Evaluation of (101)Rh as a brachytherapy source.

PURPOSE: Recently a number of hypothetical sources have been proposed and evaluated for use in brachytherapy. In the present study, a hypothetical (101)Rh source with mean photon energy of 121.5 keV and half-life of 3.3 years, has been evaluated as an alternative to the existing high-dose-rate (HDR) sources. Dosimetric characteristics of this source model have been determined following the recommendation of the Task Group 43 (TG-43) of the American Association of the Physicist in Medicine (AAPM), and the results are compared with the published data for (57)Co source and Flexisource (192)Ir sources with similar geometries.
MATERIAL AND METHODS: MCNPX Monte Carlo code was used for simulation of the (101)Rh hypothetical HDR source design. Geometric design of this hypothetical source was considered to be similar to that of Flexisource (192)Ir source. Task group No. 43 dosimetric parameters, including air kerma strength per mCi, dose rate constant, radial dose function, and two dimensional (2D) anisotropy functions were calculated for the (101)Rh source through simulations.
RESULTS: Air kerma strength per activity and dose rate constant for the hypothetical (101)Rh source were 1.09 ± 0.01 U/mCi and 1.18 ± 0.08 cGy/(h.U), respectively. At distances beyond 1.0 cm in phantom, radial dose function for the hypothetical (101)Rh source is higher than that of (192)Ir. It has also similar 2D anisotropy functions to the Flexisource (192)Ir source.
CONCLUSIONS: (101)Rh is proposed as an alternative to the existing HDR sources for use in brachytherapy. This source provides medium energy photons, relatively long half-life, higher dose rate constant and radial dose function, and similar 2D anisotropy function to the Flexisource (192)Ir HDR source design. The longer half-life of the source reduces the frequency of the source exchange for the clinical environment.

Purpose

Brachytherapy procedures have been used for treatment of cancer patients by radiation emitted from small encapsulated sources. In this treatment modality, the sources are placed directly into the tumor or adjacent to the treatment volume. The common radionuclides of brachytherapy sources are 60Co, 137Cs, 192Ir, 198Au, 125I, and 103Pd [1-4]. 192Ir can be accounted as the most commonly radionuclide used for both low-dose-rate (LDR) and high-dose-rate (HDR) brachytherapy sources. Recently, new radionuclides have been considered as alternatives to the above noted sources in brachytherapy treatments. These sources include 131Cs, 169Yb, 57Co, 153Gd, and 170Tm [5-12]. 131Cs was proposed as low dose rate seeds [5]. Radial dose function of 131Cs is similar to those of 125I sources but it has radiobiological advantages in permanent brachytherapy implants over other sources due to its shorter half-life [5, 6]. 169Yb and 170Tm with average energies of 93 keV and 66.39 keV and half-lives of 32 days and 128.6 days, respectively, are interesting for use as HDR brachytherapy sources [7-10]. 57Co emits photons with energies of 122 and 136 keV and half-life of 272 days [11]. Two dimensional (2D) anisotropy function of 57Co is comparable to that of 192Ir and its radial dose function has an increasing trend in some distances, which could lead to larger dose to deeper tissue [11]. 153Gd radionuclide with average energy of 60.9 keV was proposed as a low dose rate or pulsed dose rate brachytherapy source [12]. Enger et al. have shown that this source can be used for interstitial brachytherapy with rotating shield in needle [12]. These new radionuclides, despite having average energies lower than 192Ir, have high enough energy to minimize photoelectric interactions in soft tissue. With these sources, a smaller shielding is required relative to 192Ir sources. The drawbacks of some of the new hypothetical isotopes can be related to their short half-life, beta contamination, and bremsstrahlung contributions. For example, the yield of the photons emission of 170Tm is lower than the yield its electron emission (6 photons per 100 electrons emitted).

The energy of the photon emission and half-life of Rhodium-101 (101Rh) make this radionuclide a possible candidate as a brachytherapy source. 101Rh can be produced by nuclear reactions of 101Ru(d, 2n)101Rh or 102Ru(d, 3n)101Rh in a cyclotron. Prototype radionuclides have been produced at the isochronous cyclotron at the HISKP (Helmholtz-Instituts für Strahlen-und Kernphysik) of the University of Bonn (Germany) [13]. A natural ruthenium foil (with 20 mm2 cross-section and 0.3 mm thickness) was applied as target, which included 101Ru and 102Ru isotopes. A number of typical factors used in this nuclear reaction are: 1) deuteron energy: 27 MeV; cross section: 1000 mbarn; deuteron beam intensity: 1 µA; 2) irradiation time: 2 days; 101Rh activity after cool-down period of 43 days: 1 MBq [13]. Advantages of 101Rh over the existing 192Ir may be due to its relatively longer half-life of 3.3 years, higher specific activity of 397 TBq/g, and lower mean photon energy of 121.5 keV as shown in Tables 1 and 2. These adequate physical characteristics made this radioisotope interesting for this project for evaluation as a possible alternative for the HDR brachytherapy source. The purpose of this work is to evaluate the dosimetric parameters of a hypothetical 101Rh source using task group No. 43 (TG-43) recommendations and verify if it would be a potential HDR brachytherapy source.

Table 1

The photon energy spectrum emitted by 101Rh radionuclide [15]

Assignment	Energy (keV)	Prevalence (%)
γ1(101Ru)	110.94	0.04415
γ2(101Ru)	127.227	68.07
γ3(101Ru)	184.1	0.0591
γ4(101Ru)	197.99	73
γ5(101Ru)	295.01	0.59518
γ6(101Ru)	325.23	11.8311
γ7(101Ru)	422.1	0.19911
Ru Ll	2.253	0.10318
Ru Lη	2.382	0.0477
Ru Lα2	2.554	0.274
Ru Lα1	2.558	2.44
Ru Lβ1	2.683	1.1818
Ru Lβ4	2.741	0.09123
Ru Lβ6	2.763	0.015624
Ru Lβ3	2.763	0.133
Ru Lβ2	2.836	0.224
Ru Lγ1	2.965	0.08613
Ru Lγ2	3.181	0.0154
Ru Lγ3	3.181	0.0277
Ru Kα3	18.893	0.000653
Ru Kα2	19.15	22.88
Ru Kα1	19.279	43.315
Ru Kβ3	21.634	3.5512
Ru Kβ1	21.657	6.8723
Ru Kβ5	21.836	0.04112
Ru Kβ2	22.074	1.615
Ru Kβ4	22.115	0.33616

Table 2

Characteristics of the 57Co, 101Rh and 192Ir sources

	57Co	101Rh	192Ir
Specific activity (TBq/g)	312	397	341
Half-life	272 days	3.3 years	74 days
Photon energy (keV)	122, 136	121.5 (average)	360 (average)
Yield (photons/disintegration)	1.61	2.37	2.36
Air kerma strength per activity (U/mCi)	0.46	1.09 ± 0.01	3.62
Dose rate constant (cGy/(hU))	1.215	1.18 ± 0.08	1.114
Half-value layer in lead (mm)	0.298	0.0331	2.97

Material and methods

Design for the hypothetical 101Rh source

In this project, the geometric structure of the hypothetical 101Rh source was designed to be similar to the Flexisource 192Ir HDR source [14], just for the ease of the comparison of the dosimetric parameters between the two sources. The design and dimensions of 101Rh source is shown in Figure 1. The active core of the source is a pure 101Rh cylinder (density of 12.41 g/cm3) with an active length of 3.5 mm and an active diameter of 0.6 mm. The active core is covered by a 304 stainless steel capsule (composition by weight – Fe: 67.92%, Cr: 19%, Ni: 10%, Mn: 2%, Si: 1% and C: 0.08%; density: 8.0 g/cm3). The outer dimensions of the source are 0.85 mm in diameter and 4.6 mm in total length. The 304 stainless steel cable (density of 8.0 g/cm3) has been considered as a cylinder with a length of 5 mm and a diameter of 0.5 mm. The energy spectrum of the 101Rh photons considered in this study is listed in Table 1, [15].

Fig. 1

Geometric design of the 101Rh hypothetical source (all dimensions are in millimeters). This figure is not to scale

Calculation of TG-43 dosimetric parameters

Based on updated TG-43 formalism [2], dose rate at a point is calculated from the following formula:

In the above formula, S , Λ, G(r,θ), g(r) and F(r,θ) present air kerma strength, dose rate constant, geometry function, radial dose function, and 2D anisotropy function, respectively. The MCNPX Monte Carlo (MC) code (version 2.4.0) [16] was used for obtaining dosimetric parameters of the new source design. In the calculations, only photons emitted by the 101Rh were defined in the source definition card. It was assumed that the beta particles emitted by the source (ranging 118.78-541 keV with the most probable energy of 215.77) are absorbed by the encapsulation of source, therefore they were ignored.

For estimation of air kerma strength, a torus with a minor diameter of 0.5 mm and major radii of 20 cm was defined on the transverse plane of the hypothetical source. F6 tally, which is a commonly used tally in MCNP simulations, was used for these calculations and the results were obtained in terms of kerma of the source in MeV/(g.photon). This torus is composed of air and the surrounding phantom material is composed of vacuum. Air kerma strength then was calculated from the air kerma strength formula presented in the updated TG-43 formalism using the kerma value and the corresponding units and other conversion factors (such as eV to joule, grams to kilograms, photon yields, etc.).

For determination of dose rate constant, the hypothetical source was positioned at the center of a spherical water phantom with radius of 50 cm in the simulations. A torus with 0.5 mm minor diameter and major radii of 1.0 cm was defined on the transverse plane and *F8 tally was calculated. This torus was also composed of water. Using *F8 tally (in terms of MeV) in MCNP it is feasible to score the absorbed energy inside a tally cell. Therefore, dose rate is output of *F8 tally divided by the mass of the torus considering the units for conversion factors (for example, eV to joule, grams to kilograms, yields, etc.). Dose rate constant then was obtained as the dose rate divided by the air kerma strength.

As an example for conversion of the MC output (for *F8 tally divided per mass in terms of: MeV/g per photon) to absorbed dose rate in units of cGy/(h.mCi) the following formula can be used:

with this regard, the value of photon yield of the 101Rh radionuclide is equal to 2.37 photons/dis. This value can be obtained from summation and normalization of the prevalences listed in Table 1. The activity of 101Rh source, used in this conversion was equal to 1.0 mCi.

In order to calculate radial dose function of the new source design, a spherical water phantom with 50 cm radius was defined with the source located at the center of the phantom. Torus voxels were considered at distances of 0.1-15.0 cm on the transverse plane of the source in the phantom. The thicknesses of the tori were 0.4 mm at distances up to 1.0 cm, and 1.0 mm for other distances up to 15.0 cm. The outputs of *F4 tally was used for calculation of dose rate at various radial distances. To have an acceptable level of uncertainty, this type of tally was used to speed up the calculations in the MCNP simulations. A line-source approximation was applied in calculation of two dimensional geometry functions. Finally, radial dose function was calculated from dose rate and geometry function values at various distances.

For calculation of 2D anisotropy function, the same phantom (i.e. spherical water phantom with 50 cm radius) was defined as that described in calculation of radial dose function. For these simulations, the calculation points were located on a circular path around the source with the same radial distances, and different polar angles, ranging from 0° to 180° angles at 5 degrees intervals. These points were divided into two groups, those that were along the longitudinal axis of the source, and those that were off the longitudinal axis. Spherical voxels with 0.1 cm diameter were utilized to score the dose on the longitudinal axis of the source. For the other points, tori with minor diameter of 0.4 mm were defined at radial distances of 0.5 cm and 1.0 cm. The minor diamaters of these tori were 1 mm for larger distances (larger than 1.0 cm, up to 15.0 cm). The reason for different tally voxel sizes at various distances is due to the presence of steep dose gradient as a function of distance from the source, which is an inherent effect in brachytherapy. Therefore, using larger voxel size at shorter distances may introduce artifacts in the calculations by volume-averaging. The outputs of *F4 was used for calculation of 2D anisotropy functions. From the selected tallies at various distances, the 2D anisotropy function values were calculated for distances ranging between 0.5-15.0 cm and for angles between 0° to 180° with 5.0 degrees intervals. In calculation of TG-43 parameters, toroidal cells were used to comply with the cylindrical symmetry of the dose distribution. An attempt was also made to use toroid cells not having large dimensions at various reference directions to avoid volume-averaging artifacts. As an example, for calculation of anisotropy function at 2.0 cm distance and 30° angle, a torus with minor diameter of 1 mm at horizontal and vertical directions was utilized. It should be noticed that in these calculations, the source delivery cable is coincident on the angle of 180 degrees.

In all calculations, the energy cut off of 1 keV was defined for photons and electrons. For calculation of TG-43 dosimetric parameters, except for 2D anisotropy function, the input files were run for 2.5 × 107 particles. For calculation of 2D anisotropy function, in order to reduce the calculation uncertainties at larger distances from the source, the input file was run for 108 photons. The maximum type A errors or statistical uncertainties in Monte Carlo calculations for air kerma strength, dose rate constant and radial dose function over all the evaluated distances were 0.3%, 3.32% and 0.44%, respectively. The reason for having larger uncertainty in dose rate constant than the other two parameters was due to the use of *F8 tally in this case compared to F6 or *F4 tallies for the others. For similar particle scoring numbers, using *F8 tally in MCNP normally incorporates to larger uncertainty compared to the other tallies. The maximum type A uncertainty in calculation of 2D anisotropy function was 4.42%. It should be noted that these values are type A uncertainties, by having ignored type B uncertainties and with a coverage factor (k) of 2.0, the expanded uncertainties (U; U = ku c, in which u c is the combined uncertainty) will be twice these values. A coverage factor of 2.0 corresponds to 95% confidence level. With 95% confidence level, the true value is in the “calculated value – 2U” – “calculated value + 2U” interval with the probability of 0.95. The input programs were run using a personal computer having 64-bit Windows 7.0 operating system, 3.20 GHz Intel (R) Core i7 CPU, and 2.00 GB RAM. With this computer, the running time for the air kerma strength, dose rate constant, radial dose function, anisotropy function, with the aforementioned numbers of particles histories, was 5.25 h, 24.75 h, 25.50 h, and 134.25 h, respectively. Finally, the TG-43 dosimetric parameters of the hypothetical 101Rh source were compared with the data for another hypothetical 57Co source [17] and Flexisource 192Ir source [14].

Results

Air kerma strength per activity and dose rate constant values for the hypothetical 101Rh source were obtained, and found to be 1.09 ± 0.01 U/mCi and 1.18 ± 0.08 cGy/(hU), respectively. The errors in these values are reported with a coverage factor of 2σ. A comparison of other characteristics of these three sources is shown in Table 2. These characteristics include specific activity, half-life, average photons energy, etc. It can be noticed that the specific activity was calculated theoretically from: S = (λN A)/M.

Radial dose function of the hypothetical 101Rh source as a function of radial distance is shown in Table 3. 2D anisotropy function values of the hypothetical 101Rh as a function of radial distance and polar angles were listed in Table 4.

Table 3

Radial dose function for 101Rh hypothetical source

Radial distance r (cm)	g L(r)
0.1	0.944
0.2	0.953
0.3	0.961
0.4	0.966
0.5	0.968
0.6	0.973
0.7	0.978
0.8	0.985
0.9	0.995
1.0	1.000
1.5	1.021
2.0	1.045
2.5	1.062
3.0	1.080
3.5	1.099
4.0	1.118
4.5	1.126
5.0	1.139
5.5	1.147
6.0	1.154
6.5	1.161
7.0	1.162
7.5	1.164
8.0	1.158
8.5	1.163
9.0	1.153
9.5	1.149
10.0	1.139
12.0	1.103
15.0	1.013

Table 4

2D anisotropy function for 101Rh hypothetical source

θ (degree)	Distance r (cm)
	0.5	1.0	3.0	4.0	5.0	7.0	10.0	12.0	15.0
0	–	0.681	0.738	0.775	0.795	0.847	0.892	0.903	0.909
5	0.709	0.708	0.770	0.798	0.818	0.843	0.899	0.869	0.922
10	0.756	0.751	0.814	0.835	0.845	0.862	0.894	0.910	0.918
15	0.812	0.812	0.860	0.873	0.884	0.902	0.912	0.923	0.930
20	0.862	0.859	0.893	0.900	0.909	0.920	0.933	0.934	0.945
25	0.896	0.892	0.918	0.920	0.922	0.927	0.947	0.939	0.958
30	0.917	0.915	0.936	0.941	0.941	0.941	0.957	0.945	0.959
35	0.940	0.936	0.948	0.946	0.945	0.956	0.964	0.964	0.967
40	0.955	0.951	0.963	0.963	0.960	0.967	0.973	0.968	0.975
45	0.966	0.963	0.973	0.974	0.967	0.975	0.983	0.974	0.985
50	0.976	0.972	0.982	0.981	0.979	0.981	0.988	0.979	0.980
55	0.984	0.976	0.979	0.981	0.979	0.977	0.989	0.985	0.987
60	0.989	0.984	0.993	0.989	0.986	0.982	0.994	0.984	0.993
65	0.992	0.988	0.994	0.996	0.991	0.987	0.989	0.990	0.994
70	0.997	0.989	0.997	0.995	0.994	0.995	1.001	0.990	0.995
75	0.999	0.994	1.000	0.996	0.995	0.991	1.000	0.993	0.997
80	1.004	0.995	1.000	0.998	0.997	0.994	1.003	0.998	1.001
85	1.001	0.991	1.000	0.998	0.995	0.991	1.002	0.993	0.995
90	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
95	1.005	0.997	0.997	0.997	0.993	1.001	1.004	0.991	0.996
100	1.001	0.994	1.003	1.000	1.000	0.994	1.004	0.995	0.998
105	0.998	0.993	0.998	0.997	0.996	0.992	0.995	0.992	1.001
110	0.996	0.988	0.997	0.996	0.992	0.993	1.003	0.991	0.997
115	0.994	0.988	0.993	0.994	0.991	0.993	0.996	0.985	0.995
120	0.989	0.979	0.989	0.990	0.988	0.992	0.992	0.988	0.992
125	0.981	0.975	0.987	0.988	0.985	0.987	0.991	0.976	0.986
130	0.975	0.970	0.978	0.978	0.977	0.972	0.989	0.973	0.981
135	0.967	0.962	0.970	0.971	0.969	0.970	0.982	0.981	0.978
140	0.955	0.947	0.961	0.959	0.956	0.963	0.972	0.967	0.972
145	0.935	0.928	0.949	0.948	0.949	0.952	0.960	0.955	0.963
150	0.924	0.921	0.928	0.939	0.940	0.945	0.955	0.950	0.964
155	0.897	0.889	0.916	0.916	0.923	0.935	0.949	0.952	0.957
160	0.863	0.858	0.896	0.900	0.910	0.919	0.936	0.933	0.953
165	0.821	0.816	0.862	0.874	0.880	0.903	0.918	0.921	0.954
170	0.760	0.761	0.808	0.827	0.836	0.864	0.897	0.888	0.900
175	–	0.659	0.746	0.767	0.781	0.828	0.848	0.891	0.901
180	–	0.509	0.635	0.678	0.720	0.763	0.820	0.834	0.869

A comparison of the radial dose function of the hypothetical 101Rh source with the published data for the hypothetical 57Co source, and Flexisource 192Ir source is shown in Figure 2. Moreover, Figure 3 shows the comparison between the 2D anisotropy function of the hypothetical 101Rh source and the values from hypothetical 57Co source and Flexisource 192Ir source, at the distances of 0.5, 1.0, 5.0, and 10.0 cm from the source. The data presented in Figures 2 and 3 for the hypothetical 57Co source was obtained via personal communication [17]. The data for the Flexisource 192Ir source in these figures were extracted from the data reported by Granero et al. [14].

Fig. 2

A comparison of the radial dose function for the hypothetical 101Rh source, hypothetical 57Co source and Flexisource 192Ir source

Fig. 3

A comparison of the 2D anisotropy function for the hypothetical 101Rh source, hypothetical 57Co source and Flexisource 192Ir source at the distances of (A): 0.5 cm; (B): 1.0 cm; (C): 5.0 cm; (D): 10.0 cm from the source

Discussion

Dosimetric comparison with other radionuclides

In the present study, TG-43 dosimetric parameters of a hypothetical 101Rh source were calculated and compared with those of a hypothetical 57Co and a commercially available Flexisource 192Ir sources. Based on the data in Table 2, air kerma strength per mCi activity for the hypothetical 101Rh source is higher (factor of 2.37) than that of the hypothetical 57Co source and less (factor of 3.32) than that of 192Ir source. Since all the three sources have the same geometrical structure, this effect is due to the differences in yield and specific activities, energy spectra of photons emitted from the radionuclides and self-absorptions inside the active cores. Higher air kerma strength per mCi is an advantage for 192Ir brachytherapy source over 101Rh. However, this source (101Rh) has more than twice air kerma strength per mCi higher than that for 57Co. Air kerma strength per mCi activity indicates the level of self-absorption inside a source, and depends on the energy spectrum of a radionuclide and source design. On the other hand, dose rate constant of the hypothetical 101Rh source is approximately the same as than those for 57Co and 192Ir sources. Therefore, it can be concluded that per U of air kerma strength, these sources can release the same dose rates at the reference distance (1.0 cm) from the source in water.

Based on the data presented in Figure 2, radial dose function for the hypothetical 101Rh source is larger than the values for 192Ir source for distances greater than 1.0 cm. This comparison indicates that there would be a larger dose delivered to the tissues located at larger distances from the hypothetical 101Rh source than the 192Ir source. For example, at a distance of 8.0 cm, the radial dose functions of the 101Rh and 192Ir sources are 1.16 and 0.96, respectively. Therefore, at this distance, 101Rh delivers approximately 21% larger dose than 192Ir source. This is an excellent advantage for brachytherapy treatment of many different cancer patients, such as cervical or deep seated vaginal cancer. Moreover, the skin sparing characteristics of the 101Rh source may become superior to the 192Ir sources for some treatments such as the breast cancer with AccuBoost system [18]. Interestingly, this graph indicates that the 57Co source delivers larger dose than both 101Rh and 192Ir sources, which can be accounted as an advantage for this source.

Figure 3 shows that, except at very small angles (< 20 degrees), the 2D anisotropy function data of 101Rh very similar to the 192Ir and 57Co data for all distances. 2D anisotropy function explains the non-isotropic feature of dose distribution around the source due to self-absorption within the source and distribution of the activity with a linear pattern.

It should be clear that the present results of TG-43 parameters for the 101Rh source are only valid for the geometry design that is defined in this study and they cannot be used for the clinical purposes. Therefore, the dose distribution and the TG-43 parameters will change by variation in the source design for the hypothetical 101Rh sources. In other words, the results of this work, especially anisotropy, depend on the structural details of the source design. Moreover, anisotropy cannot be used as a criterion for advocating the use of a new radionuclide for brachytherapy. There are cases, in which highly anisotropic sources can be used effectively, provided their anisotropy function is accurately known. This leaves air kerma strength, dose rate constant, and radial dose function as TG-43 quantities meaningful for the evaluation of a candidate radionuclide for brachytherapy. Their appropriateness, however, should be assessed in the context of a specific application. As a sample, the study by Lymperopoulou et al. [19] compares 169Yb with 192Ir for use as sources in prostate brachytherapy. The same evaluations can be performed for the 101Rh radionuclide for application in prostate brachytherapy. In the following text, some aspects of use of this source in interstitial rotating shield brachytherapy (I-RSBT) are described.

Based on the data presented in Table 2, the energy of photons emitted by 101Rh is on average lower than 192Ir. Therefore, for medium energy photons emitted by 101Rh, the required thickness of the shielding (with half-value layer of 0.0331 mm lead [20]) is much lower than that needed for protection against 192Ir brachytherapy source (2.97 mm lead). It should be noticed that having compared only the average photon energies of these sources (Table 2), the HVL differences are not justified. Therefore, considering the energy spectra of the sources can be illuminating. The lower thickness for shielding requirement will reduce the costs of treatment room shielding for the 101Rh source. Other advantages of 101Rh are relatively long half-life of 3.3 years and high specific activity of 397 TBq/g. These suitable characteristics made 101Rh radionuclide interesting for application as a brachytherapy source. On the other hand, air kerma strength per mCi activity of the 101Rh source with the proposed geometry is lower than that of 192Ir source. While this effect will depend on the geometries of these two sources with the assumed geometries, this can be an advantage of 192Ir over 101Rh. This is because that with the same and specific activities for these two sources, with 192Ir the source strength will be higher and this is corresponded to a lower treatment time duration in a single treatment session.

As listed in Table 2, dose rate constant of the 101Rh is 1.18 cGy/(h.U), which is approximately 6% larger than the 192Ir value of 1.114. Therefore, the 101Rh with the specified source design is able to deliver approximately 6% larger dose per air kerma strength than the 192Ir source. The significance of this difference should be evaluated by considering the radiobiology of the treatment site.

Enger et al. [12] has reported a dose rate at 1.0 cm in water of 4.18 × 104 cGy/h for a VariSource 192Ir source with activity of 370 GBq (i.e. equivalent to 10 Ci). This amounts to a dose rate per activity of 112.97 cGy/(h.GBq). Based on our Monte Carlo calculations, this quantity for the 101Rh hypothetical source with a Flexisource design is equal to 35.29 cGy/(h.GBq). While this variable for 101Rh source is lower than that of 192Ir source, the source designs should be taken into account, since VariSource 192Ir source has an active length of 10.0 mm but Flexisource design have a 3.5 mm one. This comparison was performed roughly for these two sources and a precise comparison should be performed with sources with the same geometries because self-absorption inside a source and encapsulation geometry and composition will affect the dose rate at 1.0 cm distance inside a water phantom. Generally, a lower dose rate per activity at 1.0 cm will need to higher activity or higher treatment time for a standard prescribed dose regime, and, therefore, for having a reasonable treatment time a multiple-source choice may be relevant. For having a higher activity, the production costs will be higher due to longer irradiation time needed for the target in a cyclotron or nuclear reactor. With this regard, a comparison more consistent with TG-43 formalism can be performed with comparison of dose rate constants of the sources. Since in dose rate constant, dose rate is normalized to air kerma strength (instead of activity in mCi) and having mCi to U conversion factors for the 192Ir and 101Rh sources is necessary to have a conversion from dose rate at 1.0 cm per activity to dose rate constant.

In a study by Lymperopoulou et al. [21], Monte Carlo simulation was utilized and 169Yb was compared to 192Ir for breast HDR brachytherapy with multiple catheter implants. The results using dose volume histogram indicated that 169Yb could be at least as effective as 192Ir in delivering the same dose to the lung and with slightly less dose to the skin on breast. The finding implied that for the sources with intermediate photon energy such as 169Yb, there is a need for the modification of calculation algorithms used in clinical treatment planning applied in particular brachytherapy practices. Lymperopoulou et al. [19] assumed a hypothetical 169Yb source in evaluation of the use of this radionuclide for prostate HDR brachytherapy. 169Yb proved to be at least equivalent to 192Ir, independent to the prostate volume in the situation that the radiation scattering be overcompensated for absorption in intermediate energies and distances in prostate HDR brachytherapy. Having reviewed the methods used in the aforementioned studies, and the point that the 101Rh source was evaluated only from TG-43 and general dosimetric specifications, it is recommended that this radionuclide be evaluated further in comparison with 192Ir in brachytherapy applications for various cancers.

Production of 101Rh

One may need to consider two other aspects of the 101Rh source in application of this source in brachytherapy. One aspect is the costs of production of this radioisotope. Normally, radioisotopes are produced by a cyclotron or a nuclear reactor. It is mentioned in the introduction section that 101Rh is produced by deuteron irradiation of isotopes of Ru. Theoretically, there are also other possible reactions for 101Rh production. Some examples are: 103Rh(n, 3n)101Rh; 102Pd(n, np)101Rh and 102Pd(n, d)101Rh. 103Rh and 102Pd are natural isotopes of ruthenium and palladium, respectively. These reactions can be achieved by neutron irradiation of the target in a nuclear reactor. It should be noticed that these reactions require fast neutrons and are not accounted as fission reactions. The cross sections of these deuteron or neutron reactions are of importance when having enough activity of the 101Rh product is aimed. Recently, a study was performed and it was disclosed that with 100 h cyclotron run time having 1 µA beam current, about 900 MBq of 101mRh can be produced [21]. In other words, the yield for production of 101mRh is 8.7 MBq/µAh [21]. 101mRh is metastable and decays to 101Rh following gamma emission. A maximum beam current of 1 µA could be achieved in the experiments with Bonn cyclotron. However, cyclotrons with higher currents are also available. ISPRA cyclotron is an example having higher beam current. 101mRh decays to 101Rh with a half-life of 4.34 days, and due to the short half-life of 101mRh and long half-life of 101Rh, production of 101mRh with a cyclotron can be an alternative for achievement of 101Rh [22]. As another pathway for production of 101Rh, it was also reported that irradiation of 89Y by 12C can be proposed as a pathway to produce 101Rh. 89Y is the natural isotope of yttrium with 100% abundance. This reaction was obtained by 12C beam with average energy of 49.5 MeV [23]. An issue, which could be considered with this process is the contamination of 101Rh by other radionuclides and its influence on the TG-43 parameters. The cross section of 101Rh as compared to those of 192Ir is relevant in these considerations. 192Ir is produced as a fission product in a nuclear reactor by irradiation of the target by thermal neutrons. Production of a radioisotope in a nuclear reactor result to lower production costs compared to a cyclotron. Although, due to lower energy photons emitted by 101Rh relative to 192Ir, the construction costs of shielding for treatment room may be lower, the production costs for 101Rh source may be more than that of 192Ir source. This issue should be considered before application of the 101Rh radioisotope in brachytherapy.

Use of 101Rh in interstitial rotating shield brachytherapy

I-RSBT method was evaluated for 153Gd radioisotope in a study by Adams et al. [24]. In that study a novel needle, catheter and source system was presented for (I-RSBT) application in brachytherapy of prostate. Their justification for application of a shielded source and catheter system was their aim to reduce the dose received by urethra, rectum, and bladder. The reason for use of 153Gd source instead of 125I and 192Ir sources was that 125I sources has normally rapid dose fall-off in soft tissue and the thickness for shielding of 192Ir sources would be large and cannot be fitted in catheters, which are normally used in prostate brachytherapy. Similar to 153Gd, the same notifications can be considered for 101Rh source. However, 101Rh emits a number of relatively high energy photons albeit with lower prevalence (Table 1). The average photon energy and half-value layer (HVL) in lead for 101Rh is much lower than that of 192Ir. A comparison between the HVL for 153Gd and 101Rh may be interesting for use in I-RSBT. The HVL (in mm Pb) for 153Gd and 101Rh are 0.0783 and 0.0331, respectively [20]. It is evident that with 101Rh a lower thickness of shielding is required in this method. However, a quantitative evaluation of shielding requirements for application in I-RSBT as a subject of further research in this field and will be interesting.

Limitations and suggestions for further research

The evaluations on the 101Rh radionuclide could be accomplished simply using point source approximation that is time efficient but still meaningful with the purpose of only indication of an effect. A number of previous studies have used this approximation in their evaluations [25]. In a report on dose calculation in brachytherapy, it was proposed as a consideration for high energy photon emitting brachytherapy sources that modeling of sources using point source approximation is facilitated by averaging dose anisotropy over all angles. This method of calculation can be used in permanent prostate brachytherapy dose calculation, in which seed orientation is not distinguishable for clinical non-stranded application due to the large number of seed orientations [26]. Utilizing a point source could be easier and faster to perform but we preferred to execute a precise evaluation on this source similar to other articles on hypothetical sources by simulation of the source in its complete geometry [10-12].

Based on the AAPM and ESTRO report on high-energy sources [26], spherical phantom radius of 40 cm was recommended. The size of the spherical phantom used in this study is 50 cm and therefore the full scatter condition is obtained. This report also announced recommendations on maximum voxel sizes, which are being used for scoring the dosimetric variables to minimize the volume-averaging artifacts: (0.1 mm)3 voxels for r ≤ 1 cm; 0.5 × 0.5 × 0.5 mm3 voxels for 1 cm < r ≤ 5 cm; 1 × 1 × 1 mm3 voxels for 5 cm < r ≤ 10 cm and 2 × 2 × 2 mm3 voxels for 10 cm < r ≤ 20 cm. In the present study, voxels are not cubic and are in the form of toroidal cells with 0.4 mm in thickness for r < 1 cm and with 1 mm thickness for the other distances. There are minor differences between our methodology for the thickness of the tally cells and those recommended by the AAPM and ESTRO report at close distances from the source, and this point should be noticed in the further methodologies.

The spectral data of 101Rh radionuclide used in the simulations in this study (Table 1) were extracted from a database presented by Lund University [15]. On the other hand, a joint report by American Association of Physicists in Medicine (AAPM), and European Society for Therapeutic Radiology and Oncology (ESTRO) [26] recommends that National Nuclear Data Center (NNDC) data be used in application of brachytherapy sources, which are clinically related. NNDC includes three datasets for energy spectra of the 101Rh radionuclide [26]. In a study by Rivard et al. [28], the influence of the choice of energy spectrum on kerma and dose distributions of three brachytherapy sources was evaluated. It was observed that there were water-kerma differences of about 2%, 2%, and 0.7% with various spectrum choices for 192Ir, 125I, and 103Pd sources, respectively. Furthermore, the influence of photon spectrum on the dose rate constant and the radial dose function ranged from 0.1-2%. As a rough evaluation, the photon yields from the spectrum used in this study (Table 1) and the dataset No. 3 of NNDC [27] for 101Rh are 2.3705 photons/dis. and 2.3670 photons/dis., respectively. The details are not presented herein but the spectrum used in the present study for the 101Rh has more detailed energies than that announced by NNDC. While it is predicted that relatively the same results will be obtained with various choices of reference spectra for the hypothetical 101Rh source, it is recommended that the spectrum from the NNDC website be used in the future studies on this source.

Treatment planning in brachytherapy has advancements starting from simple look-up tables up to computerized dose calculation algorithms. The current algorithms are based on the TG-43 formalism with recent advances in calculation of dose distributions for single sources. However, this formalism has limitations for calculation of patient dose. Various dose calculation algorithms are being developed based on: Monte Carlo methods, collapsed cone, etc. The scopes of current advancements in brachytherapy include: improved dose calculation tools, planning systems to account for heterogeneities, scattering conditions, radiobiology, and image guidance brachytherapy [29]. Following literature reports announcing the deficiencies involved in the approximations of conventional brachytherapy dosimetry, model-based dosimetry algorithms were incorporated in commercial brachytherapy treatment planning systems. The primary calculations of these algorithms are defined, having criteria established by the developers with the purpose of optimization of computation speed and accuracy. On the other hand, a basic realization of the limitations of these algorithms in commissioning step and their further evaluations compared to the conventional ones is necessary [30]. Task Group No. 186 provided guidance for early adopters of model-based dose calculation algorithms for brachytherapy users. Dose calculation accuracy in brachytherapy highly depends on the scatter conditions and photoelectric cross-sections relative to water. In some situations, differences between the TG-43 and model-based algorithms can lead to dose differences exceeding a factor of 10. Model-based dose calculation algorithms raise the major aspects, which are not addressed by current guidelines: dose sensitivity to the dose specification medium, dose calculation for the local medium in heterogeneous medium, and the dose in a small volume of water in heterogeneous medium. These issues are changed as patient-specific [31]. While in the present study 101Rh was evaluated from only general and TG-43 dosimetric parameters, further evaluation of this source from the view of model-based dose calculation algorithms can be a subject of more complementary studies.

Conclusions

Advantages of 101Rh to 192Ir are having relatively longer half-life (3.3 years versus 74 days), higher specific activity (397 TBq/g versus 341 TBq/g)k, and very low half-value layer (0.0331 mm in lead versus 2.97 mm of lead). These adequate physical characteristics make this radionuclide interesting as a possible brachytherapy source. Air kerma strength per activity for hypothetical 101Rh source is about twice than that of hypothetical 57Co source and it has a dose rate constant comparable to hypothetical 57Co and 192Ir sources. Radial dose function for the 101Rh hypothetical source is greater than that of 192Ir source for distances greater than 1.0 cm. The 2D anisotropy function of 101Rh is very similar to that of 192Ir, which can be taken into account as another advantage of this new proposed source. With these suitable physical properties, 101Rh could be considered as potential candidate in brachytherapy.