Computational Modelling of Myocardial Metabolism in Patients with Advanced Heart Failure

Although perturbations of myocardial metabolism and energy depletion are fundamental characteristics of heart failure (HF) pathophysiology, interindividual variability in bioenergetic phenotypes and their clinical implications remain poorly understood. Here, we used large-scale proteomics data from 137 cardiac biopsies to generate personalized computational models of myocardial metabolism in patients with advanced HF and nonfailing sex-matched controls. Myocardial bioenergetics were highly heterogeneous in HF, ranging from severely diminished to normal ATP production capacity. In patients undergoing left-ventricular assist device implantation, preserved myocardial substrate preference was associated with a favorable myocardial response upon mechanical circulatory support. Moreover, system level simulations identified two mitochondrial proteins that represent potential targets to efficiently restore myocardial ATP production capacity in HF. Collectively, computational modelling provides functional, quantitative insights into impaired myocardial metabolism in HF. Personalized reconstruction of cardiac bioenergetics may improve understanding of disease heterogeneity, individual risk stratification, and guidance of personalized clinical decision-making in the future.