Evolution of dose fractionation in radiotherapy

Radiotherapy is a cornerstone modality for the comprehensive management of tumors. Approximately 50% of patients with cancer receive radiotherapy during their treatment.1 Fractionation regimens are continuously evolving and being refined to achieve better tumor control while protecting normal tissues. Temporal fractionation, which delivers the total dose in a series of smaller daily fractions, protects normal tissues by allowing time for recovery between fractions and simultaneously increasing tumor radiosensitivity; thus, it has become the mainstream treatment regimen. Temporal fractionation is based on the classic 6Rs theory of radiobiology, which includes repair of sublethal injury, redistribution within the cell cycle, reoxygenation, repopulation, radiosensitivity, and reactivation of the antitumor immune response. Recently, FLASH radiotherapy, or the delivery of ultra-high dose rates of radiation (>40 Gy/s), has demonstrated unique biological advantages in normal tissues owing to its ultrahigh dose rates.2 Fundamentally, it can be viewed as a form of temporal segmentation and therefore represents a strategy for optimizing the temporal structure of radiotherapy. Although the temporal fractionation regimen has demonstrated favorable outcomes, it continues to encounter significant challenges in treating advanced, large-volume tumors and those exhibiting pronounced radioresistance. In this context, spatial fractionation, also known as spatially fractionated radiation therapy (SFRT), has recently received renewed attention.3 Distinct from temporal fractionation, which aims for uniform dose distribution across the target volume, SFRT is characterized by the deliberate creation of spatially modulated regions of extremely high dose (“peaks”) and low dose (“valleys”) within the tumor. Notably, tumor cells located in “valley” regions have been observed to exhibit radiobiological responses comparable to those in “peak” regions. This observation suggests that the underlying mechanisms of SFRT may extend beyond the classical 6Rs of radiobiology, potentially involving nontargeted biological effects such as the bystander effect and immune system activation. Combining temporal and spatial fractionation can create a novel radiotherapy fractionation mode, namely, spatiotemporal fractionated radiation therapy. A key feature of this mode is that each treatment fraction can deliver a high single dose to different subregions of the tumor target while maintaining a relatively consistent cumulative dose distribution in the surrounding normal tissues through reasonable planning, thereby fully exploiting the fractionation effect in radiobiology. Current research has mainly focused on the combined application of FLASH radiotherapy and SFRT (FLASH-SFRT).4 As a specialized implementation of spatiotemporal fractionation, the integration of these two modalities not only combines the technical benefits of ultrahigh dose rates and nonuniform dose distributions but may also elicit synergistic biological effects through distinct mechanisms, such as oxygen depletion, the bystander effect, and immune activation, thereby offering novel strategies to overcome the therapeutic limitations of conventional radiotherapy. In this study, we examined fractionation regimens in three dimensions: temporal, spatial, and spatiotemporal fractionation. We proposed a spatiotemporal domain map to integrate various dose-fractionation regimens, providing a unified theoretical framework and a visualization tool for exploring potential radiotherapy fractionation regimens. Temporal dose fractionation: The concept of temporal fractionation in radiotherapy originated from observations made by Leopold Freund in 1896, who discovered that dividing X-ray irradiation into fractions could effectively reduce the incidence of skin necrosis. Modern radiotherapy typically uses a conventional fractionation regimen consisting of 25–35 fractions, with daily doses ranging from 1.8 to 2.0 Gy. Building on this, various modified fractionation regimens have been developed, including hyperfractionation, accelerated fractionation, continuous hyperfractionated-accelerated radiotherapy, hypofractionation, and adaptive fractionation. A summary of the temporal fractionation regimens is provided in Supplementary Table 1, https://links.lww.com/CM9/C837. The common principle underlying these regimens is the modulation of four key parameters: total dose, dose per fraction, number of fractions, and frequency of fractions—aimed at optimizing tumor control while minimizing radiation-induced toxicity in normal tissues. Compared with conventional fractionation, hyperfractionation reduces the dose per fraction (1.1–1.3 Gy) while increasing the total number of fractions, enabling escalation of the total radiation dose with relative sparing of late-responding tissues, thereby improving tumor control probability. Hypofractionation uses a higher dose per fraction (e.g., 8 Gy), which significantly reduces the number of treatment sessions. The theoretical rationale for this approach lies in the low α/β ratio of certain tumor types, such as prostate cancer, where higher doses per fraction can more effectively induce tumor cell death and enhance biological effectiveness. Both accelerated fractionation and continuous hyperfractionated-accelerated radiotherapy use a strategy of multiple daily irradiations to complete the treatment course before the onset of accelerated tumor repopulation in rapidly proliferating tumors, thereby enhancing local tumor control rates. However, this regimen may exacerbate acute radiation toxicity. Adaptive radiotherapy introduces the concept of dynamic plan adjustment, allowing real-time optimization of treatment plans based on changes in tumor morphology and volume, reflecting the evolving direction toward individualized and precise treatment. FLASH radiotherapy, a distinct modality within temporal fractionation, does not rely on the modulation of the four key parameters, but instead involves modulation of the dose rate. It confers substantial protection to normal tissues by delivering radiation at ultrahigh dose rates. Its mechanism of action transcends the biological framework of conventional fractionated radiotherapy, engaging pathways such as oxygen depletion, free radical effects, immune activation, and other physicochemical mechanisms, thereby offering the potential for novel radiobiological regimens. However, the mechanism of the FLASH effect remains unclear, and further research is required to determine the maximum FLASH effect achievable with temporal fractionation. Spatial dose fractionation: The concept of spatial fractionation dates back to 1909, when German scholar Alban Köhler proposed the “grid therapy” (GRID). Modern evolutions include lattice radiation therapy (LRT), minibeam radiation therapy (MBRT), and microbeam radiation therapy (MRT), which are based on differences in the size and spacing of subbeams. A summary of the spatial fractionation regimens is presented in Supplementary Table 2, https://links.lww.com/CM9/C837. The GRID and LRT operate on a centimeter scale and can be implemented in clinical practice using conventional linear accelerators. The core difference between the two lies in the irradiation dimensions; GRID uses a two-dimensional planar irradiation method, which provides limited dose coverage for deep-seated tumors and relatively weak protection for normal tissues. In contrast, LRT uses a three-dimensional stereotactic irradiation mode, enabling more precise control of dose distribution and significantly enhancing the protection of surrounding normal tissues. The MBRT and MRT fall within the submillimeter-to-micrometer scale range. MBRT is regarded as an important transitional form of MRT for clinical applications. It uses a relatively wide beam width (approximately 1 mm) and a moderate peak-to-valley dose ratio (10–20), and can be implemented on a conventionally modified linear accelerator. MRT represents a radiotherapy mode under extreme physical conditions, featuring extremely fine beams (25–100 µm), small inter-beam spacing, an extremely high peak-to-valley dose ratio (>50), and tolerates extremely high single-peak doses (>500 Gy). Its significant normal tissue protection effect is mainly attributed to the biological response induced by the extremely low “valley” dose regions. However, this technology currently relies heavily on synchrotron radiation sources, which have strict equipment requirements, and faces significant technical and implementation challenges in treating deep-seated tumors. Spatiotemporal dose fractionation: Historically, the moving strip technique has been regarded as the earliest form of spatiotemporal fractionation. This method is characterized by daily shifts in the irradiation field, enabling comprehensive abdominal irradiation while maintaining radiation doses to normal tissues within tolerable limits. Currently, clinical applications of spatiotemporally fractionated radiotherapy remain limited. Preliminary studies have explored spatiotemporal fractionated optimization methods for the brain, liver, and spinal metastases; however, their clinical efficacy requires further validation. FLASH-SFRT represents a unique integration of spatial and temporal fractionation modalities, and its combined implementation may elicit a synergistic effect that transcends the classical 6Rs framework of radiobiology, potentially enhancing the therapeutic index, particularly in radioresistant malignancies such as glioblastoma and hypoxic sarcomas. However, this synergistic effect arises from the combination of two cutting-edge technologies with distinct and still unclear mechanisms. Although pioneering studies using the synchrotron-generated microbeam FLASH in murine models have demonstrated remarkable tumor control and normal tissue sparing, the synergistic effect resulting from their combined application may be comparable to, or potentially worse than, the biological effects of conventional irradiation. Additionally, it may lead to an increase in tumor radioresistance owing to uneven prescription dose distribution (such as low-dose valleys). Further research is needed to clarify their synergistic effects. To date, few clinical trials have been conducted. The clinical translation of FLASH-SFRT faces engineering challenges in synchronizing spatial and temporal dose delivery with high precision. Currently, the equipment that supports this type of fractionation regimen includes proton-beam systems that use slit collimators or magnetic focusing to achieve submillimeter dose gradients, large-scale synchrotron facilities, line-focus X-ray tubes, and modified linear accelerators that combine robotic multileaf collimators with FLASH electron beams. However, these devices are still in the research stage. Summary and future perspectives: This article systematically elucidates the evolution of dose-fractionation regimens in radiotherapy. As demonstrated in the spatiotemporal domain map Figure 1, radiotherapy fractionation can be comprehensively categorized as temporal, spatial, and spatiotemporal. Color variation represents the maturity of different fractionation regimens (darker shades indicate greater maturity). Maturity refers to the quantity and depth of published research in each field. Future research should focus on relatively immature and underexplored areas. This diagram provides a robust theoretical foundation for optimizing therapeutic ratios through multidimensional dose modulation.Figure 1: Spatiotemporal domain map of radiotherapy. Color variation represents the maturity of different fractionation regimens (darker shades indicate greater maturity). The n represents the approximate number of searchable studies associated with each radiotherapy fractionation regimen. Distances on the horizontal and vertical coordinate scales are for illustrative purposes only and do not reflect actual numerical proportions. The dimensional boundaries of each fractionation regimen are not absolute, and overlapping transition regions exist in practical applications. CHART: Continuous hyperfractionated-accelerated radiotherapy.Current clinical practice primarily focuses on experience-driven sequential fractionation schemes. However, research methods based on the spatiotemporal domain map hypothesis are more effective than blind trial-and-error approaches. Five core research directions warrant attention. First, expediting the clinical translation of preclinical breakthroughs, particularly FLASH radiotherapy and lattice beam radiotherapy, requires establishing standardized protocols for beam parameter optimization and normal tissue sparing. Second, developing advanced treatment planning systems that integrate real-time predictions of biological effectiveness could effectively bridge the gap between theoretical models and clinical applications. Third, MBRT and MRT require advanced collimation systems and novel dosimetry tools for precise submillimeter beam characterization. Fourth, the potential synergy between spatial fractionation and immunotherapy warrants further exploration. Fifth, the need to advance the application of artificial intelligence to optimize personalized segmentation regimens, including the dynamic adjustment of treatment parameters informed by radiomics or multiomics data. These core directions are imperative for fostering cross-disciplinary collaboration among medical physicists, radiation biologists, and artificial intelligence specialists. Ultimately, integrating advanced fractionation regimens with precise image guidance and biological targeting may be useful in a new era of ablative radiotherapy characterized by unparalleled therapeutic precision. Funding This study was supported by the National High Level Hospital Clinical Research Funding (No. 2025-LYZX-Z-B03). Conflicts of interest None.