This topic might seem a bit unusual for the theme of gene-environment interactions — but responses of genes (genetic susceptibility) to any drug (an environmental signal) seems clear. And any screening-model system that attempts to predict clinical drug response would be an important advance in the field. Among different types of drug-induced toxicity, cardiovascular toxicity is responsible for 17.4% of drugs removed from the pipeline before they are clinically available, as well as 46% of drugs removed after they have been on the market. Considering this (health and economic) impact, the pharmaceutical industry has implemented more stringent cardiovascular toxicity testings during early drug-development stages.
During early phases of drug development, cell-culture systems [e.g. human induced pluripotent stem cell cardiomyocytes (hiPSC-CMs)] are the traditional model for high-throughput screens of chemical libraries for effects on heart. As human-derived cells, hiPSC-CMs retain full genetic conservation of cardiac drug targets; hence, hiPSC-CMs can provide insights into altered cellular processes such as calcium transport, mitochondrial regulation, or ion channels function. Despite these relevant features — cellular systems do not recapitulate the structural complexity of the heart or its interplay with the vasculature. In addition, cell-culture models are removed from other organs and their metabolism and thus cannot address the cardiotoxic impact of drug metabolites. Although hiPSC-CMs might provide a good translational model for on-target effects, their reductionist nature (finding associations between phenomena, which can be described in terms of other simpler or more fundamental phenomena) might diminish their predictive power when it comes to assessment of cardiotoxic adverse effects.
In addition to “public relations” concerns (about screening tests in the dog, rat or mouse), use of mammals for cardiotoxicity-testing has a negative impact on the industry efforts for implementing animal research “replacement, reduction, and refinement” (3R) measures. Limitations of cellular and mammalian models reveal the need for alternative screening methods to streamline the prediction of drug-induced cardiotoxicity in early/mid-stage drug development phases. Along these lines, using zebrafish larvae [see attached article] can become a major asset for selecting safer candidates for preclinical phases and to avoid late drug attrition. Two reasons make the zebrafish an attractive model for pharmaceutical research: [a] The zebrafish genome is highly evolutionarily conserved — which results in most drug-metabolism components and on-/off-drug targets being conserved with humans. [b] Zebrafish are well suited for high-throughput-screening approaches: larvae are small, large numbers of progeny develop outside the mother. Moreover, tissue transparency and rapid life cycle of the larvae make them an ideal model for non-invasive image-based in vivo assessment of organs, and the availability of transgenic lines facilitates observing specific organs by fluorescent microscopy. Finally, zebrafish larvae younger than 5 days post-fertilization are considered cell-culture systems by animal welfare regulation within the European Union — allowing 3R implementation by the pharmaceutical industry.
Authors [see attached article] developed a high-throughput screening platform to measure many relevant cardiovascular parameters: heart rate, arrhythmia, atrial-ventricular blockage, ejection fraction, and blood flow, etc. Authors tested a chemical library of 92 compounds with known clinical cardiotoxicity profiles. Cross-comparison with clinical data and data acquired from hiPSC-CMs calcium-imaging showed that zebrafish larvae allow a more reliable prediction of cardiotoxicity than cell culture systems. Analysis with zebrafish yields similar predictive performance as previous validation meta-studies performed with dogs — which are the standard regulatory preclinical model for predicting cardiotoxic liabilities prior to clinical phases.
DwN
Toxicol Sci Aug 2019; 171: 283–295