P. Lindsay

Clinical implementation of electron energy changes of Varian linear accelerators

Modern dual photon energy linear accelerators often come with a few megavoltage electron beams. The megavoltage electron beam has limited range and relative sharp distal falloff in its depth dose curve compared to that of megavoltage photon beam. Its radiation dose is often delivered appositionally to cover the target volume to its distal 90% depth dose (d90), while avoiding the normal ‐ sometimes critical ‐ structure immediately distal to the target. Varian linear accelerators currently offer selected electron beams of 4, 6, 9, 12, 16 and 20 MeV electron beam energies. However, intermediate electron energy is often needed for optimal dose distribution. In this study we investigated electron beam characteristics and implemented two intermediate 7 and 11 MeV electron beams on Varian linear accelerators. Comprehensive tests and measurements indicated the new electron beams met all dosimetry parameter criteria and operational safety standards. Between the two new electron beams and the existing electron beams we were able to provide a choice of electron beams of 4, 6, 7, 9, 11, 12, 16 and 20 MeV electron energies, which had d90 depth between 1.5 cm and 6.0 cm (from 1.5 cm to 4.0 cm in 0.5 cm increments) to meet our clinical needs. PACS number: 87.56.bd.

SU‐FF‐T‐501: The Effect of Monte Carlo‐Based Dose Calculations On Tumor Control Probability Modeling

Purpose: It has not been established whether accurate lungdosimetry really has an impact on metrics and modeling related to outcome prediction. We studied the impact of Monte Carlo‐based dose calculations on TCP modeling of lungcancertumors.Methods and Materials: All analyzable patients treated for NSCLC between 1991 and 2001 at Washington University in St. Louis (n=56) had TCP modeled using water equivalent dosimetry and Monte‐Carlo corrected dosimetry in turn. Clinical factors in the modeling included age, gender, chemotherapy, performance status, weight loss, and smoking. Dosimetric variables included Vx, Dx, maximum dose, mean dose and minimum dose covering the primary GTV. The best multivariate logistic models for both treatment planning system dose (homogeneous, uncorrected for tissue heterogeneities) and Monte‐Carlo corrected dose were obtained. Results: The best multivariate model for homogeneous dose calculations is a six‐parameter model (Spearmans rank correlation coefficient, R=0.62) including V75_TCPgtvPlan, V80_TCPgtvPlan, Mindose_TCPgtvPlan, age, PreTxChemo and gtvVol. However, there was not a strong preference for specific dose‐volume terms, and many models had a similar predictive power. The best multivariate model for Monte‐Carlo corrected dose is a two‐parameter model (R=0.527) including V75_TCPgtvMC and gtvVol, and the modeling strongly indicated that only two terms were necessary to explain the data. Interestingly, the Monte Carlo model has the superior ‘face validity” while the water‐based model resulted in a higher overall predictive power. The reasons for this are currently unknown. Conclusions: Monte‐Carlo corrected dose significantly impacts resulting TCP modeling, which demonstrates the need accurate modeling. Differences in the resulting models between water‐based and Monte Carlo based dosimetry remain poorly understood and will require further research. Partially supported by NIH R01 grant CA85181.

SU‐GG‐T‐404: The Impact of Breathing‐Motion and Tumor Regression On Dose‐Volume Metrics Used for Outcome Analyses

Purpose: Breathing motion and tumor regression may affect the correlation between derived treatment planning dose‐volume metrics and local control, especially for lungtumors. We investigated this effect for 3‐D plan archives that use older computed‐tomography scanning protocols that effectively averaged over multiple breathing cycles. Method and Materials: To simulate tumor motion due to breathing, we convolved the static planned dose distribution with a probability distribution function (PDF), which describes the nature of GTV motion due to breathing, thereby creating breathing motion weighted dose distributions. To simulate tumor regression, a mathematical model based on experimental observations of NSCLCtumor regression (Ramsay et al., IJROBP (2006) 64(4):1237–44) was applied, with the additional needed assumption that cell loss is uniform, to describe the decrease of GTV volume and track dose through the treatment. Datasets from two patients, one with a large GTV volume (315.3736cc), the other with a GTV volume (0.6644cc) were used. Monte Carlo recomputed dose distributions were used for improved dosimetric accuracy. GTV DVHs incorporating breathing motion, and incorporating both breathing motion and tumor regression were derived to estimate ‘true’ DVHs. Results: For the small tumor, the effect of breathing dominated, and the ‘true’ GTV was worse than the planned GTV (D98 decreased by 2.27%). For the large tumor, the effect of regression dominated and the ‘true’ dose distribution was significantly better than planned dose distribution (D98 increased by 1.2%). Of these two effects, tumor regression generally had larger impact. Conclusion: Breathing and tumor regression are likely to be confounding factors when retrospectively analyzing lung treatment plans for tumor control probability analyses. A model has been developed to account for these effects in an approximate fashion. This model will be tested for an ability to potentially improve correlations between derived metrics and local control. Partially supported by NIH R01 grant CA85181.

The denoising of Monte Carlo dose distributions using convolution superposition calculations.

Monte Carlo (MC) dose calculations can be accurate but are also computationally intensive. In contrast, convolution superposition (CS) offers faster and smoother results but by making approximations. We investigated MC denoising techniques, which use available convolution superposition results and new noise filtering methods to guide and accelerate MC calculations. Two main approaches were developed to combine CS information with MC denoising. In the first approach, the denoising result is iteratively updated by adding the denoised residual difference between the result and the MC image. Multi-scale methods were used (wavelets or contourlets) for denoising the residual. The iterations are initialized by the CS data. In the second approach, we used a frequency splitting technique by quadrature filtering to combine low frequency components derived from MC simulations with high frequency components derived from CS components. The rationale is to take the scattering tails as well as dose levels in the high-dose region from the MC calculations, which presumably more accurately incorporates scatter; high-frequency details are taken from CS calculations. 3D Butterworth filters were used to design the quadrature filters. The methods were demonstrated using anonymized clinical lung and head and neck cases. The MC dose distributions were calculated by the open-source dose planning method MC code with varying noise levels. Our results indicate that the frequency-splitting technique for incorporating CS-guided MC denoising is promising in terms of computational efficiency and noise reduction.

WE‐C‐AUD‐10: A Comparison of DPM and VMC++ Mont Carlo Codes Applied to Heterogeneous Media

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 5×5cm2field size at 100cm SSD, DPM and VMC++ agreed within 1%, except in the penumbra region. For 0.5×0.5cm2field 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.

SU‐FF‐T‐195: Effect of Respiratory Gating On the Dose Distributions of Dynamic Wedge and IMRT Treatment Fields

Purpose: To determine the magnitude of error introduced into treatmentdelivery due to timing effects between respiratory‐gated delivery and dynamic beam delivery (enhanced dynamic wedge (EDW) and step‐and‐shoot intensity modulated radiation therapy(IMRT)).Method and Materials: EDW and IMRT fields were delivered on a linear accelerator (Trilogy, Varian Medical Systems). Gating of the beam was achieved using a commercial respiratory monitoring system (RPM, Varian Medical Systems), and respiratory motion was simulated using a commercial respiratory motion phantom (RPM phantom, Varian Medical Systems). All fields were delivered with no gating, a 1‐mm gating window, and a 2‐mm gating window. Film and ion chamber measurements were made for both EDW and IMRT fields in a water‐equivalent IMRT QA phantom, which was stationary during all measurements. Film measurements used EDR‐2 film, while ion chamber measurements were made with a 0.04 cc volume pin‐point chamber. Leakage current was subtracted from the ion chamber readings. Wedged fields of 15 and 60 degrees were delivered at a dose rate of 600 MU/min. Results were evaluated for the full IMRT plan and for one of the individual beams, at dose rates of 400 and 600 MU/min. Results: All ion chamber measurements were analyzed as percent difference between the gated (1‐mm or 2‐mm gating window) and non‐gated delivery. All differences were less than 1%. There were no significant differences between 1‐mm and 2‐mm gated delivery. Film results, as analyzed from isodose curves and the gamma metric, showed no substantial differences between gated and non‐gated beam delivery.Conclusion: The use of gated treatmentdelivery of EDW and step‐and‐shoot IMRT fields does not introduce clinically significant differences into the resulting dose distributions, as quantified by film and ion chamber measurements.

SU‐FF‐J‐21: A Comparison of a Point Based Tool with An Image Overlay Tool for Fiducial Based Setup

Purpose: To compare interactive image overlay tools for fiducial based setup with a point based system. Method and Materials: A single implanted fiducial was used for localization. Gated kVp images were acquired for daily setup; if a shift was made, a 2nd set of images were acquired subsequent to the shift. Shifts were determined using both a commercial software package designed for point matching (ISOLOC, CIVCO, Kalona, IA) and also by dragging the projection of the fiducial contour over the image of the fiducial using tools integrated into the treatment delivery console(4DTC, Varian, Palo Alto, CA). Data was collected for 29 shifts on 15 different days. Results: The mean magnitude of the initial shifts were 1.5, 5.7, and 1.5 mm and the means of the differences between the shifts were 0.2, 1.1 and 0.8 mm in the lateral, cranial‐caudal, and AP directions respectively. On many days there was a discrepancy of 1–2 mm between the AP and lateral gated kVp X‐rays. The interactive tools allowed the user to compromise between the two projections, while the automatic system forced a compromise. The interactive tools also displayed a projection of the tumor allowing validation of the position of the fiducial with respect to the bulk tumor. The interactive tools were quicker to use and automatically applied the shifts; these tools are somewhat limited by the accuracy of the contouring of the fiducials. Conclusion: Both the interactive tools and the automatic system are practical for daily setup of patients with fiducials, but the integration into the console, the speed of localization, and most importantly the ability to correlate the fiducials with radioopaque anatomy make the interactive tools superior.

SU‐FF‐J‐04: Uncertainties in Respiratory Gating for Lung Tumors

Purpose: The purpose of this work is to quantify the uncertainties in tumor location during respiratory‐gated radiation delivery when using implanted fiducial‐based setups and respiratory gating to deliver the treatment.Method and Materials: A patient treated for non‐small‐cell lungcancer was set up daily based on a fiducial implanted in the tumor. The fiducial was visualized using AP and Lateral kilovoltage x‐rays gated on end expiration. Audio prompting was used to ensure regular breathing. Following alignment, a second pair of gated images was acquired to verify the shift. The uncertainties in the gating system were quantified by the residual error in fiducial location after the initial shift. In addition, cine megavoltage images were acquired of an AP field during the gated treatment (6–10/sec) and the fiducial positions were measured to quantify uncertainties in position between successive respiratory gates. Results: Gating uncertainties measured by the residual error of the patient shift were determined to be 1.4, 3.3, and 1.3 mm for the LR, SI, and AP directions, respectively. The variation each day in the fiducial position averaged over each respiratory gate was determined to be 1.0 and 2.2 mm in the LR and SI directions respectively. The variation in fiducial position in successive respiratory gates over the course of treatment was measured to be negligible in the LR direction and 1.8 mm in the SI direction. Conclusion: The uncertainties in respiratory gating were 3 mm over the course of treatment and 2 mm measured each day on successive gates used to deliver the treatment. Although respiratory gating may reduce tumor motion, the uncertainties in tumor position during respiratory gating must be considered when the internal margin is being designed.

SU‐FF‐T‐165: Dose Measurement Accuracy of TLD‐100 in the Bragg Peak Region of a Therapeutic Proton Beam

Purpose: Measurements made in a proton beam indicated differences in dose calculated by the Eclipse (Varian, Milpitas, CA) treatment planning system and TLD measurements in a lung phantom. The differences were exceptionally large in the distal fall‐off region of the spread out Bragg peak. This work investigates the accuracy of the TLD in regions of high LET as a potential cause for the dose differences seen. Methods and Materials: All measurements were made along the central axis of an unmodulated 200 MeV proton beam, at a source‐to‐axis distance of 270 cm, 10×10 cm2field size, at varying depths along the Bragg peak. Measurements were made using TLD‐100 powder flat packs, placed in a virtual water slab phantom. To validate our TLD results, the measurements were repeated using a parallel plate ionization chamber.Results: The dose measurements using TLD‐100 in a proton beam were accurate to within ±5.0% of the expected dose typically seen in photon and electron measurements. The ionization chamber and the TLD relative dose measurements agreed well with each other. Absolute dose measurements using TLD agreed with ionization chamber measurements to within ± 3.0 cGy for an exposure of 100 cGy. Conclusion: The accuracy of the TLD is not a potential cause for the dose differences observed between the Eclipse calculations and our TLD measurements. This was further substantiated by the agreement of our ionization chamber measurements with TLD. Additional work must be done to identify the exact cause of dose discrepancy.