Mimicking the Heart In Vitro and Ex Vivo: A Critical Comparative Analysis of Self-Organizing, Engineered, and Native Cardiac Model Systems

Understanding human cardiac development, disease, and therapeutic response requires experimental models that reproduce defined aspects of myocardial organization and function. However, no single in vitro or ex vivo preparation captures the complete structural, electrophysiological, contractile, metabolic, vascular, stromal, and immune context of the human heart. This review critically compares five partly overlapping model families: cardiac spheroids and aggregate-based microtissues, developmentally self-organizing cardiac organoids and assembly-guided constructs, engineered heart tissues, heart-on-a-chip systems, and living myocardial slices. The comparison distinguishes how tissue organization is generated or preserved, which variables can be controlled, how function is measured, and which conclusions are supported by the resulting evidence. Across model families, interpretation is shaped by species and cell source, maturation state, cellular composition, conditioning, transport, measurement design, and donor variables. Representative studies illustrate applications in cardiac development, functional perturbation, fibrosis, immune -cardiac interactions, genotype attribution, and retained adult tissue physiology. Major limitations include domain-specific and incomplete maturation, restricted oxygen delivery, variable cellular persistence, technical reproducibility, limited access to human tissue, and insufficient consideration of donor sex and hormonal context. Rather than identifying a universally superior platform, this review proposes a claim-validation alignment in which model selection is guided by the biological question, the required tissue features, the relevant endpoint, and the level of inference supported by validation.