A directional 103Pd brachytherapy device: Dosimetric characterization and practical aspects for clinical use

Abstract

Purpose: A brachytherapy (BT) device has been developed with shielding to provide directional BT for preferentially irradiating malignancies while sparing healthy tissues. The CivaSheet is a flexible low-dose-rate BT device containing CivaDots with ¹⁰³Pd shielded by a thin Au disk. This is the first report of a clinical dosimetric characterization of the CivaSheet device.

Methods and materials: Radiation dose distributions near a CivaDot were estimated using the MCNP6 radiation transport code. CivaSheet arrays were also modeled to evaluate the dose superposition principle for treatment planning. The resultant data were commissioned in a treatment planning system (TPS) (VariSeed 9.0), and the accuracy of the dose superposition principle was evaluated for summing individual elements comprising a planar CivaSheet.

Results: The dose-rate constant (0.579 cGy/h/U) was lower than for ¹⁰³Pd seeds due to Au L-shell x-rays increasing the air-kerma strength. Radial dose function values at 0.1, 0.5, 2, 5, and 10 cm were 1.884, 1.344, 0.558, 0.088, and 0.0046, respectively. The two-dimensional anisotropy function exhibited dramatic reduction between the forward (0°) and rearward (180°) directions by a factor of 276 at r = 0.1 cm, 24 at r = 1 cm, and 5.3 at r = 10 cm. This effect diminished due to increasingly scattered radiation. The largest gradient in the two-dimensional anisotropy function was in contact with the device at 92° due to the Au disk shielding. TPS commissioning and dose superposition accuracies were typically within 2%.

Conclusions: Simulations of the CivaDot yielded comprehensive dosimetry parameters that were entered into a TPS and deemed acceptable for clinical use. Dosimetry measurements of the CivaSheet are also of interest to the BT community.

Keywords: CivaSheet; Dosimetry; Monte Carlo methods.

Fig. 1

Cross sectional view of the cylindrically symmetric CivaDot ¹⁰³Pd directional low-dose-rate brachytherapy source, which is a component of the CivaSheet brachytherapy device. The light blue peripheral region is a portion of the flexible polymer base containing the CivaDot. The orange region is the plastic disk (0.253 cm diameter) housing the depression containing the active region (dark blue) that is shielded by a 0.005 cm thick Au disk (yellow). The crosshair centered in the Au disk depicts the coordinate system origin centered on the radio-opaque marker. The unshielded direction is on top while the CivaDot backside is towards the bottom.

Fig. 2

Based on the NIST WAFAC geometry, the photon spectrum of the CivaDot (blue curve) was estimated using the MCNP6 radiation transport code with fluence results expressed in native units (MeV/g/history) for 35% Pd loading. The principal x rays generated following ¹⁰³Pd source disintegration occur at 20 and 23 keV, and the weak emission at 2.7 keV was observed. Also evident were the numerous Au L-shell characteristic x rays from 8 to 15 keV (principally 9.7 keV and 11.5 keV), which contributed to the total air-kerma strength (red curve) and the total number of photons (green curve) by approximately 15% and 4%, respectively.

Fig. 3

Behavior of Monte Carlo-derived radial dose function results using the point-source geometry function as a function of the percentage of Pd loading in the active region inside a CivaDot.

Fig. 4

2D anisotropy function for the ¹⁰³Pd CivaDot brachytherapy source with the reference 35% Pd loading, normalized to unity at θ=0°. The high gradient near the source (small r) and transverse plane (θ~90°) is evident, where the largest gradient was observed at P(0.15 cm, θ=92°). Compared to conventional LDR low-energy seeds, these results were more uniform as a function of r.

Fig. 5

Comparison of the isodose distribution from the VariSeed 9.0 treatment planning system (left) to the overlaid Monte Carlo-derived isodose distribution (right) for a commissioned CivaDot. A distance scale with centimeter increments is at the bottom with the lowest isodose level (1.0%) being about 8 cm wide.

Fig. 6

a) Comparison of the isodose distribution for a CivaSheet comprising a 6×6 array of CivaDots for Monte Carlo simulations (upper) and a test of the dose superposition principle when overlaying dose distributions for 36 CivaDots (lower) positioned at the same locations as in the CivaSheet array. Results are in MCNP6 native units (MeV/g/history) without normalizations or other corrections. b) ratio of the Monte Carlo simulation results to the dose superposition results. Results in the lower image are reflected on X=0 due to geometric symmetry and to facilitate visual comparison with the isodoses depicted in part a).