Abstract
Ferroptosis and pyroptosis are two distinct forms of regulated cell death that play crucial roles in cancer, neurodegeneration, and inflammatory diseases. Ferroptosis is characterised by iron-dependent lipid peroxidation, while pyroptosis is an inflammatory cell death mediated by gasdermin proteins. Recent studies reveal extensive crosstalk between these pathways. This review establishes the first hierarchical framework coupling the autophagy bridge function (ferritinophagy-mitophagy-cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) axis) with the p53/signal transducer and activator of transcription 3 (STAT3)/Nuclear factor erythroid 2-related factor 2 (NRF2) transcriptional hub, creating a unified decision-making network absent in prior reviews. Crosstalk mechanisms include the reactive oxygen species (ROS)-NOD-like receptor protein 3 (NLRP3) positive feedback loop, caspase cross-activation, and iron metabolism-inflammasome integration. Preclinically, the transferrin-targeted nanosystem Tf-LipoMof@PL increased intratumoral iron/ROS 3-5-fold, inducing robust antitumour immunity, while Ginsenoside Rh3 suppressed colorectal cancer growth in vivo via STAT3/p53/NRF2-mediated dual death induction. We critically address STAT3's paradoxical roles-promoting Gasdermin E (GSDME)-mediated pyroptosis in oesophageal cancer while suppressing NLRP3 via suppressor of cytokine signalling 3 (SOCS3) feedback in acute respiratory distress syndrome (ARDS)-highlighting cell type-specific feedback architectures that dictate phenotypic outcomes. For therapeutic translation, we propose a Translational Priority Matrix ranking nanodelivery systems (Tf-LipoMof@PL) and dual-function small molecules (N6F11) as the highest priority for intrahepatic cholangiocarcinoma (iCCA)/triple-negative breast cancer (TNBC), while deprioritising metal photosensitizers pending resolution of cardiac retention toxicity (0.8 μg/g myocardium in Good Laboratory Practice (GLP) studies). The "registration gap" stems from iron burst-release (> 80% within 30 min) and species-specific biomarker failures. We advocate replacing murine malondialdehyde (MDA)/glutathione (GSH) ratios with human-anchored metrics (ferritin heavy chain 1 (FTH1)/solute carrier family 40 member 1 (SLC40A1) expression, serum ferritin) and propose a "Cross-Death AI Platform" integrating network pharmacology (OmniPath/STRING), GraphSAGE deep learning (AlphaFold2 structures), and organoid validation to stratify patients and predict optimal drug combinations. By resolving spatiotemporal heterogeneity and implementing AI-guided precision medicine, we can transform multi-target interventions from empirical strategies into rational, patient-specific regimens, bridging the gap between preclinical promise and clinical success in cancers and neurodegenerative diseases.