Monte Carlo Evaluation of Tissue Assignment Variation In Low-Energy Lung Brachytherapy
Abstract
Purpose
To quantify the effect of tissue assignment variation on dose distributions for ¹⁰³Pd, ¹²⁵I, and ¹³¹Cs seeds in lung brachytherapy.
Methods
CT images from three patients with non-small cell lung cancer and the corresponding contours of the planning target volume (PTV), ipsilateral lung, and heart were investigated. Three tissue assignment schemes were examined: coarse (5 categories), intermediate (12 categories), and fine (20 categories). TOPAS Monte Carlo simulations were performed for each patient, source type, and tissue assignment scheme using identical implant geometries. Dose metrics including D90 and V100 for the PTV, mean lung and heart dose were calculated. Percent differences between tissue schemes were analyzed.
Results
The Low-energy sources, particularly ¹⁰³Pd, demonstrated the highest sensitivity to tissue assignment, with D90 varying by up to 12% between coarse and fine schemes. Mean lung dose differed by 6–7% for ¹⁰³Pd, while ¹³¹Cs showed smaller variations (< 4%). Mean heart dose was only slightly affected (<3%), though slightly higher for finer tissue assignment scheme due to the more accurate modeling of soft tissue. Coarse tissue assignments overestimated the dose in low-density regions and underestimated it near high-density structures.
Conclusion
Tissue assignment variation can result in clinically significant differences in lung brachytherapy, particularly for low-energy sources. Model-based Monte Carlo calculations with refined tissue definitions are recommended for treatment planning to ensure accurate target coverage and organ-at-risk dose calculation.