Presenter(s)
A. Liles1,2, and J. S. Bredfeldt1; 1Department of Radiation Oncology, Brigham and Women’s Hospital, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, 2Department of Physics, University of Massachusetts Lowell, Lowell, MA
Purpose/Objective(s): Intrafraction motion degrades target dose coverage during radiotherapy delivery, particularly for prolonged beam on time (BOT). While accelerated delivery techniques are increasingly feasible, the degree to which reduced BOT mitigates dose loss depends strongly on the underlying motion pattern. This study quantifies the dosimetric impact of BOT across four clinically relevant motion classes and evaluates the tradeoff between BOT and required treatment margin using dose-based robustness metrics.
Materials/Methods: A Monte Carlo simulation framework was developed to model intrafraction target motion during treatment delivery. Four motion patterns were evaluated: baseline drift (i.e., bladder filling), random motion (gross patient motion), cyclic respiratory motion (breathing or cardiac motion), and random cyclic motion (swallowing or rectal gas). BOTs ranged from 10 - 240 s at 0.1 s temporal resolution and a logistic dose penumbra model was used to approximate realistic beam edge falloff (80–20% penumbra width = 4 mm). Delivered dose was computed continuously as the target translated relative to the beam aperture with mean delivered dose and D95% evaluated for each delivery. Margin vs. BOT tradeoffs were quantified by determining the minimum margin required to maintain D95 = 95% of prescription dose. Each condition was averaged across 1000 simulated fractions.
Results: Mean delivered dose exhibited modest degradation with increasing BOT across all motion types. D95 demonstrated substantially greater sensitivity to delivery duration and motion characteristics. Baseline drift produced the most pronounced degradation with D95 declining sharply with increasing BOT, requiring margins to increase from approximately 2 mm for BOTs <30 s to >5 mm for BOTs =180 s. Random motion showed progressive but slower deterioration, with required margins increasing gradually as BOT lengthened. Cyclic respiratory motion exhibited minimal dependence on BOT beyond one breathing cycle with required margins remaining largely stable per delivery time. In contrast, random cyclic motion produced stable mean dose but highly variable D95, particularly at longer BOTs where the probability of treating through one movement excursion increased. Margin requirements for this pattern increased moderately with BOT but plateaued once excursion treatment probability saturated.
Conclusion: BOT reduction substantially improves target dose robustness for non-periodic motion patterns, particularly baseline drift and transient excursions, and potentially enables margin reduction. Periodic respiratory motion is comparatively insensitive to delivery duration beyond one motion cycle. These findings suggest that accelerated delivery primarily mitigates stochastic and drift-like intrafraction motion and should be considered a complementary strategy to motion management rather than a substitute for respiratory gating or tracking.