Unconjugated Hyperbilirubinemia in Acetaminophen-Related Acute Liver Failure

Introduction

Acetaminophen overdose, either accidental or voluntary, is characterized by a progressive increase of liver function test results, including bilirubin. When the liver damage is at its maximum, the peak value for bilirubin is often reached between the third and fourth day. However, the presence of unconjugated bilirubin can be detected as early as 12 h after overdose [1]. On the other hand, persistence of unconjugated bilirubin for a longer period can be observed in some patients who recovered or died, as in the present case observation. Among the mechanisms explaining the increase in unconjugated bilirubin, damage to endoplasmic reticulum appears likely in relationship with the centrilobular necrosis caused by the toxic metabolism of acetaminophen.

Case Report

A 52-year-old man (87 kg weight) was admitted to the Intensive Care Unit (ICU) after having been transferred from a general hospital for fulminant liver failure. His past medical history included chronic analgesic abuse and alcohol use disorder. In addition, he was a chronic user of cocaine and cannabis. However, there was no evidence for chronic liver disease. At laboratory investigation performed a few months before the current hospital admission, no changes in liver function tests were noted, and there were no biological signs of hemolysis. Due to chronic back pain, the patient was also daily treated with clonazepam, pregabalin, oxycodone, and acetaminophen. During the last months, the daily consumption of acetaminophen reached frequently 6 g, or even more. The patient was addressed by his general practitioner to the psychiatric department for drug and ethanol withdrawal. As he was found stuporous at admission in the psychiatric ward, a drug overdose was suspected. The patient admitted a greater acetaminophen ingestion (6 g/ day) over the last 3 days. At physical examination, the patient was icteric and flapping tremor was noted. Laboratory investigations revealed a glucose level of 27 mg/dL (reference range, 70–100), aspartate aminotransferase >7000 IU/L (10–40), ala-nine aminotransferase of 5600 IU/L (10–40), international normalized ratio >7 (0.80–1.20), ammonia of 197 µmol/L (18–72), creatinine of 2.8 mg/dL (0.6–1.30), total bilirubin of 7.94 mg/dL (<1.2), and conjugated bilirubin of 5.8 mg/dL (<0.3). There was no evidence (either on admission or during follow-up) of hemolysis, according to the laboratory test results for hemoglobin, reticulocytes, schistocytes, lactate dehydrogenase, and haptoglobin (Figure 1). The enzymatic activity of glucose-6-phosphate-dehydrogenase (measured at the peak level of bilirubin) was 7 IU/g Hb (5–8). Serum acetaminophen level on ICU admission (at least 24 h after the last exposure) was 24 µg/mL. Extensive toxicological investigations (blood and urine) did not reveal the presence of ethanol, drugs, or illicit substances. There was no sign of steatosis on liver ultrasound and computed tomography (CT) examination. The patient received the classical N-acetylcysteine treatment regimen for acetaminophen overdose. The clinical course was characterized by a progressive worsening of the neurological condition, evolving to grade IV encephalopathy. Coagulation disorders persisted, with factor V level <10%. The evolution of laboratory values is shown in Figure 1. During the clinical course, there was a marked shift from conjugated to unconjugated bilirubin. After a careful psychiatric and social evaluation, it was concluded that liver transplantation was not appropriate for this patient. He died on hospital day 8 from the multiple complications of fulminant liver failure. Genetic study of the promotor region of the uridine di-phosphate-glucuronosyltransferase enzyme gene (UGT1A1) showed that the patient was UGT1A1 *1/*1, excluding the diagnosis of Gilbert syndrome. On a previous laboratory analysis, before the current episode of acetaminophen overdose, the serum total bilirubin level was 0.28 mg/dL, with a conjugated bilirubin level of 0.1 mg/dL.

Discussion

Several mechanisms could explain the increase of unconjugated bilirubin after acetaminophen overdose. Hemolysis is not a classical complication of acetaminophen overdose except in some patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency [3]. It was also advocated that acetaminophen might impair the uptake of bilirubin from the blood by the liver, either by competing with bilirubin for the binding sites on the liver cell membrane or in the cytoplasm or by damaging the uptake mechanism [4]. More probably, the increase of unconjugated bilirubin level during the course of acetaminophen poisoning is linked to an acquired deficit of bilirubin glucuronidation. The most affected liver area after acetaminophen poisoning is the centrilobular zone (zone III), which possess the greater distribution and content of CYP450; however, with extremely toxic doses, necrosis can also extend to zones I and II [5]. When the less-efficient renal excretion becomes the main route of elimination for the less water-soluble unconjugated bilirubin, unconjugated bilirubin can also accumulate in the blood [6].

The predisposition of some populations to acetaminophen toxicity is debatable [7]. Gilbert syndrome is a benign inherited disorder of bilirubin glucuronidation that affects between 5% and 10% of Western Europeans. Unconjugated bilirubin is then caused by a deficit in bilirubin-uridine diphosphate-glucuronyltransferase (bilirubin-UGT), the enzyme that conjugates bilirubin to glucuronic acid, converting bilirubin into a soluble form excretable in bile [8]. The diagnosis is often made incidentally on routine liver biochemistry testing, but definitive diagnosis is obtained by genetic testing with a mutation in the promotor region UGT1A1. Serum bilirubin levels in Gilbert syndrome are usually less than 3 mg/dL, but can increase to higher levels and reach 6 mg/dL under certain stress or physiologic conditions. Acetaminophen requires bilirubin-UGT mediated glucuronidation and about 55% is excreted as glucuronide. There is no direct evidence that patients with Gilbert syndrome are more susceptible to liver damage with therapeutic or supra-therapeutic doses of acetaminophen [9–12]. In our case, the hypothesis of an associated Gilbert syndrome could be definitely excluded by the results of genotyping.

Older observations, before the era of routine practice of liver transplantation, showed a rapid increase of unconjugated bilirubin in the blood of paracetamol-poisoned patients. This was first noted in a series of 38 patients out of a group of 60 patients who had taken a acetaminophen overdose [1]. The rise in unconjugated bilirubin started as early as 12 h after overdose. Studies performed on rats given an acetaminophen overdose demonstrated a striking reduction in the activity of bilirubin glucuronyl-transferase and the liver content of cyto-chrome P450 [1]. However, the direct implication on acetaminophen glucuronidation in humans is not known; all the more, acetaminophen is likely to be substrate of a UGT isoform other than the UGT1A1 [13].

The peak serum level of total bilirubin is usually not considered as a predictive factor in acetaminophen poisoning, even if high serum bilirubin levels are often associated with a poorer outcome [2]. In a previous paper from our institution, the investigation of unconjugated bilirubinemia in an acetaminophen-poisoned population found that admission value was not able to discriminate between patient outcome (mortality, transplantation, or composite endpoint of mortality and transplantation) [2]. However, maximal value within the first week was independently related to transplantation (4.1 mg/dL ±2.8 vs 2.3 mg/ dL ±1.9; P=0.0042), and to a composite endpoint of transplantation or death (3.5 mg/dL ±2.6 vs 2.3 mg/dL ±1.9; P=0.0095).

Bilirubin, and particularly unconjugated bilirubin, has been found to be a biomarker for short-term mortality in acute-on-chronic liver failure [14]. The increase in unconjugated bilirubin is not specific for acetaminophen-induced liver injury but has not been investigated in details in other etiologies of fulminant liver failure. In the late course of acute acetaminophen poisoning, when liver transplantation is not indicated or feasible, this elevation can reflect the severity of functional injury, despite the apparent improvement of other laboratory values (aspartate aminotransferase, international normalized ratio, creatinine).

Conclusions

A high unconjugated/conjugated bilirubin ratio several days after acetaminophen poisoning may reflect a persisting defect in bilirubin conjugation in relationship with liver centrilobular injury, but the influence on paracetamol-glucuronidation is not documented and it could not be considered as a prognostic factor.