Beltran, Chris

ABSTRACT Background Radiotherapy is a cornerstone of treatment for pediatric central nervous system (CNS) tumors, but dose selection remains uniform despite interpatient variability in tumor radiosensitivity. This is consequential in children, for whom toxicity can have lifelong effects. Genomic-adjusted radiation dose (GARD) integrates tumor genomics with delivered radiation dose to quantify the biological effect of radiotherapy and has been validated across adult malignancies. Its relevance in pediatric CNS tumors is unknown. Methods We performed a cohort study using gene expression and data from 246 pediatric patients with high-grade glioma, medulloblastoma, or ependymoma from the Children’s Brain Tumor Network. GARD was calculated using a sequencing-adapted radiosensitivity index integrated with radiation dose through the linear–quadratic model. Associations of GARD and physical radiation dose with event-free and overall survival were evaluated using Cox proportional hazards models stratified by tumor type and anatomic location. Patients who did not receive radiotherapy comprised a negative control cohort (sham-GARD). Results Among patients receiving radiotherapy, physical dose was uniform, whereas GARD showed substantial interpatient variability in biological effect. Higher GARD was associated with improved event-free survival (hazard ratio [HR] 0·90, 95% CI 0·83–0·97; p=0·004) and overall survival (HR 0·90, 95% CI 0·83–0·99; p=0·018). Physical dose was not associated with either endpoint. Among patients who did not receive radiotherapy, sham-GARD was not associated with outcomes, supporting GARD as a treatment-specific predictor rather than a prognostic biomarker. Conclusions In pediatric CNS tumors, the biological effect of radiotherapy quantified by GARD was associated with outcomes, whereas physical dose was not. These findings challenge uniform radiotherapy dosing and support genomically informed dose individualization. Prospective evaluation of GARD-guided radiotherapy is warranted to optimize tumor control while minimizing long-term toxicity in children.

Abstract Purpose Modeling relative biological effectiveness (RBE) is central to carbon ion radiotherapy treatment planning. The modified Microdosimetric Kinetic Model (mMKM) is a clinically established RBE framework that has guided treatment protocols at many existing carbon centers, while the Mayo Clinic Florida Microdosimetric Kinetic Model (MCF MKM) is a recently developed alternative. This work aims to implement the MCF MKM in the open‐source treatment planning system matRad and to quantitatively compare its RBE‐weighted dose predictions with those of the clinically established mMKM using identical physical dose distributions across multiple disease sites. These findings will help assess their dosimetric equivalence and inform protocol development for carbon ion radiotherapy at MCF. Methods Monte Carlo simulations of the MCF carbon beamline were performed to generate physical (IDD, LET, lateral spread) and biological base data for integration of each RBE model into matRad . Treatment plans were generated for six patients, each corresponding to a different disease site, using clinical beam configurations with carbon‐reference dose prescriptions, and plans were optimized using the MCF MKM. To isolate differences attributable solely to the RBE model, the resulting physical dose distributions were held fixed and RBE‐weighted doses were recalculated using the mMKM. Dose volume histogram (DVH) metrics and spatial dose‐difference maps were used to compare target coverage and organ‐at‐risk doses between the two models. Results Across patient cases, RBE‐weighted dose distributions from MCF MKM and mMKM showed strong agreement. Differences in target coverage were small, with CTV D95% differing by less than 1.6% across all disease sites and maximum target dose differences not exceeding 0.88%. Organ‐at‐risk dose deviations were limited, with differences of 3.0% or less across evaluated DVH metrics. Spatial dose‐difference maps showed that the largest discrepancies occurred in regions of steep dose gradients near target to organ‐at‐risk interfaces, while overall dose conformity and plan quality remained comparable between the two models. Conclusions: This study served as the first systematic model comparison of the MCF MKM and mMKM within a treatment planning environment. These findings suggest that the MCF MKM and mMKM produce dosimetrically consistent RBE‐weighted dose predictions under realistic planning conditions using carbon‐reference parameters. Accordingly, the fractionation schemes developed from years of clinical experience with mMKM implementations may serve as a practical foundation for protocol development at MCF.

The sharp dose gradients that underpin the dosimetric advantage of particle therapy over photon therapy can be undermined by the interplay effects due to intra-fraction motion in modern pencil beam scanning systems. Fluoroscopy-Guided Particle Therapy (FGPT) offers a promising path to improved motion management through real-time tracking of tumors or surrogate signals. The advent of flat-panel detector (FPD)-based technology has enabled tighter integration of fluoroscopy/fluorography into treatment units and accelerated clinical adoption and research, with commercial systems such as Hitachi’s Real-time Gated Particle Therapy (RGPT) now available. However, the need for implanted fiducial markers, with the associated invasiveness and risk of complications, limits the utility of RGPT to a few anatomic sites in selected patients. The full potential of FGPT, therefore, depends on reliable marker-less tumor tracking, which remains challenging because soft-tissue targets are obscured by overlapping anatomy along the X-ray path, leading to reduced reliability of traditional image-registration algorithms in the projection domain. Recent advances in deep learning and AI-driven image registration have renewed hope for overcoming these barriers, enabling real-time marker-less tracking for particle therapy. This review outlines the evolution of fluoroscopy technology from image intensifier (II) to FPD-based systems, summarizes historical and recent vendor-supported FGPT strategies, and surveys emerging AI-based algorithms in the literature. A general review of machine learning-based image registration is provided, challenges in generalizability and interpretability are highlighted, and potential paths toward reliable, clinically deployable FGPT are discussed.

Introduction Proton dose driven continuous scanning (DDCS) is a form of proton pencil beam scanning (PBS) in which the beam may remain on during transitions between successive spots. Mayo Clinic Florida (MCF) is preparing to commission a synchrotron-based proton therapy system designed for DDCS, incorporating novel irradiation control features. Given the novelty and complexity of this system, a systematic risk-based analysis was performed to inform upcoming commissioning strategies. Methods A prospective risk analysis was conducted using the American Association of Physicists in Medicine (AAPM) Task Group 100 (TG 100) framework. A comprehensive process map of the DDCS irradiation control workflow was developed. Failure modes and effects analysis (FMEA) was performed to identify potential delivery system failure modes and to score each failure mode based on occurrence, severity, and lack of detectability. Targeted mitigation and commissioning strategies were then developed to address the highest-risk items. Results Twenty-nine delivery system failure modes were identified. The highest-risk failure modes were specific to DDCS operation. Targeted commissioning strategies were defined, including scan path measurement and modeling, validation of move and flap dose, dose monitor collection efficiency testing, spot position monitor performance evaluation, and verification of control system intervention behavior. Incorporation of these strategies into the commissioning plan reduced the mean RPN by more than 50%. Discussion This work provides a detailed description of the MCF proton DDCS irradiation control system and uses a TG 100 based approach for identifying system-specific risks. As new systems incorporate technologies for which limited guidance exists, institutions must perform their own risk analyses to determine appropriate quality management strategies.