Acute Ethanol Coingestion Confers a Lower Risk of Hepatotoxicity after Deliberate Acetaminophen Overdose

Authors


Address for correspondence and reprints: William S. Waring, PhD; e-mail: s.waring@ed.ac.uk.

Abstract

Objectives:  Little is known about the clinical significance of acute ethanol coingestion around the time of acetaminophen (paracetamol) overdose. This study prospectively examined the effect of acute ethanol coingestion on risk of hepatotoxicity among patients admitted to hospital for N-acetylcysteine (NAC) therapy after deliberate acetaminophen overdose.

Methods:  This was a prospective observational study and included sequential patients who presented within 24 hours of acute acetaminophen ingestion and required NAC therapy. Significant hepatotoxicity was defined by alanine transaminase > 1,000 U/L or the international normalized ratio > 1.3 after a standardized intravenous administration of 300 mg/kg NAC.

Results:  There were 362 patients, including 178 (49.2%) who coingested ethanol acutely. The prevalence of hepatotoxicity was 5.1% (95% CI = 2.6% to 9.5%) in those who ingested ethanol, compared to 15.2% (95% CI = 10.7% to 21.2%) in those who did not (p = 0.0027 by chi-square proportional test). Acute ethanol intake conferred a lower risk of hepatotoxicity in patients who had acetaminophen concentrations above or below the “200-line” and was independent of the interval between ingestion and assessment.

Conclusions:  Acute ethanol intake is associated with a lower risk of hepatotoxicity after acetaminophen overdose. This apparent protective effect cannot be explained solely by lower exposure to acetaminophen in this group, nor differences in the interval between ingestion and initiation of treatment. Further work is required to establish mechanisms by which ethanol might confer protection against hepatotoxicity, so as to identify novel strategies for reducing risk after acute acetaminophen ingestion.

N-Acetylcysteine (NAC) is capable of preventing hepatotoxicity if given sufficiently early after acute acetaminophen ingestion.1 It is normally reserved for patients who have serum acetaminophen concentrations higher than the Rumack-Matthew nomogram, the so-called “200-line,” which is extrapolated exponentially from 200 mg/L at 4 hours postingestion, assuming a half-life of four hours.2 A lower threshold for treatment is advocated for individuals thought to be at increased risk of toxicity, for example, those who regularly consume large quantities of ethanol, malnourished patients, and patients receiving medications capable of inducing hepatic enzymes. For example, a “150-line” has been adopted in the United States for high-risk patients, and a “100-line” is used in Australia, the United Kingdom, and other parts of Europe.3,4

Patients who present to hospital after deliberate drug ingestion have often coingested ethanol acutely. Despite this, the effects of acute ethanol consumption on the risk of acetaminophen-induced hepatotoxicity are not clearly defined. Basic research suggests that acute ethanol consumption at the time of acetaminophen ingestion is capable of protecting against hepatotoxicity.5,6 Acute ethanol intake suppresses cytochrome P450 activity and alters hepatocellular redox state, and these mechanisms have been proposed to protect against hepatotoxicity by minimizing formation of toxic intermediate metabolites, for example N-acetyl-p-benzoquinoneimine.6,7 There are limited clinical data to support these basic research findings. A lower proportion of patients who coingest ethanol at the time of acute acetaminophen ingestion appear to develop subsequent hepatotoxicity.8,9 However, ingestion of lesser acetaminophen quantities and earlier presentation to hospital after ingestion are important confounding factors. A multivariate analysis that considered the stated acetaminophen dose and interval between ingestion and hospital attendance found that acute ethanol intake had no independent effect on the risk of acetaminophen-induced hepatotoxicity.10 The potential effects of acute ethanol consumption are highly relevant to risk assessment and initial triage in this important patient group.

This study prospectively examined whether acute ethanol consumption might influence the risk of hepatotoxicity in patients who presented to hospital within 24 hours of acute acetaminophen ingestion and required NAC treatment. The aim was to determine whether acute ethanol intake around the time of acetaminophen overdose might be capable of influencing the risk of hepatotoxicity in a high-risk population.

Methods

Study Design

This was a prospective, observational study of patients who attended the emergency department (ED) within 24 hours of acute acetaminophen ingestion. The primary outcome variable was hepatotoxicity, defined by either alanine transaminase > 1,000 U/L or international normalized ratio (INR) > 1.3 after NAC administration in accordance with national guidelines for management of acetaminophen poisoning. The study protocol was reviewed and approved by the Lothian research ethics committee.

Study Setting and Population

The study was conducted in the ED and Toxicology Unit of the Royal Infirmary of Edinburgh, a teaching hospital that admits approximately 3,000 poisoned patients annually. A standard operating procedure is used for patients who have ingested acetaminophen, so clinical care is delivered in a consistent manner and in accordance with TOXBASE (Edinburg, Scotland), the national resource for poisoning management advice in the United Kingdom. In brief, intravenous NAC therapy is indicated if serum acetaminophen concentration is above the “200-line” nomogram or above the “100-line,” and the patient is considered at high risk due to chronic ethanol excess, use of enzyme-inducing drugs (carbamazepine, phenobarbital, phenytoin, rifampicin, St. John’s wort), chronic liver disease, or malnutrition. Chronic ethanol excess is defined by regular consumption of more than 21 units (168 g) per week in men or 14 units (112 g) in women.11

Intravenous 300 mg/kg NAC is administered over 20.25 hours, based on the patient’s weight up to 110 kg. At the end of the infusion, if serum creatinine and liver biochemistry are normal, and INR is ≤ 1.3, then the patient may be considered safe for discharge.12

Study Protocol

The study was conducted between March 1, 2005, and June 30, 2006. All patients who presented to the ED after suspected acetaminophen ingestion were identified to the study investigators and were managed according a standardized clinical protocol. Inclusion criteria were men or women, aged 18 to 80 years, in whom NAC therapy was indicated on the basis of standard clinical criteria. Exclusion criteria were presentation to hospital more than 24 hours after ingestion or staggered acetaminophen ingestion (>2 hours). These inclusion and exclusion criteria were selected to identify a high-risk population. Consecutive patients were included to minimize the risk of selection bias.

Measurements

A standardized data collection sheet was used to record patient age, gender, stated date and time of overdose, quantity of acetaminophen ingested, coingested drugs and alcohol, the presence of risk factors for hepatotoxicity, serum acetaminophen concentration, interval between ingestion and acetaminophen determination, and whether the measured acetaminophen concentration was higher or lower than the “200-line.” Acute ethanol consumption was defined by self-reported ethanol consumption within 4 hours of acetaminophen ingestion. A breath alcometer (SD400 Lion alcolmeter, Lion Laboratories Limited, South Glamorgan, UK) was used to determine the presence or absence of ethanol in all patients who presented after acute acetaminophen overdose; data were available for 357 patients (98.6% of the study population) and corresponded with the patient history in all cases. Blood or urine ethanol concentrations were not recorded.

Serum alanine transaminase activity and INR were measured by the local hospital laboratory, which is an accredited reference facility. In brief, venous blood samples were collected into separate Monovette tubes containing serum gel and citrate (Sarstedt Limited, Leicester, UK). Serum alanine transaminase activity was determined by an automated ultraviolet kinetic method (Olympus, Southall, UK), and INR was determined using Innovin thromboplastin (Dade Behring, Marburg, Germany) by CA6000 automated analyzer (Sysmex, Milton Keynes, UK).

Data Analysis

Data are presented as median and interquartile ranges (IQRs). Proportions are expressed as ratios and 95% confidence intervals (CIs) constructed using the modified Wald method.13 Convenience dosage groups used were 0–4, 4.1–8, 8.1–12, 12.1–16, 16.1–24, and >24 g, which were based on the maximum normal daily dose of 4 g. Between-group comparisons were made by two-tailed Yate’s corrected chi-square proportional tests (MedCalc statistical software v.9.1.0.1, Mariakerke, Belgium), and p-values <0.05 were accepted as statistically significant in all cases.

Acknowledgments

The authors thank Janice Pettie and Margaret Dow of the Scottish Poisons Information Bureau for assistance with identification and retrieval of clinical data and Lisa Galloway of the Royal Infirmary of Edinburgh for assistance with collection of laboratory data.

Results

During the study period, 1,045 patients presented to the Royal Infirmary of Edinburgh within 24 hours of acute acetaminophen overdose. Of these, 683 did not require NAC and, therefore, the study population consisted of 362 patients (240 women) with median (IQR) age 35 years (IQR = 22 to 44 years). The stated quantity of acetaminophen ingested was 17 g (IQR = 13 to 25 g), and the interval between ingestion and serum acetaminophen measurement was 4.9 hours (IQR = 4.2 to 8.3 hours). Acute ethanol coingestion was documented in 178 patients (49.2%). There were 194 patients (53.6%) considered to be at high risk due to chronic alcohol excess (151), malnutrition (33), prior use of enzyme-inducing drugs (8), recent acetaminophen ingestion (1), and the presence of established liver disease (1).

Hepatotoxicity occurred in 10.2% (95% CI = 7.5% to 13.8%) of the study population. The proportion who developed hepatotoxicity was 5.1% (95% CI = 2.6% to 9.5%) in those who consumed ethanol acutely compared to 15.2% (95% CI = 9.8% to 20.0%) in those who did not (p = 0.0027). Acute coingestion of ethanol was associated with a lower proportion of patients who developed hepatotoxicity in the subgroup that had acetaminophen concentrations above the “200-line” (8.0% vs. 23.2%, p = 0.018) and tended to be associated with lower risk at different intervals between ingestion and acetaminophen determination (Figure 1). In the subgroup of patients that regularly consumed excess ethanol, the risk of hepatotoxicity was 3.0% (95% CI = 0.7% to 8.9%) in those with acute ethanol coingestion, and 13.5% (95% CI = 6.4% to 25.6%) in those without (p = 0.0338; Table 1). The proportion of “high-risk” patients that had consumed ethanol acutely was 56.6%, compared to 40.7% in those without additional risk factors (p = 0.0027). None of the patients in this series developed fulminant hepatic failure.

Figure 1.

 Patients treated with NAC after acute acetaminophen ingestion characterized by acute ethanol coingestion, serum acetaminophen concentration, and interval between ingestion and NAC (0–8, 8–16, and 16–24 hours). The Rumack-Matthew nomogram (“200-line”) is shown in bold, and the numbers of patients who developed hepatotoxicity in each subgroup are shown. *p = 0.018 versus no ethanol coingestion by chi-square proportional test after Yate’s correction.

Table 1.   Prevalence of Hepatotoxicity after Acute Acetaminophen Ingestion in Patients Treated with NAC in Subgroups Defined by Coingestion of Ethanol and the Presence or Absence of Conventional Risk Factors, Namely, Chronic Ethanol Consumption and Others (Malnutrition, Enzyme-inducing Drugs, Existing Liver Disease)
Ethanol CoingestionNo Risk FactorsChronic Ethanol ExcessOther Risk FactorsWhole Group
  1. ALT = alanine transaminase.

  2. * p < 0.05 versus no ethanol coingestion by chi-square proportional tests with Yate’s correction.

  3. † p < 0.005 versus no ethanol coingestion by chi-square proportional tests with Yate’s correction.

None
 Number995233184
 ALT > 1000 U/L 4 (4.0%)2 (3.8%)0 (0.0%)6 (3.3%)
 INR > 1.316 (16.2%)7 (13.5%)4 (12.1%)27 (14.7%)
 Either17 (17.2%)7 (13.5%)4 (12.1%)28 (15.2%)
Acute
 Number689911178
 ALT > 1000 U/L3 (5.6%)0 (0.0%)0 (0.0%)3 (1.7%)
 INR > 1.35 (7.4%)3 (3.0%)*1 (9.1%)9 (5.1%)†
 Either5 (7.4%)3 (3.0%)*1 (9.1%)9 (5.1%)†

Discussion

In a high-risk group of patients requiring NAC therapy after deliberate acetaminophen ingestion, the proportion of patients who developed hepatotoxicity was lower in those who coingested ethanol. This is in keeping with retrospective clinical data that show acute ethanol consumption around the time of acetaminophen overdose confers a lower risk of subsequent hepatotoxicity.14 Acute ethanol consumption was also found to reduce risk in patients who had acetaminophen concentrations higher than the “200-line.” The apparent protective effect of acute ethanol consumption cannot be explained solely on the basis of lower acetaminophen exposure. A pharmacokinetic interaction is unlikely, given that acute ethanol intake does not influence acetaminophen absorption after a single oral administration.15

The ability of acute ethanol consumption to lower the risk of hepatotoxicity might be explained by a number of possible mechanisms. For example, acute ethanol consumption is capable of inhibiting tissue cytochrome 2E1 activity, thereby lessening the extent of acetaminophen metabolism to potentially toxic metabolites.16 Acute inhibition of cytochrome 2E1 activity might, at least in part, explain the present findings and offers a plausible explanation for why the effects were so pronounced in patients with chronic ethanol excess, who would have been expected to have increased cytochrome 2E1 activity.17 An alternative hypothesis is that ethanol is capable of directly enhancing NAD(P)H:quinone reductase activity, thereby reconverting quinone metabolites back to native acetaminophen and limiting the potential for accumulation of toxic metabolite concentrations.18

The proportion of patients that developed hepatotoxicity was no higher amongst those who regularly consumed excess ethanol, in contrast to other reports.9,10,19 Preclinical research shows that chronic ethanol consumption increases acetaminophen-induced hepatic injury via glutathione depletion and induction of cytochrome P450 2E1 activity.16,20–22 However, there is some uncertainty about the significance of chronic ethanol consumption as an independent risk factor, because the association is confounded by delayed presentation to hospital and initiation of NAC.23–26 Exclusion of patients who presented more than 24 hours after ingestion does not allow us to address whether chronic ethanol excess might influence the development of hepatotoxicity in patients who present late. Induction of cytochrome 2E1 requires chronic exposure to ethanol concentrations >250 mg/dL, which might be encountered only in patients with extremely heavy drinking patterns.16 Therefore, the conservative definition of chronic ethanol excess in this study is likely to have given rise to a heterogenous group, and only a proportion of these patients would be expected to have significantly induced cytochrome 2El activity. These findings indicate that the risk of hepatotoxicity was greatest in patients who presented to hospital more than 8 hours after ingestion and underscore the importance of prompt initiation of NAC therapy in this situation, even before acetaminophen concentrations are known.

Limitations

Our protocol adopted a qualitative approach to acute ethanol ingestion that relied on patient self-reporting and the breath alcometer reading being positive or negative. The quantity of ethanol consumed was not estimated and, therefore, a dose–response relationship or threshold effect cannot be established. A further potential limitation is that INR > 1.3 was used to define hepatotoxicity, in accordance with current UK guidelines for management of acetaminophen poisoning. This is a more conservative criterion than used to define acute liver failure in other settings and, therefore, these findings might not be directly applicable to other patient groups. The demographic characteristics of the study population were typical of poisoned patients in our region, namely young adults with a prevalence of chronic ethanol excess of around 40%. These data might not be applicable to populations with substantially different patterns of ethanol intake.

Conclusions

Ethanol consumption around the time of acute acetaminophen overdose confers a lower risk of hepatotoxicity, which is relevant to initial patient triage. This effect cannot be explained by lesser acetaminophen exposure, suggesting a direct pharmacologic effect. High ethanol concentrations might indicate the appropriateness of a higher treatment threshold after acute acetaminophen overdose; however, further work is required to examine the validity of this approach. Further attention is required to explore the underlying mechanisms so that pathways involved in acetaminophen-induced hepatotoxicity can be better understood.

Ancillary