Reversing the pipeline: a ‘human-first’ multi-omics approach to cardiovascular discovery

Abstract

Despite significant biomedical advances, cardiovascular disease remains a leading cause of global mortality. The traditional ‘bench-to-bedside’ research pipeline, which heavily relies on animal models to guide clinical translation, is often hampered by species-specific physiological differences, leading to high rates of clinical trial failures. In this Review, we advocate for a paradigm shift toward a ‘human-first’ discovery approach. We explore how integrating human cardiac tissue resources obtained from biopsies, surgical explants and autopsies with high-resolution multi-omics technologies (including transcriptomics, proteomics and metabolomics) is transforming our understanding of complex heart diseases. We highlight recent breakthroughs in heart failure, ischemic heart disease, arrhythmias, myocarditis and valvular stenosis achieved through direct profiling of human samples. Finally, we propose a ‘bedside-to-bench-and-back’ strategy, in which therapeutic targets are identified in human tissues and then validated in animal models to accelerate the development of precise, clinically relevant cardiovascular treatments. The graphical abstract contrasts the conventional bench-to-bedside model with the human-first framework proposed in this Review. In the traditional pipeline, hypotheses are usually generated from literature and preclinical animal models, followed by mechanistic studies, therapeutic testing and eventual clinical translation. Although this approach has produced major advances, it is limited by species-specific differences in cardiovascular physiology and by the inability of many animal models to fully recapitulate complex human disease. In the proposed human-first model, discovery begins with clinically derived human cardiovascular tissues, including biopsies, surgical specimens, explanted hearts, donor tissues and autopsy samples. These samples are analyzed using high-resolution molecular technologies such as transcriptomics, proteomics, metabolomics, epigenomics, single-cell sequencing and spatial omics. Candidate mechanisms identified directly in human tissue are then prioritized for experimental validation in animal models, organoids, engineered tissues or other perturbable systems. The goal is not to eliminate animal models, but to reposition them as causal-validation platforms downstream of human molecular discovery. This iterative bedside-to-bench-and-back strategy is intended to improve translational fidelity, identify clinically relevant mechanisms and accelerate the advancement of precision cardiovascular medicine. This summary was initially drafted using artificial intelligence, then revised and fact-checked by the author.