Ferroptosis in cancer: integrating multi-omics, AI-driven biomarkers, and translational challenges

Ferroptosis, an iron-dependent regulated cell death driven by lipid peroxidation, has emerged as a potential target in cancers resistant to apoptosis and other programmed cell death pathways. This review integrates classical mechanistic insights with multi-omics and artificial intelligence (AI) approaches to examine ferroptosis regulation, biomarker identification, and therapeutic potential across malignancies. Core regulators (GPX4, SLC7A11, ACSL4, FSP1, labile iron pool) interact with oncogenic pathways (RAS, p53, NRF2, YAP/TAZ) to influence sensitivity. Multi-omics integration and AI models have generated prognostic signatures and supported drug sensitivity predictions in retrospective cohorts, while circulating biomarkers (e.g., MDA, 4-HNE, exosomal FRGs) show promise for non-invasive monitoring. Ferroptosis also modulates the tumor microenvironment through immune-metabolic interactions, with preclinical evidence of synergy with immunotherapy and radiotherapy. However, clinical translation remains at an early stage, with major challenges in biomarker standardization, model generalizability, context-dependent effects, and the lack of mature prospective trials. AI-assisted multi-omics strategies offer opportunities for future precision oncology applications, but rigorous validation and standardization are required before routine clinical implementation.