813
immunoprophylaxis after liver transplantation in HBsAg-positive patients Passive
110 HBsAg-positive patients underwent orthotopic liver transplantation and received long-term antihepatitis B virus (HBV) passive immunoprophylaxis with anti-HBs immunoglobulin. During a mean follow-up period of 20 months, all patients became HBsAg negative after transplantation but circulating HBsAg reappeared in 25 (22·7%). Overall 1-year survival was 83·6% and overall 2-year actuarial recurrence of HBsAg was 29% (59% after posthepatitis B cirrhosis, 13% after posthepatitis B-delta cirrhosis, and 0% after fulminant hepatitis B). Patients with HBV cirrhosis who were HBV-DNA positive had a much greater risk of H BsAg recurrence than patients who were H BV-DNA negative (96% vs 29% at 2 years). Reappearance of HBsAg was associated with evidence of HBV replication and abnormal histological findings in the graft. Longterm passive anti-HBV immunoprophylaxis significantly reduced HBV reinfection and improved survival in patients without evidence of active HBV replication before orthotopic liver transplantation. Introduction
Orthotopic liver transplantation in patients with cirrhosis fulminant hepatitis caused by hepatitis B virus (HBV) infection is complicated by high rates of HBV infection of the transplanted liver,t.2 which have been blamed for disappointing survival rates in these patients.1 As a result, some surgeons are reluctant to transplant livers into patients positive for hepatitis B surface antigen (HBsAg). We describe a prospective study of the efficacy of long-term passive immunoprophylaxis against HBV for prevention of reinfection of the graft in a series ofHBsAg-positive patients with end-stage liver disease who received a transplant regardless of the HBV status (ie, presence of serum HBeAg, HBV DNA, or anti-delta antibody).3
or
Patients and methods an orthotopic liver transplant at Paul Brousse Hospital between June, 1986, and September, 1989, were studied prospectively (85 men, 25 women; mean age 37 [SD 10] years, range 15-56). The initial liver disease was posthepatitis B cirrhosis in 40, posthepatitis B-delta cirrhosis in 49, fulminant hepatitis B in 17, and fulminant hepatitis B-delta in 4. According to our modified Pugh’s classification4 65 (73%) of the 89 patients with cirrhosis had grade C disease and 24 (27%) had grade B involvement. 15 of the patients with posthepatitis B cirrhosis and 3 with posthepatitis B-delta cirrhosis also had hepatocellular carcinoma. 16 of 40 patients with posthepatitis B cirrhosis and 1 of 4
110
HBsAg-positive patients who received
with fulminant hepatitis B-delta were also HBV-DNA positive. HBV status was determined before transplantation by use of standard, commercially available, enzyme-linked immunosorbent assays (ELISAs) (HBsAg, anti-HBc, HBeAg, and anti-HBe, Wellcome Diagnostics, Kent, UK; anti-HBs, Hoechst Behring
Diagnostics, Marburg, Germany; HDV-Ag, Organon Teknika, Durham, Netherlands;
and
anti-HDV, Diagnostics Pasteur, was tested for by a standard molecular hybridisation spot test.5 After transplantation,
Mames La Coquette, France). HBV DNA
all patients were screened for HBsAg, anti-HBs, and HBV DNA at the end of the first postoperative week. Subsequently, serum anti-HBs was measured weekly, serum HBsAg and HBV DNA monthly, and serum HBeAg every 3 months. The first 19 patients received 10 000 IU intravenous anti-HBs immunoglobulin (Ig) perioperatively during the anhepatic phase, and 10 000 IU intravenously on every postoperative day until HBsAg disappeared from the serum; 3000 IU were then given intramuscularly every month (intravenous polyclonal anti-HBs Ig [Biotransfusion, Les Ulis, France] was administered as 10 000 IU in 200 ml over 2 h; intramuscular polyclonal anti-HBs Ig [Centre de Transfusion Sanguine de Nancy, Nancy-Vandoeuvre, France] was prepared as 10 000 IU in a 10 ml injection; both products were obtained from purified human serum). The other 91 patients received 10 000 IU intravenous anti-HBs Ig during the anhepatic phase and 10 000 IU intravenously during each of the first 6 postoperative days. 10 000 IU anti-HBs Ig were given intravenously whenever the circulating and-HBs antibody concentration fell below 100 IU/1 during follow-up. If HBsAg persisted in the serum despite readministration of anti-HBs Ig, further administration of anti-HBs Ig was held to be unjustifiable and was discontinued; such patients were considered to be HBsAg positive. With the second regimen patients received an average of fifteen 10 000 IU doses of anti-HBs Ig during the first postoperative year and 8 during the second year. Post-transplant immunosuppressive therapy consisted of a combination of prednisone, cyclosporin, and azathioprine .6 For long-term immunosuppression a combination of prednisone and cyclosporin or of prednisone, cyclosporin, and azathioprine was used. Post-transplantation liver biopsies were done at 6 months, 1 year, and at yearly intervals thereafter; a biopsy was also taken whenever graft dysfunction was suspected. Mean follow-up was for 19-6 (SD 11) months (range 0-50). A BMDP statistical software package (University of California Press, Berkeley, 1990) was used for statistical analyses. Survival and recurrence of HBV infection rates were calculated by actuarial life-table analysis. Survival rates were compared by the log-rank test and relative risks calculated with the Cox model.
Results All patients became HBsAg negative and anti-HBs positive by the end of the first postoperative week, and all patients who had been HBV-DNA positive became HBVDNA negative. Administration of these high doses of anti-HBs Ig was well tolerated, with only symptoms of mild arthralgia in some patients. HBsAg reappeared in the serum in 25 patients (22 7 % ), with a mean delay of 9-1months (SD 55, range 2-24). The actuarial recurrence rate of HBsAg in serum was 6 (2) %,17 (4) %, and 29 (6) % at 6 months,I year, and 2 years, respectively. The mean delay for reappearance of HBsAg was 55 months (SD 08, range 4-6) in HBV-DNA negative posthepatitis B cirrhosis, 7-5 months ADDRESSES: Hepatobiliary Surgery and Liver Transplantation Research Unit, Paris South University (D. Samuel, MD, Prof H. Bismuth, MD), Haematology Laboratory (A. Bismuth, MD, J.-L Arulnaden, MD), Microbiology Laboratory (D. Mathieu, PhD), and Pathology Laboratory (M. Reynes, MD), Hopital Paul Brousse,
Villejuif; Hepatology Unit, Hôpital Beaujon, Clichy (J.-P. Benhamou, MD); and Hybridotest Laboratory, Institut Pasteur, Paris, France (C. Brechot, MD). Correspondence to Dr D. Samuel, Hepatobiliary Surgery and Liver Transplantation Research Unit, Paris South University, Hôpital Paul Brousse, 94800 Villejuif, France.
814
TABLE
1-HBsAg REAP PEARANCEAFTERTRANS PLANT RELATED TO INITIAL LIVER DISEASE
(SD 4-3, range 2-12) in HBV-DNA positive posthepatitis B cirrhosis, and 16-8 months (SD 5, range 12-24) in posthepatitis B-delta cirrhosis. The actuarial 2-year recurrence rate of HBV infection was much higher in patients with posthepatitis B cirrhosis than in those with fulminant hepatitis B (p < 001) or with posthepatitis B-delta cirrhosis (p < 0-001; relative risk [RR]=72; table I). In patients with posthepatitis B cirrhosis, the actuarial rate of recurrent HBV infection at 2 years in those who were HBV-DNA positive before transplantation was significantly higher than for those who were HBV-DNA negative (96 [6]% vs 29 [10]%, p < 0-01; RR=47; table n) but there was no significant overall effect for preoperative HBeAg status (63 [15]% in HBeAg-positive vs 55 [11]% in HBeAg-negative patients; table II). Circulating HBsAg reappeared in 25 patients, always followed by reappearance of HBV DNA after a mean delay of 0-8 months (range 0-3). HBV DNA persisted throughout follow-up in all but 3 (12%; 2 with posthepatitis B-delta cirrhosis and 1 with posthepatitis B cirrhosis) who cleared both HBsAg and HBV DNA from the serum; all other patients with HBV recurrence remained HBsAg positive. The outcome of the liver graft was studied in the 93 patients for whom liver biopsy findings were available (12 patients died perioperatively, 2 could not undergo liver biopsy because of thrombocytopenia and 2 because of arterial thrombosis with intrahepatic strictures, and 1 patient refused biopsy). Histological appearances were normal in 1 of 21 (4-8%) patients in whom HBsAg recurred compared with 60 of 72 (83 3%) who remained HBsAg negative (p < 0-001; table III). In 20 of the 21 patients with recurrent circulating HBsAg an episode of acute hepatitis occurred within 3 months and led to chronic hepatitis in 9 and to cirrhosis in 3 (at 18 months, 18 months, and 36 months; table III). The 3 patients who developed cirrhosis and 3 patients who had acute hepatitis which led to liver failure and repeat transplantation had posthepatitis B cirrhosis as the initial liver disease; no patient with posthepatitis B-delta cirrhosis or fulminant hepatitis B or hepatitis B-delta had post-transplantation cirrhosis or liver failure during follow-up. 72 patients remained HBsAg negative and had liver biopsies: 9 (125%) had an episode of acute hepatitis but had no serological or hepatic markers of recurrent HBV infection. 25 patients died during follow-up, 20 from causes other than HBV reinfection (12 died perioperatively from brain death [3], surgical complications [3], sepsis [5], and myocardial infarction [1]; 8 died more than 4 months postoperatively from recurrent hepatocellular carcinoma [3], sepsis [2], chronic rejection [1], Kaposi’s sarcoma [1], and cerebrovascular accident [1]). 5 patients died from recurrent HBV infection at 7-21 months: 1 from cirrhosis in the absence of a suitable donor and 4 after retransplantation. No patient with fulminant hepatitis B or posthepatitis
B-delta cirrhosis died or
required retransplantation.
TABLE ∥-HBsAg REAPPEARANCE IN POSTHEPATITIS B CIRRHOSIS RELATED TO PRETRANSPLANT HBV STATUS
TABLE III-HISTOLOGICAL FINDINGS RELATED TO RECURRENCE
HBsAg
Overall survival after transplantation was 83’6% at 1 year (80%, 76-4%, and 91-8% after posthepatitis B cirrhosis, fulminant hepatitis B, and posthepatitis B-delta cirrhosis, respectively). Overall actuarial survival at 2 years was 74% (57 [9] %, 71 [ 11] %, and 89 [5]%, after posthepatitis B cirrhosis, fulminant hepatitis B, and posthepatitis B-delta cirrhosis, respectively). 2-year survival was similar after fulminant hepatitis B and posthepatitis B cirrhosis (RR=1 1), but significantly better after posthepatitis Bdelta cirrhosis (p 0-05, RR =1 33).
Discussion Patients with liver disease caused by HBV who undergo liver transplantation have a high risk of recurrent HBV infection in the graft. Furthermore, immunosuppressive agents facilitate HBV replication and may increase the liver damage caused by HBV infection.7-11 Results after liver transplantation for HBsAg-positive liver disease in patients not given passive anti-HBV immunoprophylaxis or who received such prophylaxis only perioperatively and during the first few days postoperatively (to neutralise circulating HBV particles and avoid immediate reinfection of the graft) have been disappointing.2,12 Lauchart et al13 reported promising results in patients who received high doses of anti-HBs Ig during the immediate postoperative period and for 6 months postoperatively to maintain serum anti-HBs concentrations above 100 IU/1: all patients became HBsAg negative and reappearance of HBsAg in the serum during the first 6 months only occurred in patients who were HBV-DNA positive before transplantation. 13 However, longer follow-up indicated that 6 months of passive immunoprophylaxis only delayed HBV recurrence. We thought it likely that extrahepatic focuses of HBV infection could persist for longer than 6 months, and decided to regularly protect all our patients with anti-HBs Ig without
815
discontinuation; after the first 19 patients the regimen of long-term prophylaxis was modified as it seemed appropriate to relate anti-HBs Ig administration to serum anti-HBs antibody concentration. We found that recurrent post-transplant infection was much more likely in patients who had had posthepatitis B cirrhosis than those with fulminant hepatitis B or posthepatitis B-delta cirrhosis. Pretransplant evidence of HBV replication, indicated by serum HBV DNA, significantly influenced the outcome, with recrudescent HBV infection in 96% of HBV-DNApositive patients compared with 29% of those who were HBV-DNA negative. The presence of HBeAg before transplantation was a less accurate predictor of risk, although in patients with posthepatitis B cirrhosis the lowest rate of HBV reinfection (25 %) was seen in those who were HBV-DNA and HBeAg negative. All our patients with posthepatitis B-delta cirrhosis and fulminant hepatitis B were HBV-DNA negative before transplantation, which may explain their low incidence of HBV reinfection. Recurrent HBV infection was more likely in patients with posthepatitis B cirrhosis than in patients with fulminant hepatitis B or posthepatitis B-delta cirrhosis, independent of evidence of preoperative HBV replication: in fulminant hepatitis B, HBV replication is usually interrupted14 and few who recover spontaneously progress to chronic hepatitis B. The low rate of HBV reinfection in patients who had had posthepatitis B-delta cirrhosis may reflect inhibition of HBV replication by the delta virus. 15,16
Reappearance of HBsAg in serum was invariably by HBV-DNA positivity, indicative of HBV replication, and by a high incidence of chronic hepatitis. The severity of post-transplant reinfection and the rapid evolution of cirrhosis in 3 of our patients could be influenced by immunosuppressive treatment,10,11 but the occurrence of cirrhosis as early as 10-18 months after transplantation in 2 of our patients and in a patient in another series12 indicates followed
that other mechanisms may be involved. Our findings indicate that long-term passive immunisation of HBsAg-positive patients reduces the incidence of HBV reinfection of the liver graft after transplantation and improves patient and graft survival, with a 2-year actuarial incidence of HBV reinfection of 29% compared with 83% (5 of 6)2 and 89% (8 of 9)12 in other series and 9 of 10 patients treated at our unit before the introduction of this immunisation regimen (unpublished observations). Prevention of HBV reinfection improved patient survival: indeed, we have found similar 1-year survival rates after transplantation for HBsAg-negative and HBsAg-positive liver disease3-a striking contrast to other series.1 However we also found that late HBV reinfection could occur more than a year after transplantation. All patients became HBsAg negative during the first week of immunisation, and for patients with reappearance of serum HBsAg the mean interval was 9-1months. 83% of patients who remained HBsAg negative during immunisation had normal liver histology in biopsy specimens, compared with only 1 of 21 in whom HBV infection recurred. With a polymerase chain reaction assay we have found that HBV DNA may be absent from the liver graft ofHBsAg-negative patients treated with anti-HBs Ig but was detectable in peripheral blood mononuclear cells and serum,17 and have found replicative forms of HBV in peripheral blood mononuclear cells when they were not found in serum or the liver graft.18 These findings indicate that viral reinfection of the graft may arise from extrahepatic sites rather than from
graft by circulating HBV, and may explain the disappointing long-term results of passive immunoprophylaxis for only 6 months after transplantation. We also found that immunoprophylaxis failed to prevent reinfection in most patients who had evidence of active viral replication, with circulating HBV DNA, before transplantation. Results may be improved in these patients if HBV replication could be arrested preoperatively: indeed, the role of concurrent antiviral chemotherapy in prevention of HBV reinfection in all patients who undergo liver transplantation needs to be assessed. Use of long-term immunoprophylaxis must be balanced against potential side-effects and cost. No patient had severe immune-complex disease; mild arthralgia was the only symptom noted. 10 000 IU intravenous anti-HBs Ig cost approximately L500 ($1000 US), and in 12 months of our regimen average cost was 7500 for the first year and around £4000 for the second. Such sums seem justifiable when compared with the overall costs of transplantation and the strikingly reduced incidence of HBV reinfection. immediate infection of the
We thank Dr R. Padbury, Dr D. Azoulay, and Dr T. Diamond for helpful criticism, Mrs B. Ducot for statistical analyses, and Mrs M. Gigou and Miss C. Serres for technical assistance.
REFERENCES 1. Starzl TE, Demetris AJ, van Thiel D. Liver transplantation. N Engl J Med 1989; 321: 1092-99. 2. Lauchart W, Müller R, Pichlmayr R. Immunoprophylaxis of hepatitis B virus reinfection in recipients of human liver allografts. Transplant Proc 1987; 19: 2387-89. 3. Bismuth H. Liver transplantation. The Paul Brousse experience. Transplant Proc 1988; 20S: 486-89. 4. Bismuth H, Adam R, Mathur S, Sherlock D. Options for elective treatment of portal hypertension in cirrhotic patients in the
transplantation era. Am J Surg 1990; 160: 105-10. 5. Scotto J, Hadchouel M, Hery C, Yvart J, Tiollais P, Brechot C. Detection of hepatitis B virus DNA in serum by a simple spot hybridization technique: comparison with results of other viral markers. Hepatology
1983; 3: 279-84.
Gugenheim J, Samuel D, Saliba F, Castaing D, Bismuth H. Use of flexible triple drug immunosuppressive therapy in liver transplantation. Transplant Proc 1987; 19: 3805-07. 7. Nagington J, Cossart YE, Cohen BJ. Reactivation of hepatitis B after transplantation operations. Lancet 1977; i: 558-60. 8. Hoofnagle JH, Dusheiko GM, Shafer DF, et al. Reactivation of chronic active hepatitis B virus infection by cancer chemotherapy. Ann Intern 6.
Med 1982; 96: 447-49. 9. Lam KC, Lai CL, Trepo C, Wu PC. Deleterious effect of prednisolone in HBs Ag positive chronic active hepatitis. N Engl J Med 1981; 304: 380-86. 10. Degos F, Lugassy C, Degott C, et al. Hepatitis B virus and hepatitis B related viral infection in renal transplant recipients. A prospective study of 90 patients. Gastroenterology 1988; 94: 151-56. 11. Parfrey PS, Forbes RDC, Hutchinson TA, et al. The impact of renal transplantation on the course of hepatitis B liver disease. Transplantation 1985; 39: 610-15. 12. Demetris AJ, Jaffe R, Sheahan DG, et al. Recurrent hepatitis B in liver allograft recipients. Differentiation between viral hepatitis B and
rejection. Am J Pathol 1986; 125: 161-72. 13. Lauchart W, Müller R, Pichlmayr R. Long-term immunoprophylaxis of hepatitis B virus (HBV) reinfection in recipients of human liver allografts. Transplant Proc 1987; 19: 4051-53. 14. Brechot C, Bernuau J, Thiers V, et al. Multiplication of hepatitis B virus in fulminant hepatitis B. Br Med J 1984; 288: 270-71. 15. Rizzetto M. The delta agent. Hepatology 1983; 3: 729-37. 16. Krogsgaard K, Kryger P, Aldershvile J, et al. Delta infection and suppression of hepatitis B virus replication in chronic HBsAg carriers. Hepatology 1987; 7: 42-45. 17. Feray C, Zignego AL, Samuel D, et al. Persistent hepatitis B virus infection of mononuclear cells without concomitant liver infection: the liver transplantation model. Transplantation 1990; 49: 1155-58. 18. Zignego AL, Samuel D, Gugenheim J, et al. Hepatitis B virus replication and mononuclear blood cell infection after liver transplantation. In: Zuckermann AJ, ed. Viral hepatitis and liver disease. New York: Alan R. Liss, 1988: 808-09.
immunoprophylaxis after liver transplantation in HBsAg-positive patients Passive
110 HBsAg-positive patients underwent orthotopic liver transplantation and received long-term antihepatitis B virus (HBV) passive immunoprophylaxis with anti-HBs immunoglobulin. During a mean follow-up period of 20 months, all patients became HBsAg negative after transplantation but circulating HBsAg reappeared in 25 (22·7%). Overall 1-year survival was 83·6% and overall 2-year actuarial recurrence of HBsAg was 29% (59% after posthepatitis B cirrhosis, 13% after posthepatitis B-delta cirrhosis, and 0% after fulminant hepatitis B). Patients with HBV cirrhosis who were HBV-DNA positive had a much greater risk of H BsAg recurrence than patients who were H BV-DNA negative (96% vs 29% at 2 years). Reappearance of HBsAg was associated with evidence of HBV replication and abnormal histological findings in the graft. Longterm passive anti-HBV immunoprophylaxis significantly reduced HBV reinfection and improved survival in patients without evidence of active HBV replication before orthotopic liver transplantation. Introduction
Orthotopic liver transplantation in patients with cirrhosis fulminant hepatitis caused by hepatitis B virus (HBV) infection is complicated by high rates of HBV infection of the transplanted liver,t.2 which have been blamed for disappointing survival rates in these patients.1 As a result, some surgeons are reluctant to transplant livers into patients positive for hepatitis B surface antigen (HBsAg). We describe a prospective study of the efficacy of long-term passive immunoprophylaxis against HBV for prevention of reinfection of the graft in a series ofHBsAg-positive patients with end-stage liver disease who received a transplant regardless of the HBV status (ie, presence of serum HBeAg, HBV DNA, or anti-delta antibody).3
or
Patients and methods an orthotopic liver transplant at Paul Brousse Hospital between June, 1986, and September, 1989, were studied prospectively (85 men, 25 women; mean age 37 [SD 10] years, range 15-56). The initial liver disease was posthepatitis B cirrhosis in 40, posthepatitis B-delta cirrhosis in 49, fulminant hepatitis B in 17, and fulminant hepatitis B-delta in 4. According to our modified Pugh’s classification4 65 (73%) of the 89 patients with cirrhosis had grade C disease and 24 (27%) had grade B involvement. 15 of the patients with posthepatitis B cirrhosis and 3 with posthepatitis B-delta cirrhosis also had hepatocellular carcinoma. 16 of 40 patients with posthepatitis B cirrhosis and 1 of 4
110
HBsAg-positive patients who received
with fulminant hepatitis B-delta were also HBV-DNA positive. HBV status was determined before transplantation by use of standard, commercially available, enzyme-linked immunosorbent assays (ELISAs) (HBsAg, anti-HBc, HBeAg, and anti-HBe, Wellcome Diagnostics, Kent, UK; anti-HBs, Hoechst Behring
Diagnostics, Marburg, Germany; HDV-Ag, Organon Teknika, Durham, Netherlands;
and
anti-HDV, Diagnostics Pasteur, was tested for by a standard molecular hybridisation spot test.5 After transplantation,
Mames La Coquette, France). HBV DNA
all patients were screened for HBsAg, anti-HBs, and HBV DNA at the end of the first postoperative week. Subsequently, serum anti-HBs was measured weekly, serum HBsAg and HBV DNA monthly, and serum HBeAg every 3 months. The first 19 patients received 10 000 IU intravenous anti-HBs immunoglobulin (Ig) perioperatively during the anhepatic phase, and 10 000 IU intravenously on every postoperative day until HBsAg disappeared from the serum; 3000 IU were then given intramuscularly every month (intravenous polyclonal anti-HBs Ig [Biotransfusion, Les Ulis, France] was administered as 10 000 IU in 200 ml over 2 h; intramuscular polyclonal anti-HBs Ig [Centre de Transfusion Sanguine de Nancy, Nancy-Vandoeuvre, France] was prepared as 10 000 IU in a 10 ml injection; both products were obtained from purified human serum). The other 91 patients received 10 000 IU intravenous anti-HBs Ig during the anhepatic phase and 10 000 IU intravenously during each of the first 6 postoperative days. 10 000 IU anti-HBs Ig were given intravenously whenever the circulating and-HBs antibody concentration fell below 100 IU/1 during follow-up. If HBsAg persisted in the serum despite readministration of anti-HBs Ig, further administration of anti-HBs Ig was held to be unjustifiable and was discontinued; such patients were considered to be HBsAg positive. With the second regimen patients received an average of fifteen 10 000 IU doses of anti-HBs Ig during the first postoperative year and 8 during the second year. Post-transplant immunosuppressive therapy consisted of a combination of prednisone, cyclosporin, and azathioprine .6 For long-term immunosuppression a combination of prednisone and cyclosporin or of prednisone, cyclosporin, and azathioprine was used. Post-transplantation liver biopsies were done at 6 months, 1 year, and at yearly intervals thereafter; a biopsy was also taken whenever graft dysfunction was suspected. Mean follow-up was for 19-6 (SD 11) months (range 0-50). A BMDP statistical software package (University of California Press, Berkeley, 1990) was used for statistical analyses. Survival and recurrence of HBV infection rates were calculated by actuarial life-table analysis. Survival rates were compared by the log-rank test and relative risks calculated with the Cox model.
Results All patients became HBsAg negative and anti-HBs positive by the end of the first postoperative week, and all patients who had been HBV-DNA positive became HBVDNA negative. Administration of these high doses of anti-HBs Ig was well tolerated, with only symptoms of mild arthralgia in some patients. HBsAg reappeared in the serum in 25 patients (22 7 % ), with a mean delay of 9-1months (SD 55, range 2-24). The actuarial recurrence rate of HBsAg in serum was 6 (2) %,17 (4) %, and 29 (6) % at 6 months,I year, and 2 years, respectively. The mean delay for reappearance of HBsAg was 55 months (SD 08, range 4-6) in HBV-DNA negative posthepatitis B cirrhosis, 7-5 months ADDRESSES: Hepatobiliary Surgery and Liver Transplantation Research Unit, Paris South University (D. Samuel, MD, Prof H. Bismuth, MD), Haematology Laboratory (A. Bismuth, MD, J.-L Arulnaden, MD), Microbiology Laboratory (D. Mathieu, PhD), and Pathology Laboratory (M. Reynes, MD), Hopital Paul Brousse,
Villejuif; Hepatology Unit, Hôpital Beaujon, Clichy (J.-P. Benhamou, MD); and Hybridotest Laboratory, Institut Pasteur, Paris, France (C. Brechot, MD). Correspondence to Dr D. Samuel, Hepatobiliary Surgery and Liver Transplantation Research Unit, Paris South University, Hôpital Paul Brousse, 94800 Villejuif, France.
814
TABLE
1-HBsAg REAP PEARANCEAFTERTRANS PLANT RELATED TO INITIAL LIVER DISEASE
(SD 4-3, range 2-12) in HBV-DNA positive posthepatitis B cirrhosis, and 16-8 months (SD 5, range 12-24) in posthepatitis B-delta cirrhosis. The actuarial 2-year recurrence rate of HBV infection was much higher in patients with posthepatitis B cirrhosis than in those with fulminant hepatitis B (p < 001) or with posthepatitis B-delta cirrhosis (p < 0-001; relative risk [RR]=72; table I). In patients with posthepatitis B cirrhosis, the actuarial rate of recurrent HBV infection at 2 years in those who were HBV-DNA positive before transplantation was significantly higher than for those who were HBV-DNA negative (96 [6]% vs 29 [10]%, p < 0-01; RR=47; table n) but there was no significant overall effect for preoperative HBeAg status (63 [15]% in HBeAg-positive vs 55 [11]% in HBeAg-negative patients; table II). Circulating HBsAg reappeared in 25 patients, always followed by reappearance of HBV DNA after a mean delay of 0-8 months (range 0-3). HBV DNA persisted throughout follow-up in all but 3 (12%; 2 with posthepatitis B-delta cirrhosis and 1 with posthepatitis B cirrhosis) who cleared both HBsAg and HBV DNA from the serum; all other patients with HBV recurrence remained HBsAg positive. The outcome of the liver graft was studied in the 93 patients for whom liver biopsy findings were available (12 patients died perioperatively, 2 could not undergo liver biopsy because of thrombocytopenia and 2 because of arterial thrombosis with intrahepatic strictures, and 1 patient refused biopsy). Histological appearances were normal in 1 of 21 (4-8%) patients in whom HBsAg recurred compared with 60 of 72 (83 3%) who remained HBsAg negative (p < 0-001; table III). In 20 of the 21 patients with recurrent circulating HBsAg an episode of acute hepatitis occurred within 3 months and led to chronic hepatitis in 9 and to cirrhosis in 3 (at 18 months, 18 months, and 36 months; table III). The 3 patients who developed cirrhosis and 3 patients who had acute hepatitis which led to liver failure and repeat transplantation had posthepatitis B cirrhosis as the initial liver disease; no patient with posthepatitis B-delta cirrhosis or fulminant hepatitis B or hepatitis B-delta had post-transplantation cirrhosis or liver failure during follow-up. 72 patients remained HBsAg negative and had liver biopsies: 9 (125%) had an episode of acute hepatitis but had no serological or hepatic markers of recurrent HBV infection. 25 patients died during follow-up, 20 from causes other than HBV reinfection (12 died perioperatively from brain death [3], surgical complications [3], sepsis [5], and myocardial infarction [1]; 8 died more than 4 months postoperatively from recurrent hepatocellular carcinoma [3], sepsis [2], chronic rejection [1], Kaposi’s sarcoma [1], and cerebrovascular accident [1]). 5 patients died from recurrent HBV infection at 7-21 months: 1 from cirrhosis in the absence of a suitable donor and 4 after retransplantation. No patient with fulminant hepatitis B or posthepatitis
B-delta cirrhosis died or
required retransplantation.
TABLE ∥-HBsAg REAPPEARANCE IN POSTHEPATITIS B CIRRHOSIS RELATED TO PRETRANSPLANT HBV STATUS
TABLE III-HISTOLOGICAL FINDINGS RELATED TO RECURRENCE
HBsAg
Overall survival after transplantation was 83’6% at 1 year (80%, 76-4%, and 91-8% after posthepatitis B cirrhosis, fulminant hepatitis B, and posthepatitis B-delta cirrhosis, respectively). Overall actuarial survival at 2 years was 74% (57 [9] %, 71 [ 11] %, and 89 [5]%, after posthepatitis B cirrhosis, fulminant hepatitis B, and posthepatitis B-delta cirrhosis, respectively). 2-year survival was similar after fulminant hepatitis B and posthepatitis B cirrhosis (RR=1 1), but significantly better after posthepatitis Bdelta cirrhosis (p 0-05, RR =1 33).
Discussion Patients with liver disease caused by HBV who undergo liver transplantation have a high risk of recurrent HBV infection in the graft. Furthermore, immunosuppressive agents facilitate HBV replication and may increase the liver damage caused by HBV infection.7-11 Results after liver transplantation for HBsAg-positive liver disease in patients not given passive anti-HBV immunoprophylaxis or who received such prophylaxis only perioperatively and during the first few days postoperatively (to neutralise circulating HBV particles and avoid immediate reinfection of the graft) have been disappointing.2,12 Lauchart et al13 reported promising results in patients who received high doses of anti-HBs Ig during the immediate postoperative period and for 6 months postoperatively to maintain serum anti-HBs concentrations above 100 IU/1: all patients became HBsAg negative and reappearance of HBsAg in the serum during the first 6 months only occurred in patients who were HBV-DNA positive before transplantation. 13 However, longer follow-up indicated that 6 months of passive immunoprophylaxis only delayed HBV recurrence. We thought it likely that extrahepatic focuses of HBV infection could persist for longer than 6 months, and decided to regularly protect all our patients with anti-HBs Ig without
815
discontinuation; after the first 19 patients the regimen of long-term prophylaxis was modified as it seemed appropriate to relate anti-HBs Ig administration to serum anti-HBs antibody concentration. We found that recurrent post-transplant infection was much more likely in patients who had had posthepatitis B cirrhosis than those with fulminant hepatitis B or posthepatitis B-delta cirrhosis. Pretransplant evidence of HBV replication, indicated by serum HBV DNA, significantly influenced the outcome, with recrudescent HBV infection in 96% of HBV-DNApositive patients compared with 29% of those who were HBV-DNA negative. The presence of HBeAg before transplantation was a less accurate predictor of risk, although in patients with posthepatitis B cirrhosis the lowest rate of HBV reinfection (25 %) was seen in those who were HBV-DNA and HBeAg negative. All our patients with posthepatitis B-delta cirrhosis and fulminant hepatitis B were HBV-DNA negative before transplantation, which may explain their low incidence of HBV reinfection. Recurrent HBV infection was more likely in patients with posthepatitis B cirrhosis than in patients with fulminant hepatitis B or posthepatitis B-delta cirrhosis, independent of evidence of preoperative HBV replication: in fulminant hepatitis B, HBV replication is usually interrupted14 and few who recover spontaneously progress to chronic hepatitis B. The low rate of HBV reinfection in patients who had had posthepatitis B-delta cirrhosis may reflect inhibition of HBV replication by the delta virus. 15,16
Reappearance of HBsAg in serum was invariably by HBV-DNA positivity, indicative of HBV replication, and by a high incidence of chronic hepatitis. The severity of post-transplant reinfection and the rapid evolution of cirrhosis in 3 of our patients could be influenced by immunosuppressive treatment,10,11 but the occurrence of cirrhosis as early as 10-18 months after transplantation in 2 of our patients and in a patient in another series12 indicates followed
that other mechanisms may be involved. Our findings indicate that long-term passive immunisation of HBsAg-positive patients reduces the incidence of HBV reinfection of the liver graft after transplantation and improves patient and graft survival, with a 2-year actuarial incidence of HBV reinfection of 29% compared with 83% (5 of 6)2 and 89% (8 of 9)12 in other series and 9 of 10 patients treated at our unit before the introduction of this immunisation regimen (unpublished observations). Prevention of HBV reinfection improved patient survival: indeed, we have found similar 1-year survival rates after transplantation for HBsAg-negative and HBsAg-positive liver disease3-a striking contrast to other series.1 However we also found that late HBV reinfection could occur more than a year after transplantation. All patients became HBsAg negative during the first week of immunisation, and for patients with reappearance of serum HBsAg the mean interval was 9-1months. 83% of patients who remained HBsAg negative during immunisation had normal liver histology in biopsy specimens, compared with only 1 of 21 in whom HBV infection recurred. With a polymerase chain reaction assay we have found that HBV DNA may be absent from the liver graft ofHBsAg-negative patients treated with anti-HBs Ig but was detectable in peripheral blood mononuclear cells and serum,17 and have found replicative forms of HBV in peripheral blood mononuclear cells when they were not found in serum or the liver graft.18 These findings indicate that viral reinfection of the graft may arise from extrahepatic sites rather than from
graft by circulating HBV, and may explain the disappointing long-term results of passive immunoprophylaxis for only 6 months after transplantation. We also found that immunoprophylaxis failed to prevent reinfection in most patients who had evidence of active viral replication, with circulating HBV DNA, before transplantation. Results may be improved in these patients if HBV replication could be arrested preoperatively: indeed, the role of concurrent antiviral chemotherapy in prevention of HBV reinfection in all patients who undergo liver transplantation needs to be assessed. Use of long-term immunoprophylaxis must be balanced against potential side-effects and cost. No patient had severe immune-complex disease; mild arthralgia was the only symptom noted. 10 000 IU intravenous anti-HBs Ig cost approximately L500 ($1000 US), and in 12 months of our regimen average cost was 7500 for the first year and around £4000 for the second. Such sums seem justifiable when compared with the overall costs of transplantation and the strikingly reduced incidence of HBV reinfection. immediate infection of the
We thank Dr R. Padbury, Dr D. Azoulay, and Dr T. Diamond for helpful criticism, Mrs B. Ducot for statistical analyses, and Mrs M. Gigou and Miss C. Serres for technical assistance.
REFERENCES 1. Starzl TE, Demetris AJ, van Thiel D. Liver transplantation. N Engl J Med 1989; 321: 1092-99. 2. Lauchart W, Müller R, Pichlmayr R. Immunoprophylaxis of hepatitis B virus reinfection in recipients of human liver allografts. Transplant Proc 1987; 19: 2387-89. 3. Bismuth H. Liver transplantation. The Paul Brousse experience. Transplant Proc 1988; 20S: 486-89. 4. Bismuth H, Adam R, Mathur S, Sherlock D. Options for elective treatment of portal hypertension in cirrhotic patients in the
transplantation era. Am J Surg 1990; 160: 105-10. 5. Scotto J, Hadchouel M, Hery C, Yvart J, Tiollais P, Brechot C. Detection of hepatitis B virus DNA in serum by a simple spot hybridization technique: comparison with results of other viral markers. Hepatology
1983; 3: 279-84.
Gugenheim J, Samuel D, Saliba F, Castaing D, Bismuth H. Use of flexible triple drug immunosuppressive therapy in liver transplantation. Transplant Proc 1987; 19: 3805-07. 7. Nagington J, Cossart YE, Cohen BJ. Reactivation of hepatitis B after transplantation operations. Lancet 1977; i: 558-60. 8. Hoofnagle JH, Dusheiko GM, Shafer DF, et al. Reactivation of chronic active hepatitis B virus infection by cancer chemotherapy. Ann Intern 6.
Med 1982; 96: 447-49. 9. Lam KC, Lai CL, Trepo C, Wu PC. Deleterious effect of prednisolone in HBs Ag positive chronic active hepatitis. N Engl J Med 1981; 304: 380-86. 10. Degos F, Lugassy C, Degott C, et al. Hepatitis B virus and hepatitis B related viral infection in renal transplant recipients. A prospective study of 90 patients. Gastroenterology 1988; 94: 151-56. 11. Parfrey PS, Forbes RDC, Hutchinson TA, et al. The impact of renal transplantation on the course of hepatitis B liver disease. Transplantation 1985; 39: 610-15. 12. Demetris AJ, Jaffe R, Sheahan DG, et al. Recurrent hepatitis B in liver allograft recipients. Differentiation between viral hepatitis B and
rejection. Am J Pathol 1986; 125: 161-72. 13. Lauchart W, Müller R, Pichlmayr R. Long-term immunoprophylaxis of hepatitis B virus (HBV) reinfection in recipients of human liver allografts. Transplant Proc 1987; 19: 4051-53. 14. Brechot C, Bernuau J, Thiers V, et al. Multiplication of hepatitis B virus in fulminant hepatitis B. Br Med J 1984; 288: 270-71. 15. Rizzetto M. The delta agent. Hepatology 1983; 3: 729-37. 16. Krogsgaard K, Kryger P, Aldershvile J, et al. Delta infection and suppression of hepatitis B virus replication in chronic HBsAg carriers. Hepatology 1987; 7: 42-45. 17. Feray C, Zignego AL, Samuel D, et al. Persistent hepatitis B virus infection of mononuclear cells without concomitant liver infection: the liver transplantation model. Transplantation 1990; 49: 1155-58. 18. Zignego AL, Samuel D, Gugenheim J, et al. Hepatitis B virus replication and mononuclear blood cell infection after liver transplantation. In: Zuckermann AJ, ed. Viral hepatitis and liver disease. New York: Alan R. Liss, 1988: 808-09.
