Acute storm versus chronic siege: divergent signalling networks and cell death programmes in liver-associated myocardial injury

Abstract Background Acute and chronic liver disease are associated with differing cardiac phenotypes and recovery trajectories, raising the question of how the underlying signalling networks and cellular injury programmes compare across these settings. We propose an “acute storm” versus “chronic siege” framework that compares two dominant injury programmes across signalling architecture, immune-cell programming, and injury mechanisms, while allowing overlap and transition. Main body In the acute storm, DAMPs and endotoxin trigger rapid, high-intensity TLR4/NF-κB activation and mitochondrial oxidative stress, with ferroptosis and pyroptosis implicated as prominent regulated cell-death programmes in preclinical models of acute cardiac injury. The associated cardiac phenotype is predominantly functional depression with greater potential for recovery. In the chronic siege, sustained gut-liver-heart axis disruption delivers persistent low-grade LPS, metabolic stressors, and microbial metabolites that may favour profibrotic macrophage states, contribute to impaired β-adrenergic reserve, and promote TGF-β-dependent fibroblast activation. The siege thus operates as a self-reinforcing intercellular communication loop, culminating in progressive structural remodelling with variable and often incomplete reversibility. A transitional grey zone, exemplified clinically by acute-on-chronic liver failure and mechanistically by the hypothesis of trained immunity, frames the boundary as a risk continuum rather than a fixed line. Conclusions The framework suggests different therapeutic research priorities: rapid control of acute inflammatory and injury pathways for the storm, and sustained metabolic and anti-fibrotic approaches for the siege. Prospective studies should test whether these dominant injury programmes correspond to different cardiac cell states and longitudinal trajectories of recovery and remodelling.