S604 I. J. Radiation Oncology d Biology d Physics Volume 72, Number 1, Supplement, 2008

CT numbers by transforming the functions to relative electron density for the Pinnacle3 treatment planning system. Three CT phan-toms of Catphan-503, including CTP486 for imaging uniformity, CTP515 for low-contrast performance, and CTP401 for CT num-bers, were scanned by both CT modes and analyzed for dose difference using the straight and horizontal beams. The Randophantom was scanned with both CT modes, and planned with a simple intensity modulated radiation therapy (IMRT) plan for do-simetric comparison of targets and organs at risk. Finally, one representative patient with nasopharyngeal cancer (NPC) had thedosimetric comparison of IMRT plan between FBCT and the on-board CBCT before the first treatment.

Results: With 360� rotational acquisition of CBCT and high-resolution reconstruction, the poor uniformity was improved andlarge standard deviations of CT numbers for background and high-density materials were reduced significantly. When comparingFBCT with adjusted CBCT, the average dose difference was 1.1 ± 0.6% in 3 CT phantoms. The dose-volume histogram of targets/organs correlated well between two CT images of Rando phantom, including target (correlation factor [CF] 0.99, p \0.001), leftparotid (CF 1.0, p\0.001), right parotid (CF 1.0, p\0.001), spinal cord (CF 0.99, p\0.001), and brain stem (CF 1.0, p\0.001).The dose curves of IMRT plans on adjusted CBCT of the NPC patient were similar to those on FBCT. The dose curves covering$90% prescribed dose on FBCT and adjusted CBCT were not different for more than 2 mm. The dose-volume histogram of targets/organs correlated well between two CT images of the NPC patient, including target (CF 0.98, p\0.001), left parotid (CF 0.99, p\0.001), right parotid (CF 1.0, p \ 0.001), and brain stem (CF 1.0, p \ 0.001).

Conclusions: Imaging adjustment program with mathematical processing on CBCT images was able to generate accurate on-boardimages volumetrically and dosimetrically for re-planning purpose.

Author Disclosure: C. Hu, None; J. Wu, None; H. Chao, None; C. Wang, None; C. Tsai, None; J. Cheng, None.

2970 Evaluation of Tumor Response and an Adaptive Treatment Strategy in Lung Radiotherapy using Cone-

beam CT

T. P. Wong, K. McCune, J. Ye, D. Cao, M. Afghan, D. Shepard, V. Mehta

Swedish Cancer Institute, Seattle, WA

Purpose/Objective(s): Kilovoltage cone-beam CT imaging serves primarily as a tool to improve the accuracy of patient position-ing in radiation therapy. In addition, cone-beam CT images can be used to observe tumor response for non-small-cell lung cancerpatients over their course of external beam radiation therapy. In this study, we have evaluated the tumor regression for NSCLCduring radiotherapy, and we have developed an adaptive treatment strategy designed to increase the dose delivered to the residualsub-volume of the tumor. The strategy also seeks to maintain the normal tissue sparing and deliver the prescribed dose to the orig-inal tumor volume in order to avoid increasing the risk of marginal recurrence.

Materials/Methods: Fifteen NSCLC patients treated between 4/2006 and 12/2007 with a conventional fractionation scheme wereanalyzed in this study. The patients were treated with a motion-encompassing treatment plan and daily CBCT-based alignmentswere performed. 4D-CT imaging was used for treatment planning. Patients were scanned in the supine position during normal re-laxed free-breathing and were immobilized with a wing-board. The internal target volume (ITV) was determined using the max-imum-intensity-projection (MiP) CT data to account for organ motion. A 5-mm PTV margin was used to account for setup error.Online setup correction was based on the daily CBCT imaging prior to treatment. To evaluate tumor regression during the course ofradiation therapy, the ITV was contoured on the first and the subsequent weekly CBCT. For patients with a decrease in ITV of morethan 50%, their treatment was re-planned to boost the residual sub-volume of the ITV while maintaining the prescription dose to theoriginal ITV and normal tissue sparing.

Results: Three out of the 15 patients had more than 25% tumor shrinkage over the course of treatment. One patient had tumorshrinkage of 71% half way through the course of treatment (18/36 fx). For this patient, a new treatment was planned for the remain-ing fractions. The 7-field IMRT plan with 95% of the ITV covered by 65 Gy was modified to include a boost to the residual sub-volume of the ITV to 70.1 Gy, with minimal changes in prescription dose to the original ITV (from 65.0 Gy to 65.7 Gy) and normaldose sparing. For example, the change in the V20 and the mean lung dose for the ipsilateral lung were 0.9% (from 6.0% to 6.9%) and2.0 Gy (from 5.2 Gy to 7.2 Gy) respectively. The maximum dose to the spinal cord remained unchanged at 42.0 Gy.

Conclusions: In addition to IGRT, CBCT can also be used to document and evaluate tumor regression during a course of radiationtherapy of non-small-cell lung cancer. An adaptive treatment strategy can be used to boost the residual sub-volume of the ITV whilemaintaining tissue sparing and avoiding the risk of increased marginal recurrence.

Author Disclosure: T.P. Wong, Elekta, B. Research Grant; K. McCune, None; J. Ye, Elekta, B. Research Grant; D. Cao, None; M.Afghan, None; D. Shepard, Elekta, B. Research Grant; V. Mehta, Elekta, B. Research Grant.

2971 Dosimetric Variability in Lung Cancer IMRT/SBRTamong Institutions with Identical Treatment Planning

Systems

I. J. Das1, C. M. Desrosiers2, S. P. Srivastava3, K. L. Chopra4, K. O. Khadivi5, M. S. Taylor6, Q. Zhao2, P. A. S. Johnstone2

1University of Pennsylvania Medical Center, Philadelphia, PA, 2Indiana University School of Medicine, Indianapolis, IN, 3ReidHospital & Health Care Service, Richmond, IN, 4Kennedy Health System, Sewell, NJ, 5Harold Leever Regional Cancer Center,Waterbury, CT, 6Kaiser Permanente, Portland, OR

Purpose/Objective(s): Lung cancer represents nearly 1/3 of the cancer burden in the USA, and treatment outcomes are relativelypoor. To improve outcome, dose escalation is being proposed; however clinical data lack dosimetric uniformity due to differencesin treatment planning systems (TPS) and algorithms such as equivalent path length (EPL), pencil beam (PB), collapsed cone (CC)and analytical anisotropic algorithm (AAA). Unlike 3D-CRT, IMRT fields are divided into segments consists of small beamlets.Small field dosimetry is challenging due to electronic disequilibrium and becomes uncertain in low density medium (lung) andespecially for high energy beams. Most of the studies that evaluated inhomogeneity correction used big fields in 3D-CRT. Dosi-metric variability among TPS is expected and has been reported in the literature. However, intra-TPS variability among institutionsusing IMRT/SBRT is not available and presented in this study.

Proceedings of the 50th Annual ASTRO Meeting S605

Materials/Methods: A widely available TPS (Eclipse; Varian Medical Systems) was evaluated in this study. Measurements wereperformed in a lung phantom made out of cork (p = 0.25 g/cm3) sandwiched between solid water for 1 x 1-10 x 10 cm2 fields rep-resentative of IMRT beamlets or SBRT fields. The phantom was scanned and images were sent to the participating institutions fortreatment planning. The inhomogeneity correction factor (CF = ratio of dose with and without inhomogeneity) was calculated atvarious depths and field sizes for 6 and 15 MV beams. The calculated CF data were sent to a central location for analysis.

Results: The CF based on TPS showed significant variability (220%) among institutions with identical TPS. The differencesamong algorithms are more dramatic compared to the measured CF. In general, differences are more pronounced for small fields,deeper depths, high energy beams and non-electron transport algorithms. The range of CF differences between measured and TPSof +90.3% to -19.0% and 237% to -11.8% were observed for 6 MV and 15 MV beams respectively. Differences are smaller for .4x 4 cm2 fields: 10.5% to -11.1%, 10.9% to -18.8% for AAA and PB respectively for 6 MV and 8.6% to -9.1% and 16.7% to -11.8%for 15 MV beam.

Conclusions: Intra-TPS dosimetric variability is extremely large and cannot justify comparison of outcome data among centers ina clinical trial. Advance algorithm with 3D electron transport (AAA) is required to produce accurate dosimetry in lung where dif-ferences are minimum (\5%) and agrees in entire range of field sizes to ±10%. It is concluded that AAA should be the algorithm ofchoice in Eclipse TPS and used over PB and EPL in lung treatment planning for IMRT/SBRT. The large intra-TPS variations couldbe reduced with proper CT to electron density table, smaller grid size and proper bench marking in small fields during TPS com-missioning.

Author Disclosure: I.J. Das, None; C.M. Desrosiers, None; S.P. Srivastava, None; K.L. Chopra, None; K.O. Khadivi, None; M.S.Taylor, None; Q. Zhao, None; P.A.S. Johnstone, None.

2972 Motion Compensated Volumetric Modulated Arc Therapy (4D VMAT)

O. Keshet1, L. Lee2, Y. Ma2, L. Xing2

1Stanford University, Stanford, CA, 2Stanford University School of Medicine, Stanford, CA

Purpose/Objective(s): 4D-CT provides temporal information that can be explicitly incorporated in modulated arc therapy doseoptimization to accommodate temporal changes in anatomy that occur during treatment. Here we present a concept and proofof feasibility of organ motion compensated VMAT (4D VMAT) optimization.

Materials/Methods: A cylindrical digital phantom with a moving target was designed. The involved structures move cyclically in10-phase with the target movement range and period of 1.5 cm and 6 sec, respectively. The 4D VMAT plan was constrained toa single pass of 360o gantry rotation. The target cyclic motion trajectory was replicated to form a ‘breathing’ curve to which a cor-relation with the full gantry rotation was established. A reference phase was chosen to which all spatial and dosimetric informationfrom other phases were mapped onto through the use of a deformable registration model, and the inverse planning goals were de-fined in this reference phase. The full rotation was initially modeled as a series of fixed source positions for every 5o starting from0o. The initial aperture at each gantry angle was chosen to be the beam’s eye view of the target at the phase determined from thephase/gantry-angle correlation. An aperture-based optimization using simulated annealing algorithm was adopted to optimizea weighted least-squares cost function with volumetric constraints. The dose deposited for each angle was mapped to the referencephase for the calculation of cumulative dose and assessment of the cost value. The process terminated when the cost function con-verged. The resultant 4D dose from the complete rotation was calculated by mapping the dose deposited at various angles (corre-lated to different phases) onto the reference phase. The 4D VMAT plan and dose volume histograms were studied and comparedwith those of the 3D VMAT plans optimized with the target volume encompassing the whole motion range. Two lung cases with4D-CT scans were also selected and tested using the developed 4D VMAT procedure.

Results: A motion compensated VMAT plan optimization framework has been formulated that enables arc delivery with a 100%duty cycle. The 4D VMAT plan of both the phantom and lung cases yielded significantly improved coverage of the target andsparing of the surrounding normal tissues when compared to the 3D VMAT plan.

Conclusions: 4D VMAT is feasible based on (1) the aperture-based optimization that takes ‘time’ as a parameter expressed ina form of phase/gantry-angle correlation and (2) the spatial/dosimetric mapping between phases using deformable registration.Real-time image guided adaptation of the 4D VMAT plan to account for a possible breathing irregularity is in progress. The currentstudy helps in assessing the degree of improvement of 4D VMAT over its 3D counterpart and providing a theoretical benchmark for4D VMAT.

Author Disclosure: O. Keshet, None; L. Lee, None; Y. Ma, None; L. Xing, None.

2973 Evaluation of Target Volume Reconstruction from 4D-PET/CT Imaging using a Dynamic Phantom and

Patient Data

D. Ionascu1, S. Park1, M. Mamede2, J. Killoran1, V. Gerbaudo2, R. Berbeco1

1Dana Farber/Brigham and Women’s Cancer, Boston, MA, 2Brigham and Women’s Hospital/Division of Nuclear Medicine,Department of Radiology, Boston, MA

4D PET/CT imaging is a powerful tool to assist clinicians in accurate definition of gross tumor volumes (GTV) in the lung andabdomen. We present the results of phantom studies designed to evaluate the accuracy of 4D PET/CT at reproducing the true tri-dimensional trajectory of a given GTV based on pre-recorded data from real patients.

Measurements were conducted using a 4D dynamic phantom control system (developed at Washington University in SaintLouis) capable of reproducing a time-dependent 3D tumor motion, previously measured in lung patients using implanted fiducials.In addition, a decoupled 1D stage moved concurrently with the internal 3D data to reproduce the rise and fall of the patient’s ab-dominal surface. This motion was also based on pre-recorded patient data. To represent a tumor, an 8 ml spherical vial of FDG wasused within a water-FDG mixture adjusted to produce a target-to-background activity ratio of 8:1. The image data was obtainedusing a GE Discovery CT-PET scanner and reconstructed into 10 phase bins over the respiratory cycle. The scanner gating system