Etiology of Pediatric Acute Liver Failure
The Pediatric Acute Liver Failure (PALF) Study Group, a multicenter and multinational consortium, suggested the following definition: (1) no known evidence of chronic liver disease, (2) biochemical evidence of liver injury, and (3) hepatic-based coagulopathy defined by international normalized ratio (INR) over 1.5 not corrected by vitamin K administration in the presence of hepatic encephalopathy (HE) or INR over 2.0 regardless of the presence of HE.
The etiology of PALF is age dependent ( Table 36.1 ). In infants younger than 3 months, metabolic, infectious, and gestational alloimmune liver disease are the common causes. Galactosemia, tyrosinemia type 1, and mitochondrial disorders are the common etiology of metabolic disease in infants and young children. Herpes simplex virus (HSV) is an important treatable cause in infants and children. In older children, drug toxicity, viruses, metabolic diseases, and autoimmune hepatitis are the main causes. The etiology remains indeterminate in 30% to 50% of the patients, partly because of the lack of proper diagnostic evaluation. About 10% of indeterminate PALF patients can develop bone marrow failure, and some progress to aplastic anemia, even after successful liver transplantation (LT).
|• Metabolic Galactosemia Tyrosinemia Mitochondrial disorderNiemann-Pick type C||• Metabolic Wilson disease Mitochondrial disorder|
|• Autoimmune Gestational alloimmune liver disease||• Autoimmune Autoimmune hepatitis|
|• Infectious Herpes simplex virus Enterovirus||• Infectious Hepatitis A EBV|
|• Vascular Heart failure||• Vascular Veno-occlusive disease|
|• MalignantHLH||• Toxic Acetaminophen|
|• Indeterminate||• Indeterminate|
Diagnostic Procedures Necessary to Prepare for Liver Transplantation
Diagnostic evaluations must be extensive to evaluate the etiology of PALF ( Table 36.2 ). Early establishment of the etiology provides the predictor of outcomes and identifies conditions that are treatable without LT or contraindicated for LT. Potentially treatable diseases are HSV infection, gestational alloimmune liver disease, acute acetaminophen toxicity, autoimmune hepatitis, and Wilson disease. Gestational alloimmune liver disease is confirmed by demonstration of extrahepatic iron deposits by a buccal salivary gland biopsy or magnetic resonance imaging (MRI). Autoimmune hepatitis presenting with PALF is mostly of type 2, with positive anti–liver kidney microsome antibodies. Coombs-negative hemolytic anemia, a high bilirubin-to-alkaline phosphatase ratio, and renal failure are characteristic features of ALF attributed to Wilson disease. An elevated ratio of lactate to pyruvate is a sensitive but nonspecific marker of mitochondrial disorders, and it is important to rule out neuromuscular involvement. Brain MRI, cerebral spinal fluid analysis, muscle enzymes, muscle biopsy for mitochondrial and respiratory chain analysis, echocardiogram, and assessment of renal tubular function are important to exclude systemic disease. Hepatocerebral mitochondrial DNA depletion syndromes are autosomal recessive diseases caused by mutations in DGUOK , MPV17 , and POLG . A next-generation sequencing panel for candidate genes may be the best diagnostic strategy to identify these disorders. Recognition of hemophagocytic lymphohistiocytosis (HLH) as a cause of PALF is important because LT is not curative.
|Viral||HSV PCR, ant-HSV IgM, HSV culture of blood or CSF|
|Anti-HAV IgM, HBsAg, anti-HBc IgM, anti-HCV, anti-HEV, EBV VCA IgM, EBV PCR|
|Viral study for cytomegalovirus, adenovirus, enterovirus, human herpes virus-6, parvovirus, influenza|
|Metabolic||Newborn screen for galactosemia (< 3 months)|
|Urine succinylacetone for tyrosinemia (< 3 months)|
|Plasma acylcarnitine profile and urine organic acid for fatty acid oxidation defect|
|Plasma lactate and pyruvate for mitochondrial disorder|
|Serum amino acid for urea cycle defect|
|Urine orotic acid for ornithine transcarbamylase deficiency|
|Serum ceruloplasmin, 24-hour urine copper (> 3 years)|
|Autoimmune||Antinuclear antibody, anti–smooth muscle antibody, anti-LKM antibody, immunoglobulin G|
|Ferritin, iron, total iron-binding capacity (< 3 months)|
|HLH||Ferritin, triglycerides, fibrinogen, IL2R|
|Drug||Serum acetaminophen level|
|Vascular||Abdominal ultrasound, echocardiography|
The role of liver biopsy in PALF is controversial. Severe coagulopathy makes percutaneous liver biopsy high risk, and sampling error is possible. Liver histology does not increase the diagnostic yield and may not differentiate autoimmune PALF from indeterminate PALF. In contrast, liver biopsy can be safely performed by the transjugular approach, and histological findings may guide diagnosis and therapy.
Clinical tools to predict irreversible brain injury are lacking. Neuroimaging is frequently normal until late in the course of neurological injury. Brain-computerized tomography or MRI was abnormal in only 13% of PALF patients and was not associated with outcome. Brain imaging may be used to exclude other causes of decline in mental status, such as intracranial hemorrhage. Children with electroencephalogram abnormalities, including slowing and epileptiform discharges on admission to the pediatric intensive care unit (PICU), were more likely to require LT or die.
General Management Awaiting Liver Transplantation
Management of PALF should be performed in PICU within an LT center, where continuous monitoring and multidisciplinary team are available. Oral intake should be encouraged but should be avoided if HE progresses. Protein restriction under 1 g/kg is not recommended unless the patient has an underlying urea cycle defect or organic academia. There is insufficient evidence to prescribe branched-chain amino acids. Lipids can be started unless fatty acid oxidation disorder or mitochondrial disease is suspected.
The fundamental hemodynamic abnormality in ALF is systemic vasodilation with reduced effective central blood volume. Patients with ALF are at risk for hypovolemia owing to decreased oral intake and decreased effective central volume. The initial treatment of hypotension is effective with intravenous normal saline. In volume refractory hypotension, norepinephrine may augment peripheral organ perfusion while minimizing tachycardia and preserving hepatic blood flow. Despite the lack of pediatric data, norepinephrine is a logical choice in patients with PALF.
In the PALF Study Group database, 10% of children with ALF needed hemofiltration support. When dialysis is needed, continuous renal replacement therapy (CRRT) is preferred because of less hemodynamic instability and less risk of HE compared with intermittent hemodialysis. CRRT is useful for managing fluid overload and acute kidney injury.
Patients with ALF are at risk for infection or sepsis, which may preclude LT or complicate the postoperative course. Regular periodic surveillance cultures should be performed to detect bacterial and fungal pathogens. A prospective adult study has suggested an association of infection and the systemic inflammatory response (SIRS) with progression to deeper stages of HE. Empiric antibiotics should be initiated promptly at signs of SIRS, refractory hypotension, or progression to grade 3 to 4 HE.
Despite elevated INR, most patients with ALF maintain normal hemostasis. In the absence of bleeding, plasma transfusion is not advised because it influences the PT, an important marker of prognosis. Plasma transfusion is recommended before invasive procedures or in the setting of active bleeding. Patients who develop significant bleeding with a platelet count below 50,000/mm 3 should be transfused. During LT, optimal hematocrit is 23% to 30%, and patients may need platelet transfusions if the platelet count is less than 20,000/mm 3 in the absence of active bleeding.
The degree of HE is the best predictor of outcome for children with ALF ( Table 36.3 ). Mortality was highest in patients with PALF with severe or worsening HE. Cerebral edema and intracranial hypertension can manifest as systemic hypertension, bradycardia, hypertonia, decerebrate posturing, hyperreflexia, and pupil dilatation in patients who progress to grade 3 or 4 HE. In patients with grade 3 to 4 HE, endotracheal intubation and mechanical ventilation should be undertaken. Sedation should be avoided unless needed in the intubated patient because of aggravating HE or precipitating respiratory failure. There is insufficient data to recommend the routine use of intracranial pressure (ICP) monitoring in patients with PALF. In the absence of ICP monitoring, frequent neurological evaluation is recommended to identify early evidence of intracranial hypertension. If intracranial hypertension develops, mannitol is recommended.
|1||Confusion, mood change, altered sleep habits, inconsolable crying, not acting like self|
|2||Drowsy, inappropriate behavior, disorientation, inability to interact with or recognize parents|
|3||Stupor, somnolence but arousable to vocal stimuli, hyperreflexia, positive Babinski sign|
|4||Comatose, response to pain or no response to pain, decerebrate or decorticate posturing|
Children with acetaminophen overdose should be treated with N-acetylcysteine (NAC). A pediatric randomized controlled study does not support the use of NAC in children with nonacetaminophen PALF. Newborns with ALF may have HSV infection and should be treated with acyclovir until the results for HSV are known. Gestational alloimmune liver disease is treated with intravenous immunoglobulin and exchange transfusion. Nucleos(t)ide analogues should be considered for hepatitis B–associated PALF and for prevention of post-transplant recurrence. Some children with autoimmune hepatitis may respond well to steroid therapy, and others may require LT. Steroids may be detrimental in children with a known diagnosis other than autoimmune hepatitis. The management of HLH is immunosuppressive therapy or hematopoietic stem cell transplantation. Children in whom Wilson disease is the likely cause of ALF should be promptly listed for LT. A patient showing a lack of progressive deterioration or evidence of improved coagulation parameters while awaiting graft availability has a prediction for spontaneous recovery, so the decision to LT is reversed.
Liver Support Devices
CRRT is the mainstay treatment for managing acute kidney injury in PALF. CRRT may be offered to manage hyperammonemia, high lactate, and HE. CRRT was initiated for ammonia over 200 μmol/L or HE greater than grade 2.
In children with Wilson disease presenting ALF, released copper can cause hemolysis and renal tubular damage. Plasma exchange can be used to remove serum copper, limit hemolysis, and treat coagulopathy. Plasma exchange in a randomized controlled study has been shown to improve transplant-free survival in adult patients with ALF and to decrease SIRS and multiorgan dysfunction. Similar studies are needed in children.
In patients with PALF, a molecular adsorbent recirculating system showed improvement in ammonia, bilirubin, and creatinine levels, but no survival benefit.
Indication and Timing for Liver Transplantation
LT is the only definitive treatment for patients who are unable to achieve regeneration of sufficient hepatocyte mass to sustain life. Recent reports have shown that 22% to 31% of patients with PALF undergo LT. LT is indicated in ALF where prognostic indicators suggest a high likelihood of death. Acute respiratory failure, acute kidney injury, and cerebral edema are associated with increased risk of mortality.
Decisions to proceed to LT in PALF are complicated by difficulties in predicting outcome. Several prognostic laboratory variables have been identified, and the incorporation of these variables into scoring systems has been attempted to select candidates for LT ( Table 36.4 ). A recent meta-analysis of the King’s College Hospital criteria, which is used extensively in adult patients with ALF, showed good specificity but poor sensitivity in predicting LT-free mortality. The King’s College Hospital criteria do not reliably predict death in nonacetaminophen-induced PALF and may lead to overuse of LT in children who would otherwise survive despite meeting criteria. The Clichy criteria were developed in a cohort of French patients with viral hepatitis. The sensitivity of the Clichy criteria was low and could be improved by the integration of bilirubin and creatinine clearance for nonacetaminophen ALF. The Clichy criteria have not been validated in pediatric patients. The pediatric end-stage liver disease (PELD) score has been used to aid the prioritization of LT to children with chronic liver disease. Experiences with the PELD score in PALF are limited in small cohorts. Separate prioritizing systems for LT are needed in children with chronic liver disease and ALF. Liver Injury Units (LIU) score using peak values of bilirubin, INR, and ammonia was proposed for PALF. The validation study for the PALF Study database revealed that the highest daily LIU score–calculating laboratory values obtained on the same day had good discrimination for both death and LT. Prognostic models often combine death and LT as one outcome, but because some patients undergoing LT would have spontaneously recovered, models comparing spontaneous survival with death without LT are more appropriate. The revised Wilson disease prognostic index, a disease-specific score, has good predictive values for mortality without LT. The ideal prognostic model should reflect the dynamic nature of PALF. The ALF early dynamic (ALFED) model is based on dynamicity of variables during 3 days of admission. When the ALFED model was applied to PALF patients, specificity was good but sensitivity was poor in predicting LT. A prognostic scoring system using the change of total bilirubin and INR can predict the mortality in patients with PALF.