|
| |||||||||||||||||||||||||||||||||||||||
Background The role of liver transplantation in patients positive for the hepatitis B surface antigen (HBsAg) is controversial because of the high rate of recurrent hepatitis B virus (HBV) infection. It has not been determined whether this risk is greater for certain patients and whether the administration of anti-hepatitis B surface antigen (anti-HBs) immune globulin is beneficial.
Methods We conducted a retrospective study at 17 European centers of 372 consecutive HBsAg-positive patients who underwent liver transplantation between 1977 and 1990. Recurrence of HBV infection was defined as the reappearance of HBsAg in serum.
Results For all 334 patients with follow-up data, the mean (±SE) three-year actuarial risk of recurrence of HBV was 50 ±3 percent. The risk was 67 ±4 percent among 163 patients with HBV-related cirrhosis, 32 ±5 percent among 110 patients with cirrhosis related to hepatitis delta virus, 40 ±16 percent among 14 patients with fulminant hepatitis delta infection, and 17 ±7 percent among 39 patients with fulminant HBV infection (P<0.001). Among the patients with HBV-related cirrhosis, the risk of HBV recurrence was greatest (83 ±6 percent) in those who were seropositive for HBV DNA at the time of transplantation and lowest (58 ±7 percent) in those with neither HBV DNA nor hepatitis B e antigen detectable in serum. With respect to the use of passive prophylaxis with anti-HBs immune globulin, the risk of HBV recurrence was 75 ±6 percent among the 67 patients given no immunoprophylaxis, 74 ±5 percent among the 83 treated for two months, and 36 ±4 percent among the 209 treated for six months or longer (P<0.001). In a multivariate analysis the predictors of a lower risk of HBV recurrence were the long-term administration of the immune globulin, hepatitis delta virus superinfection, and acute liver disease. For the entire study cohort, survival was 75 percent at one year and 63 percent at three years, but for those in whom HBV infection recurred, survival was 68 percent at one year and 44 percent at three years.
Conclusions In this retrospective study of HBsAg-positive patients, liver transplantation had better results in those who had fulminant hepatitis or delta virus superinfection. An absence of viral replication at the time of transplantation and long-term immunoprophylaxis were associated with a reduced risk of recurrent HBV infection and reduced mortality.
As part of the European Concerted Action on Viral Hepatitis (EUROHEP) project,23 we evaluated the results of liver transplantation in HBsAg-positive patients in European centers to determine the factors that predict HBV recurrence and survival after liver transplantation and to assess the effects of passive anti-HBs immunoprophylaxis on recurrence and survival.
Methods
Patients
A questionnaire was sent in March 1990 to the 69 liver-transplant centers that were members of the European Liver Transplant Registry,1 asking for data on all HBsAg-positive patients who had undergone liver transplantation as of that date. The following data were requested for each patient: age and sex; the type of liver disease diagnosed; status for HBV and hepatitis delta virus (HDV) before transplantation (indicated by the presence of HBV DNA detected by standard molecular hybridization technique and the presence in serum of HBsAg, HBeAg, antibody to HBeAg, HDV antigen, or antibody to HDV); methods used to prevent HBV reinfection before, during, and after transplantation (details on the use of polyclonal or monoclonal anti-HBs immune globulin, active anti-HBs immunization, and other antiviral therapy); date of reappearance of HBsAg and HBV DNA; current status for HBV; current status of the graft; duration of follow-up; dates of and reasons for repeated transplantation; and date and cause of death. Eighteen centers replied to the questionnaire. All HBsAg-positive patients undergoing transplantation at each center from the initiation of a program of liver transplantation at the center to March 1990 were described by the centers and included in this study. A second questionnaire was sent in June 1991 to 17 of the centers (the center in Helsinki, Finland, reported that no HBsAg-positive patients had undergone transplantation) to update to June 1991 follow-up data on the patients already included.
Hbv recurrence was defined as the reappearance of HBsAg in serum. With respect to the administration of anti-HBs immune globulin, patients were classified into three groups: those who received no immune globulin at all (no immunoprophylaxis), those who received it for less than six months after liver transplantation (short-term therapy), and those who received it for at least six months after transplantation (long-term therapy). In each group, some patients died before completing the course of immunoprophylaxis. These patients were included in the analyses according to the intention-to-treat principle.
Statistical Analysis
Statistical analyses were performed with a BMDP statistical software package24. Means were compared by analysis of variance. All patients were followed for the recurrence of HBV from the date of transplantation to death, documentation of positivity for HBsAg, or the last disease-free visit. Follow-up for survival analysis began on the date of liver transplantation and continued until death or the last visit. The status of all patients was updated by each center in June 1991. Patients who received a second transplant, whether or not they had had a recurrence of HBV, were included in the analysis until their death or last visit. Survival and the recurrence rate for HBV infection were calculated with actuarial life-table analyses. Survival curves were compared by the log-rank test. Multivariate analyses were performed with the Cox model (with the Wald test for assessing the effect of the entered variables on survival and recurrence rates for HBsAg and for calculating the relative risks of death and HBV recurrence). The following variables were entered in the multivariate analyses: anti-HBs immune globulin treatment (long-term therapy vs. short-term therapy vs. no immunoprophylaxis), year of liver transplantation (before 1986, 1986, 1987, 1988, and 1989), course of liver disease (acute vs. chronic disease; acute disease included fulminant HBV and fulminant HDV, and chronic disease included HBV-related and HDV-related cirrhosis), serologic status for HBV DNA before liver transplantation (presence vs. absence), serologic status for HBeAg before transplantation (presence vs. absence), and superinfection with hepatitis (HDV vs. non-HDV; HDV superinfection included HDV-related cirrhosis and fulminant HDV; non-HDV superinfection included HBV-related cirrhosis and fulminant HBV).
Results
Liver transplantation was performed in 372 consecutive HBsAg-positive patients between 1977 and March 1990 at 17 centers. The patients' diagnoses are shown in Table 1. Among the 201 patients with HBV-related cirrhosis, the pretransplantation status for HBV DNA was known in 163 (58 were seropositive and 105 seronegative). The pretransplantation status for HBeAg was known in 191 patients (73 were seropositive and 118 seronegative). The mean (±SE) length of follow-up in the 372 patients was 24.8 ±19 months.
|
|
|
Survival of all patients at one year and three years was 75 ±2 percent and 63 ±3 percent, respectively. The actuarial one-year survival of patients undergoing transplantation before 1986, in 1986, in 1987, in 1988, and in 1989 was 39 ±8 percent, 57 ±8 percent, 78 ±5 percent, 77 ±4 percent, and 83 ±3 percent, respectively (P<0.001). The actuarial survival of patients in relation to their initial liver disease (fulminant HBV, fulminant HDV, HBV-related cirrhosis, or HDV-related cirrhosis) is shown in Figure 3; the differences among these four groups were significant (P<0.001). The actuarial survival of patients receiving no immunoprophylaxis, short-term therapy, and long-term therapy is shown in Figure 4. Actuarial survival at one year and three years was 73 percent and 54 percent, respectively, among patients with HBV recurrence and 90 percent and 83 percent, respectively, among those who remained HBsAg-negative. The survival curves differed significantly (P<0.001). Among patients with recurrence of HBV, actuarial survival at one year and three years was 68 percent and 44 percent, respectively, among those undergoing transplantation for HBV-related cirrhosis and 86 percent and 83 percent, respectively, among those undergoing transplantation for HDV-related cirrhosis (P<0.001). In a univariate analysis, the presence of HDV superinfection (fulminant HDV or HDV-related cirrhosis) (P<0.001), long-term administration of anti-HBs immune globulin (P<0.001), the absence of HBV DNA before liver transplantation (P<0.001), and the absence of HBeAg before transplantation (P = 0.05) were predictive of better survival. In a multivariate analysis, long-term administration of anti-HBs immune globulin (relative risk, 2.22; 95 percent confidence interval, 1.13 to 4.33; P<0.001) and HDV superinfection (relative risk, 6.25; 95 percent confidence interval, 3.13 to 12.42; P<0.001) were independently predictive of better survival among all patients studied. Long-term administration of immune globulin (relative risk, 2.41; 95 percent confidence interval, 1.06 to 5.48; P<0.05) was independently predictive of better survival among patients with HBV-related cirrhosis.
|
|
Forty of the 372 patients (10.8 percent) died as a direct result of recurrence of HBV, with a mean survival of 12.0 ±7.2 months (range, 0.5 to 30). Thirty-one (15.4 percent) of the 201 patients with HBV-related cirrhosis died of recurrence of HBV, as compared with 3 (2.7 percent) of the 110 with HDV-related cirrhosis, 2 (14.3 percent) of the 14 with fulminant hepatitis HDV, and none of those with fulminant HBV.
Discussion
Liver transplantation in HBsAg-positive patients remains controversial2,7,11. Some centers consider the presence of HBsAg a contraindication to transplantation, and most centers are reluctant to perform the procedure if HBV replication is ongoing. Our study offers an opportunity to analyze the results of liver transplantation in the largest series of HBsAg-positive patients ever to our knowledge described. The inclusion of 17 European centers permitted us to compare outcomes in relation to initial liver disease and to compare the results of the use of anti-HBs passive immunoprophylaxis on the rate of infection of the graft with HBV and on survival. However, this study is retrospective, which may be a limiting factor in the analysis.
The main finding of this study was that 372 HBsAg-positive liver recipients had an actuarial survival of 75 percent at one year and 63 percent at three years, and that the overall actuarial risk of HBV recurrence was 50 percent.
The three-year actuarial risk of recurrence of HBsAg was significantly higher among patients undergoing transplantation for HBV-related cirrhosis (67 percent) than among those undergoing transplantation for HDV-related cirrhosis (32 percent) or fulminant HBV (17 percent) (Figure 1). These results agree with those of previous studies, which showed that the type of liver disease at presentation influenced the rate of recurrence of HBV8,10,16,17. This study also confirms the effect of HBV replication before transplantation on the risk of recurrence after transplantation. The three-year actuarial risk of recurrence of HBV infection among patients with HBV-related cirrhosis who were seropositive for HBV DNA before transplantation was 83 percent, as compared with 66 percent among those seronegative for HBV DNA and seropositive for HBeAg, and 58 percent in those seronegative for both HBV DNA and HBeAg (P<0.05). Univariate analysis demonstrated that the presence of HBV DNA or HBeAg before transplantation, irrespective of the type of liver disease, was a factor predictive of recurrence of HBV. Multivariate analysis demonstrated that these two markers of HBV replication appeared to be independent predictive factors of recurrence among patients undergoing transplantation for HBV-related cirrhosis. In a previous study, Samuel et al. showed that HBV DNA was a prognostic indicator of recurrence of HBV in patients undergoing transplantation for HBV-related cirrhosis, but that a patient's status for HBeAg was not such an indicator16. However, other reports indicated that HBeAg status had a role as a marker of recurrence of HBV after transplantation10,17. It is clear from the present study that a patient's status for HBV replication, determined either by a simple hybridization technique for HBV DNA or by serologic examination for HBeAg, is important for predicting HBV recurrence. Both markers of HBV replication should be determined since many patients, especially those in Mediterranean areas, are positive for the antibody to HBeAg as well as for HBV DNA25,26.
In contrast, the three-year actuarial recurrence rate of HBV infection was much lower among patients undergoing transplantation for fulminant HDV, HDV-related cirrhosis, or fulminant HBV (40 percent, 32 percent, and 17 percent, respectively) (Figure 1). The lower risk of recurrence among the patients with HDV-related cirrhosis could be explained by the seronegativity for HBV DNA in most of these patients before transplantation and by the inhibitory effect of HDV on HBV replication9,27. The lower risk of HBV recurrence among the patients with fulminant HBV could be explained by the absence of detectable HBV replication according to molecular hybridization studies at the time of transplantation in most patients28. The three-year actuarial risk of HBV recurrence among patients with fulminant HDV was 40 percent, higher than the risk among patients with fulminant HBV. There is no clear explanation for this finding; we could not differentiate patients with HDV superinfection from those coinfected with HBV and HDV.
A second finding of this study was that anti-HBs immune globulin had a significant effect on the risk of HBV recurrence after transplantation. The patients given long-term therapy with this agent had a significant overall reduction in HBV recurrence, as compared with the patients given short-term therapy and those given no immunoprophylaxis (Figure 2). Interestingly, there was no difference in the risk of HBV recurrence between the two latter groups (Figure 2). In addition, long-term administration of immune globulin significantly reduced the recurrence of HBV after transplantation in three groups of patients with different types of liver disease -- HBV-related cirrhosis, HDV-related cirrhosis, and fulminant HBV. The efficacy of long-term treatment with anti-HBs immune globulin had been suggested by several nonrandomized studies from Paris, Hannover, and Berlin16,17,18. In contrast, investigators who gave patients either no immunoprophylaxis6,8,10 or short-term immunoprophylaxis10,13,14 observed a high risk of HBV recurrence -- about 80 percent. The beneficial effect of long-term passive prophylaxis with anti-HBs immune globulin supports the hypothesis that reinfection of the graft with HBV is related to the presence and replication of this virus in extrahepatic sites29,30,31. It should be emphasized that most cases of HBV reinfection occurred during the first year after transplantation. The marked difference in the risk of HBV recurrence at three months between the group given long-term therapy and the group given short-term therapy could be due to the higher amount of anti-HBs immune globulin administered to the former group and to the discontinuation of this drug in the latter group two months after transplantation. In the group receiving long-term immunoprophylaxis, most cases of HBV reinfection occurred during the first year after transplantation, and a few cases occurred after two years (Figure 2). This interesting observation raises questions about the proper duration of anti-HBs passive immunoprophylaxis. In this study, some investigators discontinued the therapy six months or one year after transplantation, and some groups continued it indefinitely. The issue of whether therapy can be discontinued after one year or two years should be addressed.
Our study showed that overall actuarial survival among HBsAg-positive patients one year and three years after liver transplantation was 75 percent and 63 percent, respectively. As among patients undergoing the procedure for other indications, outcome has improved continuously, with one-year survival of 57 percent and 83 percent of patients undergoing transplantation in 1986 and 1989, respectively. It is difficult to determine whether this improvement is related to improvement in the techniques of liver transplantation or to greater use of anti-HBs passive immunoprophylaxis in recent years, as suggested by Figure 4. We think that both factors are relevant. In this series of patients, 10.8 percent died of graft infection with HBV, which occurred more frequently in patients undergoing transplantation for HBV-related cirrhosis than in those undergoing it for HDV-related cirrhosis or fulminant HBV; this higher frequency may be due to the higher risk of HBV reinfection among patients with HBV-related cirrhosis but also -- intriguingly -- to a more severe course of HBV reinfection in this group. Thus, it is very important to long-term outcome that patients remain HBsAg-negative after transplantation, since the overall actuarial survival at one year and three years was significantly higher among patients who remained negative for HBsAg (90 percent and 83 percent, respectively) than among those who became HBsAg-positive again (73 percent and 54 percent, respectively).
We are indebted to Professor Solko W. Schalm, Project Leader of EUROHEP.
Source Information
From the Hepatobiliary Surgery and Liver Transplantation Research Unit, Paul Brousse Hospital, Villejuif, France (D.S., H.B.); Abteilung Gastroenterologie und Hepatologie, Medizinische Hochschule Hannover, Hannover, Germany (R.M.); the Department of Medicine, Addenbrooke's Hospital, Cambridge, United Kingdom (G.A.); Centro Trapianti, Ospedale Policlinico, Istituto Ricovero e Cura a Carattere Scientifico, Milan, Italy (L.F.); INSERM U 292, Le Kremlin-Bicetre, France (B.D.); and Service d'Hepatologie, Hopital Beaujon, Clichy, France (J.-P.B.). The investigators and institutions participating in the European Concerted Action on Viral Hepatitis (EUROHEP) study are listed in the Appendix.
Address reprint requests to Dr. Samuel at the Hepatobiliary Surgery and Liver Transplantation Research Unit, Paul Brousse Hospital, 14 Av. PV Couturier, 94800-Villejuif, France.
References
The following investigators and institutions were members of the EUROHEP Study Group on Hepatitis B Virus and Liver Transplantation: Hospital Universitario "Princeps D'Espanya," Barcelona, Spain -- T. Casanovas and E. Jaurrieta; Universitatsklinikum Rudolf Virchow, Berlin, Germany -- G. Blumhardt and P. Neuhaus; Queen Elizabeth Hospital, Birmingham, United Kingdom -- J. Neuberger and P. McMaster; Hopital Saint Andre, Bordeaux, France -- J. Saric and S. Winnock; Universite Catholique de Louvain, Brussels, Belgium -- J. Lerut and B. Otte; Addenbrooke's Hospital, Cambridge, United Kingdom -- G. Alexander and R.Y. Calne; S. Martino Hospital, Genoa, Italy -- M. Bertocchi and R. Pellicci; Chirurgische Universitat Klinik, Hamburg, Germany -- B. Kremer; Medizinische Hochschule Hannover, Hannover, Germany -- B. Ringe and R. Pichlmayr; King's College Hospital, London -- J.G. O'Grady and R. Williams; Hospital "12 de Octubre," Madrid, Spain -- E. Moreno-Gonzales and C. Loinaz Segurola; Istituto Ricovero e Cura a Carattere Scientifico, Milan, Italy -- L.R. Fassati and D. Galmarini; Chirurgische Klinik und Poliklinik der Technischen Universitat, Munich, Germany -- J. Adolf; Services de Chirurgie et d'Hepatologie, Hopital Bicetre, Kremlin-Bicetre, France -- O. Bernard and J. Valayer; Paul Brousse Hospital, Villejuif, France -- H. Bismuth and D. Samuel; University Hospital Rotterdam, Rotterdam, the Netherlands -- H.J. Metselaar and S.W. Schalm; and University Hospital, Zurich, Switzerland -- M. Decurtins and R. Schlumpf.
Executive Team of EUROHEP on liver transplantation: G. Alexander, J.P. Benhamou, H. Bismuth, L.R. Fassati, R. Muller, and D. Samuel.
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Liver Transplantation in Hepatitis B
Fabrega A. J., Pollack R., Wei M., Didier S., Alexander G., Bismuth H.
Extract |
Full Text
N Engl J Med 1994;
330:1317-1318, May 5, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |