Monte Carlo-based dose calculation for (32)P patch source for superficial brachytherapy applications.

Skin cancer treatment involving (32)P source is an easy, less expensive method of treatment limited to small and superficial lesions of approximately 1 mm deep. Bhabha Atomic Research Centre (BARC) has indigenously developed (32)P nafion-based patch source (1 cm × 1 cm) for treating skin cancer. For this source, the values of dose per unit activity at different depths including dose profiles in water are calculated using the EGSnrc-based Monte Carlo code system. For an initial activity of 1 Bq distributed in 1 cm(2) surface area of the source, the calculated central axis depth dose values are 3.62 × 10(-10) GyBq(-1) and 8.41 × 10(-11) GyBq(-1)at 0.0125 and 1 mm depths in water, respectively. Hence, the treatment time calculated for delivering therapeutic dose of 30 Gy at 1 mm depth along the central axis of the source involving 37 MBq activity is about 2.7 hrs.

Introduction

Basal cell carcinoma is one of the most common skin cancers, occurs mostly in middle aged people, and is more probable for the fair complexion people.[1] The treatment modalities for skin cancers are surgical excision, radiotherapy and chemotherapy. Each treatment modality has its own advantages and disadvantages. Removing the affected area by surgical excision is usually preferred in many cases, but the recurrence rates after treatment are high. Radiotherapy treatment using external beam therapy is too expensive and it also delivers unnecessary dose to underlying normal tissues. Chemotherapy has its own side effects.

Mould or superficial brachytherapy is a promising alternative treatment method for such skin cancers, where high-energy beta emitting radio-nuclides such as 32P,90Sr/90Y,188Re are used to overcome the disadvantages of radiotherapy and surgery. In superficial brachytherapy, prescribed dose can be delivered to the affected area without excessive damage to the neighboring normal tissues. This technique is simple, less trauma to patients, and less expensive as compared to external beam therapy.

Lee et al., introduced the treatment of skin cancer and Bowen's disease using beta emitting 165Ho-impregnated patch sources.[2] Successful tumor destruction was observed both in animal and human studies. Mukherjee et al., in their studies evaluated 90Y skin patches and 188Re radioactive bandages for therapy of superficial tumors in mice.[34] Treatment of skin cancer using 188Re-labeled paper patches has been reported by Jeong et al.[5]

Pandey et al., reported the use of 32P cellulose-based adsorbent paper skin patches to control the tumor regression in C57BL6 mice bearing melanoma.[6] Park et al.,[7] studied the use of 32P ophthalmic applicator after pterygium and glaucoma surgeries. They demonstrated that dose distributions obtained using the 32P source is beneficial for reducing the incidence rate of radiation-induced cataract and it can deliver therapeutic doses to the surface of the conjunctiva while sparing the lens better than the 90Sr/90Y applicators. Xu et al.,[8] investigated the therapeutic effects of the chromic phosphate particle-based 32P source in a rabbit VX 2 lung tumor animal model and found that the tumor volume significantly decreased after implantation of source particle.

Salguerio et al., designed 32P brachytherapy patch source (1 mm in height × 5 mm in dia.) for skin diseases using phosphoric acid and chromic phosphate in combination with natural rubber or silicone and evaluated its therapeutic efficacy.[910] They reported arrest of tumor growth and complete regression of tumor in some cases with 40 Gy of single-dose scheme in animal studies. They estimated the dose rate at selected depths (0.0001, 0.01, 4 and 7.5 mm) using the Monte Carlo-based MCNP5 code.[1011] The activity per unit area considered in their calculations was 10.6 MBqcm-2. The surface area of the source was 0.196 cm2. Hence, the total activity of the source considered in their work was 2.081 MBq. We repeated their study using the DOSRZnrc user-code.[12] The dose rate values showed a good agreement for 0.0001 and 0.01 mm depths. For 4 and 7.5 mm depths, the published values were higher by a factor of about 22 and 3.6 × 104, respectively. We concluded that this large discrepancy in the dose rate values at 4 and 7.5 mm depths published by Salguerio et al.,[9] was due to possible systematic error in their Monte Carlo calculations.[13]

32P is a suitable radioisotope for such therapeutic application due to many advantages over other beta emitting radioisotopes. It is a pure beta emitter with maximum energy of 1.71 MeV. Its half-life is 14.2 days. Hence it is less hazardous material from transportation, storage, and waste disposal point of view. The maximum range of 32P beta particle in soft tissue is 8 mm (the average range is 3 mm).[14] Due to its short range, there will be negligible radiation dose to the underlying healthy normal tissues and bone.

Isotope Production and Applications Division, BARC has indigenously developed nafion–zirconium phosphate film-based 32P patch source for superficial brachytherapy applications.[15] A nafion-117 membrane of thickness 100 μm is treated with ZrOCl2 solution, and subsequently dipped in orthophosphoric acid. These radioactive 32P patches are cut in to 1 cm × 1 cm sizes and then subsequently laminated with thermoplastic polyurethane sheets of thickness 40 μm. The above preparation method is robust, inexpensive and reproducible and complies with the safety standard stipulated by Atomic Energy Regulatory Board, India.[16] The detailed preparation of nafion-117 patches is explained by Saxena et al.[15]

The present study is aimed at calculating central axis depth dose and dose profiles in water phantom for the indigenously developed 32P-nafion-based patch source. For this purpose, the EGSnrc-based Monte Carlo code system is used.[17] Based on the calculated dose rate data, the treatment time to deliver a therapeutic dose of 30 Gy at reference depth is also calculated, as per the IAEA-tecdoc-1274.[18]

Materials and Methods

Monte Carlo calculations

DOSXYZnrc user-code[19] of the EGSnrc-code system[17] is used to calculate central axis depth doses and dose profiles in the unit density water medium for simulating the 1 cm × 1 cm 32P-nafion-patch source. The 1 cm × 1 cm 32P-nafion-patch source is positioned on 2 × 2 ×2 cm3water phantom. The thickness of source is 100 μm. The geometry and co-ordinate system used in the Monte Carlo calculations is shown in Figure 1. The elemental composition and density of phosphorous-loaded zirconium-nafion-117 composite membrane used in the Monte Carlo calculation is given in Table 1.[15]

Figure 1

(a) Schematic diagram of the 32P-nafion-patch source and water phantom used in the DOSXYZnrc Monte Carlo simulation. (b) Coordinate system used in the simulation

Table 1

Elemental composition of phosphorous-loaded zirconium-nafion- 117 composite membrane (density= 1620 kgm−3)

The 32P beta spectrum [see Figure 2] needed for the Monte Carlo calculation is based on ICRU Report No. 56.[20] In the Monte Carlo calculations, it is considered that the source particles are uniformly distributed in the nafion patch of dimensions 1 cm × 1 cm × 100 μm. The water phantom was divided in to voxels of dimension of 0.25 × 0.25 × 0.25 mm3 for generating dose profiles. Dose distributions in water are scored in these voxels. Separate simulation is carried out to score central axis depth dose by using bigger voxel dimensions (2 × 2× 0.25 mm3).

Figure 2

32P beta spectrum used in the Monte Carlo simulation

In the Monte Carlo calculations, all secondary particles such as knock on electrons and secondary bremsstrahlung photons produced by primary source electrons are completely followed in the simulation. The PEGS4 dataset needed for EGSnrc calculations is based on the XCOM compilations.[21] The PEGS4 dataset is generated by setting AE = ECUT = 0.521 MeV and AP = PCUT = 0.01 MeV, where the parameters AE and AP are the low energy thresholds for the production of knock-on electrons and secondary bremsstrahlung photons, respectively.

All Monte Carlo simulations utilized the PRESTA-II algorithm. The electron step-size parameter is set at ESTEP = 0.25. To increase the speed of the calculations, electron range rejection technique was used by setting ESAVE = 2 MeV. Auxiliary simulation was also carried without range rejection. The calculations suggest that using the range rejection with ESAVE = 2 MeV, improves the efficiency of the calculations by about 40%. In order to know the effect of boundary crossing algorithm on computational time, both PRESTA-I and EXACT boundary crossing algorithms were used in the calculations. The study shows that the using the PRESTA-I option results in improving the efficiency of the calculations by a factor 2 when compared to using the EXACT boundary-crossing algorithm. This observation is consistent with the findings by Walters and Kawrakow in their EGSnrc-based Monte Carlo study involving radiotherapy electron beams.[22]

Results and Discussion

The variation of the dose values per unit activity (GyBq-1) as a function of depth (mm) in water for the 32P-nafion-patch source is shown in Figure 3. The dose decreases rapidly with increasing depth in water. Central axis dose at 4 mm depth in water is only 0.08% of the central axis surface dose. Such a rapid decrease in dose will result in better sparing of the normal tissues.

Figure 3

Depth dose distribution of 32P-nafion-patch source along the central axis of the source

Table 2 compares the values of central axis depth dose per unit activity (GyBq-1) of 32P-nafion-patch source with the corresponding values of 32P-silicon-patch[9] for different depths in water. Higher dose rate values are observed in the case of 32P-silicon-patch source, because the radioactivity is distributed in lesser surface area (0.196 cm2) as compared to 32P-nafion-patch source, where surface area is 1 cm2.

Table 2

Comparison of dose values per unit activity (GyBq−1) presented as a function of depth in water

For treatment time calculation, 1 mm depth from the surface along the central axis of the source is considered as reference depth.[18] The value of dose in water calculated at 1 mm from the source surface is 8.41 × 10-11 GyBq-1. Hence, the time required to deliver a therapeutic dose of 30 Gy for a 37 MBq of radioactivity distributed in 1 cm2of 32P-nafion-patch source is about 2.7 hours.

Figure 4 presents the dose rate profiles along the x-axis of 32P-nafion-patch source for three different depths z = 0.5, 1 and 2 mm from the source surface. Figure 5 presents normalized dose values along the x-axis at depth 1 mm. The central axis dose value at 1 mm depth is used for normalization. Dose rate value at 3.5 mm away from the central axis is about 91% of the central axis value. Whereas dose rate at 5 mm away from the central axis is only 50% of the central axis value. Figures 6–8 show isodose profiles of the 32P-nafion-patch source at depths of 0.5, 1, and 2 mm. About 3.25–3.5 mm distance around the central axis is covered by about 90% isodose line for depths of 0.5, 1, and 2 mm. Hence, the 32P-nafion-patch source is effective for treatment of approximately 6.5–7.0 mm diameter lesions.

Figure 4

Dose profile along the x-axis of 32P-nafion-patch source for different depths, z = 0.5 mm, z = 1 mm and z = 2 mm

Figure 5

Normalized dose profile along the x-axis of 32P-nafion-patch source for depth z = 1 mm

Figure 6

Isodose profiles of the 32P-nafion-patch source on xy-plane at a depth of 0.5 mm

Figure 8

Isodose profiles of the 32P-nafion-patch source on xy-plane at a depth of 2 mm

Isodose profiles of the 32P-nafion-patch source on xy-plane at a depth of 1 mm

Conclusions

Dose distributions for the indigenously developed 1 cm × 1 cm32P-nafion skin patch source are calculated using the Monte Carlo-based EGSnrc code system. The calculated treatment time for delivering therapeutic dose of 30 Gy at 1 mm depth along the central axis of the source involving 37 MBq activity is about 2.7 hrs. This source is effective for treatment of approximately 6.5–7.0 mm diameter lesions.