Preclinical liver toxicity models: Advantages, limitations and recommendations.

Experimental animal models are crucial for elucidating the pathophysiology of liver injuries and for assessing new hepatoprotective agents. Drugs and chemicals such as acetaminophen, isoniazid, valproic acid, ethanol, carbon tetrachloride (CCl), dimethylnitrosamine (DMN), and thioacetamide (TAA) are metabolized by the CYP2E1 enzyme, producing hepatotoxic metabolites that lead to both acute and chronic liver injuries. In experimental settings, acetaminophen (centrilobular necrosis), carbamazepine (centrilobular necrosis and inflammation), sodium valproate (necrosis, hydropic degeneration and mild inflammation), methotrexate (sinusoidal congestion and inflammation), and TAA (centrilobular necrosis and inflammation) are commonly used to induce various types of acute liver injuries. Repeated and intermittent low-dose administration of CCl, TAA, and DMN activates quiescent hepatic stellate cells, transdifferentiating them into myofibroblasts, which results in abnormal extracellular matrix production and fibrosis induction, more rapidly with DMN and CCL than TAA (DMN > CCl > TAA). Regarding toxicity and mortality, CCl is more toxic than DMN and TAA (CCl > DMN > TAA). Models used to induce metabolic dysfunction-associated liver disease (MAFLD) vary, but MAFLD's multifactorial nature driven by factors like obesity, fatty liver, dyslipidaemia, type II diabetes, hypertension, and cardiovascular disease makes it challenging to replicate human metabolic dysfunction-associated steatohepatitis accurately. From an experimental point of view, the degree and pattern of liver injury are influenced by various factors, including the type of hepatotoxic agent, exposure duration, route of exposure, dosage, frequency of administration, and the animal model utilized. Therefore, there is a pressing need for standardized protocols and regulatory guidelines to streamline the selection of animal models in preclinical studies.