I topic liver transplantation (OLT) to avoid organ

ORIGINAL ARTICLES Long-Term Immunosuppression Without Corticosteroids After Orthotopic Liver Transplantation: A Positive Therapeutic Aim Gerald M. Fraser, * Kons tantinos Grammous tianos, Jayendravandan Reddy, Keith RolZes, Brian Davidson, and Andrew K. Burroughs Long-term treatment with corticosteroids after orthotopic liver transplantation (OLT) may cause adverse effects, particularly hypertension, diabetes, and bone disease. The results of steroid withdrawal from long-term immunosuppression in patients after OLT was reviewed. Initial treatment was with corticosteroids, azathioprine, and cyclosporine A in 76.3% and with antithymocyte globulin in 7.5%. Corticosteroids were stopped in 96 patients (8.2%) during mean follow-up of 6.7 2 3.9 months, and acute rejection subsequently developed in 8. By comparison 7 of 8 patients, in whom corticosteroids were continued, developed acute rejection. Six of these had received blood group (ABO)-compatible nonidenti- cal grafts. Rates for retransplantation in the steroid withdrawal and nonwithdrawal groups were.2% and 22.2%, respectively, and mortality in the two groups was.6% and.%, respectively. Azathioprine was not given or withdrawn in 28 patients in the group from which corticosteroids were also withdrawn, with no adverse effect. Diabetes mellitus improved following corticosteroid withdrawal, but there was no improvement in hypertension. We conclude that corticosteroids can be safely withdrawn in the majority of patients after OLT. Copyright o 7 996 by the American Association for the Study of Liver Diseases mmunosuppressive therapy is required after ortho- I topic liver transplantation (OLT) to avoid organ rejection. Since the first transplants were performed there has been considerable evolution in the pharmacology of immunosuppressive therapy, especially since the introduction of cyclosporine A and more recently, tacrolimus. Corticosteroids, in combination with cyclosporine and azathioprine, are still used for induction and initial maintenance therapy in most liver transplant centers.2 Subsequent tapering of immunosuppression is guided by a number of factors, which include maintenance of adequate immunosuppression on the one hand and avoidance of toxic side effects and exposure to infections or reinfection on the other. Corticosteroids are well known to have detrimental effects on glucose tolerance, osteoporosis, and hypertension. In children avoidance of long-term corticosteroid treatment has beneficial effects on gr~wth.~ Very few studies have assessed steroid withdrawal after OLT in adults. In a recent study of 77 liver transplantation patients surviving longer than 3 months, it was indicated that corticosteroids can be completely withdrawn in most patients and that a substantial proportion can be maintained on cyclosporine m~notherapy.~ As a result, complications normally associated with corticosteroid therapy were low, and there did not appear to be an increase in acute or long-term rejection or graft loss. However, in this study, the follow-up details of patients who continued treatment with corticosteroids is not described. In a small prospective randomized study in 66 patients comparing corticosteroid maintenance with withdrawal, similar conclusions were reached. However, 79% of the patients had either long-term hepatitis B or C.* By contrast, it has been suggested that azathioprine withdrawal increases the risk of long-term rejection.6 In renal transplant patients, quality-of-life scales were better in patients maintained without corticosteroids. In the present study we reviewed immunosuppressive therapy given to our patients undergoing OLT. In our program we endeavor to taper corticosteroid From the Liver Transplantation and Hepatobiliary Medicine Department, Royal Free Hospital and Medical School, London, UK. *Recipient of a Lewis Foundation Scholarship. Present address: Department of Gastroenterology, Beilinson Medical Center, Petach- Tihvah, Israel 9. Address reprint requests to Dr. A.K. Burroughs, Liver Transplantation and Hepatobiliary Medicine, Royal Free Hospital, Pond Street, London, NW3 2QG, UK. Copyright 996 by the American Association for the Study of Liver Diseases 7-322/96/26-$3./ Liver Tranrphntation and Surgery, Vol2, No 6 (Navember), 996: pp - 7

~ 2 Fvasev et al therapy and stop by 3 months. We were particularly interested to assess the outcome of this policy on organ rejection and steroid-induced side effects. Methods Study population. All patients who underwent their first OLT at the Royal Free Hospital in London from October 988 to June 99 constitute the study population. Of these 29 transplant recipients the following patients were excluded: - Fifty-seven patients followed for 3 months or less because they died or were retransplanted within that period of time. - Twenty-two patients for whom follow-up details were not available, such as patients followed in other countries. - Eight patients who never received corticosteroids and who received induction therapy with antithymocyte globulin. - Four patients not treated with long-term cyclosporine A, in 3 because of neurotoxic side effects. - Four patients treated from the outset with tacrolimus. The hospital charts and previously collected database of the remaining transplant patients were reviewed and abstracted, and details were recorded on standardized forms. Items recorded were demographic details, etiology of liver disease, ABO groups of recipient and donor, initial treatment regimen, rejection episodes and their management, infectious complications, diabetes mellitus, hypertension, presence of abnormal liver rests, severity of liver biopsy findings, dose of cyclosporine A per kilogram at last follow-up, withdrawal of azathioprine, death, and cause of death. Immunosuppression Protocol. Immunosuppression was induced with a bolus of g methylprednisolone intraoperatively, followed by a daily dose of.8 mg/kg intravenously. Azathioprine was given at a daily dose of.5 mg/kg. Cyclosporine therapy was introduced between the first and third days, mg/kg intravenously (IV), divided into two daily doses. In a group of 2 patients ATG was used for induction at a dose of 2.5 mg/kg IV for days. Methylprednisolone,.8 mg/kg, was given from day, and cyclosporine was introduced as described. Corticosteroid Tnerapy. All patients were entered into a protocol in which the aim was to stop corticosteroid therapy by 3 months after transplantation. This entailed reducing the daily dose by 5 mg every to 2 weeks. However, the clinical status of the patient was the ultimate determinant for adjustment of corticosteroid dosage. Cyclosporine A. Oral cyclosporine, mg/kg/day in two daily doses, was introduced when absorption was ensured, and dosage was adjusted according to blood levels. Blood levels of cyclosporine A were trough levels measured in plasma by high performance liquid chromatography. The therapeutic range was defined as 5 to 5 ng/ml. Follow-up. Patients were followed up depending on their clinical status. The protocol usually involved weekly visits for month, followed by visits at 2-week intervals for a further month, then monthly for 2 months, and subsequently every three months. Liver biopsies were routinely performed at 5 days, days, and year. Additional biopsies were performed according to clinical necessity. The follow-up period was determined from the date of transplantation to the last follow-up prior to December 3, 99, or patient death or retransplantation. Definitions Rejection episodes were defined as acute or chronic according to the presence of appropriate clinical, biochemical, and histological findings. Acute rejection was defined as the presence of a mixed portal infiltrate, predominantly lymphocytes with eosinophils and neutrophils; clearly defined bile duct damage; endothelitis of portal or terminal hepatic vein branches in liver biopsy specimens; and biochemical evidence of liver damage.8 Chronic rejection was defined as foam cell obliterative artenopathy or convincing evidence of bile duct loss with cytokeratin staining (> 5% of the triads) and evidence of liver dysfunction. Cytomegalovirus disease was regarded as proven when there was a clinical picture compatible with this infection, together with confirmation by culture of the virus or detection of the antigen in biopsy specimens or bronchial washings or a positive polymerase chain reaction for CMV-DNA. Pneumoqstis carinii pneumonia was diagnosed from the clinical picture supported by the results of bronchiolar lavage and transbronchial biopsy specimen. Bacterial or Fungal Infection. For this study we evaluated use of antibacterials and antifungals for proven or presumed infection as a marker of infection, even if subsequently no infection was proven by clinical events. Diabetes Mellitus and Hypertension. Patients were defined as diabetic if plasma venous blood glucose levels were greater than. mmol/l on two or more occasions or if the patient was being actively treated for diabetes. Patients were defined as hypertensive if diastolic blood pressure was greater than 95 mm Hg on two or more occasions or if the patient was being actively treated with antihypertensive mediations. Abnormal liver tests were defined as those in which the transaminases or bilirubin were more than twice normal and/or the alkaline phosphatase was more than.5 times normal. Abnormalities in liver biopsy specimens were graded as normal (, with minimal nonspecific changes (I), and with a definite abnormality (2). Results Data were available for patients who had survived more than 3 months after their first liver transplant (Table ). The mean follow-up time for these patients was 2.3? 8. months (mean? SD), with 75% being foilowed for at least months and 5% for least 8 months. Subgroups The patients were divided into two major groups; those in whom corticosteroids were stopped (96 patients, 8.2%) and those in whom corticosteroids

Immunosuppression Without Corticosteroids 3 Table. Details of Patients Undergoing OLT at the Royal Free Hospital October 988-June 99 Who Survived More Than 3 Months Total number Indications HCV PBC PSC Alcohol-related cirrhosis Cryptogenic cirrhosis FHF Other Follow-up (mean f SD) 75% 5% 26 (23%) 23 (2%) 8 (6%) 3 (%) 9 (8%) 3 (3%) 22 ( 9%) 2.3? 8. months months 8 months Abbreviations: HCV, hepatitis C virus; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; FHF, fulminant hepatic failure. were continued (8 patients, 5.8%; Table 2). Patients who continued corticosteroids were followed for 27. k 8.5 months, which was considerably longer than the 8.3 -t. months for those who were still taking them ( p <.OOl). In 76.3% of cases the initial immunosuppression regimen after OLT was a combination of cyclosporine A ( mg/kg IV daily), azathioprine (.5 mg/kg IV daily), and corticosteroids (methylprednisolone.8 mg/kg daily). There was, however, a period in which Table 2. Initial Immunosuppression Therapy and Outcome After OLT in Patients for Whom Corticosteroids Were Stopped or Continued Corticosteroids Corticosteroids Stopped Continued Patients 96 (8%) 8 (6%) FoIIOW-UP 27 f 9 8+ (mean t SD) months months* Initial treatment CyA, Aza, Pred 77 (8%) (78%) Aza, Pred (%) (6%) Pred, CyA (%) (OYO) ATG 7 (8%) 3 (6%) Outcome Chronic rejection 3 (3%) 3 (7%) Retransplanted (%) (22%)* Death (%) 8 (%)* Abbreviations: CyA, cyclosporine A; Aza, azathioprine; Pred, prednisone; ATG, antithymocyte globulin. *Statistically significant. we used ATG (2.5 mg/kg daily for days). Altogether, 2 patients received this therapy, and in 7 steroids were withdrawn, 5 without further episodes of rejection. Corticosteroid Withdrawal and Acute Cellular Rejection Within each major subgroup the patients were divided into those who did or did not experience episodes of acute cellular rejection more than month after transplantation. Details of the patients, including length of follow-up and duration of corticosteroid treatment in each group, are shown in Table 3. Acute cellular rejection was diagnosed in 8 of 96 (8.3%) patients after steroid withdrawal, despite adequate dose and blood levels of cyclosponne. Thus, 77.2% of the entire study population stopped steroids without subsequently developing acute cellular rejection. However, in 7 of 8 (38.9%) patients who were still being treated with steroids, there was evidence of acute rejection month or more after OLT. This latter group received the highest number of -g bolus doses of methylprednisolone and received more days of Om3 therapy than the other groups (Table ). Corticosteroid Withdrawal, Retransplantation, Mortality, and Prognostic Factors Retransplantation was required in 8 patients (7%): 6 for chronic rejection, following a complicated course after hepatobiliary reconstruction, and for severe cholestasis of unknown cause. Four of these patients subsequently died. In patients in whom corticosteroids were stopped, of 96 (.2%) required retransplantation compared with of 8 patients (22.2%) in whom they were continued. In addition to the patients who died after retransplantation, an additional 8 patients died, so that overall mortality was 9.3%. The combined mortality in patients for whom corticosteroid withdrawal was associated with rejection or in whom corticosteroids were continued was 6.2%, compared with.% in patients in whom they were withdrawn without rejection. Malignancy accounted for six patient deaths over all groups. We examined patient files to look for factors associated with patient prognosis. As expected, abnormal liver tests and biopsy specimens were more frequent in the patient groups with higher mortality rates. In patients in whom corticosteroids were withdrawn without subsequent rejection, 38 of 8

Fraser et al Tsble 3. Details of Patients According to Corticosteroid Treatment and Development d Acute Cellular Rejectlon Cellular Rejection Cellular Rejection Cellular Rejection Cellular Rejection Absent (n = 88) Present (n = 8) Absent (n = ) Present (n = 7) Age.9 2 2.3.6 f 6.8 9.3 t 2.5 36.?. Sex (MIF) 9/39 35 83 25 Duration of follow-up (months) 27.6 t 8.7 9.7? 8.8 8. f. 8.7 2. Steroid treatment (months) 6.7? 3.9 5.6 5.9 8. f. 8.7 t. Fulminant hepatic failure 2 (6.9%) had grade 2 changes compared with 2 of 22 (9.9%) of patients in the remaining groups. An ABO-compatible nonidentical graft was received by 6 of the 7 patients who developed rejection while continuing corticosteroid therapy. Four of these patients required retransplantation, and 2 subsequently died. There was nothing in the medical background of the remaining patients that might predict a complicated course, and only patient in this group was transplanted for fulminant hepatitis. Complications of Immunosuppression Therapy Complications associated with immunosuppression therapy for each of the follow-up groups are shown in Table 5. Antibiotic use was markedly lower following steroid withdrawal, and there also appeared to be a beneficial effect in reducing viral, heurnoqstis and fungal infection. Diabetes mellitus was newly diagnosed in 6 patients after OLT but resolved in all but two cases after withdrawal of steroids. However, patients with hypertension did not benefit, presumably because of continuing treatment with cyclosporine A. Further Details of Immunosuppression Therapy Azathioprine. Six patients did not receive azathioprine at any time and continued immunosuppression with cyclosporine alone (Table 6). An additional 3 Table. Details of Rejection Episodes and Their Treatment After OLT Cellular Rejection Cellular Rejection Cellular Rejection Cellular Rejection Absent (n = 88) Present (n = 8) Absent (n = ) Present (n = 7) Eplsodes of acute rejection On steroids Number (mean 2 SO).2 2.8.6 2.7.5 2. 2. f.7 Median (range) (-5) (-3) (-5) (-3) Off steroids Number (mean? SD). f. - Median (range) (-2) - Antlrejectlon treatment On steroids Methylprednisolone No. -g doses.2 2 2.7 7.? 3.7 5. t.3 8.3 f 2.8 Median (range) 3 (-5) 6.5 (3-2) 3 (-5) 7 (6-) OKr3 (days of treatment).8 t.9 2. 2 3.9 2.7 f 5.3 ATG (days of treatment).8 It 2.6 2.9 2.9 2. f 3.6 Off steroids Methylprednisolone No. -g doses.3 2.8 - - Median (range) 3 (3-7) - -

Immunosuppression Without Corticosteroids 5 Table 5. Complications Associated with OLT During Treatment wlth Corticosteroids and After Withdrawal and in Patients who Continued Corticosteroids Cellular Rejection Absent Cellular Rejection Present Cellular Cellular Before After Before After Rejection Rejection Stopping Stopping Stopping Stopping Absent Present (n = 88) (n = 8)* (n = 8) (n = 8) (n = ) (n = 7) Cytomegalovirus proven Cytomegalovirus not proven Ganciclovir used Pneumocystis carinii P. carinii treated Antibacterial use Antifungal use Other Hypertension Diabetes pre-olt Diabetes post-olt 3 3 65 7 9 22 2 23 9 3 2 2 2 patients stopped azathioprine during follow-up, 2 of them after stopping steroids. One of these patients subsequently developed acute rejection. In patients the reasons for stopping azathioprine were the development of malignancy (2), cytomegalovirus and herpes simplex virus infection with Haernophilus influenzae pneumonia (l), and recurrent hepatitis D (). The remaining 5 patients had been transplanted early in the program and had been stable on cyclosporine and azathioprine for a considerable period of time (7.2 k 2 months) after stopping corticosteroids. Azathioprine was stopped before corticosteroids in patients. In of these, rejection episodes subsequently developed. Azathioprine was stopped in 3 of 6 patients who subsequently developed longterm rejection; in patient this was before stopping corticosteroids, and in 2 patients it was after. Cyclosporine. Cyclosporine was stopped in 6 patients, of whom 5 were in the corticosteroid withdrawal group (Table 6). In the subgroup with no rejection, patient developed a lymphoma and died, and the other had a severe cholestatic syndrome and was retransplanted. Of the 3 patients in the corticoste- Table 6. Details of Azathioprine and Cyclosporine A Therapy in Patient Groups After OLT Cellular Rejection Cellular Rejection Cellular Rejection Cellular Rejection Absent (n = 88) Present (n = 8) Absent (n = ) Present (n = 7) Azathioprine Never received 6 Stopped On steroids 3 2 3 2 Off steroids 9 Cyclosporine A Stopped 2 3 Dose/ kg*. & 2. 5.8 %. 8.2 f.3 6.2? 5.5 <5 ng/mlt 3. f 2.8 2.?.8 2.7 %..7 %. *Daily dose of cyclosporine A taken by patient at time of last follow-up. tnurnber of times that the blood plasma level was 5 ng/mlduring follow-up (therapeutic range, 5-5 ng/rnl).

6 Fraser et al roid withdrawal rejection subgroup, stopped because of poor compliance, developed long-term rejection, was retransplanted, and later died. The other 2 were transferred to tacrolimus: with recurrent acute cellular rejection who is alive and with long-term rejection who was retransplanted and is alive. The 6th patient developed acute cellular rejection while on steroids, responded to bolus methylprednisolone, and was then maintained on corticosteroids without evidence of rejection. Discussion Corticosteroids were withdrawn in 96 of patients (8.2%) surviving for more than 3 months after OLT. We had anticipated that withdrawal would be completed by this time; however, because this was a clinical guideline and not part of a formal protocol in practice, the withdrawal was delayed. In the remaining patients, steroids could not be withdrawn because of a complicated clinical course. Although the benefits of this policy are difficult to assess precisely because there was no true control group for comparison, we did find that bacterial infections were uncommon after withdrawal, and 3 patients with diabetes mellitus in the postoperative period returned to normal glucose tolerance, although hypertension did not improve significantly. Moreover, episodes of acute cellular rejection developed in only 8 of 96 patients (8.3%) after corticosteroid withdrawal, which is very similar to the 7.8% with rejection episodes following corticosteroid withdrawal reported by Padbury et al. The figures for long-term rejection were also very similar, being.2% and.5%, respectively. We have, therefore, confirmed that corticosteroids can be withdrawn safely in the majority of cases. In addition to examining the effect of corticosteroid withdrawal, we abstracted data from files of patients who remained on corticosteroids. Mortality or retransplantation was high in patients in these groups. Patients who developed rejection while on corticosteroids were younger than those in other groups, had more episodes of acute cellular rejection, and required more doses of bolus -g methylprednisolone and more days of Om3 treatment. Six of 7 of these patients received ABO nonidentical but compatible grafts, but only patient was transplanted for fulminant hepatic disease. By contrast, no particular problems were reported by Padbury et ai for this matching. Caution may be needed before reducing immunosuppression in patients with ABO-compatible nonidentical grafts. Corticosteroid tapering appears to be possible when there is good graft function and patient stability, and this in turn allows the patient to enjoy the benefits of that withdrawal. Rubin has described patients that require continuing intensive immunosuppression as the chronic ne er do wells, and these patients are especially prone to late infections.9 Specifically, infections more than 6 months after OLT are not related to the immediate immunosuppressive regimen but more to the area under the curve, that is, the sum of immunosuppression after transplantation. This may also be true for other complications of organ transplantation. Thus, failure to withdraw corticosteroids appears to be of negative prognostic significance over and above the side effects associated with this treatment. Quality of life has been assessed in renal transplant patients given different long-term immunosuppressive regimen^.^ All patients received similar treatment of up to 3 months of cyclosporine and prednisone, and then half the patients continued with cyclosporin alone and half were converted to treatment with azathioprine and prednisone without cyclosporine. Quality-of-life measurements were made just before randomization and at 6 and 2 months after OLT, and scores were high in both patient groups and compared favorably with the general population. Analysis of patients adhering strictly to the protocols indicated that patients treated with cyclosporine alone had better psychosocial scores on the Sickness Impact Profile than the azathioprinecorticosteroid group. The authors concluded that this benefit was due to corticosteroid withdrawal. A recent randomized prospective study comparing steroid withdrawal with long-term steroid therapy was conducted in patients who suffered mainly from hepatitis B or hepatitis C cirrh~sis.~ There were 32 and 3 patients in each group, respectively, but unfortunately, due to small numbers, the occurrence of each specific steroid-related complication was small, and differences did not reach statistical significance. However, when all steroid-related complications were compared in the two groups, they were significantly more frequent in the patients maintained on. mg/kg/day methylprednisolone. There were no differences in the incidence of episodes of acute or chronic rejection between the two groups. We conclude that corticosteroid withdrawal can be achieved in the majority of cases and benefits the patient by reducing infections, diabetes mellitus, and probably osteoporosis. An increase in acute and long-term rejection does not appear to be a consequence of this policy.

Immunosuppression Without Corticosteroids 7 References 5. Romani F, Belli LS, De Carlis L, Rondenara A, Alberti A, Zetterman RK, McCashland TM. Long-term follow-up of the orthotopic liver transplantation patient. Semin Liver Dis 995;5:73-8. Lake JR, Roberts JP, Ascher NL. Maintenance immunosuppression after liver transplantation. Semin Liver Dis 992;2:73-79. Dunn SP, Falkenstein K, Lawrence JP, Meyers R, Jincaur C, Billmire D. Monotherapy with cyclosporin for chronic immunosuppression in pediatric liver transplant recipients. Transplantation 99575-57. Padbury RTA, Gunson BK, Dousset B, Hubscher S, Buchels J, Neuberger J, et al. Steroid withdrawal from long-term immunosuppression in liver allograft recipients. Transplantation 993;55:789-79. Sansalone C, et al. Cyclosporin monotherapy (after 3 months) in liver transplant patients: a prospective randomized trial. Transplant Proc 99;26:2683-2685. 6. Van Hoek B, Wiesner RH, Ludwig J, Gores GJ, Moore B, Krom RAF. Combination immunosuppression with azathioprine reduces the incidence of ductopenic rejection and vanishing bile duct syndrome after liver transplar+ation. Transplant Proc 99 ;23:3-6. 7. Hilbrands LB, Hoitsma AJ, Koene RAP. The effect of immunosuppression drugs on quality of life after renal transplantation. Transplantation 995;59:263-27. 8. International Working Party. Terminology for hepatic allograft rejection. Hepatology 995:22;69-65. 9. Tolkoff-Rubin NE, Rubin RH. The interaction of immunosuppression with infection in the organ transplant recipient. Transplant Proc 99;26:6-9.