Multiscale modelling and safe-by-design strategies for iron oxide nanoparticles in diagnosis and therapy

Abstract

Iron oxide nanoparticles (IONPs), particularly superparamagnetic iron oxide nanoparticles (SPIONs), represent one of the most clinically relevant nanoplatforms for magnetic resonance imaging (MRI), targeted drug delivery, magnetic hyperthermia, and integrated theranostic applications. Their multifunctional adaptability across oncology, neurodegenerative disease, cardiovascular medicine, immunomodulation, and regenerative therapy has stimulated extensive translational research. However, clinical advancement remains constrained by complex nano-bio interactions, protein corona formation, oxidative reactivity, biodistribution variability, immune recognition, and uncertainties surrounding long-term iron metabolism and safety. This review examines how predictive multiscale computational modelling frameworks can address these challenges and support rational safe-by-design optimization of IONPs prior to extensive experimental and clinical implementation. We discuss density functional theory (DFT) approaches for investigating Fe₃O₄ surface energetics, magnetic anisotropy, facet-dependent reactivity, and dissolution behaviour; molecular dynamics (MD) and coarse-grained (CG) simulations for understanding protein corona evolution, membrane interactions, aggregation dynamics, and cellular uptake; and nano-quantitative structure-activity relationship (nano-QSAR), nano-informatics, and machine learning (ML) approaches for predicting cytotoxicity, inflammatory signalling, and biological response using physicochemical and electronic descriptors. At the organism level, physiologically based pharmacokinetic (PBPK) and PBPK-toxicodynamic models are highlighted for linking nanoscale properties with biodistribution, organ accumulation, clearance kinetics, and long-term safety prediction. By integrating quantum-level, molecular, cellular, tissue, and systemic modelling approaches, this review emphasizes how computational nanomedicine can transition from descriptive characterization toward predictive theranostic design. Rather than treating imaging performance, therapeutic efficacy, biodistribution, and nanotoxicity as isolated challenges, the reviewed modelling strategies provide a unified framework for mechanistically informed, clinically translatable, and safe-by-design iron oxide nanomedicine.