Risk factors for consequent kidney impairment and differential impact of liver transplantation on renal function

Transcription

1 Nephrol Dial Transplant (2010) 25: doi: /ndt/gfq093 Advance Access publication 5 March 2010 Risk factors for consequent kidney impairment and differential impact of liver transplantation on renal function Jung Pyo Lee 1, Nam Ju Heo 2, Kwon Wook Joo 2, Nam Joon Yi 3, Kyung-Suk Suh 3, Kyung Chul Moon 4, Seong Gyun Kim 5 and Yon Su Kim 1,2 1 Seoul National University Kidney Research Institute, Seoul, Korea, 2 Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Korea, 3 Department of Surgery, Seoul National University College of Medicine, Seoul, Korea, 4 Department of Pathology, Seoul National University College of Medicine, Seoul, Korea and 5 Department of Internal Medicine, Hallym University College of Medicine, Chunchon, Korea Correspondence and offprint requests to: Yon Su Kim; yonsukim@snu.ac.kr Jung Pyo Lee collected data, analysed results, performed the study, wrote the manuscript and revised it. Nam Ju Heo, Kwon Wook Joo, Nam Joon Yi, Kyung-Suk Suh and Seong Gyun Kim designed the study and collected data. Kyung Chul Moon reviewed the kidney histopathology. Yon Su Kim designed the study, analysed the results and wrote, edited and revised the manuscript. Abstract Background. Chronic kidney disease (CKD) develops frequently after liver transplantation (LTx), and it is important to identify and correct risk factors that negatively affect kidney function. Risk factors have not been well evaluated in Asian countries where hepatitis B virus (HBV) infection is a dominant cause. Methods. Four hundred thirty-one Korean recipients who underwent LTx between 1997 and 2008 were analysed. CKD was defined as a sustained decrease in estimated glomerular filtration rate (egfr) of <60 (ml/min/1.73 m 2 ) for at least three consecutive months using an abbreviated Modification in Renal Disease (MDRD) formula. Results. Eighty percent of the patients had HBV-related underlying diseases. The recipients whose pretransplant egfr had been low (<30 ml/min/1.73 m 2 ) improved their renal function after LTx, but significant functional decline occurred in recipients whose pretransplant egfr was high ( 60 ml/min/1.73 m 2 ). A multivariate Cox regression analysis revealed that the overall risk of CKD development (egfr < 60 ml/min/1.73 m 2 ) was associated with old age of recipients, cyclosporine, posttransplant acute renal failure (ARF), cause [calcineurin inhibitor (CNI) nephrotoxicity] and severity of posttransplant ARF, low pretransplant egfr, pretransplant hepatorenal syndrome, pretransplant proteinuria, high Child Pugh score and high Model for End-Stage Renal Disease (MELD) score. Especially in recipients whose pre-operative egfr was high ( 60 ml/min/ 1.73 m 2 ), rapid progression of kidney disease was associated with high tacrolimus level, non-hbv disease, posttransplant ARF, cause (CNI nephrotoxicity) and severity of posttransplant ARF and Child Pugh score. CNI toxicity and focal segmental sclerosis, but not immune-complex disease, were revealed as significant contributors to CKD after LTx in HBV recipients. Conclusion. Judicious use of CNIs should be applied to liver recipients to prevent kidney dysfunction. Keywords: calcineurin inhibitor toxicity; hepatitis B virus; kidney biopsy; kidney dysfunction; liver transplantation Introduction Chronic kidney disease (CKD) is a common problem in long-term survivors after liver transplantation (LTx). The prevalence of CKD ranges from approximately 20% to 70%, and the occurrence of CKD significantly increases the mortality in liver recipients [1,2]. Recently, several studies were conducted to clarify the risk factors for CKD progression, such as the use of calcineurin inhibitors (CNIs) and preoperative kidney function [1,3,4]. Asian ethnicity, in comparison to other ethnic groups, has been suggested as a good prognostic factor for the development of CKD after LTx [1], but a reason for this has never been documented. In Western countries, liver disease caused by hepatitis C virus (HCV) linked with glomerulonephritis is the leading indication for liver transplantation, accounting for 30% to 50% of cases [1,5,6]. However, an Asian institute, specifically the Korean Society, developed a distinctive strategy for the institution of liver transplantation in treating liver failure [7]. Hepatitis B virus (HBV) infection is more prevalent in Korea as well as in other Asian countries where the vertical transmission of HBV is a major route of infection. It implies the long duration of infection, which has a potential to cause serious liver diseases such as liver cirrhosis and hepatocellular carcinoma. These distinctions between Asians and Caucasians may explain not only the difference in frequency but also in the main cause of kidney dysfunction after LTx. We previously reported the features of CKD after LTx in The Author Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 CKD after liver transplantation 2773 Korea [8], but this report had limitations due to the insufficient number of recipients examined and the short duration of follow-up. The purpose of this study was to delineate the risk factors associated with progressive kidney dysfunction after LTx in recipients with HBV infection, the main indication for liver transplantation in Korea. Additionally, we reviewed kidney biopsy cases performed after liver transplantation in recipients with HBV infection to explain the above-mentioned ethnic differences because biopsy serves as a unique way to understand the exact cause of kidney dysfunction [5,9,10]. Materials and methods Study population We studied the clinical records of 604 consecutive Korean LTx patients who underwent LTx between 1997 and 2008 at our institute. Data from 173 recipients who received multi-organ transplantation, had a short follow-up duration (<6 months), died within 6 months after transplantation, or were <18 years old were excluded. Medical records were reviewed retrospectively. Clinical parameters that could have influenced development of chronic kidney disease were collected, i.e. age at transplantation, gender, era, history of hypertension and diabetes mellitus, underlying disease, history of hepatocellular carcinoma, body mass index (BMI), pretransplant proteinuria, Child Pugh score, Model for End-Stage Disease (MELD) score, hepatorenal syndrome, posttransplant acute renal failure (ARF), aetiologies of ARF, history of cardiovascular disease, donor type and immunosuppressant. Also, information on the agents used to treat hepatitis B virus infection such as adefovir or tenofovir, which are known to be nephrotoxic in some patients was collected. Hepatitis B viraemic status was measured with serum HBV DNA polymerase chain reaction (PCR) at and after transplantation. Measurement of kidney function and definition of CKD and ARF Serum creatinine was measured with a commercially available assay based on the modified Jaffé method reported by Larsen [11]. CKD was defined as a sustained decrease in estimated glomerular filtration rate (egfr) of <60 ml/min/1.73 m 2 for at least three consecutive months using an abbreviated Modification in Renal Disease (MDRD) formula [=186 (serum creatinine mg/dl) (age) (0.742 if female)]. Severe CKD was defined as an egfr < 30 ml/min/1.73 m 2 for at least three consecutive months. Posttransplant ARF was defined using the RIFLE criteria [12]. The aetiological categories of ARF, i.e. prerenal ARF, ischaemic acute tubular necrosis (ATN) and CNI nephrotoxicity were defined as described previously [13]. Hepatorenal syndrome was considered when all the major criteria of the International Ascites Club were fulfilled [14]. immunosuppression was preferable to tacrolimus-based immunosuppression. In cases receiving cyclosporine-based treatment, the trough level was maintained between 200 and 300 ng/ml for the first 3 months and was reduced to 100~200 ng/ml thereafter. In the tacrolimus-based protocol, the trough level was 8~13 ng/ml for the first 3 months and decreased to 5~100 ng/ml thereafter. When the recipient had poor renal function at the time of operation, the trough level of CNIs was maintained at a relatively low level, especially in recipients with egfr <30 ml/min/1.73 m 2. Prednisolone was started at 20 mg/day and gradually tapered. Routine liver biopsy was performed on posttransplant day 10. Purine synthesis inhibitors such as mycophenolate mofetil were used as an initial immunosuppressive treatment based on a clinical decision that considered the pretransplant kidney function, results of the protocol biopsy and risk factors of rejection [e.g. human leukocyte antigen (HLA) mismatch]. It was more frequently used from the initiation of immunosuppression when the renal function was significantly compromised. Kidney biopsy and histopathologic lesion Kidney biopsy was performed after the need for percutaneous kidney biopsy was evaluated by nephrologists in the case of significant proteinuria (>1.0 g/day), persistent microscopic haematuria or a progressive deterioration of renal function. Biopsy tissue was examined by light-, electron-, and immunofluorescent (IF) microscopy. All specimens were reviewed by an experienced kidney pathologist at the same time. Diabetic glomerulosclerosis was determined by light microscopy as well as IF using specific linear staining for IgG and albumin along the glomerular capillary walls and tubular basement membranes. CNI toxicity was diagnosed when biopsy specimens revealed the pattern of interstitial fibrosis with a striped appearance, nodular arteriolar hyalinosis and, later, tubular atrophy with glomerulosclerosis and arteriosclerosis. The number of focally or completely sclerosed glomeruli was scored as a percentage of the total number of glomeruli seen on each biopsy. Tubulointerstitial fibrosis was graded as +1 (0 24%), +2 (25 49%), +3 (50 74%), and +4 (75 100%). Vascular damage was also graded semiquantitatively as 0 to +3. Statistical methods The statistical analysis was performed using SPSS 12.0 K (SPSS Inc, Chicago, IL, USA). Data were expressed as means ± SD. Continuous and categorical data were compared using the independent t-test and the chi-squared test, respectively. Cumulative incidence of CKD was calculated using the Kaplan Meier method. The Cox regression model was used to calculate unadjusted and adjusted hazard ratio (HR) and 95% confidence interval (CI) for factors that affect the development of CKD. The interactions among variables for adjustments were determined using the general linear model. To compare the effects of variables on graft survival, we used the backward stepwise selection in the Cox regression model. Logistic regression analysis was used to discover risk factors for rapid progression of kidney dysfunction. P-values <0.05 were considered significant. Definitions of other clinical parameters Hypertension was defined as a systolic blood pressure over 140 mmhg, diastolic pressure over 90 mmhg or the use of anti-hypertensive medication. Presence of diabetes mellitus was diagnosed if patients had symptoms and had a random blood glucose concentration 200 mg/dl or if fasting plasma glucose was 126 mg/dl on at least two separate measurements or if patients used antidiabetic medication. MELD scores were calculated for each patient using the following equation: (0.957 log (e) (serum creatinine mg/dl) log (e) (serum bilirubin mg/dl) log (e) (PT International normalized ratio) 0.643) 10. Proteinuria was defined as 1 or more on the dipstick urinalysis. Immunosuppressive treatment protocols A standardized immunosuppression protocol involving a combination of CNIs and steroids was initiated within 24 h of surgery. Choice of calcineurin inhibitor between cyclosporine and tacrolimus was determined by a transplantation team. In recipients with preoperative renal insufficiency, hepatitis C virus infection, or presence of diabetes, cyclosporine-based Results Baseline characteristics A total of 431 adult recipients [male:female 312:119, mean age (±SD) 48 ± 9.2 years old], were included, and the follow-up duration was 46 ± 31.4 months. Eighty percent of LTx cases were due to HBV-related diseases (liver cirrhosis and hepatocellular carcinoma) (Table 1). The proportion of viraemic patients at transplantation was 82.3% in the HBV recipients. Almost all of them were treated with antiviral agents such as lamivudine and adefovir. Hepatocellular carcinoma was found in 40% of the recipients. The deceased donor transplantation rate was 23%. During the follow-up period, biopsy-proven acute re-

3 2774 J.P. Lee et al. Table 1. Demographic characteristics between cyclosporine-based and tacrolimus-based immunosuppression Total (n = 431) Cyclosporine (n = 72) Tacrolimus (n = 359) P-value Male recipient 72.4% 68.1% 73.3% Recipient age (years) 48.3 ± ± ± Era < % 8.3% 7.5% % 83.3% 40.9% % 8.3% 51.5% Body mass index 21.8 ± ± ± Pretransplant diabetes mellitus 14.4% 29.2% 11.4% <0.001 Duration of diabetes (years) a 4.9 ± ± ± Duration of treated diabetes (years) a 4.7 ± ± ± Pretransplant hypertension 4.4% 5.6% 4.2% Duration of anti-hypertensive therapy (years) b 5.5 ± ± ± History of cardiovascular disease 0.9% 1.4% 0.8% Pretransplant proteinuria 8.4% 6.9% 8.6% Pretransplant egfr (ml/min/1.73 m 2 ) 81.1 ± ± ± Hepatorenal syndrome 12.1% 15.3% 11.4% Posttransplant ARF 27.4% 36.1% 25.6% Cause of ARF c Ischaemic ATN 44.1% 42.3% 44.6% Prerenal ARF 27.1% 26.9% 27.2% CNI toxicity 16.9% 15.4% 17.4% Persistent hepatorenal syndrome 7.6% 7.7% 7.6% Others 4.2% 7.7% 3.3% RIFLE criteria c Risk 23.7% 3.8% 29.3% Injury 51.7% 69.2% 46.7% Failure 18.6% 19.2% 18.5% Loss 5.9% 7.7% 5.4% Posttransplant haemodialysis or haemofiltration 3.2% 2.8% 3.3% Deceased donor 23.0% 12.5% 25.1% Underlying disease HBV infection 79.6% 77.8% 79.9% HCV infection 3.9% 9.7% 2.8% Alcoholic liver disease 4.9% 4.2% 5.0% Others 11.6% 8.3% 12.3% HBV viraemia at transplantation d 82.3% 77.4% 83.0% HBV viraemia after transplantation d 8.1% 15.2% 6.8% Anti-HBV therapy d None used 9.8% 23.2% 7.3% Lamivudine 66.8% 71.4% 65.9% Lamivudine adefovir 4.4% 1.8% 4.9% Adefovir 4.4% 0.0% 5.2% Entecavir 14.6% 3.6% 16.7% Hepatocellular carcinoma 39.9% 43.1% 39.3% Child Pugh score 10.6 ± ± ± MELD score 21.0 ± ± ± Inhibitors of purine synthesis <0.001 None used 56.1% 13.9% 64.6% Mycophenolate 43.4% 83.3% 35.4% Azathioprine 0.5% 2.8% 0.0% Trough level of CNIs (ng/ml) [% of patients with trough levels above the target range e ] Posttransplant 1 month ± 75.1 [8.6%] 10.0 ± 2.5 [9.6%] f Posttransplant 3 months ± 62.3 [1.4%] 8.5 ± 2.2 [2.9%] f Posttransplant 6 months ± 50.4 [15.4%] 6.8 ± 2.2 [6.9%] f Posttransplant 12 months ± 53.0 [1.5%] 5.4 ± 1.9 [1.7%] f Mean ± SD, P-value (cyclosporine vs. tacrolimus), independent t-test for continuous variables and chi-squared test or Fisher s exact test for categorical variable. egfr, estimated glomerular filtration rate by MDRD formula; ARF, acute renal failure; HBV, hepatitis B virus; HCV, hepatitis C virus. a In the recipients with pretransplant diabetes mellitus. b In the recipients with pretransplant hypertension. c In the posttransplant ARF(+) recipients. d In the HBV recipients. e Target range of cyclosporine: ng/ml at 1 and 3 months, ng/ml at 6 and 12 months; target range of tacrolimus: 8 13 ng/ml at 1 and 3 months, 5 10 ng/ml at 6 and 12 months. f Fisher s exact test.

4 CKD after liver transplantation 2775

5 2776 J.P. Lee et al. Table 2. Univariate and multivariate analyses of the hazard ratios for chronic kidney disease (egfr < 60 ml/min/1.73 m 2 ) after liver transplantation in all recipients Univariate Cox regression Multivariate Cox regression a HR (95% CI) P-value HR (95% CI) P-value Male recipient (vs. female) 0.61 ( ) ( ) a Recipient age (years) 1.08 ( ) < ( ) <0.001 a Era (referent) 1 (referent) ( ) ( ) a ( ) ( ) a Body mass index 0.98 ( ) ( ) a Pretransplant diabetes mellitus 1.98 ( ) ( ) a Duration of diabetes (years) d 1.03 ( ) ( ) a Duration of treated diabetes (years) d 1.03 ( ) ( ) a Pretransplant hypertension 2.09 ( ) ( ) a Duration of anti-hypertensive therapy (years) 0.97 ( ) (0.00 ) a History of cardiovascular disease 1.86 ( ) ( ) a Deceased donor (vs. living donor) 0.95 ( ) ( ) a Underlying disease HBV infection 1 (referent) 1 (referent) HCV infection 2.39 ( ) ( ) a Alcoholic liver disease 0.76 ( ) ( ) a Others 1.09 ( ) ( ) a HBV-related liver disease (vs. non-hbv-related disease) 0.81 ( ) ( ) b Hepatocellular carcinoma 0.88 ( ) ( ) a Use of adefovir 1.01 ( ) ( ) a Calcineurin inhibitor Tacrolimus 1 (referent) 1 (referent) Cyclosporine 2.22 ( ) < ( ) a Inhibitors of purine metabolism None used 1 (referent) 1 (referent) Mycophenolate mofetil 1.81 ( ) ( ) a Azathioprine 1.14 ( ) ( ) a Acute rejection 1.44 ( ) ( ) a Posttransplant ARF (vs. no ARF) 1.46 ( ) ( ) a Cause of ARF No ARF 1 (referent) 1 (referent) Ischaemic ATN 1.66 ( ) ( ) c Prerenal ARF 0.48 ( ) ( ) c CNIs toxicity 2.98 ( ) < ( ) <0.001 c Persistent hepatorenal syndrome 6.81 ( ) < ( ) c RIFLE criteria No ARF 1 (referent) 1 (referent) Risk 0.99 ( ) ( ) c Injury 1.81 ( ) ( ) c Failure 1.55 ( ) ( ) c Loss 2.70 ( ) ( ) c Pretransplant haemoglobin (g/dl) 0.86 ( ) < ( ) a Pretransplant egfr (ml/min/1.73 m 2 ) 0.97 ( ) < ( ) <0.001 a Pretransplant proteinuria 2.32 ( ) ( ) a Hepatorenal syndrome 5.67 ( ) < ( ) <0.001 d Child Pugh score 1.13 ( ) ( ) a MELD score 1.05 ( ) < ( ) <0.001 e HR, hazard ratio; ARF, acute renal failure; egfr, estimated glomerular filtration rate by MDRD formula; MELD, Model for End-Stage Liver Disease. a Adjusted for the following variables: recipient age, gender, pretransplant diabetes mellitus, pretransplant hypertension, underlying disease, calcineurin inhibitor, inhibitors of purine metabolism, posttransplant ARF, pretransplant haemoglobin, pretransplant egfr, pretransplant proteinuria and Child Pugh score. b Adjusted for the variables listed in footnote (a) except for underlying disease. c Adjusted for the variables listed in footnote (a) except for posttransplant ARF. d Adjusted for the variables listed in footnote (a) except for pretransplant egfr. e Adjusted for the variables listed in footnote (a) except for pretransplant egfr and Child Pugh score. Fig. 1. Kidney function after liver transplantation and incidence of CKD. (A) Overall egfr after liver transplantation (paired t-test, *P < 0.05). (B) Cumulative incidence of CKD (egfr < 60 ml/min/1.73 m 2 ). (C) Cumulative incidence of severe CKD (egfr < 30 ml/min/1.73 m 2 ).

6 CKD after liver transplantation 2777 Fig. 2. Relationship between posttransplant ARF and CKD. Cumulative incidence of CKD (egfr < 60 ml/min/1.73 m 2 ) according to (A) causes of posttransplant ARF (log-rank test, *P < 0.05; **P < 0.001) and (B) RIFLE criteria of posttransplant ARF (log-rank test, **P < 0.01).

7 2778 J.P. Lee et al. Fig. 3. Impact of pretransplant egfr and progression of CKD. Kidney function at sequential time points in comparison to the pretransplant egfr (mean ± SD, paired t-test, *P < 0.05); pre-tx, pretransplant. jections were found in 9.2% of patients. The proportion of patients with hepatorenal syndrome was 12.1%. The mean Child Pugh and MELD scores (±SD) were 10.6 ± 2.6 and 21.0 ± 10.0, respectively. Posttransplantation ARF was documented in 27.4% of the recipients. The most common cause was ischaemic ATN (44.1%), followed by prerenal ARF (27.1%) and CNI nephrotoxicity (16.9%). Fourteen recipients (3.2%) had ARF of sufficient severity as to require renal replacement therapy after liver transplant surgery. Twenty-three (5.3%) deaths were observed during the follow-up period. The causes of death were as follows: 8 from infection, 10 from recurrence of hepatocellular carcinoma, 4 from hepatic failure due to rejection and 1 from cardiovascular disease. Kidney function after liver transplantation and incidence of chronic kidney disease The mean egfr (ml/min/1.73 m 2,±SD)was82± 30.0 ml/min/1.73 m 2 before transplantation, 67 ± 18.6 ml/min/1.73 m 2 at 6 months, 68 ± 17.8 ml/min/ 1.73 m 2 at 1 year, 69 ± 16.3 ml/min/1.73 m 2 at 3 years and 68 ± 16.8 ml/min/1.73 m 2 at 5 years (Figure 1A). Kidney function gradually deteriorated until 6 months after transplantation but was maintained thereafter. The cumulative incidence of CKD (egfr < 60 ml/min/1.73 m 2 )was 17.6% at 1 year, 23.7% at 3 years and 27.5% at 5 years (Figure 1B). An egfr < 30 ml/min/1.73 m 2 occurred in only five patients during follow-up (Figure 1C). One patient was on long-term renal replacement therapy. Risk factors for CKD A multivariate Cox regression analysis revealed that the overall risk of CKD (egfr <60) was associated with old age of recipients, cyclosporine, posttransplant ARF, the cause and severity of posttransplant ARF, low pretransplant egfr, pretransplant hepatorenal syndrome, pretransplant proteinuria, high Child Pugh score and high MELD score (Table 2). However, diabetes, hypertension, HCV infection, type of donor, acute rejection and pretransplant existence of hepatocellular carcinoma were not associated with CKD development. The risk for CKD was increased by 1.07 times as the age of recipient increased by 1 year. Cyclosporine-based immunosuppression increased the risk by 1.89 times in comparison to tacrolimus-based treatment. The posttransplant ARF increased the risk by a factor of Additionally, the cause and severity of posttransplant ARF were related to the risk of CKD (Table 2, Figure 2). The Risk group did not increase the risk of CKD, but the Injury group increased the risk by 3.04 times in comparison to the non-arf group. Prerenal ARF or ischaemic ATN could not increase the risk, but CNI nephrotoxicity significantly increased the risk by a factor of Impact of preoperative egfr and progression of chronic kidney disease We observed a change in posttransplant kidney function depending on pretransplant egfr. Recipients with low egfr (pretransplant egfr <60 ml/min/1.73 m 2 )were older and had higher Child Pugh scores, and the propor-

8 CKD after liver transplantation 2779 Table 3. Univariate and multivariate analyses of the hazard ratios for chronic kidney disease (egfr < 60 ml/min/1.73 m 2 ) after liver transplantation in the recipients with pretransplant good kidney function (egfr 60 ml/min/1.73 m 2 ) Univariate Cox regression Multivariate Cox regression a HR (95% CI) P-value HR (95% CI) P-value Male recipient (vs. female) 0.63 ( ) ( ) a Recipient age (years) 1.10 ( ) < ( ) <0.001 a Era (referent) 1 (referent) ( ) ( ) a ( ) ( ) a Body mass index 0.96 ( ) ( ) a Pretransplant diabetes mellitus 1.45 ( ) ( ) a Duration of diabetes (years) d 1.07 ( ) ( ) a Duration of treated diabetes (years) d 1.06 ( ) ( ) a Pretransplant hypertension 0.98 ( ) ( ) a Duration of anti-hypertensive therapy (years) 1.13 ( ) History of cardiovascular disease 4.79 ( ) ( ) a Deceased donor (vs. living donor) 0.76 ( ) ( ) a Underlying disease HBV infection 1 (referent) 1 (referent) HCV infection 3.65 ( ) ( ) a Alcoholic liver disease 0.87 ( ) ( ) a Others 1.41 ( ) ( ) a HBV-related liver disease (vs. non-hbv-related disease) 0.59 ( ) ( ) b Hepatocellular carcinoma 1.15 ( ) ( ) a Use of adefovir 0.74 ( ) ( ) a Calcineurin inhibitor Tacrolimus 1 (referent) 1 (referent) Cyclosporine 2.63 ( ) < ( ) a Inhibitors of purine metabolism None used 1 (referent) 1 (referent) Mycophenolate mofetil 1.52 ( ) ( ) a Azathioprine 1.79 ( ) ( ) a Acute rejection 1.34 ( ) ( ) a Posttransplant ARF (vs. no ARF) 2.02 ( ) ( ) a Cause of ARF No ARF 1 (referent) 1 (referent) Ischaemic ATN 2.00 ( ) ( ) c Prerenal ARF 0.72 ( ) ( ) c CNI toxicity 4.72 ( ) < ( ) <0.001 c RIFLE criteria No ARF 1 (referent) 1 (referent) Risk 1.50 ( ) ( ) c Injury 2.41 ( ) ( ) c Failure 2.02 ( ) ( ) c Pretransplant haemoglobin (g/dl) 0.91 ( ) ( ) a Pretransplant proteinuria 1.27 ( ) ( ) a Child Pugh score 1.05 ( ) ( ) a MELD score 1.01 ( ) ( ) d HR, hazard ratio; ARF, acute renal failure; MELD, Model for End-Stage Liver Disease. a Adjusted for the following variables: recipient age, gender, pretransplant diabetes mellitus, pretransplant hypertension, underlying disease, calcineurin inhibitor, inhibitors of purine metabolism, posttransplant ARF, pretransplant haemoglobin, pretransplant proteinuria and Child Pugh score. b Adjusted for the variables listed in footnote (a) except for underlying disease. c Adjusted for the variables listed in footnote (a) except for posttransplant ARF. d Adjusted for the variables listed in footnote (a) except for Child Pugh score. tion of males was lower at the time of transplantation (Supplement Table 1). In addition, CNIs were maintained at a lower trough level, and the use of mycophenolate mofetil was relatively higher in these patients compared to patients with higher egfr. Kidney function was significantly improved in the low egfr group (pretransplant egfr 29 ml/min/1.73 m 2 ) during the first 6 months (from 18 ± 7.1 to 48 ± 16.0 ml/min/1.73 m 2 ), which was maintained afterwards. In the low-intermediate egfr group (pretransplant egfr ml/min/1.73 m 2 ), there was no significant change in kidney function. Recipients with pretransplant egfr 60 ml/min/1.73 m 2 showed a significant reduction in kidney function during the first 6 months (Figure 3). Therefore, we assumed that it was critical to identify factors related to the decline of kidney function during the first 6 months after transplantation, especially in the recipients with good preoperative kidney function.

9 2780 J.P. Lee et al. Fig. 4. Comparison of kidney function between recipients treated with tacrolimus and with cyclosporine after liver transplantation (mean ± SD, independent t-test, *P < 0.05; **P < 0.01); pre-tx, pretransplant. Subgroup analysis in recipients with good preoperative kidney function First, we performed a subgroup analysis of factors associated with development of posttransplant CKD in recipients with pretransplant egfr 60 ml/min/1.73 m 2.A multivariate Cox regression analysis revealed that pretransplant low haemoglobin level and non-hbv disease as well as the previously described risk factors including old age of recipients, cyclosporine, posttransplant ARF and causes (ischaemic ATN and CNI nephrotoxicity) and severity of posttransplant ARF were revealed as risk factors (Table 3). We compared serial egfr changes between recipients treated with either tacrolimus or cyclosporine in recipients with a pretransplant of egfr 60 ml/min/1.73 m 2.Although both groups had similar pretransplant egfr and the trough level of cyclosporine was not higher than that expected, the posttransplant egfr of recipients treated with cyclosporine was significantly decreased after 6 months (Figure 4). Next, according to the rate of deterioration in renal function (ΔGFR) during the first 6 months after transplantation, we divided the recipients using tacrolimus into rapid progression group (ΔGFR 40, n = 55), slow progression group (ΔGFR 20 39, n = 78) and stable group (ΔGFR 19, n = 158). The trough level of the tacrolimus was significantly higher in the rapid progressive group compared to other two groups (Figure 5). When the rapid progression group and the stable group were compared by multivariate logistic regression, non-hbv disease, posttransplant ARF, cause (CNI nephrotoxicity) and severity of posttransplant ARF and Child Pugh score were acknowledged as significant risk factors for rapid progression in addition to the tacrolimus trough level (Table 4). This subgroup analysis was not done on those treated with cyclosporine due to the low number of each group [rapid progression group (n = 11), slow progressive group (n = 22) and stable group (n = 17)]. Kidney histopathology after LTx in patients with hepatitis B virus-associated liver disease Kidney biopsy was performed in nine recipients with HBV-related liver disease (Table 5). Eight recipients had progressive azotaemia [median serum creatinine at kidney biopsy 2.3 (1.9~3.0) mg/dl], and one (case 8) had newly developed proteinuria (random urine protein/creatinine 2.67 mg/mg, serum creatinine 1.2 mg/dl). The median time from transplantation to kidney biopsy was 11.3 (4.7~77.9) months. Chronic CNI toxicity was detected in five cases (55.6%), and none of these had overt proteinuria. In the remaining four cases with proteinuria over 1 mg/mg (spot urine protein/creatinine), two had

10 CKD after liver transplantation 2781 Fig. 5. Trough tacrolimus level depending on the rate of kidney dysfunction progression during first 6 months posttransplant. ΔGFR, rate of kidney dysfunction progression during first 6 months posttransplant; stable, ΔGFR 19 (n = 150); slow progressor, ΔGFR (n = 76); rapid progressor, ΔGFR 40 (n = 54) (mean ± SD, ANOVA test, *P < 0.05; **P < 0.01; ***P < 0.001); prog, progressor. IgA nephropathy (considered as associated with liver cirrhosis), one had prominent diabetic glomerular change, and one had severe sclerotic change and tubulointerstitial fibrosis similar to end-stage renal disease. Three cases had focal segmental sclerosis. However, HBV-related immunecomplex glomerulonephritis (GN) [e.g. membranous nephropathy or membranoproliferative GN (MPGN)] was not found (Table 5). Based on the results of the kidney biopsy, either a change in immunosuppressant or the addition of angiotensin receptor blockers was applied to the recipients. Only one recipient progressed to end-stage renal disease after the kidney biopsy, and the remainder demonstrated improvement in kidney function or complete remission of proteinuria. Discussion This work evaluated the incidence and risk factors of chronic kidney disease following liver transplantation in a cohort of transplant patients of a single Asian centre where the prevalence of HBV infection is high. Although there have been several studies examining kidney dysfunction after liver transplantation, deceased donor transplantation and liver failure caused by non-hbv disease were the main features of the study populations [4,15 18]. In previous work by Ojo et al., more than liver recipients were analysed, but only 5.6% of them were HBV(+) [1]. In this setting, HBV infection was not a significant risk factor compared to other causes comprising the rest of the recipients (94.4%). HBV infection, however, was the major cause of liver disease (80%) in our study population. In our cohort, we may compare the significance of specific viral infection such as HBV or HCV for the development of CKD. Therefore, this study provides important information that should be considered in the treatment of patients with different ethnic and epidemiologic backgrounds. We used creatinine-based GFR calculated with the abbreviated MDRD formula to assess renal function. Recently, several studies have tried to modify the formula with the Asian s own coefficients [19 22]. We also obtained Korean racial coefficients [four- and six-variable isotope dilution mass spectrometry (IDMS) MDRD] from the study Measurement of glomerular filtration rate and calculation of GFR estimates for Korea, granted by the Korean Society of Nephrology. Racial coefficients of four- and six-variable IDMS MDRD study equations using recalibrated serum creatinine were and , respectively (manuscript in preparation). Therefore, the usage of creatinine for the estimation of GFR in Asians, especially in Koreans, is appropriate and validated. The overall incidence of CKD in the present study was much lower than that in other studies [1,4,15,16], even though the definitions for CKD varied. It is consistent with the finding that Asian ethnicity may be a good prognostic factor for the development of CKD [1]. The decline of kidney function during the first 6 months was faster than that of later periods, and this finding is also similar to previous studies [3,16,17,23]. Through observing kidney function at sequential time points in comparison to the pretransplant egfr, we noted that kidney function declined to a greater degree in the high egfr subgroup (egfr 90 ml/min/1.73 m 2 ). On the other hand, recipients with low kidney function (egfr 29 ml/min/1.73 m 2 ) demonstrated a significant improvement in kidney function. Consequently, the mean egfr in each subgroup progressed toward a common mean. A unique finding of our work is that liver transplantation showed a differential impact on kidney function depending on the pretransplant egfr. This study identified old age, low pretransplant egfr, use of cyclosporine, low pretransplant haemoglobin, posttransplant ARF, cause (CNI nephrotoxicity) and severity of posttransplant ARF, pretransplant hepatorenal syndrome, pretransplant proteinuria, high Child Pugh score, high MELD score, non-hbv infection and high trough level of CNIs as independent risk factors from multivariate risk analysis and the subgroup analysis in the pretransplant high egfr recipients (egfr 60 ml/min/1.73 m 2 ).

11 2782 J.P. Lee et al. Table 4. Univariate and multivariate logistic regression of the odds ratios for rapid progression after liver transplantation in the recipients with pretransplant good kidney function (egfr 60 ml/min/1.73 m 2 ) using tacrolimus Risk factors Stable a (n = 158) Rapid progressor b (n = 55) Univariate analysis Multivariate analysis OR (95% CI) P-value OR (95% CI) P-value Male recipient (vs. Female) (%) ( ) ( ) d Recipient age (years) 47.8 ± ± ( ) ( ) d Era (%) P for trend P for trend d (referent) 1.00 (referent) ( ) ( ) d ( ) ( ) d Pretransplant diabetes mellitus (%) ( ) ( ) d Pretransplant hypertension (%) ( ) ( ) d Donor type (%) Living donor (referent) 1.00 (referent) Deceased donor ( ) ( ) d Underlying disease (%) HBV infection (referent) 1.00 (referent) HCV infection ( ) ( ) d Alcoholic liver disease ( ) ( ) d Others ( ) ( ) d HBV-related liver disease ( ) ( ) e (vs. non-hbv-related disease) (%) Hepatocellular carcinoma (%) ( ) ( ) d Inhibitors of purine metabolism (%) None used (referent) 1.00 (referent) Mycophenolate mofetil ( ) ( ) d Use of adefovir (%) ( ) ( ) d Posttransplant ARF (%) ( ) < ( ) d Cause of ARF No ARF 1.00 (referent) 1.00 (referent) Ischaemic ATN c ( ) ( ) f Prerenal ARF c ( ) ( ) f CNI toxicity c ( ) ( ) f RIFLE criteria P for trend <0.001 P for trend f No ARF 1.00 (referent) 1.00 (referent) Risk c ( ) ( ) f Injury c ( ) < ( ) f Failure and loss c ( ) ( ) f Acute rejection (%) ( ) ( ) d Pretransplant proteinuria (%) ( ) ( ) d Pretransplant haemoglobin (g/dl) 11.1 ± ± ( ) ( ) d MELD score 18.1 ± ± ( ) ( ) g Child Pugh score tacrolimus 9.8 ± ± ( ) < ( ) d trough level (ng/ml) at 1 month 9.6 ± ± ( ) < ( ) <0.001 d at 3 months 8.0 ± ± ( ) < ( ) <0.001 h at 6 months 6.0 ± ± ( ) < ( ) <0.001 h OR, odds ratio; ARF, acute renal failure; MELD, Model for End-Stage Liver Disease. a ΔGFR (rate of kidney dysfunction progression during posttransplant first 6months) 19 ml/min/1.73 m 2. b ΔGFR 40 ml/min/1.73 m 2. c % in the posttransplant ARF (+) recipients. d Adjusted for the following variables: recipient age, gender, era, donor type, underlying disease, hepatocellular carcinoma, inhibitors of purine metabolism, posttransplant ARF, posttransplant acute rejection, pretransplant proteinuria, pretransplant haemoglobin, Child Pugh score and tacrolimus trough level at 1 month. e Adjusted for the variables listed in footnote (d) except for underlying disease. f Adjusted for the variables listed in footnote (d) except for posttransplant ARF. g Adjusted for the variables listed in footnote (d) except for Child Pugh score. h Adjusted for the variables listed in footnote (d) except for tacrolimus trough level at 1month. Our results revealing old age, posttransplant ARF, low pretransplant egfr, pretransplant hepatorenal syndrome, pretransplant proteinuria, high CNI trough levels and use of cyclosporine as risk factors of CKD after liver transplantation corresponded with the previous studies [1,3,4]. As with Ojo et al. [1], this study presents cyclosporine as a significant risk factor for CKD in the multivariate Cox regression analysis. Although tacrolimus may be less nephrotoxic by diminished vasoconstriction and by reduced expression of cytokines such as TGF-β [24], the beneficial effect of tacrolimus is still controversial [25 27]. Preoperative MELD score or Child Pugh score have been known to be related with recovery from hepatorenal syndrome, posttransplant haemodialysis and early acute renal failure [13,18,28]. But the finding that the severity of liver disease is related with development of CKD after liv-

12 CKD after liver transplantation 2783 Table 5. Characteristics and histopathologies of the recipients conducted kidney biopsy after liver transplantation Cases Sex/ Underlying age b disease Time a from Tx to Bx Type of CNIs Pre-Tx DM Spot urine protein/cr (mg/dl) Cr (mg/dl) Time a from Bx to lowest Pre-Tx Bx Lowest c Cr Histopathology of kidney Glob. Scls (%) Seg. Scls (%) T-I fibrosis Arteriopathy Arteriolopathy Diabetic lesion Chronic CNI arteriopathy TMA IgAN Management Result 1 M/48 HBV LC 21.6 Tac CNI withdrawal d Improved 2 M/46 HBV LC 16.5 Tac CNI withdrawal d Improved 3 M/50 HBV HCC 11.3 Tac nf 2 CNI withdrawal d Improved 4 M/46 HBV LC 4.7 Tac Steroid dose and CNI withdrawal d 5 F/41 HBV LC 77.9 Tac CNI withdrawal d Improved 6 M/51 HBV LC 35.6 CsA CNI withdrawal d Progression to ESRD 7 M/51 HBV LC 5.8 CsA CNI withdrawal d Improved 8 F/55 HBV LC 4.7 Tac ARB addition Remission of proteinuria 9 M/47 HBV LC 9.4 Tac ARB addition Improved Improved tpl, transplantation; Bx, biopsy; CNIs, calcineurin inhibitors; DM, diabetes mellitus; Cr, creatinine; Glob. Scls, global sclerosis; Seg. Scls, segmental sclerosis; T-I, tubulointerstitial; TMA, thrombotic microangiopathy; IgAN, IgA nephropathy; HBV, hepatitis B virus; LC, liver cirrhosis; HCC, hepatocellular carcinoma; Tac, tacrolimus; nf, not found; CsA, cyclosporine; CNIs, calcineurin inhibitors; ARB, angiotensin receptor blocker. a Expressed in months. b At the time of transplantation. c Lowest value after kidney biopsy. d Mycophenolate mofetil addition with CNI dose reduction or mycophenolate mofetil monotherapy.

13 2784 J.P. Lee et al. er transplantation is new. Hypoalbuminaemia and low haemoglobin, since they modify Starling s forces in the systemic capillaries, may deteriorate kidney perfusion, which is an important mechanism of CNI nephrotoxicity [29]. In addition, they can alter the pharmacokinetics of potentially nephrotoxic drugs such as CNIs through raising the level of the unbound form. Recently, there has been an effort to standardize the definition of ARF to the RIFLE criteria [12]. Association severity or cause of posttransplant ARF with CKD has never been reported. In this study, more severe posttransplant ARF causes CKD well. Ischaemic ATN on univariate analysis and CNI nephrotoxicity on univariate and multivariate analyses increased the risk of CKD compared with the prerenal ARF or no ARF group. Cause or severity of posttransplant ARF as well as occurrence of ARF are important for development of CKD in liver transplantation. HBV infection, compared to other causes of liver disease, favourably affected kidney outcome in this study. LikeHCV,HBVisalsoawell-knownpathogenthat can cause glomerulonephritis [30]. HCV is the most common cause of liver failure in Western countries, and it increases CKD risk in liver transplantation primarily because it may cause glomerulonephritis [1,6]. Until now, only three small-sized studies have exploited renal histology after liver transplantation. In the study by McGuire et al., 30 HCV-cirrhosis patients received renal biopsy after liver engraftment [5], in which immune-complex glomerulonephritis was common (83% of patients). All kidney biopsy cases in this report had liver disease caused by HBV infection, but none of the patients had MPGN or membranous GN, although the case number was small. One of the reasons for low incidence of CKD in Asians may be the high prevalence of HBV infection instead of HCV. A large-number renal biopsy study in HBV-related liver disease will be necessary in future investigations. The use of adefovir, which is known to be nephrotoxic in HBV recipients, did not increase the risk of CKD. Although our results are informative, our study has some limitations. First, we used creatinine-based GFR to assess renal function. Interference occurs during measurement of serum creatinine when patients have hyperbilirubinaemia. Therefore, assessing egfr in cirrhotic patients using any formula has limitations, even if previous several studies with similar design have also used the MDRD formula [1,3,31,32]. Ideally, either inulin clearance or isotopic clearance is necessary for assessing the true measure of GFR [33]. Second, the study design was retrospective and did not involve protocol kidney biopsy. In conclusion, the judicious use of CNIs during the early posttransplant period could preserve kidney function. Additionally, we suggest that the prevention of posttransplant ARF and correction of anaemia would improve kidney outcome. Acknowledgements. This study was supported by a grant of the Korea Health 21 R&D project, Ministry of Health and Welfare, Republic of Korea (03-PJ10-PG13-GD ). Conflict of interest statement. None declared. Supplementary data Supplementary data is available online at oxfordjournals.org. References 1. Ojo AO, Held PJ, Port FK et al. Chronic renal failure after transplantation of a nonrenal organ. N Engl J Med 2003; 349: O'Riordan A, Wong V, McCormick PA et al. Chronic kidney disease post-liver transplantation. Nephrol Dial Transplant 2006; 21: Pawarode A, Fine DM, Thuluvath PJ. Independent risk factors and natural history of renal dysfunction in liver transplant recipients. Liver Transpl 2003; 9: Guitard J, Ribes D, Kamar N et al. Predictive factors for chronic renal failure one year after orthotopic liver transplantation. Ren Fail 2006; 28: McGuire BM, Julian BA, Bynon JS Jr et al. Brief communication: glomerulonephritis in patients with hepatitis C cirrhosis undergoing liver transplantation. Ann Intern Med 2006; 144: Tsui JI, Vittinghoff E, Shlipak MG et al. Association of hepatitis C seropositivity with increased risk for developing end-stage renal disease. Arch Intern Med 2007; 167: de Villa V, Lo CM. Liver transplantation for hepatocellular carcinoma in Asia. Oncologist 2007; 12: Kim SG, Kim HJ, Lee JP et al. Incidence and risk factors of renal dysfunction after liver transplantation in Korea. Transplant Proc 2004; 36: Pillebout E, Nochy D, Hill G et al. Renal histopathological lesions after orthotopic liver transplantation (OLT). Am J Transplant 2005; 5: O'Riordan A, Dutt N, Cairns H et al. Renal biopsy in liver transplant recipients. Nephrol Dial Transplant 2009; 24: Larsen K. Creatinine assay by a reaction-kinetic principle. Clin Chim Acta 1972; 41: Bellomo R, Ronco C, Kellum JA et al. Acute renal failure definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004; 8: R204 R Cabezuelo JB, Ramirez P, Rios A et al. Risk factors of acute renal failure after liver transplantation. Kidney Int 2006; 69: Arroyo V, Gines P, Gerbes AL et al. Definition and diagnostic criteria of refractory ascites and hepatorenal syndrome in cirrhosis. International Ascites Club. Hepatology 1996; 23: Gonwa TA, Mai ML, Melton LB et al. End-stage renal disease (ESRD) after orthotopic liver transplantation (OLTX) using calcineurin-based immunotherapy: risk of development and treatment. Transplantation 2001; 72: Cohen AJ, Stegall MD, Rosen CB et al. Chronic renal dysfunction late after liver transplantation. Liver Transpl 2002; 8: Velidedeoglu E, Bloom RD, Crawford MD et al. Early kidney dysfunction post liver transplantation predicts late chronic kidney disease. Transplantation 2004; 77: Sanchez EQ, Gonwa TA, Levy MF et al. Preoperative and perioperative predictors of the need for renal replacement therapy after orthotopic liver transplantation. Transplantation 2004; 78: Zuo L, Ma YC, Zhou YH et al. Application of GFR-estimating equations in Chinese patients with chronic kidney disease. Am J Kidney Dis 2005; 45: Jafar TH, Schmid CH, Levey AS. Serum creatinine as marker of kidney function in South Asians: a study of reduced GFR in adults in Pakistan. J Am Soc Nephrol 2005; 16: Imai E, Horio M, Nitta K et al. Estimation of glomerular filtration rate by the MDRD study equation modified for Japanese patients with chronic kidney disease. Clin Exp Nephrol 2007; 11: Kang YS, Han KH, Han SY et al. Characteristics of population with normal serum creatinine impaired renal function and: the validation

14 CKD after liver transplantation 2785 of a MDRD formula in a healthy general population. Clin Nephrol 2005; 63: Al Aly Z, Abbas S, Moore E et al. The natural history of renal function following orthotopic heart transplant. Clin Transplant 2005; 19: Artz MA, Boots JM, Ligtenberg G et al. Improved cardiovascular risk profile and renal function in renal transplant patients after randomized conversion from cyclosporine to tacrolimus. J Am Soc Nephrol 2003; 14: Ovuworie CA, Fox ER, Chow CM et al. Vascular endothelial function in cyclosporine and tacrolimus treated renal transplant recipients. Transplantation 2001; 72: Khanna A, Plummer M, Bromberek C et al. Expression of TGF-beta and fibrogenic genes in transplant recipients with tacrolimus and cyclosporine nephrotoxicity. Kidney Int 2002; 62: A comparison of tacrolimus (FK 506) and cyclosporine for immunosuppression in liver transplantation. The U.S. Multicenter FK506 Liver Study Group. N Engl J Med 1994; 331: Alessandria C, Ozdogan O, Guevara M et al. MELD score and clinical type predict prognosis in hepatorenal syndrome: relevance to liver transplantation. Hepatology 2005; 41: Campistol JM, Sacks SH. Mechanisms of nephrotoxicity. Transplantation 2000; 69: SS5 SS Johnson RJ, Couser WG. Hepatitis B infection and renal disease: clinical, immunopathogenetic and therapeutic considerations. Kidney Int 1990; 37: Kim DY, Lim C, Parasuraman R et al. Renal disease burden following liver transplantation. Transplant Proc 2006; 38: Cicinnati VR, Yu Z, Klein CG et al. Clinical trial: switch to combined mycophenolate mofetil and minimal dose calcineurin inhibitor in stable liver transplant patients assessment of renal and allograft function, cardiovascular risk factors and immune monitoring. Aliment Pharmacol Ther 2007; 26: Poge U, Gerhardt T, Stoffel-Wagner B et al. Calculation of glomerular filtration rate based on cystatin C in cirrhotic patients. Nephrol Dial Transplant 2006; 21: Received for publication: ; Accepted in revised form: