Pretransplant Risk Factors for New-Onset Diabetes Mellitus After Transplant in Pediatric Liver Transplant Recipients

Transcription

1 LIVER TRANSPLANTATION 16: , 2010 ORIGINAL ARTICLE Pretransplant Risk Factors for New-Onset Diabetes Mellitus After Transplant in Pediatric Liver Transplant Recipients Hung-Tien Kuo, 1,2 Christine Lau, 1 Marcelo Santos Sampaio, 1,3 and Suphamai Bunnapradist 1 1 Department of Medicine, David Geffen School of Medicine, University of California, Los Angeles, CA; 2 Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, Kaohsiung, Taiwan; and 3 Department of Medicine, Pedro Ernesto University Hospital, Rio de Janeiro State University, Rio de Janeiro, Brazil Our objectives are to examine the incidence of new-onset diabetes mellitus after transplant (NODAT) and to identify its risk factors in pediatric liver transplant recipients using the Organ Procurement and Transplant Network/United Network for Organ Sharing database. Between July 2004 and December 2008, a total of 1214 children (2-20 years old) received their first liver transplant alone, and had at least 1 follow-up report of posttransplant diabetic status. Among these, 1161 recipients were identified as not having diabetes mellitus before transplant. Risk factors for NODAT were examined using classification and regression tree and multivariate Cox regression analysis. Diabetes mellitus was newly reported in 10.1% of the 1161 recipients over the median follow-up time of 770 days. The cumulative incidences of NODAT at 1, 2, and 3 years after transplant were 5.9%, 8.3%, and 11.2%, respectively. More than 50% of recipients with cystic fibrosis developed NODAT. In recipients without cystic fibrosis, independent risk factors for NODAT included increased recipient age (compared to 2-5 years, hazard ratio ¼ 3.09 for 5-13 years, p ¼ 0.02; 7.14 for 13 years, p < 0.001), African American race (1.97, p ¼ 0.003), and primary diagnosis of primary sclerosing cholangitis (2.24, p ¼ 0.02) and acute hepatic necrosis (1.89, p ¼ 0.04). In conclusion, NODAT occurred in one-tenth of pediatric liver transplant recipients in the United States during the median follow-up of 2 years. Some of the risk factors for NODAT in pediatric liver transplant recipients are similar to those reported in other solid organ transplants. Underlying liver disease of cystic fibrosis, primary sclerosing cholangitis, and acute hepatic necrosis are independent risk factors for NODAT in pediatric liver transplant recipients. Liver Transpl 16: , VC 2010 AASLD. Received April 29, 2010; accepted July 15, New-onset diabetes mellitus after transplant (NODAT) is an important complication occurring after solid organ transplantation. The incidence of NODAT was shown to be 26.4% in adult liver transplant recipients with a median follow-up time of 685 days in our recent analysis of the Organ Procurement and Transplant Abbreviations: AA, African American; AHN, acute hepatic necrosis; Ale, alemtuzumab; BMI, body mass index; CNI, calcineurin inhibitors; CsA, cyclosporine A; egfr, estimated glomerular filtration rate; ESLD, end-stage liver disease; HCV, hepatitis C virus; HR, hazard ratio; ICU, intensive care unit; IL, interleukin; MMF, mycophenolate mofetil; MPA, mycophenolic acid; mtori, mammalian target of rapamycin inhibitor; NODAT, new-onset diabetes after transplant; NODM, new-onset diabetes mellitus; OPTN/UNOS, Organ Procurement and Transplant Network/United Network for Organ Sharing; PSC, primary sclerosing cholangitis; Tac, tacrolimus. Data reported here was taken from the Organ Procurement Transplant Network/United Network for Organ Sharing (OPTN/UNOS) as of November This work was supported in part by Health Resources and Services Administration contract C. The content is responsibility of the authors alone and does not necessarily reflect the views of policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. Participation in this work by M.S.S. has been made possible through a fellowship funded by the International Society of Nephrology. Address reprint requests to Suphamai Bunnapradist, M.D., 1033 Gayley Avenue, Suite 208, Los Angeles, CA Telephone: ; FAX: ; bunnapradist@mednet.ucla.edu. DOI /lt View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI /lt. Published on behalf of the American Association for the Study of Liver Diseases VC 2010 American Association for the Study of Liver Diseases.

2 1250 KUO ET AL. LIVER TRANSPLANTATION, November 2010 Network/United Network for Organ Sharing (OPTN/ UNOS) database. 1 Risk factors for NODAT in the studied population included increased age, African American (AA) race, high body mass index (BMI), hepatitis C virus (HCV) infection, recipient cirrhosis history, increased donor age, diabetic donor, and immunosuppressive regimen. In adult liver recipients, NODAT has been shown to be associated with impaired long-term graft function and reduced survival. 2 Studies evaluating the incidence and risk factors associated with NODAT in the pediatric transplant population have focused mostly on heart and kidney recipients. 3-5 Our previous study of pediatric renal recipients with NODAT identified from the OPTN/ UNOS database showed an incidence of 4.6%. 5 The information about incidence and risk factors of NODAT in pediatric liver transplant recipients are limited. 6 Since June 30, 2004, the OPTN/UNOS has been collecting data on the presence of diabetes during follow up via recipient follow-up form. The objectives of our study are to identify the incidence of NODAT and donor-, recipient-, and transplant-related factors associated with development of NODAT in the recent cohort of pediatric liver recipients in the United States using the OPTN/UNOS databases. PATIENTS AND METHODS OPTN/UNOS has been collecting data on the presence of diabetes through the recipient follow-up forms since June 30, A total of 1701 pediatric patients (2-20 years old) who received their first liver transplant between July 1, 2004, and December 31, 2008, were identified in the OPTN/UNOS database as of November 9, Multiorgan recipients (n ¼ 180), retransplant recipients (n ¼ 215), and those who did not have at least one follow-up report with diabetes status (n ¼ 92) were excluded. Among 1214 pediatric liver recipients, those with documented pretransplant diabetes (n ¼ 26 or 2.1%) or unknown status (n ¼ 27 or 2.2%) were excluded. A total of 1161 (95.7%) nondiabetic recipients were identified into the study population, of whom 117 (10.1%) developed NODAT during the median follow-up period of 770 days after transplant (25%, 380 days; 50%, 770 days; 75%, 1431 days). All work was performed in accordance with the regulations of UCLA s Institutional Review Board. The presence of diabetes mellitus prior to transplantation was identified according to reports to OPTN/ UNOS if diabetes was listed as a comorbid condition (coded as 2 through 5 in DIAB). The presence of NODAT was identified in nondiabetic recipients (coded as 1 in DIAB) if they had at least one record of diabetes after liver transplant on the follow-up reports (DIA- BETES_DURING in the follow-up files). The date of onset of NODAT was assumed to be the date of the earliest report of diabetes after liver transplant. The correlation of NODAT with recipient-, donor-, and transplant-related factors was analyzed. Recipient-related factors included sex, age, BMI percentile (calculated from age, sex, body weight, and body length using growth chart program from the U.S. Centers for Disease Control and Prevention), race, presence of HCV infection, underlying cause of end-stage liver disease (ESLD), status 1/1A, medical condition at registration (hospitalized in the intensive care unit [ICU], non-icu hospitalized, or not hospitalized), serum bilirubin level at transplant, and estimated glomerular filtration rate at transplant. Donor-related factors included sex, age, and donor type (living or deceased). Transplant-related factors included cold ischemia time, induction regimen, choice of calcineurin inhibitor (tacrolimus [Tac] or cyclosporine A [CsA]) at discharge, mycophenolate mofetil (MMF)/mycophenolic acid (MPA) use at discharge, mammalian target of rapamycin inhibitor (mtori) use at discharge, and steroid use at discharge. Continuous variables, including age, BMI percentile, serum bilirubin level, estimated glomerular filtration rate, and cold ischemia time were categorized because their effects on outcome were not linear. Baseline recipient, donor, and transplant factors according to the presence of NODAT were compared using the chi-squared test or Fisher s exact test. The Kaplan-Meier product limit method was used to estimate the survival rate. The log-rank test was used for comparison of the survival curves. Risk algorithms for NODAT were developed by using classification and regression tree (CART) analysis. 7 CART is a nonparametric method of identifying predictor variables by using binary recursive partitioning: subsets of recipients were formed by examining each possible cut point of each variable to identify the cut point that resulted in maximum discrimination between subgroups of recipients with respect to the probability of NODAT development during follow-up. Significant predictors of NODAT (p < 0.05) were included in the algorithms. Univariate and multivariate analysis for NODATfree survival were performed in recipients who did not have cystic fibrosis, using the Cox proportional hazard model. The primary endpoint was time until reporting of diabetes during follow-up, as a time-varying event: P(T > t) ¼ S(t) where P is the probability of a patient whose graft is functioning through time (T) without the development of NODAT (t) and S is the survival function. Recipients were followed until graft loss, patient death, retransplant, or loss to follow-up. Potential confounding factors in the univariate analysis (p < 0.10) were analyzed in the multivariate analysis. All reported p values were 2-tailed, and a p value < 0.05 was considered statistically significant. All analyses were performed using STATA Statistical Software, Release 10 (StataCorp LP, College Station, TX). RESULTS A total of 1161 nondiabetic pediatric liver recipients was identified in the OPTN/UNOS databases and received their first transplant from July 1, 2004, until December 21, Of these 1161 patients, 117

3 LIVER TRANSPLANTATION, Vol. 16, No. 11, 2010 KUO ET AL Figure 1. NODAT-free survival in pediatric liver transplant recipients, 95% confidence intervals are shown. The cumulative incidence of NODAT development at 1, 2, and 3 years after transplant was 5.9%, 8.3%, and 11.2%, respectively. (10.1%) developed NODAT during the median followup period of 770 days after transplant (25%, 380 days; 50%, 770 days; 75%, 1431 days). Figure 1 shows the Kaplan-Meier survival curve for NODATfree survival in the study population in relation to months after transplant. The cumulative incidence of NODAT development at 1 year, 2 years, and 3 years after transplant was 5.9%, 8.3%, and 11.2%, respectively. The prevalence of NODAT according to recipient, donor, and transplant factors is presented in Table 1. A significantly higher proportion of patients did develop NODAT if they were older in age; were of AA race; had ESLD secondary to acute hepatic necrosis (AHN), cirrhosis, primary sclerosing cholangitis (PSC), or cystic fibrosis; or had serum bilirubin 3 mg/dl. Recipient sex, donor sex, donor age, and donor type (living versus deceased) did not influence the rates of NODAT. The percentage of NODAT development in recipients with and without HCV infection was 15.2% and 9.9%, respectively. However, the difference was not statistically significant. None of the transplant factors conferred any significant difference in the development of NODAT, including the choice of immunosuppressive regimen. The associations of NODAT with posttransplant factors are shown in Table 2. The percentage of NODAT development in those who developed acute rejection with steroid treatment in 1 year was higher than that in those with other treatment, and recipients who were free of acute rejection. Maintenance steroid treatment at 1 year was not significantly associated with NODAT development. For posttransplant BMI percentile, the report time is, on average, 6 months after transplant. Interestingly, those who were underweight after undergoing transplantation tended to have a higher percentage of NODAT development. However, the percentage of missing value is more than 50%, and hence posttransplant BMI was not included into subsequent analysis. The results of CART analysis are shown in Fig. 2. Identified predictors (nodes) for NODAT development (p < 0.05) include: cystic fibrosis, age of 13 years or greater, AA race, hospitalized at liver transplant, AHN, and cirrhosis as cause of ESLD. NODAT-free survival curves according to the 3 major risk factors for development of NODAT (race, age, and causes of ESLD) are shown in Figs. 3 to 5. The unadjusted NODAT-free survival rates at 12, 24, and 36 months for recipients with AA race were 91.2%, 86.2% and 79.1%; for other races were 94.7%, 92.8% and 90.7% (p < 0.001; Fig. 3). The unadjusted NODAT-free survival rates at 12, 24, and 36 months for recipients 2-5 years old were 99.2%, 98.7%, and 98.7%; for 5-13 years old were 96.7%, 94.4%, and 91.0%; for age 13 years old were 88.8%, 85.0%, and 81.1% (p < 0.001; Fig. 4). Primary diagnosis for ESLD was associated with development of NODAT in our study population. The unadjusted NODAT-free survival rates at 12, 24, and 36 months for recipients with AHN were 89.6%, 86.1%, and 86.1%; for those with cirrhosis were 93.4%, 91.8%, and 80.9%; for those with PSC were 90.3%, 86.2%, and 77.7%; for those with cystic fibrosis were 59.7%, 49.8%, and 42.6%; for those with other causes of ESLD were 97.5%, 95.7%, and 94.3% (p < 0.001; Fig. 4). The results of univariate and multivariate Cox regression analysis calculating hazard ratios (HRs) for development of NODAT in non cystic fibrosis recipients are shown in Table 3. Significant risk factors in the univariate analysis included recipient age 5-13 years old (versus 2-5 years old), 13 years old (versus 2-5 years old), AA race (versus others), cause of ESLD (AHN, cirrhosis, and PSC; versus others), hospitalized at transplant (versus nonhospitalized), bilirubin at transplant 3 mg/dl (versus <3 mg/dl), estimated glomerular filtration rate 60 ml/minute at transplant (versus <60 ml/minute), and donor age 40 years old (versus <20 years old). Independent risk factors in the multivariate analysis were found to include recipient age 5-13 years old (versus 2-5 years old), 13 years old (versus 2-5 years old), AA race, and ESLD secondary to AHN and PSC. Cirrhosis was associated with a modest but statistically insignificant trend (p ¼ 0.06) toward a higher risk of NODAT. DISCUSSION The incidence of NODAT in pediatric liver recipients was found to be 10.1% with a median follow-up of 770 days after transplant. This result is similar to a previous study by Hathout et al. that looked at posttransplant diabetes/glucose intolerance in pediatric liver patients who showed an incidence of 13%. 6 However, these rates in pediatric liver recipients are higher than those found in pediatric renal recipients in our

4 1252 KUO ET AL. LIVER TRANSPLANTATION, November 2010 TABLE 1. Prevalence of NODAT According to Baseline Recipient, Donor, and Transplant Factor Factors Total (%) Percent with NODAT p Values Recipient Factors Sex Male 599 (51.6%) 10.0% 0.9 Female 562 (48.4%) 10.2% Age 2-5 years 278 (23.9%) 4.3% < years 413 (35.6%) 7.0% 13 years 470 (40.5%) 17.7% BMI percentile Underweight (<5%) 91 (7.9%) 11.0% 0.9 Normal (5%-85%) 698 (60.5%) 10.5% Overweight (85%-95%) 193 (16.7%) 9.3% Obesity (95%) 172 (14.9%) 8.7% Race Caucasian 676 (58.2%) 8.3% AA 192 (16.6%) 17.1% Hispanic 221 (19.0%) 9.5% others 72 (6.2%) 8.3% Cause of ESLD AHN 230 (19.8%) 14.4% <0.001 Cirrhosis 152 (13.1%) 14.5% PSC 71 (6.1%) 18.3% Biliary atresia 197 (17.0%) 3.6% Cystic fibrosis 26 (2.2%) 57.7% Metabolic disease 160 (13.8%) 4.4% Malignancy 118 (10.2%) 3.4% others 207 (17.8%) 7.7% HCV infection Yes 33 (2.8%) 15.2% 0.3 No/unknown 1128 (97.2%) 9.9% Status 1/1A Yes 333 (28.7%) 10.8% 0.6 No 828 (71.3%) 9.8% Medical condition In ICU 307 (26.4%) 13.0% 0.06 Hospitalized, non-icu 124 (10.7%) 12.1% Nonhospitalized 730 (62.9%) 8.5% Serum bilirubin <2 mg/dl 378 (32.6%) 6.4% mg/dl 102 (8.8%) 7.8% 3 mg/dl 681 (58.6%) 12.5% egfr 60 ml/min/1.73 m 2 81 (7%) 9.8% 0.3 <60 ml/min/1.73 m 2 1,080 (93%) 13.6% Donor Factors Sex Female 485 (41.8%) 9.3% 0.4 Male 676 (58.2%) 10.7% Age <20 years 724 (62.4%) 8.8% years 279 (24.0%) 11.5% 40 years 158 (13.6%) 13.3% Donor type Living 100 (8.6%) 8% 0.5 Deceased 1061 (91.4%) 10.3% Diabetic donor Yes 18 (1.6%) 11.1% 0.9 No 1143 (98.4%) 10.1% Transplant Factors Cold ischemia time <6 hours 346 (29.8%) 10.1% hours 624 (53.7%) 10.3% 12 hours 74 (6.4%) 5.4% Unknown 117 (10.1%) 12.0% Induction None 812 (75.2%) 9.6% 0.2 Thymoglobulin 75 (6.5%) 12% IL-2 inhibitors 192 (16.6%) 9.4% Ale 21(1.8%) 23.8% CNI CsA 29 (2.5%) 13.8% 0.9 Tac 1091 (94.8%) 9.9% None 31 (2.7%) 12.9% MMF/MPA Yes 492 (42.4%) 11.4% 0.2 No 669 (57.6%) 9.1% mtori Yes 19 (1.6%) 5.3% 0.5 No 1142 (98.4%) 10.2% Steroid Yes 1063 (91.6%) 10.3% 0.5 No 98 (8.4%) 8.2%

5 LIVER TRANSPLANTATION, Vol. 16, No. 11, 2010 KUO ET AL TABLE 2. Association of Posttransplant Factors with NODAT Development Factor Total (%) Percent with NODAT p Values Acute rejection in 1 year Yes, steroid treatment 106 (9.1%) 18.9% Yes, other/no treatment 210 (18.1%) 12.9% No/missing 845 (59.8%) 8.3% Maintenance steroid in 1 year Yes 940 (81.0%) 10.2% 0.8 No/missing 221 (19.0%) 9.5% BMI percentile at 6 months Underweight (<5%) 41 (3.5%) 17.1% 0.06 Normal (5%-85%) 281 (24.2%) 12.8% Overweight (85%-95%) 93 (8.0%) 5.4% Obesity (95%) 92 (7.9%) 5.4% Unknown/missing 654 (56.4%) 9.8% Figure 2. Risk algorithms for developing NODAT in the CART analysis. prior study that evaluated the OPTN/UNOS database. In the pediatric renal study, NODAT was reported in 4.6% of patients with a median follow-up of 693 days. 5 In addition, we demonstrated that the incidence of pretransplant diabetes (2.1%) in pediatric liver recipients is higher than in the general pediatric population. Results from the National Health and Nutrition Examination Survey (NHANES) from estimated that 0.5% of U.S. adolescents years of age have diabetes, and 29% of these children have type 2 diabetes mellitus. 8 Thus, the patients underlying liver disease may be contributing to the higher incidence of diabetes as compared to pediatric renal recipients and the general pediatric population. Cystic fibrosis was the cause of ESLD in 2.2% of the pediatric recipients studied. Diabetes is the most common comorbid condition in patients with severe cystic fibrosis, and 50%-70% develop diabetes by the age of 30 through mechanisms still unidentified but thought to be related to destruction of beta cells from exocrine pancreas malfunction, malnutrition, chronic inflammation, infection, and exogenous glucocorticoids. 9 Based on our results showing that 57.7% of patients with liver failure from cystic fibrosis developed NODAT during follow-up and their elevated risk of diabetes at baseline, patients with cystic fibrosis should be more frequently and carefully screened for diabetes after undergoing transplantation. PSC was also found to be an independent risk factor for NODAT. PSC was the cause of ESLD in 6.1% of the pediatric recipients studied. In the multivariate analysis, PSC was associated with a 1.26-fold increased risk for NODAT. PSC is a chronic cholestatic liver disease of unknown origin that mostly affects male patients with inflammatory bowel disease. The immune system is believed to be involved in the etiology and pathogenesis, and PSC has been shown to be associated with autoimmune diseases, including type 1 diabetes. A human leukocyte antigen haplotype found in both PSC and type 1 diabetes could be the bond linking the two diseases. 10 Based on our results, we suggest that close monitoring for

6 1254 KUO ET AL. LIVER TRANSPLANTATION, November 2010 NODAT is indicated in pediatric liver recipients with PSC. Compared to adult liver recipients, the incidence of NODAT in pediatric liver recipients was much lower (26.4% in adults versus 10.1% in children). It may be partially explained by the difference in underlying liver disease between adults and children. Pretransplant Figure 3. NODAT-free survivals of recipient according to recipient race (AA, African American race; and other races). The survival difference was statistically significant. (log-rank test, p < 0.001) Figure 4. NODAT-free survivals of recipient according to recipient age (2-5, 5-13, and 13 years). The survival difference was statistically significant. (log-rank test, p < 0.001) TABLE 3. Hazard Ratios for NODAT Development in Non Cystic Fibrosis Recipients Univariate Analysis Multivariate analysis Reference HR (95% CI) p Value HR (95% CI) p Value Recipient age, 5-13 yrs 2-5 years 3.32 ( ) ( ) yrs 2-5 years 10.1 ( ) < ( ) <0.001 BMI percentile <5% 5%-85% 1.06 ( ) % 5%-85% 1.07 ( ) % 5%-85% 0.96 ( ) 0.9 AA Race Other races 2.61 ( ) < ( ) Cause of ESLD AHN Other diseases 3.02 ( ) < ( ) 0.04 Cirrhosis Other diseases 3.16 ( ) < ( ) 0.06 PSC Other diseases 3.96 ( ) < ( ) 0.02 HCV Non-HCV 1.67 ( ) 0.2 Hospitalized Nonhospitalized 1.60 ( ) ( ) 0.4 Bilirubin at transplant 3 mg/dl <3 mg/dl 2.38 ( ) ( ) 0.4 egfr (ml/min) at transplant < ( ) ( ) 0.2 Living donor Deceased 0.53 ( ) 0.2 Donor age years <20 years 1.42 ( ) ( ) years <20 years 1.95 ( ) ( ) 0.9 Medication at discharge Induction No induction 1.14 ( ) 0.6 Tac CsA 0.66 ( ) 0.4 MMF/MPA No MMF/MPA 1.41 ( ) ( ) 0.5 mtori No mtori 0.52 ( ) 0.5 Steroid No steroid 1.64 ( ) 0.2 Potential covariates (p < 0.1) in the univariate analysis were analyzed in the multivariate analysis. Values in bold indicate statistical significance.

7 LIVER TRANSPLANTATION, Vol. 16, No. 11, 2010 KUO ET AL Figure 5. NODAT-free survivals of recipient according to underlying liver disease (AHN, cirrhosis, PSC, cystic fibrosis, and other diseases). The survival difference was statistically significant. (log-rank test, p < 0.001) HCV infection is common in adult liver recipients (48.9% in our previous study using the OPTN/UNOS database) and it conferred a higher risk of NODAT. 1 On the other hand, HCV infection is relatively rare in children. In the current study, pretransplant HCV infection was reported in 2.8% of pediatric liver recipients. The reported rate of NODAT in recipients with and without HCV infection was 15.2% and 9.9%, respectively. However, the difference was not statistically significant. The small number of recipients with HCV infection may contribute to the insignificant result. The association of chronic liver disease with impaired glucose metabolism is well known. 11 In our previous study, cirrhosis was found to be associated with an increased risk of NODAT in adult liver transplant recipients. 1 In the pediatric cohort, the percentage of cirrhosis is 13%, which was lower than that of 56.7% in adult liver recipients. In the current study, cirrhosis was associated with a 75% increased risk of NODAT in non cystic fibrosis recipients, with a p value of 0.06 after adjusting for other factors. Older recipient age at transplant was an independent risk factor for NODAT. This finding is consistent with other studies that report older age as a risk factor in pediatric transplant recipients. 3,5,6 This may be related to the higher rates of type 2 diabetes in the adolescent population, especially among AA, Hispanic, and Native Americans, compared to younger children, given the increasing rates of obesity and metabolic syndrome. 12 AA race was found to be an independent risk factor with a 0.97-fold increased risk of developing NODAT after pediatric liver transplant. In our previous study in adult liver recipients, AA race was also found to be an independent risk factor for development of NODAT. 1 Although inherent insulin resistance has been thought to account for the higher rates of diabetes in AA youth, a recent study showed that obesity was the most important risk factor for insulin resistance, independent of ethnicity. 13 Obesity has been shown to be an independent risk factor for NODAT in solid organ transplantation. 1,5,14 Interestingly, our study found that BMI percentile was not associated with an increased risk for NODAT. However, BMI may be overestimated in patients with liver failure due to their ascites. In a report by Leonard et al., they stated that correcting for ascites results in 11%-20% of patients moving into a lower BMI classification. 15 The impact of BMI percentile on NODAT may be therefore underestimated in the current study. Posttransplant BMI percentile, measured after improvement of ascites and malnutrition, may be more appropriate than pretransplant BMI percentile in the analysis of NODAT. However, further analysis was limited by the high missing value proportion of posttransplant BMI percentile in the OPTN/UNOS database. Serum bilirubin 3 mg/dl was also found to confer an increased risk for developing NODAT in the univariate analysis but not in the multivariate analysis. In our study population, pretransplant bilirubin level was significantly correlated with underlying liver disease. The percentage of recipients with bilirubin 3 mg/dl was 91.7% in those with AHN, 67.1% in those with cirrhosis, 70.4% in those with PSC, 61.4% in those with biliary atresia, 38.5% in those with cystic fibrosis, 34.4% in those with metabolic disease, 7.6% in those with malignancy, and 59.4% in those with other diseases (chi-squared test, p < 0.001). After adjusting for other confounding factors, including underlying liver disease, hyperbilirubinemia was not an independent risk factor for developing NODAT. Interestingly, previous studies have shown that higher bilirubin levels may have protective effects on atherosclerosis through its anti-inflammatory properties. 16,17 Analysis of NHANES (National Health and Nutrition Examination Survey) data from showed that serum bilirubin levels were inversely related to the prevalence of metabolic syndrome in adolescents aged years, 18 but this cohort did not have liver failure. The effects of elevated bilirubin on metabolic syndrome in pediatric patients with liver failure need further research. A specific immunosuppressive regimen may influence the development of NODAT following solid organ transplantation. Our previous study in adult liver recipients showed that tacrolimus (versus cyclosporine A [CsA], HR ¼ 1.24) and steroid at discharge (HR ¼ 1.59) were associated with an increased risk for developing NODAT, whereas induction therapy was associated with an decreased risk (HR ¼ 0.82). 1 However, the risk of developing NODAT was not significantly influenced by choice of immunosuppression at discharge in the current study of a pediatric population. Because dosages of medications were not recorded in the OPTN/ UNOS database, their impacts on development of NODAT cannot be analyzed. In our study, the

8 1256 KUO ET AL. LIVER TRANSPLANTATION, November 2010 prevalence of NODAT was higher in those who developed acute rejection in 1 year and received steroid treatment (18.9%), compared to that in those who received other treatment (12.9%) and that in recipients free of acute rejection (8.3%). The high steroid dosage in treating acute rejection may contribute to the higher NODAT rate. Because the cause and effect relationship between acute rejection and NODAT cannot be clarified in the OPTN/UNOS database, we do not include acute rejection in the multivariate Cox regression model. Limitations in this study include those that are inherent with retrospective analyses of databases, such as reporting bias or error. The definition of diabetes also may vary among transplant centers. In addition, the date of NODAT reported to OPTN/UNOS may not have been the actual date of onset. Steroid-related hyperglycemia and diabetes related to metabolic syndrome cannot be distinguished in the current analysis. Resolution of new-onset diabetes mellitus (NODM) after the manipulation of immunosuppressive medication, weight control, and lifestyle changes was not captured in our analysis. Because the diagnosis of diabetes in the current study was based on the diagnosis reported on the UNOS form by the transplant physicians, it may not have captured some patients who may have fulfilled the more rigorous criteria of diabetes as defined by the World Health Organization/ American Diabetes Association. It is possible that the UNOS data may have underreported the incidence of NODAT, although these patients would fall in the milder category of hyperglycemia In conclusion, NODAT was reported in one-tenth of pediatric liver recipients in the United States with a median follow-up of 2 years after transplant. Identified independent risk factors include older age, AA race, and specific cause of ESLD. More than 50% of recipients with cystic fibrosis developed NODAT during follow-up. In recipients without cystic fibrosis, presence of AHN and PSC conferred a higher risk. Screening for NODAT is recommended in pediatric liver recipients, especially in those with increased risk. REFERENCES 1. Kuo HT, Sampaio MS, Ye X, Reddy P, Martin P, Bunnapradist S. Risk factors for new-onset diabetes mellitus in adult liver transplant recipients, an analysis of the Organ Procurement and Transplant Network/United Network for Organ Sharing database. Transplantation 2010;89: Pageaux GP, Faure S, Bouyabrine H, Bismuth M, Assenat E. Long-term outcomes of liver transplantation: diabetes mellitus. Liver Transpl 2009;15(Suppl 2):S79-S Hathout EH, Chinnock RE, Johnston JK, Fitts JA, Razzouk AJ, Mace JW, et al. Pediatric post-transplant diabetes: data from a large cohort of pediatric heart-transplant recipients. Am J Transplant 2003;3: Al-Uzri A, Stablein DM, Cohn AR. Posttransplant diabetes mellitus in pediatric renal transplant recipients: a report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS). Transplantation 2001;72: Kuo HT, Poommipanit N, Sampaio M, Reddy P, Cho YW, Bunnapradist S. Risk factors for development of newonset diabetes mellitus in pediatric renal transplant recipients: an analysis of the OPTN/UNOS database. Transplantation 2010;89: Hathout E, Alonso E, Anand R, Martz K, Imseis E, Johnston J, et al. Post-transplant diabetes mellitus in pediatric liver transplantation. Pediatr Transplant 2009;13: Breiman L, Friedman J, Olshen R, Stone C. Classification and Regression Trees. Pacific Grove, CA: Wadsworth; Duncan GE. Prevalence of diabetes and impaired fasting glucose levels among US adolescents: National Health and Nutrition Examination Survey, Arch Pediatr Adolesc Med 2006;160: O Riordan SM, Dattani MT, Hindmarsh PC. Cystic fibrosis-related diabetes in childhood. Horm Res Paediatr 2010;73: Saarinen S, Olerup O, Broome U. Increased frequency of autoimmune diseases in patients with primary sclerosing cholangitis. Am J Gastroenterol 2000;95: Picardi A, D Avola D, Gentilucci UV, Galati G, Fiori E, Spataro S, et al. Diabetes in chronic liver disease: from old concepts to new evidence. Diabetes Metab Res Rev 2006;22: De Ferranti SD, Osganian SK. Epidemiology of paediatric metabolic syndrome and type 2 diabetes mellitus. Diab Vasc Dis Res 2007;4: Lee JM, Okumura MJ, Davis MM, Herman WH, Gurney JG. Prevalence and determinants of insulin resistance among U.S. adolescents: a population-based study. Diabetes Care 2006;29: Shah T, Kasravi A, Huang E, Hayashi R, Young B, Cho YW, et al. Risk factors for development of new-onset diabetes mellitus after kidney transplantation. Transplantation 2006;82: Leonard J, Heimbach JK, Malinchoc M, Watt K, Charlton M. The impact of obesity on long-term outcomes in liver transplant recipients-results of the NIDDK liver transplant database. Am J Transplant 2008;8: Bulmer AC, Blanchfield JT, Toth I, Fassett RG, Coombes JS. Improved resistance to serum oxidation in Gilbert s syndrome: a mechanism for cardiovascular protection. Atherosclerosis 2008;199: Vitek L, Jirsa M, Brodanova M, Kalab M, Marecek Z, Danzig V, et al. Gilbert syndrome and ischemic heart disease: a protective effect of elevated bilirubin levels. Atherosclerosis 2002;160: Lin LY, Kuo HK, Hwang JJ, Lai LP, Chiang FT, Tseng CD, et al. Serum bilirubin is inversely associated with insulin resistance and metabolic syndrome among children and adolescents. Atherosclerosis 2009;203: