Impact of tissue heterogeneity on intracranial Cesium‐131 brachytherapy dosimetry: A patient‐specific Monte Carlo Study

Abstract Background AAPM TG‐43‐based dose calculations for intracranial cesium‐131 ( 131 Cs) brachytherapy ignore patient‐specific heterogeneities, which can affect dose accuracy. This study investigates the dosimetric impact of these heterogeneity effects using Monte Carlo (MC) simulations and patient‐specific imaging data. Purpose To quantify the impact of tissue heterogeneity on dose distributions in intracranial brachytherapy using 131 Cs sources and to evaluate the accuracy of AAPM TG‐43‐based treatment planning system (TPS) calculations. Methods TG‐43 dosimetric parameters for the 131 Cs source were independently characterized and benchmarked against published consensus data. MC simulations using MCNP were performed in both homogeneous water phantoms and heterogeneous voxelized lattice phantoms derived from CT DICOM RT data. Single and multiple‐seed configurations were investigated. TPS and MC dose distributions were compared using dose‐volume histogram (DVH) metrics such as (V 60 , V 100 , D 2cc, and mean dose), isodose distributions, dose difference maps, and local and global dose difference analyzes. Spatial agreement was quantified using gamma analysis. Results Strong agreement was observed between TPS and MC calculations in homogeneous media. However, heterogeneous patient‐specific MC simulations revealed systematic differences in DVH metrics for the target volume and organs at risk, including the skull and skin. Local dose differences revealed pronounced deviations near high‐density bony regions, where dose enhancement effects were observed, while global dose difference analyzes demonstrated an overall dosimetric bias. Gamma pass rates decreased in heterogeneous regions, consistent with these spatial dose discrepancies. Compared with the MC simulations, the TG‐43‐based TPS overestimated the mean dose to the target by approximately , V 60 and V 100 by approximately , and , respectively, and the mean dose to the skin by approximately , while underestimating the mean dose to the skull by . Conclusion Although the TG‐43 formalism provides accurate dose estimates in homogeneous media, it introduces clinically meaningful local and global dosimetric deviations in heterogeneous intracranial environments. Careful consideration is required when reporting doses to the target and OARs, particularly given the potential for substantially higher doses to the irradiated skull in certain cases (up to 3.2‐fold in this study), as well as the possibility of target underdosage within the high‐risk clinical target volume (HR‐CTV). Patient‐specific MC modeling highlights the importance of incorporating tissue heterogeneity into intracranial brachytherapy dose calculations to improve dosimetric accuracy and optimize treatment planning.