ORIGINAL ARTICLE Year : 2021  Volume : 46  Issue : 3  Page : 197203 Validation of American association of physicists in medicine tg 43 dosimetry data in commercial treatment planning system Sathiyan Saminathan, K Maheshraja, KM Ganesh Department of Radiation Physics, Kidwai Memorial Institute of Oncology, Bengaluru, Karnataka, India Correspondence Address: Aims This study aimed to validate the dosimetric data of lowenergy photonemitting lowdose rate (LELDR) brachytherapy seed sources in commercial treatment planning system (TPS). Materials and Methods The LELDR seed sources dosimetric data were published in the American Association of Physicists in Medicine (AAPM) Task Group reports TG43 (1995), TG43U1 (2004), TG43U1S1 (2007), and TG43U1S2. The Bhabha Atomic Research centre (BARC) ^{125}I OcuProsta seed dosimetry data are also available in the literature. The commercially available TPSs are using both twodimensional (cylindrically symmetric linesource) and onedimensional (1D) (point source) dosecalculation formalisms. TPS used in this study uses only 1D dosecalculation formalism for permanent implant dosimetry. The pointdose calculation, dose summation, isodose representation, and dose–volume histogram quality assurance tests were performed in this study. The pointsource dosecalculation tests were performed for all the available sources in the literature. The others tests were performed for the I125 BARC OcuProsta seeds. The TPScalculated doses were validated using manual calculation. Results and Discussion In pointsource calculation test, the TPScalculated pointdose values are within ±2% agreement with manually calculated dose for all the seeds studied. The agreement between the TPS and manually calculated dose is 0.5% for the dose summation test. The isodose line pass through the grid points at an equal distance was verified visually on the computer screen for seed used clinically. In dose–volume histogram test, the TPSdetermined volume was compared with the real volume. Conclusion Misinterpretation of the TPS test and/or misunderstanding of the TG43 dosecalculation formalism may cause large errors. It is very important to validate the TPS using literature provided dosimetric data. The dosimetric data of BARC ^{125}I OcuProsta Seed are validated with other AAPM TG43recommended seeds. The dose calculation of Best® NOMOS permanent implant TPS is accurate for all permanent implant seeds studied.
Introduction The American Association of Physicists in Medicine (AAPM) published the Task Group No. 43 (TG43)[1] in the year 1995 by introducing a new brachytherapy dosecalculation formalism based on findings of the Interstitial Collaborative Working Group.[2] AAPM TG43 presented the dosimetry formalism for three lowenergy photonemitting sources. The source models are shown in [Table 1]. The dosimetry formalism was updated in 2004 and was termed the AAPM TG43U1 report.[3] [Table 2] shows that the source models were presented in AAPM TG43U1 report. Several additional sources were introduced in the market, to include the dosimetric datasets, the TG43 supplement report was introduced in the year 2007 and termed as AAPM TG43U1S1 report.[4] AAPM TG43U1S1 report presented source models which are shown in [Table 3]. The AAPM and the Groupe Européen de Curiethérapie–European Society for Radiotherapy and Oncology jointly introduced the TG43 supplement report 2 in the year 2017 and termed as AAPM TG43U1S2 report.[5] In this report, 11 lowenergy photonemitting brachytherapy source dosimetry datasets were introduced. [Table 4] shows the source models presented in AAPM TG43U1S2 report. The 131Cs source model were also discussed in this report in addition to 125I and 103Pd source models. {Table 1}{Table 2}{Table 3}{Table 4} In India, BARC 125I OcuProsta seed is in clinical use, mostly used for ophthalmic application.[6],[7] Only a few Radiotherapy centers are using this source for permanent prostate implants. The BARC 125I OcuProsta seed dosimetry data are also available in the literature.[8],[9] The commercial treatment planning system (TPS) used for the permanent implant can be configured either with line source approximation or pointsource approximation or both. In our center, we have Best® NOMOS permanent implant TPS. The Best® NOMOS permanent implant TPS uses the pointsource approximation for dose calculation.[10],[11] In this study, 27 lowenergy photonemitting lowdose rate (LELDR) brachytherapy sources (20 125I, 6 103Pd, and 1 131Cs) dosimetric data were evaluated using commercial TPS. The test cases recommended in the Netherlands Commission on Radiation Dosimetry Subcommittee “Dosimetry and quality control of brachytherapy with lowenergy photon sources (125I)” are used in this study to evaluate the permanent implant TPS algorithm.[12] Materials and Methods The LELDR sources dosimetric data such as dose rate constant (^), geometry function G (r, ɵ), radial dose function g(r), onedimensional (1D) anisotropy function (ᶲan(r)), and twodimensional (2D) anisotropy function F (r, ɵ) were published in AAPM TG43 (1995), TG43U1 (2004), TG43U1S1 (2007), and TG43U1S2 reports. The BARC 125I OcuProsta seed dosimetry data are also available in the literature.[8],[9] The commercially available TPSs are using both 2D (cylindrically symmetric line source) and 1D (point source) dosecalculation formalisms. However, some TPS uses 1D dosecalculation formalism for permanent implant dosimetry. General twodimensional formalism This formalism assumes the dose to be distributed symmetrically with respect to the longitudinal axis of the source. A polar coordinate system is used with its origin located at the center of the active source and zero angles coinciding with the longitudinal axis of the source. The reference point P (r0,ɵ0) is taken on the source transverse plane (ɵ0 = 90°) at the reference distance of 1 cm [Figure 1].{Figure 1} The dose rate to water at a point P (r, θ) can be expressed as [INLINE:1] where Ḋ(r, θ ) is the absorbed dose rate to water at (r, q) (in units of cGy h−1), r is the radial distance from the source center to the point of interest, θ is the polar angle, Sk is the airkerma strength of the source (in units of U = μGy m2 h−1 = cGy cm2 h−1), Λ is the dose rate constant (in units of cGy h−1 U−1 which reduces to cm−2), GL (r, q) is the geometry function, gL (r) is the radial dose function, F (r, q) is the 2D anisotropy function General onedimensional formalism When evaluating implants with a large number of seeds in TPS, it is often difficult to determine the exact orientation of the source longitudinal axis for each individual seed. Moreover, many TPSs assume the seeds to be oriented parallel to the longitudinal axis of the image set, i.e. perpendicular to the transverse images. The 1D pointsource approximation (Eq. 2) eliminates the need to determine the orientation of the sources. This model will only approximate the true 2D dose distribution, it applies an average correction for the anisotropy effects and treat the source as isotropic.[12] The dose rate to water at a point can be expressed as [INLINE:2] For dose calculation, the equation (2) can be used as [INLINE:3] or [INLINE:4] TG43U1 (2004) recommends the use of Eq.(4) for dose calculation because of an improved accuracy at small distances (r <1 cm). However, many TPS systems use Eq. (3) for dose calculation. Using the Eq.(3), the initial dose rate is converted to total dose when multiplying by the mean lifetime, Ʈ. The mean lifetime is defined as Ʈ =1.44 × t1/2, where t1/2 represents the halflife. Treatment Planning System In our center, we use the Best® NOMOS (DBA: Best Medical International, Inc., Pittsburgh, PA, USA), TPS, its dual Activity Module creates the treatment plans using multiple activity sources. Octant TherapyTM Module allows the view of DVH values of specific volumes. Slice Shifter™ provides the ability to correct image slice positioning errors. The concurrent twodimensional (2D) and 3D visualization allows instant visualization of seed placement, dose distribution, and anatomical structures. The pattern loading feature can be used for new plans with user definable seed pattern loading. This TPS supports both the volume plan (Version 4.9n) and computed tomography (CT) plan (Version 4.01). In volume plan, we can perform the preplan or operating room dosimetry (OR) dosimetry. The TPS can be connected to an ultrasound unit to access the live video image. From the live video, the images can be acquired to TPS with a set of Zvalues for preplan dosimetry. Realtime planning is not possible with this TPS. However, in the operating room (OR), the intraoperative planning can be performed is known as realtime planning. The patient remains stationary between the time of the volume study and the implant procedure. The postimplant CT provides the postplan dosimetry with the help of auto seed detection method. The Best® NOMOS permanent implant TPS uses the 1D dosecalculation formalism for permanent implant dosimetry. The source can be configured in the dose kernel model. The input parameters are source name, manufacturer name, source length, isotope type, default activity, halflife, airkerma rate constant, dose rate constant, radial dose function, and anisotropy function. The configured source model can be used for both preplan and postplan CT dosimetry. In this TPS, BARC 125I OcuProsta seed and other AAPM TG43recommended (26 LELDR) seed sources were configured for dosimetric study. Quality assurance tests The following quality assurance tests were reported in the literature.[12] Test 1, pointsource calculation Calculation of the dose from a single seed in a number of points at different distances perpendicular to the seed axis using the pointsource approximation. Test 2, dose summation Summation of the dose from two seeds in a number of points at different distances perpendicular to the seed axis using the pointsource approximation. Test 3, isodose representation Representation of the isodose lines around a single source using the pointsource approximation. Test 4, dose–volume histogram Calculation of the DVH from a single source using the pointsource approximation Test 5, linesource calculation Calculation of the dose from a single seed in a number of points at different distances along the transverse and longitudinal axis of the seed and in three points at angles in between using the linesource approximation. Tests 1–4 were performed using a 1D anisotropy calculation model. Test 5 is not considered in this study, since this TPS is not supporting line source approximation. The test procedure is as follows. Test 1: Pointsource calculation test In this test, the dose was calculated in a number of points at different distances perpendicular to the seed axis of a single seed using the pointsource approximation. Test procedure The test patient was created in the TPS with ultrasound prostate phantom (CIRS, USA). The phantom is shown in [Figure 2]The source strength of 100 U was placed at image 5 (z = 2 cm) at the leftmost column in the middleThe dose points add in the image at grid point on right to the source at distances of 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, and 6.0 cmThe dose value of TPS was compared with the manually calculated dose valuesThe manual calculation was performed using the equation (3).{Figure 2} Test 2: Dose summation test In this test, the dose calculation was done using two seeds in a number of points (dose points) at different distances perpendicular to the seed axis. Test procedure The same plan was used from the test 1At 1 cm right from the first source, the second source strength of 100U was added at image z = 2 cmThe sum of the dose was verified at the distance of 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 cm. Test 3: Isodose representation test In this test, the isodose representation around a single source was verified. Test procedure Using the test plan 1, the source strength of 100 U was added at the central image (z = 5 cm) of center gridThe isodose levels equal to dosecalculation points of 1.0, 2.0, and 3.0 cm were selectedThe isodose lines passing through the grid points were verified. Test 4: Dose–volume histogram test In this test, the dose–volume calculation of the TPS was verified using geometrical DVH values. Test procedure In a new test study, a volume (prostate) of 4 cm × 4 cm × 6 cm was createdA source was placed at the central image (z = 3.5 cm) of center gridThe source strength was determined to get the 100 Gy point dose at a distance of 1.0, 2.0, and 3.0 cm. Also using the test 1 (pointsource calculation), the dosimetry of 27 LELDR sources was studied. For all the sources, the TPScalculated point doses were validated using manually calculated dose. Results and Discussion Test 1: Pointsource calculation test The calculated point doses in the TPS at various grid point distances of 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, and 6.0 cm were validated using manually calculated dose value. The equation (3) was used for the manual calculation. [Figure 3]a and [Figure 3]b shows the variation between TPS and manually calculated pointdose values for BARC OcuProsta (125I) and AAPM TG43U1 (2004)recommended seeds. [Figure 4]a and [Figure 4]b shows the variation between TPS and manually calculated pointdose values of AAPM TG43U1S1 (2007)recommended seeds. [Figure 5]a, [Figure 5]b, [Figure 5]c shows the variation between TPS and manually calculated pointdose values for AAPM TG43U1S2 (2017)recommended seeds. The results show the variation between TPS and manually calculated pointdose values are within ± 2% for all the seeds studied.{Figure 3}{Figure 4}{Figure 5} Test 2: Dose summation test In this test, the dose calculation was done using 2 BARC OcuProsta seeds in a number of points (dose points) at different distances perpendicular to the seed axis. The sum of the doses was verified at various distances of 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 cm in TPS. The TPScalculated pointdose values are with 0.5% compared to manually calculated dose value. [Figure 6] shows the dose summation test.{Figure 6} Test 3: Isodose representation test This test was performed using BARC OcuProsta seed placed at center grid (Z = 5 cm). The isodose levels were passing the grid points at 1.0, 2.0, and 3.0 cm distances. The standard uncertainty of mean for the grid points was within 1%. [Figure 7] shows the image of isodose lines passing through the grid points at a distance of 1.0, 2.0, and 3.0 cm.{Figure 7} Test 4: Dose–volume histogram test In the TPS, the volume was created using the contouring tools, which was 4 cm × 4 cm × 6 cm (96.0 cm3) and TPScalculated volume was 96.04 cm3. To get the 100 Gy dose at a distance of 1.0 cm, the source strength used was 5.5U. The source strength was 26.0 U to get the 100 Gy dose at a distance of 2.0 cm and the source strength was 75.0 U to get the 100 Gy dose at a distance of 3.0 cm. Very often, TPSs come with preloaded TG43 data for most source models/brands. The user should be well aware of the version of algorithm used in TPS for dose calculation (preplanning, online planning, and/or postplanning procedures). The user should verify the TPS for the correct seed model to be used for dose calculation. We verified the dose calculation in a number of points on the longitudinal and transverse axis of the source and compared the calculated dose with manual calculation using data provided in the literature.[3],[4],[5] The deviation between the TPScalculated and manually calculated dose is within ± 2%. It is well agree with the literature recommendation.[12] The typical implantation involves many seeds (up 50 or more). The TPS calculates the dose distribution from each individual seed and sums up the dose distribution to the total dose. In this process, the sourcetosource shielding effects are neglected by the system. The sum dose should be numerically accurate. The tested dose sum accuracy is 0.5%. The isodose line pass through the grid points at an equal distance was verified visually on the computer screen for seed used clinically. In dose–volume histogram test, the TPSdetermined volume was compared with the real volume. Conclusion Misinterpretation of the TPS test and/or misunderstanding of the TG43 dosecalculation formalism may cause large errors.[12] It is very important to validate the TPS using literature provided dosimetric data. Furthermore, it is important to recognize the limitation of TPS; in this TPS, the source orientation is not clearly visible in the postimplant clinical image and also, this TPS cannot compute 2D anisotropy calculation. The dosimetric data of BARC 125I OcuProsta seed are validated on for with other AAPM TG43recommended seeds. The dose calculation of Best® NOMOS permanent implant TPS is accurate for the all permanent implant seeds studied. In case of major system upgrades or improvements in the basic TG43 data, it is good practice to study the influence of changes for a number of clinical cases by comparing results of previous calculations with the new ones. The TPS should be tested at time of commissioning, at every software upgrade, when basic TG43 data are modified or updated or when a new source model is being used. Acknowledgment The authors would like to thank Dr.Sanjay Kumar Saxena, Scientific Officer (F) Radiopharmaceuticals Division, BARC, Mumbai, for providing the BARC OcuProsta I125 seed source, and Vineet Gupta, R & D Director, Best NOMOS (Pittsburgh, PA), for TPS technical support. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest. References


