APAP Hepatotoxicity After Therapeutic Doses

Paracetamol (acetaminophen, APAP) is a ubiquitous painkiller and antipyretic available worldwide in numerous over-the-counter and prescription medications. At supra-therapeutic doses, APAP is a well described hepatotoxic agent and a significant public health concern since 30'000 hospitalizations are estimated to be related to APAP drug induced liver injury (DILI) each year in the US. APAP is indeed the commonest cause of acute liver failure (ALF) in the US and Europe with an estimated overall mortality of 28%. Half APAP overdoses are unintentional and the poor ability of patients to identify products with APAP has been documented. In hospitals, 1% all drug prescriptions have contained an overdosed APAP prescription highlighting the need for increased awareness on unintentional APAP overdose.

At the recommended therapeutic dose of 4g/d, APAP is usually considered safe. However, associations between APAP therapeutic dosing and alanine aminotransferase (ALT) elevations have been demonstrated. Indeed, up to 1/3 healthy volunteers treated with therapeutic doses of APAP experienced 3x ALT elevation (up to 14x) after 3 days of treatment for up to 11 days. These results were confirmed in non-drinker patients. The multinational case-population SALT study reported 81 cases of ALF (49 in France) leading to transplantation after non-overdose APAP exposure in a 3 year follow-up and non-overdose APAP was associated with a 2x higher rate of ALF than NSAIDs. In the US, 17% of APAP induced ALF over a 41 month period were reported with an APAP dose inferior to 4g/day.

Malnutrition as a risk factor Some authors have suggested that therapeutic doses of APAP may be hepatotoxic in the presence of malnutrition and low body weight as well as chronic alcohol consumption and drugs inducing cytochromes P450 (CYPs). As glutathione (GSH) is synthesized from 3 amino-acids (aa)(Cys, Glu and Gly), protein or aa deficiency may result in GSH depletion. In rats, fasting was associated with increased APAP hepatotoxicity as a consequence of hepatic GSH decrease, and a 16h fasting period was sufficient to deplete GSH stores. GSH levels have been shown to be reduced in anorexic female patients as compared to controls and a positive correlation between GSH levels and BMI was observed. A small retrospective study (n=10) showed that severe hepatotoxicity after moderate APAP dose (4 to 10g/day) was preceded in 80% of cases by malnutrition. Several case reports of severe hepatotoxicity (some fatal) after therapeutic APAP doses in malnourished adults have been published. In the pediatric population, ALF cases were reported after therapeutic APAP doses after viral infection and low nutrition status. However, no well conducted study has aimed to prospectively assess the impact of malnutrition on the toxicity to therapeutic doses of APAP.

FDA experts pointed out that APAP is frequently used in cachexia patients. Cachexia is a complex syndrome characterized by several homeostatic perturbations including progressive involuntary weight loss, accompanied by wasting, early satiety, weakness and anorexia. A broad spectrum of clinical disease is associated with cachexia. The prevalence of malnutrition has been observed in 20-29% of hospitalized patients in Europe, and 33% in surgical wards. Pickering et al. demonstrated that APAP metabolism shifts toward the toxic oxidative pathways after major aortic surgery suggesting that those patients may be particularly susceptible. Drug consumption data also indicates that APAP utilization is high is post-operative settings. The British National Formulary has recommended a maximal dose of 60mg/kg IV APAP for use in adults whose weight is less than 50kg and 3g/day IV in chronic malnutrition or dehydration. Surprisingly, no such recommendation is available for oral APAP and which is commonly prescribed independently of the nutritional status. It is likely that cases of liver damage secondary to normal recommended dose are under-represented as the dose is not perceived as excessive and not described as such in international guidelines for pain management. Considering the widespread use of APAP and the prevalence of malnutrition in surgery wards, it is of crucial clinical importance to clarify whether malnutrition predispose to APAP induced hepatotoxicity at the recommended dosage of 4g/day.

Biomarker of APAP hepatotoxicity:

APAP is metabolized by glucuronidation (55%) and sulfatisation (40%). The metabolites of APAP are excreted out of the liver by multidrug resistance-associated proteins (36-38). APAP is oxidized (5%) by cytochromes P450 (CYP) 2E1, 3A and 1A2 to the highly reactive N-acetyl-p-benzo-quinone (NAPQI), electrophilic and cytotoxic metabolite, responsible for APAP liver toxicity.

At therapeutic dose, APAP is usually rapidly detoxified by conjugation with GSH, cleared from the liver and excreted in urine. Slattery et al. have shown that GSH depletion begins over the range of 0.5-3g APAP. After excessive APAP intake, both sulfonation and glucuronidation pathways become saturated in favor of the oxidation pathway. This results in the formation of large amounts of NAPQI and liver GSH depletion. NAPQI covalently binds to macromolecules, reacting with sulfur groups in hepatic proteins, and is responsible for the histopathological hepatic centrilobular necrosis with periportal sparing. N-acetylcysteine (NAC) is a scavenger for NAPQI. Within 24h of a single acute ingestion, APAP plasma concentrations are used to predict the likelihood of hepatotoxicity and the need for NAC antidote. However the normogram is not relevant to patients presenting later than 24h after ingestion or after a chronic ingestion. Developing and validating an early and easily accessible marker of hepatotoxicity would especially be useful in higher risk and fragile populations to improve diagnosis and management of APAP induced hepatotoxicity. Indeed, unrecognized cases of APAP hepatotoxicity carry a poor prognosis as antidote administration will not be instituted or delayed. APAP covalently binds to protein as a result of a reaction between NAPQI and cysteine residues to produce APAP-CYS protein adducts. In APAP related ALF, peak concentrations of APAP-CYS adducts were shown to correlate with peak aminotransferase concentrations and detected up to 12 days post-ingestion. In 157 adolescents and children with APAP overdose, peak APAP-CYS adducts correlated with peak hepatic transaminases, time to treatment with NAC and risk determination using the normogram. It has also been demonstrated that APAP-CYS concentrations varied according to the degree of exposure and APAP-CYS is specific for APAP exposures. However, no direct detection of APAP-CYS on full-length proteins or long polypeptides was performed. Detection was carried out after digestion with a non-specific endopeptidase. Thus, the exact adduct position and the identity of modified proteins is unknown.

The binding of chemical substances on hemoglobin and plasma albumin is a well-known phenomenon and most chemicals acting via reactive metabolites form such adducts. Hemoglobin and albumin adducts are easily accessible from a blood sample and have a well-defined life span due to the absence of repair. Recently, direct protein analysis protocols relying on LC-MS/MS analysis of intact globin chains or peptides were described for the analysis of hemoglobin and albumin adducts. The clinical utility of APAP-CYS modified albumin and hemoglobin as biomarkers of APAP hepatotoxicity will be evaluated and compared to the serum level of transaminases, APAP, APAP-CYS and metabolites.

Microvesicles (MVs) have progressively emerged as potential fruitful biomarker holders. MVs are circulating vesicles released from almost all cell types, and are composed of a huge variety of biomolecules such as messenger RNAs, micro RNAs, proteins and lipids. Interestingly, their composition is related to their original cell, tissue or organ, and is influenced by stimulation and micro-environmental changes of the donor cell. This gives them "signatures" of a physiological state. MVs are rapidly released in the blood after a stimuli or a change of condition and are thus potential early indicators of a physiological state, containing precious information for the monitoring of pathologies. Specific molecules released directly in the blood via MVs from hepatocytes, could contain interesting biomarker candidates to improve patient treatment management following APAP intoxication. Quantitative proteomics strategies will be used to isolate new protein biomarkers for APAP-induced hepatotoxicity from MVs. Alternatively, circulating micro-RNAs have been shown to be powerful potential biomarkers for a variety of diseases including hepatotoxicity following APAP-overdoses. Micro-RNAs are small ~22 nt long non-coding regulatory molecules affecting the expression levels of hundreds of genes that are stable in circulating fluids. In cases of APAP-overdoses, several studies have shown that the plasmatic concentration of the liver specific miR122 correlates and even slightly precedes the increase in blood levels of the classic hepatotoxicity markers. Transcriptomic screening from plasma and MVs will be used to identify potential new miRNA candidates for APAP induced liver toxicity.

Detection of new biomarkers can be a tedious process that can be facilitated by using extreme samples as filters to identify the most relevant candidates whether on protein or nucleic acid level. In the context of this study, extreme samples will be provided by 6 patients arriving at hospital after ingestion of an APAP overdose.

Genetic marker of susceptibility Gene polymorphisms in drug metabolizing enzymes (DME) and transporters involved in the pharmacokinetics of APAP might be used as biomarkers of susceptibility to APAP liver toxicity. The principle routes of elimination of APAP are phase II DMEs UDP-glucuronosyltransferases (UGT) and sulfotransferases (SULT). Three UGT isoforms appear to be involved in APAP glucuronoconjugation and up to 15x variability in APAP glucuronoconjugation has been demonstrated. UGT1A1*28 and *6 are associated with reduced enzymatic activity and increased irinotecan toxicity. In animals, UGT deficient Gunn rats had an increased susceptibility to APAP as compared to controls. Gilbert syndrome is a hereditary hyperbilirubinemia due to UGT1A1*28 leading to a 40% enzymatic activity reduction, reduced APAP glucuronidation and increased active metabolite production. Sulfatisation of APAP is catalyzed in human by SULT isoforms that can have up to 50x difference in activity. However the impact of SULT polymorphisms on APAP toxicity is unknown. Bioactivation of APAP into NAPQI is mediated by the CYP family wiht CYP2E1 and 2D6 appearing as the most relevant isoforms. Both are highly polymorphic and can undergo gene duplication. Cyp2e1knockout mice are less sensitive to APAP hepatotoxic effects than wild-type animals. . In humans, the impact of CYP2D6 and 2E1 polymorphisms on APAP toxicity is unknown. NAPQI is detoxified by GSTP1 in the liver for which two common single nucleotide polymorphisms (SNPs) have been described, one of them reducing enzymatic activity. Finally, up-regulation of efflux transporters has been described after toxic APAP ingestion. This project will aim to characterize in vitro the metabolic pathways involved in APAP metabolism and reactive metabolite formation, as well as their impact on metabolites production.