• DOI:10.7326/0003-4819-137-12-200212170-00007
• Corpus ID: 11390513

Results of a Prospective Study of Acute Liver Failure at 17 Tertiary Care Centers in the United States

@article{Ostapowicz2002ResultsOA,
  title={Results of a Prospective Study of Acute Liver Failure at 17 Tertiary Care Centers in the United States},
  author={George Ostapowicz and Robert John Fontana and Frank Schidt and Anne M. Larson and Timothy J. Davern and Steven‐Huy B. Han and Timothy Mccashland and A. Obaid Shakil and J. Eileen Hay and Linda S. Hynan and Jeffrey S. Crippin and Andres T. Blei and Grace Samuel and Joan S. Reisch and William M. Lee},
  journal={Annals of Internal Medicine},
  year={2002},
  volume={137},
  pages={947-954}
}

• G. Ostapowicz, R. Fontana, +12 authors William M. Lee
• Published 17 December 2002
• Medicine
• Annals of Internal Medicine

Context Acute liver failure is a catastrophic condition affecting about 2000 persons in the United States each year. Contribution Among 308 consecutive patients with liver failure admitted to 1 of 17 referral centers between 1998 and 2001, acetaminophen overdose (39%) and idiosyncratic drug reactions (13%) were the most common cause of disease. Sixty-seven percent of patients were still living 3 weeks after presentation (44% without liver transplantation and 23% after transplantation). Survival was associated with cause of liver failure (persons with acetaminophen overdose fared best) and depth of coma at presentation. Implications Drug toxicities cause most cases of acute liver failure in the United States. Cause and depth of coma at presentation predict patient outcomes. The Editors Acute liver failure is characterized by severe and sudden liver cell dysfunction leading to coagulopathy and hepatic encephalopathy in previously healthy persons with no known underlying liver disease (1). Sometimes termed fulminant hepatic failure, acute liver failure is thought to affect approximately 2000 persons per year in the United States (2). This catastrophic illness can rapidly progress to coma and death from cerebral edema and multiorgan dysfunction (2, 3). Over the past 30 years, the most frequently identified cause of acute liver failure has been viral hepatitis, especially hepatitis B but also hepatitis A and, in tropical populations, hepatitis E (4-8). Smaller numbers of cases have been related to drug-induced liver injury and other causes (9). The frequency of identified causes of acute liver failure has varied greatly worldwide. A very high prevalence of suicidal overdoses of acetaminophen (73%) has been recorded in the United Kingdom (10), and no cases of acetaminophen-related disease have been observed in developing countries (7). Until recently, only limited data have been available on the cause and outcome of acute liver failure in the United States, presumably because of the relative rarity of the condition, referral patterns in North America, and lack of a centralized data registry (11-14). The U.S. Acute Liver Failure Study Group was formed in 1997 as a consortium of liver centers interested in better defining causes and outcomes of acute liver failure. The primary aim of the current study was to prospectively describe, with emphasis on the role of transplantation, the causes of acute liver failure and short-term outcomes of patients with this condition who presented to participating liver centers between 1998 and 2001. Our secondary aim was to compare presenting clinical features and liver transplantation in patients with acute liver failure related to acetaminophen hepatotoxicity, other drugs, indeterminate factors, and other causes. Methods Between January 1998 and May 2001, a 41-month period, 308 patients 15 years of age or older were enrolled from 17 tertiary care liver centers around the United States, all but one of which has a liver transplantation program. The local institutional review board of each center approved the study. All enrolled patients met entry criteria for acute liver failure: the presence of coagulopathy (prothrombin time > 15 seconds or international normalized ratio 1.5) and any hepatic encephalopathy within 26 weeks of the first symptoms without previous underlying liver disease (15). Informed consent was obtained from patients' next of kin before enrollment, according to the guidelines of local institutional review boards. On the basis of records of patients considered for enrollment but excluded because of lack of informed consent or other reasons, more than 65% of eligible patients were enrolled. Under code to protect confidentiality, each center provided detailed demographic, clinical, laboratory, and outcome information for all enrolled patients. Hepatic coma was graded on a standard scale of I to IV, as described elsewhere (1, 2). Etiologic diagnoses, which were made at each study center, were based on accepted diagnostic criteria circulated to all investigators. These criteria involved history; laboratory values; imaging studies; and, in some cases, histopathologic characteristics. Acute liver failure was considered to be of indeterminate cause when extensive clinical, radiographic, and laboratory evaluation (including toxicology screens and serologic markers for viral hepatitis A, B, and C and antinuclear and antismooth-muscle antibodies) was inconclusive. Investigators tested and searched for other viruses by using RNA methods as clinically indicated, but these results were not uniformly available. Clinical guidelines for patient management, although determined at each site, were uniform (16). Candidacy for liver transplantation was determined at individual centers according to the guidelines of the United Network of Organ Sharing. Shortly after patients were admitted to the study, an initial case report form was completed and forwarded to the central study site (University of Texas Southwestern Medical Center, Dallas, Texas). In addition, an outcome case report form detailing the hospital course and final laboratory results not available at admission was completed within 3 weeks of discharge, death, or transplantation. Outcome end points, which were determined 3 weeks after study admission, included liver transplantation, death, and survival without transplantation. Case report forms were checked for missing values and inconsistencies; queries were sent to the participating centers, and corrections were made at the data coordinating center. All data were then entered into an Access database (Microsoft Corp., Redmond, Washington). Annual site visits by researchers from the data coordinating center confirmed the accuracy of the data. Statistical Analysis Results are expressed as medians and ranges. Statistical significance was not used in the analysis because the data were descriptive. Role of the Funding Sources The funding sources had no specific role in the design, conduct, or reporting of the study. Results Demographic Characteristics and Clinical Data Of the 308 patients with acute liver failure, 224 (73%) were women. The median age of the group was 38 years (range, 15 to 78 years); women were older than men (39 years vs. 32.5 years). Two hundred twenty-seven patients (74%) were white, 31 (10%) were Hispanic, 29 (9%) were African American, 10 (3%) were Asian, 5 (2%) were Native American, and 6 (2%) were of other ethnic backgrounds. Extensive variation was seen in the median duration of symptoms (6 days [range, 0 to 74 days]) and jaundice (2 days [range, 0 to 61 days]) before encephalopathy onset. Coma grade at study admission was relatively equally distributed (Table 1). Forty percent of patients had a serum creatinine concentration of 177 mol/L or greater ( 2.0 mg/dL), and 14% had acidosis (arterial pH < 7.30). Nearly half of patients (44%) acquired a culture-proven infection. Most patients (260 [84%]) were referred from other hospitals, and the remainder were admitted directly to the tertiary care centers. Table 1. Characteristics of Patients at Admission according to Presumed Cause of Acute Liver Failure Causes of Acute Liver Failure The Figure displays the presumed cause of acute liver failure and the outcome data for all study patients. Acetaminophen overdose was the most commonly implicated reason for acute liver failure, accounting for 120 patients (39%). The median dosage of acetaminophen ingested was 13.2 g/d (range, 2.6 to 75 g/d), and 99 of 120 patients (83%) had ingested more than 4 g/d (the maximum package recommendation). Among these 120 patients, 44 (37%) were considered to have ingested acetaminophen with suicidal intent (single ingestion with suicide attempt admitted) and 68 (57%) had accidental toxicity (many ingestions for pain relief without suicidal intent). In 8 patients (7%), the reason for overdose could not be determined. Figure. Presumed causes and outcomes in 308 patients with acute liver failure. Idiosyncratic drug reactions were deemed responsible for another 40 cases of acute liver failure (13%). Hepatitis A and B infection led to 14 cases (4%) and 22 cases (7%), respectively. Presumptive diagnoses of ischemic hepatitis (shock liver), autoimmune hepatitis, Wilson disease, and the BuddChiari syndrome accounted for 17 (6%), 13 (4%), 8 (3%), and 5 (2%) cases, respectively. Six patients with pregnancy-associated acute liver failure were listed as having acute fatty liver (n = 2), the HELLP (Hemolysis, Elevated Liver enzymes, Low Platelets) syndrome (n = 3), or eclampsia (n = 1). Acute liver failure was attributed to cancer (metastatic carcinoma or lymphoma) in 4 patients and to another cause (heat stroke, sepsis, or giant-cell hepatitis) in 6 patients. Fifty-three cases (17%) of indeterminate cause were equally divided among the 12 centers (median cases per center, 4.5 [range, 1 to 9]). Comparison of Clinical and Laboratory Features by Presumed Cause Table 1 shows between-group differences in clinical and laboratory data at admission, based on presumed cause of acute liver failure. All groups had a similar preponderance of women. Patients with acetaminophen-related liver failure and those with indeterminate liver failure were younger than patients with liver failure that was drug induced or related to other causes. White persons were overrepresented in the acetaminophen group compared with the other three groups. Patients with acetaminophen-related disease had shorter disease duration; higher serum alanine aminotransferase, aspartate aminotransferase, and creatinine levels; and lower bilirubin levels and arterial pH. The drug-induced and indeterminate groups had longer disease duration, lower alanine aminotransferase levels, and higher bilirubin levels. The groups did not differ significantly in other measures of disease severity at presentation (for example, coma grade, mean arterial pressure, prothrombin time, or international normalized