Nonalcoholic fatty liver disease (NAFLD) is rapidly

Transcription

1 Lifestyle Intervention and Antioxidant Therapy in Children with Nonalcoholic Fatty Liver Disease: A Randomized, Controlled Trial Valerio Nobili, 1 Melania Manco, 1 Rita Devito, 2 Vincenzo Di Ciommo, 3 Donatella Comparcola, 1 Maria Rita Sartorelli, 1 Fiorella Piemonte, 4 Matilde Marcellini, 1 and Paul Angulo 5 No proven treatment exists for nonalcoholic fatty liver disease (NAFLD) in children and adolescents. We sought to determine the efficacy of lifestyle intervention with or without antioxidant therapy in pediatric NAFLD. A total of 53 patients (age years, 37 boys) were included. Lifestyle intervention consisting of a diet tailored to the patient s calorie needs, and increased physical activity was prescribed in all. Patients were concomitantly randomized to alpha-tocopherol 600 IU/day plus ascorbic acid 500 mg/day (n 25) or placebo (n 28), and treated for 24 months. The study was an extension of a previous study aimed at evaluating the effect of 12-month lifestyle intervention and antioxidant therapy on serum levels of aminotransferases. The primary end point of the present study was change in liver histology on repeated biopsy at 24 months. Secondary end points were changes in body weight, liver enzymes, and insulin sensitivity indices on 2-hour oral glucose tolerance test. The amount of weight lost at 24 months was similar in the placebo and antioxidant groups ( 4.75 [range, ] versus 5.5 [range, ] kg, respectively, P 0.9). A significant improvement occurred in the grade of steatosis, lobular inflammation, and hepatocyte ballooning, and in the NAFLD activity score in both groups. Levels of aminotransferases, triglycerides, cholesterol, fasting glucose, and insulin, and insulin sensitivity indices improved significantly as well. The improvement in all these parameters was not significantly different between the two groups. Conclusion: Lifestyle intervention with diet and increased physical activity induces weight loss and is associated with a significant improvement in liver histology and laboratory abnormalities in pediatric NAFLD. Alpha-tocopherol plus ascorbic acid does not seem to increase the efficacy of lifestyle intervention alone. (HEPATOLOGY 2008; 48: ) Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; FPG, fasting plasma glucose; HOMA, homeostatic model assessment; IGT, impaired glucose tolerance; ISI, insulin sensitivity index; NAFLD, nonalcoholic fatty liver disease; OGTT, oral glucose tolerance test. From the 1 Liver Unit, 2 Pathology Department, 3 Epidemiology Unit, and 4 Molecular Medicine Bambino Gesù Children s Hospital and Research Institute, Rome, Italy; and 5 Department of Medicine, Division of Gastroenterology & Hepatology, Mayo Clinic, Rochester, MN. Received October 2, 2007; accepted March 11, Address reprint requests to: Valerio Nobili, M.D., Liver Unit, Research Institute, Bambino Gesù Children s Hospital, S. Onofrio 4 Square, Rome, Italy. nobili66@yahoo.it; fax: Copyright 2008 by the American Association for the Study of Liver Diseases. Published online in Wiley InterScience ( DOI /hep Potential conflict of interest: Nothing to report. Nonalcoholic fatty liver disease (NAFLD) is rapidly becoming one of the most important chronic liver diseases in children and adolescents. NAFLD affects 2.6% to 9.8% of children and adolescents, 1-3 but this figure increases up to 77% among obese individuals. 4 The long-term prognosis of pediatric NAFLD remains incompletely defined, but some series have reported well-documented cases of cirrhotic stage disease. 3,5 Rapid progression to cirrhosis can occur in some children with NAFLD, 6 and some series have reported cases of children with NAFLD who developed cirrhosis in young adulthood. 7,8 Thus, identifying and treating NAFLD in children and adolescents is expected to prevent the development of advanced liver disease later in life. Treatment trials in pediatric NAFLD are scarce, and nothing has proved effective. Weight loss improved or normalized liver enzymes and the ultrasonographic features in small case series and uncontrolled trials of children with NAFLD. 2,9-11 Alpha-tocopherol and metformin improved liver enzymes in small, open-label pilot studies of few months duration However, those stud- 119

2 120 NOBILI ET AL. HEPATOLOGY, July 2008 ies lacked a control group and the effect on liver histology was not evaluated, precluding meaningful conclusions. Recently, we reported that in pediatric NAFLD, 1 year of lifestyle intervention with diet and increased physical activity induced weight loss, and improvement in liver enzymes and markers of insulin resistance. 15 In that study, the addition of alpha tocopherol and ascorbic acid was not associated with an additional benefit as compared to placebo. The rationale for adding antioxidant therapy with alpha tocopherol plus ascorbic acid to lifestyle intervention was based on the knowledge that injury from oxidative stress contributes to the progression from simple steatosis to steatohepatitis, fibrosis, and cirrhosis. 16 Oxidative stress enhances severity of insulin resistance. 16,17 In in vitro studies, 18,19 alpha tocopherol inhibited proinflammatory cytokine production and attenuated the release of profibrogenic agents and liver collagen. Alpha tocopherol supplementation restores significantly the production of glutathione, which plays a key role decreasing oxidative stress and inhibiting hepatic fibrogenesis and collagen deposition 20,21 Obese subjects may be more susceptible to oxidative injury due to lower levels of circulating alpha tocopherol. 22 A diet rich in fat and low in vitamins, particularly alpha tocopherol and ascorbic acid, can contribute to reduce levels of circulating and stored vitamin E. Further, ascorbic acid enhances regeneration of oxidized vitamin E. 23 No therapeutic trial in pediatric NAFLD including liver histology as the primary end point has been reported to date. Further, given the rather poor correlation of changes in liver enzymes with changes in liver histology, 24 it remains uncertain whether the improvement in liver enzymes occurring with treatment reflects improvement of liver histological features. Because we observed a significant biochemical improvement and a high compliance to lifestyle intervention and pharmacological treatment in our prior 12-month study, 15 we decided to extend this trial for an additional year to deal with these issues and to determine whether the addition of alpha tocopherol and ascorbic acid increases the beneficial effects of lifestyle intervention in children with NAFLD using liver histology as the primary efficacy end point. Materials and Methods Patient Selection and Evaluation The study lasted from January 2003 to October In our prior double-blind, placebo-controlled trial, children or adolescents with NAFLD were enrolled. They were randomized to treatment with alpha tocopherol 600 IU/day plus ascorbic acid 500 mg/day (n 45) or an identical placebo (n 45) given orally. All patients were included in a lifestyle intervention program consisting of a diet tailored on the individual requirements and increased physical exercise. Treatment was given for 12 months. Details of the results of this study have been reported. 15 Briefly, lifestyle intervention led to significant weight loss, improvement in liver enzymes, and insulin resistance. The improvement in all these parameters was not significantly different between the placebo and alpha tocopherol/ascorbic acid groups. At the end of the 12- month treatment period, the parents or legal guardians were informed about the results of the study as well as the treatment the patients had been assigned to. They were asked to continue the assigned treatment in an open-label fashion for an additional 12 months and undergo liver biopsy at the end of the 24 months. Written informed consent was obtained from the parents or legal guardians. The study was approved by the Ethics Committee at the Bambino Gesù Hospital, Rome, Italy. Inclusion criteria were persistently elevated serum aminotransferase levels, diffusely echogenic liver on imaging studies suggestive of fatty liver, and biopsy consistent with the diagnosis of NAFLD. Exclusion criteria were hepatic virus infections (HCV RNA polymerase chain reactionnegative; hepatitis A, B, C, D, E, and G; cytomegalovirus; and Epstein-Barr virus), alcohol consumption, history of parenteral nutrition, and use of drugs known to induce steatosis (for example, valproate, amiodarone, or prednisone) or to affect body weight and carbohydrate metabolism. Autoimmune liver disease, metabolic liver disease, Wilson s disease, and -1-antitrypsin-associated liver disease were ruled out. The body mass index (BMI) and BMI z-score were calculated. 25 Obesity was defined as BMI 95th percentile for age and gender. 26 Patients underwent a medical evaluation every 3 months with the laboratory tests repeated during the 24-month study period. Evaluation of Glucose Metabolism and Insulin Sensitivity A 2-hour oral glucose tolerance test (OGTT) was performed at baseline and at 24 months with the standard 1.75 g of glucose per kg, or maximum of 75 g. According to the American Diabetes Association, fasting plasma glucose (FPG) levels up to 99 mg/dl are considered normal; impaired fasting glucose was defined by an FPG of mg/dl; impaired glucose tolerance (IGT) was defined by a 2-hour plasma glucose of mg/dl; diabetes mellitus was defined by a FPG 126 mg/dl, or a 2-hour plasma glucose 200 mg/dl. 27 Plasma glucose was measured in triplicate by the glucose oxidase technique on a Beckman glucose analyzer (Beckman, Fullerton, CA); and plasma insulin measured by a specific radioimmunoassay (MYRIA Technogenetics, Milan, Italy).

3 HEPATOLOGY, Vol. 48, No. 1, 2008 NOBILI ET AL. 121 The degree of insulin resistance (IR) and sensitivity was determined, respectively, by the homeostatic model assessment (HOMA-IR) using the formula: IR (insulin*glucose)/22.5; 28 and by the insulin sensitivity index (ISI) derived from OGTT using the formula: ISI (10,000/square root of [fasting glucose fasting insulin] [mean glucose mean insulin during OGTT]). 29 Liver Biopsy Liver biopsy was performed at baseline and repeated at 24 months of treatment. Biopsies were routinely processed and analyzed as described. 15,30,31 Liver biopsies were reviewed and scored by a single pathologist (R.D.) who was unaware of the assigned treatment, patients clinical and laboratory data, and liver biopsy sequence. The main histological features of NAFLD, including steatosis, inflammation, hepatocyte ballooning, and fibrosis, were scored using the scoring system for NAFLD recently proposed by Kleiner et al. 32 The features of steatosis (0-3), lobular inflammation (0-3), and hepatocyte ballooning (0-2) were then combined in a score that goes from 0 to 8, called the NAFLD activity score. 32 The stage of fibrosis was scored in a 5-point (stage 0 to 4) scale. 32 To determine the intraobserver agreement, the pathologist scored the liver biopsies blindly twice and the weighted kappa coefficients were calculated. The pathologist had a long time experience in the field of liver pathology and she had good-to-excellent intraobserver agreement between readings as demonstrated by a weighted kappa coefficient of 0.89 for steatosis, 0.72 for lobular inflammation, 0.75 for portal inflammation, 0.82 for ballooning, and 0.87 for fibrosis. Therapeutic Trial Lifestyle Intervention. Patients and responsible guardians underwent monthly 1-hour nutritional counseling by the same experienced dietician during the 24- month treatment period. At every visit, participants were encouraged to continue adherence to treatment. Diet was hypocaloric (25-30 calories/kg/day) in overweight and obese children; and isocaloric (40-45 calories/kg/day) in children with normal BMI. The number of calories needed was determined taking into account anthropometrics, and physical and daily activities. Diet composition consisted of carbohydrate (50%-60%); fat (23%- 30%); and protein (15%-20%); fatty acid composition was two-thirds unsaturated, and one-third saturated; the 6/ 3 ratio was 4:1 as recommended by the Italian Recommended Dietary Allowances. Diet was tailored on individual preferences and requirement of calories. The goal of weight management was to induce a negative calorie balance in overweight and obese patients and to allow a neutral calorie balance in normal weight subjects. The above diet regimen was prescribed with a recommendation to engage in a moderate daily exercise program (45 minutes/day aerobic physical exercise). At each visit, subjects or their responsible guardians were asked to fill out a 3-day dietary and physical activity recall to evaluate adherence to lifestyle recommendations. A multidisciplinary team including dieticians, hepatologists, endocrinologists, psychologists, and cardiologists evaluated and closely followed-up patients participating in this study as done routinely in overweight and obese children and adolescents treated in our institution. To increase compliance with the recommended treatment and weight maintenance during follow-up, the weight loss program focused on long-term dietary modification (low-fat meals, decrease of nutrient dense foods and consequent increase of fruits and vegetables, ingestion of small-size to moderate-size portions of meals throughout the day), increment in daily physical activity and reduction of sedentary activities, and behavior change skills (self-monitoring, familybased reinforcement systems, identification of high-risk situations, self-awareness, stimulus controls, and cognitive behavior strategies). Participants and their family were instructed on how to exercise and maintain adherence to the exercise program by a skilled exercise physiologist as part of this multidisciplinary program. Pharmacological Treatment. A computer-generated randomization sequence assigned participants in a 1:1 ratio to treatment with alpha tocopherol 600 IU/day plus ascorbic acid 500 mg/day (vitamin group) or placebo (placebo group). A statistician (V.D.C.), who was blinded to participants clinical data and did not participate in patients clinical care, generated the allocation sequence and assigned participants to their group. Only the statistician had access to the treatment codes. Vitamins and placebo pills were of identical appearance and taste, and prepared by the same pharmacist at the Bambino Gesù Children s Hospital. Pills were stored at the hospital pharmacy and dispensed at the baseline visit (randomization) and bimonthly thereafter. Participants and investigators were blinded to drug treatment assignments for the first 12 months, and then the study continued in an open-label fashion for an additional 12 months as described above. Participants were instructed to record daily the intake of pills. Compliance to the treatment was evaluated by pill count at every visit, review of the study medication intake records, and directly interviewing the participants. Power and Sample Size Calculation The primary end point of this study was change in liver histological features. We used the change in the NAFLD activity score at 24 months of treatment to determine

4 122 NOBILI ET AL. HEPATOLOGY, July 2008 Table 1. Clinical, Laboratory, and Histological Characteristics of Patients Completing 24 Months of Treatment and Those Who Did Not at Baseline, and 12 and 24 Months Baseline 12 Months 24 Months Completing (n 53) Not Completing (n 35) P* Completing (n 53) Not Completing (n 33) P* Completing (n 53) Not Completing (n 33) P* Sex (M/F) 37/16 15/ /16 15/20 37/16 15/20 Age (year) 11.9 (3.6 18) 12.4 ( ) 0.5 Weight (kg) 62.5 (26 101) 58.8 (30 101) (26 94) 55 ( ) (25 95) 54.7 ( ) 0.7 BMI (kg/m 2 ) 25.2 ( ) 25.9 ( ) ( ) 24.8 ( ) ( ) 23.3 ( ) 0.7 BMI Z-score 1.7 ( ) 1.8 ( ) ( ) 1.6 ( ) ( ) 1.5 ( ) 0.9 ALT (IU/L) 63 (18 89) 68 (14 192) (15 52) 34 (18 70) (14 46) 37 (20 67) 0.6 AST (IU/L) 40 (19 70) 45 (12 41) (22 90) 36 (21 87) (20 127) 33 (21 56) 0.7 Cholesterol (mg/dl) 150 (90 222) 156 (75 222) (87 152) 135 (90 164) (90 156) 130 (75 170) 0.8 Triglycerides (mg/dl) 78 (33 348) 56 (30 209) (37 188) 72 (30 170) (28 177) 67 (31 152) 0.6 Fasting glucose (mg/dl) 78 (60 106) 85 (70 138) (62 85) 73 (64 92) (60 82) 74 (65 112) 0.7 Fasting insulin ( IU/mL) 9.7 (4.5 22) 10.5 ( ) (4 20) 8.3 (3.5 21) ( ) 7 ( ) 0.4 HOMA-IR 1.93 ( ) 2.43 ( ) ( ) 1.42 ( ) ( ) 1.88 ( ) 0.4 ISI 3.89 ( ) 3.8 ( ) ( ) 5.55 ( ) ( ) 5.74 ( ) 0.6 Steatosis 1(1 3) 2 (1 3) 0.7 NA NA 1 (0 1) NA Lobular inflammation 1 (0 2) 0 (0 1) 0.7 NA NA 1 (0 2) NA Portal inflammation 0 (0 2) 0.5 (0 2) 0.1 NA NA 0 (0 1) NA Ballooning 1 (0 2) 1 (0 2) 0.7 NA NA 1 (0 2) NA NAFLD activity score 4 (2 7) 3 (1 7) 0.07 NA NA 2 (0 5) NA Fibrosis stage 1 (0 1) 1 (0 3) 0.2 NA NA 0 (0 2) NA Data are expressed as median (minimum-maximum). *P value at the Mann-Whitney U test (intergroup comparison) at baseline, and 12 and 24 months. NA, not available. treatment efficacy. The null hypothesis was that there was no relationship between treatment and change in the NAFLD activity score. However, as no data on the potential effect of treatment on liver histological features in pediatric NAFLD exist in the literature, we arbitrarily chose an improvement in 2 points of the NAFLD activity score (at least one of these 2 points had to be on inflammation or ballooning) in 80% of patients in the antioxidant group and in 40% of patients in the placebo group. To detect a difference of 40% between groups at an alpha (two-tailed) value of 0.05 and power of 80%, 23 patients per group would be needed. Secondary end points included changes in body weight, liver enzymes, markers of insulin resistance, and changes in individual histological features. Statistical Analysis Data are summarized as frequencies or percentages for categorical variables, and as mean standard error of the mean or median and minimum-maximum ranges for continuous variables. Continuous variables were compared using standard nonparametric statistics for paired and unpaired data. Frequency data were compared using the chi-square test or Fisher s exact test where appropriate or the McNemar s test for paired frequency data. Relationships between variables were sought by linear correlation analysis (Spearman s r 0 ) and regression analysis performed using standard techniques. A two-tailed P value 0.05 was considered statistically significant. The data were analyzed using the intention-to-treat principle, that is, values recorded at baseline were compared to values recorded at 24 months in all patients regardless of treatment duration. Results By 12 months of therapy, two of the 90 randomized patients (both in the vitamin group) had withdrawn from the study. 15 Of the 88 patients completing the first 12 months of treatment, 35 did not agree to take part of the additional 12-month study extension (33 due to reluctance to undergo liver biopsy at 24 months, and two patients who were lost to follow-up). As shown in Table 1, the baseline characteristics of these 35 patients were similar to those of the 53 patients who agreed to have a liver biopsy at 24 months. The 33 patients who did not agree to have a 24-month liver biopsy continued the same lifestyle intervention program and were followed-up for the additional 12 months. The 53 study patients (mean age years; range, ; 37 boys) included 28 randomized to placebo and 25 to antioxidants. Table 2 reports their anthropometric, laboratory, and histological data at baseline and at 24 months. Characteristics of the Groups at Baseline. At baseline, patients in the vitamin and placebo groups were well

5 HEPATOLOGY, Vol. 48, No. 1, 2008 NOBILI ET AL. 123 Table 2. Clinical, Laboratory, and Histological Characteristics of Patients Completing the Study at Baseline and 24 Months Placebo (n 28) Alpha-tocopherol Ascorbic acid (n 25) Baseline 24 months P* Baseline 24 months P* P** Sex (M/F) 23/5 14/11 Age (year) 11.7 ( ) 12.8 ( ) Weight (kg) 61 (40 101) 57 (37 89) (30 84) 50 (29 76) Height (m) 1.56 ( ) 1.59 ( ) ( ) 1.54 ( ) BMI (kg/m 2 ) 26.8 ( ) 23.7 ( ) ( ) 21.2 ( ) ALT (IU/L) 63 (14 126) 34 (21 45) (14 192) 40 (20 67) AST (IU/L) 49 (21 87) 34 (21 60) (20 127) 32 (21 56) GT (IU/L) 21 (11 130) 23 (11 88) (11 67) 22 (12 61) Cholesterol (mg/dl) 156 ( ) 132 ( ) (75 222) 127 (75 151) Triglycerides (mg/dl) 89 (28 145) 62 (37 152) (33 177) 67 (31 123) Fasting glucose (mg/dl) 87 (70 130) 75 (65 112) (74 99) 72 (69 89) Fasting insulin ( IU/mL) 12.8 ( ) 8.2 ( ) (3.5 18) 6.8 (2.0 15) HOMA-IR 2.68 ( ) 1.90 ( ) ( ) 1.40 ( ) ISI 3.43 ( ) 5.52 ( ) ( ) 6.25 ( ) Steatosis 1 (1 3) 1 (0 1) (1 3) 1 (0 3) Portal inflammation 1 (0 2) 0 (0 2) (0 2) 1 (0 1) Lobular inflammation 1 (0 2) 0 (0 1) (0 2) 1 (0 2) Ballooning 1 (0 2) 0 (0 2) (0 1) 2 (0 2) Fibrosis 1 (0 1) 2 (0 1) (0 2) 1 (0 2) NAFLD activity score 4 (2 7) 2 (0 5) (2 7) 2 (1 5) Data are expressed as median (minimum-maximum). *P value for the intragroup comparison (baseline versus 24-month); **P value for the intergroup comparison (changes from baseline between group). At baseline, patients in the vitamin and placebo groups were different for values of fasting insulin, HOMA-IR, and ISI (P 0.05 for all the three parameters). However values of these variables at month 24 were independent of values at baseline in a regression model analysis. balanced except for values of fasting insulin (12.8 [range, ] versus 9.2 [range, ] IU/L; P 0.05), and consequently HOMA-IR (2.68 [range, ] versus 2.10 [range, ], P 0.05) and ISI (3.43 [range, ] versus 4.9 [range, ], P 0.05). Male/female distribution was 23/5 in the placebo and 14/11 in the vitamin group (P 0.07). All patients, but two (both in the antioxidant group) were insulin resistant with both HOMA 1 28 and ISI The proportion of children with severe insulin resistance (for example, HOMA 3) was not different between the two groups (13/28 versus 9/25, P 0.6). The number of obese children was 19 (68%) in the placebo and 13 (52%) in the antioxidant group (P 0.3). A total of six patients had abnormal OGTT at baseline, including four in the placebo group (one with impaired fasting glucose, two with IGT, and one with diabetes) and two in the antioxidant group both with IGT (P 0.8). The liver histological features are summarized in Table 2. At entry, neither the severity of the individual histological features nor the proportion of patients distributed in the three NAFLD activity score categories was significantly different between the two groups. Also, the NAFLD activity score was not significantly different between the placebo and vitamin groups ( versus , respectively, P 0.2). Outcomes at 24 Months. Compliance with the recommended diet and exercise program was excellent in both groups during the 24-month study period. However, two patients in the placebo group and four patients in the antioxidant group took the study medication for 21 months or less. The 53 participants, however, agreed to complete the follow-up and undergo a liver biopsy at 24 months, allowing the intention-to-treat analysis of the data. No adverse events were recorded in any case. Changes in Body Weight and Laboratory Parameters. Figure 1 shows the changes in BMI-z score over the 24-month study period. The median weight loss was similar in the placebo and antioxidant groups ( 4.75 [range, 16 to 4.0] versus 5.5 [range, 12.2 to 0.4] kg, respectively, P 0.9) or BMI ( 2.88 [range, 7.6 to 0.69] versus 3.5 [range, 5.32 to 0.52] kg/m 2, respectively, Fig. 1. Changes in BMI z-score during the 24-month study period in children treated with placebo (clear circles) and antioxidants (black circles). The figure shows mean values and 95% confidence intervals for the error bars. The change in BMI-z score was not significantly different between the groups at any time interval.

6 124 NOBILI ET AL. HEPATOLOGY, July 2008 Table 3. Spearman Correlation Analysis Between Percent Changes in BMI and Metabolic Parameters Percent changes Percent Change in BMI (kg/m 2 ) ALT (IU/L) AST (IU/L) GT (IU/L) Cholesterol (mg/dl) Triglycerides (mg/dl) Fasting insulin ( IU/mL) HOMA-IR ISI P 0.3). Paralleling the decrease in body weight, laboratory variables: namely, alanine aminotransferase (ALT), aspartate aminotransferase (AST), fasting glucose, insulin, cholesterol, triglycerides, HOMA-IR, and ISI improved significantly in both groups (Table 2). Body weight declined logarithmically with small oscillation (within 1.0 kg) in the last 6 months of follow-up. Correlation coefficients between changes in BMI and those of some metabolic variables are reported in Table 3. When changes between groups were compared, the vitamin group had a significantly higher decrease in levels of cholesterol ( 35 6 versus 21 4 mg/dl, P 0.02) whereas the placebo group showed a significantly higher increase in ISI ( versus , P 0.02). Otherwise, the changes in body weight and in all other laboratory parameters were similar between the two groups (Table 2). ALT levels reached the normal range in a higher proportion of patients in the placebo group than in the antioxidant group (22/28 [79%] versus 13/25 [52%], respectively, P 0.05). AST levels reached the normal range in a similar proportion of patients in the placebo and antioxidant groups (22/28 [79%] versus 19/23 [76%], respectively, P 1.0); AST levels improved in 3/28 and 4/25 patients, and remained the same in 3/28 and 2/25 patients. ALT or AST levels did not increase in any case. The number of patients with severe insulin resistance (HOMA-IR 3) decreased from 13 to three in the placebo group and from nine to zero in the antioxidant group. At 24 months, only three patients remained with abnormal OGTT, including the one with diabetes and two with IGT (one in each group). Changes in Liver Histology. As summarized in Table 2 and illustrated in Fig. 2, a significant improvement was noted in both groups in grade of steatosis, lobular inflammation, and hepatocyte ballooning, and in the mean NAFLD activity score. The degree of improvement in all these histological features, however, was not significantly r o P different between the two groups. There was some improvement in the grade of portal inflammation in both groups, but it did not reach statistical significance. The mean stage of fibrosis remained essentially unchanged. The NAFLD activity score improved in 25 patients in the placebo group, did not change in two, and worsened in one patient from three to five points (Table 4). In the antioxidant group, the NAFLD activity score improved in 24 patients and did not change in the other one. However, it did not worsen in any patient from this group. The proportion of patients with improvement, no change, or worsening of the NAFLD activity score was not significantly different between the two groups (P 0.9). The number of patients that reached the primary end point, that is, improvement in the NAFLD activity score by 2 points was identical in the placebo and antioxidant groups (19/28 [68%] versus 17/25 [68%], respectively, P 1.0). At 24 months of treatment, criteria for nonalcoholic steatohepatitis diagnosis were present in only two patients in the placebo group, and in one patient in the antioxidant group. The only patient with stage 2 fibrosis at entry remained on stage 2 despite improvement of 2 points (one for inflammation, one for ballooning), that is, from 6 to 4 in the NAFLD activity score. One patient with fibrosis stage 1c on baseline liver biopsy had stage 2 fibrosis on the 24- month liver biopsy; her NAFLD activity score improved only 1 point (on ballooning) from 6 to 5. The 13 patients without fibrosis at entry remained without fibrosis at 24 months whereas three patients with stage 1 at entry (one stage 1b, two stage 1c) did not have fibrosis on 24-month liver biopsy. Discussion In this study, 2 years of lifestyle intervention with a diet tailored on the individual calorie requirement and increased physical activity was associated with a significant improvement in the severity of steatosis, inflammation, and hepatocyte ballooning, and consequently in the NAFLD activity score. Paralleling the decrease in body weight, secondary end points such as insulin resistance and levels of lipids and liver enzymes improved significantly. The study represents the first clinical trial ever reported using liver histology as the primary end point in children and adolescents with NAFLD. The study suggests that the addition of alpha tocopherol and ascorbic acid for 24 months may not be associated with a greater beneficial effect than that achieved by lifestyle intervention alone. We chose changes in the NAFLD activity score as the primary end point as liver histology represents the most appropriate end point to determine efficacy in clinical

7 HEPATOLOGY, Vol. 48, No. 1, 2008 NOBILI ET AL. 125 Fig. 2. Changes in liver histological features from baseline (clear bars) to 24 months (black bars) of treatment in the two groups. Bars represent mean and error bars represent standard deviation. The scores were calculated according to the criteria of Kleiner et al. 32 trials in NAFLD. A total of 68% of our patients reached the primary end point of the trial, whereas the NAFLD activity score improved or remained stable in all others, but one patient. Although it remains uncertain what prognostic implications the features composing the NAFLD activity score have in the long-term, features of the NAFLD activity score may be more suitable to change in a clinical trial of 24-months duration, as compared to fibrosis stage. In our study, portal inflammation improved in both groups, but this improvement did not reach statistical significance. This suggests that in pediatric NAFLD the improvement in portal injury may follow a slower course as compared to features composing the NAFLD activity score. We decided to use the scoring

8 126 NOBILI ET AL. HEPATOLOGY, July 2008 Table 4. Hepatic Histological Findings Placebo (n 28) Alpha Tocopherol Ascorbic Acid (n 25) Grade Baseline At 24 months Baseline At 24 months Macrovesicular steatosis Microvesicular steatosis Lobular inflammation Portal inflammation Ballooning Fibrosis stage a b c NAFLD activity score 2 (not NASH) (borderline) (NASH) The scores were calculated according to the criteria of Kleiner et al. 32 None of the baseline histological features were significantly different between the two groups (P 0.05). The proportion of patients in the three NAFLD activity score categories at baseline was not significantly different between the two groups (P 0.15). system proposed by Kleiner at al. 32 because it is the only scoring that has been specifically developed for use in clinical trials in pediatric NAFLD. Interestingly, no significant changes in stage of fibrosis occurred in our study despite a clear improvement in the NAFLD activity score. This can be explained by the fact that all but one patient had either no fibrosis at all (stage 0) or very mild (stage 1) fibrosis at entry. It remains uncertain whether a longer treatment period beyond 24 months is necessary to determine the effect of treatment on fibrosis stage, and what effect lifestyle intervention has on more advanced stages of fibrosis such as septal fibrosis (stage 3) or cirrhosis (stage 4) in pediatric NAFLD. The addition of alpha-tocopherol and ascorbic acid was not associated with a greater histological or biochemical improvement as compared to placebo. This is consistent with the negative results of the only two trials reported that have included a comparative control group The dosage used in our study was as twice as the average amount of antioxidants commonly prescribed in different diseases (lipid malabsorption, kidney diseases, age-related maculopathy, and cognitive decline) and in prevention of cardiovascular disease and cancer, but lower than the dosages prescribed in previous clinical trials for the treatment of adult NAFLD that revealed negative results. 33,35 Notwithstanding the recent evidence that the use of several antioxidants is not innocuous, but associated with an increased risk of death and heart failure. 36,37 Conversely, the lifestyle intervention approach used in this study, which is better structured and more intense than a conventional approach, can be recommended in the treatment of pediatric NAFLD. In fact, this approach has shown a good retention of the weight lost up to the second year of follow-up in agreement with the results obtained in a recent study in obese children. 38 Our study has some limitations. First, our trial was double-blind for only the first 12 months and open-label for an additional 12 months. The rational for disclosing the treatment codes at the end of the first year was to determine whether a significant biochemical improvement occurred with treatment; if such improvement had not been seen then we would have stopped the trial. How-

9 HEPATOLOGY, Vol. 48, No. 1, 2008 NOBILI ET AL. 127 ever, as a significant biochemical improvement was seen with treatment at 12 months, 15 but improvement in liver enzymes do not accurately correlate with changes in liver histology, patients were offered to continue the assigned treatment and undergo liver biopsy at 24 months. Second, nutritional counsel favored the consumption of more fruits and vegetables, which may have augmented the intake of natural antioxidants in both groups and minimized the effects of the vitamin supplement in the vitamin group. Third, all patients in our trial were intervened with a diet and increased physical activity and thus it remains unclear whether alpha tocopherol/ascorbic acid may be better than no intervention at all in pediatric NAFLD, or in those patients who do not adhere to lifestyle recommendations. In summary, our study demonstrates that lifestyle intervention consisting of a diet tailored on the individual requirement and increased physical activity improves significantly the liver histological injury of children and adolescents with NAFLD. Lifestyle intervention also significantly improved the underlying insulin resistance, serum levels of aminotransferases, and lipid levels. The addition of alpha-tocopherol and ascorbic acid did not seem to add an extra benefit to that achieved by lifestyle changes alone. However, future studies are required to vary aspects of diet and physical activity and to assess their joint or separate effects. Acknowledgment: We thank Dr. Eugenio Ciacco and Dr. Rodolfo Fruwhirt for their expert technical assistance. We also thank all the nursing staff for their excellent work and dedication. We are also indebted to all the children and their legal guardians who participated in this study. References 1. Tominaga K, Kurata JH, Chen YK, Fujimoto E, Miyagawa S, Abe I, et al. prevalence of fatty liver in Japanese children and relationship to obesity. An epidemiological ultrasonographic survey. Dig Dis Sci 1995;40: Franzese A, Vajro P, Argenziano A, Puzziello A, Iannucci MP, Saviano MC, et al. Liver involvement in obese children. Ultrasonography and liver enzymes levels at diagnosis and during follow-up in an Italian population. Dig Dis Sci 1997;42: Schwimmer JB, Deutsch R, Kahen T, Lavine JE, Stanley C, Behling C. Prevalence of fatty liver in children and adolescents. Pediatrics 2006;118: Chan DF, Li AM, Chu WC, Lavine JE, Stanley C, Behling C. Hepatic steatosis in obese Chinese children. Int J Obes Relat Metab Disord 2004; 28: Kinugasa A, Tsunamoto K, Furukawa N, Sawada T, Kusunoki T, Shimada N. Fatty liver and its fibrous changes found in simple obesity of children. J Pediatr Gatroenterol Nutr 1984;3: Molleston JP, White F, Teckman J, Fitzgerald JF. Obese children with steatohepatitis can develop cirrhosis in childhood. Am J Gastroenterol 2002;97: Adams LA, Feldstein A, Lindor KD, Angulo P. Nonalcoholic fatty liver disease among patients with hypothalamic and pituitary dysfunction. HEPATOLOGY 2004;39: Suzuki D, Hashimoto E, Kaneda K, Tokushige K, Shiratori K. Liver failure caused by non-alcoholic steatohepatitis in an obese young male. J Gastroenterol Hepatol 2005;20: Manton ND, Lipsett J, Moore DJ, Davidson GP, Bourne AJ, Couper RT. Non-alcoholic steatohepatitis in children and adolescents. Med J Aust 2000;173: Vajro P, Fontanella A, Perna C, et al. Persistent hyperaminotransferasemia resolving after weight reduction in obese children. J Pediatr 1994;125: Koçak N, Yüce A, Gürakan F, Ozen H. Obesity: a cause of steatohepatitis in children. Am J Gastroenterol 2000;95: Lavine JE. Vitamin E treatment of nonalcoholic steatohepatitis in children: a pilot study. J Pediatr 2000;136: Vajro P, Mandato C, Franzese A, Ciccimarra E, Lucariello S, Savoia M, et al. Vitamin E treatment in pediatric obesity-related liver disease: a randomized study. J Pediatr Gastroenterol Nutr 2004;38: Schwimmer JB, Middleton MS, Deutsch R, Lavine JE. A phase 2 clinical trial of metformin as a treatment for non-diabetic paediatric non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2005;21: Nobili V, Manco M, Devito R, Ciampalini P, Piemonte F, Marcellini M. Effect of vitamin E on aminotransferase levels and insulin resistance in children with non-alcoholic fatty liver disease. Aliment Pharmacol Ther 2006;24: Day CP. From fat to inflammation. Gastroenterology 2006;130: Paolisso G, Giugliano D. Oxidative stress and insulin action: is there a relationship? Diabetologia 1996;39: Hill DB, Devalarajan R, Joshi-Barve S, Barve S, McClain CJ. Antioxidants attenuate nuclear factor-kappa B activation and tumor necrosis factoralpha production in alcohol hepatitis patient monocytes and rat kupffer cells, in vitro. Clin Biochem 1999;32: Lee KS, Buck M, Houglum K, Chojikier M. Activation of hepatic stellate cells by TGF and collagen type I is mediated by oxidative stress through c-myb expression, J Clin Invest 1995;96: Phung N, Farrel G, Robertson G. Vitamin E but not glutathione precursors inhibits hepatic fibrosis in experimental NASH exhibiting oxidative stress and mitochondrial abnormalities. HEPATOLOGY 2001;34:361A [Abstract]. 21. Parola M, Muraca R, Dianzani I, Barrera G, Leonarduzzi G, Bendinelli P, et al. Vitamin E dietary supplementation inhibits transforming growth factor beta 1 gene expression in the rat liver. FEBS Lett 1992;308: Strauss RS Comparison of serum concentrations of alpha-tocopherol and beta-carotene in a cross-sectional sample of obese and nonobese children (NHANES III). National Health and Nutrition Examination Survey. J Pediatr 1999;134: Chan AC. Partners in defense, vitamin E and vitamin C. Can J Physiol Pharmacol 1993;1: Suzuki A, Lymp J, Sauver JS, Angulo P, Lindor K. Values and limitations of serum aminotransferases in clinical trials of nonalcoholic steatohepatitis. Liver Int 2006;26: Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320: Centers for Disease Control. BMI Body Mass Index. About BMI for Children and Teens. Available at: childrens_bmi/about_childrens_bmi.htm. Accessed March 10, The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care 2003;26: Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28: Matsuda M, De Fronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 1999;22: Nobili V, Manco M, Raponi M, Marcellini M. Case management applied to a multispecialist pediatric outpatients clinic for patients affected by nonalcholic fatty liver disease (NAFLD). J Paediatr Child Health 2007;43:

10 128 NOBILI ET AL. HEPATOLOGY, July Nobili V, Marcellini M, Devito R, Ciampalini P, Piemonte F, Comparcola D, et al. Nonalcoholic fatty liver disease in children: a prospective clinicalpathological study and effect of lifestyle advice. HEPATOLOGY 2006;44: Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. HEPATOLOGY 2005;41: Harrison SA, Torgerson S, Hayashi P, Ward J, Schenker S. Vitamin E and vitamin C treatment improves fibrosis in patients with non-alcoholic steatohepatitis. Am J Gastroenterol 2003;98: Adams LA, Angulo P. Vitamins E and C for the treatment of NASH: duplication of results but lack of demonstration of efficacy. Am J Gastroenterol 2003;98: Bugianesi E, Gentilcore E, Manini R, Natale S, Vanni E, Villanova N, et al. A randomized controlled trial of metformin versus vitamin E or prescriptive diet in nonalcoholic fatty liver disease. Am J Gastroenterol 2005;100: Guallar E, Hanley DF, Miller ER, 3rd. An editorial update: annus horribilis for vitamin E. Ann Intern Med 2005;143: Bjelakovic G, Kikolova D, Gludd LL, Simonetti RG, Gluud C. Mortality in randomized trials of antioxidant supplements for primary and secondary prevention. Systematic review and meta-analysis. JAMA 2007;29: Wilfrey D, Stein R, Saelens BE, Mockus D, Hayden-Wade H, Welch RR, et al. Efficacy of maintenance treatment approaches for childhood overweight. A randomized controlled trial. JAMA 2007;298: