Abstract

Background: Paracetamol is a major cause of poisoning. Treatment decisions are predominately based on the dose ingested and a timed blood paracetamol concentration because most patients present to hospital soon after overdose, before hepatotoxicity can be confirmed/excluded using serum alanine transaminase (ALT). Nonetheless, ALT is measured at hospital presentation; we investigated its value in predicting hepatotoxicity.

Methods: From March 2011 to May 2012, patients admitted to the Royal Infirmary of Edinburgh for paracetamol overdose treatment were identified. We determined the value of admission ALT (below or above our upper limit of normal—50 IU/l) at predicting three endpoints: 1—doubling of ALT; 2—peak ALT >1000 IU/l; 3—peak international normalized ratio (INR) >2.

Results: From 500 patients, 410 met the entry criteria; 264 presented within 8 h of overdose, 54 between 8 and 24 h, 53 after 24 h and 39 were staggered ingestions. Admission ALT was increased in 71. For endpoint 1 (ALT doubling), the positive predictive value (PPV) of admission ALT was 19% [95% confidence interval (CI) 12–30] with a negative predictive value (NPV) of 98% (95% CI 96–99); endpoint 2 (ALT >1000 IU/l: PPV 23% (95% CI 14–34) and NPV 100% (95% CI 99–100) and for endpoint 3 (INR >2): PPV 14% (95% CI 7–25) and NPV of 100% (95% CI 99–100). The NPV remained high when only late presenters were included.

Conclusion: Admission ALT within the normal range has a high NPV and could be used, alone or in combination with newer biomarkers, to identify lower risk patients at hospital presentation.

Introduction

Paracetamol (acetaminophen) is the most commonly overdosed drug in UK and a leading cause of acute liver failure in the Western world.1 It accounts for >40% of poisoning episodes that are admitted to hospital: 38 000 emergency hospital admissions in the financial year 2010–11 in England alone.2 In UK, guidance on the management of paracetamol overdose is available through resources such as TOXBASE (www.toxbase.org) and the British National Formulary, with need for treatment being largely determined by the dose ingested and timed plasma paracetamol concentration.3

With the mechanism of paracetamol hepatotoxicity extensively investigated, the current mainstay of treatment available for paracetamol poisoning is replenishment of cellular glutathione through the administration of acetylcysteine (AC). Although early treatment with AC effectively prevents hepatotoxicity, its efficacy declines if treatment is delayed >8 h after overdose; patients presenting late to hospital are at an increased risk of significant liver injury.4

It is essential to triage patients presenting with paracetamol overdose into those in whom hepatic damage is likely and those in whom it is unlikely. Such hospital ‘front door’ stratification would allow new therapeutic approaches to be targeted at lower or higher risk groups and increase the likelihood of clinical trials being successful. At present, a timed blood paracetamol concentration is measured to stratify patients into two groups; those needing AC treatment and those who do not require antidote. However, a significant proportion of patients are misclassified as needing treatment as they would not get liver injury if untreated.5 Furthermore, within the treated group further stratification could lead to the AC dose being individualized depending on risk of liver injury. Although it is rare for a patient to develop toxicity if their blood paracetamol concentration is below the ‘at-risk’ line on the treatment nomogram, such cases do occur and have resulted in potentially avoidable deaths and, therefore, great concern for both clinicians and medicine regulators.6 Alanine transaminase (ALT) has the potential to aid treatment decisions at the hospital ‘front door’. In this study, we tested the performance of serum ALT activity at hospital admission as a predictor of subsequent injury in a large cohort of patients who received AC treatment.

Methods

The Edinburgh Clinical Toxicology Unit has a database of all patients admitted to the toxicology ward at the Royal Infirmary of Edinburgh (RIE), UK, with excess paracetamol ingestion that required treatment with intravenous AC. This does not include patients directly admitted to the Scottish Liver Transplantation Unit, also based at RIE. This database was the source of patient identification. Exclusion criteria were as follows: patients who did not complete AC treatment; patients whose bloods were not repeated at end of AC, haemolysed or had no INR taken. In this study, we have included patients from March 2011 to May 2012.

The performance of hospital admission serum ALT activity (greater than the upper limit of our hospital’s normal range, 50 IU/l) was tested as a predictor of three endpoints. Endpoint 1 was doubling of ALT from admission to the end of first 20.25 h course of AC therapy. Endpoint 2 was a peak hospital stay ALT >1000 IU/l. Endpoint 3 was a peak international normalized ratio (INR) value >2. Patient blood results were extracted from RIE blood results system, which stores all patient data from NHS Lothian on an electronic database. Care was taken to maintain patients’ confidentiality through removal of all patient-identifiable data and secure storage within the National Health Service.

Positive and negative predictive values (PPV and NPV), specificities and sensitivities with 95% confidence intervals (CIs) were calculated for hospital admission ALT. Furthermore, results were divided according to time from ingestion to first blood sample—up to 8 h, 8–24 h, >24 h and staggered overdose (ingested over 2 h or more). The statistical package used to calculate these values was an online resource provided by the Canadian Institute for Health Research (www.ktclearinghouse.ca).

Results

A total of 500 patients were admitted for AC treatment in the period March 2011–May 2012. In total, 90 were excluded from the study because they fulfilled one or more of our exclusion criteria (73 patients had no end of AC treatment blood results, eight patients had haemolysed blood results and remaining nine patients had no INR measurement). Table 1 provides a summary of the patient demographics. About 264 presented to hospital within 8 h of the overdose, 54 presented within 8–24 h, 53 presented 24 h after overdose and 39 were staggered overdoses (ingested over 2 h or more) (Table 2). On admission to hospital, serum ALT activity was increased above the normal limit (>50 IU/l) in 71 patients. The breakdown of this by time interval from ingestion is presented in Table 2.

Table 1

Demographics of our cohort

 Total 
Number 410 
Sex (M:F) 151:259 
Median age (range) 37 (13–90) 
Number admission ALT < ULN 339 
Number admission ALT > ULN 71 
Number admission ALT >100 34 
Number peak ALT >1000 16 
Number normal INR 402 
Number peak INR >2 
 Total 
Number 410 
Sex (M:F) 151:259 
Median age (range) 37 (13–90) 
Number admission ALT < ULN 339 
Number admission ALT > ULN 71 
Number admission ALT >100 34 
Number peak ALT >1000 16 
Number normal INR 402 
Number peak INR >2 

ULN, upper limit of normal.

Table 2

Total number of patients at each interval following overdose and proportion who reached the endpoint

Time interval from overdose Total number of patients Number of admission ALT > ULN Endpoint Number reaching endpoint n (%) NPV (95% CI) PPV (95% CI) Sensitivity (95% CI) Specificity (95% CI) 
Overall 410 71 1. Doubling of ALT 20 (4.9) 0.98 (0.96–0.99) 0.19 (0.12–0.30) 0.65 (0.43–0.82) 0.86 (0.82–0.89) 
2. Peak ALT 1000 IU/l 16 (3.9) 1.00 (0.99–1.0) 0.23 (0.14–0.34) 1.0 (0.81–1.0) 0.86 (0.82–0.89) 
3. INR >2 8 (1.9) 1.0 (0.99–1.0) 0.14 (0.07–0.25) 1.0 (0.68–1.0) 0.87 (0.84–0.90) 
0–8 h 264 38 1. Doubling of ALT 7 (2.7) 0.98 (0.96–0.99) 0.08 (0.03–0.21) 0.43 (0.16–0.75) 0.86 (0.82–0.90) 
2. Peak ALT 1000 IU/l 3 (1.1) 1.0 (0.98–1.0) 0.08 (0.03–0.21) 1.0 (0.44–1.0) 0.87 (0.82–0.90) 
3. INR >2 3 (1.1) 1.0 (0.98–1.0) 0.09 (0.03–0.23) 1.0 (0.44–1.0) 0.88 (0.84–0.92) 
8–24 h 54 1. Doubling of ALT 3 (5.6) 0.96 (0.86–0.99) 0.17 (0.03–0.56) 0.33 (0.06–0.79) 0.90 (0.79–0.96) 
2. Peak ALT 1000 IU/l 1.0 (0.93–1.0) 0.5 (0–1.0) 0.84 (0.73–0.92) 
3. INR >2 1.0 (0.93–1.0) 0.91 (0.81–0.96) 
>24 h 53 16 1. Doubling of ALT 9 (17.0) 0.97 (0.86–0.99) 0.50 (0.28–0.72) 0.89 (0.57–0.98) 0.82 (0.68–0.91) 
2. Peak ALT 1000 IU/l 10 (18.9) 1.0 (0.90–1.0) 0.63 (0.39–0.82) 1.0 (0.72–1.0) 0.86 (0.72–0.93) 
3. INR >2 4 (7.5) 1.0 (0.91–1.0) 0.33 (0.14–0.61) 1.0 (0.51–1.0) 0.83 (0.70–0.91) 
Staggered 39 1. Doubling of ALT 1 (2.6) 1.0 (0.89–1.0) 0.13 (0.02–0.47) 1 (0.21–1.0) 0.82 (0.67–0.91) 
2. Peak ALT 1000 IU/l 3 (7.7) 1.0 (0.88–1.0) 0.38 (0.14–0.69) 1.0 (0.44–1.0) 0.85 (0.70–0.94) 
3. INR >2 1 (2.6) 1.0 (0.89–1.0) 0.13 (0.02–0.47) 1 (0.21–1) 0.82 (0.67–0.91) 
Time interval from overdose Total number of patients Number of admission ALT > ULN Endpoint Number reaching endpoint n (%) NPV (95% CI) PPV (95% CI) Sensitivity (95% CI) Specificity (95% CI) 
Overall 410 71 1. Doubling of ALT 20 (4.9) 0.98 (0.96–0.99) 0.19 (0.12–0.30) 0.65 (0.43–0.82) 0.86 (0.82–0.89) 
2. Peak ALT 1000 IU/l 16 (3.9) 1.00 (0.99–1.0) 0.23 (0.14–0.34) 1.0 (0.81–1.0) 0.86 (0.82–0.89) 
3. INR >2 8 (1.9) 1.0 (0.99–1.0) 0.14 (0.07–0.25) 1.0 (0.68–1.0) 0.87 (0.84–0.90) 
0–8 h 264 38 1. Doubling of ALT 7 (2.7) 0.98 (0.96–0.99) 0.08 (0.03–0.21) 0.43 (0.16–0.75) 0.86 (0.82–0.90) 
2. Peak ALT 1000 IU/l 3 (1.1) 1.0 (0.98–1.0) 0.08 (0.03–0.21) 1.0 (0.44–1.0) 0.87 (0.82–0.90) 
3. INR >2 3 (1.1) 1.0 (0.98–1.0) 0.09 (0.03–0.23) 1.0 (0.44–1.0) 0.88 (0.84–0.92) 
8–24 h 54 1. Doubling of ALT 3 (5.6) 0.96 (0.86–0.99) 0.17 (0.03–0.56) 0.33 (0.06–0.79) 0.90 (0.79–0.96) 
2. Peak ALT 1000 IU/l 1.0 (0.93–1.0) 0.5 (0–1.0) 0.84 (0.73–0.92) 
3. INR >2 1.0 (0.93–1.0) 0.91 (0.81–0.96) 
>24 h 53 16 1. Doubling of ALT 9 (17.0) 0.97 (0.86–0.99) 0.50 (0.28–0.72) 0.89 (0.57–0.98) 0.82 (0.68–0.91) 
2. Peak ALT 1000 IU/l 10 (18.9) 1.0 (0.90–1.0) 0.63 (0.39–0.82) 1.0 (0.72–1.0) 0.86 (0.72–0.93) 
3. INR >2 4 (7.5) 1.0 (0.91–1.0) 0.33 (0.14–0.61) 1.0 (0.51–1.0) 0.83 (0.70–0.91) 
Staggered 39 1. Doubling of ALT 1 (2.6) 1.0 (0.89–1.0) 0.13 (0.02–0.47) 1 (0.21–1.0) 0.82 (0.67–0.91) 
2. Peak ALT 1000 IU/l 3 (7.7) 1.0 (0.88–1.0) 0.38 (0.14–0.69) 1.0 (0.44–1.0) 0.85 (0.70–0.94) 
3. INR >2 1 (2.6) 1.0 (0.89–1.0) 0.13 (0.02–0.47) 1 (0.21–1) 0.82 (0.67–0.91) 

Table 2 presents the PPV, NPV, sensitivity and specificity with 95% CIs for admission ALT. For endpoint 1 (doubling of ALT from admission), there was an overall PPV of 19.1% with the highest PPV (50%) being in those presenting 24 h after the overdose. In contrast, the NPV of admission ALT for endpoint 1 was 98%.

Endpoint 2 tested the performance of admission ALT as a predictor of ALT peaking above 1000 IU/l. The overall PPV was 22.5%, with the PPV of those presenting 24 h after presentation being 62.5%. The NPV was 100% across all time intervals.

Endpoint 3 tested admission ALT against peak INR >2. PPV remained low at 13.6%; however, NPV was 100%. Once more, the PPV was uniformly low across all durations, demonstrating a peak of 33.3% for those presenting >24 h after overdose.

Increasing time between overdose and starting treatment with AC increases the risk of hepatotoxicity. For all three endpoints, the NPV of hospital admission ALT was high in patients presenting late to hospital (Table 2). Out of the 53 patients who presented >24 h after a single ingestion, 10 had a peak hospital stay ALT >1000 IU/l. In this high-risk group, a normal ALT at presentation still had a 100% NPV. Staggered paracetamol overdoses are challenging to manage clinically because the risk nomogram cannot be used. Although the numbers of these patients were small (n = 39), a normal ALT at presentation still had a 100% NPV.

Discussion

Paracetamol is the commonest drug taken in overdose in UK7,8 with an estimated hospital admission rate of 48% of all overdoses, unfortunately leading to 100–200 deaths per year.9 In Scotland, ∼40–50 patients per year develop significant hepatic dysfunction following paracetamol overdose (ALT > 1000 IU/l or INR > 2) and around 20 patients die per year as a result of overdose.10 Although the management of early paracetamol overdose is relatively straightforward, cases presenting late, following a staggered overdose or those with additional risk factors, present a more complex clinical picture. Currently, UK management of paracetamol overdose is derived from guidelines formulated by the National Poisons Information Service. As little as 10–15 g of paracetamol taken within 24 h could be enough to induce hepatocellular and renal tubular necrosis. For this reason, appropriate timely treatment is vital.10

Recently, the UK management of paracetamol overdose has been changed following reviews by the Medicines and Healthcare products Regulatory Agency.6 A single-line treatment nomogram is used to determine need for AC treatment based on a timed blood paracetamol concentration. This new nomogram may result in substantially more patients being admitted to hospital for AC treatment.11 As patients who present early to hospital typically have no symptoms or signs of hepatotoxicity, there is an urgent need to develop new circulating biomarkers that inform treatment decisions at the hospital ‘front door’. Such markers have the potential to refine patient care in a number of important ways. Future clinical trials of new therapeutic strategies could be stratified using one or more biomarkers with the objective of targeting new therapies or management pathways at lower or higher risk patients. Qualification of new biomarkers may galvanize the interest of pharma in developing new treatments as patient stratification can significantly improve the cost-effectiveness, and therefore health service use of new drugs. If such trials were successful, more individualized treatment would be possible at the hospital ‘front door’, as is already the case with the management of acute coronary syndromes (being routinely stratified by sensitive cardiac biomarkers such as troponin). Currently, patients who have taken a single paracetamol overdose, who present late to hospital, can only have liver injury confidently excluded if their standard liver function tests are normal around 24–36 h after overdose. Biomarkers may allow earlier exclusion of injury which would have an impact on hospital bed occupancy and avoid adverse AC reactions by reducing unnecessary treatment. Patients with a timed blood paracetamol concentration below the at-risk line on the treatment nomogram are usually not treated. Sensitive biomarkers of hepatotoxicity may increase physician confidence in discharging the patient from hospital and could reduce the number of patients being incorrectly classified as not needing treatment. Serum ALT activity is the most frequently used laboratory indicator of hepatotoxicity.12,13 It is commonly measured on first presentation to hospital, but at present, it is not used to guide treatment decisions because most patients present to hospital soon after drug ingestion (70% within 4 h of overdose).2 Unfortunately, this remains a timeframe when ALT cannot diagnose or exclude liver injury.

Our study has demonstrated that a normal admission ALT has a high NPV for subsequent liver injury if AC course is completed. This is consistent with a previous study of 94 patients, all of whom had a peak transaminase activity >1000 IU/l.14 In this Canadian cohort of patients with significant hepatotoxicity, transaminase activity was already increased at hospital admission. As already stated, ALT has important limitations as a biomarker, so our group has recently investigated a panel of new circulating biomarkers that have enhanced liver selectivity (microRNA 122) or provide mechanistic information (HMGB1 and keratin-18) in plasma from patients following paracetamol overdose.15 In patients who developed hepatotoxicity, these markers were substantially increased on first presentation to hospital at a time when ALT was still in the normal range. This demonstrated that patients who develop paracetamol-induced hepatotoxicity have evidence of cellular damage soon after overdose and is ‘proof of concept’ that front door biomarker development is possible for this common poison. In the present study, the NPV of ALT was high even in patients presenting >8 h after overdose, although these patient numbers were relatively small. Therefore, this straightforward, routinely available test could be used to identify lower risk patients. There are two scenarios where this could be useful. Currently, all patients are treated with the same dose of AC. It is probable that this leads to some patients being over-treated16 resulting in avoidable adverse drug reactions to AC and unnecessary hospital bed occupancy. Hospital admission ALT within the normal range identifies a low-risk group and could be used as inclusion criteria for clinical trials of new targeted therapeutic strategies. New circulating biomarkers are in clinical development that appears more sensitive than ALT.17,18 These may make it possible to better identify lower risk patients and allow earlier rule in/rule out of liver cell death following paracetamol overdose.

Our findings suggest that admission ALT can be used to inform the treating clinician about patient prognosis; if the ALT activity is within the normal range and the patient completes treatment, then liver injury is unlikely, even in patients who present late to hospital. In contrast, an increased ALT activity at presentation to hospital had a low PPV. This can probably be explained through the knowledge that serum ALT activity is increased in co-morbidities such as alcohol excess, biliary disease, diabetes, obesity, heart failure and myopathy. Therefore, our findings indicate that an increased ALT alone cannot be used to identify higher risk patients at first presentation to hospital, whereas a normal ALT may be useful in identifying patients at lower risk.

Conclusion

Our study tested the predictive value of admission serum ALT activity in paracetamol overdose. It demonstrated that on first presentation to hospital, normal serum ALT is a predictor of very low risk of subsequent liver injury on completion of AC. If confirmed in larger clinical studies, this simple test could be used, alone or in combination with new biomarkers, to identify low-risk patients for clinical trials of novel therapeutic strategies.

Funding

J.W.D. acknowledges the funding of NHS Research Scotland through NHS Lothian.

Conflict of interest: None declared.

Acknowledgements

J.W.D. acknowledges the contribution of the British Heart Foundation Centre of Research Excellence Award.

References

1
Marudanayagam
R
Shanmugam
V
Gunson
B
Mirza
DF
Mayer
D
Buckels
J
, et al.  . 
Aetiology and outcome of acute liver failure
HPB
 , 
2009
, vol. 
11
 (pg. 
429
-
34
)
2
Prescott
K
Stratton
R
Freyer
A
Hall
I
Le Jeune
I
Detailed analyses of self-poisoning episodes presenting to a large regional teaching hospital in the UK
Br J Clin Pharamacol
 , 
2009
, vol. 
68
 (pg. 
260
-
8
)
3
Pettie
JM
Dow
DA
Sandilands
EA
Thanacoody
HK
Bateman
DN
An integrated care pathway improves the management of paracetamol poisoning
Emer Med J
 , 
2012
, vol. 
29
 (pg. 
482
-
6
)
4
Prescott
LF
Treatment of severe acetaminophen poisoning with intravenous acetylcysteine
Arch Intern Med
 , 
1981
, vol. 
141
 (pg. 
386
-
9
)
5
Prescott
LF
Paracetamol (Acetaminophen). A Critical Bibiliographic Review
 , 
1996
London, Taylor & Francis
6
Medicine and Healthcare products Regulatory Agency
Paracetamol overdose: simplification of the use of intravenous acetylcysteine
  
7
Thomas
SHL
Bevan
L
Bhattacharyya
S
Bramble
MG
Chew
K
Connolly
J
, et al.  . 
Presentation of poisoned patients to accident and emergency departments in the North of England
Hum Exp Toxicol
 , 
1996
, vol. 
15
 (pg. 
466
-
70
)
8
Bialas
MC
Reid
PG
Beck
P
Lazarus
JH
Smith
PM
Scorer
RC
, et al.  . 
Changing patterns of self-poisoning in a UK health district
Q J Med
 , 
1996
, vol. 
89
 (pg. 
893
-
901
)
9
Fagan
E
Wanna
G
Reducing paracetamol overdoses
BMJ
 , 
1996
, vol. 
313
 (pg. 
1417
-
8
)
10
Bateman
DN
Poisoning: Focus on paracetamol
J R Coll Physicians Edin
 , 
2007
, vol. 
37
 (pg. 
332
-
4
)
11
McQuade
DJ
Dargan
PI
Keep
J
Wood
DM
Paracetamol toxicity: what would be the implications of a change in UK treatment guidelines?
Eur J Clin Pharmacol
 , 
2012
, vol. 
68
 (pg. 
1541
-
7
)
12
Amacher
DE
Serum transaminase elevations as indicators of hepatic injury following the administration of drugs
Regul Toxicol Pharmacol
 , 
1998
, vol. 
27
 (pg. 
119
-
30
)
13
Amacher
DE
A toxicologist’s guide to biomarkers of hepatic response
Hum Exp Toxicol
 , 
2002
, vol. 
21
 (pg. 
253
-
62
)
14
Green
TJ
Sivilotti
MLA
Langmann
C
Yarema
M
Juurlink
D
Burns
MJ
, et al.  . 
When do the aminotransferases rise after acute acetaminophen overdose?
ClinToxicol
 , 
2010
, vol. 
48
 (pg. 
787
-
92
)
15
Antoine
DJ
Dear
JW
Starkey-Lewis
P
Platt
V
Coyle
J
Masson
M
, et al.  . 
Mechanistic biomarkers provide early and sensitive detection of acetaminophen-induced acute liver injury at first presentation to hospital
Hepatology
  
doi: 10.1002/hep.26294 [Epub 6 February 2013]
16
Rumack
BH
Bateman
DN
Acetaminophen and acetylcysteine dose and duration: past, present and future
Clin Toxicol
 , 
2012
, vol. 
50
 (pg. 
91
-
8
)
17
Starkey Lewis
PJ
Dear
J
Platt
V
Simpson
KJ
Craig
DG
Antoine
DJ
, et al.  . 
Circulating microRNA as potential markers of human drug-induced liver injury
Hepatology
 , 
2011
, vol. 
54
 (pg. 
1767
-
76
)
18
Antoine
DJ
Jenkins
RE
Dear
JW
Williams
DP
McGill
MR
Sharpe
MR
, et al.  . 
Molecular forms of HMGB1 and Keratin-18 as mechanistic biomarkers for mode of cell death and prognosis during clinical acetaminophen hepatotoxicity
J Hepatol
 , 
2012
, vol. 
56
 (pg. 
1070
-
9
)