Comparative Analysis of Classical and Emerging Experimental Animal Models of Liver Injury: A Comprehensive Review

Liver injury remains a significant global health concern with diverse etiologies including drug-induced hepatotoxicity, chemical exposure, and metabolic disorders. Despite decades of research using animal models of liver injury, several critical gaps remain unaddressed in the literature. First, while individual models have been well described, there is no comprehensive, comparative framework that simultaneously evaluates classical hepatotoxin models, dietary models, and emerging systems (e.g., humanized mice, inflammation-sensitized models, organoids) side-by-side. This fragmentation forces researchers to consult multiple disparate sources, leading to inconsistent model selection and reduced reproducibility. Second, the translational validity of these models-particularly for idiosyncratic drug-induced liver injury (iDILI), non-alcoholic steatohepatitis (NASH), and herb-induced liver injury (HILI)-remains poorly benchmarked against human disease. Third, there is no evidence-based decision framework that guides researchers toward the most appropriate model for a given mechanistic or therapeutic question (e.g., "Which model best recapitulates mitochondrial hepatotoxicity?" or "How can I model immune-mediated DILI?"). Fourth, recent advances in omics technologies (single-cell RNA sequencing, spatial transcriptomics, proteomics) and humanized platforms have not been systematically integrated into traditional model comparisons. This review aims to close these four gaps by providing a side-by-side comparative analysis of 18 distinct animal models across 12 parameters (induction method, injury pattern, mechanism, species/strain differences, time course, regenerative response, advantages, limitations, translational relevance, cost, throughput, and ethical considerations), mapping each model to its closest human disease correlate with explicit discussion of similarities and limitations, proposing a practical, decision-tree-based model selection framework and integrating emerging technologies (humanized mice, CRISPR screens, organoids) into the classical landscape. Understanding the distinctive features of each model is essential for translational relevance and reproducibility in hepatotoxicity research.