The evolution of surrogate microspheres in transarterial radioembolization: from SPECT prediction to multimodal AI-guided precision dosimetry.

Surrogate materials enable pre-treatment prediction of intra-hepatic distribution of therapeutic radioactive microsphere, facilitating patient-specific dosimetry for transarterial radioembolization (TARE). Technetium-99m macroaggregated albumin (⁹⁹ᵐTc-MAA) has long constituted the clinical standard for SPECT-based surrogates, however, its fundamental physicochemical difference from therapeutic microspheres limits the reliability of MAA-based dosimetry and prediction for personalized treatment planning. Next-generation surrogate technology includes biodegradable polymeric microspheres engineered to closely replicate the therapeutic microsphere while affording post-procedural arterial recanalization through controlled biodegradation. Theranostic microspheres incorporating diverse radionuclides, positron-emitting PET compatible surrogates, and multimodal CT and MRI-visible surrogates have collectively expanded same-particle pre-treatment dosimetry capabilities. These material-level advances are synergistically coupled with computational dosimetry tools including artificial intelligence/deep learning algorithms for automated segmentation/dose-prediction, and Monte Carlo simulations that enhance voxel-based dosimetry precision. This review examines recent advances in surrogate materials design and integrated imaging/dosimetry methodologies that enhance TARE efficacy, safety, and personalized treatment outcomes.