Successful treatment of de novo autoimmunehepatitis and cirrhosis after pediatricliver transplantation

Nelson E. Gibelli, Uenis Tannuri,Evandro S. Mello, Eduardo R. Cancado,Maria M. Santos, Ali A. Ayoub, Joao G.Maksoud-Filho, Manoel C. P. Velhote,Marcos M. Silva, Maria L. Pinho-Apezzato and Joao G. MaksoudLiver Transplantation Unit, Children Institute, Hospitaldas Clinicas, University of Sao Paulo, Sao Paulo, Brazil

Gibelli NE, Tannuri U, Mello ES, Cancado ER, Santos MM, AyoubAA, Maksoud-Filho JG, Velhote MCP, Silva MM, Pinho-ApezzatoML, Maksoud JG. Successful treatment of de novo autoimmunehepatitis and cirrhosis after pediatric liver transplantation.Pediatr Transplantation 2006: 10: 371–376. r 2006 BlackwellMunksgaard

Abstract: Over a 15-yr period of observation, among the 205 childrenwho underwent liver transplantations, one of them developed aparticular type of late graft dysfunction with clinical and histologicalsimilarity to autoimmune hepatitis. The patient had a1-antitrypsindeficiency and did not previously have autoimmune hepatitis or anyother autoimmune disease before transplantation. Infectious andsurgical complications were excluded. After repeated episodes ofunexplained fluctuations of liver function tests and liver biopsiesdemonstrating reactive or a biliary pattern, without any correspondingalteration of percutaneous cholangiography, a liver-biopsy sampletaken 4 yr after the transplant showed active chronic hepatitisprogressing to cirrhosis, portal lymphocyte aggregates, and a largenumber of plasma cells. At that time, autoantibodies (gastric parietalcell antibody, liver–kidney microsomal antibody, and anti-hepaticcytosol) were positive and serum IgG levels were high. Based on thesefindings of autoimmune disease, a diagnosis of ‘de novo autoimmunehepatitis’ was made. The treatment consisted of reducing the dose ofcyclosporine, reintroduction of corticosteroids, and addition ofmycophenolate mofetil. After 19 months of treatment, a new liver-biopsy sample showed marked reduction of portal and lobularinflammatory infiltrate, with regression of fibrosis and of thearchitectural disruption. At that time, serum autoantibodies becamenegative. The last liver-biopsy sample showed inactive cirrhosis anddisappearance of interface hepatitis and of plasma cell infiltrate.Presently, 9 yr after the transplantation, the patient is doing well, withnormal liver function tests and no evidence of cirrhosis. Herimmunosuppressive therapy consists of tacrolimus, mycophenolatemofetil, and prednisolone. In conclusion, the present case demonstratesthat de novo autoimmune hepatitis can appear in liver-transplantpatients despite appropriate anti-rejection immunosuppression, andtriple therapy with tacrolimus, mycophenolate mofetil, andprednisolone could sustain the graft and prevent retransplantation.

������������������������������������������������������������������������

������������������������������������������������������������������������

Late graft dysfunction after pediatric liver trans-plantation is always a reason of concern, andsometimes it does not result from easily recog-nized causes, such as acute or chronic rejection,infection, vascular or biliary complications, recur-rence of the original liver disease, and lymphopro-liferative disease related to the Epstein–Barr virus

(1–5). We report a particular type of unexplainedlate graft dysfunction that occurs in associationwith serology and histology features that are com-patible with autoimmune hepatitis in a patientwho did not suffer from autoimmune hepatitis orany other autoimmune disease before transplan-tation. This special type of graft dysfunction has

Pediatr Transplantation 2006: 10: 371–376. DOI: 10.1111/j.1399-3046.2005.00470.xPrinted in Singapore. All rights reserved.

Copyright r 2006 Blackwell Munksgaard

Pediatric Transplantation

Key words: liver transplantation – autoimmunehepatitis – immunosuppression

Professor Uenis Tannuri, Faculdade de Medicina daUniversidade de Sao Paulo, Avenida Dr. Arnaldo 455,41 andar, sala 4106, Sao Paulo, CEP: 01246-903,SP, BrazilTel.: 0055-11-30812943Fax: 0055-11-32556285E-mail: uenist@usp.br

Accepted for publication 19 October 2005

371

been reported in children (6, 7) and adults (8, 9),and its recognition is important because promptappropriate treatment can prevent liver graft lossand retransplantation.

Case report

Our liver transplantation program was started in1989, and till now (July 2005) 256 liver transplan-tations have been performed in 205 pediatricpatients utilizing grafts from living or deceaseddonors. Over this period of observation, onerecipient presented a characteristic form of graftdysfunction. W.L.B.S, female, white, 9 yr old,40 kg, blood type O1 , underwent an orthotopicliver transplantation with a deceased donor graftin December 6, 1995. The indication for livertransplantation was a1-antitrypsin deficiency(phenotype determination was not performed)and hepatic failure. Pretransplant autoantibodiesand serological tests were negative, with the ex-ception of HCV positivity by a second-generationenzyme-linked immunoabsorbent assay, probablyrelated to previous blood transfusions. No poly-merase chain reaction for hepatitis C virus (PCR–HCV) was performed. The donor was a 12-yr-old,female, 45 kg, with the same blood type (O1); se-rological tests were negative and a whole grafttransplant was performed. The patient had a verygood post-operative course andwas discharged byday 21with normal liver function tests (LFTs) andbilirubin values, with standard immunosuppres-sion consisting of cyclosporine, in a dose to main-tain trough concentrations of 200–250 mg/L in thefirst 2 months and 150–200 mg/L thereafter, andprednisolone at 2mg/kg daily for the first 5 daysafter the transplantation tapered to a medianmaintenance dose of 0.07mg/kg daily over the fol-lowing weeks. PCR–HCV was negative in the im-mediate post-operative period. Starting from thethird month, there were repeated episodes of LFTalterations with spontaneous resolution. At thattime, anti-rejection immunosuppression wasappropriate. All investigations were negative, theabdominal sonography was normal, and liver bi-opsy at the 15th month showed only a reactivepattern, suggestive of viral infection or drug tox-icity. Serological viral tests were negative, includ-ing a new PCR–HCV. In the 22ndmonth after thetransplantation, the patient developed a new epi-sode of LFT alterations associated with fever. Aliver biopsy sample showed a biliary pattern, withperipheral ductal proliferation and neutrophilicportal inflammatory cell infiltrate. In view of thesefindings, a percutaneous cholangiography wasperformed and showed a normal biliary tree.

After this, LFTs showed aspartate aminotrans-ferase and alanine aminotransferase fluctuationsuntil September 1999, when a large increase ofthese values occurred. A new liver biopsy sampleshowed moderately active chronic hepatitis pro-gressing to cirrhosis, portal lymphocyte aggre-gates, and a large number of plasma cells. Wecould not identify characteristic changes attribut-ed to acute or chronic rejection. The patient had ahistory of good compliance and serum cyclospor-ine concentrations remained adequate. Totalserum protein and albumin levels were normal,autoantibodies were positive (gastric parietal cellantibody, liver–kidney microsomal antibody1:160, anti-hepatic cytosol 1:320), and IgG levelsraised to 4447mg/dL (normal value, 800–1200mg/dL). At that time, autoantibody determi-nations had been tested in normal controls andwere considered positive if the titles were 1:80 orhigher. Based on these histological and serologicalfindings of autoimmune disease, a diagnosis of‘de novo autoimmune hepatitis’ was made (Fig. 1).Cyclosporine dose was initially reduced and

corticosteroids were reintroduced, with an initialdosage of 1mg/kg/day for a month, with gradualreduction to 20mg every other day. Mycopheno-late mofetil was added at an initial dosage of 1 g/day, and then increased to 2 g/day. After 4monthsof treatment, LFTs were normal, and IgG levelhad normalized; however, repeat liver biopsyrevealed cirrhosis, interface hepatitis, and intenseplasma cell infiltrate. After 19 months of treat-ment, a new liver-biopsy sample showed markedreduction of portal and lobular inflammatory in-filtrate, with regression of fibrosis and of the ar-chitectural disruption (Figs. 2 and 3). At that time,autoantibody determinations demonstrated thatliver–kidney microsomal antibody and anti-hepatic cytosol became negative. The main histo-logical features of liver-biopsy samples are shownin Table 1. It is important to note that the lastliver-biopsy sample showed inactive cirrhosisand disappearance of interface hepatitis and ofplasma cell infiltrate. During all the stages of thiscase, the clinical features of cirrhosis and portalhypertension like splenomegaly, leukopenia, orthrombocytopenia were not present.Currently, 9 yr after transplantation, the patient

is doing well, with normal liver function tests andno evidence of cirrhosis, although the presence offibrosis continued to be quite important in thelast liver-biopsy sample. It is unknown whetherthis fibrosis may be a cause for future graft loss.As all our children who underwent liver trans-plantation and were receiving cyclosporinehave been switched to tacrolimus, the present

372

Gibelli et al.

immunosuppressive therapy of this patient con-sists in tacrolimus, mycophenolate mofetil, andprednisolone. She has never received any biolog-ical immunosuppressant.

Discussion

De novo autoimmune hepatitis after liver trans-plantation, autoimmune hepatitis, and autoim-mune sclerosing cholangitis are recognized as the3 liver disorders in childhood in which liverdamage is likely to arise from an autoimmuneattack (10). Herein, we report a case of de novoautoimmune hepatitis after liver transplantationthat occurred in 0.5% (1/205) of children who un-derwent transplantation in our unit in the last16 yr. According to previous reports (6–8, 11, 12),graft dysfunction is characterized by increased

concentrations of serum aminotransferase, asso-ciated with the presence of autoantibodies, raisedIgG concentrations, chronic hepatitis on liver bi-opsy specimens, and good response to immuno-suppressive used to treat autoimmune hepatitis.Pathophysiologically, whether the liver damage

in this patient is a special form of rejection or aconsequence of an autoimmune injury, triggeredby drugs or viral infection, has yet to be estab-lished. Although our patient presented somehistological changes observed in severe acuterejection, the unusual prominent plasma cellcomponent within the inflammatory plasma-cellinfiltrate and the absence of other features of cel-lular rejection indicated the diagnosis of de novoautoimmune hepatitis. Different from other pub-lications, we could not demonstrate any previousviral infection as a possible mechanism leading to

Fig. 1. Pre-treatment liver biopsy. (A) Masson trichrome staining disclosing a cirrhotic liver with small nodules and thick septae.(B) Severe inflammation with intense interface hepatitis. (C) A large number of plasma cells is present.

373

Successful treatment of de novo autoimmune hepatitis and cirrhosis

the de novo autoimmunity to the liver. Salcedoet al. (11) reported that all their patients developedde novo autoimmune hepatitis in relation to infec-tion with cytomegalovirus, Epstein–Barr virus, orparvovirus. These infections, which are frequentafter liver transplantation, may lead to autoim-munity through mechanisms such as polyclonalstimulation, enhancement, and induction ofmem-brane expression of major histocompatibilitycomplex class I and II antigens and interferencewith immunoregulatory cells. Another interestingobservation based on an animal experimentalmodel of transplantation is that the use of calci-neurin inhibitors predisposes to autoimmunityand autoimmune disease, possibly by interferingwith the maturation of T lymphocytes or with thefunction of regulatory T cells, with consequentemergence and activation of autoreactive T-cellclones (13, 14).Similar to previous reports (6–12), our patient

presented the disorder 4 yr after the transplant.Other possible causes of graft dysfunction, mainlythe viral infections thatmay occur at the same timeafter the transplant, were carefully investigatedand excluded. Biliary complications were ruledout by percutaneous cholangiographic study, andvascular complications were excluded by findingsof normal blood flow across the hepatic artery,hepatic, and portal vessels.In our center, all liver-biopsy samples are re-

viewed by a team of 6 specialized pathologists, in

order to avoid errors of interpretation. The mostimportant differential diagnosis has to be madewith a form of acute or chronic rejection that is themost common cause of graft dysfunction afterliver transplantation. About 85% of our patientswho underwent transplantation had at least 1episode of acute reversible cellular rejection, and10% developed chronic rejection (3 of themrequired re-transplantation) (3). There are someimportant histological features of acute andchronic rejection distinguishable from de novoautoimmune hepatitis. First of all, the main his-tological change of acute rejection consists of adense mixed portal inflammatory infiltration ofthe biliary epithelium and endothelitis. This in-flammatory infiltrate contains activated lymph-ocytes and sometimes eosinophil cells can beidentified. The characteristic features of chronicrejection are loss of bile ducts and obliterative ar-teriopathy (15). The liver histological findings ofour patient were quite different. There was a denselymphocytic portal infiltrate with a predominanceof plasma cells, interface hepatitis, and porto-portal fibrosis progressing to cirrhosis, similar towhat is described in patients with autoimmunehepatitis (16). Finally, our patient was a female,according to the preponderance observed inclassical autoimmune hepatitis (17).There are some important differences between

the treatment ofde novoautoimmunehepatitis andcellular rejection in post-liver transplantation.

Fig. 2. Mid-treatment liver biopsy. (A) Cirrhotic liver, as shown in Masson trichrome staining. (B) A significant decrease in theinflammation and in the interface hepatitis.

374

Gibelli et al.

Although prednisone is effective in both autoim-munehepatitis and rejection, the dose andmodeoftreatment are different in the 2 disorders. Cellularrejection responds promptly to a short course of ahigh dose of intravenous prednisone, with nor-malization of liver function within a few days. Au-toimmune hepatitis responds to a lower dose ofprednisone, which needs to be gently tapered overseveral weeks to a dose at which normal liver func-tion can be maintained, a dose that is generallyhigher than that used after liver transplantation.

According to several reports, azathioprine isadded to the treatment as a steroid-sparing agent(6, 9, 11, 18). The unique feature about our patientis that she presented a good response to theaddition of mycophenolate mofetil, with inactiva-tion of cirrhosis, disappearance of inflammatoryinfiltrate (Fig. 3), and normalization of liverfunction tests.In conclusion, the present case demonstrates

that diagnosis of de novo autoimmune hepatitisafter liver transplantation has important clinical

Fig. 3. Post-treatment liver biopsy. (a) Note the thin septae at Masson trichrome staining. (b) Almost complete disappearance ofinflammation.

Table 1. Main biopsy findings

March 1997 October 1998 September 1999 September 2000 April 2001 July 2002

Diagnosis Mild reactivehepatitis

Biliary portalreaction

Chronic activehepatitis,pre-cirrhotic

Chronic activehepatitis, cirrhotic

Chronic activehepatitis, cirrhotic

Inactivecirrhosis

Lobularinflammation

None None Moderate Intense Mild None

Plasma cellinfiltration

None None Moderate Intense None None

Interfacehepatitis

Mild Mild Moderate Intense Mild None

Medicationtreatments

Cyclosporineprednisolone

Cyclosporine Cyclosporineprednisolonemycophenolate

Cyclosporineprednisolonemycophenolate

Cyclosporineprednisolonemycophenolate

Tacrolimusprednisolonemycophenolate

375

Successful treatment of de novo autoimmune hepatitis and cirrhosis

implications since its appropriate triple therapywith tacrolimus, mycophenolate mofetil, andprednisolone could save the graft and preventretransplantation.Addendum: Since the submission of this paper to

the Pediatric Transplantation, 1 more pediatricpatient was diagnosed as having de novo autoim-mune hepatitis. She is a 4-yr-old girl who under-went a living related donor liver transplantationfrom her father (segments I and II) 3 yr ago.The indication for transplant was extra hepaticbiliary atresia. In the last 2 months, she has beenpresenting LFT alterations with serum autoanti-bodies. A liver-biopsy sample showed a pattern ofchronic hepatitis, portal mononuclear infiltratewith predominance of plasma cells, and interfacehepatitis.

References

1. RYCKMAN FC, ALONSO MH, BUCUVALAS JC, BALISTRERI WF.

Long-term survival after liver transplantation. J Pediatr Surg1999: 34: 845–849.

2. PAPPO O, RAMOS H, STARZL TE, et al. Structural integrity and

identification of causes of liver allograft dysfunction occurring

more than 5 years after liver transplantation. Am J Surg Pathol1995: 19: 192–206.

3. TANNURI U, VELHOTE MC, SANTOS MM, et al. Pediatric liver

transplantation: fourteen years of experience at the children

Institute in Sao Paulo, Brazil. Transplant Proc 2004: 36:941–942.

4. SLAPAK GI, SAXENA R, PORTMANN B, et al. Graft and systemic

disease in long-term survivors of liver transplantation.Hepatology 1997: 25: 195–202.

5. TANNURI U, MELLO ES, CARNEVALE FC, et al. Hepatic venous

reconstruction in pediatric living-related donor liver trans-

plantation – experience of a single center. Pediatr Transplant2005: 9: 293–298.

6. KERKARN,HADZICN, DAVIES ET, et al. De-novo autoimmunehepatitis after liver transplantation. Lancet 1998: 351:

409–413.

7. PETZ W, SONZOGNI A, BERTANI A, et al. A cause of late graft

dysfunction after pediatric liver transplantation: de novoauto immune hepatitis. Transplant Proceed 2002: 34:

1958–1959.

8. HANEGAN MA, PORTMANN BC, NORRIS SM, et al. Graft

dysfunction mimicking autoimmune hepatitis following livertransplantation in adults. Hepatology 2001: 36: 464–470.

9. MIELI-VERGANIG,VERGANID.De-novo autoimmune hepatitis

after liver transplantation. J Hepatol 2004: 40: 3–7.10. MIELI-VERGANIG,VERGANID. Immunological liver diseases in

children. Semin Liver Dis 1998: 18: 271–279.

11. SALCEDOM,VAQUERO J, BANARESR, et al. Response to steroids

in de novo autoimmune hepatitis after liver transplantation.Hepatology 2002: 35: 349–356.

12. MIYAGAWA-HAYASHINO A, HAGA H, EGAWA H, et al. Outcome

and risk factors of de novo autoimmune hepatitis in living-

donor liver transplantation. Transplantation 2004: 78:128–135.

13. GAO E, LO D, CHENEY R, KANAGAWA O, SPRENT J. Abnormal

differentiation in thymocytes in mice treated with cyclosporineA. Nature 1988: 336: 176–179.

14. WU DY, GOLDSCHNEIDER I. Cyclosporine A-induced auto-

logous graft-versus-host disease: a prototypical model of

autoimmunity and active (dominant) tolerance coordinatelyinduced by recent thymic emigrants. J Immunol 1999: 162:

6926–6933.

15. GHOBRIAL RM, AMERSI F, MCDIARMID SV, BUSUTIL RW.

Pediatric liver transplantation. In MADDREY WC, SCHIFF ER,SORREL MF, eds. Transplantation of the Liver. Philadelphia:

Lippincott Williams &Wilkins, 2001: 79–99.

16. ALVAREZ F, BERG PA, BIANCHI FB, et al. InternationalAutoimmune Hepatitis Group Report: review of criteria for

diagnosis of autoimmune hepatitis. J Hepatol 1999: 31:

929–938.

17. MIELI-VERGANI G, VERGANI D. Autoimmune hepatitis.Minerva Gastroenterol Dietol 2004: 50: 113–123.

18. GUPTA P, HART J, MILLIS JM, et al. De novo hepatitis with au-

toimmune antibodies and atypical histology. Transplantation

2001: 71: 664–668.

376

Gibelli et al.