Quantitative Systems Toxicology Modeling of Acetaminophen Pharmacokinetics and Hepatic Biomarkers After Overdoses of Extended‐Release and Immediate‐Release Formulations in Adults With Chronic Alcohol Use or Low Glutathione

Abstract

ABSTRACT

Acetaminophen (APAP), an over‐the‐counter analgesic and antipyretic, can cause hepatotoxicity when ingested in large overdoses. APAP has multiple formulations including immediate‐release (IR) and extended‐release (ER) preparations. A recently published consensus statement on the management of APAP poisoning indicated that management of APAP‐ER overdose is the same as that for APAP‐IR overdose. Consistent with this consensus, it was previously reported that quantitative systems toxicology (QST) modeling using DILIsym predicted similar pharmacokinetic (PK) and hepatic biomarker profiles for the APAP‐ER and APAP‐IR formulations after overdose in healthy adults. Hepatic injury from APAP is caused by the reactive metabolite, N‐acetyl‐ρ‐benzoquinone imine (NAPQI), which is formed predominantly by CYP2E1‐mediated metabolism and eliminated by hepatic glutathione. As such, conditions that can increase NAPQI production (e.g., CYP2E1 induction by alcohol) or decrease hepatic glutathione stores (e.g., underling liver disease) may impact PK and susceptibility to hepatotoxicity after overdose of APAP‐IR and APAP‐ER. In the current study, APAP‐IR and APAP‐ER models in chronic alcohol users and individuals with low hepatic glutathione were developed and verified within DILIsym. Simulations using verified models predicted similar PK and hepatic biomarker profiles for the APAP‐ER and APAP‐IR formulations in moderate and excessive chronic alcohol users and adults with low hepatic glutathione levels after single acute overdoses up to ~100 g and repeat supratherapeutic ingestions (up to 7.8 g/day for 10 days). These results further support that approaches to manage APAP‐IR overdoses can be applied to manage APAP‐ER overdoses in adults with chronic alcohol consumption or lower hepatic glutathione levels.

---

Summary

• What is the current knowledge on the topic?

  • ○

    Acetaminophen (APAP), an over-the-counter analgesic and antipyretic, can cause hepatotoxicity when taken in overdose. Recently updated guidelines on APAP management suggested that approaches to manage APAP immediate-release (IR) overdose can be applied to manage APAP extended-release (ER) overdose. Quantitative systems toxicology (QST) modeling using DILIsym predicted similar PK and hepatic biomarker profiles for the APAP-ER and APAP-IR formulations after overdose in healthy adults.

• What question did this study address?

  • ○

    Would APAP-ER formulation have similar PK and hepatic biomarker profiles compared to APAP-IR formulation after overdose in adults with chronic alcohol use or low hepatic glutathione?

• What does this study add to our knowledge?

  • ○

    QST modeling predicted similar PK and hepatic biomarker profiles for the APAP-ER and APAP-IR formulations in moderate and excessive chronic alcohol users and adults with low hepatic glutathione levels after single acute overdoses and repeat supratherapeutic ingestions.

• How might this change drug discovery, development, and/or therapeutics?

  • ○

    Approaches to manage APAP-IR overdoses can be applied to manage APAP-ER overdoses in chronic alcohol users and individuals with low glutathione due to stable liver disease.

---

Introduction

Acetaminophen (APAP) is an over-the-counter analgesic and antipyretic that has been widely used in patients with fever and other conditions causing mild to moderate pain. When ingested in overdose, particularly large intentional overdose, APAP can cause hepatotoxicity, which may result in fulminant hepatic failure and death in severe cases. Hepatic injury from APAP is caused by its reactive metabolite, N-acetyl-ρ-benzoquinone imine (NAPQI), which is formed predominantly by CYP2E1-mediated metabolism of APAP. APAP is mostly metabolized to glucuronide and sulfate conjugates that are excreted in the urine, and only a small percentage (< 10%) of APAP is converted to NAPQI. As such, NAPQI is detoxified by hepatic glutathione without any toxicological consequences at therapeutic dose of APAP. However, if hepatic glutathione stores are depleted after overdose of APAP, the excess NAPQI can cause oxidative stress and mitochondrial dysfunction resulting in liver injury [1]. Patients with acute APAP overdose are treated with N-acetyl cysteine (NAC), which is a precursor of glutathione, if their blood APAP concentration is above the treatment line on the Rumack-Matthew nomogram [2]. In the Rumack-Matthew nomogram, serum concentrations of APAP are plotted in a log scale against the time, starting from 4 h after APAP ingestion at which time absorption of APAP is likely to be complete. The original Rumack-Matthew line started at 200 μg/mL at 4 h and follows a slope representing a half-life of 4 h; a treatment line that is 25% below the original, which starts at 150 μg/mL at 4 h, has been used to guide NAC treatment in APAP overdose patients [3].

APAP has multiple formulations including immediate-release (IR), modified-release (MR), and extended-release (ER) preparations. APAP-IR products are labeled for use on a 4-h basis. APAP-MR and APAP-ER products are designed to achieve prolonged absorption and efficacy for an 8-h period, but MR and ER formulations use different technologies. APAP-MR products, which were previously marketed in Europe and Australia, were associated with markedly delayed absorption, potentially due to formation of a bezoar, and unexpected toxicity [4]. There was also a concern that current treatment guidelines developed for APAP overdose were inappropriate for these MR formulations. As a result, the European Medicines Agency recommended suspending medicines containing APAP-MR from the market in 2017. On the other hand, clinical experiences do not support such concerns for APAP-ER products. APAP-ER products, designed with an IR layer and an erodible ER layer, are marketed in the US and Canada. A recently published consensus statement on the management of APAP poisoning indicated that management of APAP-ER overdose is the same as that for APAP-IR overdose, although an additional blood APAP measurement is recommended for APAP-ER if the APAP concentration from samples drawn 4–12 h after ingestion is below the treatment line but above 10 μg/mL [5].

Consistent with clinical experiences, we have previously reported that quantitative systems toxicology (QST) modeling using DILIsym predicted similar PK and hepatic biomarker profiles for the APAP-ER and APAP-IR formulations in healthy adults after single acute overdoses and repeat supratherapeutic ingestions (RSTI) [6]. DILIsym represents drug exposure predicted by physiologically-based pharmacokinetic (PBPK) models, liver biochemistry, and mechanisms contributing to drug-induced liver injury (DILI) [7]. A mechanistic model of APAP-IR was previously developed within DILIsym and verified with clinical and animal datasets [8–10]. In order to address clinical inter-individual variability, a simulated population (SimPops) that represents variability in model parameters related to APAP exposure and mechanistic hepatotoxicity pathways in healthy subjects was also developed and verified based on existing data [8–10]. In our prior work, the PBPK model of APAP-ER was developed and verified within DILIsym to evaluate potential differences in the PK and metabolism of APAP-ER and APAP-IR by simulation over a range of acute and chronic overdoses [6]. In addition, simulations of PK and hepatic biomarkers used to monitor patients, specifically plasma alanine aminotransferase (ALT), total bilirubin (TB), and international normalized ratio (INR), associated with these overdoses were explored in healthy adults. ALT is the primary biomarker of acute liver injury and an elevation of serum TB in the absence of other causes or cholestasis indicates a decrease in hepatic function [11]. Similar to TB, rising INR indicates functional impairment and represents substantial liver injury. Food and Drug Administration (FDA)'s guidance for DILI suggests that increased incidence of ALT > 3X upper limit of normal (ULN) compared to control indicates that a potential for liver injury is high, and Hy's Law cases, defined as ALT > 3X ULN and TB > 2X ULN, are indicative of severe DILI [12].

While simulations predicted similar PK and hepatic biomarker profiles for the APAP-ER and APAP-IR formulations in healthy adults, it remained unanswered if this similarity would be impacted by patient conditions. Specifically, conditions that can increase NAPQI production, such as induction of CYP2E1 by alcohol [13–15], or conditions that can decrease hepatic glutathione store, such as underling liver disease [16, 17], may impact PK and susceptibility to hepatotoxicity after overdose of APAP-IR and APAP-ER. In the current study, DILIsym was employed to evaluate the PK and hepatic responses of APAP-IR and APAP-ER in adults with chronic alcohol use or low hepatic glutathione (Figure 1). APAP-IR and APAP-ER models in moderate chronic alcohol users (MCAU) and excessive chronic alcohol users (ECAU) with no/mild hepatic impairment were developed by updating physiological parameters of previously developed and verified healthy adult APAP models [6] using physiology data from chronic alcohol users. APAP-IR and APAP-ER models in adults with low hepatic glutathione compared to healthy volunteers were developed by updating physiological parameters of previously developed and verified healthy adult APAP models [6] using physiology data from cirrhotic patients. The verified APAP-IR and APAP-ER models in three adult populations (i.e., MCAU, ECAU, low glutathione) were used to simulate PK and hepatic biomarkers after single acute overdose and RSTI of APAP. Simulated PK and hepatic biomarker profiles of APAP-ER were compared with those of APAP-IR by ingested dose for each population.

[IMAGE OMITTED. SEE PDF]

Methods

Software

DILIsym version 8A patch 2 (Simulations Plus, Research Triangle Park, North Carolina) was refined as described in the Methods section and used for APAP PK and hepatotoxicity simulations. GastroPlus version 9.7 Patch 2 (Simulations Plus, Lancaster, CA) was used for alcohol PBPK modeling.

Development and Verification of APAP-IR and APAP-ER Models in Adults With Chronic Alcohol Use

APAP-IR and APAP-ER models in MCAU and ECAU with no/mild hepatic impairment were developed by updating physiological parameters of previously developed and verified healthy adult APAP models [6] using physiology data from chronic alcohol users (Table S1). Representation of physiological changes in MCAU and ECAU focused on chronic effects of alcohol on hepatic glutathione and expression of APAP metabolizing enzymes, as these parameters are the most relevant to NAPQI-mediated liver injury. Of note, simulated chronic alcoholics do not represent underlying alcohol-induced liver injury (e.g., elevated hepatic biomarkers). The simulated baseline MCAU, defined as individuals who chronically consume moderate amounts of alcohol, represents induced CYP2E1 by 1.78-fold compared to healthy subjects, which is the weighted average of the reported fold changes in MCAU (Table S1). The simulated ECAU, defined as individuals who chronically consume excessive amounts of alcohol, represents reduced hepatic glutathione levels by 38.2% and induced CYP2E1 by 2.0-fold compared to healthy subjects, which are the weighted averages of the reported fold changes in ECAU (Table S1). Details regarding the development and validation of MCAU and ECAU populations are provided in Supporting Information S1. Comparison of parameter distribution in MCAU, ECAU, and healthy adult SimPops is presented in Figure S1.

In addition to chronic effects, alcohol has been shown to competitively inhibit CYP2E1 (acute effects), which might offset the effect of CYP2E1 induction [18]. To simulate the effect of ethanol-mediated competitive inhibition of CYP2E1, a PBPK model of ethanol disposition was developed and verified using GastroPlus, and alcohol was simulated as a co-ingestant with an inhibition constant of 16.5 mM for CYP2E1 [18]. Details regarding the development and verification of the PBPK model are provided in Supporting Information S1, Table S3, and Figure S2. Clinical data used to verify APAP models in MCAU and ECAU are summarized in Table S2 [16, 19–24].

Development and Verification of APAP-IR and APAP-ER Models in Adults With Low Glutathione

APAP-IR and APAP-ER models in adults with low hepatic glutathione were developed by updating physiological parameters of previously developed and verified healthy adult APAP models [6] using physiology data from cirrhotic patients (Table S1). Representation of physiological changes in adults with low glutathione focused on altered hepatic glutathione levels and expression of APAP metabolizing enzymes as these parameters are the most relevant to NAPQI-mediated liver injury. Of note, simulated adults with low glutathione do not represent underlying liver injury (e.g., elevated hepatic biomarkers). The simulated baseline adult with low glutathione represents reduced hepatic glutathione, CYP2E1, CYP3A4, CYP2A6, sulfotransferases, and renal clearance by 22%, 62%, 67%, 57%, 76%, and 43%, respectively, compared to healthy subjects which were weighted-average values of the reported fold-changes in cirrhotic patients (Table S1). Details regarding the development of low glutathione SimPops are provided in Supporting Information S1. Comparison of parameter distribution in low glutathione and healthy adult SimPops is presented in Figure S1. Clinical data used to verify APAP models in adults with low glutathione are summarized in Table S2 [20, 25–28].

Simulations of PK and Hepatic Biomarkers With APAP-ER and APAP-IR Overdoses

The verified DILIsym APAP models in adults with chronic alcohol use or low glutathione, as described above, were used to simulate PK and hepatic biomarkers (i.e., plasma ALT and INR) of APAP-ER and APAP-IR after single acute overdoses and repeat supratherapeutic ingestions (RSTI). Very importantly to note, simulations did not include clinical interventions (e.g., activated charcoal, lavage) and NAC treatment, such that the simulations represent the worst-case scenarios and many of the scenarios simulated (especially the extreme, high doses) are unlikely clinical scenarios. Despite this limitation, the outcomes for APAP-IR versus APAP-ER can still be compared effectively within each population and condition, which is the main purpose of the current work. Additional details about APAP dosing scenarios, characteristics of the SimPops, and simulation outputs are described below. Details about the representation of alcohol consumption patterns in MCAU and ECAU are described in Supporting Information S1.

Acute Overdose

Single acute overdoses of APAP-ER at 3.9, 9.75, 19.5, 32.5, 65, 78, and 100.1 g were simulated over 1 week, which represent the ingestion of 6, 15, 30, 50, 100, 120, and 154 ER caplets, respectively. The same doses of APAP IR were simulated for comparison of PK and the hepatic biomarkers between formulations.

Repeat (Supra)therapeutic Ingestion (R(S)TI)

APAP-ER and APAP-IR were simulated over 2 weeks for three 10-day repeat-ingestion scenarios:

• Scenario A (therapeutic, 3.9 g/day): APAP-ER two 650-mg caplets q8hr, or APAP-IR two 325-mg caplets q4hr
• Scenario B (supra-therapeutic, 5.2 g/day): APAP-ER two 650-mg caplets q6hr, or APAP-IR as four 325-mg caplets q6hr
• Scenario C (supra-therapeutic, 7.8 g/day): APAP-ER two 650-mg caplets q4hr, or APAP-IR four 325-mg caplets or 1.3 g q4hr

Results

Verification of APAP-IR and APAP-ER Models in Chronic Alcohol Users

A PBPK model of ethanol was developed and verified using clinical PK data. Simulated plasma concentration-time profiles of ethanol reasonably recapitulated clinical data; the simulated AUC and C_max values were within 0.81–1.12-fold of observed data (Supporting Information S1, Figure S2).

Simulated acute (i.e., inhibition of CYP2E1) and chronic (i.e., induction of CYP2E1) effects of alcohol on APAP PK were verified using clinical PK data. Simulations of a single dose of 20 mg/kg APAP-IR in ECAU SimPops predicted that concomitant ingestion of alcohol (a loading dose of 0.6 g/kg and maintenance doses of 0.16 g/kg hourly for 8 h) decreased urinary recovery of NAPQI by 45% via competitive inhibition of CYP2E1 while it caused minimal changes in urinary recovery of other metabolites, consistent with the clinical data reported by Critchley et al. (Figure 2A) [19]. Simulations of a single dose of 1 g APAP-IR in newly abstinent ECAU SimPops (without concomitant alcohol ingestion) predicted a 1.7-fold increase in NAPQI recovery in urine via induction of CYP2E1 compared to healthy adults, which was within clinically observed ranges (Figure 2B) [16, 20, 21].

[IMAGE OMITTED. SEE PDF]

It has been reported that repeat therapeutic ingestion of APAP led to no or only slight increases in ALT and no changes in functional biomarkers (i.e., TB, INR) in actively drinking MCAU and newly abstinent ECAU [22–24]. Simulations of repeat therapeutic ingestion of APAP-IR and APAP-ER in MCAU and newly abstinent ECAU SimPops predicted only modest, transient ALT elevations in a subset of simulated individuals without total bilirubin increase, consistent with clinical data (Figure 2C–H). Details regarding the verification of APAP models in chronic alcohol users are provided in Supporting Information S1 and Table S2.

Refinement and Verification of APAP-IR and APAP-ER Models in Individuals With Low Glutathione

Simulations of a single dose of 1 g APAP-IR in a baseline individual with low glutathione predicted increased urinary recovery of APAP-glucuronide and decreased urinary recovery of APAP and other metabolites. By contrast, clinical data showed no significant changes in urinary recovery of APAP and all metabolites despite a decrease in APAP oral clearance, suggesting that all the metabolic pathways were decreased to a similar extent in cirrhotic patients with moderate/severe impairment compared to healthy subjects [28]. Thus, the simulated individual with low glutathione was further refined to represent reduced APAP metabolism and renal clearance pathways to the same extent (i.e., 50% compared to healthy subjects to reproduce average 50% decrease in APAP oral clearance) to retain the relative ratio of metabolic flux and urinary recovery of APAP in its metabolites. Simulations of 1 g APAP-IR with the refined low glutathione individual reasonably recapitulated decreased APAP clearance and unaltered urinary recovery of APAP and metabolites observed in multiple APAP PK studies (Figure 3A,B). It has been reported that repeat therapeutic ingestion of APAP led to no changes in ALT and functional biomarkers (i.e., total bilirubin, INR) in cirrhotic patients with modest/severe hepatic impairment [25]. Simulations of APAP-IR 1 g QID (4 g/day) for 14 days in the low glutathione SimPops predicted no ALT increase in the majority of simulated individuals and no changes in functional biomarkers (i.e., total bilirubin, INR) in all simulated individuals, generally consistent with clinical data in patients with liver disease (Figure 3C,D) [25]. Details regarding the refinement and verification of APAP models in adults with low glutathione are provided in Supporting Information S1 and Table S2.

[IMAGE OMITTED. SEE PDF]

Simulations of PK and Hepatic Biomarkers With APAP-ER and APAP-IR Acute Overdoses

Simulated APAP PK profiles and hepatic biomarker responses after acute overdoses of APAP-ER and APAP-IR in MCAU, ECAU, and adults with low glutathione are presented in Figure 4 (averaged profiles), Figure S3 (population profiles), and Table 1. Simulated APAP PK profiles showed no apparent Rumack-Matthew treatment line crossing differences between APAP-ER and APAP-IR within each population. On average, APAP exposure after acute overdoses was predicted to be lower for APAP-ER compared to APAP-IR. AUC_∞ ER/IR ratio ranges were 0.85–0.98, 0.85–0.98, and 0.88–0.99 in MCAU, ECAU, and adults with low glutathione, respectively. Lower APAP exposure for APAP-ER compared to APAP-IR was attributed to a lower fraction absorbed (Fa) as presented in Table 1, which was informed by in vitro dissolution assays and mechanistic modeling of absorption using GastroPlus as previously reported [6].

[IMAGE OMITTED. SEE PDF]

TABLE 1 Comparison of simulated plasma APAP pharmacokinetic parameters and hepatic biomarkers after acute overdoses of APAP-ER and APAP-IR in MCAU, ECAU, and individuals with low glutathione.

Ingested Dose (g)	AUC∞ (μg∙h/mL)	Cmax (μg/mL)	Fa	ALTmax (U/L)	INRmax
APAP-ER	APAP-IR	APAP-ER	APAP-IR	APAP-ER	APAP-IR	APAP-ER	APAP-IR	APAP-ER	APAP-IR
Moderate chronic alcohol users
3.9	185.4 (23.3%)	190.2 (23.1%)	32.4 (19.7%)	45.9 (18.7%)	1.00 (1.00–1.00)	1.00 (1.00–1.00)	30.0 (30.0–30.1)	30.0 (30.0–30.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
9.75	509.3 (24.3%)	517.4 (24.4%)	83.5 (19.8%)	102.1 (19.2%)	0.97 (0.95–0.99)	1.00 (0.99–1.01)	153.7 (30.0–1469.5)	184.6 (30.0–1885.8)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
19.5	976.6 (25.7%)	1096.0 (25.9%)	153.6 (20.4%)	198.8 (19.2%)	0.87 (0.81–0.94)	0.97 (0.89–1.00)	2063.9 (30.5–8523.3)	2813.9 (40.3–8735.9)	1.5 (1.0–7. 8)	1.9 (1.0–7.9)
32.5	1528.7 (27.0%)	1792.5 (27.7%)	231.4 (20.6%)	308.2 (20.4%)	0.77 (0.69–0.87)	0.88 (0.74–0.96)	4762.6 (193.1–8861.8)	5327.3 (314.1–8733.8)	2.7 (1.0–7.9)	3.0 (1.0–7.9)
65	2682.3 (29.3%)	3085.2 (31.3%)	389.0 (20. 9%)	491.3 (23.0%)	0.61 (0.51–0.74)	0.70 (0.50–0.88)	5401.2 (1658.5–8629.2)	5186.5 (1631.3–8603.1)	2.5 (1.0–7.9)	2.3 (1.0–7.8)
78	3066.8 (30.1%)	3470.6 (32.5%)	440.0 (21.1%)	541.2 (23.9%)	0.57 (0.46–0.70)	0.64 (0.44–0.84)	5254.1 (2440.6–8755.0)	5047.4 (2105.7–8748.4)	2.3 (1.1–7.9)	2.1 (1.1–7.9)
100.1	3642.2 (31.4%)	4012.2 (34.2%)	514.5 (21.6%)	607.8 (25.1%)	0.51 (0.39–0.65)	0.56 (0.36–0.79)	5000.7 (2665.1–8670.9)	4860.3 (2612.1–8388.4)	2.0 (1.1–7.9)	1.9 (1.1–7.7)
Excessive chronic alcohol users
3.9	201.5 (24.2%)	205.7 (24.1%)	33.3 (19.6%)	46.5 (18.6%)	1.00 (1.00–1.00)	1.00 (1.00–1.00)	30.0 (30.0–30.0)	30.0 (30.0–30.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
9.75	553.0 (25.3%)	563.0 (25.4%)	85.2 (19.6%)	104.0 (19.0%)	0.97 (0.95–0.99)	1.00 (0.99–1.00)	41.1 (30.0–350.4)	45.3 (30.0–406.3)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
19.5	1054.7 (26.6%)	1181.6 (26.8%)	156.2 (20.2%)	201.6 (19.0%)	0.87 (0.81–0.94)	0.97 (0.89–0.99)	384.7 (30.0–6203.8)	589.8 (30.0–7645.7)	1.0 (1.0–3.5)	1.0 (1.0–6.2)
32.5	1636.5 (27.7%)	1912.1 (28.3%)	234.4 (20.4%)	311.5 (20.2%)	0.77 (0.69–0.87)	0.89 (0.74–0.97)	1813.0 (36.7–8533.6)	2655.6 (64.9–8783.7)	1.5 (1.0–7.7)	1.7 (1.0–7.8)
65	2825.7 (29.6%)	3234.1 (31.3%)	392.1 (20.8%)	495.1 (22.8%)	0.61 (0.51–0.74)	0.70 (0.50–0.88)	4955.8 (255.6–8679.2)	5271.6 (283.9–8810.3)	2.8 (1.0–7.9)	2.9 (1.0–7.9)
78	3215.2 (30.2%)	3620.6 (32.2%)	443.2 (21.0%)	545.1 (23.7%)	0.57 (0.46–0.70)	0.64 (0.44–0.84)	5325.0 (340.7–8852.7)	5325.3 (340.2–8631.5)	2.9 (1.0–7.9)	2.7 (1.0–7.8)
100.1	3792.5 (31.2%)	4158.9 (33.6%)	517.8 (21.5%)	611.8 (24.9%)	0.51 (0.39–0.65)	0.56 (0.36–0.79)	5408.3 (465.3–8650.8)	5295.0 (416.9–8743.4)	2.7 (1.0–7.9)	2.5 (1.0–7.9)
Individuals with low glutathione
3.9	411.2 (26.4%)	416.2 (26.3%)	40.4 (19.3%)	52.8 (18.5%)	1.00 (1.00–1.00)	1.00 (1.00–1.00)	30.0 (30.0–30.0)	30.0 (30.0–30.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
9.75	1147.4 (28.0%)	1176.4 (28.0%)	101.8 (19.5%)	121.1 (18.8%)	0.97 (0.95–0.99)	1.00 (0.99–1.00)	33.8 (30.0–184.0)	34.8 (30.0–196.8)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
19.5	2223.0 (29.4%)	2495.0 (29.6%)	185.0 (20.0%)	234.4 (18.9%)	0.87 (0.81–0.94)	0.96 (0.89–0.99)	188.5 (30.0–2058.1)	276.4 (30.0–3424.6)	1.0 (1.0–1.0)	1.0 (1.0–1.3)
32.5	3469.7 (30.5%)	3936.8 (30.2%)	275.7 (20.4%)	360.4 (19.9%)	0.77 (0.69–0.87)	0.89 (0.74–0.97)	810.2 (30.0–8274.0)	1323.5 (30.0–8573.4)	1.1 (1.0–7.1)	1.2 (1.0–7.8)
65	6020.0 (32.4%)	6892.9 (37.9%)	453.5 (21.0%)	570.7 (22.3%)	0.61 (0.51–0.74)	0.70 (0.50–0.88)	3200.5 (54.3–8686.7)	3923.9 (104.9–8894.3)	2.0 (1.0–7.9)	2.4 (1.0–7.9)
78	6855.6 (33.0%)	7725.9 (34.9%)	510.4 (21.2%)	628.1 (23.1%)	0.57 (0.46–0.70)	0.64 (0.44–0.84)	3859.1 (92.1–8811.6)	4430.3 (145.7–8929.9)	2.4 (1.0–7.9)	2.6 (1.0–7.9)
100.1	8092.2 (33.9%)	8879.4 (36.2%)	593.7 (21.5%)	705.1 (24.2%)	0.51 (0.39–0.65)	0.56 (0.36–0.79)	4610.3 (153.4–8986.7)	4877.2 (204.6–8785.2)	2.7 (1.0–7.9)	2.8 (1.0–7.8)

Similar ALT and INR profiles as well as peak ALT (ALT_max) and peak INR (INR_max) values were predicted per overdose level for APAP-ER and APAP-IR after acute overdoses within each population. The ER/IR ratios of all biomarker values across all overdose levels were within 0.73–1.10, 0.65–1.08, and 0.61–1.00 in MCAU, ECAU, and adults with low glutathione, respectively. Maximum biomarker values are not necessarily the highest for the highest overdose levels, which can be explained by an increasing number of early simulation terminations (i.e., simulated “death” which is defined as the fraction of viable hepatocytes < 0.15 in DILIsym) with increasing overdose levels in those instances. Details regarding early simulation terminations are provided in Supporting Information S1.

Simulations of PK and Hepatic Biomarkers With APAP-ER and APAP-IR Repeat Supratherapeutic Ingestion (RSTI)

Simulated APAP PK profiles and hepatic biomarker responses after 10 days of repeat doses of APAP-ER and APAP-IR in MCAU, ECAU, and adults with low glutathione are presented in Figure 5 (averaged profiles), Figure S4 (population profiles), and Table 2. No Rumack-Matthew treatment line crossing was observed for the three repeat dosing scenarios in MCAU and ECAU. Simulated APAP PK profiles showed no apparent Rumack-Matthew treatment line crossing differences between APAP-ER and APAP-IR in adults with low glutathione. On average, total steady-state APAP exposure (AUC_τ) as well as the accumulation ratio (based on PK after the first repeat dose and after repeat doses) were similar between APAP-ER and APAP-IR after repeat doses. Similar ALT and INR profiles as well as ALT_max and INR_max values were predicted per overdose level for APAP-ER and APAP-IR after repeat overdoses within each population. The ER/IR ratios of all biomarker values across all overdose levels were within 0.99–1.01, 1.00–1.00, and 1.00–1.01 in MCAU, ECAU, and adults with low glutathione, respectively. No early simulation terminations were predicted for the three repeat dosing scenarios in MCAU, ECAU, and adults with low glutathione.

[IMAGE OMITTED. SEE PDF]

TABLE 2 Comparison of simulated plasma APAP pharmacokinetic parameters and hepatic biomarkers after repeat dose of APAP-ER and APAP-IR in MCAU, ECAU, and individuals with low glutathione.

Steady-state parameter	Scenario A (3.9 g/day)	Scenario B (5.2 g/day)	Scenario C (7.8 g/day)
APAP-ER	APAP-IR	APAP-ER	APAP-IR	APAP-ER	APAP-IR
Moderate chronic alcohol users
AUCτ (μg∙h/mL)	51.8 (24.9%)	51.4a (24.9%)	53.2 (25.8%)	54.1 (25.5%)	56.6 (28.7%)	57.4 (28.5%)
Cmax, ss (μg/mL)	11.4 (21.2%)	10.4 (21.1%)	13.1 (22.5%)	17.9 (20.3%)	17.3 (26.1%)	22.4 (23.2%)
AccR	0.99 (0.86–1.18)	1.02 (0.86–1.27)	1.01 (0.86–1.27)	0.98 (0.85–1.22)	1.06 (0.86–1.63)	1.03 (0.85–1.57)
ALTmax (U/L)	38.4 (30.0–272.9)	37.9 (30.0–268.3)	52.9 (30.0–352.6)	53.3 (30.0–357.9)	236.6 (30.0–3401.2)	237.7 (30.0–3435.3)
INRmax	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.4)	1.0 (1.0–1.4)
Excessive chronic alcohol users
AUCτ (μg∙h/mL)	57.3 (26.2%)	56.8a (26.3%)	58.6 (27.1%)	59.5 (26.8%)	61.3 (28.4%)	62.0 (28.1%)
Cmax, ss (μg/mL)	12.2 (21.7%)	11.2 (21.8%)	14.1 (23.2%)	19.0 (20.7%)	18.6 (25.8%)	23.7 (22.9%)
AccR	1.01 (0.87–1.25)	1.04 (0.87–1.35)	1.03 (0.87–1.34)	1.00 (0.86–1.28)	1.07 (0.87–1.50)	1.04 (0.86–1.43)
ALTmax (U/L)	30.2 (30.0–69.0)	30.1 (30.0–63.9)	32.9 (30.0–191.7)	32.9 (30.0–192.7)	45.3 (30.0–308.4)	45.3 (30.0–308.1)
INRmax	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)
Individuals with low glutathione
AUCτ (μg∙h/mL)	122.8 (29.4%)	122.7a (29.4%)	127.4 (30.2%)	128.0 (30.1%)	134.7 (31.2%)	135.2 (31.1%)
Cmax, ss (μg/mL)	21.0 (24.4%)	19.8 (25.0%)	26.1 (26.5%)	31.2 (23.8%)	37.3 (29.1%)	42.8 (26.5%)
AccR	1.02 (0.87–1.36)	1.07 (0.86–1.50)	1.05 (0.87–1.47)	1.01 (0.85–1.39)	1.11 (0.88–1.64)	1.07 (0.86–1.54)
ALTmax (U/L)	32.6 (30.0–170.7)	32.6 (30.0–170.6)	37.2 (30.0–281.9)	37.1 (30.0–281.1)	56.5 (30.0–377.1)	56.1 (30.0–376.5)
INRmax	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)

Discussion

QST models use a mechanistic, mathematical approach to integrate physiology, drug properties, and biochemical processes to evaluate drug safety. In this study, QST modeling using DILIsym was employed to evaluate the PK exposure and hepatic biomarker responses after overdoses of APAP-IR and APAP-ER in moderate and excessive chronic alcohol users and adults with low hepatic glutathione levels. Simulations predicted that APAP-ER had similar PK and hepatic biomarker profiles compared to APAP-IR in these populations after single acute overdoses and repeat supratherapeutic ingestions.

Chronic alcohol users reportedly have reduced glutathione after excessive consumption and induced CYP2E1 after moderate and excessive consumption [13–15, 17, 29]. However, ethanol has been shown to competitively inhibit CYP2E1 [18], which might offset the effect of CYP2E1 induction. Therefore, both chronic and acute effects of alcohol on the formation and elimination of NAPQI, the toxic reactive metabolite of APAP, have to be considered when evaluating the susceptibility of APAP-mediated hepatotoxicity in chronic alcohol users. Mathematical, mechanistic modeling provides an effective framework to quantitatively predict the net effects of alcohol on these competing and alternating pathways. A previously reported mathematical model suggested that the risk of APAP-induced hepatotoxicity (represented as decrease in fraction viable hepatocytes) after acute overdose was increased if APAP was ingested shortly after alcohol is eliminated from the body in chronic alcohol users with CYP2E1 induction, whereas concurrent consumption of alcohol showed protective effects [30]. While this work provided a quantitative assessment of alcohol effects on hepatotoxicity induced by acute overdose of APAP, the authors did not explore the combined effects of chronic and acute alcohol effects nor represent the alcohol effect on hepatic glutathione. Also, they did not represent specific APAP formulations and clinically relevant outcomes such as hepatic biomarkers. In the current study, APAP-IR and APAP-ER formulations were specifically represented in simulated chronic alcohol users who represent both acute and chronic effects of alcohol on APAP metabolizing enzymes and hepatic glutathione to compare clinically relevant outcomes (i.e., PK and hepatic biomarkers such as plasma ALT and INR) for two formulations.

CYP2E1 enzyme degrades rapidly after alcohol-mediated induction if alcohol consumption is discontinued. It has been reported that CYP2E1 levels were normalized within 3–8 days in newly abstinent heavy drinkers (i.e., ECAU) [14]. As such, newly abstinent ECAU, who still have induced CYP2E1 but do not actively consume alcohol that can directly inhibit CYP2E1, are expected to be the most susceptible to NAPQI-mediated hepatotoxicity among chronic alcohol users. Retrospective studies and case reports suggested that alcohol use in association with APAP overdose may be a risk factor for hepatotoxicity [31, 32]. However, a systematic review of randomized controlled trials demonstrated that repeat therapeutic doses (≤ 4 g/day) of APAP-IR and APAP-ER did not predispose chronic alcohol users, including newly abstinent heavy drinkers who were presumed to have increased risk, to liver injury [28]. Consistent with these data, simulations of repeat therapeutic dosing (4 g/day) of APAP-IR and APAP-ER formulations in newly abstinent ECAU predicted minimal hepatic injury. Simulation outcomes suggested that while newly abstinent ECAU have increased NAPQI synthesis and reduced hepatic glutathione levels, hepatocytes still have sufficient hepatic glutathione to eliminate NAPQI at therapeutic doses of APAP.

Glutathione is a reduced-thiol and tripeptide molecule concentrated in hepatocytes that helps maintain a balanced intracellular redox state by preferentially binding electrophilic molecules such as NAPQI [33]. As such, individuals with reduced hepatic glutathione due to underlying liver disease, malnutrition, and aging may be predisposed to APAP-mediated hepatotoxicity after overdose [34]. The glutathione sub-model within DILIsym represents glutathione synthesis from its precursor, steady-state turnover, depletion, and nuclear factor-like 2 (NRF2)-mediated upregulation of glutathione synthesis [8], which allows for evaluation of APAP-induced hepatotoxicity in individuals with reduced glutathione. In the current work, APAP-ER and APAP-IR representations in adults with low glutathione were developed and verified using data from cirrhotic patients in terms of their hepatic glutathione levels and expression of APAP metabolizing enzymes.

Cirrhotic patients with modest-to-severe hepatic impairment have reduced hepatic glutathione as well as CYP2E1 and other APAP metabolizing enzymes [16, 17, 35–40]. Interestingly, although cirrhotic patients showed increased APAP exposure and reduced systemic clearance compared to healthy adults, the overall biotransformation (e.g., percent urinary recovery of each metabolite) remained unchanged compared to healthy adults [20, 26–28]. In our simulations, the amount of NAPQI formed (i.e., urinary recovery) was similar in healthy adults and individuals with low glutathione, while the formation rate of NAPQI (and other metabolites) was reduced in individuals with low glutathione compared to healthy adults, consistent with clinical data. Simulations of repeat therapeutic dosing (4 g/day) of APAP predicted minimal ALT elevations and no functional impairment in cirrhotic adults with low glutathione, which is consistent with data from prospective clinical studies where stable chronic liver disease does not pose a risk for developing APAP hepatotoxicity at therapeutic doses [25]. Quantitative modeling representing the dynamics of NAPQI formation and elimination, and glutathione dynamics, demonstrated that while cirrhotic patients have a reduced level of hepatic glutathione, the rate of NAPQI formation is also reduced, which led to sufficient NAPQI elimination at a therapeutic dose of APAP-IR and APAP-ER.

The effect of the APAP formulation on its overdose hepatotoxicity in adults with chronic alcohol use and low hepatic glutathione has not been fully characterized. Retrospective studies and case reports are informative but do not provide firm conclusions due to multiple methodological weaknesses and the existence of confounding factors which are not well documented. Also, reliance on the patient's self-reported dose, timing, and formulation introduces great uncertainties. For ethical reasons, it is unlikely that any large randomized controlled trial will be undertaken to understand formulation-specific hepatotoxicity outcomes after APAP overdose in chronic alcohol users (or any patient groups). In the absence of controlled trial data, such information should be extrapolated from a combination of existing PK and hepatic response data by quantitative mechanistic modeling. In our prior work, PK and hepatic response in healthy volunteers after therapeutic dose and overdose (acute and repeat-supratherapeutic ingestions) were developed and verified using clinical data. In the current work, representation of altered physiology in chronic alcohol users and individuals with low hepatic glutathione was verified using APAP PK and hepatic response data after repeated therapeutic doses. Integrating APAP dose-responses and population-specific physiology, modeling was able to evaluate the PK and hepatic biomarker responses to APAP-IR and APAP-ER after acute and repeat supratherapeutic overdoses in special populations.

The modeling and simulation work presented in this study have several assumptions and limitations. First, models and subsequent simulations described in this report did not include concomitant clinical interventions or NAC treatment. Interventions used in the clinic to prevent APAP overdose-associated hepatotoxicity include treatments such as activated charcoal and gastric lavage, each of which has the potential to diminish the fraction of APAP that is absorbed [41–43]. Furthermore, the simulations did not include the administration of the glutathione precursor NAC, an important and standard therapy, after an acute APAP overdose, which decreases APAP hepatotoxicity by restoring glutathione stores and detoxifying the toxic NAPQI metabolite. As a result, APAP hepatotoxicity predictions (e.g., biomarker levels) would very likely have been decreased substantially in the presence of NAC [10].

Second, simulated chronic alcohol users and adults with low glutathione did not represent underlying liver injury. Thus, simulated hepatic effects from individuals with low glutathione do not necessarily represent other causes of underlying liver injury; hepatic biomarker change from the baseline/pretreatment value was assessed to determine APAP-mediated effects. As such, the simulated hepatic biomarker responses represent APAP effects only in stable liver disease and cannot be extrapolated to decompensated liver disease. Lastly, specific alcohol consumption patterns were assumed for the purpose of representing direct competitive inhibition of CYP2E1 by alcohol in chronic alcohol users. The extent of competitive inhibition depends on alcohol exposure, which varies depending on the amount of alcohol consumed and the relative timing of alcohol consumption and APAP ingestion. Despite these limitations, the outcomes for APAP-IR versus APAP-ER can still be compared effectively within each population and condition, as the current modeling and simulations were not meant for cross-population comparisons for the same formulation.

In summary, QST modeling using DILIsym predicted similar PK and hepatic biomarker profiles for the APAP-ER and APAP-IR formulations in moderate and excessive chronic alcohol users and adults with low hepatic glutathione levels after single acute overdoses and repeat supratherapeutic ingestions. These modeling results further support that approaches to manage APAP-IR overdoses can be applied to manage APAP-ER overdoses, as indicated in the recently published consensus statement on the APAP overdose management [5].

Author Contributions

K.Y., J.J.B., B.A.H., J.M., E.A., J.C.K.L., C.K.G., S.S., and E.A. wrote the manuscript; K.Y., J.J.B., B.A.H., J.M., E.A., J.C.K.L., C.K.G., S.S., and E.A. designed the research; K.Y., J.J.B., J.M., and E.A. performed the research; K.Y., J.J.B., J.M., and E.A. analyzed the data.

Disclosure

The authors have nothing to report.

Conflicts of Interest

Kyunghee Yang, James J. Beaudoin, Brett A. Howell, and James Mullin are employees of Simulations Plus Inc. John C.K. Lai, Sury Sista, and Evren Atillasoy are employees of Kenvue. Cathy K. Gelotte is a consultant of Kenvue. Elham Amini has no competing interests to declare.

References

A. Ramachandran and H. Jaeschke, “Acetaminophen Toxicity: Novel Insights Into Mechanisms and Future Perspectives,” Gene Expression 18 (2018): 19–30.

A. Mutsaers, J. P. Green, M. L. A. Sivilotti, et al., “Changing Nomogram Risk Zone Classification With Serial Testing After Acute Acetaminophen Overdose: A Retrospective Database Analysis,” Clinical Toxicology (Philadelphia, Pa.) 57, no. 6 (2019): 380–386.

B. H. Rumack and H. Matthew, “Acetaminophen Poisoning and Toxicity,” Pediatrics 55 (1975): 871–876.

H. Salmonson, G. Sjöberg, and J. Brogren, “The Standard Treatment Protocol for Paracetamol Poisoning May Be Inadequate Following Overdose With Modified Release Formulation: A Pharmacokinetic and Clinical Analysis of 53 Cases,” Clinical Toxicology (Philadelphia, Pa.) 56, no. 1 (2018): 63–68.

R. C. Dart, M. E. Mullins, T. Matoushek, et al., “Management of Acetaminophen Poisoning in the US and Canada: A Consensus Statement,” JAMA Network Open 6 (2023): e2327739.

J. J. Beaudoin, K. Yang, B. A. Howell, et al., “Modeling and Simulation of Acetaminophen Pharmacokinetics and Hepatic Biomarkers After Overdoses of Extended‐Release and Immediate‐Release Formulations in Healthy Adults Using the Quantitative Systems Toxicology Software Platform DILIsym,” CPT: Pharmacometrics & Systems Pharmacology 14 (2025): 681–694, https://doi.org/10.1002/psp4.13304.

P. B. Watkins, “Quantitative Systems Toxicology Approaches to Understand and Predict Drug‐Induced Liver Injury,” Clinics in Liver Disease 24 (2020): 49–60.

B. A. Howell, Y. Yang, R. Kumar, et al., “In Vitro to In Vivo Extrapolation and Species Response Comparisons for Drug‐Induced Liver Injury (DILI) Using DILIsym: A Mechanistic, Mathematical Model of DILI,” Journal of Pharmacokinetics and Pharmacodynamics 39 (2012): 527–541.

B. A. Howell, S. Q. Siler, and P. B. Watkins, “Use of a Systems Model of Drug‐Induced Liver Injury (DILIsym()) to Elucidate the Mechanistic Differences Between Acetaminophen and Its Less‐Toxic Isomer, AMAP, in Mice,” Toxicology Letters 226 (2014): 163–172.

J. L. Woodhead, B. A. Howell, Y. Yang, et al., “An Analysis of N‐Acetylcysteine Treatment for Acetaminophen Overdose Using a Systems Model of Drug‐Induced Liver Injury,” Journal of Pharmacology and Experimental Therapeutics 342 (2012): 529–540.

J. R. Senior, “Alanine Aminotransferase: A Clinical and Regulatory Tool for Detecting Liver Injury‐Past, Present, and Future,” Clinical Pharmacology and Therapeutics 92 (2012): 332–339.

P. Clinical, “Guidance for Industry Drug‐Induced Liver Injury: Premarketing Clinical Evaluation Guidance for Industry Drug‐Induced Liver Injury: Premarketing Clinical Evaluation,” 2009.

S. Liangpunsakul, D. Kolwankar, A. Pinto, J. C. Gorski, S. D. Hall, and N. Chalasani, “Activity of CYP2E1 and CYP3A Enzymes in Adults With Moderate Alcohol Consumption: A Comparison With Nonalcoholics,” Hepatology 41 (2005): 1144–1150.

C. M. Oneta, C. S. Lieber, J. J. Li, et al., “Dynamics of Cytochrome P4502E1 Activity in Man: Induction by Ethanol and Disappearance During Withdrawal Phase,” Journal of Hepatology 36 (2002): 47–52.

C. Girre, D. Lucas, E. Hispard, C. Menez, S. Dally, and J. F. Menez, “Assessment of Cytochrome P4502E1 Induction in Alcoholic Patients by Chlorzoxazone Pharmacokinetics,” Biochemical Pharmacology 47 (1994): 1503–1508.

B. H. Lauterburg and M. E. Velez, “Glutathione Deficiency in Alcoholics: Risk Factor for Paracetamol Hepatotoxicity,” Gut 29 (1988): 1153–1157.

E. Altomare, G. Vendemiale, and O. Albano, “Hepatic Glutathione Content in Patients With Alcoholic and Non Alcoholic Liver Diseases,” Life Sciences 43 (1988): 991–998.

K. Thummel, “Ethanol and Production of the Hepatotoxic Metabolite of Acetaminophen in Healthy Adults,” Clinical Pharmacology & Therapeutics 67, no. 6 (2000): 591–599.

J. A. Critchley, E. H. Dyson, A. W. Scott, D. R. Jarvie, and L. F. Prescott, “Is There a Place for Cimetidine or Ethanol in the Treatment of Paracetamol Poisoning?,” Lancet 1 (1983): 1375–1376.

J. P. Villeneuve, G. Raymond, J. Bruneau, L. Colpron, and G. Pomier‐Layrargues, “Pharmacokinetics and Metabolism of Acetaminophen in Normal, Alcoholic and Cirrhotic Subjects,” Gastroentérologie Clinique et Biologique 7 (1983): 898–902.

M. H. Skinner, R. Matano, W. Hazle, and T. F. Blaschke, “Acetaminophen Metabolism in Recovering Alcoholics,” Methods and Findings in Experimental and Clinical Pharmacology 12 (1990): 513–515.

K. Heard, J. L. Green, J. E. Bailey, G. M. Bogdan, and R. C. Dart, “A Randomized Trial to Determine the Change in Alanine Aminotransferase During 10 Days of Paracetamol (Acetaminophen) Administration in Subjects Who Consume Moderate Amounts of Alcohol,” Alimentary Pharmacology & Therapeutics 26 (2007): 283–290.

R. C. Dart, J. L. Green, E. K. Kuffner, K. Heard, B. Sproule, and B. Brands, “The Effects of Paracetamol (Acetaminophen) on Hepatic Tests in Patients Who Chronically Abuse Alcohol—A Randomized Study: Effects of Paracetamol on Hepatic Tests in Chronic Alcohol Abusers,” Alimentary Pharmacology & Therapeutics 32 (2010): 478–486.

S. Bartels, M. Sivilotti, D. Crosby, and J. Richard, “Are Recommended Doses of Acetaminophen Hepatotoxic for Recently Abstinent Alcoholics? A Randomized Trial,” Clinical Toxicology (Philadelphia, Pa.) 46, no. 3 (2008): 243–249.

G. D. Benson, “Acetaminophen in Chronic Liver Disease,” Clinical Pharmacology and Therapeutics 33 (1983): 95–101.

P. Zapater, R. Moreu, and J. F. Horga, “The Diagnosis of Drug‐Induced Liver Disease,” Current Clinical Pharmacology 1, no. 2 (2006): 207–217.

J. A. Forrest, J. A. Clements, and L. F. Prescott, “Clinical Pharmacokinetics of Paracetamol,” Clinical Pharmacokinetics 7, no. 2 (1982): 93–107.

K. L. Hayward, E. E. Powell, K. M. Irvine, and J. H. Martin, “Can Paracetamol (Acetaminophen) be Administered to Patients With Liver Impairment?,” British Journal of Clinical Pharmacology 81 (2016): 210–222.

S. A. Jewell, D. di Monte, A. Gentile, A. Guglielmi, E. Altomare, and O. Albano, “Decreased Hepatic Glutathione in Chronic Alcoholic Patients,” Journal of Hepatology 3 (1986): 1–6.

A. Ghosh, I. Berger, C. H. Remien, and A. Mubayi, “The Role of Alcohol Consumption on Acetaminophen Induced Liver Injury: Implications From a Mathematical Model,” Journal of Theoretical Biology 519 (2021): 110559.

G. P. Bray, C. Mowat, D. F. Muir, J. M. Tredger, and R. Williams, “The Effect of Chronic Alcohol Intake on Prognosis and Outcome in Paracetamol Overdose,” Human & Experimental Toxicology 10 (1991): 435–438.

N. Pholmoo and C. Bunchorntavakul, “Characteristics and Outcomes of Acetaminophen Overdose and Hepatotoxicity in Thailand,” Journal of Clinical and Translational Hepatology 7 (2019): 132–139.

S. C. Lu, “Regulation of Glutathione Synthesis,” Molecular Aspects of Medicine 30 (2009): 42–59.

S. S. Kalsi, P. I. Dargan, W. S. Waring, and D. M. Wood, “A Review of the Evidence Concerning Hepatic Glutathione Depletion and Susceptibility to Hepatotoxicity After Paracetamol Overdose,” Open Access Emergency Medicine: OAEM 3 (2011): 87–96.

K. Dilger, J. Metzler, J. C. Bode, and U. Klotz, “CYP2E1 Activity in Patients With Alcoholic Liver Disease,” Journal of Hepatology 27 (1997): 1009–1014.

J. George, K. Byth, and G. C. Farrell, “Age but Not Gender Selectively Affects Expression of Individual Cytochrome P450 Proteins in Human Liver,” Biochemical Pharmacology 50 (1995): 727–730.

F. P. Guengerich and C. G. Turvy, “Comparison of Levels of Several Human Microsomal Cytochrome P‐450 Enzymes and Epoxide Hydrolase in Normal and Disease States Using Immunochemical Analysis of Surgical Liver Samples,” Journal of Pharmacology and Experimental Therapeutics 256 (1991): 1189–1194.

R. Frye, N. K. Zgheib, G. R. Matzke, et al., “Liver Disease Selectively Modulates Cytochrome P450–Mediated Metabolism,” Clinical Pharmacology and Therapeutics 80 (2006): 235–245.

E. A. Sotaniemi, A. Rautio, M. Bäckstrom, P. Arvela, and O. Pelkonen, “CYP3A4 and CYP2A6 Activities Marked by the Metabolism of Lignocaine and Coumarin in Patients With Liver and Kidney Diseases and Epileptic Patients,” British Journal of Clinical Pharmacology 39 (1995): 71–76.

E. B. Yalcin, V. More, K. L. Neira, et al., “Downregulation of Sulfotransferase Expression and Activity in Diseased Human Livers,” Drug Metabolism and Disposition 41 (2013): 1642–1650.

P. J. Yeates and S. H. Thomas, “Effectiveness of Delayed Activated Charcoal Administration in Simulated Paracetamol (Acetaminophen) Overdose,” British Journal of Clinical Pharmacology 49, no. 1 (2000): 11–14.

G. Levy and J. B. Houston, “Effect of Activated Charcoal on Acetaminophen Absorption,” Pediatrics 58 (1976): 432–435.

M. Davis, C. J. Simmons, N. G. Harrison, and R. Williams, “Paracetamol Overdose in Man: Relationship Between Pattern of Urinary Metabolites and Severity of Liver Damage,” Quarterly Journal of Medicine 45 (1976): 181–191.