Author
Comparators
N
Population
Findings
ALT
Larrieta-Carrasco et al. [58]
ALT ≤ or >35 U/L
1037
Overweight Mexican children aged 6–12
PNPLA3 GG genotype had 3.7 times the odds for ALT > 35
(95 % CI 2.3–5.9, p = 3.7 × 10−8)
Giudice et al. [57]
ALT ≤ or >40 U/L
1048
Obese Italian children aged 2–16
PNPLA3 GG genotype had 2.97 times the odds for ALT > 40
(95 % CI 1.80–4.18)
Romeo et al. [55]
ALT ≤ or >30 U/L
475
Overweight Italian children mean age 10 years ±3
ALT > 30 U/L in 32 % with GG genotype and 10 % with CC genotype
Lin et al. [56]
Mean ALT
520
Obese Taiwanese children aged 6–18
Mean ALT was 31 U/L for GG genotype and 22 for CC genotype
Imaging via MRI
Goran et al. [25]
MRI-determined signal fat fraction ≤ or >5.5 %
188
Hispanic American children aged 8–18
Mean signal fat fraction was 11 % for GG genotype and was 4.7 % for CC genotype
Santoro et al. [59]
MRI-determined signal fat fraction ≤ or >5.5 %
85
Clinically obese American children aged 8–18
Signal hepatic fat fraction >5.5 % overall; of those, 7 had CC vs. 32 with either CG or GG
Histologic severity
Rotman et al. [26]
Histologic parameters
223
NASH CRN
PNPLA3 genotype was not associated with severity of steatosis, presence of NASH, or severity of fibrosis
Valenti et al. [60]
NASH and fibrosis
149
Italian children mean age 10.2 ± 2.6
PNPLA3 G allele had 1.9 times the odds for fibrosis (95 % CI 1.14–3.45 per number of G alleles)
Diet
Davis et al. [24]
MRI signal fat fraction ≤ or >5.5 %
153
Hispanic American children aged 8–18
With GG genotype, hepatic fat positively correlated with carbohydrate intake (r = 0.38) and total sugar intake (r = 0.33) but not with CG or CC genotypes
Santoro et al. [62]
Signal hepatic fat fraction ≤ or >5.5 %
127
Clinically obese American children mean age 14.7 years ±3.3
With GG genotype, hepatic fat positively correlated with n-6/n-3 PUFA intake (r 2 = 0.45) but not with CG or CC genotypes
ALT and PNPLA3
Several studies have evaluated interactions between ALT and the variant allele in children. In a study of 475 obese or overweight children, 32 % of subjects with homozygous minor alleles for PNPLA3 (GG) had ALT values >30 U/L versus ALT > 30 U/L in only 10 % of subjects who had the homozygous wild-type allele (CC) [55]. In a subsequent study of 520 obese Taiwanese children, higher ALT levels were found in children who were homozygous for the minor allele [56]. In a large Italian study, Giudice et al. evaluated 1048 obese children and found that there was a significant positive interaction between the variant allele and waist circumference with respect to risk for elevated serum ALT. Homozygotes showed a stronger correlation between ALT and waist to height ratio than heterozygotes [57]. In another large study evaluating 1037 Mexican children aged 6–12, the variant allele was found to be significantly associated with ALT levels >35 U/L. Once stratified by weight classification, there was a significant interaction between weight status and risk for ALT > 35 U/L. In normal-weight children with the CC genotype, ALT elevation was extremely rare, but in normal-weight children with the GG genotype, the rate of ALT elevation was similar to obese children with the CC genotype [58].
Imaging, Histology, and PNPLA3
A study by Goran et al. evaluated 188 Hispanic children at the University of Southern California (USC) using MRI to evaluate hepatic signal fat fraction and found signal fat fraction in GG subjects was 1.7 and 2.4 times higher than GC and CC subjects (11.1 ± 0.8 % in GG vs. 6.6 ± 0.7 % in GC and 4.7 ± 0.9 % in CC; p < 0.0001) [25]. Santoro and colleagues reported similar findings in a study of 85 children from the Yale Pediatric Obesity Clinic. They reported a significant effect of the variant G allele such that those with at least one G allele had substantially greater hepatic signal fat fraction [59].
Conflicting evidence exists regarding the association between histologic severity and genotype in children. In 2010, Rotman and colleagues evaluated 223 children from the NASH CRN. In this study, there was no association of PNPLA3 with the histologic severity of NAFLD [26], contrary to what they saw in adults. In Italy, a study of 149 children with biopsy-proven NAFLD reported an association with the variant PNPLA3 allele and histologic severity. Valenti and colleagues reported that the variant G allele was associated with several key features of NAFLD including the severity of steatosis and the presence of steatohepatitis. In addition, the G allele was associated with the presence of pericentral fibrosis but not portal fibrosis [60]. Given the sample sizes of these studies and some conflicting observations, larger and more diverse studies will be needed to better evaluate the role of PNPLA3 on liver histology in children with NAFLD.
Diet and PNPLA3
Several groups have performed subsequent analysis of their data to evaluate the effect of diet on PNPLA3. The USC group evaluated the influence of PNPLA3 variants on hepatic fat modulated by carbohydrate and sugar intake in 153 Hispanic children. There was a mild but significant correlation between carbohydrate and/or sugar intake and hepatic signal fat fraction only in those children who were homozygous for the G allele [24]. However, when histologic variables were assessed in a study of 149 children with biopsy-proven NAFLD, sugar-sweetened beverage consumption was not associated with the histologic features nor the severity of NAFLD [61]. The Yale group evaluated 127 pediatric participants with a median age of 14.7 ± 3.3 whose dietary composition was assessed for essential omega polyunsaturated fatty acid (PUFA) intake. In this study there was a moderate but significant correlation between dietary PUFA intake and hepatic signal fat fraction only in those children who were homozygous for the G allele [62].
Influence of Maternal Factors and Breastfeeding on NAFLD
The possibility that NAFLD may start at birth has been explored by looking at infants of obese and diabetic mothers. In a study of 25 neonates born to normal-weight mothers (n = 13) and obese mothers with gestational diabetes (n = 12), Brumbaugh and colleagues evaluated differences in neonatal fat distribution. Neonates underwent MRI measurement of subcutaneous and intra-abdominal fat and magnetic resonance spectroscopy (MRS) for signal hepatic fat fraction at 1–3 weeks of age. Infants born to obese mothers with gestational diabetes had a mean 68 % greater liver hepatic signal fat fraction compared to infants born to normal-weight mothers. In all infants, signal hepatic fat fraction correlated with maternal prepregnancy BMI but not with subcutaneous adiposity [63].
Modi and colleagues studied 105 mother/neonate pairs to determine whether neonatal liver signal fat fraction measured by MRS was influenced by maternal BMI. They found a strong relationship between prepregnancy BMI and infant signal fat fraction even after adjusting for infant sex and postnatal age; women with higher BMI had babies with higher signal fat fraction [64].
Once a child is born, neonatal overfeeding may have a long-term effect on de novo lipogenesis [65]. Breastfeeding may be protective. In an investigation of 191 Caucasian children with biopsy-proven NAFLD, the distribution of steatosis, inflammation, hepatocyte ballooning, and fibrosis were all worse among children who were not breastfed compared to breastfed children [66].
Clinical Features
Obesity
There is a strong association between obesity and NAFLD in the pediatric population. As many as 70–90 % of children with NAFLD are obese [67]. In SCALE, the prevalence of NAFLD was 5 % among normal-weight children, 16 % among overweight children, and 38 % among obese children [1]. The distribution of adiposity is also important. Manco and colleagues reported that in 197 children with biopsy-proven NAFLD, 84 % had a large waist circumference (>90th percentile for age and gender). The only risk factor for liver fibrosis in this study was a large waist, and thus, body fat distribution may also be an important prognostic factor for disease severity [68].
The relationship between NAFLD, NASH, and morbid obesity in adolescents is less clear. In the Teen LABS study of 242 severely obese adolescents (mean BMI 50.5 kg/m2) undergoing weight loss surgery, NAFLD was diagnosed or suspected in 37 % [69]. In a separate smaller study, Holterman and colleagues compared morbidly obese adolescents to morbidly obese adults. Among 24 severely obese adolescents undergoing weight loss surgery compared to 24 adults with similar BMI, severely obese adolescents had a significantly higher prevalence of NASH (62.5 % versus 25 %) and fibrosis (83 % versus 29 %) [69, 70].
Acanthosis Nigricans
Acanthosis nigricans, a marker of hyperinsulinemia, is common in children with NAFLD. In acanthosis nigricans, the skin at the nape of the neck and sometimes in the axillary area appears hyperpigmented and thickened. Acanthosis nigricans has been found in 36–49 % of children with biopsy-proven NAFLD [34, 74].
Other Physical Exam Findings
Many patients with NAFLD have hepatomegaly. Thus, percussing and palpating the liver and estimating its size are important parts of the pediatric abdominal exam. The excess abdominal girth makes this difficult in obese children, but can still be performed effectively. There may be right upper quadrant tenderness as well as stigmata of chronic liver disease, such as scleral icterus and jaundice, gynecomastia, spider angiomata, palmar erythema, and asterixis.
Diagnostic Approaches
Screening
Screening children for NAFLD has been addressed by multiple societies (Table 17.2). In 2007, an expert committee on childhood obesity published guidelines on the assessment of overweight and obese children. They recommended screening for NAFLD in children who are ≥10 years old and obese or overweight with additional risk factors by measuring ALT and AST levels. The recommendations suggest consultation with a pediatric gastroenterologist should these results be ≥2 times the upper limit of normal (ULN) [75].
Table 17.2
Society guidelines regarding screening overweight and obese children for NAFLD
Society | Recommend screening children for NAFLD | |||
|---|---|---|---|---|
Yes | No | Uncertain | Not stated | |
American Academy of Family Physicians | X | |||
American Academy of Pediatrics | X | |||
American Association for the Study of Liver Disease | X | |||
American College of Gastroenterology | X | |||
American Gastroenterological Association | X | |||
Endocrine Society | X | |||
European Society for Pediatric Gastroenterology, Hepatology, and Nutrition | X | |||
National Association of Pediatric Nurse Practitioners | X | |||
North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition | X | |||
Since screening relies in part on lab values, understanding the “normative” range for these values is important. The median value for ALT upper limit of normal in children’s hospitals nationwide is 53 U/L [13]. However, the SAFETY study identified that among healthy-weight children who did not have liver disease, a more biologically based 95th percentile for ALT was 25.8 U/L in boys and 22.1 U/L in girls. Thus, a lab value of 53 U/L that is considered within the normal reference range at many hospitals would be two times the ULN of the biology-based value. Since many labs continue to use inappropriate normal reference ranges, calculating two times the upper limit of “normal” as screening threshold to detect liver disease is problematic. Moreover, the optimal value of ALT that has sufficient accuracy to be used explicitly for NAFLD screening has not been agreed upon.
In a study of 347 overweight and obese children ≥10 years of age referred from primary care to pediatric gastroenterology for suspected NAFLD identified by screening, nearly 85 % of the children with a screening ALT ≥ 80 U/L (a value most often considered to be 2× ULN) had some form of liver disease. Although a majority of all of the children screened had NAFLD (53 %; 193/347), a nontrivial minority (18 %) of those referred had other forms of liver disease. In addition, 11 % of children referred were found to have advanced fibrosis at diagnosis [18]. Thus, implementing the recommended screening strategy identifies many children with previously undetected liver disease. However, it casts a wide net, as elevated ALT is not specific for the diagnosis of NAFLD. Moreover, ALT of ≥80 U/L was found to have a sensitivity of 57 % and a specificity of 71 % for diagnosing NAFLD. In contrast, when two times the biologically based thresholds of ALT were used, ALT ≥ 50 in boys and ≥44 in girls, there was a substantial increase in the sensitivity of diagnosis of NAFLD to 88 %; however, the specificity declined to 26 % [18].
This recent data does support the practice of screening with ALT to identify liver disease, although it illustrates that not every obese child with elevated ALT will have NAFLD. In addition, the optimal ALT cutoff for screening with sufficient accuracy to detect NAFLD in all children, including those with advanced fibrosis, is not clear. Careful history, physical exam, lab evaluation, and histologic evaluation are extremely important for making a correct diagnosis. Consideration of other causes of chronic liver disease in children, such as autoimmune hepatitis, drug toxicity, infectious hepatitis, Wilson disease, alpha-1-antitrypsin deficiency, celiac disease, hemochromatosis, and metabolic disease, must be considered in clinically appropriate situations.
Biomarkers
A noninvasive, sensitive, and specific biomarker for NAFLD would be helpful as many children with NAFLD go undiagnosed in part due to the requirement for a diagnostic liver biopsy. The most helpful biomarker would not only accurately reflect the presence of disease, but also classify severity of the disease along with the stage of fibrosis. Many biomarkers have been studied in children thus far. As shown in Table 17.3, there are several molecules that are associated with various histologic features of NAFLD; however, none of these have a sufficiently strong enough relationship to be considered clinically useful. For example, in the study of cytokeratin-18 (CK-18), information gathered in cross section or longitudinally does not appear to provide additional information different from what can be gained from measuring ALT [76]. Hyaluronic acid may be a better marker than human cartilage glycoprotein-39 (YKL-40) [77] and has a significant relationship to fibrosis [78], but it is unclear whether these findings may be clinically useful. Adipokines such as chemoattractant protein-1 (MCP-1) and plasminogen activator inhibitor-1 (PAI-1) have a significant relationship to fibrosis [42], yet it may not be strong enough for clinical utility. And finally, fibroblast growth factor-21 (FGF-21) had a mild positive correlation with liver fat (r 2 = 0.278) [79]. Although there is a significant association with biomarkers and histologic features in group aggregate, the findings are not sensitive nor specific enough to be used clinically.
Table 17.3
Potential biomarkers in pediatric NAFLD
Author | Marker(s) | Sample size | Population | Comparator | Key findings |
|---|---|---|---|---|---|
Lee et al. [77] | Hyaluronic acid (HA) Human cartilage glycoprotein-39 (YKL-40) | 128 | 128 children and young adults aged 1.4 months to 27.6 years | Histology | HA: median 74.7 ng/mL in F3–F4 vs. 17.7 ng/mL in F0–F2 fibrosis (p < 0.0001) YKL-40: median 31.5 ng/mL in F3–F4 fibrosis vs. 34.2 ng/mL in F0–F2 fibrosis (p = 0.85) |
Giannini et al. [79] | Fibroblast growth factor-21 (FGF-21) | 217 | Lean and obese adolescents | Fast gradient MRI to measure hepatic signal fat fraction (SFF) | FGF-21 levels correlated with HF r 2 = 0.278, p < 0.001 |
Fitzpatrick et al. [42] | Monocyte chemoattractant protein-1 (MCP-1) Plasminogen activator inhibitor-1 (PAI-1) | 40 | Children recruited from a tertiary care pediatric hepatology unit | Histology | Predictors of advanced fibrosis : MCP-1: AUC 0.76 (95 % CI 0.62, 0.91) PAI-1: AUC 0.78 (95 % CI 0.6, 0.91) |
Vuppalanchi et al. [76] | Cytokeratin-18 | 152 | Children with NAFLD in TONIC | Histology | Change in histology: AUC 0.72 (95 % CI 0.63–0.81) |
Lebensztejn et al. [78] | Cytokeritin-18 M30 Hyaluronic acid | 52 | Children | Histology | CK-18 M30: 177.5 U/L without fibrosis, 311 U/L with fibrosis (p = 0.05) HA: 18.5 ng/mL without fibrosis, 20.5 ng/mL (p = 0.04) HA (cutoff 19.1): sensitivity 84 %, specificity 55 %, PPV 52 %, NPV 86 %. AUC with fib = 0.672, AUC without fib = 0.666 HA + CK-18: sensitivity 74 %, specificity 79 %, PPV 56 %, NPV 63 %, AUC 0.73) |
Liver Imaging
There are a limited number of studies in children comparing radiologic techniques to the reference standard of histology in children with NAFLD. Currently, there are no radiologic techniques that have been shown to be diagnostic, though some show promise for screening and monitoring. The radiologic evaluation of hepatic fibrosis has also been attempted in limited studies involving children, but the abnormalities identified are of very advanced disease. Thus, their evaluation does not contribute to early detection, which is an important consideration in pediatrics.
Ultrasound Compared to Liver Histology
Ultrasound technology uses characteristics of high-frequency sound wave propagation in attempts to differentiate fatty tissue from normal hepatic tissue. Fatty tissue scatters the ultrasound beam, causing more echoes to return to the ultrasound transducer and resulting in a brighter more echogenic liver.
Although ultrasound is commonly used in routine practice to determine the presence and degree of fatty liver, a systematic review revealed its limitations in both of these areas [80]. The positive predictive value of liver ultrasound for the detection of fatty liver in children was found to be between 47 and 62 % [81, 82]. Thus, it is not an optimal modality to be used as a diagnostic test. This limitation is due to an inherent property of ultrasound, in that it does not measure fat directly, but relies on a subjective and nonquantitative interpretation of the echogenicity. Therefore, relying on ultrasound as a semiquantitative measure of hepatic steatosis is unreliable. In addition, the common practice of using ultrasound to exclude fatty liver has insufficient evidence. There has been only one study evaluating children who had a negative ultrasound in addition to a liver biopsy for comparison [82]. In that study, most of the negative ultrasounds turned out to be falsely negative. This was an artifact of the study design, as the study only included children with known NAFLD. Thus, data are also lacking regarding the ability of ultrasound to exclude fatty liver in children. The primary role of ultrasound in pediatrics is in the evaluation of structural problems within the liver or gallbladder. Future studies using ultrasound should consider the evaluation of emerging quantitative ultrasound techniques using liver biopsy as the reference standard for diagnosis and grading hepatic steatosis in children.
MRI Compared to Liver Histology
In pediatric clinical research studies, MRI has overtaken ultrasound as the modality of choice for the noninvasive measurement of hepatic steatosis. In some institutions, MRI is now used in standard clinical practice to measure hepatic signal or proton density fat fraction. The increasing use of MRI is due to its growing availability and promise as a quantitative measure of hepatic steatosis. However, the evidence base is extremely limited. One study evaluated 25 obese children in Rome, Italy, with biopsy-proven NAFLD who underwent MRI prior to liver biopsy to explore the accuracy of the MRI-determined hepatic signal fat fraction [83]. The MRI method used in this study was a modification of the 2-point Dixon method [84]. The MRI-determined hepatic signal fat fraction was strongly correlated (r = 0.88) with the histological grade of steatosis (grade 0 <5 %, grade 1 5–33 %, grade 2 34–65 %, grade 3 ≥66 %). The small sample size precluded more specific determinations of accuracy.
An advancement in the field beyond common use of the modified 2-point Dixon method has identified the importance of performing MRI correctly to compensate for confounders that introduce error into both the accuracy and precision of conventional MRI [85–90]. Such methods are referred to here as advanced MRI. A recent study of advanced MRI compared to histology in 174 children demonstrated that an advanced MRI measure of steatosis called proton density fat fraction (PDFF) correlated well with histologically determined steatosis grade in children. MRI-estimated liver PDFF was significantly (p < 0.01) correlated (0.725) with steatosis grade. The correlation was significantly (p < 0.01) stronger in girls (0.86) than in boys (0.70) and was significantly (p < 0.01) weaker in children with stage 2–4 fibrosis (0.61) than children with no fibrosis (0.76) or stage 1 fibrosis (0.78). This study also evaluated published magnetic resonance-derived threshold values intended to discriminate between no steatosis and mild steatosis (1.8 %, 5.5. %, 6.4 %, and 9 %). Sensitivity ranged from 42 to 98 %, while specificity ranged from 54 to 96 % [91]. Achieving a distinct separation between having and not having a fatty liver based upon a single MRI-based cutoff point remains challenging.
Treatment
Lifestyle Interventions
Pediatric NAFLD is often associated with obesity ; thus, dietary and exercise treatments are often recommended. NAFLD intervention trials have focused on weight loss, but data for lifestyle modification specifically in children with NAFLD are limited [7, 92–98]. Moreover, the specific amount of weight loss required that will result in an improvement in NAFLD in children is not clear. Histology is the cornerstone of diagnosis and should be used as an outcome measure to assess the efficacy of interventions, but current outcome measures in lifestyle treatment studies vary. One of the early studies by Nobili and colleagues enrolled 84 obese or overweight children with biopsy-proven NAFLD to evaluate the effect of low-calorie diet and individually tailored moderate exercise. A total of 52 participants completed the trial, which included medical examinations and laboratory assessment at 3-month intervals and an ultrasound at 12 months. Seventeen participants lost >10 % body weight. Only 5 of the 17 children with >10 % weight loss had a normal ultrasound at 1 year [7]. However, histologic change was not assessed. The greatest decrease in ALT was seen in those who lost 5 % or more of their body weight. At this time, weight loss with diet and exercise should continue to be recommended for overweight and obese children. However, weight loss alone may not be sufficient to improve NAFLD in all children, and further research with histology as an outcome measure is certainly needed.
There is interest in the role of fructose in the pathogenesis of NAFLD. Fructose consumption has been suggested to be a risk factor for NAFLD [99, 100]. A pilot study evaluated the efficacy of a low-fructose diet in decreasing ALT in children with NAFLD. Ten children with NAFLD or suspected NAFLD were placed on either a low-fructose diet (n = 6) or low-fat diet (n = 4) for 6 months. There was no significant change in ALT in either group. Histologic changes in NAFLD were not assessed [94]. Subsequently, Vos et al. conducted a 4-week, double-blind, randomized, controlled intervention study of 21 Hispanic children aged 11–18 with BMI ≥ 85 % who regularly consumed sweet beverages. All participants had hepatic fat quantification using MRS signal fat fraction. They were randomized to drink 24 fluid ounces of fructose or glucose beverage per day. After 4 weeks, there was no significant change in hepatic fat in either group [101]. In summary, while fructose can cause NAFLD in animal models, the degree to which fructose is a relevant factor in children with NAFLD remains uncertain.
Pharmacologic Therapy
Antioxidant and Hepato-protective Agents
Mitochondrial dysfunction and damage by reactive oxygen species are implicated in the pathogenesis of NAFLD as described in previous chapters, and thus, antioxidants have been evaluated as a potential therapy. Treatment with vitamin E resulted in decreased ALT in a pilot study of 11 children with increased liver echogenicity determined by ultrasound [102]. Vitamin E as an intervention was again studied in a larger cohort of 88 children with biopsy-proven NAFLD who also participated in monthly sessions with dieticians. In this study, Nobili and colleagues compared children receiving vitamin E versus vitamin C versus placebo for 1 year. Treatment with vitamin E did not significantly improve ALT compared to vitamin C or placebo [103]. In a follow-up study by the same group, about 60 % of these patients underwent liver biopsy. Although there was significant improvement in steatosis, hepatocellular ballooning, and lobular inflammation among study participants, there was no significant difference in histology between vitamin E and placebo groups [104]. Subsequently, the Treatment of NAFLD in Children (TONIC) trial, a large NASH CRN multicenter, randomized, double-blind, placebo-controlled trial, was completed in 2010 [105]. In this trial, 173 children with biopsy-proven NAFLD were randomized to receive metformin, high-dose vitamin E, or placebo for 96 weeks. The primary outcome measure was a decrease in ALT by 50 % compared to baseline or a decrease to less than 40 U/L. In that study, treatment with vitamin E did not result in significant decrease in ALT compared to placebo. In addition, the features of steatosis inflammation, ballooning, and fibrosis were evaluated after 2 years of treatment. There was no significant improvement in steatosis, inflammation, or fibrosis with vitamin E. However, hepatocyte ballooning was shown to improve in 38 % (22/58) of children taking vitamin E and only 17 % (10/58) of children taking placebo.
Ursodeoxycholic acid (UDCA) , a cytoprotective agent, is a secondary bile acid formed by intestinal bacteria. In one study of 31 obese children with elevated ALT and increased echogenicity as determined by liver ultrasound, the effect of UDCA on ALT was determined. Participants were divided into four groups: diet alone (n = 11), UDCA treatment (n = 7), UCDA and diet (n = 7), and untreated controls (n = 6) [106]. UDCA alone was not effective in lowering ALT.
Cysteamine is an aminothiol agent that acts as an antioxidant by scavenging reactive oxygen intermediates as well as increases glutathione, the most abundant intracellular antioxidant agent, and thus is a candidate therapy for NAFLD [107]. In an open-label pilot study of 11 children with biopsy-proven NAFLD and serum ALT ≥ 60 U/L, Dohil et.al evaluated the effect of twice-daily enteric-coated cysteamine for 24 weeks on serum ALT [107]. At the 24-week time point, 64 % of subjects had a decrease in serum ALT by at least 50 % of baseline. This effect was maintained at 48 weeks. This cohort was evaluated to assess the impact of cysteamine treatment on multimerization. After 24 weeks of therapy, there was an increase in total adiponectin (49.3 %, p = 0.05) from baseline [108]. Currently, the NASH CRN is conducting a multicenter, placebo-controlled clinical trial of children aged 8–17 years with biopsy-confirmed moderate to severe NAFLD. The primary objective is to evaluate whether 52 weeks of treatment with cysteamine bitartrate delayed-release capsules will result in an improvement in liver disease severity [109].
Targeting Insulin Resistance with Metformin
Insulin resistance is believed to be a key component in the pathogenesis of NAFLD. Therefore, the efficacy of metformin has been studied as a treatment option for pediatric NAFLD. In an open-label pilot study of ten children with biopsy-proven NASH, metformin treatment was evaluated using MRS signal fat fraction [110]. Normalization of ALT occurred in 40 % of subjects. There was also a significant reduction in hepatic signal fat fraction in 90 % of subjects from a mean of 30–23 % after 24 weeks of treatment. Nobili and colleagues also conducted an open-label pilot study. In this study, 30 children with biopsy-proven NAFLD underwent 24 months of metformin treatment [111]. Of the 40 % of participants who had a follow-up biopsy, several histologic features including steatosis, ballooning, and lobular inflammation improved after metformin treatment. In the TONIC trial, oral metformin treatment of 500 mg twice daily for 96 weeks did not result in a significant decrease in ALT compared to placebo. In addition, the features of steatosis inflammation, ballooning, and fibrosis were evaluated after 2 years of treatment. There was no significant improvement in steatosis, inflammation, or fibrosis with metformin. However, hepatocyte ballooning was shown to improve in 39 % (22/57) of children taking metformin and only 17 % (10/58) of children taking placebo [105] .
Dietary Supplements
Limited studies are available assessing the efficacy of dietary supplements, such as omega-3 polyunsaturated fatty acids and probiotics, for treatment of NAFLD in children. One randomized clinical trial evaluating treatment with docosahexaenoic acid (DHA) enrolled 60 children with biopsy-proven NAFLD for 6 months. Subjects were randomized to one of three groups: DHA 250 mg/day, DHA 500 mg/day, or placebo. Treatment with DHA did not improve serum ALT or BMI, but was noted to improve insulin sensitivity [112].
Whether manipulation of the microbiome can impact pediatric NAFLD status is unclear, as there are only a few double-blind, placebo-controlled studies of probiotics in children with NAFLD that have been carried out to date. One such study enrolled 20 obese children with abnormal hepatic echotexture on ultrasound and elevated transaminases that persisted for >3 months. They were randomized to treatment with Lactobacillus GG (12 billion CU/day) for 8 weeks or placebo. ALT was noted to decrease significantly from 70 U/L ± 35 to 40 U/L ± 22 in the lactobacillus group over the 8-week study period, whereas in the control group, ALT remained unchanged [113]. The authors noted there was no change in echogenicity measured by ultrasound in either group. In another randomized double-blind placebo-controlled clinical trial, 22 Caucasian children with biopsy-confirmed NAFLD whose median age was 10 years were treated with a proprietary blend of probiotics and compared to 22 children given placebo. There was no significant change in ALT or insulin sensitivity noted in either group. Interestingly, a decrease in BMI of 8 % was noted in the treatment group, while there was no change in BMI in the placebo group [114]. Whether any liver-related feature of NAFLD improves with probiotics is yet to be determined.
Surgery
Bariatric surgery has become an important treatment option for morbidly obese patients and is increasingly used in the adolescent population. Since NAFLD is associated with obesity, bariatric surgery may provide a mechanism for treating NAFLD. Studies in the pediatric population are available that report on significant weight loss and decreased liver chemistries in adolescents who underwent laparoscopic adjustable gastric banding [115, 116], but they do not report data on histologic resolution of NAFLD after bariatric surgery. More studies are needed to determine the efficacy and safety of bariatric surgery to treat pediatric NAFLD.
Outcomes
Children with NAFLD have many associated comorbidities. Outcome data, however, are lacking due to the small number of longitudinal pediatric studies performed to date.
Mortality
Outcome data for pediatric NAFLD are limited with respect to mortality. One study following 66 children with NAFLD for a mean follow-up time of 6.4 years found children with NAFLD to be at higher risk for mortality compared to the general population with a standardized mortality ratio of 13.6 %; two of these children required liver transplant for decompensated cirrhosis, and two died from non-liver-related conditions [117].
Advanced Fibrosis and Cirrhosis
Advanced fibrosis has been reported in 5–15 % of children with biopsy-proven NAFLD at the time of diagnosis [23, 60]. Fibrosis can progress rapidly in some children. For example, in 102 children with NAFLD, the prevalence of advanced fibrosis increased to 20 % after a median follow-up of 2.2 years [118]. Cirrhosis and its sequelae have been observed in children with NAFLD [119].
Hepatocellular Carcinoma
When NAFLD begins in childhood, the long duration of disease raises concern for the future risk of hepatocellular carcinoma . HCC in association with NAFLD has been reported as young as age 7 [41]. Screening efforts should rightfully be directed at those children with NAFLD who have advanced fibrosis. Important clinical issues to address include deciding which children to screen for HCC and which method to use for screening and determining the optimal frequency of screening.
Related posts:
Diagnostic Considerations and Clinical End Points for Nonalcoholic Steatohepatitis
Animal Models of Nonalcoholic Fatty Liver Disease
Clinical Features, Disease Modifiers, and Natural History of Alcoholic Liver Disease
Hepatic and Extrahepatic Malignancies in Alcoholic Liver Disease
Animal Models of Alcoholic Liver Disease
Epidemiology and Risk Factors for Alcoholic Liver Disease
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
