A significant health problem in children, hepatobiliary disorders include anatomic disorders (biliary atresia, choledochal cyst), autoimmune diseases (autoimmune hepatitis, primary sclerosing cholangitis), viral hepatitis (hepatitis A, B, C, delta agent, E), metabolic liver diseases and drug-induced hepatic toxicity.
Extrahepatic biliary atresia (EHBA), which affects 1 in 8,000 to 1 in 15,000 children, is a leading cause of morbidity and mortality in children with gastrointestinal (GI) or liver disease. EHBA is characterized by fibro-obliterative destruction of all or part of the extrahepatic biliary system. The etiology of EHBA is not known. Putative factors include viral infections, a genetic predisposition, immune-mediated cholangiopathy and developmental defects.
Two consensus conferences have been convened by the National Institutes of Health (NIH) to focus on extrahepatic biliary atresia (1,2). Although significant progress has clearly been made in the care of children with EHBA, the field nevertheless is only “halfway there” (3). EHBA is the leading indication for liver transplantation in the pediatric age group (4). Even though patients have relatively high survival rates as a result of management with hepatoportoenterostomy or liver transplantation (5,6), much remains to be done before the highest quality of care can be attained.
Acute liver failure. In most pediatric series, 10% of liver transplantations are performed in patients with acute liver failure. As such, between 50 and 75 children each year undergo liver transplantation for acute liver failure. In such patients, extensive hepatocyte damage has resulted in diminished liver synthetic function and encephalopathy. The condition is considered fulminant when the patient has no prior history of liver disease. Depending on the definition of acute liver failure, estimates of the incidence of this disorder range from 200 to 1000 cases each year in the US pediatric population. Viral hepatitis and drug toxicity can cause acute liver failure in children, but in many pediatric cases, the cause is unknown. The case-fatality rate for idiopathic fulminant liver failure associated with encephalopathy can reach 90%. Current therapy is limited to supportive care and liver transplantation.
Chronic viral hepatitis affects more than 300,000 children in North America. As many as 150,000 have chronic hepatitis C (HCV); yet we know little about the natural history nor is there effective therapy for pediatric patients.
Worldwide, the most common form of acute and chronic hepatitis is hepatitis B (HBV). In the US alone, 150,000 new cases are estimated each year, and about 1 million people are chronically infected with HBV. Children are much more prone to develop chronic HBV if they acquire an acute infection prior to 1 year of age. Consequently, childhood infection accounts for a disproportionate number of chronically infected individuals with HBV. Most children chronically infected with HBV will not respond to currently available therapy.
The incidence of hepatitis A peaks in children between the ages of 5 and 14 years, reaching 17 per 100,000 population (7). In 1997 a total of 30,000 cases of hepatitis A were reported to the National Notifiable Diseases Surveillance System in the US. About 7% of all children and adolescents with hepatitis A were hospitalized, resulting in an average cost of $1500 per case.
Metabolic pathways in the liver. The liver is not physiologically mature at birth. The metabolism of bilirubin is not fully developed, and numerous pathways for the disposition of drugs and toxins are still immature. While some metabolic pathways are the focus of active research, interactions between the developing liver and exposure to drugs and toxins in the environment have not been well studied. With recent advances in our understanding of hepatic drug metabolism, the stage is set for a developmental approach to investigations of adverse effects of medications and the role of the liver.
A major limitation in addressing key clinical questions has been the relatively small numbers of patients cared for at any given center. Assessment of treatment approaches and development of trial designs are predicated on the availability of high-quality, prospectively collected data on the natural history of a disease, such as could be provided by a multicenter database. For example, the natural history of portal hypertension, especially variceal hemorrhage, could be defined prior to implementation of a multicenter interventional trial. Ascending cholangitis could be studied and its prevalence correlated with the progression of disease. Growth parameters in biliary atresia could be prospectively assessed and correlated with disease status.
AREAS OF EMPHASIS
Evaluate Efficacy of Corticosteroids After Portoenterostomy for EHBA
To date, no randomized treatment trials in EHBA have been performed. However, recent descriptions of center-specific excellence in the care of children with EHBA suggest that optimal clinical approaches do exist. Important research questions in the area of EHBA include the effect of steroids in post-portoenterostomy patients.
Some centers have advocated the use of corticosteroids in the postoperative care of children with EHBA. The rationale for corticosteroid use is to reduce inflammation in the bile ductules at the level of the porta, thereby enhancing bile flow.
For any particular hepatobiliary disorder, no single medical center has enough patients to enable detailed study of the etiology and management. It is clear that for pediatric hepatobiliary research to advance, a mechanism for collaborative clinical research needs to be established.
A Children's Liver Study Group (CLSG) would provide the infrastructure to conduct multicenter clinical trials and collaborative research investigations. Such an organization would comprise a scientific advisory board, a central data coordination site, and a central biological-materials storage facility. Centers that are CLSG members would be eligible to participate in CLSG-sponsored clinical trials, contribute to a CLSG database and receive funding for a clinical research coordinator.
Members of the scientific advisory board would be selected on the basis of demonstrated expertise and commitment to pediatric hepatology. Physicians, nurses, statisticians and lay persons would be represented on the board. Board responsibilities would include organization of an annual meeting; the review and approval of research proposals, including requests to use either the central database or biological-materials bank; and the review and selection of centers to be funded.
The data coordination site is a fundamental requirement for an effective study group. This site would be a central repository for longitudinal information collected by CLSG member centers. The site would also be responsible for subsequent data analysis and for assisting member centers with research questions. The central biological-materials facility would collect and organize liver, bile, serum, fibroblast and other samples obtained from children with liver diseases. Separate supervisory panels would set up the two sites and initially review requests for materials.
Study design. With a CLSG in place, a multicenter randomized trial is recommended to evaluate the use of corticosteroids after portoenterostomy for EHBA. The proposed study would enroll infants with EHBA in whom portoenterostomy was performed before they were 10 weeks old. Subjects with known hypersensitivity to corticosteroids would be excluded.
Infants could be randomized prior to portoenterostomy to one of three arms of this study: 1) empiric and interventional corticosteroid administration, 2) interventional corticosteroid administration only, and 3) no corticosteroid administration. In the first treatment arm, empiric therapy would consist of prednisolone taper over 4 weeks after portoenterostomy. Interventional corticosteroid administration would consist of 7-day prednisolone pulse for all episodes of cholestasis, including poor postoperative drainage (acholic stools or no reduction in serum bilirubin level) and subsequent episodes of cholestasis (i.e., development of acholic stools or increase in total bilirubin levels to >5 mg/dL).
The primary study endpoint would be liver transplantation or death. Secondary endpoints would include total bilirubin levels at 12 and 24 months of age, blood culture-positive episodes of ascending cholangitis, and linear growth velocity at 12, 24 and 36 months of age.
Projected Timetable and Funding Requirements
Funding of a CLSG would be required to cover costs associated with major activities and administrative expenses in support of a scientific advisory board, central data coordination site and biological-materials bank.
A single source of funding would likely prove insufficient to cover CLSG expenses. Potential funding sources might include the NIH, organizations such as the American Liver Foundation, and pharmaceutical firms.
Evaluate Efficacy of Beta-Adrenergic Blockade in Preventing Complications of Portal Hypertension in Children With EHBA
In a number of adult studies, nonspecific beta-blocker therapy decreased the risk of variceal hemorrhage. Interestingly, the effect of beta-blockade on portal pressure is greatest in patients without varices. Animal models have indicated that beta-blockade may actually prevent the development of portal hypertension.
Children with EHBA universally develop some stigmata of portal hypertension. Thus, they are an ideal study population for the investigation of novel approaches to the treatment and prevention of portal hypertension. Preliminary nonrandomized data suggest that beta-blockade is safe in children and may be useful in preventing variceal hemorrhage (8).
A multicenter clinical trial would enroll subjects who had had a portoenterostomy for biliary atresia prior to 10 weeks of age. Subjects would be excluded if they have a known hypersensitivity to Inderal (propranolol), known reactive-airways disease, or an underlying heart condition that is a contraindication to beta-blockade.
The infants could be randomized 1 month after successful portoenterostomy to receive either propranolol or placebo. Baseline studies should include electrocardiography and Holter monitoring to assess baseline heart rate. The starting dose of propranolol could be 1.0 mg/kg administered bid or tid, with a 50% dose escalation every week until a 25% reduction in heart rate is achieved.
Primary study endpoints would be a) complications related to portal hypertension, including variceal hemorrhage, ascites, hepatopulmonary syndrome and hepatic encephalopathy; and b) either liver transplantation or death. A secondary endpoint could be the presence of esophageal and gastric varices, as documented by upper GI endoscopy, at the time of liver transplantation.
Prospectively Analyze the Etiologic Role of Reovirus, Rotavirus and Cytomegalovirus in EHBA
Perhaps the most fundamental unresolved issue regarding EHBA is its etiology. Reports have described fetal and perinatal forms of EHBA, which may be distinct diseases. The recent description of an entity akin to biliary atresia in a mouse with a defect in sidedness supports such a contention (9). A number of viruses are thought to play an etiologic role, including reovirus 3, cytomegalovirus and group C rotavirus. Recent advances have indicated that immune responses may also be critical in the pathogenesis of EHBA. Clearly, a better understanding of the etiology is essential for better diagnostic accuracy, improved treatment regimens and ultimately prevention of EHBA.
Infants younger than 6 months of age would be recruited into a multicenter clinical study. Inclusion criteria would include EHBA, neonatal cholestasis, choledochal cyst and pyloric stenosis (sepsis would be ruled out). In subjects with a qualifying diagnosis, serum and whole-blood DNA samples would be obtained at clinical presentation and 4 months later.
Biliary tissue should be obtained in subjects with a qualifying diagnosis who undergo a clinically indicated procedure in which such biopsies might be routine (e.g., choledochal cyst repair or hepatoportoenterostomy). Liver tissue should be obtained in infants with a qualifying diagnosis who undergo a laparotomy (e.g., exploratory laparotomy for neonatal cholestasis or choledochal cyst repair). Serologic and DNA diagnostic procedures should be performed in an investigative reference laboratory that is blinded to the clinical diagnosis of the subject in question. Similarly, investigations of immune function (e.g., lymphocyte activities, antibody production) should be performed in a blinded fashion by appropriate investigative laboratories.
Analyze Molecular Events Leading to Development of Fibrosis and Cirrhosis in EHBA
Potential research issues of general relevance to children with liver disease include the prevention of variceal hemorrhage, the role of choleretic agents in treatment of cholestatic liver disease, and mechanisms for the prevention of fibrosis and cirrhosis. Over the past ten years, much has been learned about the molecular events contributing to the development of fibrosis (10). This knowledge needs to be examined in a comprehensive and systematic fashion in infants with EHBA. With further advances in our knowledge base, therapies might be developed that can slow or arrest the development of cirrhosis in EHBA.
Children who have received a diagnosis of EHBA and undergone laparotomy would be enrolled in a multicenter study. Patients would be excluded who have uncorrectable coagulopathy or thrombocytopenia precluding liver biopsy.
Wedge liver biopsy should be obtained in children with EHBA who undergo laparotomy. Similarly, liver tissue should be obtained in children with EHBA who undergo liver transplantation. Molecular and immunohistochemical analyses of these samples would be performed.
Identify Risk Factors for Development of Acute Liver Failure of Unknown Etiology
There is mounting evidence that the causative agent of acute liver failure is a virus. The illness is preceded by a viral infection-like prodrome, including fatigue, loss of appetite and GI complaints. Episodes of this illness tend to cluster in geographic areas and are more common in the winter months. Between 5% and 25% of patients develop hypoplastic/aplastic anemia, which is a known complication of other forms of viral hepatitis (5). An alternative theory is that the liver injury is caused by an atypical host response to a common viral pathogen. Thus far, attempts to define the epidemiology of this rare disorder have been unsuccessful.
Potential areas of investigation include the identification of host factors that predispose an individual to the development of severe liver injury. Studies of potential host factors might include, for example:
Screening for HLA subtypes that are more prevalent in patients who develop acute liver failure
Testing for specific genetic defects that precipitate a metabolic crisis following periods of fasting
Identifying subtle immune deficiencies that cause the host to be more susceptible to virus-mediated liver injury
Collection of serum and liver specimens from affected individuals would allow analysis of novel viral particles and future testing of the frequency and clinical relevance of newly discovered hepatotrophic viruses.
A large collaborative research group focused on studying acute liver failure in children would be essential to accomplishing these goals. A multicenter case-control study of this condition in the pediatric population has yet to be accomplished.
For a matched case-control study, it is recommended that patients between the ages of 6 weeks and 18 years be enrolled who have acute liver failure, defined as progression of acute hepatitis within an 8-week period to liver dysfunction. The inclusion criteria would include a) prothrombin time > 20 seconds or international normalized ratio (INR) > 2, b) hepatic encephalopathy of any grade, c) negative serology for known hepatotropic viruses, and d) a negative workup for Wilson's disease and autoimmune hepatitis. The matched controls could be parents available for a phone interview.
Subjects would be excluded who have a history of toxic ingestion, including acetaminophen use exceeding 150 mg/kg/day, or a known history of illicit drug use. Exclusion criteria for controls would include a history of significant liver disease, toxic ingestion (including excessive acetaminophen use) and illicit drug use.
Cases would be identified by principal investigators at participating centers when patients present for management of acute liver failure. Case report forms, including demographic and exposure data, would be completed by family interview. The family would be asked to identify two peer children as control subjects. Controls would be matched for age, area of residence and school (if applicable). The parents of control subjects would be interviewed by telephone to collect the same demographic and exposure data collected for the case subject.
Blood would be collected at enrollment for a bank of serum, genomic DNA and lymphocytes. In those children who undergo liver transplantation of postmortem analysis, liver, bile and skin fibroblasts would be collected. These samples would be available for future analysis of potential etiology.
Evaluate Efficacy of Prostaglandin Infusions for Severe Acute Liver Injury
Prostaglandins play an important role in regulating cell growth and immune function in the liver. The therapeutic use of prostaglandin E (PGE) infusion in the setting of serious liver injury has yielded conflicting results. There appears to be some benefit, particularly in patients with drug toxicity who are treated early during the course of the illness. However, the benefit of PGE infusion has never been carefully studied in a pediatric population with acute liver failure.
A multicenter randomized trial would enroll infants and children (up to 18 years of age) who have severe acute liver injury, based upon the inclusion criteria listed above. Potential participants would be excluded if they had contraindications to PGE infusion, such as congenital heart disease or demonstrated hypersensitivity.
Children meeting the inclusion criteria would be enrolled and stratified according to age group, cause of liver injury, and degree of coagulopathy and encephalopathy. Enrollees would be randomized to treatment with PGE infusion or placebo.
Primary study endpoints could include hospital discharge, liver transplantation and death. Potential secondary endpoints include number of days in the intensive care unit, need for dialysis, length of hospital stay and blood product requirements.
Screen for Presence of Fatty Acid Oxidation Defects in Patients With Acute Liver Failure of Unknown Etiology
The association of fatty acid oxidation defects with isolated liver failure is a relatively new observation (11). Previously healthy children presenting with acute liver failure frequently are not screened for fatty acid oxidation defects. By measuring fatty acid uptake and metabolism in cultured fibroblasts of patients who present with acute liver failure, the frequency of fatty acid oxidation defects as a cause of acute liver failure in children can be defined.
Study subjects would meet the inclusion criteria of acute liver injury listed above. In addition, subjects would be negative on serologic testing for known hepatotrophic viruses, autoimmune liver disease, and other common forms of metabolic liver disease, including Wilson's disease, tyrosinemia and α1-antitrypsin deficiency. Exclusion criteria would include a history of toxic ingestion, including acetaminophen use exceeding 150 mg/kg/day.
Data and specimens to be collected from study subjects should include a urine sample for organic acid analysis, a serum sample for carnitine profile and free fatty acid analysis, a whole-blood sample for DNA analysis of known genetic defects in fatty acid oxidation, a skin biopsy specimen for fibroblast culture and subsequent measurement of fatty acid oxidation, and liver tissue (biopsy, explant or autopsy specimen), if available, for analysis of fatty acids and carnitine content.
Prospectively Analyze Acetaminophen Hepatotoxicity Related to Therapeutic Misadventure in Children
Because acetaminophen is so widely given to children, a determination of its role in liver injury would have important public health implications. A recommended research topic is the prospective analysis of the frequency of acetaminophen hepatotoxicity related to therapeutic misadventure in children. Reliable indications of acetaminophen-induced liver injury include positive staining of fixed liver tissues and serum immunoassays for acetaminophen/protein adducts (12,13).
Study subjects would be children identified with acute liver failure as described above. There would be no specific exclusion criteria as patients with toxic acetaminophen ingestion would serve as a positive control.
Careful historical data would be collected for each study subject to ascertain any acetaminophen use in the 3 months prior to enrollment in the study. Serum for analysis of 3-(cystein-S-yl)-acetaminophen protein adducts (3-cys-A) would be collected at enrollment and at 24-hour intervals. Available liver tissue would be fixed in formalin and processed for immunohistochemistry of 3-cys-A. Liver specimens obtained from patients who require liver transplantation or at autopsy would be snap-frozen and subsequently analyzed by immunoblotting to quantify 3-cys-A tissue levels. Levels of 3-cys-A in serum and liver specimens would be correlated with outcome measures such as survival, liver transplantation and time to recovery of normal liver function.
Describe the Natural History of Chronic Hepatitis
Chronic HCV appears to progress slowly. Children who acquire the infection are likely to be at increased risk of long-term complications, including cirrhosis and hepatocellular carcinoma. There is a pressing need to understand the natural history of HCV acquired in infancy. It is possible, for example, that HCV acquired by vertical transmission may be relatively quiescent, and similar to the course of vertically acquired chronic hepatitis B. However, because the immunologic mechanisms underlying liver injury with HCV and HBV do differ, it is also possible that vertical transmission of HCV will result in an inexorable progression to cirrhosis.
Studies are recommended to analyze the immunologic response to hepatitis C in children. What immunologic factors contribute to the clearance of hepatitis C in some patient populations, while the infection persists in those who acquire the virus vertically? Studies could include an evaluation of immunologic competency, tolerance and viral factors.
As with HCV, the natural history of chronic HBV has not been well characterized to date. Factors need to be identified that can lead to adverse outcomes with chronic infection. This information will assist in targeting therapies for children at increased risk of developing end-stage liver disease.
Finally, non-alcoholic fatty liver disease (or non-alcoholic steatohepatitis [NASH]) has recently been identified as one of the most prevalent forms of liver disease in the US. The natural history of this disorder in children is unknown. In fact, diagnostic criteria and standardized evaluations for children have not been determined.
Multicenter epidemiologic trials are recommended to study the natural history of chronic hepatitis.
Evaluate Potential Antiviral Therapies for Children With Chronic Hepatitis Infection
There is a pressing need for clinical trials of antiviral therapy in children with chronic hepatitis infection. Although antiviral therapy, with interferon or interferon plus ribavirin, is currently approved for use in adults with chronic HCV, no therapy has been approved for use in infected children. Interferon therapy is approved for use in children with chronic HBV, but response rates do not exceed 25% to 30%. 14 Effective therapies for HBV and HCV have yet to be developed.
Multicenter clinical trials are needed because no single center has enough patients to adequately power a study. Long-term follow-up studies must also be designed, especially in the treatment of Hepatitis B. The outcome of hepatitis B e antigen (eAg) seroconversion in adults is just now being assessed and this information is unavailable for children. The influence of seroconversion on long-term risk of the development of cirrhosis and hepatocellular carcinoma will require 10 to 20 years' follow-up.
Evaluate the Natural History of NASH in Children
A multitude of questions exist regarding NASH in children. Many are quite fundamental, including issues of prevalence and natural history.
As an initial step in understanding this disorder, prospective information would be collected in a multicenter study of the natural history of NASH in children. Data to be collected would include demographic, vital statistics, standard diagnostic investigation, histology and response to empiric therapies (weight loss, vitamin E, ursodeoxycholic acid). Preliminary analysis of this information will permit design of prospective investigations of this important disorder.
HEALTH AND ECONOMIC OUTCOMES
Surveillance studies have not yet been conducted to determine the precise prevalence of liver disease in children. A recent report suggested that 0.2% of all children in the US younger than 12 years of age were infected with hepatitis C virus. 15 The economic impact of pediatric liver disease is largely unknown. An evaluation of the cost of liver transplantation in adults concluded that charges for the first year of care routinely exceed $150,000 (16). Assuming similar costs for pediatric liver transplantation, approximately $82.5 million of health care charges are generated each year to provide this therapy for children. The cost of care for patients with chronic liver dysfunction who do not advance to end-stage liver disease is substantially less, but because the number of these patients far outweighs the number requiring transplantation, the financial burdens are still considerable.
The total cost of caring for children with liver disease includes not only the direct cost of health care delivery but also the cost to society of caring for the sick individual. Economic analyses of health care costs frequently include indirect costs of illness, such as absenteeism from work, decreased earning ability and the value of quality-of-life-years lost. Such calculations are difficult in children who have not yet demonstrated their potential societal contribution as adults. Nevertheless, it would appear that indirect costs are substantially greater for children. Caring for a seriously ill child can involve parents, siblings, and extended family members. Many parents must take an extended leave of absence from work. The impact of disease on the family is not routinely measured in adult studies, but should be a central focus of quality-of-life research in the pediatric population.