAM or ADV, entecavir is a more potent antiviral agent with a very high genetic barrier to resistance. This agent has been used since 2007 as pretransplant and posttransplant monotherapy to prevent recurrent HBV in Hong Kong. That group recently reported the long-term results of 75 patients undergoing transplantation on entecavir. The major difference between that study and our own is that only half of the Hong Kong patients were started on antiviral therapy before transplantation, whereas all patients were in our series. As a result, the median HBV DNA level was much higher (3.6 log10 IU/mL in the Hong Kong series versus only 1.9 log10 IU/mL in our series). The rate of HBsAg clearance after transplantation was much slower in the Hong Kong series, with a median time to HBsAg loss of 4 weeks versus 7 days in this study and with a 12-month cumulative rate of HBsAg loss of 88% versus 100% in this study.
The other possible alternative to LAM/ADV as an HBIG-free strategy is tenofovir disoproxil fumarate (TDF). Like ADV, TDF is an acyclic nucleoside phosphonate analogue that is active against both wild-type and LAM-resistant HBV. In patients with compensated CHB disease, TDF showed superior antiviral efficacy in comparison with ADV. Long-term TDF use is associated with excellent efficacy and tolerability, with no resistance seen after 5 years. In the future, LAM/ADV is likely to be replaced by LAM/TDF or Truvada, the commercially available combination of tenofovir and emtricitabine (Gilead Pharmaceuticals, Foster City, CA). LAM/TDF or Truvada would also provide effective HBIG-free antiviral prophylaxis for patients with established LAM resistance before transplantation.
The combination of LAM and ADV is cost-effective posttransplant prophylaxis in comparison with low-dose IM HBIG/LAM ($8290 versus $13,718 per year) and high-dose IV HBIG/LAM, which is approximately $50,000 per year (US dollars). Furthermore, this regimen is significantly less burdensome to patients, allows clinic visits every 3 months rather than monthly clinic visits, and obviates the requirement for parenteral therapy.
Although the regimen was well tolerated, ADV can lead to the deterioration of renal function, and this must be monitored. However, the change in renal function observed over the course of this study (a median increase in the serum creatinine level of 37 μmol/L over a median of 74 months after transplantation) was not greater than might be expected in patients receiving calcineurin inhibitor therapy after liver transplantation, and this suggests that to date any possible additive renal toxicity from ADV has been minor.
In conclusion, the combination of LAM and ADV initiated at the time of wait listing and continued for the long term after transplantation provides safety and efficacy similar to those of current prophylaxis regimens with LAM plus HBIG but without the cost and inconvenience associated with monthly HBIG administration. In the future, either combination therapy with 2 oral nucleos(t)ides without cross-resistance (LAM/TDF or Truvada) or monotherapy with a potent nucleos(t)ide with a high genetic barrier (entecavir or TDF) should confine HBIG administration to perioperative use in those few patients with high viral loads at the time of transplantation. HBIG-free prophylaxis will dramatically reduce the costs of managing HBsAg-positive liver transplant recipients.