ARTICLE |
a Pediatrics
b Pharmacology and Toxicology, University of Arkansas for Medical Sciences and Arkansas Children's Hospital Research Institute, Little Rock, Arkansas
c Department of Pediatrics, Siragusa Transplant Center, Children's Memorial Hospital and Northwestern University Feinberg School of Medicine, Chicago, Illinois
d University of Texas Southwestern Medical Center at Dallas, Dallas, Texas
e University of California at San Francisco, San Francisco, California
f Children's Hospital of Pittsburgh, Pittsburgh, Pennsylvania
| ABSTRACT |
|---|
|
|
|---|
METHODS. Serum samples from children with acute liver failure were measured for acetaminophen cysteine protein adducts using high-performance liquid chromatography with electrochemical detection. For comparison, samples from children with well-characterized acetaminophen toxicity and children with known other causes of acute liver failure also were measured for acetaminophen cysteine protein adducts. The analytical laboratory was blinded to patient diagnoses.
RESULTS. Acetaminophen cysteine protein adduct was detected in 90% of samples from children with acute liver failure that was attributed to acetaminophen toxicity, 12.5% of samples from children with acute liver failure of indeterminate cause, and 9.6% of samples from children with acute liver failure that was attributed to other causes. Adduct-positive patients from the indeterminate cause subgroup had higher levels of serum aspartate aminotransferase and alanine aminotransferase and lower levels of bilirubin. Adduct-positive patients also had lower rates of transplantation and higher rates of spontaneous remission.
CONCLUSIONS. A small but significant percentage of children with acute liver failure of indeterminate cause tested positive for acetaminophen cysteine protein adducts, strongly suggesting acetaminophen toxicity as the cause of acute liver failure. An assay for the detection of acetaminophen cysteine protein adducts can aid the diagnosis of acetaminophen-related liver injury in children.
Key Words: acetaminophen liver failure alanine aminotransferase
Abbreviations: APAPacetaminophen ALFacute liver failure NAPQIN-acetylbenzoquinoneimine APAP-CYSacetaminophen cysteine adducts PALFPediatric Acute Liver Failure PRprothrombin time INRInternational Normalized Ratio HPLC-EChigh-performance liquid chromatography with electrochemical detection ALTalanine aminotransferase NACN-acetylcysteine
Acetaminophen (APAP) is the most widely used drug for the treatment of pain and fever in the world today. Although the drug has a remarkable record of safety when used at recommended doses, in overdose, the drug may cause fulminant hepatotoxicity. Recent data suggest that APAP currently is the most common cause of acute liver failure (ALF) in adults in the United States.1,2
The initial stages of APAP toxicity were well characterized in the 1970s.35 At therapeutic doses, the parent drug undergoes sulfation and glucuronidation to form conjugates that undergo renal elimination. A minor portion of the drug undergoes P450 activation, primarily through CYP2E1, forming N-acetylbenzoquinoneimine, (NAPQI), an electrophilic metabolite that binds to cysteine groups on protein. With toxic doses of APAP, increased amounts of NAPQI are produced, resulting in formation of cysteine adducts (APAP-CYS). Whereas the critical events that occur subsequent to adduct formation are poorly understood, formation of APAP-CYS is a widely recognized correlate of APAP-mediated hepatic injury in animal models of toxicity.68 These studies68 have shown that APAP-CYS adducts are formed concurrent with glutathione depletion, a critical early event in the toxicity; APAP-CYS adducts are formed in the hepatocytes that subsequently undergo APAP-mediated necrosis; and serum APAP-CYS levels are reflective of adducts that are formed in the liver.
We previously reported the development of a highly sensitive and specific assay for the measurement of APAP-CYS in human blood samples.9 In adults with well-characterized APAP-related ALF, APAP-CYS was present in sera for up to 7 days after presentation to the medical facility, a time point when APAP may no longer be detectable in serum or plasma.10 Therefore, the assay has the potential to identify patients with APAP-related ALF when presentation to the medical facility is delayed. In this study, we used this assay to measure adducts in sera from children who were enrolled in the multicenter Pediatric Acute Liver Failure (PALF) Study Group to determine the role of APAP toxicity in ALF of indeterminate cause in this population.
| METHODS |
|---|
|
|
|---|
18 years of age at participating centers were enrolled in the PALF Study when they met the following criteria: (1) no known evidence of chronic liver disease, (2) biochemical evidence of acute liver injury, and (3) coagulopathy not corrected by vitamin K. The presence of encephalopathy was required for diagnosis of ALF when the prothrombin time (PT) was between 15 and 19.9 seconds or the International Normalized Ratio (INR)11 was between 1.5 and 1.9 but was not required when the PT was
20 seconds or INR was
2.0. A coma grade scale was developed for infants and children who were younger than 4 years, and standard adult coma scales were used for older children. Informed consent and assent were obtained from the parents and the study children who were enrolled in the registry.
Diagnostic Categories
Diagnostic criteria for acute APAP toxicity included a toxic serum APAP level as defined by the Rumack nomogram12 or a history of an acute ingestion of 100 mg/kg APAP within a 24-hour period and exclusion of other common causes of acute hepatitis. The diagnosis of autoimmune hepatitis was established when a patient had 1 or more positive autoantibody tests (anti-nuclear antibody
1:80, smooth muscle antibody
1:20, or liver-kidney microsomal antibody
1:20) and no evidence of serologically defined viral hepatitis.13 Drug-induced hepatitis was diagnosed when a temporal relationship between exposure to a suspected drug and the onset of ALF was established and other common causes were excluded. Hepatitis A, B, or C infection was confirmed serologically or by polymerase chain reaction. Evidence of other viral infections required a positive IgM antibody, evidence of virus in liver tissue, or a positive polymerase chain reaction. Metabolic diseases were diagnosed by laboratory tests (eg,
-1-antitrypsin phenotype of ZZ) or analysis of liver tissue (eg, fatty acid oxidation defect). Perfusion abnormalities that result in ALF, such as Budd-Chiari malformation, veno-occlusive disease, and ischemia related to hypotension, were established by characteristic history and/or Doppler ultrasound findings. When the site investigator suspected an infection or metabolic disease but lacked supporting evidence or when a specific diagnosis could not be established, the final diagnosis was registered as indeterminate ALF.
Data Collection
After informed consent was obtained, a detailed case report form was completed by the local site investigator and sent by telefax to the central site. At 3 weeks after admission to the study, a separate case report form was completed to record the short-term outcome, either death or transplantation.
Study Population
This analysis included patients who were registered between February 3, 2000, and April 23, 2004, and selected on the basis of available serum and diagnosis. A total of 151 patients with ALF of indeterminate cause were enrolled during the study period. Samples from 64 (42.4%) patients from this subgroup representing 15 centers were available for analysis of APAP-CYS. The remaining 87 patients with indeterminate cause either were enrolled in the registry without serum sample collection or had a single, small aliquot obtained, which was reserved for future viral isolation studies. No differences in median age, admission laboratory values, or clinical outcomes were present between the patients in the indeterminate subgroup with and without samples available for APAP-CYS analysis. In addition, serum samples from 10 patients who were coded as having APAP toxicity and 30 children with other causes of ALF were selected as control subjects. The diagnosis of ALF for the 30 patients with other causes included metabolic liver disease (9), shock (5), other viruses (4), autoimmune hepatitis (3), other (8), and Wilson's disease (1).
Sample Collection
At the time of study enrollment, study-related serum samples were obtained concomitantly with other daily laboratory samples during a 7-day period. Samples were divided into 100- to 250-µL aliquots and stored at 80°C for later shipment to the central site. In all but 7 cases, study day 1 samples were used for APAP-CYS analysis. Because of sample availability in the remaining cases, study day 2 and 3 samples were used for 6 and 1 subjects, respectively.
Analytical Method
Samples were analyzed for APAP-CYS by a previously reported high-performance liquid chromatography with electrochemical detection (HPLC-EC) method,9 and the analytical laboratory was blinded to patient diagnoses. Briefly, samples were treated with protease, followed by sample dialysis. The resulting supernatant was injected on the HPLC column. Coefficients of variation for the assay were <10% at the known standards of 1, 6, and 30 µmol/L APAP-CYS. On the basis of the linearity of the assay, a positive assay was defined at an APAP-CYS concentration of
1 µmol/L APAP-CYS.10 Consistent with previous reports, final adduct concentrations were expressed relative to the protein content of the sample in units of nmol APAP-CYS/mg protein.14 The median (range) protein values for the study population were 17.3 mg/mL protein (10.038.0). The final range of positive values was 0.07 to 5.08 nmol APAP-CYS/mg protein.
Statistical Analysis
Analyses were conducted with SPSS V13.0 (SPSS, Inc, Chicago, IL). Mann-Whitney U tests were used to compare adduct-positive and adduct-negative cases from the indeterminate group. Fisher's exact tests were used to compare dichotomous variables in the indeterminate group.
| RESULTS |
|---|
|
|
|---|
Adducts were detected in 9 of the 10 children with known APAP-related ALF (median: 2.02 nmol APAP CYS/mg protein; range: 0.859.40; Fig 1). The serum sample from the APAP case that tested negative for adducts was obtained on the second day of follow-up, approximately 4 days after an acute ingestion of APAP. At that point, serum transaminases and coagulation were improving and adducts likely had cleared from the circulation.10
|
Adduct-Positive Patients From Indeterminate Cause Subgroup
Eight (12.5%) of the 64 samples from patients with ALF of indeterminate cause were positive for APAP-CYS (median: 1.98 nmol APAP-CYS/mg protein; range: 0.075.08). A history of APAP exposure was present in 5 of these 8 patients. The dosage of APAP received was either undocumented or within the therapeutic range for these 5 patients. Parental report indicated that 1 of these patients received APAP at 45 mg/kg per day for 8 days before presentation. APAP levels were measured in 6 of these patients and were detectable in 3 patients. APAP levels at the time of study entry were 3.1 mg/L, 7.0 mg/L, and 18.8 mg/L for these 3 patients. One additional patient had a positive urine screen for APAP with a negative serum level. Of these 8 adduct-positive patients, 4 presented with advanced encephalopathy (stage 3 or 4), 1 patient died, and none underwent liver transplantation. The single patient who died was a 3-year-old Asian boy who presented with stage 3 to 4 encephalopathy, an INR of 2.8, and respiratory failure. The patient had exposure to an unknown quantity of APAP for 3 days before presentation and had an APAP level of 18.8 mg/L
24 hours after the last exposure. Liver enzymes and coagulation seemed to be improving during the first 3 days of supportive care, but the patient died of multisystem organ failure on day 4 of follow-up. Postmortem examination revealed hemorrhagic, necrotizing pneumonia and hepatic injury with extensive centrilobular congestion and necrosis and macrovesicular steatosis. Viral culture of lung tissue was positive for adenovirus.
Liver biopsy or autopsy was performed in 4 of the 8 adduct-positive indeterminate cases. In all 4 cases, including the case described above, liver histology revealed centrilobular hepatocyte necrosis, a pattern of injury that classically is associated with APAP toxicity. Figure 2 represents a photomicrograph of liver tissue that was obtained from a 14-year-old white boy who had indeterminate ALF and whose serum was positive for APAP-CYS.
|
|
| DISCUSSION |
|---|
|
|
|---|
In the present study, adduct levels were measured in a population of children who presented to tertiary care medical centers with ALF. APAP-CYS were detected in a small but significant proportion of patients (12.5%) for whom the cause of ALF was indeterminate. These findings are important in that previous studies suggested APAP toxicity as the cause of ALF in
14% of children who present with this diagnosis.18 The results of our study suggest that APAP toxicity may be a more significant factor in pediatric ALF than previously recognized. In addition to known APAP overdose cases, our data suggest that there may be a previously unrecognized group of children with serious liver injury that is secondary to APAP toxicity.
The assay that was used in the present study has improved sensitivity than previously reported methods for the measurement of APAP-CYS.9,14,19 Using the HPLC-EC assay, adducts were detected in the vast majority (90%) of patients with ALF attributed to APAP ingestion. For patients in this subgroup, the diagnosis of APAP-related ALF had been made on the basis of historical data providing a clear history of ingestion of toxic doses of APAP. In the patient from this subgroup who was negative, the serum sample was obtained later in the progression of liver injury, when adducts likely had cleared from the circulation. In a recent, similar study in adults with APAP-mediated ALF,10 peak adduct values were obtained with the first study day sample, declined over time in parallel with resolution of the hepatic injury (serum ALT), and were present for up to 7 study days in some patients. In addition, APAP-CYS was not detected in the blood samples of patients who had acute APAP overdoses and received early treatment with NAC and had protection from toxicity.10 Additional study will be needed to delineate fully the duration of APAP-CYS in sera in pediatric APAP-related ALF. Nevertheless, the assay may provide an important diagnostic tool in cases of ALF with late presentation to medical centers, when APAP in plasma or serum may no longer be detectable and the history of APAP dosing is unclear.
A small fraction of the patients with an established cause of ALF also tested positive for adducts. In 1 of these 3 cases, APAP exposure at therapeutic doses was reported by history. The clinical presentation and outcomes of these patients were similar to the adduct-positive patients in the indeterminate cause subgroup. For example, aspartate aminotransferase values were higher than ALT levels and total bilirubin was <4 mg/dL at presentation in all 3 patients. All had liver injury that would have altered hepatic microcirculation, and 2 had ischemic liver injury that has been associated with depletion of hepatic glutathione content under experimental conditions.20 These alterations could have rendered the hepatocytes more sensitive to APAP injury. It is possible that APAP toxicity in these cases was a secondary event, occurring after the primary liver insult.
It is impossible to determine whether the patients who had indeterminate cases and tested positive for adducts experienced APAP toxicity as a primary or secondary event. Indeed, APAP toxicity has been implicated as a "second hit" in adults with ALF that is caused by viral infection with hepatitis A and B.21 Pediatric hepatologists have postulated that the majority of ALF of indeterminate cause in children is attributable either to seronegative viral infection or to undiagnosed metabolic diseases.22,23 If this postulation is correct, then APAP may potentiate injury in patients with these primary insults. It also is possible that a combination of minor physiologic alterations, such as dehydration and fasting, may significantly increase the risk for primary APAP toxicity in some children. Additional study is needed to elucidate the potential impact of these factors in the development of toxicity. In addition, the role of genetic variability in glutathione detoxification, glucuronidation, sulfation, and the cytochrome P450 enzymes that are involved in the metabolism of APAP needs to be examined rigorously in the context of APAP toxicity.
| CONCLUSION |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
Participating investigators and centers in the Pediatric Acute Liver Failure Study Group were as follows: Norberto Rodriguez-Baez, MD, University of Texas Southwestern Medical Center; Karen F. Murray, MD, University of Washington; Ross W. Shepherd, MD, Washington University; Phillip Rosenthal, MD, University of California at San Francisco; Benjamin L. Schneider, MD, and Sukru Emre, MD, Mount Sinai School of Medicine; Simon Horslen, MD, University of Nebraska, Robert H. Squires, Children's Hospital of Pittsburgh; Estella M. Alonso, MD, Northwestern University; Martin G. Martin, MD, University of California at Los Angeles; M. James Lopez, MD, PhD, University of Michigan; Brendan M. McGuire, MD, University of Alabama; Michael Narkewicz, MD, University of Colorado; Maureen Jonas, MD, Boston Children's Hospital; Kathleen Schwarz, MD, Johns Hopkins University; Steven Lobritto, MD, Columbia Presbyterian University; Daniel W. Thomas, MD, University of Southern California; Liz Rand, MD, University of Pennsylvania; Anil Dhawan, MD, King's College Hospital; Vicky Ng, MD, Hospital for Sick Children; Deirdre A. Kelly, MD, Birmingham Children's Hospital; Ruben E. Quiros, MD, Baylor College of Medicine; Joel E. Lavine, MD, University of California, San Diego; and Humberto Soriano, MD, Drexel University.
The expertise of Linda Letzig, BS (Department of Pediatrics, University of Arkansas for Medical Sciences) and Dr Hector Melin-Aldana (Department of Pathology, Children's Memorial Hospital) is acknowledged.
| FOOTNOTES |
|---|
Address correspondence to Laura James, MD, Arkansas Children's Hospital, Pediatric Clinical Pharmacology and Toxicology, 800 Marshall Street, Little Rock, AR 72202. E-mail: jameslaurap{at}uams.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||