103Pd strings: Monte Carlo assessment of a new approach to brachytherapy source design

PURPOSE: A new type of (103)Pd source (CivaString and CivaThin by CivaTech Oncology, Inc.) is examined. The source contains (103)Pd and Au radio-opaque marker(s), all contained within low-Zeff organic polymers that permit source flexibility. The CivaString source is available in lengths L of 10, 20, 30, 40, 50, and 60 mm, and referred to in the current study as CS10-CS60, respectively. A thinner design, CivaThin, has sources designated as CT10-CT60, respectively. The CivaString and CivaThin sources are 0.85 and 0.60 mm in diameter, respectively. The source design is novel and offers an opportunity to examine its interesting dosimetric properties in comparison to conventional (103)Pd seeds. METHODS: The MCNP5 radiation transport code was used to estimate air-kerma rate and dose rate distributions with polar and cylindrical coordinate systems. Doses in water and prostate tissue phantoms were compared to determine differences between the TG-43 formalism and realistic clinical circumstances. The influence of Ti encapsulation and 2.7 keV photons was examined. The accuracy of superposition of dose distributions from shorter sources to create longer source dose distributions was also assessed. RESULTS: The normalized air-kerma rate was not highly dependent on L or the polar angle θ, with results being nearly identical between the CivaString and CivaThin sources for common L. The air-kerma strength was also weakly dependent on L. The uncertainty analysis established a standard uncertainty of 1.3% for the dose-rate constant Λ, where the largest contributors were μen/ρ and μ/ρ. The Λ values decreased with increasing L, which was largely explained by differences in solid angle. The radial dose function did not substantially vary among the CivaString and CivaThin sources for r ≥ 1 cm. However, behavior for r < 1 cm indicated that the Au marker(s) shielded radiation for the sources having L = 10, 30, and 50 mm. The 2D anisotropy function exhibited peaks and valleys that corresponded to positions adjacent to (103)Pd wells and Au markers, respectively. Dose distributions of both source types had minimal anisotropy in comparison to conventional (103)Pd seeds. Contributions by 2.7 keV photons comprised ≤ 0.1% of the dose from all photons at positions farther than 0.13 mm from the polymer source surface. Differences between absorbed dose to water and prostate became more substantial as distance from the sources increased, with prostate dose being about 13% lower for r = 5 cm. Using a cylindrical coordinate system, dose superposition of small length sources to replicate the dose distribution for a long length source proved to be a robust technique; a 2.0% tolerance compared with the reference dose distribution did not exceed 0.1 cm(3) for any of the examined source combinations. CONCLUSIONS: By design, the CivaString and CivaThin sources have novel dosimetric characteristics in comparison to Ti-encapsulated (103)Pd seeds. The dosimetric characterization has determined the reasons for these differences through analysis using Monte Carlo-based radiation transport simulations.