Objectives: In pediatric patients, renal dysfunction after living-donor liver transplant is a major issue that is difficult to evaluate. Recently, predictive equations for Japanese children have been introduced.
Materials and Methods: We conducted a retrospective study by prospectively collecting data on 26 patients under 16 years old who underwent living-donor liver transplant between June 2004 and March 2015. Serum creatinine and cystatin C levels were measured. Paired t tests and Bland-Altman plots were used to compare the following formulas for estimated glomerular filtration rate: the Schwartz formula and 3 formulas that were matched with Japanese children (polynomial, simple, and cystatin C formulas).
Results: Average estimated glomerular filtrations rates (in mL/min/1.73 m2) were 143.46, 122.90, 121.58, and 123.31 using the Schwartz, polynomial, simple, and cystatin C formulas, respectively. The estimated glomerular filtrations rate for biliary atresia was 141.53 ± 31.37 versus 109.95 ± 19.52 for other diseases, with significant differences only noted with the cystatin C formula. The formulas tailored for Japanese children showed significantly lower estimated glomerular filtrations rates than those obtained using the Schwartz formula (P < .01).
Conclusions: The use of formulas for measuring estimated glomerular filtrations rates that are based on race may allow early detection of deteriorating renal function.
Key words : Chronic kidney disease; Creatinine; Cystatin C; Renal function
Recent advances in surgical techniques, improvements in postoperative care, and administration of immunosuppressants have led to significantly improved prognosis after living-donor liver transplant in children.1 Because long-term survival has become more common in children, as in adults,2-6 renal dysfunction has become a major issue during the postoperative management of pediatric liver transplant.7-13 Measuring inulin clearance is the criterion standard for evaluating renal function, but this method is complex and is not performed at most medical facilities. Instead, serum creatinine has been used to estimate glomerular filtration rate (GFR).14 However, the use of serum cystatin C (CysC) to evaluate renal function has been recently advocated because of certain inaccuracies with the use of serum creatinine for estimating GFR in adults.14-24
The evaluation of renal function in children remains difficult because of varying age and physique and differences between sexes.22,25-29 The most widely used method for estimating GFR in pediatric patients has been the Schwartz formula30; however, recently, CysC has also been used.22
It has been highlighted that these predictive equations cannot be used in Japanese patients due to ethnic/racial differences. In response, recent revisions in the standard values of serum creatinine and CysC levels that are relevant to Japanese children have been published.25,26 This has led to the release of modified predictive equations using creatinine (polynomial and simple formulas) and CysC.27-29
The purpose of this study was to assess the extent of differences between the Schwartz formula and the other formulas that are relevant to Japanese children in the evaluation of renal function after liver transplant.
Materials and Methods
Study design and population
The data of pediatric patients who underwent living-donor liver transplant at the Fujita Health University Hospital between June 2004 and March 2015 were prospectively collected. The inclusion criteria were age < 16 years and ongoing follow-up. The collected data were then retrospectively examined.
In all cases, the donors were the participants' parents. The postoperative immunosuppressant protocol (Figure 1) was initiated with continuous intravenous infusion of cyclosporine or tacrolimus with methylprednisolone at a dose of 10 mg/kg on postoperative days 1 to 3, 5 mg/kg on postoperative days 4 to 6, and 2.5 mg/kg on postoperative days 7 to 9. Once the patients could receive enteral nutrition, each medication was switched to the oral form, and mycophenolate mofetil was initiated at a dose of 10 to 25 mg/kg. After surgery, prednisolone was continuously administered orally for 3 months at a dose of 0.5 mg/kg on postoperative days 10 to 12, 1 mg/kg on postoperative days 13 to 15, and 0.05 to 0.1 mg/kg thereafter. Mycophenolate mofetil administration was continued for 1 year after surgery. The trough serum levels were maintained at 40 to 60 μg/L for the cyclosporine group and at 2 to 6 μg/L for the tacrolimus group. In suspected cases of organ rejection, prednisolone was administered in small doses.
The creatinine and CysC levels, age at transplant, follow-up duration after transplant, age at the time of study, and height, weight, and body mass index of all participants were measured. The enzyme test was used to measure creatinine levels, and the latex coagulation test (Labospect 008, Hitachi, Tokyo, Japan) was used to measure CysC. The Schwartz formula30 and the 3 formulas relevant to Japanese children (polynomial, simple, and CysC formulas)27-29 were used to measure estimated glomerular filtration rate (eGFR) (Table 1). To determine chronic kidney disease (CKD), GFR categories as per the Kidney Disease Improving Global Outcomes were used (Table 2).31 We defined renal dysfunction as a GFR of < 90 mL/min/1.73 m2.
All calculations were performed using JMP version 10 statistical software (SAS Institute Inc., Cary, NC, USA). Data are expressed as means, standard deviation, and range. Paired t tests, Bland-Altman plots, and Mann-Whitney U tests were used to compare the groups. P < .01 was considered as significant.
In this analysis, 26 patients (12 males and 14 females) were included; their characteristics are shown in Table 3. The average age at transplant was 43.38 months (range, 7-160 mo), and the average duration after transplant was 61.07 months (range, 3-134 mo). There were 2 groups that received different immunosuppressants: the cyclosporine group (n = 8) and the tacrolimus group (n = 18).
The average eGFR values were 143.46 ± 48.59 mL/min/1.73 m2 using the Schwartz formula, 122.90 ± 37.87 mL/min/1.73 m2 using the polynomial formula, 121.58 ± 41.17 mL/min/1.73 m2 using the simple formula, and 123.31 ± 29.66 mL/min/1.73 m2 using the CysC formula. Comparisons among these eGFR values showed significantly lower values using each formula tailored for Japanese children versus results obtained using the Schwartz formula (P < .01; Table 4). Mean biases between the Schwartz formula and the other formulas were significantly different from 0 (P < .01). There were no differences among the formulas tailored for Japanese children (Table 5). The Bland-Altman plot is shown in Figure 2. There are significant differences between the formulas tailored for Japanese children and the Schwartz formula (P < .01). Using the Schwartz formula, we found 1 patient with CKD stage 2 (3.8%) and 1 patient with stage 3 (3.8%). According to the polynomial and simple formulas, there were 4 patients with CKD stage 2 (15.3%). According to the CysC formula, there were 3 patients with CKD stage 2 (11.5%) and 0 patients with CKD stage 3. None of the patients was classified as CKD stage 4 or 5 using any of the formulas (Table 6).
The median values of GFR using the CysC formula in biliary atresia and other diseases were 138.81 mL/min/1.73 m2 and 117.62 mL/min/1.73 m2, respectively. When we compared the average GFR values between patients with biliary atresia and patients with other diseases, only GFR measured using CysC showed a significant difference (P < .01); no significant differences were observed for GFR measured using serum creatinine levels (Table 7). When we compared patients according to follow-up duration after transplant (< 5 years vs ≥ 5 years), we found no significant differences in eGFR (Table 8).
The average cyclosporine dose was 3.31 ± 1.76 mg/kg/day (range, 1.03-6.57 mg/kg/d), and the average tacrolimus dose was 0.068 ± 0.044 mg/kg/day (range, 0.023-0.183 mg/kg/d). The average trough levels of the cyclosporine and tacrolimus groups were 69.98 ± 45.83 μg/L and 3.81 ± 1.69 μg/L, respectively. Although there were no significant differences in eGFR between the cyclosporine and tacrolimus groups, the cyclosporine group had a tendency toward a lower eGFR (Table 9).
There have been several studies on renal dysfunction after liver transplant.2-13 In recent years, the long-term survival of pediatric liver transplant patients has been increasing; consequently, more patients with renal dysfunction have been reported.7-13 The morbidity of renal dysfunction after pediatric liver transplant has been estimated to be 10% to 35% and continues to increase annually.7,9-12 Acute renal dysfunction was reportedly associated with perioperative fluid and cardiovascular management, as well as the use of loop diuretics and dopamine2,6; calcineurin inhibitors (CNIs) can also be a cause of chronic renal dysfunction.2-7,32,33 The adverse reactions of CNIs include increased susceptibility to infection, nephropathy, central nervous system symptoms, hyperglycemia, cardiotoxicity, and development of malignant tumors; in particular, nephropathy is one of the most significant adverse reactions.5,34-36 Introduction of mycophenolate mofetil and mammalian target of rapamycin inhibitors, as well as protocols for reducing or discontinuing CNIs, have been proposed to prevent CNI-induced renal dysfunction in both adults and children.37-46 However, although these measures have contributed to the improvement of renal function, cases of organ rejection have increased; therefore, a satisfactory outcome has not been achieved.
Although creatinine has been used to assess pediatric renal dysfunction at many medical facilities,4 differences in age and physique of children make it less reliable. In addition, children have little muscle mass, which may lead to overestimation of renal function.14 The modified Schwartz formula was released in 200930; however, it was reported to be unsuitable for use in Japanese children.27-29 By 2014, standard values of creatinine and CysC, as well as predictive equations, were tailored for Japanese children.25-29 It was previously thought that Japanese and other Asian people experience less renal dysfunction after liver transplant.4 However, our present study using predictive equations for Japanese children indicated an increased number of cases with renal dysfunction (GFR < 90 mL/min/1.73 m2). Because no other study has focused solely on Japanese children, these predictive equations should be used in future assessments of GFR.
Inulin clearance provides the most accurate assessment of GFR; however, this test is complex and difficult to conduct in all cases.4 Several studies have utilized radioisotopes to assess GFR,17,19-21,30 but this method should be used in cases of decreased GFR to investigate the validity of the predictive equations used in this study. Many studies have reported CysC as an effective serologic marker for renal dysfunction.14-16,18 Cystatin C is a cysteine protease inhibitor that is derived from nucleated cells throughout the body and is a basic protein with a molecular weight of 13 kDa. A specific amount of CysC is constantly produced extracellularly by all cells. It is not influenced by muscle mass and is mostly reabsorbed by the proximal tubules after glomerular filtration. Therefore, blood levels of CysC are thought to depend on GFR.8,15,17-20 Thus, the use of CysC may provide a more sensitive assessment of renal function versus the use of creatinine. Although CysC has been reported to be affected by thyroid disease, asthma, steroids, cyclosporins, and other factors,47,48 it has been shown to be effective when used in pediatric patients.8,20,22 We believe that its use will become widespread in the future.
In this study, no significant differences were demonstrated between the use of creatinine and CysC in the predictive equations for Japanese children in both paired t tests and Bland-Altman plots. The Bland-Altman plot showed that, compared with each formula tailored for Japanese children, the Schwartz formula overestimated renal function using creatinine and CysC. Several studies have indicated that evaluation of both creatinine and CysC may allow a more accurate assessment of GFR,14,16,21,23 and we concur with this conclusion. Because the predictive equation using creatinine is limited for patients who are from 2 to 19 years old,29 a predictive equation using CysC would be a more preferable choice for patients of any age.
This study had a retrospective observational design with a small number of patients; consequently, a lack of comparison with the true value of GFR was its limitation. In conclusion, the assessment of renal function using methods that are tailored for race/ethnicity can allow earlier detection of renal function deterioration.
Volume : 16
Issue : 6
Pages : 708 - 713
DOI : 10.6002/ect.2017.0159
From the the 1Department of Pediatric Surgery, Fujita Health University,
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare. A. Sugioka, Y. Kato, M. Tokoro, and N. Tanahashi performed surgery on the liver donor. T. Suzuki, F. Hara, S. Watanabe, N. Uga, A. Naoe, and Y. Kondo performed surgery on the liver recipient, participated in postoperative care, and collected data. T. Yoshikawa, Y. Kawamura, and H. Miura designed the study.
Corresponding author: Toshihiro Yasui, Department of Pediatric Surgery, Fujita Health University, 1-98 Dengakugakubo, Kutsukake-cho, Toyoake, Aichi 470-1192, Japan
Phone: +81 562 93 9247
Table 1. Formulas for Evaluating Renal Function in Children After Liver Transplant
Table 2. Chronic Kidney Disease Stages as per Kidney Disease Improving Global Outcomes According to Glomerular Filtration Rate
Table 3. Baseline Characteristics of Patients in Whom Renal Function Was Evaluated After Liver Transplant (N = 26)
Table 4. Comparison of the Different Predictive Formulas for Estimated Glomerular Filtration Rate
Table 5. Comparison of Each Estimated Glomerular Filtration Rate Using the Bland-Altman Plot
Table 6. Distribution of Chronic Kidney Disease Stages According to Glomerular Filtration Rate Estimated Using Different Formulas
Table 7. Comparison of Estimated Renal Function After Liver Transplantation in Pediatric Patients According to the Presence of Biliary Atresia
Table 8. Comparison of Estimated Renal Function in Pediatric Patients According to Follow-up Duration After Liver Transplant
Table 9. Comparison of Estimated Renal Function After Liver Transplant in Pediatric Patients According to Immunosuppressants Received
Figure 1. Protocol for the Initiation of Immunosuppressive Therapy After Liver Transplantation in Pediatric Patients
Figure 2. Statistical Analysis Using Bland-Altman Plot