Beneficial Effects of Short-Term Lamivudine Treatment for de novo Hepatitis B Virus Reactivation After Liver Transplantation


*Corresponding author: Tsutomu Chiba,


Clearance of hepatitis B surface antigen (HBsAg) by lamivudine is achieved in only a small proportion of patients with chronic hepatitis B virus (HBV) infection. We investigated the effect of lamivudine on de novo HBV reactivation after living-donor liver transplantation when the number of HBV was expected to be very small. Thirty-eight HBV-naive recipients who received liver grafts from antibodies to core antigen-positive donors receiving hepatitis B immunoglobulin (HBIG) were studied. HBsAg appeared in nine cases (23.7 %) despite receiving HBIG for 12–71 months (mean: 35.1 months) after transplantation. Lamivudine treatment was started in six recipients during the acute phase of HBV reactivation. Five of the six recipients achieved complete clearance of HBsAg in sera at a median of 4.6 months (ranging from 21 to 330 days) after lamivudine administration. Although lamivudine was stopped in four cases, all remained negative for HBsAg. Our findings suggested that short-term lamivudine treatment during acute phase of HBV reactivation could achieve complete clearance of HBsAg in a significant number of liver transplant recipients.


There is growing recognition that the majority of healthy individuals who are negative for hepatitis B surface antigen (HBsAg) but positive for antibodies to core antigen (anti-HBc), who had once been assumed to denote previous exposure to hepatitis B virus (HBV), have persistent viral infection of the liver tissues (1,2). Recently, we demonstrated that latent HBV infection is accompanied by ongoing viral replication in the livers, but not in the sera, of healthy anti-HBc-positive liver transplant donors (3). In support of the concept of occult HBV infection, HBV was transmitted from anti-HBc-positive donors to HBV-naive recipients at a high frequency via liver grafts in living-donor liver transplantation (LDLT) and orthotopic liver transplantation (4–8). Because of the persistent shortage of organs and the increasing number of patients awaiting transplantation, the use of liver grafts from anti-HBc-positive donors cannot be avoided, especially in areas where the prevalence of HBV is high (4). Thus, adequate prophylaxis against HBV is required to prevent viral reactivation in HBV-naive recipients after liver transplantation.

To date, strategies to prevent viral breakthroughs in the recipients of anti-HBc-positive livers have been empirical, and hepatitis B immunoglobulin (HBIG) has been widely used as the standard prophylaxis after liver transplantation (9). Several reports, including ours, suggested that HBIG prophylaxis was effective for preventing HBV exacerbation in recipients who received hepatic allografts from anti-HBc-positive donors (4,10). However, long-term passive immunization with HBIG is associated with problems, such as high cost, limited availability, and selection of viral strains containing mutations in the surface gene of HBV-DNA (9). Moreover, the difficulty of maintaining serum antibodies to HBsAg (anti-HBs) titer and the poor compliance of HBIG could result in viral reactivation after liver transplantation (7).

The aim of this study was to evaluate the efficacy of short-term administration of lamivudine for the treatment of de novo HBV exacerbation in transplant recipients with anti-HBc-positive donors.

Patients and Methods


From July 1995 to July 2004, 902 patients underwent LDLT at Kyoto University Hospital. Before operations, serological evaluation for HBV markers, including HBsAg, anti-HBs, anti-HBc, hepatitis B e antigen (HBeAg), and antibodies to HBeAg (anti-HBe), was carried out using commercial enzyme immunoassay kits (Dainabot, Tokyo, Japan). HBV-DNA was analyzed using a commercial polymerase chain reaction (PCR) assay (Amplicor HBV Monitor, Roche, Branchburg, NJ, USA). Among 902 donors, 121 (13.4%) were positive for anti-HBc in the absence of HBsAg. Of these, all recipients fulfilling both of the following criteria were included: (a) none of the HBV-related serological markers positive before transplantation; (b) post-operative survival and well-tolerated prophylaxis with HBIG for longer than 6 months after LDLT. Accordingly, 55 recipients were excluded because they were positive for HBV-serological markers before LDLT. Twenty-six patients were also excluded from the study because of their short duration of survival, and all of these patients died from causes not related to the HBV reactivation. Two patients refused to receive HBIG for financial reasons, and were lost to follow-up. A total of 38 recipients were considered eligible for this study. These HBV-naive recipients with anti-HBc-positive donors underwent LDLT for the following liver diseases: biliary atresia (n = 21), hepatitis C virus (HCV)-related chronic liver disease (n = 3), primary biliary cirrhosis (n = 3), primary sclerosing cholangitis (n = 2), chronic rejection (n = 2), post-LDLT graft failure (n = 2) and others (n = 5). The male/female ratio was 1/1 and the age range was 0–58 years (mean age: 15.9 years, age ≤18 years: n = 28, age >18 years: n = 10). HBV reactivation after LDLT was diagnosed by confirming the appearance of HBsAg in sera of the recipients. Liver tissue and serum samples of all anti-HBc-positive donors were obtained at the time of operation and subjected to analysis for HBV-DNA. All subjects provided written informed consent, and the study was conducted in accordance with the principles of the Declaration of Helsinki.

Prophylaxis with HBIG and immunosuppressive protocol

HBIG monotherapy was given to all the recipients with grafts from anti-HBc donors, as reported previously (4). The first dose of HBIG at 200 IU/kg body mass was administered during the anhepatic phase of LDLT, and the same dose was given every day during the first 6 post-operative days. Subsequently, HBV-serological markers were examined at monthly intervals after the transplant operation and 1000 IU of HBIG was periodically administered to maintain serum anti-HBs titers at more than 200 IU/L throughout the follow-up period.

Immunosuppressive therapy for all recipients consisted of tacrolimus and low-dose steroids. Target trough levels of tacrolimus in whole blood were 10 to 15 ng/mL in the first week, and then 5 to 10 ng/mL during the first month after transplantation. Methylprednisolone (10 mg/kg) was administered intravenously (IV) during the anhepatic phase of surgery, followed by 2 mg/kg administered IV for the first 3 days, then tapered to 1 mg/kg for 3 days and converted to 0.3 mg/kg/day of prednisone, which was decreased gradually and discontinued between 3 and 6 months.

PCR amplification of HBV-DNA and sequencing of the surface gene

Preparation of DNA samples and detection of HBV genomes by nested PCR have been described previously (3). The nucleotide sequence spanning the S region was amplified by PCR using specific primer sets, followed by subcloning of PCR products using a pGEM-T Easy Vector System I (Promega, Madison, WI, USA). A total of 15 clones derived from each serum specimen were subjected to sequencing analyses (3).


HBV reactivation despite HBIG prophylaxis

Post-operative HBIG prophylaxis was given to 38 HBV-naive recipients with grafts from anti-HBc-positive donors. Among them, 29 showed no evidence of HBV recurrence during the follow-up period (mean: 41.1 months, range: 10 months to 9.5 years). Unfortunately, in 9 of 38 cases (23.7 %), anti-HBs titer decreased concurrently with the appearance of HBsAg in the serum despite HBIG prophylaxis after LDLT. Table 1 shows the serological characteristics of the donors and pre-transplant status of the recipients who suffered from HBV reactivation despite HBIG administration. Baseline characteristics including age, gender and HBV-related serology were similar between these nine recipients and the remaining recipients without HBV recurrence (data not shown). Consistent with our previous analyses, 31 of the 38 donors with anti-HBc (81.6%) were positive for HBV-DNA in the liver specimens, indicating a high frequency of latent HBV infection in the livers of anti-HBc-positive individuals (3). In contrast, HBV-DNA was negative in the sera of all anti-HBc-positive donors. All allografts of nine recipients with HBsAg appearance were positive for HBV-DNA by PCR analyses, suggesting that the de novo HBV reactivation originated from the liver graft with latent HBV infection after LDLT.

Table 1.  HBV-serological status of recipients with HBV reactivation post-LDLT and of their corresponding donors with anti-HBc
Case #Age/SexIndication for LDLTHBsAg/ Anti-HBsAnti-HBc Anti-HBsHBsAg/ Anti-HBsHBeAg/ Anti-HBeHBV-DNA in liver graft
  1. LDLT = living-donor liver transplantation; HBsAg = hepatitis B surface antigen; anti-HBs = antibody to HBsAg; anti-HBc = antibody to hepatitis B core antigen; HBeAg = hepatitis B e antigen; anti-HBe = antibody to HBeAg; BA = biliary atresia; PSC = primary sclerosing cholangitis.


To define the factors associated with HBV reactivation in these nine recipients, variables related to the donors and recipients, transplant procedures, and HBIG prophylaxis were analyzed. The indications for LDLT in these recipients with HBV reactivation were biliary atresia (n = 7), Wilson's disease (n = 1) and primary sclerosing cholangitis (n = 1). The mean period between LDLT and de novo HBsAg appearance was 35.1 months (range: 12–71 months; Table 2). A liver biopsy was performed on five of the nine patients at the time of the de novo HBV recurrence, and all exhibited evidence of chronic active hepatitis accompanied by mild inflammatory activity and mild fibrosis. HBV appearance was attributed to the decrease in serum anti-HBs titer despite HBIG prophylaxis in four of nine recipients. Among them, three recipients (cases #1, #2 and #3) were considered to have suffered from HBV recurrence because of non-compliance of HBIG. Although post-operative HBIG prophylaxis was given to these three patients, they had a transient cessation of HBIG treatment for personal reasons 12, 7 and 11 months after LDLT. They experienced a decrease in anti-HBs titer and, consequently, HBsAg became detectable in the sera after cessation of HBIG treatment. The mean period between the cessation of HBIG treatment and the emergence of HBsAg in their sera was 12.3 months (range, 3–26 months). An immunosuppressive condition was presumed to be associated with viral activation, with consequent decreases of anti-HBs in one case (case #4). Recipient #4 showed an anti-HBs titer of less than 23.8 IU/L 2 weeks after the 2000 IU-HBIG infusion, followed by the appearance of HBsAg. Continuous medication with prednisolone for the treatment of chronic rejection suggested the underlying possible immunocompromised condition in this case.

Table 2.  Clinical features of recipients with HBV reactivation
Case #Anti-HBs titer1 (mIU/mL)Duration until HBV reactivation2 (months)Clinical features at the time of HBV reactivationPossible reasons for HBV reactivation
ALT (IU/L)Histology
  1. anti-HBs = antibody to HBsAg; ALT = alanine aminotransferase; N.D. = not determined; CAH = chronic active hepatitis; noncompliance = noncompliance of HBIG; escape mutant = emergence of surface escape mutant.

  2. 1Anti-HBs titer before HBsAg appearance.

  3. 2Period between liver transplantation and HBsAg appearance.

5140.61524N.D.Escape mutant
611730153CAHEscape mutant

In the remaining five cases (cases #5, #6, #7, #8 and #9), HBsAg eventually became positive despite the continuous treatment with periodical HBIG prophylaxis. HBV clones comprising mutations in the S gene have been reported in OLT recipients who developed recurrent hepatitis B despite HBIG prophylaxis (11). To ascertain whether the HBIG failure in these cases was associated with changes in antigenicity of the S protein, the S gene sequence of HBV-DNA was determined in the HBV strain of two patients whose sera at the acute phase of HBV exacerbation were available for further analyses (cases #5 and #6). The sequence analyses of the two cases revealed that the detected HBV clone contained several mutations, including G- to -A substitutions at nucleotide 586 (subtype adr) and 587 (subtype adw) within the ‘a’ determinant region on the HBsAg-encoding gene, suggesting that the cloned HBV variants might be responsible for HBV recurrence despite HBIG administration in these two cases. Unfortunately, we could not determine the factors that were related to HBV reactivation in cases #7, #8 and #9.

Short-term lamivudine treatment

Among recipients with de novo HBV recurrence, six (cases #1, #2, #4, #6, #7 and #9) of nine cases had elevated levels of serum alanine aminotransferase, suggesting that recurrent active hepatitis was present. Moreover, of these six cases, all four cases that were examined exhibited histological evidence of inflammation with lymphocytes infiltration around the portal area at the time of de novo HBV reactivation (Table 2). Lamivudine (100 mg) was given in six cases to suppress the viral activity (Figure 1). Of them, five patients started the treatment immediately after HBsAg appearance (average: 27 days; range: 1 day to 2 months). One patient (case #1) did not take lamivudine at the time of HBsAg appearance because of personal reasons, but finally received lamivudine therapy 23 months after HBV reactivation.

Figure 1.

Clinical course of the six liver transplant recipients with HBV reactivation who received lamivudine treatment. The bold line represents HBsAg and the fine line represents anti-HBs. The treatment with HBIG is shown as arrows and treatment with lamivudine as shaded boxes. HBIG = hepatitis B immunoglobulin; Lam = lamivudine.

After the administration of lamivudine, HBsAg decreased in the sera of five out of the six recipients, and in all these five cases (cases #1, #2, #4, #7 and #9), HBsAg disappeared from the sera at a median of 4.6 months (range: 21–330 days) after the beginning of lamivudine treatment. Suppression of HBsAg by lamivudine treatment was invariably associated with a decline in serum transaminase levels in these five cases. After confirming the stable seroconversion to anti-HBs-positive status, lamivudine treatment was stopped in four of the five recipients after 60 months, 10 months, 1.5 months and 4 months (cases #1, #2, #7 and #9). The remaining individual (case #4) is currently receiving lamivudine treatment because she started treatment only 4 months ago; her most recent blood test was negative for HBsAg. The presence of circulating exogenous anti-HBs derived from HBIG complicates detection of the endogenous anti-HBs, which reflect the development of anti-HBV immunity in recipients. Thus, we also gradually withdrew the HBIG prophylaxis after the administration of lamivudine, and confirmed the sustained positivity of anti-HBs without HBIG treatment in two recipients (cases #2 and #7). Since then, they consistently showed evidence of immunity against HBV with endogenous anti-HBs titers greater than 100 IU/L without any prophylaxis. Consequently, HBIG was not given to these two recipients, even after termination of lamivudine treatment. Serum HBsAg of these two recipients remained negative during the follow-up periods of 14 and 35 months after the withdrawal of HBIG and lamivudine. Recipient #7 died of progressive deterioration of liver function 35 months after the complete loss of HBsAg in sera, and the histological autopsy findings were consistent with chronic rejection in the liver graft. The remaining recipient (case #6) is currently being treated with lamivudine monotherapy. None of the recipients developed tyrosine-methionine-aspartate-aspartate (YMDD) mutants during the course of lamivudine therapy (data not shown).

Three patients (recipients #3, #5 and #8) did not receive lamivudine treatment after HBsAg appearance because the alanine aminotransferase levels of these patients showed no evidence of active hepatitis.


In this study, we demonstrated that short-term lamivudine therapy for LDLT recipients with de novo HBV reactivation could achieve a significant HBsAg seroconversion rate without inducing drug-resistant YMDD mutants. The important point to note is that after the transient use of lamivudine, some of the recipients who had been naive against HBV infection before transplantation maintained anti-HBs positivity despite withdrawal of HBIG even during immunosuppressive therapy, indicating that they were likely to acquire endogenous immunity to HBV infection.

Passive immunoprophylaxis with HBIG has been used to prevent de novo HBV recurrence after liver transplantation in HBV-naive recipients who received liver grafts from anti-HBc-positive donors (8). However, not only the efficacy, but also the safety of long-term HBIG treatment after liver transplantation has not been determined, including the risk of development of hepatitis B ‘surface’ escape mutants during passive immunoprophylaxis. In this study, we demonstrated that a considerable number of patients developed de novo HBV reactivation during the long-term course of HBIG prophylaxis. Our findings suggested that factors affecting the recurrence of de novo HBV in those recipients included non-compliance of HBIG, an immunosuppressive condition, and possible emergence of ‘surface’ escape mutants.

Lamivudine is a potent inhibitor of HBV replication, and several investigators have reported that lamivudine therapy is effective for hepatitis B treatment following liver transplantation (12,13). The present data show that early initiation of lamivudine treatment for post-transplant HBV de novo reactivation efficiently induced a sustained loss of HBsAg, suggesting that the effect of lamivudine on de novo HBV reactivation is stronger than its effect on chronic HBV infection. In general, HBV replication decreases immediately after administration of lamivudine; however, it has been shown that lamivudine treatment achieves clearance of HBsAg in only a very small proportion of cases. Indeed, loss of HBsAg occurred in only 0–2% of patients with chronic HBV infection that were treated with lamivudine (14). In transplant recipients with HBV-associated liver disease, lamivudine treatment induced the disappearance of HBsAg in only 3 of 52 cases (6%) by week 52, 2 of whom acquired anti-HBs (12). The better results observed in this study can be attributed to the fact that lamivudine treatment was commenced during the acute phase of de novo HBV reactivation when the number of HBV was expected to be very small. The beneficial effect of early lamivudine therapy for acute hepatitis has been reported previously (15,16). We suggest that the timing of the lamivudine administration in patients with HBV activation, specifically in the acute phase of HBV reactivation, is important to achieve complete viral suppression and successful seroconversion from HBsAg to anti-HBs.

It is well recognized that prolonged administration of lamivudine can lead to viral breakthrough because of the emergence of viral variants with reduced sensitivity to the drug resulting from one or more mutations in the YMDD locus of the HBV polymerase gene (17). Many investigators have reported a high rate of virological breakthrough with prolonged lamivudine therapy in liver transplant recipients with active HBV infection (17,18). In contrast, we confirmed the beneficial effect of lamivudine on de novo HBV reactivation and showed that it does not cause the emergence of YMDD mutants. Reduced risk for virological breakthrough during lamivudine therapy may be attributed to the short-term use of the drug and the low levels of HBV-DNA at the acute phase of de novo HBV reactivation, as prolonged use of lamivudine and high HBV-DNA levels before treatment were shown to be associated with the emergence of the drug-resistant mutants (19). Prevention of drug-resistant viral clones by transient use of lamivudine has also been demonstrated in patients with chronic HBV infection (20). Although short-term lamivudine therapy could reduce the incidence of YMDD mutants, the relapse rate was high after the withdrawal of lamivudine in HBV-related chronic liver disease patients (20). Thus, short-term lamivudine treatment might be applicable specifically to the acute phase of de novo HBV reactivation.

In conclusion, we have shown in this study that short-term use of lamivudine resulted in complete clearance of HBsAg in the majority of patients with de novo HBV reactivation, and that the effects of lamivudine were stronger in patients with de novo HBV reactivation than in patients with HBV-related chronic liver disease. More importantly, sustained clearance of HBV was obtained in two cases who stopped both HBIG and lamivudine administration, suggesting acquired immunity against HBV; however, a study of a large number of patients with de novo HBV reactivation is needed to reach any definitive conclusion. Thus, further studies with greater patient numbers are required to determine whether short-term lamivudine administration induces clearance of HBsAg in the sera, as well as acquired immunity against HBV in HBV-naive recipients receiving anti-HBc-positive allografts.


This study was supported by a grant-in-aid for scientific research (15209024 and 16790378) from the Japan Society for the Promotion of Science (JSPS).