Ultra-high-dose-rate pulsed irradiation and the conditional FLASH effect: dosimetry, multiscale biology, and clinical...

Ultra-high-dose-rate (UHDR) pulsed irradiation can deliver therapeutic doses within an extremely short time window, but UHDR is a physical delivery state, whereas the FLASH effect is a conditional comparative phenotype in which normal-tissue injury is reduced relative to a spatially dose-comparable conventional-dose-rate (CDR) irradiation comparator without loss of tumor control. This Review integrates evidence across four levels: physical exposure characterization, dosimetric measurement, biological comparison, and clinical translation. We examine the temporal structure and quality-assurance requirements of electron, very-high-energy electron, photon, proton, and heavy-ion UHDR platforms; synthesize positive, negative, and conditional findings across normal-tissue and tumor models; and integrate mechanistic studies spanning multiple time scales. Current evidence does not support 40 Gy/s, or any single average dose-rate metric, as either a sufficient or a necessary threshold for the FLASH effect. Effect magnitude and reproducibility are more likely determined jointly by dose per pulse, pulse structure, total delivery time, fractionation, irradiated volume, radiation quality, tissue-specific radiobiological characteristics, and biological state. Mechanistic data likewise favor multiscale coupling across time rather than a single explanatory pathway, although several causal links from early oxygen and radical dynamics to cellular and tissue outcomes still lack direct interventional validation. Human studies have established the feasibility of clinical UHDR delivery on selected electron and proton platforms and have provided preliminary safety data, but comparative clinical benefit over CDR irradiation has not yet been demonstrated. Further translation should therefore prioritize pulse-resolved traceable dosimetry, organ-specific normal-tissue dose-response relationships, late toxicity, comparable tumor control, and concurrent comparative clinical endpoints, with expansion guided by prespecified stage-gated translational criteria.