Christopher S. Melhus

Supplement to the 2004 Update of the AAPM Task Group No. 43 Report

Since publication of the 2004 update to the American Association of Physicists in Medicine (AAPM) Task Group No. 43 Report (TG-43U1), several new low-energy photon-emitting brachytherapy sources have become available. Many of these sources have satisfied the AAPM prerequisites for routine clinical use as of January 10, 2005, and are posted on the Joint AAPM/RPC Brachytherapy Seed Registry. Consequently, the AAPM has prepared this supplement to the 2004 AAPM TG-43 update. This paper presents the AAPM-approved consensus datasets for these sources, and includes the following 125I sources: Amersham model 6733, Draximage model LS-1, Implant Sciences model 3500, IBt model 1251L, IsoAid model IAI-125A, Mentor model SL-125/ SH-125, and SourceTech Medical model STM1251. The Best Medical model 2335 103Pd source is also included. While the methodology used to determine these data sets is identical to that published in the AAPM TG-43U1 report, additional information and discussion are presented here on some questions that arose since the publication of the TG-43U1 report. Specifically, details of interpolation and extrapolation methods are described further, new methodologies are recommended, and example calculations are provided. Despite these changes, additions, and clarifications, the overall methodology, the procedures for developing consensus data sets, and the dose calculation formalism largely remain the same as in the TG-43U1 report. Thus, the AAPM recommends that the consensus data sets and resultant source-specific dose-rate distributions included in this supplement be adopted by all end users for clinical treatment planning of low-energy photon-emitting brachytherapy sources. Adoption of these recommendations may result in changes to patient dose calculations, and these changes should be carefully evaluated and reviewed with the radiation oncologist prior to implementation of the current protocol.

Approaches to calculating AAPM TG‐43 brachytherapy dosimetry parameters for , , , , and sources

Underlying characteristics in brachytherapy dosimetry parameters for medical radionuclides 137Cs, 125I, 192Ir, 103Pd, and 169Yb were examined using Monte Carlo methods. Sources were modeled as unencapsulated point or line sources in liquid water to negate variations due to materials and construction. Importance of phantom size, mode of radiation transport physics--i.e., photon transport only or coupled photon:electron transport, phantom material, volume averaging, and Monte Carlo tally type were studied. For noninfinite media, g(r) was found to degrade as r approached R, the phantom radius. MCNP5 results were in agreement with those published using GEANT4. Brachytherapy dosimetry parameters calculated using coupled photon:electron radiation transport simulations did not differ significantly from those using photon transport only. Dose distributions from low-energy photon-emitting radionuclides 125I and 103Pd were sensitive to phantom material by upto a factor of 1.4 and 2.0, respectively, between tissue-equivalent materials and water at r =9 cm. In comparison, high-energy photons from 137Cs, 192Ir, and 169Yb demonstrated +/- 5% differences in dose distributions between water and tissue substitutes at r=20 cm. Similarly, volume-averaging effects were found to be more significant for low-energy radionuclides. When modeling line sources with L < or = 0.5 cm, the two-dimensional anisotropy function was largely within +/- 0.5% of unity for 137Cs, 125I, and 192Ir. However, an energy and geometry effect was noted for 103Pd and 169Yb, with Pd-103F(0.5,0 degrees)=l.05 and yb-169F(0.5,0 degrees)=0.98 for L=0.5 cm. Simulations of monoenergetic photons for L=0.5 cm produced energy-dependent variations in F(r, theta) having a maximum value at 10 keV, minimum at 50 keV, and approximately 1.0 for higher-energy photons up to 750 keV. Both the F6 cell heating and *F4 track-length estimators were employed to determine brachytherapy dosimetry parameters. F6 was found to be necessary for g(r), while both tallies provided equivalent results for F(r, theta).

Brachytherapy dosimetry parameters calculated for a new 103Pd source

A new brachytherapy source having 103Pd adsorbed onto silver beads has been designed. The dose distributions of this source have been characterized using version 5 of the MCNP Monte Carlo radiation transport code available from Oak Ridge National Laboratory. These results are presented in terms of the updated AAPM Task Group No. 43 (TG-43U1) formalism, dosimetry parameters, and recommended calculation methodology.

COMS eye plaque brachytherapy dosimetry simulations for 103Pd, 125I, and 131Cs.

Monte Carlo (MC) simulations were performed to estimate brachytherapy dose distributions for Collaborative Ocular Melanoma Study (COMS) eye plaques. Brachytherapy seed models 200, 6711, and CS-1 Rev2 carrying P103d, I125, and C131s radionuclides, respectively, were modeled and benchmarked against previously published values. Calculated dose rate constants ΛMC were 0.684, 0.924, and 1.052 cGy h-1 U-1 (±2.6%, k=1 uncertainty) for models 200, 6711, and CS-1 Rev2, respectively. The seeds were distributed into 10, 12, 14, 16, 18, 20, and 22 mm-diameter COMS eye plaques. Simulations were performed in both heterogeneous and homogeneous environments, where the latter were in-water and the former included the silastic seed carrier insert and gold-alloy plaque. MC-based homogenous central axis dose distributions agreed within 2%±1% (±1 s.d.) to hand-calculated values. For heterogeneous simulations, notable photon attenuation was observed, with dose reduction at 5 mm of 19%, 11%, and 9% for P103d, I125, and C131s, respectively. A depth-dependent correction factor was derived to correct homogenous central-axis dose distributions for plaque component heterogeneities, which were found to be significant at short radial distances.

Approaches to calculating AAPM TG-43 brachytherapy dosimetry parameters for 137Cs, 125I, 192Ir, 103Pd, and 169Yb sources.

Underlying characteristics in brachytherapy dosimetry parameters for medical radionuclides 137Cs, 125I, 192Ir, 103Pd, and 169Yb were examined using Monte Carlo methods. Sources were modeled as unencapsulated point or line sources in liquid water to negate variations due to materials and construction. Importance of phantom size, mode of radiation transport physics--i.e., photon transport only or coupled photon:electron transport, phantom material, volume averaging, and Monte Carlo tally type were studied. For noninfinite media, g(r) was found to degrade as r approached R, the phantom radius. MCNP5 results were in agreement with those published using GEANT4. Brachytherapy dosimetry parameters calculated using coupled photon:electron radiation transport simulations did not differ significantly from those using photon transport only. Dose distributions from low-energy photon-emitting radionuclides 125I and 103Pd were sensitive to phantom material by upto a factor of 1.4 and 2.0, respectively, between tissue-equivalent materials and water at r =9 cm. In comparison, high-energy photons from 137Cs, 192Ir, and 169Yb demonstrated +/- 5% differences in dose distributions between water and tissue substitutes at r=20 cm. Similarly, volume-averaging effects were found to be more significant for low-energy radionuclides. When modeling line sources with L < or = 0.5 cm, the two-dimensional anisotropy function was largely within +/- 0.5% of unity for 137Cs, 125I, and 192Ir. However, an energy and geometry effect was noted for 103Pd and 169Yb, with Pd-103F(0.5,0 degrees)=l.05 and yb-169F(0.5,0 degrees)=0.98 for L=0.5 cm. Simulations of monoenergetic photons for L=0.5 cm produced energy-dependent variations in F(r, theta) having a maximum value at 10 keV, minimum at 50 keV, and approximately 1.0 for higher-energy photons up to 750 keV. Both the F6 cell heating and *F4 track-length estimators were employed to determine brachytherapy dosimetry parameters. F6 was found to be necessary for g(r), while both tallies provided equivalent results for F(r, theta).

Comparison of dose calculation methods for brachytherapy of intraocular tumors

PURPOSE To investigate dosimetric differences among several clinical treatment planning systems (TPS) and Monte Carlo (MC) codes for brachytherapy of intraocular tumors using 125I or 103Pd plaques, and to evaluate the impact on the prescription dose of the adoption of MC codes and certain versions of a TPS (Plaque Simulator with optional modules). METHODS Three clinical brachytherapy TPS capable of intraocular brachytherapy treatment planning and two MC codes were compared. The TPS investigated were Pinnacle v8.0dp1, BrachyVision v8.1, and Plaque Simulator v5.3.9, all of which use the AAPM TG-43 formalism in water. The Plaque Simulator software can also handle some correction factors from MC simulations. The MC codes used are MCNP5 v1.40 and BrachyDose/EGSnrc. Using these TPS and MC codes, three types of calculations were performed: homogeneous medium with point sources (for the TPS only, using the 1D TG-43 dose calculation formalism); homogeneous medium with line sources (TPS with 2D TG-43 dose calculation formalism and MC codes); and plaque heterogeneity-corrected line sources (Plaque Simulator with modified 2D TG-43 dose calculation formalism and MC codes). Comparisons were made of doses calculated at points-of-interest on the plaque central-axis and at off-axis points of clinical interest within a standardized model of the right eye. RESULTS For the homogeneous water medium case, agreement was within approximately 2% for the point- and line-source models when comparing between TPS and between TPS and MC codes, respectively. For the heterogeneous medium case, dose differences (as calculated using the MC codes and Plaque Simulator) differ by up to 37% on the central-axis in comparison to the homogeneous water calculations. A prescription dose of 85 Gy at 5 mm depth based on calculations in a homogeneous medium delivers 76 Gy and 67 Gy for specific 125I and 103Pd sources, respectively, when accounting for COMS-plaque heterogeneities. For off-axis points-of-interest, dose differences approached factors of 7 and 12 at some positions for 125I and 103Pd, respectively. There was good agreement (approximately 3%) among MC codes and Plaque Simulator results when appropriate parameters calculated using MC codes were input into Plaque Simulator. Plaque Simulator and MC users are perhaps at risk of overdosing patients up to 20% if heterogeneity corrections are used and the prescribed dose is not modified appropriately. CONCLUSIONS Agreement within 2% was observed among conventional brachytherapy TPS and MC codes for intraocular brachytherapy dose calculations in a homogeneous water environment. In general, the magnitude of dose errors incurred by ignoring the effect of the plaque backing and Silastic insert (i.e., by using the TG-43 approach) increased with distance from the plaques central-axis. Considering the presence of material heterogeneities in a typical eye plaque, the best method in this study for dose calculations is a verified MC simulation.

Evaluation of high-energy brachytherapy source electronic disequilibrium and dose from emitted electrons

PURPOSE The region of electronic disequilibrium near photon-emitting brachytherapy sources of high-energy radionuclides (60Co, 137CS, 192Ir, and 169Yb) and contributions to total dose from emitted electrons were studied using the GEANT4 and PENELOPE Monte Carlo codes. METHODS Hypothetical sources with active and capsule materials mimicking those of actual sources but with spherical shape were examined. Dose contributions due to source photons, x rays, and bremsstrahlung; source beta-, Auger electrons, and internal conversion electrons; and water collisional kerma were scored. To determine if conclusions obtained for electronic equilibrium conditions and electron dose contribution to total dose for the representative spherical sources could be applied to actual sources, the 192Ir mHDR-v2 source model (Nucletron B.V., Veenendaal, The Netherlands) was simulated for comparison to spherical source results and to published data. RESULTS Electronic equilibrium within 1% is reached for 60Co, 137CS, 192Ir, and 169Yb at distances greater than 7, 3.5, 2, and 1 mm from the source center, respectively, in agreement with other published studies. At 1 mm from the source center, the electron contributions to total dose are 1.9% and 9.4% for 60Co and 192Ir, respectively. Electron emissions become important (i.e., > 0.5%) within 3.3 mm of 60Co and 1.7 mm of 192Ir sources, yet are negligible over all distances for 137Cs and 169Yb. Electronic equilibrium conditions along the transversal source axis for the mHDR-v2 source are comparable to those of the spherical sources while electron dose to total dose contribution are quite different. CONCLUSIONS Electronic equilibrium conditions obtained for spherical sources could be generalized to actual sources while electron contribution to total dose depends strongly on source dimensions, material composition, and electron spectra.

An approach to using conventional brachytherapy software for clinical treatment planning of complex, Monte Carlo-based brachytherapy dose distributions

Certain brachytherapy dose distributions, such as those for LDR prostate implants, are readily modeled by treatment planning systems (TPS) that use the superposition principle of individual seed dose distributions to calculate the total dose distribution. However, dose distributions for brachytherapy treatments using high-Z shields or having significant material heterogeneities are not currently well modeled using conventional TPS. The purpose of this study is to establish a new treatment planning technique (Tufts technique) that could be applied in some clinical situations where the conventional approach is not acceptable and dose distributions present cylindrical symmetry. Dose distributions from complex brachytherapy source configurations determined with Monte Carlo methods were used as input data. These source distributions included the 2 and 3 cm diameter Valencia skin applicators from Nucletron, 4-8 cm diameter AccuBoost peripheral breast brachytherapy applicators from Advanced Radiation Therapy, and a 16 mm COMS-based eye plaque using 103Pd, 125I, and 131Cs seeds. Radial dose functions and 2D anisotropy functions were obtained by positioning the coordinate system origin along the dose distribution cylindrical axis of symmetry. Origin:tissue distance and active length were chosen to minimize TPS interpolation errors. Dosimetry parameters were entered into the PINNACLE TPS, and dose distributions were subsequently calculated and compared to the original Monte Carlo-derived dose distributions. The new planning technique was able to reproduce brachytherapy dose distributions for all three applicator types, producing dosimetric agreement typically within 2% when compared with Monte Carlo-derived dose distributions. Agreement between Monte Carlo-derived and planned dose distributions improved as the spatial resolution of the fitted dosimetry parameters improved. For agreement within 5% throughout the clinical volume, spatial resolution of dosimetry parameter data < or = 0.1 cm was required, and the virtual brachytherapy source data set included over 5000 data points. On the other hand, the lack of consideration for applicator heterogeneity effect caused conventional dose overestimates exceeding an order of magnitude in regions of clinical interest. This approach is rationalized by the improved dose estimates. In conclusion, a new technique was developed to incorporate complex Monte Carlo-based brachytherapy dose distributions into conventional TPS. These results are generalizable to other brachytherapy source types and other TPS.

Equivalent phantom sizes and shapes for brachytherapy dosimetric studies of Ir192 and Cs137

The impact of phantom size and shape in brachytherapy dosimetry was assessed using Monte Carlo methods in liquid water for Ir192 and Cs137 point sources. This is needed since differences in published dosimetry data, both measurements and simulations, employ a variety of phantom sizes and shapes which can cause dose differences exceeding 30% near the phantom periphery. Spheres of radius, Rsph, 10-40cm were examined to determine the equivalent spherical phantom size to a variety of cylinder and cube sizes, Rcyl and Rcube, respectively. These sizes ranged from 10to30cm. The equivalent Rsph for a given size cylinder or cube was determined using a figure of merit (FOM) function to minimize differences between radial dose functions, g(r). Using the FOM approach, a linear fit (R2>0.99) was obtained for the equivalent Rsph for a given size cylinder or cube. The equivalent phantom for a cylinder, of 40cm diameter and length 40cm, is a sphere of 21cm in radius and the equivalent phantom for a cube of 30cm on each side is a sphere of 17.5 in radius. When normalizing all results to r=1cm for g(r) comparisons of phantom shape, the absolute dose rates were equivalent within 0.1% for Rsph⩾10cm for both Ir192 and Cs137. Correlation factors to permit comparisons of unbounded g(r) data for r⩽15cm in 20 published datasets resulted in agreement generally within 2%. Residual differences with four datasets were attributed to methodological uncertainties in the published references.

Approaches to calculating AAPM TG-43 brachytherapy dosimetry parameters for {sup 137}Cs, {sup 125}I, {sup 192}Ir, {sup 103}Pd, and {sup 169}Yb sources

Underlying characteristics in brachytherapy dosimetry parameters for medical radionuclides 137Cs, 125I, 192Ir, 103Pd, and 169Yb were examined using Monte Carlo methods. Sources were modeled as unencapsulated point or line sources in liquid water to negate variations due to materials and construction. Importance of phantom size, mode of radiation transport physics--i.e., photon transport only or coupled photon:electron transport, phantom material, volume averaging, and Monte Carlo tally type were studied. For noninfinite media, g(r) was found to degrade as r approached R, the phantom radius. MCNP5 results were in agreement with those published using GEANT4. Brachytherapy dosimetry parameters calculated using coupled photon:electron radiation transport simulations did not differ significantly from those using photon transport only. Dose distributions from low-energy photon-emitting radionuclides 125I and 103Pd were sensitive to phantom material by upto a factor of 1.4 and 2.0, respectively, between tissue-equivalent materials and water at r =9 cm. In comparison, high-energy photons from 137Cs, 192Ir, and 169Yb demonstrated +/- 5% differences in dose distributions between water and tissue substitutes at r=20 cm. Similarly, volume-averaging effects were found to be more significant for low-energy radionuclides. When modeling line sources with L < or = 0.5 cm, the two-dimensional anisotropy function was largely within +/- 0.5% of unity for 137Cs, 125I, and 192Ir. However, an energy and geometry effect was noted for 103Pd and 169Yb, with Pd-103F(0.5,0 degrees)=l.05 and yb-169F(0.5,0 degrees)=0.98 for L=0.5 cm. Simulations of monoenergetic photons for L=0.5 cm produced energy-dependent variations in F(r, theta) having a maximum value at 10 keV, minimum at 50 keV, and approximately 1.0 for higher-energy photons up to 750 keV. Both the F6 cell heating and *F4 track-length estimators were employed to determine brachytherapy dosimetry parameters. F6 was found to be necessary for g(r), while both tallies provided equivalent results for F(r, theta).