Plasma paracetamol concentration at hospital presentation has a dose-

dependent relationship with liver injury despite prompt treatment with

intravenous acetylcysteine

David G Cairney1, Hannah KS Beckwith2, Khalid Al-Hourani1, Michael Eddleston1,3,

D Nicholas Bateman1 & James W Dear*1,3.

1. NPIS Edinburgh, Royal Infirmary of Edinburgh, Edinburgh UK

2. Department of Renal Medicine, Imperial College, London, UK.

3. Pharmacology, Toxicology and Therapeutics, University/BHF Centre for

Cardiovascular Science, University of Edinburgh, Edinburgh UK

*corresponding author:

Dr James Dear Pharmacology, Toxicology and TherapeuticsEdinburgh University/BHF Centre for Cardiovascular Science Queen's Medical Research Institute 47 Little France Crescent Edinburgh, EH16 4TJ. UK Tel 44 131 2429236 Email: james.dear@ed.ac.uk

Keywords: hepatotoxicity, paracetamol, nomogram, acetylcysteine

Written in UK English

1

Abstract

Context: paracetamol (acetaminophen) overdose is a common reason for

emergency hospital admission in the UK and the leading cause of acute liver

failure in the Western world. Currently, the antidote acetylcysteine (NAC) is

administered at a dose determined only by body weight without regard for the

body burden of paracetamol.

Objective: to determine whether higher plasma paracetamol concentrations are

associated with increased risk of liver injury despite prompt treatment with

intravenous NAC.

Methods: patients admitted to hospital for treatment with intravenous NAC

following a single acute paracetamol overdose entered the study if NAC was

commenced within 24 hours of drug ingestion (N=727 hospital presentations).

Based on the plasma paracetamol concentration at first presentation to hospital,

a series of nomograms were created: 0-100, 101-150, 151-200, 201-300, 301-

500 and over 501 mg/L. The primary endpoints were acute liver injury (ALI -

peak serum ALT activity > 150U/L and double the admission value) and

hepatotoxicity (peak ALT > 1000U/L).

Results: ALI and hepatotoxicity were more common in patients with higher

admission plasma paracetamol concentrations despite NAC treatment (ALI:

nomogram 0-100: 6%, 101-150: 3%, 151-200: 3%, 201-300: 9%, 301-500: 13%,

over 501mg/dL: 27%. P<0.0001). This dose-response relationship between

paracetamol concentration and ALI persisted even in patients treated with NAC

2

within 8 hours of overdose (nomogram 0-100: 0%, 101-150: 0.8%, 151-200: 2%,

201-300: 3.6%, 301-500: 12.5%, over 501mg/L: 33%. P<0.0001) and in patients

with normal ALT activity at first presentation (nomogram: 0-100: 0%, 101-150:

1.2%, 151-200: 1.5%, 201-300: 5.3%, 301-500: 10.8% P<0.0001).

Discussion: patients with increased concentrations of plasma paracetamol at

hospital presentation are at higher risk of liver injury even when intravenous

NAC is promptly administered before there is biochemical evidence of toxicity.

Conclusion: this study supports theoretical concerns that the current

intravenous dose of NAC may be too low in the setting of higher paracetamol

exposure.

3

Introduction

Paracetamol (acetaminophen) overdose is a common reason for emergency

hospital attendance and the leading cause of acute liver failure in the Western

world.(1) Annually, in the UK, paracetamol overdose results in approximately

100,000 Emergency Department presentations and 50,000 acute hospital

admissions(2) and is the direct cause of death in around 150 people.(3) The

mechanism of liver injury after paracetamol overdose is well defined and can be

translated from rodent models to humans using mechanistic biomarkers.(4) In

overdose, the reactive metabolite N-acetyl-p-benzoquinone imine (NAPQI) is

generated in excess, depleting cellular glutathione (GSH) then binding covalently

to cellular proteins resulting in oxidative stress and hepatocyte death,

predominately by necrosis.(5) The current antidote, acetylcysteine (NAC),

replenishes cellular GSH and is highly effective at preventing liver injury if

administered orally or intravenously soon after overdose.(6, 7) However,

improved stratification of patients is needed to target NAC at the correct dose to

patients who stand to benefit because NAC is commonly associated with

unpleasant and occasionally severe dose-related adverse drug reactions (ADRs).

(8) Furthermore, current NAC treatment regimens are time-consuming so result

in significant hospital bed occupancy (around 65,000 bed days per year in UK).

(2)(9) With regard to ADRs, in a recent randomised trial nausea/vomiting

occurred in more than half of intravenous NAC recipients and anaphylactoid

reactions in about a third.(9) These severe reactions result in treatment

interruption, treatment refusal and extended hospital stays.

4

In a patient who has overdosed on paracetamol the decision to commence

treatment with NAC is commonly based on a timed plasma paracetamol

measurement. However, once NAC treatment is deemed necessary, its dose is

solely based on the patient’s weight without regard for the body’s drug burden

(as quantified by the reported dose of drug ingested or paracetamol

concentration) or presence of liver injury. Using the initial paracetamol

concentration as a measure of exposure, in those patients above the 200-line on

the current nomogram, Smilkstein et al. reported no increase in the incidence of

hepatotoxicity (defined as a serum AST activity >1000U/L) with increased

paracetamol concentration in patients treated within 8 hours of overdose with

oral NAC.(6) Below the 200-line, hepatotoxicity was rare when oral NAC

treatment started within 8 hours of overdose (around 1%).(10) Patient numbers

with hepatotoxicity were small and a type II error is possible, but these data may

suggest that the oral dose of NAC is sufficient for treatment of large paracetamol

overdoses when administered soon after overdose. Theoretically, the standard

intravenous NAC dose has been estimated to be sufficient to treat a paracetamol

overdose of around 16g but may not be high enough to prevent all liver damage

for larger body burdens.(11) It is uncertain whether to increase the dose of NAC

in patients with larger overdoses, what the cut off level for use of higher doses of

NAC should be, and how much additional NAC to give (some clinicians suggesting

a doubling of the dose in the third bag of the standard infusion for patients with

plasma paracetamol above 450mg/L).(12) The objective of this study was to

determine whether higher plasma paracetamol concentrations on admission are

associated with increased risk of acute liver injury in patients treated promptly

with standard dose intravenous NAC.

5

Methods

The Edinburgh Clinical Toxicology Unit has a prospective 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

NAC. 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. The database is part of an ongoing clinical audit of patient care

and is fully-anonymised and exempt from needing review by our ethics

committee but operates with appropriate approval of the local Caldicott

Guardian. The inclusion criterion for this study was a clear history of a single

acute paracetamol overdose (tablets ingested over less than 2h) with a patient-

reported timing that was treated with intravenous NAC within 24 hours of drug

ingestion. Before September 2012, the regimen for NAC was 150 mg/kg over 15

min, 50 mg/kg over 4 h then 100 mg/kg over 16 h. After September 2012 the

first NAC bag was given over 1h rather than 15min, the total dose of NAC being

the same, but given over 21 h.(2) Blood sampling earlier than 4h post-overdose

was considered a 4h result if our database recorded the time as 3h 30min to 4h.

Exclusion criteria were as follows: patients who did not complete NAC treatment

and patients whose blood tests were not repeated at end of NAC. The decision to

treat was made by the treating clinician using the contemporaneous UK

paracetamol treatment nomogram (prior to September 2012 the lines were 200

without risk factors for toxicity and 100 with risk factors, since September 2012

there has been a single 100 line that determines need for NAC). From February

2005 to August 2013 data were retrospectively recorded from patients’ records

6

and the initial plasma paracetamol concentration was plotted against time from

overdose. A series of patient groups were defined, prior to commencing the

study, using nomograms based on the standard Rumack-Matthew and Prescott

nomograms derived in the 1970s from untreated patients managed in

Edinburgh(13, 14): 0-100, 101-150, 151-200, 201-300, 301-500 and over 501

mg/L. These concentrations refer to the back extrapolation of the nomograms to

the 4h concentration using a 4h drug half-life as originally suggested. The

primary endpoints - whose frequency was determined in each nomogram group

- were acute liver injury (ALI), defined as peak serum ALT activity > 150U/L and

double the admission value, and hepatotoxicity defined conventionally as a peak

ALT > 1000U/L. At the time of planning the study, this definition of ALI was

chosen as it was the UK indication for continued NAC therapy at the end of the

standard 21h regimen.

Data are expressed as median and inter-quartile range. Chi-square test for trend

and Kruskal–Wallis test were used to compare the incidence of injury and peak

ALT across groups, respectively. Nominal significance was P<0.05.

7

Results

In total 727 presentations with paracetamol overdose ingested over less than 2

hrs and receiving NAC within 24 hrs were included in this study. The patient

demographics are presented in table 1. The presentations were spread across

the range of paracetamol concentration nomogram groups. The patients in the

over 501mg/L nomogram group presented significantly later to hospital,

compared with other groups. There was no significant difference in time from

overdose to commencing NAC across the lower nomogram groups (table 1) and

no difference in those patients who started NAC within 8 hours of overdose

(nomogram group 0-100: 6.8h (6.2-8), 101-150: 6.9h (6.0-7.9), 151-200: 6.6h (6-

7.7), 201-300: 6.5h (6-7.7), 301-500: 6.9h (6.1-7.8) over 501mg/dL: 7.1h (6.4-

8.7)). A higher proportion of patients in the 0-100 nomogram group had an

elevated ALT at first presentation to hospital in comparison with the other

groups (table 1). This is likely to explain why these low-risk patients were

treated with NAC, at least in part. Fifty-three patients developed ALI and 30

developed hepatotoxicity during the index hospital admission (figure 1).

ALI and hepatotoxicity was more common with higher admission paracetamol

concentrations despite NAC treatment (figure 2). Patients in the 0-100

nomogram group had a higher rate of ALI and hepatotoxicity compared with the

101-150 and the 151-200 groups. This increased rate of ALI in the 0-100

nomogram group did not persist when those with elevated ALT on presentation

were excluded (figure 3). The dose-response relationship between paracetamol

concentration and ALI persisted even when NAC treatment was started within 8

8

hours of overdose (nomogram 0-100: 0%, 101-150: 0.8%, 151-200: 2%, 201-

300: 3.6%, 301-500: 12.5%, over 501mg/L: 33%. P<0.0001) and in patients with

normal serum ALT activity, defined as 50U/L or less, at hospital presentation

(nomogram: 0-100: 0%, 101-150: 1.2%, 151-200: 1.5%, 201-300: 5.3%, 301-

500: 10.8% P<0.0001) (figure 3). There were too few patients in the >501mg/L

nomogram group with normal ALT at presentation for analysis with regard to

ALI. There were also too few patients with hepatotoxicity when NAC treatment

was started within 8 hours of overdose or with a normal admission ALT for

robust analysis. Analysis with ALT as a continuous variable demonstrated that

the peak hospital stay ALT was higher in patients in higher paracetamol

nomogram groups, but with marginal significance given our pre-defined nominal

cut-off value (P=0.055) (figure 4).

9

Discussion

This study shows that patients with higher concentrations of paracetamol at

hospital presentation are at higher risk of liver injury even when intravenous

NAC is promptly administered before there is biochemical evidence of toxicity.

This effect was demonstrated with a clinically relevant but modest increase in

ALT, substantially larger studies are needed to confirm a similar dose-response

relationship with more significant hepatotoxicity and acute liver failure. This

finding is clinically important because it suggests the current intravenous dose of

NAC may be too low in the setting of higher paracetamol exposure. However,

further studies are needed to establish the benefits of higher acetylcysteine

doses in such patients. While this study cannot define an absolute paracetamol

concentration where additional NAC should be given, or the amount of additional

NAC to use in any particular patient, the data suggest that patients who lie below

the 200 mg/L nomogram line respond similarly to present doses of NAC. In

patients with blood concentrations of paracetamol above the 200 mg/L line

there would appear to be a dose response to paracetamol toxicity, with higher

toxicity above the 500 mg/L nomogram line than between 300 and 500 mg/L.

At the hospital ‘front-door’ the decision to treat with NAC following a

paracetamol overdose is largely based on the plasma paracetamol concentration

interpreted using a nomogram that plots concentration against time after

overdose, because liver injury cannot be confidently confirmed or excluded by

standard clinical chemistry such as serum ALT activity. This approach to patient

stratification results in the overtreatment of low-risk patients as paracetamol

10

concentration is only a surrogate marker for risk of hepatotoxicity and, if left

untreated, a substantial proportion of ‘at-risk’ patients will not develop

significant liver injury.(15) The data presented in this paper indicate that there

may also be patients who are undertreated with the current ‘one size fits all’

approach to NAC prescribing. This is based on the finding that those with higher

paracetamol concentrations have a higher rate of liver injury even if treated

within 8 hours of drug ingestion and in those patients without evidence of liver

injury on routine clinical chemistry. This study demonstrates a dose-response

relationship between paracetamol concentration and liver injury in patients

treated with IV NAC. However, clinicians already recognize this is theoretically

likely and higher IV NAC dosing is recommended for patients with high

admission paracetamol concentrations in the recent Australian and New Zealand

guidelines.(16) By contrast with the data presented in this paper, Smilkstein et

al. reported no increase in hepatotoxicity (AST > 1000 IU) with increased toxic-

range blood paracetamol concentration in patients treated within 8 hours of

overdose with oral NAC.(6) Smilkstein et al. did not examine the effect of oral

NAC on smaller transaminase elevations such as ALI as defined in our paper. The

lack of an increase in hepatotoxicity may relate to the relatively small study size

(N=729), but may also reflect the higher total dose of NAC administered with the

oral protocol, an important fundamental difference. A difference in the efficacy of

the IV and oral NAC protocols was also suggested by Yarema et al. who reported

superior efficacy of the oral protocol in patients presenting 18 hour after

overdose. (17) However, the IV NAC arm of this paper was a retrospective study

compared with historical controls (a prospective cohort study of oral NAC) and

the findings have yet to be confirmed in a prospective trial.

11

The management of paracetamol overdose is changing with new biomarker

strategies emerging such as using the product of the paracetamol concentration

and serum ALT activity to determine risk of liver injury independent of time

from drug ingestion.(18) The data in this paper supports this approach as it

demonstrates that paracetamol concentration has a dose-response relationship

with injury which complements first presentation ALT activity, which has a high

negative predictive value if NAC is commenced when ALT activity is in the

normal range.(19) Novel biomarkers with enhanced specificity/sensitivity may,

with development, augment or even replace ALT for risk stratification at first

presentation to hospital after overdose and for early exclusion of injury.(4)

Although speculative, these tools may facilitate stratified clinical trials of new

treatment pathways including testing whether higher NAC doses have increased

efficacy in patients with high exposures to paracetamol or circulating biomarker

evidence of hepatocyte injury. Further clinical studies are warranted to

determine whether the intravenous NAC dose should be higher in those patients

with higher paracetamol concentration. One advantage of the recently published

12 h NAC regimen(9) is that it delivers the entire NAC dose in 12 h, with minimal

ADRs. Further increased dose in the second infusion and a prolonged third

additional infusion of NAC are possible with this regimen, with the mandatory 12

h post start of infusion blood tests facilitating more flexibility in patient NAC

dosing.

There are a few limitations to this study. The time from overdose to taking blood

and starting NAC is dependent on the accuracy of the patient history, which can

12

be variable and may add error. However, this error is likely to be spread equally

across the nomogram groups. We do not have measurements of INR in all the

patients so the relationship between paracetamol concentration and peak INR is

unknown. It is likely that those patients in the higher nomogram groups will be

at increased risk of elevations in INR, but substantially larger patient numbers

may be needed to demonstrate this relationship given that INR elevation is less

frequent than ALI. While it seems intuitive that an increased incidence of ALI in

our small study will translate to increased need for transplantation and mortality

in larger populations, this relationship is not clearly defined. A paracetamol half-

life of 4 hours was used to construct the nomograms. This estimate of the true

half-life is used in the clinical practice, however, patients with liver injury have

prolonged paracetamol clearance and a 4h half-life may be too short.(20) The

effect of this may be that patients with large overdoses are placed in a higher

nomogram group if they present late compared with if they present early.

However, the presence of a dose-response relationship in those patients who

were treated with NAC within 8h of overdose makes it unlikely that prolongation

of the drug half-life is confounding our data. Recent studies demonstrate that

almost all patients who develop liver injury subsequent to paracetamol overdose

have evidence of hepatocyte cell death at first presentation to hospital when

more sensitive circulating markers are measured.(4) The data we present

suggests that paracetamol concentration is an important factor affecting toxic

hazard, particularly if it lies above the 300mg/L nomogram. It remains to be

determined whether paracetamol concentration will remain an independent

predictor of outcome if highly sensitive markers of liver toxicity are measured. In

the UK, prior to 2012, patients with admission paracetamol concentrations

13

between the 100 – 200 lines would be treated only if risk factors for

hepatotoxicity were present (either glutathione depleting or P450 enzyme

inducing). Over the 200 line, all patients should have received NAC. This could

mean that patients below the 200 line in the present study have a higher rate of

risk factors and, therefore, the incidence of hepatotoxicity may be higher than if

our study were repeated only with patients presenting to hospital after 2012.

14

Conclusion

Higher plasma paracetamol is associated with increased risk of liver injury

despite intravenous NAC. Given that paracetamol overdose is one of the most

common medical emergencies, there is a need for further clinical studies that

explore and refine NAC dosing at both ends of the overdose spectrum, to reduce

unnecessary treatment of low-risk patients and to ensure adequate dosing of

those patients exposed to large amounts of paracetamol.

Acknowledgements:

The authors report no declarations of interest.

15

References

1. Larson AM, Polson J, Fontana RJ, Davern TJ, Lalani E, Hynan LS, et al. Acetaminophen-induced acute liver failure: results of a United States multicenter, prospective study. Hepatology. 2005;42(6):1364-72.2. Bateman DN, Carroll R, Pettie J, Yamamoto T, Elamin ME, Peart L, et al. Effect of the UK's revised paracetamol poisoning management guidelines on admissions, adverse reactions and costs of treatment. British Journal of Clinical Pharmacology. 2014;78(3):610-8.3. Hawton K, Bergen H, Simkin S, Dodd S, Pocock P, Bernal W, et al. Long term effect of reduced pack sizes of paracetamol on poisoning deaths and liver transplant activity in England and Wales: interrupted time series analyses. BMJ. 2013;346:f403.4. 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. 2013;58:777-87.5. Hinson JA, Roberts DW, James LP. Mechanisms of acetaminophen-induced liver necrosis. Handb Exp Pharmacol. 2010(196):369-405.6. Smilkstein MJ, Knapp GL, Kulig KW, Rumack BH. Efficacy of oral N-acetylcysteine in the treatment of acetaminophen overdose. Analysis of the national multicenter study (1976 to 1985). The New England Journal of Medicine. 1988;319(24):1557-62.7. Prescott LF, Park J, Ballantyne A, Adriaenssens P, Proudfoot AT. Treatment of paracetamol (acetaminophen) poisoning with N-acetylcysteine. Lancet. 1977;2(8035):432-4.8. Sandilands EA, Bateman DN. Adverse reactions associated with acetylcysteine. Clin Toxicol (Phila). 2009;47(2):81-8.9. Bateman DN, Dear JW, Thanacoody HK, Thomas SH, Eddleston M, Sandilands EA, et al. Reduction of adverse effects from intravenous acetylcysteine treatment for paracetamol poisoning: a randomised controlled trial. Lancet. 2014;383(9918):697-704.10. Rumack BH. Acetaminophen hepatotoxicity: the first 35 years. J Toxicol Clin Toxicol. 2002;40:3-20.11. Rumack BH, Bateman DN. Acetaminophen and acetylcysteine dose and duration: past, present and future. Clin Toxicol (Phila). 2012;50(2):91-8.12. Juma S, Villeneuve E, Elliot A, Palmer RB, Gosselin S. Doubling the third dose of intravenous N-acetylcysteine survey: An international practice perspective. (2015) XXXV International Congress of the European Association of Poisons Centres and Clinical Toxicologists (EAPCCT) 26–29 May 2015, St Julian's, Malta, ClinicalToxicology, 53:4, 233-403, DOI: 103109/1556365020151024953. 2015.13. Rumack BH, Matthew H. Acetaminophen poisoning and toxicity. Pediatrics. 1975;55:871-6.14. Wright N, Prescott LF. Letter: Treatment of paracetamol poisoning. Br Med J. 1975;2(5966):337.15. Prescott LF. Paracetamol (Acetaminophen): a critical bibliographic review. Revised Second Edition. London: Taylor & Francis. ; 2001. p. 587-8.

16

16. Chiew AL, Fountain JS, Graudins A, Isbister GK, Reith D, Buckley NA. Summary statement: new guidelines for the management of paracetamol poisoning in Australia and New Zealand. Med J Aust. 2015;203(5):215-8.17. Yarema MC, Johnson DW, Berlin RJ, Sivilotti ML, Nettel-Aguirre A, Brant RF, et al. Comparison of the 20-hour intravenous and 72-hour oral acetylcysteine protocols for the treatment of acute acetaminophen poisoning. Ann Emerg Med. 2009;54(4):606-14.18. Sivilotti ML, Green TJ, Langmann C, Yarema M, Juurlink D, Johnson D. Multiplying the serum aminotransferase by the acetaminophen concentration to predict toxicity following overdose. Clin Toxicol (Phila). 2010;48(8):793-9.19. Al-Hourani K, Mansi R, Pettie J, Dow M, Bateman DN, Dear JW. The predictive value of hospital admission serum alanine transaminase activity in patients treated for paracetamol overdose. QJM. 2013;106(6):541-6.20. Schiodt FV, Ott P, Christensen E, Bondesen S. The value of plasma acetaminophen half-life in antidote-treated acetaminophen overdosage. Clin Pharmacol Ther. 2002;71(4):221-5.

17

Legends

Figure 1. Paracetamol concentration versus time of ingestion for single acute

overdoses. Sloped lines are an extension of the Rumack-Matthew nomogram

corresponding to 100, 150, 200, 300 and 500 mg/L at 4 hours post-ingestion

(based on a first-order decay of paracetamol with a half-life of 4 hours). Dots are

cases with peak ALT<150U/L, triangles are cases with peak ALT>150U/L which

doubled from admission and squares are cases with peak ALT>1000U/L. The

vertical line at 8 hours separates "early" and "late" presenters.

Figure 2. The percentage of patients in each paracetamol nomogram group with

liver injury. Patients were grouped by their plasma paracetamol concentration at

first presentation to hospital and the incidence of acute liver injury (peak

ALT>150U/L which doubled from admission) and hepatotoxicity (peak

ALT>1000U/L) was determined. Bars represent the mean, error bars the 95% CI.

Figure 3. The percentage of patients in each paracetamol nomogram group with

liver injury who were at lower risk because they received NAC within 8 hours of

overdose or had a normal serum ALT activity at presentation. Patients were

grouped by their plasma paracetamol concentration at first presentation to

hospital and the incidence of acute liver injury (peak ALT>150U/L which

doubled from admission) was determined. Bars represent the mean, error bars

the 95% CI.

18

Figure 4. Peak ALT activity from index hospital admission following single acute

paracetamol overdose grouped by paracetamol nomogram group. Median and IQR

displayed as box, with 1.5x IQR whiskers and multiple outliers.

19

  Nomogram groupsPatient Characteristics Total 0-100 101-150 151-200 201-300 301-500 501+Cases 727 70 195 149 186 82 45Sex (%male) 36 34 40 37 33 32 43Median age (years) 33  33 34 34 30 33 39Median admission paracetamol level (IQR)(mg/L) 130 (80-192) 65 (18-85) 109 (80-123) 157 (122-172) 196 (105-229) 279 (179-349) 101 (49-236)Median ingestion to paracetamol level (IQR)(hr:min) 4:45 (4:04-8:31) 4:44 (4:13-9:52) 4:20 (4:01-9:52) 4:27 (4:02-6:25) 4:49 (4:04-8:41) 5:10 (4:13-8:23) 16:42 (10:33-20:03) * Median ingestion to NAC initiation (IQR)(hr:min) 7:30 (6:18-11) 8:15 (6:48-11:45) 7:06 (6:18-9:30) 7:00 (6:09-8:51) 7:18 (6:12-11) 7:42 (6:23-11) 18 (11-22) *Number with normal admission ALT (<50U/L) (% group) 629 (87%) 53 (76%) 169 (87%) 131 (88%) 170 (91%) 74 (90%) 32 (71%)Number with acute liver injury (ALT>150U/L) (% group) 53 (7%) 4 (6%) 6 (3%) 4 (3%) 16 (9%) 11 (13%) 12 (27%)Number with hepatotoxicity (ALT>1000U/L) (% group) 30 (4%) 3 (4%) 4 (2%) 3 (2%) 7 (4%) 5 (6%) 8 (18%)

Table 1. Patient demographics described by nomogram group. *P<0.01 when compared to 101-150 group.

20

Figure 1

21

Figure 2

Figure 3

22

Figure 4

23