Toward clinical harmonization of radiobiological modeling in particle-based radiotherapy: a translational review of proton, heavy-ion, and BNCT frameworks

Particle-based radiotherapy, including proton therapy, heavy-ion therapy, and boron neutron capture therapy (BNCT), enables conformal dose delivery but introduces model-dependent biological-dose uncertainty. This uncertainty reflects variations in relative biological effectiveness (RBE), linear energy transfer (LET), oxygenation, microdosimetric spectra, and boron distribution, so identical nominal prescriptions may not yield equivalent biological effects. This review summarizes representative clinical, secondary-review, and investigational model families, including fixed and variable-RBE proton models, the local effect model (LEM) and microdosimetric kinetic model (MKM) families for heavy ions, and compound biological effectiveness (CBE), photon-isoeffective-dose, microdosimetric, and nanodosimetric BNCT frameworks. We compare their assumptions, input descriptors, validation status, clinical readiness, and implementation barriers. We further distinguish prescription-embedded clinical systems from secondary-review models and investigational mechanistic frameworks to clarify the current level of clinical evidence supporting each model family. Published experimental and model-comparison studies indicate that proton RBE may increase in high-LET distal-edge regions, that LEM- and MKM-based carbon-ion systems can yield clinically relevant, site- and endpoint-dependent differences in RBE-weighted dose interpretation, and that BNCT biological dose is strongly affected by component-dose and boron-distribution assumptions. We propose an evidence-informed four-layer harmonization perspective, rather than a formal consensus guideline, consisting of minimum reporting, reference mapping, model-sensitivity quality assurance (QA), and registry-based clinical learning.