Dosimetric characterization of GMS BT‐125‐1 125I radioactive seed with Monte Carlo simulations and experimental measurement

Abstract Purpose To investigate the dosimetric characteristics of the new GMS BT‐125‐1 125I radioactive seed, including dose rate constant, radial dose functions, and anisotropy functions. Methods Dosimetric parameters of GMS BT‐125‐1 125I seed including dose rate constant, radial dose functions, and anisotropy functions were calculated using the Monte Carlo code of MCNP5, and measured with thermoluminescent dosimeters (TLDs). The results were compared with those of PharmaSeed BT‐125‐1, PharmaSeed BT‐125‐2 125I, and model 6711 125I seeds. Results The dose rate constant of GMS BT‐125‐1 125I seed was 0.959 cGy·h−1·U−1, with the difference of 0.94%, 0.83%, and 0.73% compared with the PharmaSeed BT‐125‐1 125I seed, PharmaSeed BT‐125‐2 125I seed, and Model 6711 125I seed, respectively. For radial dose function, the differences between the Monte Carlo and the experimental g(r) results were mostly within 10%. Monte Carlo results of g(r) for GMS BT‐125‐1 125I seed were found in agreement (within 3.3%) with corresponding results for the PharmaSeed BT‐125‐2 125I seed. The largest differences were 8.1% and 6.2% compared with PharmaSeed BT‐125‐1 125I seed and model 6711 125I seed, respectively. For anisotropy function, the difference between GMS BT‐125‐1 125I seed and PharmaSeed BT‐125‐2 125I seed was typically <10%. Conclusions The measured dose rate constant, radial dose functions, and two‐dimensional anisotropy functions for the GMS BT‐125‐1 125I seed showed good agreement with the Monte Carlo results. The dose rate constant of the GMS BT‐125‐1 125I seed is similar to that of the PharmaSeed BT‐125‐1 125I seed, the PharmaSeed BT‐125‐2 125I seed, and the model 6711 125I seed. For radial dose functions and two‐dimensional anisotropy functions, the GMS BT‐125‐1 125I seed is similar to the PharmaSeed BT‐125‐2 125I seed but different from the PharmaSeed BT‐125‐1 125I seed and the model 6711 125I seed.


| INTRODUCTION
Brachytherapy, using permanently implanted seeds, has become widely accepted for low-risk prostate cancer. It has been proven to be as effective as surgery or external beam radiotherapy. 1 In China, permanent seed implants have also been used in the treatment of primary and recurrent tumors of the head and neck, lungs, liver, pancreas, rectal, gynecological, and soft tissue. 2 I seed is 3.25 mm, which is different from that in the Model 6711 125 I seed that is 3.00 mm. Although the structure of the GMS BT-125-1 125 I seed is similar to that of the Model 6711 125 I seed, the different length of core markers cause different dose distributions of radioactive seeds. 16 Both GMS BT-125-1 125 I seed and PharmaSeed BT-125-1 125 I seed have a wire core with the length of 3.25 mm that was surrounded by 0.5 lm layer absorbed with 125 I. However, a silver rod is used as the wire core in GMS BT-125-1 125 I seed, while a palladium rod is used in the PharmaSeed BT-125-1 125 I seed. So, the dosimetric parameters of the PharmaSeed BT-125-1 125 I seed cannot be used as a substitute for GMS BT-125-1 125 I seed.
AAPM recommends dosimetry characterization of new low-energy seeds including dose rate constant Λ, radial dose function, g(r), and twodimensional anisotropy function Fðr; hÞ before clinical application using both experimental and Monte Carlo methods. 16,17 TG-43 U1 recommends the updated low-energy photon cross-sections in brachytherapy source dosimetry. The photon cross-sections were updated in the Monte Carlo N-Particle Transport Code 5 (MCNP5). 16,18 The aim of this study is to characterize the dosimetric parameters of a new GMS BT-125-1 125 I seed using both the experimental measurement and Monte Carlo simulation with MCNP5 code.

2.A | Radioactive seeds
The GMS BT-125-1 125 I radioactive seed was manufactured by global medical solutions (GMS) Pharmaceutical Company in Shanghai, China. The structure and dimensions of the GMS BT-125-1 125 I seed are presented in Fig. 1. The core of the radioactive seed consists of a cylindrical silver marker (10.53 g/cm 3 ), with the length of 3.25 mm and diameter of 0.50 AE 0.02 mm. The silver marker is coated with radioactive 125 I with the thickness of 0.5 lm. The length of radioactive 125 I is 3.25 mm. The radioactive 125 I is encapsulated in a hollow titanium (Ti) tube (4.5 g/cm 3 ), which is 4.5 AE 0.5 mm in length and 0.8 mm in external diameter, with a wall thickness of 0.06 mm after welding. The tube is sealed by laser welding of hemispherically shaped ends of a 0.5 mm radius. The conical angle of the end cap is 180°. The Model 6711 125 I seed consists of a 4.5 mm long welded titanium capsule with a thickness of 0.05 mm. The capsule contains a 3.0 mm long silver rod that was surrounded with absorbed 125 I. 16

2.B | Monte Carlo simulation
The Monte Carlo simulation is the most commonly used method to determine the dosimetric parameters of low or high-energy F I G . 1. Schematic diagram of the GMS BT-125-1 125 I seed. radioactive sources. 16,[18][19][20] We used a general Monte Carlo code of MCNP5 developed by Los Alamos National Laboratory in this study. 21 The photon cross-section library used in this code is MCNPLIB04, and the electron cross-section library is EL03. We performed the Monte Carlo simulation of the Model 6711 125 I seed using MCNP5 to compare with the results provided by TG-43U1 for the benchmark of Monte Carlo simulation.

2.C | Experiment set
The accurate geometry and materials of the GMS BT-125-1 125 I seed were described above. The simulated seed was positioned within the center of a water sphere with the diameter of 30 cm. The 125 I photon spectrum was taken from AAPM TG-43U1, which was based on the National Nuclear Data Center spectrum. 22 The spectrum con-  24 To determine the air kerma in the cell, the MCNP *F4 tally (MeV/cm 2 ) was used for determining the energy fluence in the cell. To determine the air kerma, the MCNP's DE/DF card was used. The simulations were performed with 1 9 10 9 histories. The dimensional sizes of the water phantoms and holes simulated in the Monte Carlo simulation were the same with those in the measurement. The conversion factors that convert the results from PMMA phantom to water phantom were obtained using the Monte Carlo simulation as well. The dose rate constant, radial dose function, and anisotropy function were determined in PMMA phantom and liquid water using the Monte Carlo simulation, respectively. The conversion factor from PMMA to liquid water was derived as the ratio of the numbers in liquid water to the numbers in PMMA phantom. 23 The geometry function was calculated using the AAPM Task Group 43 approximation for a line source [eq. (1)].
The radial dose function was determined by calculating the dose rate in water at distances of 0.5, 0.7, and 1.0-10.0 cm with an increment of 0.5 cm from the center of the source using F8 tally. According to the TG43 report, the Monte Carlo simulated radial dose functions were fit to a fifth order polynomial, g r ð Þ ¼ a 0 þ a 1 r þ a 2 r 2 þ a 3 r 3 þ a 4 r 4 þ a 5 r 5 . As described above, the MCNP *F8 tally was used to score energy fluence (MeV) at different distances. The tally voxel size is constant across the distance. Primary particles (5 9 10 9 ) were followed resulting in statistical uncertainty ranging from 0.5% to 2.5% at 0.5 and 10 cm, respectively. The dose rate was normalized to unity at the distance of r 0 = 1.0 cm, and the radial dose function was calculated according to eq. (2) using the geometry function.
The two-dimensional anisotropy function [eq. (3)] was determined by scoring the dose rate in water in spherical volumes located at distance of 0.5, 0.7, 1.0, 1.5, 2.0, 3.0, 4.0, 5.0, 6.0, and 7.0 cm from the center of the source using F8 tally. At a distance of 0.5 cm, the interval of an adjacent scoring point is 20°, while at other distances, the interval is 10°. Due to the small size of each scoring volume, it was necessary to follow 5 9 10 9 primary particles to achieve statistically meaningful results. Statistical uncertainties ranged from 0.5% to 1.1% at 0.5 and 7 cm, respectively. At each radial distance, dose rate was normalized to unity at an angle of 90°. The dosimetric parameters in the TG43U1 were evaluated in liquid water. 16 So, the TLD measured results in PMMA were corrected to those with liquid water as phantom material. The corresponding correction factor, P phant (r, h), was calculated by the following ratio: Immediately before exposure, all dosimeters were annealed for 10 min at 240°C and then rapidly cooled down to room temperature. 25 A Harshaw Model 3500 automatic reader (Thermo Fisher Scientific, Inc., Waltham, MA, USA) was used for reading the dosimeters. The reader was purged with nitrogen, and the following cycle was used: heating at 135°C for 8 s, followed by heating at a constant rate to 240°C in 20 s. Data were collected during the second phase of the readout cycle. 25 Prior to performing the experimental reading, 2-3 unexposed detectors were read to stabilize the readers. TLD readings were converted to absolute dose rate before the calculation of all the dosimetric parameters.

2.D | Thermoluminescence dosimetry
Six different seeds were used in all measurements for calculating the dosimetric parameters. The mean of the measurements was reported in this study.

2.E | Uncertainty of the study
Uncertainty includes type A error that is random error and type B error that is the systematic error. 16 The uncertainties associated with our TLD measurements were estimated by following the guidelines of the AAPM TG-138 and TG-43U1 reports. 16,26 The uncertainty associated with TLD-seed relative positioning should take into account the 0.05 mm machining Carlo, a total relative standard uncertainty was estimated to be 5.4%. Table 2, the estimation took into account the uncertainty associated with the MC Statistics including the uncertainty of air-kerma strength (3.2%) and uncertainty of dose deposition (0.5%), cross-sections (1.5%), seed geometry (2.0%), seed spectrum (0.2%), and seed/TLD positioning uncertainty (3.5%). The relative standard uncertainty for the Monte Carlo calculation was estimated from the relative uncertainties associated with the dose rates calculated in water as a function of radial distance, which increased from 0.5% at 1 cm to 0.6%, 0.8%, and 1.1% at 3, 5, and 7 cm, respectively.

| RESULTS
The calibration curve of TLD was shown in Fig. 3. We found the dose-response of the TLDs was linear below 10 Gy. The manufac-  Table 3 presents the dose rate constants for the GMS BT-125-1 125 I seed and other three commercially available radioactive seeds. 16,27,28 The differences in dose rate constants were 0.94%, 0.83%, and 0.73% when comparing the GMS BT-125-1 125 I seed with PharmaSeed BT-125-1, PharmaSeed BT-125-2 and Model 6711 125 I seeds, respectively. 16,27,28 The radial dose function, g r ð Þ accounted for dose fall-off on the transverse plane due to photon scattering and attenuation. It can also be influenced by the material and encapsulation of the seed. Figure 4 presented the radial dose function comparison of  were typically <10%. For h < 20°or r = 0.5 cm, the differences were larger than 10%. Similar to the radial dose function, the polynomial of the Monte Carlo two-dimensional anisotropy function was The coefficients of the polynomial at different distances are shown in Table 6. | 55 commercial seeds were due to the different material and length of the markers and different shape of end welding.

| DISCUSSION
Due to the limitation of the geometry of the phantom, the uncertainty of the TLD source position is about 3.5%, which is 4.0% in TG43U1 report. 16 The accuracy of the phantom was compromised by enlarging the hole of TLDs by 0.1 mm to load and remove the TLDs smoothly during the experiment. It is difficult for operation if the size of the holes of the phantom is the same as that of the TLD dosimeters.
Based on the study conducted by Safigholi et al, 31 the interseed attenuation effect in multisource implant using Monte Carlo simulation was less than 5%. The interseed attenuation effect should be considered in the clinical dose calculation.

| CONCLUSIONS
This is the first study to determine GMS BT-125-1 125 I seed using This study improved the accuracy of dose calculation using GMS BT-125-1 125 I seed during seed implantation.