Renal dysfunction is a common postoperative complication following orthotopic liver transplantation and has been reported to occur in 12% to 67% of liver transplant recipients.1–3 Early renal dysfunction is a risk factor for chronic renal failure and is associated with significant morbidity and mortality in liver transplant recipients.2, 4–7
Many risk factors such as preexisting renal impairment, hepatorenal syndrome, diabetes mellitus, liver allograft dysfunction, significant intraoperative blood loss, infection, and exposure to nephrotoxic calcineurin inhibitors (CNIs) have been reported to be associated with early renal dysfunction and acute renal failure after liver transplantation.1, 2, 8–11 Various strategies have been investigated to minimize the risk of acute renal failure after orthotopic liver transplantation, such as optimization of hemodynamic conditions in the operative and perioperative periods, but these strategies have not been successful.12–15
Delaying the initiation of the nephrotoxic CNIs cyclosporine and tacrolimus until renal function recovers appears to be a logical approach to minimize early renal dysfunction in liver transplantation. Results from recent clinical studies have demonstrated the safety and efficacy of this strategy;16–21 however, there is concern that delaying the initiation of CNI may place patients at risk for early acute rejection.17
Building on the premise that anti-thymocyte globulin allows for delayed initiation of CNI without an increased risk of acute rejection in kidney transplant recipients,22, 23 and because it has been used safely in liver transplant recipients,18, 21–24 we postulate that delaying the initiation of CNI with anti-thymocyte globulin may be safe and effective in preserving renal function in liver transplant recipients with renal dysfunction. We therefore conducted a retrospective study comparing the outcomes of liver transplant recipients with renal dysfunction who received either anti-thymocyte globulin and delayed initiation of CNI (anti-thymocyte globulin group) or no antibody and early initiation of CNI (control group).
CMV, cytomegalovirus; CNI, calcineurin inhibitor; GFR, glomerular filtration rate; HCV, hepatitis C virus; IL-2R, interleukin-2 receptor; MELD, Model for End-Stage Liver Disease; NS, not significant; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; WBC, white blood cell.
PATIENTS AND METHODS
Study Design and Patient Selection
This study was approved by the local institutional review board. This retrospective cohort study compared outcomes of 118 consecutive liver transplant recipients with renal dysfunction who received anti-thymocyte globulin and delayed initiation of CNI (anti-thymocyte globulin group) with 80 consecutive liver transplant recipients with renal dysfunction who received no antibody and early CNI initiation (control group). Renal dysfunction was defined as a baseline pretransplant serum creatinine ≥ 1.5 mg/dL or the need for dialysis at the time of transplantation.
Anti-Thymocyte Globulin Group
All 118 liver transplant patients with renal dysfunction that underwent transplantation between July 2001 and September 2006 received rabbit anti-thymocyte globulin (Thymoglobulin, Genzyme Corp., Cambridge, MA) at a dose of 0.5-1.0 mg/kg/day intravenously, starting on posttransplant day 1, and this continued daily until either the serum creatinine level had declined to ≤1.2 mg/dL or the patient had received a total of 5 doses. Anti-thymocyte globulin was infused over 4 hours for the first and all subsequent doses, with premedication given 30 to 60 minutes prior to infusion (either 40 mg of dexamethasone or 250 mg of methylprednisolone, 50 mg of intravenous or oral diphenhydramine, and 650 mg of oral acetaminophen). Tacrolimus (Prograf, Astellas Pharma US, Inc.) or cyclosporine (modified) was initiated either 1 day prior to the last of the 5 target doses of anti-thymocyte globulin or when the serum creatinine level declined to ≤1.2 mg/dL. Five hundred milligrams of mycophenolate mofetil (CellCept, Roche Pharmaceuticals) orally twice daily was initiated in all patients on posttransplant day 1 and continued for 6 months. Intraoperatively, all patients received corticosteroids either as methylprednisolone (500 mg) intravenously or as dexamethasone (100 mg) intravenously. The dose of corticosteroids was tapered and discontinued by 3 months post-transplant.
Eighty consecutive liver transplant recipients with renal dysfunction transplanted prior to July 2001 received CNI early post-transplant and did not receive any antibody therapy. In this historical control group, tacrolimus or cyclosporine was initiated within 48 hours post-transplant. Five hundred milligrams of mycophenolate mofetil orally twice daily was initiated in all patients on posttransplant day 1 and continued for 6 months. Patients received 500 mg of methylprednisolone intravenously intraoperatively. The dose of corticosteroids was tapered and discontinued by 3 months post-transplant.
All patients received the same prophylactic regimens consisting of postoperative bacterial and fungal prophylactic regimens with piperacillin/tazobactam every 8 hours for 24 hours and clotrimazole troches 5 times daily for 4 weeks, respectively. Oral ganciclovir was administered for 3 months following transplantation, regardless of donor and recipient cytomegalovirus serology. Sulfamethoxazole/trimethoprim was administered 3 times weekly for prophylaxis against Pneumocystis carinii.
Patients were evaluated at baseline, at completion of anti-thymocyte globulin therapy, and at 1, 3, 6, and 12 months post-transplant. The primary outcomes of interest were patient and graft survival at 12 months, early (30 days) biopsy-proven acute rejection, renal function, incidence of hepatitis C recurrence, and incidence of adverse events. Donor and recipient demographics, etiology of end-stage liver disease, Model for End-Stage Liver Disease (MELD) score, pretransplant hospitalization, and dialysis dependence were also collected. Collected laboratory parameters included serum creatinine levels, CD3 cell counts, white blood cell (WBC) counts, platelet counts, and hemoglobin levels. Creatinine clearance at baseline was measured by 24-hour urine collection. The glomerular filtration rate (GFR) was estimated with the Modification of Diet in Renal Disease equation.
Biopsies were performed in all suspected rejection episodes. Rejection episodes were graded according to Banff criteria, with the first episode treated with either 500 mg of methylprednisolone per day for 3 days or 100 mg of dexamethasone per day for 3 days. Refractory rejection was treated with anti-thymocyte globulin (1.5 mg/kg/day) for 7 to 10 days. Recurrence of hepatitis C was determined either by histological changes on biopsies or by significant increase in hepatitis C viral load in the presence of abnormal liver function.
Categorical variables were summarized with counts and percentages and were compared with either the chi-square test or Fisher's exact test when appropriate. Continuous variables were summarized as means and standard deviations or medians and ranges and were compared with the Student t test. Comparisons of the serum creatinine levels during the follow-up period were made by analysis of variance.
Patient demographics and baseline characteristics were similar between groups (Table 1). There were significantly more hepatitis C virus–positive patients in the anti-thymocyte globulin group [63/118 (53%)] in comparison with the control group [11/80 (14%); P < 0.001]. A high percentage of patients in both groups were hospitalized prior to transplant, and the majority of patients had MELD scores ≥ 20 at the time of transplant (Table 1). All patients had preexisting renal dysfunction as reflected in the elevated serum creatinine levels, low creatinine clearance, reduced GFR, and high number of patients on dialysis prior to liver transplant (Table 1). One hundred thirteen patients (95%) in the anti-thymocyte globulin group and all patients in the control group completed 12 months of follow-up.
Table 1. Baseline Characteristics
Anti-Thymocyte Globulin (n = 118)
Control (n = 80)
Abbreviations: CMV, cytomegalovirus; HCV, hepatitis C virus; MELD, Model for End-Stage Liver Disease; NS, not significant; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis.
Recipient age (years)
54 ± 9
Recipient gender (males)
Recipient race (Caucasian)
Etiology of end-stage liver disease
HCV and hepatocellular carcinoma
HCV and alcoholic cirrhosis
CMV donor (D)/recipient (R) serology
23.4 ± 6.1
Hospitalized prior to transplant
Dialysis prior to transplant
Serum creatinine (mg/dL)
2.8 ± 1.5
2.2 ± 0.9
Creatinine clearance (mL/min)
39.4 ± 23.9
41.4 ± 16.0
The mean total cumulative dose of anti-thymocyte globulin was 2.1 ± 1.2 mg/kg, given in a mean of 3.3 ± 1.6 doses over 4.7 ± 2.6 days. Infusion-related side effects requiring dosing adjustment and interruption occurred in 3 patients. Sixteen patients (13%) required adjustment of the anti-thymocyte globulin dose because of thrombocytopenia, leukopenia, or a combination of these reasons.
The mean time to initiation of CNI was 4.6 ± 2.3 days in anti-thymocyte globulin–treated patients and 1.9 ± 2.5 days in control patients (P < 0.001). Tacrolimus was the most common CNI administered in both groups (anti-thymocyte globulin: 90%, control: 80%, P = not significant). There were no differences in tacrolimus or cyclosporine trough levels between the 2 groups during the follow-up period (Fig. 1).
Patient Survival, Graft Survival, and Biopsy-Proven Acute Rejection
Actual patient (anti-thymocyte globulin: 90%, control: 89%) and graft survival rates (anti-thymocyte globulin: 88%, control: 86%) at 12 months post-transplant were similar between groups. The causes of death and graft failure were also similar between groups. Five patients in each group died of infection. Two patients in the anti-thymocyte globulin group and 1 patient in the control group died of a cerebral vascular event. Other causes of death in the anti-thymocyte globulin group included noncentral pontine myelinolysis (n = 1), graft failure (n = 2), hepatic artery thrombosis (n = 1), and pulmonary edema (n = 1). None of the deaths were related to anti-thymocyte globulin administration. In the control group, 1 patient died of recurrent hepatitis C and 1 patient died of hepatocellular carcinoma. Nineteen of 118 patients (16%) in the anti-thymocyte globulin group and 21 of 80 patients (26%) in the control group experienced biopsy-proven acute rejection within the first 30 days post-transplant (P = 0.08). The majority of rejection episodes were considered mild as graded by the Banff criteria.
Serum creatinine levels significantly decreased from 2.7 ± 1.5 mg/dL at baseline to 2.0 ± 1.3 mg/dL at the end of anti-thymocyte globulin treatment (P < 0.01), and serum creatinine levels continued to decline and remained stable between 1 and 12 months post-transplant (Fig. 2). Early improvements in renal function were also observed in the control group; however, this improvement was not sustained as serum creatinine levels increased starting at 3 months post-transplant and remained elevated at 12 months (Fig. 2). Anti-thymocyte globulin–treated patients had significantly higher GFR (57.4 ± 20.5 mL/min/1.73 m2) than the control group (43.7 ± 14.4 mL/min/1.73 m2) at 12 months of follow-up (P < 0.001; Fig. 3). Compared to baseline, estimated GFR at 12 months had improved by 36.6% in the anti-thymocyte globulin group, whereas estimated GFR increased by only 8.5% in the control group (P < 0.01).
Of the 19 patients (16%) in the anti-thymocyte globulin group who were dialysis-dependent prior to transplant, none of the patients required dialysis at 1 year post-transplant. One patient who was not on dialysis prior to or during anti-thymocyte globulin administration required dialysis at the end of follow-up (0.8%). In contrast, the proportion of patients in the control group requiring dialysis before transplantation (13%) remained the same post-transplantation (13%) and was significantly higher than that of the anti-thymocyte globulin group (13% versus 0.8%; P < 0.001).
Following treatment with anti-thymocyte globulin, CD3 cell counts declined from 201 ± 275 to 20 ± 23 cells/mm3 (P < 0.01). There were no statistically significant differences in WBC counts and platelet counts between the 2 groups (Fig. 4A,B). However, hemoglobin levels in the controls were significantly lower than those of the anti-thymocyte globulin–treated patients at 12 months post-transplant (Fig. 4C).
Of the 63 hepatitis C virus–positive patients in the anti-thymocyte globulin group, hepatitis C disease recurred in 25 patients (40%) during the 12 months of follow-up and occurred at a median of 133 days (range: 22-362 days) post-transplant. Similarly, recurrence of hepatitis C was observed in 7 of the 11 hepatitis C virus–positive patients (64%) in the control group, with a median time to recurrence of 110 days (range: 9-201 days) post-transplant. Cytomegalovirus disease was significantly less common in the anti-thymocyte globulin–treated patients (2.5%) compared with control patients (11.3%; P < 0.01). Overall infection incidence was also significantly less common in the anti-thymocyte globulin group (38%) compared with the control group (51%; P < 0.05). In both groups, the majority of these infections were of bacterial origin. Seven anti-thymocyte globulin–treated patients (5.9%) and 9 control patients (11.3%) experienced systemic fungal infections. One patient in the control group experienced herpes simplex viral infection. One patient in the anti-thymocyte globulin group, who had a history of previous liver transplant, developed squamous cell carcinoma of the skin during the follow-up period. There were no cases of posttransplant lymphoproliferative disorder or other malignancies observed in either group.
Early renal dysfunction in liver transplant recipients is associated with the development of end-stage renal disease and significant morbidity and mortality post-transplant.2–5, 7–9 Delaying the initiation of CNI with anti-thymocyte globulin in patients at high risk of acute renal failure was associated with a recovery in renal function in the early posttransplant period. Patients on this regimen also experienced a continual improvement in renal function, with a greater independence from dialysis, higher estimated GFR, and lower serum creatinine levels than control patients at 12 months post-transplant. This observation may have significant clinical implication as 12-month serum creatinine > 1.7 mg/dL has been reported to be a significant risk factor for development of end-stage renal disease requiring dialysis and kidney transplantation.8
One surprising observation from our study is that using anti-thymocyte globulin to delay the initiation of CNI by only 4.6 days resulted in significant improvement in estimated GFR. GFR increased by 36.6% at 12 months post-transplant in comparison with pretransplant values. In contrast, the early initiation of CNI and no antibody therapy were not associated with a continual improvement and resulted in only an 8.5% change in GFR at 12 months post-transplant compared to baseline.
Anti-thymocyte globulin was well tolerated in this cohort of acutely ill patients, with few patients experiencing any serious adverse events. Leukopenia and thrombocytopenia, which are known side effects of anti-thymocyte globulin, were uncommon and were easily managed by dose reduction or interruption. Starting at 1 month post-transplant, there were no differences in WBC and platelet counts between the 2 groups. In contrast, hemoglobin levels were significantly higher in the anti-thymocyte globulin–treated patients than the controls at 12 months of follow-up. Because anemia is a common manifestation of chronic renal insufficiency, this observation is consistent with the higher serum creatinine levels observed in the controls.
The use of any antibody therapy for induction in liver transplantation has remained controversial. The liver is considered an immunologically privileged organ; thus, the use of antibody to prevent rejection has been perceived as unnecessary and may increase the risk of overimmunosuppression. In this study, the use of anti-thymocyte globulin was associated with a numerically lower incidence of early acute rejection, a significantly lower incidence of infection, and no increased risk of malignancy in comparison with controls. Importantly, anti-thymocyte globulin was not associated with an increased incidence of hepatitis C recurrence.21, 24 This low incidence of adverse events, especially infection, may be related to the relatively low mean cumulative dose of anti-thymocyte globulin (2.2 mg/kg) used to delay the initiation of CNI in contrast to the conventional 1.5 mg/kg/day for 3 to 7 days used in kidney transplantation as induction therapy.25–28 Additionally, the control patients required additional immunosuppression to treat acute rejection, which might lead to increased overall immunosuppression load and an increased incidence of infection.
Our results are similar to those reported by Tchervenkov and colleagues18 on their experience with anti-thymocyte globulin induction in 298 consecutive liver transplant recipients. Anti-thymocyte globulin induction was found to be safe and effective in liver transplant recipients, and a subgroup analysis of patients with existing renal dysfunction (n = 59; serum creatinine ≥1.5 mg/dL) revealed that anti-thymocyte globulin induction (n = 47) allowed a significantly delayed initiation of CNI compared with no induction (n = 12): 10.8 versus 2.4 days, respectively. Despite higher baseline serum creatinine levels, patients receiving anti-thymocyte globulin showed a significant recovery in renal function. In agreement with our observations, the delayed initiation of CNI with anti-thymocyte globulin was not associated with an increased risk of acute rejection in this study.18
Although delaying initiation of CNI with interleukin-2 receptor (IL-2R) antagonists basiliximab and daclizumab has been associated with acceptable short-term outcomes, long-term benefits in renal function have not been observed.16, 17, 20, 29, 30 In comparison with our current study, the lack of apparent long-term renal function benefits from the use of IL-2R antagonists to delay CNI initiation may be related to the differences in the patient populations studied and the pharmacology of these antibodies. In previous studies that used IL-2R antagonists to delay initiation of CNI, most patients did not have preexisting renal dysfunction and were not critically ill.16, 19, 29 In this study, all patients had preexisting renal dysfunction (≥1.5 mg/dL), and most were critically ill at the time of transplant, as evident by the high MELD score and high percentage of patients hospitalized prior to transplant. Therefore, it is possible that patients at high risk for acute renal failure may derive the most benefit from antibody therapy and delayed initiation of CNI. Furthermore, because subclinical renal dysfunction can result from ischemia-reperfusion injury, improvements in renal function observed with delaying initiation of CNI with anti-thymocyte globulin may be related to the effects of anti-thymocyte globulin in minimizing ischemia-reperfusion injury.31, 32 Prospective, randomized, comparative studies are needed to determine whether there are advantages to using anti-thymocyte globulin over the IL-2R antagonists to facilitate delayed initiation of CNI in liver transplantation.
In conclusion, we observed that anti-thymocyte globulin is safe in critically ill liver transplant recipients with renal dysfunction and allows for delayed initiation of CNI. This strategy was successful in preserving renal function through 12 months of follow-up. Anti-thymocyte globulin and delayed CNI initiation were also associated with a low incidence of acute rejection and significantly less infection at 12 months post-transplant in comparison with patients receiving early initiation of CNI and no antibody. Prospective, randomized, control studies are warranted to further evaluate the safety and efficacy of this approach in liver transplant recipients at risk for renal dysfunction.