Evaluating the use of rodents as in vitro, in vivo and ex vivo experimental models for the assessment of tyrosine kinase inhibitor-induced cardiotoxicity: a systematic review.

Tyrosine kinase inhibitors (TKIs) are widely used in cancer therapy yet are strongly associated with acute and chronic cardiotoxicity in patients. There is a critical need to advance our understanding of the pathophysiology that underlies TKI-mediated cardiotoxicity, and central to this is the use of reproducible and relevant preclinical models, which are employed in the evaluation of TKIs across the drug discovery pipeline. We have conducted a systematic review of the literature to determine how rodent models are used in the measurement of TKI-induced cardiotoxicity, focusing on animal reports, physiological cardiac outputs, histopathology, and biomarkers. A PRISMA-compliant systematic review was conducted using PubMed, Scopus, and Web of Science to identify studies reporting on TKI-induced cardiotoxicity in rodents. Only controlled in vivo, primary in vitro, and ex vivo studies using rats, mice, hamsters or guinea pigs were included. Data were extracted on species, strain, sex, age, experimental design, and cardiac outcomes with risk of bias analyses performed using the SYRCLE and SciRAP tools. Among 92 studies, sunitinib, imatinib, and sorafenib were the most frequently examined TKIs, with cardiotoxicity exhibited as altered cardiac functional parameters, fibrotic changes, arrhythmias, and elevated cardiac biomarkers. Rats (51 studies) and mice (46 studies) were predominantly used to study the effects of TKIs, whilst guinea pigs were underrepresented, limiting insights into electrophysiological changes that are associated with cardiotoxicity. Most studies used male rodents, and only two studies assessed age-related effects. Comparison between species strains was rarely conducted, despite evidence of this being a contributing factor to pre-disposition to cardiotoxicity. Rodent models were shown to replicate TKI-induced cardiotoxic effects observed in humans, but risk of bias analyses revealed limited evidence for study randomisation, inconsistent blinding, lack of sex-balanced studies, and poor strain diversity. Poor methodological quality and reporting across studies compromised reproducibility and interpretation of clinical relevance. Our study highlights the need for implementation of standardised protocols, strain, sex and age-stratified analyses to better support preclinical-to-clinical translation, as well as improve the safety of TKIs for patients and ensure more ethical use of animals in research.