Purpose The Imaging and Radiation Oncology Core Houston (IROC‐H) (formerly the Radiological Physics Center) has reported varying levels of agreement in their anthropomorphic phantom audits. There is reason to believe one source of error in this observed disagreement is the accuracy of the dose calculation algorithms and heterogeneity corrections used. To audit this component of the radiotherapy treatment process, an independent dose calculation tool is needed. Methods Monte Carlo multiple source models for Elekta 6 MV and 10 MV therapeutic x‐ray beams were commissioned based on measurement of central axis depth dose data for a 10 × 10 cm2 field size and dose profiles for a 40 × 40 cm2 field size. The models were validated against open field measurements consisting of depth dose data and dose profiles for field sizes ranging from 3 × 3 cm2 to 30 × 30 cm2. The models were then benchmarked against measurements in IROC‐H's anthropomorphic head and neck and lung phantoms. Results Validation results showed 97.9% and 96.8% of depth dose data passed a ±2% Van Dyk criterion for 6 MV and 10 MV models respectively. Dose profile comparisons showed an average agreement using a ±2%/2 mm criterion of 98.0% and 99.0% for 6 MV and 10 MV models respectively. Phantom plan comparisons were evaluated using ±3%/2 mm gamma criterion, and averaged passing rates between Monte Carlo and measurements were 87.4% and 89.9% for 6 MV and 10 MV models respectively. Conclusions Accurate multiple source models for Elekta 6 MV and 10 MV x‐ray beams have been developed for inclusion in an independent dose calculation tool for use in clinical trial audits.

Purpose The Imaging and Radiation Oncology Core‐Houston (IROC‐H) Quality Assurance Center (formerly the Radiological Physics Center) has reported varying levels of compliance from their anthropomorphic phantom auditing program. IROC‐H studies have suggested that one source of disagreement between institution submitted calculated doses and measurement is the accuracy of the institution's treatment planning system dose calculations and heterogeneity corrections used. In order to audit this step of the radiation therapy treatment process, an independent dose calculation tool is needed. Methods Monte Carlo multiple source models for Varian flattening filter free (FFF) 6 MV and FFF 10 MV therapeutic x‐ray beams were commissioned based on central axis depth dose data from a 10 × 10 cm2 field size and dose profiles for a 40 × 40 cm2 field size. The models were validated against open‐field measurements in a water tank for field sizes ranging from 3 × 3 cm2 to 40 × 40 cm2. The models were then benchmarked against IROC‐H's anthropomorphic head and neck phantom and lung phantom measurements. Results Validation results, assessed with a ±2%/2 mm gamma criterion, showed average agreement of 99.9% and 99.0% for central axis depth dose data for FFF 6 MV and FFF 10 MV models, respectively. Dose profile agreement using the same evaluation technique averaged 97.8% and 97.9% for the respective models. Phantom benchmarking comparisons were evaluated with a ±3%/2 mm gamma criterion, and agreement averaged 90.1% and 90.8% for the respective models. Conclusions Multiple source models for Varian FFF 6 MV and FFF 10 MV beams have been developed, validated, and benchmarked for inclusion in an independent dose calculation quality assurance tool for use in clinical trial audits.

PURPOSE:A dose calculation tool, which combines the accuracy of the dose planning method (DPM) Monte Carlo code and the versatility of a practical analytical multisource model, which was previously reported has been improved and validated for the Varian 6 and 10 MV linear accelerators (linacs). The calculation tool can be used to calculate doses in advanced clinical application studies. One shortcoming of current clinical trials that report dose from patient plans is the lack of a standardized dose calculation methodology. Because commercial treatment planning systems (TPSs) have their own dose calculation algorithms and the clinical trial participant who uses these systems is responsible for commissioning the beam model, variation exists in the reported calculated dose distributions. Today's modern linac is manufactured to tight specifications so that variability within a linac model is quite low. The expectation is that a single dose calculation tool for a specific linac model can be used to accurately recalculate dose from patient plans that have been submitted to the clinical trial community from any institution. The calculation tool would provide for a more meaningful outcome analysis. METHODS:The analytical source model was described by a primary point source, a secondary extra-focal source, and a contaminant electron source. Off-axis energy softening and fluence effects were also included. The additions of hyperbolic functions have been incorporated into the model to correct for the changes in output and in electron contamination with field size. A multileaf collimator (MLC) model is included to facilitate phantom and patient dose calculations. An offset to the MLC leaf positions was used to correct for the rudimentary assumed primary point source. RESULTS:Dose calculations of the depth dose and profiles for field sizes 4 × 4 to 40 × 40 cm agree with measurement within 2% of the maximum dose or 2 mm distance to agreement (DTA) for 95% of the data points tested. The model was capable of predicting the depth of the maximum dose within 1 mm. Anthropomorphic phantom benchmark testing of modulated and patterned MLCs treatment plans showed agreement to measurement within 3% in target regions using thermoluminescent dosimeters (TLD). Using radiochromic film normalized to TLD, a gamma criteria of 3% of maximum dose and 2 mm DTA was applied with a pass rate of least 85% in the high dose, high gradient, and low dose regions. Finally, recalculations of patient plans using DPM showed good agreement relative to a commercial TPS when comparing dose volume histograms and 2D dose distributions. CONCLUSIONS:A unique analytical source model coupled to the dose planning method Monte Carlo dose calculation code has been modified and validated using basic beam data and anthropomorphic phantom measurement. While this tool can be applied in general use for a particular linac model, specifically it was developed to provide a singular methodology to independently assess treatment plan dose distributions from those clinical institutions participating in National Cancer Institute trials.

PURPOSE:To determine the setup error on an electron breast boost technique using daily cone beam computed tomography (CBCT). Patient and setup attributes were studied as contributing factors to the accuracy.METHODS AND MATERIALS:Reproducibility of a modified lateral decubitus position breast boost setup was verified for 33 patients using CBCT. Three-dimensional matching was performed between the CBCT and the initial planning CT for each boost fraction by matching the tumor bed and/or surgical clips. The dosimetric impact of the daily positioning error was achieved by rerunning the initial treatment plans incorporating the recorded shifts to study the dose differences. Breast compression, decubitus angle, tumor bed location and volume, and cup size were studied for their contribution to setup error.RESULTS:The range of setup errors was: 1.5 cm anterior to 9 mm posterior, 1.3 cm superior to 2.3 cm inferior, and 3.2 cm medial to 2.4 cm lateral. Seven patients had setup errors that were ≥2-cm margin placed on the tumor bed and scar. Four of those 7 patients had unacceptable coverage as defined by the volume of the tumor bed plus scar that is covered by the 90% isodose line (V90) compared with the original plan. All other patients had no discernible difference in the coverage (V90). The use of compression, tumor bed location, or volumes >20 mL showed no effect on coverage.CONCLUSIONS:In general, this study supported that a 2-cm margin was adequate (29 of 33 patients) when patients are treated under typical conditions. Care should be taken when high electron energies are selected because the coverage at depth is more difficult to maintain in the clinical environment.

Purpose: To commission a multiple‐source Monte Carlo model of Elekta linear accelerator beams of nominal energies 6MV and 10MV. Methods: A three source, Monte Carlo model of Elekta 6 and 10MV therapeutic x‐ray beams was developed in a two‐step process. Energy spectra of each of three sources, a primary source corresponding to photons created in the target, an extra‐focal source corresponding to photons originating from scattered events in the linac head, and an electron contamination source, were determined. The two photon sources were determined by an optimization process that fit the relative fluence of 0.25 MeV energy bins to the product of Fatigue‐Life and Fermi functions to match calculated percent depth dose (PDD) data with that measured in water for a 10×10cm2 field. Off‐axis effects were modeled by fitting the off‐axis fluence to a piece‐wise linear function through optimization of relative fluence to match calculated dose profiles with measured dose profiles for a 40×40cm2 field. A 3rd degree polynomial was used to describe the off‐axis half‐value layer as a function of off‐axis angle. The model was then commissioned by comparing calculated PDDs and dose profiles for field sizes ranging from 3×3cm2 to 30×30cm2 to those obtained from measurements. Results: Agreement between calculated and measured data was evaluated using 2%/2mm global gamma criterion for field sizes of 3×3, 5×5, 10×10, 15×15, 20×20, and 30×30cm2. Along the central axis of the beam 99.5% and 99.6% of all data passed the criterion for 6 and 10MV models, respectively. Dose profiles at depths of dmax, 5, 10, 20, and 25cm agreed with measured data for 95.4% and 99.2% of data tested for 6 and 10MV models, respectively. Conclusion: A Monte Carlo multiple‐source model for Elekta 6 and 10MV therapeutic x‐ray beams has been developed as a quality assurance tool for clinical trials. This work was supported by Public Health Service grants CA010953, CA081647, and CA21661 awarded by the National Cancer Institute, United States Department of Health and Human Services.

An anthropomorphic head phantom, constructed from a water-equivalent plastic shell with only a spherical target, was modified to include a nonspherical target (pituitary) and an adjacent organ at risk (OAR) (optic chiasm), within 2 mm, simulating the anatomy encountered when treating acromegaly. The target and OAR spatial proximity provided a more realistic treatment planning and dose delivery exercise. A separate dosimetry insert contained two TLD for absolute dosimetry and radiochromic film, in the sagittal and coronal planes, for relative dosimetry. The prescription was 25 Gy to 90% of the GTV, with ≤ 10% of the OAR volume receiving ≥ 8 Gy for the phantom trial. The modified phantom was used to test the rigor of the treatment planning process and phantom reproducibility using a Gamma Knife, CyberKnife, and linear accelerator (linac)-based radiosurgery system. Delivery reproducibility was tested by repeating each irradiation three times. TLD results from three irradiations on a CyberKnife and Gamma Knife agreed with the calculated target dose to within ± 4% with a maximum coefficient of variation of ± 2.1%. Gamma analysis in the coronal and sagittal film planes showed an average passing rate of 99.4% and 99.5% using ± 5%/3 mm criteria, respectively. Results from the linac irradiation were within ± 6.2% for TLD with a coefficient of variation of ± 0.1%. Distance to agreement was calculated to be 1.2 mm and 1.3mm along the inferior and superior edges of the target in the sagittal film plane, and 1.2 mm for both superior and inferior edges in the coronal film plane. A modified, anatomically realistic SRS phantom was developed that provided a realistic clinical planning and delivery challenge that can be used to credential institutions wanting to participate in NCI-funded clinical trials.

Purpose: To analyze the dosimetric impact of set-up accuracy of an electron breast boost technique using a lateral decubitus position with and without a breast compression device. Methods: Reproducibility of the breast boost set-up was verified for 19 patients and 99 fractions using Cone Beam CT (CBCT). 3D-3D matching was performed between the CBCT and the initial planning CT for each boost fraction by matching the tumor bed and clips. Shifts in all three dimensions were recorded for each fraction. The dosimetric impact of the daily positioning error was achieved by rerunning the initial treatment plans incorporating the shifts recorded for each fraction. Comparison of the tumor bed and scar coverage was analyzed for both plans. Results: The range of set-up errors based on CBCT was: 1.5 cm anterior to 8 mm posterior, 1.3 cm superior to 2.3 cm inferior, and −2.4 cm to 2.1 cm laterally. Three patients had set-up errors that were greater than or equal to 2 cm which is the normal margin placed on the tumor bed and scar. Two of these three patients had unacceptable coverage as defined by the V90 when compared to the original plan. The remaining 17 patients had no discernible difference in coverage (V90). Whether patients had breast compression seemed to have little impact on reproducibility of the set-up. Conclusion: 14 patients with a breast compression device and 5 patients without a breast compression device were studied using CBCT to quantify set-up error from 99 fractions. A 2 cm margin around the tumor bed plus scar volume seems to be adequate to account for set-up error. No appreciable difference was seen in set-up error between patients with breast compression or without. The two patients with unacceptable coverage were large breasted with deep seated tumors.

Purpose: A commercial software package (Mobius3d, Mobius Medical Systems) was acquired for use as an additional QA tool of our current treatment planning system (TPS). The system reads DICOM RT files generated from the TPS and performs an independent dose calculation for comparison to the TPS. This work summarizes our methods for commissioning a system such as this. Methods: Preliminary comparisons of our TPS and measured data to the QA softwares beam model were done by comparing PDDs and profiles with field sizes ranging from 4&×;4 to 40×40. Static fields and step and shoot IMRT plans were generated on a homogeneous medium for comparisons of 6x photons. For the homogeneous IMRT cases, ion chamber measurements were compared to the dose calculated in the TPS and in the verification system. Results: Analysis of PDD data showed that the average PDD error of the verification model (relative to the TPS) was 0.5% compared to our measured data which was 0.1%. The average infield profile error over all field sizes and depths of the TPS to our reference measured data set was 0.3% whereas the verification system relative to the TPS was 1.7%. Simple beam geometries between the TPS and the verification system were 0.8% difference on average. Ion chamber measurements on a homogeneous phantom differed from the TPS on average by 0.5% whereas the verification system was 2%. The results indicate an adjustment in the beam model of the verification system is needed to better match our current beam model/measured data. Conclusion: Preliminary commissioning results of a commercial QA dose algorithm were completed. Adjustment of the verification softwares beam model is necessary before further testing is done. Further analysis is needed to fully investigate this system to be used clinically as a QA tool of our current TPS. “Evaluation equipment provided by Mobius Medical Systems, LP”

Purpose: To compare a custom‐developed method for accurate dose recalculation of patient plans entered into clinical trials with results from a common treatment planning system. Method and Materials: A measurement‐driven multiple‐source model with the Dose Planning Method (DPM) Monte Carlo (MC) dose calculation algorithm was previously developed, validated, and benchmarked for the Varian 6 MV and 10 MV photon beams. Several patient cases have been recalculated and compared to the calculations from a Pinnacle planning system. Intensity modulated radiation therapy (IMRT) prostate, IMRT abdomen, stereotactic body radiotherapy lung, and IMRT lung patient cases were selected. Results: Field sizes from 4 cm × 4 cm to 40 cm × 40 cm were validated to within 2% of the maximum dose and 2 mm distance to agreement. At least 95% of the data tested met the validation criteria. Benchmark treatments planned using anthropomorphic phantoms were tested to within 3% of the target dose and 2 mm distance to agreement. At least 85% of the data tested met the benchmark criteria. Disagreement in the patient plan evaluation tended to occur at heterogeneity interfaces where electronic disequilibrium occurred, and in the beam penumbra, where scattered radiation was more prominent. The ratios of planning system calculation to MC calculation for the mean dose of the gross and planning target volumes for the patient plans ranged from 0.984 to 1.016. Conclusion: These results show that this MC software code generates answers similar to the Pinnacle system for IMRT and SBRT treatment plans. Differences are consistent with the superior physics modeling inherent in the Monte Carlo code. We believe the method will be useful for recalculating dose distributions for patients entered into clinical trials. Work supported by PHS CA010953, CA081647, and R01 CA85181 awarded by NCI, DHHS

The purpose of this study was to determine the accuracy of five commonly used intensity-modulated radiation therapy (IMRT) treatment planning systems (TPSs), 3 using convolution superposition algorithms or the analytical anisotropic algorithm (CSA/AAAs) and 2 using pencil beam algorithms (PBAs), in calculating the absorbed dose within a low-density, heterogeneous region when compared with measurements made in an anthropomorphic thorax phantom. The dose predicted in the target center met the test criteria (5% of the dose normalization point or 3 mm distance to agreement) for all TPSs tested; however, at the tumor-lung interface and at the peripheral lung in the vicinity of the tumor, the CSA/AAAs performed better than the PBAs (85% and 50%, respectively, of pixels meeting the 5%/3-mm test criteria), and thus should be used to determine dose in heterogeneous regions.

Purpose: Validation and benchmarking of a newly‐developed measurement‐driven source model based on Monte Carlo calculations. Method and Materials: A measurement‐driven model using the Dose Planning Method DPM dose calculation algorithm is being developed for use with Varian, Elekta, and Siemens 6 MV and 10 MV photon beams. The present work details the validation and benchmarking for the Varian 6 MV beam. The multi‐source model consists of a primary photon point source, an extra‐focal exponential disk source, and an electron contamination uniform disk source. The model accounts for fluence and off‐axis energy effects due to the flattening filter. Dose calculations for field sizes from 4 cm by 4 cm to 40 cm by 40 cm were performed and tested against the basic beam data measurements. In addition, an IMRT homogeneous plan, a stereotactic lung plan, and an IMRT lung plan were delivered to anthropomorphic phantoms housing TLD and radiographic film dosimeters for benchmark evaluations. Results: Comparisons between calculation and measurement of the PDD and dose profiles for all square field size configurations showed agreement within 2%/2 mm for 90% of the data tested. General agreement at the level of 3%/2mm for 85% of the data tested was found in both of the lung treatment plans. However, calculation of the highly modulated IMRT homogeneous plan showed an underestimation of dose of up to 8% locally in the center of the PTV. Conclusion: This work demonstrates a source model that is robust for the Varian 6 MV photon beam; however, only a simple MLC model that did not include the effects of the rounded leaf ends and interleaf leakage was used. We are currently evaluating a detailed MLC model to improve the agreement with measurement.Conflict of Interest: Work supported by PHS CA010953, CA081647, and R01 CA85181 awarded by NCI, DHHS

The authors report an error in the published paper.1 The variation in dose between the dose calculations and measurement for the low-dose critical structures was not reported correctly. The correct range for the low-dose critical structures is an underestimation of 14.9% to an overestimation of 11.2%. While this is reflected correctly in Table I of the aforementioned publication, it was not stated correctly in Sec. III on p. 5437 and in Sec. IV on p. 5438.

Purpose: We developed a Monte Carlo based IMRT recalculation tool and determined its parameters for Varian Clinac 2100C 6 MV and 18 MV photon beams. We report our comparisons with prostate and head and neck IMRT Pinnacle treatment plans. Method and Materials: Our source model components include: a primary photon point source, an extended extra‐focal source, and contamination electrons. One unique feature of the system is that it is fluence‐based, not a segment based calculation. A modified composite fluence map for each beam is built by summing the MLC segments and modifying for the effects of leakage, and rounded leaf edges. Model parameters are automatically determined by fitting to measurements. We re‐computed two 6 MV head & neck, and three 18 MV prostate, IMRT plans created by Pinnacle. For head & neck plans, we compared the DVHs of PTV, brainstem and parotid glands, as well as the mean dose to parotid glands. For the prostate plans, the DVHs for PTV, rectum, and bladder, as well as the D50, D98, and minimum doses for PTV are compared. Results: We found that our dose calculation system is comparable with Pinnacle for prostate IMRT plans, with small differences. For the prostate tests, the D50 for the PTV agrees within 0.7% with Pinnacle. DVHs for rectum and bladder all agree closely. The model predicts more pronounced dose inhomogeneity inside PTV in head and neck cases: the average reduction in the D98 value for the primary PTV was 5.5%. Conclusion: As expected, prostate IMRT recalculations agree well with the Pinnacle results. However, differences in head and neck results may be due to improved physics in the Monte Carlo system. The results support the use of the Monte Carlo tool as a treatment planning QA tool. Conflict of Interest: Work partially supported by grant PHS CA010953.

The Dose Planning Method (DPM) is one of several "fast" Monte Carlo (MC) computer codes designed to produce an accurate dose calculation for advanced clinical applications. We have developed a flexible machine modeling process and validation tests for open-field and IMRT calculations. To complement the DPM code, a practical and versatile source model has been developed, whose parameters are derived from a standard set of planning system commissioning measurements. The primary photon spectrum and the spectrum resulting from the flattening filter are modeled by a Fatigue function, cut-off by a multiplying Fermi function, which effectively regularizes the difficult energy spectrum determination process. Commonly-used functions are applied to represent the off-axis softening, increasing primary fluence with increasing angle ('the horn effect'), and electron contamination. The patient dependent aspect of the MC dose calculation utilizes the multi-leaf collimator (MLC) leaf sequence file exported from the treatment planning system DICOM output, coupled with the source model, to derive the particle transport. This model has been commissioned for Varian 2100C 6 MV and 18 MV photon beams using percent depth dose, dose profiles, and output factors. A 3-D conformal plan and an IMRT plan delivered to an anthropomorphic thorax phantom were used to benchmark the model. The calculated results were compared to Pinnacle v7.6c results and measurements made using radiochromic film and thermoluminescent detectors (TLD).

Purpose: To apply a measurement‐driven source model using the Monte Carlo Dose Planning Method (DPM) dose calculation engine to a Varian 10 MV photon beam. Method and Materials: A measurement‐driven model using the DPM dose calculation algorithm is being extended from a Varian 6 MV photon beam to include Varian 10 MV, Elekta 6 MV and 10 MV, and Siemens 6 MV and 10 MV photon beams. The present work details the model commissioning for the Varian 10 MV photon beam. The multi‐source model consists of a primary photon point source, an extra‐focal exponential disk source, and an electron contamination uniform disk source. The model accounts for fluence and off‐axis energy effects due to the flattening filter. The photon energy spectra for the primary and extra‐focal sources are modeled by the statistical fatigue‐failure function combined with a Fermi‐cutoff function. The energy spectrum of the electron contamination source is modeled as an exponential distribution. Model parameters are determined by an optimization process that minimizes the differences between measurement and calculation. The set of standard measurements used for optimizing consists of the percent depth dose (PDD) and dose profiles in water for 10×10 cm2 and 40×40 cm2 field sizes. Results: Comparisons between calculation and measurement of the PDD and dose profiles for the 10×10 cm2 field size show agreement within ±2%/2 mm except for the off‐axis low dose regions where calculations underestimate the dose by up to 3% of dmax. Conclusion: This work demonstrates that the model, previously shown to be accurate for the Varian 6 MV beam, can be successfully extended to the Varian 10 MV photon beam. Work is ongoing to further refine and validate the model to include Elekta and Siemens linear accelerators. Conflict of Interest: Work supported by PHS CA010953, CA081647, and R01 CA85181 awarded by NCI, DHHS.

Purpose: To benchmark a flexible Monte Carlo(MC) tool based on the Dose Planning Method (DPM) for use in evaluating Intensity Modulated Radiation Therapy(IMRT)treatment planning systems. Method and Materials: A dose calculation tool based on a flexible machine model using the Dose Planning Method (DPM),a “fast” Monte Carlo((MC)computer code, is being developed. Initial benchmark testing included a simple 10cm × 10cm multileaf collimator(MLC)diamond shaped pattern, a 3D conformal lung plan with the MLCs fully retracted, and an IMRTlung plan. Irradiations were performed using a 6MV photon beam from a Varian linear accelerator. Measurements were made in slab and anthropomorphic phantom geometries using thermoluminescent detectors(TLDs) and radiochromic film. The DPM calculation was then compared to measurements and also the calculation from the Pinnacle treatment planning system. Results: Profile comparisons from the MLCdiamond pattern irradiation showed good agreement in the penumbra region where MLC inter and intra leaf transmission effects were present. The point dose comparisons between the DPM calculation and measurement of the tumor for the 3D conformal and IMRTlung plans where within 2%. For the heart and spinal cord, the calculation for the 3D conformal and IMRTlung plans where within 7.5% of measurement, except in the conformal plan where the calculated dose point to the heart was positioned in a steep dose gradient and was 25% lower than measurement. Dose profiles through the center of the tumor showed good agreement in the PTV region, penumbra, and low doselung regions. Conclusion: This work demonstrates the feasibility of a source model based the DPM computer code to calculate dose distributions as part of the quality assurance program for clinical trials. Conflict of Interest: This work supported by PHS CA010953, CA081647, and CA085181 awarded by NCI, DHHS.

Purpose: Monte Carlo (MC) techniques are physically sound to provide accurate dose distributions. However, they take a large amount of CPU time compared to EGS4. Several fast MC algorithms have been developed, including VMC++ (Voxel Monte Carlo) and DPM (Dose Planning Method). For these fast MC codes, the simplifications of the underlying physics, variance reduction, and random number generation may not be equivalent. Moreover, implementation issues are complex and therefore testing and quality assurance is important. We compared these two codes as applied to heterogeneous media a quality assurance check. Methods and Materials: In this research, we conducted calculations for both codes on a standard open field water phantom, a water phantom with an air cavity, and a 5‐beam conformal therapy plan computed based on a CT‐scan of a heterogeneous anthropomorphic thorax phantom. The results were either compared with BEAM results, the Treatment Planning System (TPS; Pinnacle 7.6c), film or TLD measurements. The MC codes were integrated with CERR to facilitate CT‐based calculations. Results: In the water phantom, for 6MV 5×5cm2 field size at 100cm SSD, DPM and VMC++ agreed within 1%, except in the penumbra region. For 0.5×0.5cm2 field size of the air cavity test, they differed at the interface of air and water. For the 5‐beam 3D conformal plan on a thorax phantom, they agreed within 1% RMS ([STD of the difference larger than 5%Dmax]/Dmax); Most regions had a difference much less than 3% except at the buildup region for the two beams. Conclusions Carefully designed tests were conducted comparing DPM and VMC++. Water phantom results were almost identical. The air‐cavity‐heterogeneity results gave agreement within 1% except for the water‐air‐cavity interface. DPM appeared to be somewhat more sensitive to local material changes in the thorax phantom results.

With the advent of intensity-modulated radiation therapy (IMRT), the inclusion of heterogeneity corrections is further complicated by the conformal delivery of many small beams forming steep dose gradients. Radiation treatment planning has evolved to take into account even small changes in tissue density so that the dose to tumor can be further optimized. However, different treatment planning systems incorporate different heterogeneity correction algorithms, and it is unclear whether any of these algorithms are superior to others in terms of accurately predicting delivered radiation doses relative to measurement in a clinical setting. The purpose of this study was to determine the accuracy of heterogeneity dose calculations from two widely used IMRT treatment planning systems (Pinnacle and Corvus) against measurement. These two systems handle heterogeneity dose corrections by means of a collapsed-cone convolution superposition algorithm and a finite-size pencil-beam algorithm with one-dimensional depth scaling correction, respectively. Treatment plans were generated by each system using an anthropomorphic thorax phantom, routine clinical lung tumor constraints, and a common prescribed dose. Dose measurements made by thermoluminescent detectors (TLDs) and radiochromic film positioned within the phantom's lung and offset tumor insert were then compared with the calculated values. The collapsed cone convolution superposition dose calculation algorithm provided clinically acceptable results (+/-5% of the normalization dose or 3 mm distance to agreement) in the designed treatment plan and delivery. The pencil-beam algorithm with an effective pathlength correction showed reasonable agreement within the gross tumor volume, overestimated dose within a majority of the planning target volume, and underestimated the extent of the penumbral broadening, yielding only about 60% accuracy when judged by the above criterion. Even judged by a more generous criterion (+/-7% /7 mm), the results were clinically unfavorable (at only about 80% accuracy). To ascertain the dose in heterogeneous regions such as the tumor-lung interface and the peripheral lung dose near the tumor, the superposition convolution algorithm that accounts for lateral scatter and electron transport should be used. The use of the pencil-beam algorithm with only an effective pathlength correction may result in the dose to the target being overestimated. As a result, a full understanding of any treatment planning system's heterogeneity algorithm is required prior to clinical implementation.

Purpose: To provide a comprehensive study on the accuracy of many commonly used Intensity Modulated Radiation Therapy (IMRT) treatment planning systems using the Radiological Physics Center's (RPC) anthropomorphic thorax phantom. Method and Materials: Treatment planning systems (TPSs) from Corvus, Eclipse, Pinnacle, and Tomotherapy were evaluated using the RPC anthropomorphic phantom. Treatment plans were designed using the same clinical constraints and prescriptions so that 96% of the planning target volume (PTV) was covered by the prescription dose. The phantom is equipped with TLD located in the tumor, heart, and spinal cord and radiochromic film located in three anatomical planes intersecting the tumor center and extending into the lung. IMRT QA was performed to adjust the calculated dose distributions in order to isolate the effects of heterogeneity. Comparisons were made between each TPS calculation and measurement. In two instances, re‐calculations of the original correction based pencil beam (PB) plans were performed using the superposition convolution (SC) method. Results: TPSs employing superposition convolution algorithms predicted dose within 3.6% of the target TLD, while TPSs using correction based pencil beam algorithms predicted dose within 5.0% of the target TLD. Both algorithm types showed variations (2% to 38% in the cord and heart) in predicting low dose to normal structures. The dose distributions within the PTV and penumbra lung regions showed good agreement when using an SC algorithm. However, TPSs using the PB type algorithm overestimated dose in the PTV and underestimated the extent of penumbra broadening corresponding to the surrounding lung. Conclusion: This work demonstrated that superposition convolution algorithms found in widely used IMRT treatment planning systems are able to calculate the dose accurately to the PTV and penumbra regions when low density heterogeneities are involved. Conflict of Interest: This work supported by PHS CA010953 and CA081647 awarded by NCI, DHHS.