Head of Bioengineering Novoheart, Medera Biopharm, United States
Abstract: Cardiac screening of newly discovered drugs remains a longstanding challenge for the pharmaceutical industry. While therapeutic efficacy and cardiotoxicity are evaluated through preclinical biochemical and animal testing, 90% of lead compounds fail to meet safety and efficacy benchmarks during human clinical trials. A preclinical model more representative of the human cardiac response is needed; heart tissues engineered from human pluripotent stem cell derived cardiomyocytes offers such a platform. In this study, three functionally distinct and independently validated bioengineered cardiac tissue assays were used: human ventricular cardiac anisotropic sheet (hvCAS) designed for screening cardiac electrophysiology, cardiac tissue strip (hvCTS) designed for cardiac contractility, and cardiac organoid chamber (hvCOC) designed for cardiac pump function. Each assay was exposed to increasing concentrations of known compounds representing 5 classes of mechanistic action, Including adrenergic agonists, calcium channel blockers, cardiac glycosides, hERG potassium channel blockers, and myosin II inhibitors. This created a robust electrophysiology and contractility dataset for training and testing the machine learning model. Combining the weighted results from six individual models, comprising support vector machine (SVM) and random forest (RF) models applied to data from each of the 3 assays, the optimized ensemble algorithm was able to classify the mechanistic action of unknown compounds with 86.2% predictive accuracy, outperforming simpler single-assay and dual-assay models. This study represents the first time such an ensemble machine learning approach has been used to leverage the complementary strengths of a diversity of cardiac screening assays, all built upon the same underlying ventricular cell types. Such screening tools represent the next generation of human-specific stem cell-based assays that are poised to fulfill the promise of efforts such as the FDA Modernization Act 2.0, aiming to improve upon, and ultimately replace experimental animals to efficiently bring safer and more effective drugs to patients in need.
Funding Source: Hong Kong Innovation and Technology Commission (Enterprise Support Scheme Grant #B/E033/18)
