Successful Withdrawal of Prednisone After Adult Liver Transplantation for Autoimmune Hepatitis

Transcription

1 Successful Withdrawal of Prednisone After Adult Liver Transplantation for Autoimmune Hepatitis Thomas E. Trouillot,* Roshan Shrestha,* Igal Kam, Michael Wachs, and Gregory T. Everson* SEE EDITORIAL ON PAGE 460 Corticosteroid withdrawal after orthotopic liver transplantation (OLT) represents an attractive therapeutic option for ameliorating post-olt metabolic complications, although several reports suggest patients who undergo transplantation for autoimmune hepatitis (AIH) may have a greater incidence of acute and chronic rejection when withdrawn from corticosteroid therapy. The aim of this study is to evaluate the success of corticosteroid withdrawal in patients with AIH after OLT. Twenty-six patients underwent successful OLT for AIH. In 21 of these patients, stable maintenance immunosuppression consisted of cyclosporine (CSA) and prednisone (n 20) or tacrolimus (TAC) and prednisone (n 1). In this group, a trial of prednisone withdrawal was initiated when patients were 6 months or more post-olt, with normal liver function, and receiving an average prednisone dosage of 10 mg/d. Five additional patients treated with either TAC (n 4) or CSA (n 1) plus mycophenolate mofetil underwent a 14-day taper of prednisone. Overall, 17 of 25 patients (68%) were successfully withdrawn from corticosteroids, with a mean follow-up of 22 months (range, 1 to 34 months). Of the remaining 8 patients, 5 patients received a lower dosage of prednisone or required prednisone to control inflammatory bowel disease. Only 3 patients remained dependent on prednisone to maintain stable liver allograft function. Withdrawal from 10 to 5 mg of prednisone (n 21) resulted in four episodes of steroid-responsive and two episodes of steroid-resistant rejection in 3 patients, and 18 of 21 patients (86%) were rejection free. Withdrawal from 5 to 0 mg prednisone (n 17) resulted in eight episodes of steroid-responsive and no episodes of steroid-resistant rejection in 4 patients; 13 of 17 patients (76%) were rejection free. Of the 5 patients in the 14-day prednisonetaper group, 3 patients had steroid-responsive rejection and 1 patient required OKT3. Seventeen of 21 patients (81%) with AIH were successfully withdrawn from corticosteroids. It is notable that corticosteroid withdrawal was associated with a reduction in serum cholesterol levels, decreased use of antihypertensive agents, and reduced need for insulin or oral hypoglycemic agents. We propose corticosteroid withdrawal should be attempted in patients with underlying AIH who undergo OLT because most will benefit without significantly jeopardizing the liver allograft. Copyright 1999 by the American Association for the Study of Liver Diseases T he benefit-risk ratio of corticosteroids as immunosuppressive therapy after orthotopic liver transplantation (OLT) has created recent debate because their deleterious side effects have become more apparent with long-term follow-up. Cerebrocardiovascular events are a major cause of longterm morbidity and mortality in these patients, and the contribution of immunosuppression to posttransplantation diabetes mellitus, obesity, hypertension, and dyslipidemia is well established. 1-5 Munoz et al 6 noted an increased cardiovascular risk index in long-term survivors after OLT that was attributed to obesity and hypercholesterolemia. We previously reported improvement in glucose control, hypertension, and hypercholesterolemia in patients undergoing corticosteroid withdrawal greater than 1 year after OLT. 7,8 Patients who have undergone OLT for autoimmune hepatitis (AIH) have been characterized as difficult to wean from corticosteroids and at risk for recurrent AIH in the liver allograft. Viral hepatitis and the autoimmune-mediated liver diseases, such From the Departments of *Medicine and Surgery, University of Colorado School of Medicine, Denver, CO. Presented in part at the 1997 meeting of The American Society of Transplant Physicians, Chicago, IL, and the 1997 Fourth Congress of the International Liver Transplantation Society, Seattle, WA. Address reprint requests to Thomas E. Trouillot, MD, University of Colorado Health Sciences Center, 4200 East Ninth Ave, Campus Box B154, Denver, CO Copyright 1999 by the American Association for the Study of Liver Diseases /99/ $3.00/0 Liver Transplantation and Surgery, Vol 5, No 5 (September), 1999: pp

2 376 Trouillot et al as primary biliary cirrhosis, primary sclerosing cholangitis, and AIH, have been reported to recur in the donor liver, resulting in additional long-term liver graft failure. 9,10 In particular, AIH has been associated with a greater rate of acute and chronic ductopenic rejection after OLT. 11,12 Two reports suggest an increased frequency of acute cellular rejection (ACR) and OKT3 use for steroid-resistant rejection in patients undergoing OLT for AIH compared with non autoimmune-related liver disease. 12,13 Recurrent AIH after OLT has been correlated with HLA A1, A3, B8, and DR3 haplotypes. 14,15 A randomized prospective study of corticosteroid withdrawal after OLT excluded patients with AIH because they noted rejection was often severe and disease recurrence may result in late graft dysfunction. 15,16 In our experience, we did not notice a significant number of adult patients with AIH who had severe rejection or late graft dysfunction; therefore, it was justified to attempt corticosteroid withdrawal in this cohort. Our aim is to determine whether corticosteroid withdrawal could be successfully achieved in patients with AIH after OLT and assess the metabolic benefits and complications in these patients. Patients and Methods Patient Cohort From December 1988 to October 1996, 347 consecutive adult OLTs were performed in 314 patients at the University of Colorado Health Sciences Center (Denver, CO). Of these, 26 patients with AIH underwent OLT with greater than a 6-month follow-up, including 21 women and 5 men, with a mean age of 49 years (range, 23 to 66 years). Each patient was evaluated according to the International Autoimmune Hepatitis Group proposed scoring system for the diagnosis of AIH. 17 Scores from 15 and 11 patients qualified as definite and probable diagnosis for AIH, respectively. HLA typing was performed in 17 of 26 patients as part of the pre-olt evaluation, using a complement-mediated cytotoxicity assay, with the reactions read by a two-color fluorescence method. Immunosuppression Immunosuppressive drug regimens varied according to protocols that evolved since the inception of our liver transplant program in Induction immunosuppression consisted of cyclosporine A (CSA), azathioprine, and prednisone in 19 patients. One patient received induction with OKT3 (Orthoclon; Ortho Biotech, Raritan, NJ) and another with tacrolimus (TAC; FK506; Prograf; Fujisawa, Deerfield, IL) and prednisone. Maintenance immunosuppression for the cohort was similar (CSA and prednisone, n 20; TAC and prednisone, n 1). Corticosteroid withdrawal was initiated in patients with normal liver transaminase levels and no clinical evidence of allograft dysfunction who were greater than 6 months after OLT. Prednisone withdrawal commenced at a dose of 10 mg/d and was reduced by 2.5-mg increments over several months while monitoring liver transaminase levels. Five additional patients were enrolled onto a 14-day rapid prednisone taper protocol, treated with mycophenolate mofetil for 6 months, and maintained on TAC (n 4) or CSA (n 1). CSA and TAC whole-blood trough levels in the first 3 months were targeted for 350 to 400 pg/ml and 10 to 15 ng/ml and subsequently reduced to 200 to 300 pg/ml and 5 to 12 ng/ml, respectively. Persistently elevated liver transaminase levels prompted a percutaneous liver biopsy and intravenous steroid boluses for histological evidence of ACR, using established criteria. 18 Steroid-resistant rejection was defined as persistently elevated liver transaminase levels after bolus corticosteroid therapy for biopsyproven ACR and treated with a course of OKT3. Metabolic Parameters Metabolic outcome variables included serum creatinine level and cardiovascular risk factors, such as serum cholesterol level, need for antihypertensive medication, and need for oral hypoglycemic agents or insulin. Serum cholesterol, creatinine, and CSA and/or TAC levels were the average of four to six values obtained during each prednisone dosage interval; 10, 5, and 0 mg/d, respectively. Results Feasibility of Prednisone Withdrawal During the course of the study, 2 patients died; 1 of bilobar pneumonia with sepsis before the withdrawal of prednisone, and the other of an acute myocardial infarction 1 month after prednisone withdrawal. Seventeen of 25 patients (68%) were successfully withdrawn from corticosteroids and remain off them for a mean follow-up of 22 months (range, 1 to 34 months). These include 4 of 5 patients who received the 14-day rapid corticosteroid taper protocol. Five of 25 patients (20%) remained on prednisone therapy for other clinical indications or were undergoing a dose taper, whereas 3 of 25 patients (12%) were considered resistant to complete withdrawal. The presence of clinically significant inflammatory bowel disease (IBD) correlated with the inability to withdraw prednisone (Chi-squared, 9.560;

3 Prednisone Withdrawal After Liver Transplantation 377 Table 1. Prevalence of IBD in the Cohort Who Underwent OLT for AIH Was 24% (6 of 25 patients), Whereas the Prevalence of IBD in Patients Requiring Prednisone Was 63% (5 of 8 patients) Successful Prednisone Withdrawal Achieved Not Achieved Inflammatory bowel disease Present 1 5 Absent 16 3 NOTE. Chi-squared analysis was (P.002). Successful prednisone withdrawal was achieved in a majority of patients, including one patient with IBD. Patients who remained on prednisone had clinical indications because of underlying IBD or required it to prevent ACR. Abbreviations: IBD, inflammatory bowel disease; OLT, orthotopic liver transplantation; AIH, autoimmune hepatitis; ACR, acute cellular rejection. P.002), although this analysis was limited by a small sample size (Table 1). Of the 17 patients who were successfully withdrawn from prednisone, only one had IBD. Five of the 8 patients who were not withdrawn from prednisone had IBD. These patients required prednisone to control their colitis or liver disease. HLA typing showed that 8 of 17 patients (47%) were positive for either HLA B8 or DR3. The 3 patients considered resistant to corticosteroid withdrawal were positive for HLA DR52, whereas 2 of 3 patients were positive for HLA B8. Rejection Episodes Eighteen patients (86%) were free of rejection after prednisone withdrawal from 10 to 5 mg/d (n 21; Fig. 1). In the remaining 3 patients, there were four episodes of steroid-responsive and two episodes of steroid-resistant rejection: (1) 1 patient had three episodes of ACR while receiving 5 to 10 mg/d of prednisone requiring two courses of intravenous corticosteroid boluses and one 14-day course of OKT3; (2) 1 patient had two episodes of ACR while receiving 5 to 10 mg/d of prednisone requiring one course of intravenous corticosteroid boluses and one 10-day course of OKT3; (3) 1 patient had one episode of ACR requiring one course of intravenous corticosteroid boluses. Both patients who received OKT3 could not be weaned off corticosteroids, and they both required corticosteroids to treat IBD. The third patient is noncompliant, inaccessible outside of the United States, and has remained on maintenance prednisone therapy at 5 mg/d. Thirteen of 17 patients (76%) weaned off prednisone from a dose of 5 mg/d remain rejection free. Eight episodes of ACR occurred in the remaining 4 patients. These were steroid responsive and did not require OKT3. One patient restarted prednisone therapy after weaning, and 2 patients were converted from CSA to TAC for recurrent rejection. Three of 5 patients who underwent the 14-day rapid-taper prednisone protocol had steroid-responsive rejection, and 1 patient required OKT3. There were no complications as a result of OKT3 therapy Figure 1. Rejection episodes during prednisone withdrawal from 10 to 5 mg/day; 3 patients had four episodes of steroidresponsive and two episodes of steroid-resistant acute cellular rejection that required OKT3. Eighteen of 21 patients had no clinical evidence of acute cellular rejection. On complete prednisone withdrawal, 4 patients had eight rejection episodes that were all steroid responsive, and 13 of 17 patients were rejection free.

4 378 Trouillot et al in those patients who required antilymphocyte antibody therapy for steroid-resistant ACR. Metabolic Response to Prednisone Withdrawal To assess whether prednisone withdrawal was metabolically significant in these patients, several parameters were measured. Mean serum cholesterol levels decreased during corticosteroid withdrawal. Reduction in prednisone from 10 to 5 mg/d resulted in a downward trend in serum cholesterol levels (197 to 189 mg/dl; P not significant) that approached statistical significance on complete withdrawal (181 mg/dl; P.06; Fig. 2) Overall, there was a 10% reduction in serum cholesterol levels after prednisone withdrawal. Serum creatinine levels increased from 1.2 to 1.5 mg/dl (P.001) during prednisone withdrawal from 10 to 5 mg/d. There was no significant change in mean serum creatinine levels (1.6 mg/dl) on complete withdrawal of prednisone from 5 mg/d (Fig. 3). Based on the immunosuppression protocol, mean CSA dosages and serum levels were reduced from mg and pg/ml to mg (P.001) and pg/ml (P not significant), respectively, during prednisone withdrawal from 10 to 5 mg/d. After complete withdrawal of prednisone from 5 mg/d, mean CSA dosages and serum levels were further reduced to Figure 3. Mean serum creatinine values SE are noted in patients with autoimmune hepatitis after liver transplantation who underwent successful prednisone withdrawal. On tapering prednisone from 10 to 5 mg/d, there was a significant increase in mean serum creatinine level (P F.001, twotailed paired t-test) that remained after complete prednisone withdrawal. Subjects in the rapidtaper protocol were excluded from this analysis mg (P.01) and pg/ml (P.001), respectively. Glucose and Blood Pressure Control Risk factors for coronary artery disease were further reduced as evidenced by improvement in blood pressure and glucose control with prednisone withdrawal. Seven of 18 patients (39%) stopped or decreased the dosage of their antihypertensive medication on reduction in prednisone dosage from 10 to 5 mg/d. On complete withdrawal of prednisone, 1 of 10 patients decreased the dosage of their antihypertensive medication. Two of 3 patients with diabetes discontinued insulin or hypoglycemic agent after prednisone dosage reduction from 10 to 5 mg/d. Figure 2. Mean serum cholesterol values SE are noted in patients with autoimmune hepatitis after liver transplantation who underwent successful prednisone withdrawal. There was a downward trend in mean serum cholesterol level with prednisone withdrawal that approached statistical significance (P.06, two-tailed paired t-test). Subjects in the rapid-taper protocol were excluded from this analysis. (NS, not significant.) Discussion Our findings indicate the majority of patients with AIH who undergo OLT can be successfully weaned off corticosteroid therapy. We observed no rejection episodes in 86% and 76% of the patients on tapering prednisone therapy from 10 to 5 mg/d and 5 mg/d to off, respectively. Chronic rejection or recurrent AIH were not clinically evident in the follow-up period. The incidence of ACR during corticosteroid withdrawal (33%) was somewhat greater than our experience with corticosteroid

5 Prednisone Withdrawal After Liver Transplantation 379 withdrawal in patients who underwent transplantation for other conditions (14%) 8 or compared with other published data (4% to 27%). 5,16,19-22 In our study, most episodes of ACR were steroid responsive and there was no immunologic graft loss. Although the analysis of HLA typing was limited because of a small sample size, the apparent association of HLA B8 and DR3 in patients unable to tolerate prednisone withdrawal was of interest. Although specific alleles encompassing DR3 and DR4 were not determined in our cohort, Czaja et al 23 showed specific alleles are associated with concurrent immune diseases, including IBD, and a poor treatment response to corticosteroid therapy before OLT. Moreover, Testa et al 14 showed HLA haplotypes A1, A3, B8, and DR3 are associated with recurrent AIH in the liver allograft. Results from our study suggest post-olt prednisone dependence may be related in part to the presence of specific HLA alleles (B8 and DR52) associated with severe AIH or IBD. Furthermore, Wright et al 15 showed an association with recurrent AIH in liver transplant recipients who are HLA DR3( ) who receive an HLA DR3( ) liver allograft, although HLA B8 status did not affect disease recurrence. Together, these data from several centers support a relationship between specific HLA alleles and the phenotypic expression of AIH after OLT. We speculate HLA phenotype may facilitate the identification of post-olt patients who remain steroid dependent. A small subset of our AIH patients developed transaminasemia and histological evidence of ACR when corticosteroids were withdrawn. CSA dosage and serum levels were significantly reduced during corticosteroid withdrawal; therefore, the observed elevation in serum creatinine level during steroid withdrawal was not caused by a compensatory increase in the calcineurin inhibitor dose or level. The increase in serum creatinine level after corticosteroid withdrawal was likely caused by the cumulative effect of long-term maintenance immunosuppression with CSA or TAC, and its clinical significance needs to be further elucidated. Although we did not identify patients with histological evidence of recurrent AIH, Sempoux et al 24 described a patient with recurrent AIH based on serological and histological data who responded to resumption of corticosteroid therapy. Another group reported only a minority of patients who underwent transplantation for AIH had normal liver histopathologic results and that immunosuppressive drug withdrawal without protocol biopsies may miss significant disease. 9 Ahmed et al 25 reported a 61% incidence of unexplained chronic hepatitis in 33 patients who underwent transplantation for AIH, although 55% were associated with transaminitis. Although these data identified a significant incidence of abnormal histopathologic results in patients who underwent transplantation for AIH, the long-term outcome of these liver allografts is acceptable. 9,12,25 Because liver biopsy specimens are not routinely obtained in our program in patients with normal liver transaminase levels and liver allograft function, we may underestimate the incidence of chronic hepatitis. Further studies are needed to determine whether a steroidsparing immunosuppression regimen will affect the incidence of recurrent AIH or chronic cellular rejection compared with maintenance immunosuppression with prednisone. Overall, our findings indicate even small amounts of prednisone in the 5- to 10-mg range are metabolically significant, and that prednisone withdrawal has salutary metabolic effects. One of the most striking findings of this study was the observation that risk factors for cerebrocardiovascular events improve with corticosteroid withdrawal, including a reduction in serum cholesterol level and the use of antihypertensive and hypoglycemic medications. This is consistent with a previous report that showed a marginal benefit of corticosteroid withdrawal and the adverse effects attributed to long-term corticosteroid therapy. 16 They showed a downward trend in serum cholesterol levels in the corticosteroid withdrawal group compared with baseline values, although serum cholesterol, lowdensity lipoprotein, high-density lipoprotein, and triglyceride values were not statistically different compared with controls who remained on corticosteroid maintenance therapy. The dyslipidemia data after organ transplantation is for the most part flawed by the use of maintenance corticosteroids that affect serum lipid levels. There are some data from our program, 7,8 as well as mixed reports primarily in the pediatric literature, that show a decrease in serum cholesterol levels and improved growth curves with TAC or CSA monotherapy after corticosteroid withdrawal. 5,16,19,21,22 Belli et al 5 reported a significant reduction in the prevalence of hypertension, diabetes, and side effects of corticosteroid therapy in a prospective, randomized trial of early corticosteroid withdrawal after OLT. Our study confirms these data that risk factors for

6 380 Trouillot et al cerebrocardiovascular events are reduced after corticosteroid withdrawal without immunologic graft loss. We speculate the advantages of avoiding longterm complications from corticosteroid therapy will be more beneficial than the disadvantages of a monotherapy-based immunosuppression regimen in patients with AIH; the benefit of steroid-free long-term maintenance immunosuppression outweighs the short-term risk of treatable ACR. The optimal long-term immunosuppressive regimen for patients who undergo transplantation for AIH remains to be clarified. Contrary to clinical dogma, we report the successful withdrawal of corticosteroids in patients with AIH after OLT. References 1. Asfar S, Metrakos P, Fryer J, Verran D, Ghent C, Grant D, et al. An analysis of late deaths after liver transplantation. Transplantation 1996;61: Iwatsuki S, Starzl TE, Gordon RD, Esquivel CO, Todo S, Tzakis AG, et al. Late mortality and morbidity after liver transplantation. Transplant Proc 1987;19: Abbasoglu O, Levy M, Brkic BB, Testa G, Jeyarajah DR, Goldstein RM, et al. Ten years of liver transplantation An evolving understanding of late graft loss. Transplantation 1997;64: Padbury RTA, Toogood GJ, McMaster P. Withdrawal of immunosuppression in liver allograft recipients. Liver Transpl Surg 1998;4: Belli LS, de Carlis L, Rondinara G, Alberti AB, Bellati G, de Gasperi A, et al. Early cyclosporine monotherapy in liver transplantation: A 5-year follow-up of a prospective, randomized trial. Hepatology 1998;27: Munoz SJ, Deems RO, Moritz MJ, Martin P, Jarrell BE, Maddrey WC. Hyperlipidemia and obesity after orthotopic liver transplantation. Transplant Proc 1991;23: Stegall MD, Everson G, Schroter G, Bilir B, Karrer F, Kam I. Metabolic complications after liver transplantation Diabetes, hypercholesterolemia, hypertension, and obesity. Transplantation 1995;60: Stegall MD, Everson GT, Schroter G, Karrer F, Bilir B, Steinberg T, et al. Prednisone withdrawal late after adult liver transplantation reduces diabetes, hypertension, and hypercholesterolemia without causing graft loss. Hepatology 1997;25: Slapak GI, Saxena R, Portmann B, Gane E, Devlin J, Calne SR, et al. Graft and systemic disease in long-term survivors of liver transplantation. Hepatology 1997;25: Neuberger J. Recurrence of primary biliary cirrhosis, primary sclerosing cholangitis, and autoimmune hepatitis. Liver Transpl Surg 1995;1: Candinas D, Gunson B, Robson S, Hubscher S, McMaster P, Neuberger J. Risk factors for chronic rejection of liver allografts[abstr]. Hepatology 1995;22:142A. 12. Hayashi M, Keefe EB, Krams SM, Martinez OM, Ojogho ON, So SKS, et al. Allograft rejection after liver transplantation for autoimmune liver diseases. Liver Transpl Surg 1998;4: Herrera J, Martin P, Shackleton C, Goldstein L, Shaked A, Busuttil R. Acute cellular rejection of liver allografts is increased in frequency in patients with autoimmune idiopathic chronic active hepatitis (AICAH)[abstr]. Hepatology 1995;22:142A. 14. Testa G, Crippin J, Abbasoglu O, Goldstein RM, Husberg BS, Jeyarajah DR, et al. Does recurrent autoimmune hepatitis after liver transplant have an impact on patient and graft survival[abstr]? ASTP Sixteenth Annual Meeting 1997: Wright HL, Bou-Abboud CF, Hassanein T, Block GD, Demetris AJ, Starzl TE, et al. Disease recurrence and rejection following liver transplantation for autoimmune chronic active liver disease. Transplantation 1992;53: McDiarmid SV, Farmer DA, Goldstein LI, Martin P, Vargas JH, Tipton JR, et al. A randomized prospective trial of steroid withdrawal after liver transplantation. Transplantation 1995;60: Czaja AJ, Carpenter HA. Validation of scoring system for diagnosis of autoimmune hepatitis. Dig Dis Sci 1996;41: Snover DC, Sibley RK, Freese DK, Sharp HL, Bloomer JR, Najarian JS, et al. Orthotopic liver transplantation: A pathological study of 63 serial liver biopsies from 17 patients with special reference to the diagnostic features and natural history of rejection. Hepatology 1984; 4: Dunn SP, Falkenstein K, Lawrence JP, Meyers R, Vinocur CD, Billmire DF, et al. Monotherapy with cyclosporine for chronic immunosuppression in pediatric liver transplant recipients. Transplantation 1994;57: Padbury RTA, Gunson BK, Dousset B, Hubscher SG, Buckels JAC, Neuberger JM, et al. Steroid withdrawal from long-term immunosuppression in liver allograft recipients. Transplantation 1993;55: Margarit C, Martinez Ibanez V, Tormo R, Infante D, Iglesias H. Maintenance immunosuppression without steroids in pediatric liver transplantation. Transplant Proc 1989;21: Andrews WS, Shimaoka S, Sommerauer J, Moore P, Hudgins P. Steroid withdrawal after pediatric liver transplantation. Transplant Proc 1994;26: Czaja AJ, Strettell MDJ, Thomson LJ, Santrach PJ, Moore SB, Donaldson PT, et al. Associations between alleles of the major histocompatibility complex and type 1 autoimmune hepatitis. Hepatology 1997;25: Sempoux C, Horsmans Y, Lerut J, Rahier J, Geubel A. Acute lobular hepatitis as the first manifestation of recurrent autoimmune hepatitis after orthotopic liver transplantation. Liver 1997;17: Ahmed M, Mutimer D, Hathaway M, Hubscher S, McMaster P, Elias E. Liver transplantation for autoimmune hepatitis: A 12-year experience. Transplant Proc 1997;29:496.