Purpose: To analyze the targeting accuracy using Exac‐Trac® (BrainLAB) and Synergy® CBCT (Elekta) systems for lung SBRT, and discuss the difference between these two IGRT systems. Methods: The target localization for lung patients treated with SBRT on Novalis® with Exac‐Trac® system and on Synergy® with CBCT system was analyzed. Prior to each treatment, patient's setup correction using Exac‐Trac® is performed by fusing (bone structure matching) two oblique x‐ray images with related DRRs computed from the planning CT, the correction using the CBCT system is by fusing (grey volume registration) CBCT with the planning CT. The position error is adjusted by shifting the table in lateral, longitudinal and vertical directions. If the shifted distance is >5mm in one direction, we usually repeat the image capturing and fusing procedures to double check the position error—the “second” shifting distance. In this study, we used the “second” shifting distance to analyze the targeting accuracy for both systems. Results: Sixteen and twenty‐nine setup correction cases were studied for Exac‐Trac® and Synergy® CBCT systems, respectively. The first shifted distances were in the ranges of 8.6–22.7mm and 3.5–18.3mm, with the average of 15.2mm and 8.2mm. The “second” shifting distances were in the ranges of 0.5–4.9mm and 0.6–3.4mm, with the averages of 2.8mm and 1.9mm, respectively. For both systems there was no correlation between the first shifted and the “second” shifting distances. The average position error after table shifting in Exac‐Trac® was larger than that in Synergy® CBCT system. We think the 3D volume registration is more practical or accurate than the 2D imaging matching for soft tissue localization. Conclusion: Using volume registration, Synergy® CBCT system is more accurate compared to Exac‐Trac® system in IGRT for lung SBRT. Detailed comparison and discussion of targeting accuracy for different tumor locations between the two systems will be reported.

Although some investigators have reported acceptably low toxicity after stereotactic body radiation therapy (SBRT) for centrally located lung tumors (IJROBP 72:967, 2008), others have raised concerns about the tolerance of SBRT in this setting (JCO 24:4833, 2006). In the series in which excess toxicity was observed, dose calculations were performed without heterogeneity correction (HC), whereas in the series without toxicity HC was used. We hypothesize that unrecognized irregularities in the proximal airway dose distributions might have contributed to toxicity if dose was calculated without HC. The treatment plans of 6 patients treated with SBRT for centrally located primary or metastatic lung tumors were studied. For each the adjacent 3cm of the ipsilateral proximal bronchial tree (PBT-3cm) was contoured separately in addition to normal lung. The PTV included the GTV + 5mm radial and 10mm sup-inf margin. Multiple 6MV beam dynamic conformal arcs were used to provide PTV coverage to the prescription dose, which was first calculated Pencil Beam (PB) algorithm without HC and then re-calculated with a Monte Carlo (MC) treatment planning system incorporating HC (iPlan RT 4.0; BrainLAB AG), using the same number of monitor units for each arc. The parameters compared included the dose covering 95% of the PTV (D95), the ratio of volume covered by the 50% isodose to the PTV (R50), the max dose 2cm from the PTV (Dmax2cm), and mean and max dose to PBT-3cm (DPBT-3cm). The median PTV was 28 cc (range, 8-66). For all tumor and normal tissue dose parameters analyzed, the dose calculated with MC was higher than the dose calculated with PB without HC. The mean ratios of MC dose to PB dose without HC were as follows: D95, 1.37; R50, 1.27; Dmax2cm, 1.07; mean DPBT-3cm, 1.18; max DPBT-3cm, 1.19. For all the dosimetric parameters compared, the difference between MC and PB dose calculations was statistically significant (p < 0.05). The current analysis suggests that for centrally located lesions, SBRT doses given to tumor and important normal tissues can be roughly one fourth to one third higher than expected if the calculation algorithm does not account for tissue heterogeneity. Thus, although other factors might be contributory, for example differences in patient selection, at least one important consideration in the explanation of toxicities sometimes seen after SBRT for proximal lesions is the potential difference in dose administered relative to planned dose. Future studies to correlate dose with the incidence of proximal airway toxicity should account for this issue carefully.

Purpose: Although it has been reported that breast cancer patients have elevated risks of secondary stomach and esophagus cancers after breast radiotreatment, dosimetric information in those organs is sparse. The purpose of this study is to investigate radiation dose to specific organs at risks (OARs), specifically, stomach, esophagus and pancreas and to find factors associated with increased dose exposure. Method and Materials: Fifty four breast cancer patients treated with CT based 3D conformal treatment plans were selected to include variations in laterality (left side vs. right side), treatment volume (with or without supraclavicular fossa) and treatment techniques (whole vs. partial breast irradiation, supine vs. prone position). Treatment plans for these patients were recalculated after the OARs were contoured. Dose were calculated using the XiO (Version 4.40). Results: Patients treated to the left breast had significantly higher dose in the stomach, with average maximum dose >400 cGy when whole breast was irradiated in supine position. Prone positioning resulted in a lower stomach dose compared to supine, with an average of 200 cGy maximum. Partial breast treatment was associated with lowest dose to stomach. Similar trend is seen in the mean dose. Further regression analysis indicated that stomach dose was positively correlated with stomach volume and reversely correlated with distance to isocenter (p<0.05). Elevated esophagus is seen in patients with supraclavicular treatment, regardless of laterality. Dose to pancreas is small (maximum average < 25 cGy). Conclusion: Radiation dose to stomach and esophagus can be substantial during breast radiotherapy treatment, depending upon laterality, target volume and treatment techniques. Prone position and PBI appear to decrease dose to stomach slightly. Dosimetric information from this work provides important base for further analyses of dose-response relationship for secondary gastrointestinal malignancies in breast cancer patients and for selection of treatment techniques to lower dose in gastrointestinal organs.

Accuracy of a new Monte Carlo (MC) and a Pencil Beam (PB) dose calculation is investigated. Point dose measurements are compared to MC and PB calculations. Differences in heterogeneity corrections between MC and PB algorithms are compared to dosimetric measurements in lung. A 6MV photon beam (Novalis) and ion chamber measurements were used. MC and PB dose calculations were performed using a beta iPlan-4.0 (BrainLAB) planning system. Three isocenters were selected in a thorax heterogeneous phantom (CIRS) where lung and tissue densities are 0.21g/cm3 and 1.06g/cm3. Iso-1 is in the middle of the phantom with an irregular PTV of 177.2 cm3. Iso-2 is close to the left lung. Iso-3 is at the center of a tissue equivalent rod inserted in the left lung. Iso-2 and iso-3 have the same PTV of 12.5 cm3. Dynamic arc (3 arcs with table angles: 0°, 40° and 320°; and gantry angles: 300° - 60°) and 3D-Conformal plans (5 beams with gantry angles: 0°, 72°, 144°, 216° and 288°) are generated for each isocenter. A dose prescription of 10 Gy with 100% PTV coverage was assigned for all the plans. With the same MUs, MC and PB dose calculations at the three isocenters were compared to measurements. The dose in lung was also measured and compared to the calculations for the Iso-1 plan. For the iso-3 plans, dose ratios of PB to MC at different distances from the PTV periphery were calculated in the anterior, lateral, medial, and posterior directions, and isodose distributions were compared between MC and PB calculations for 90% and 80% dose coverage. For the dynamic arc plans, all the MC dose were within 1% agreement to measurements, while all the PB doses were <2% variation at the three isocenters. For the 3D-conformal plans, both MC and PB dose were <1% at Iso-1. The two MC dose values were <2% at both Iso-2 and Iso-3, while the PB dose was <1% at Iso-2, and was <2% at Iso-3. Comparing the measurement dose in lung, the MC dose from the dynamic arc and 3D-conformal plans were ∼0.3% and ∼3% off. The PB doses were ∼20% and ∼30% off respectively for Iso-1. For Iso-3 the doses from PB and MC in lung were calculated from 0 to 30mm from the PTV periphery in four directions. Near the PTV, PB was consistently higher by roughly 25% for dynamic arc and 20% for 3D conformal. The PB isodose lines were expanded >3mm for 90% coverage, ∼2.5mm for 80%. Farther from the PTV the differences between PB and MC were less. Film dosimetry confirmed the results. Based on our study using the BrainLAB planning system, we found that both MC and PB algorithms are accurate for the dose calculation for tumors in unit density tissue. Using the dynamic arc planning, MC was more accurate compared to PB. For lung dose calculations, MC shows much great fidelity to measurement than the PB calculations.

In this study, a 4D treatment planning tool using an analytical model accounting for breathing motion is investigated to evaluate the motion effect on delivered dose for lung cancer treatments with three-dimensional conformal radiotherapy (3DCRT). The Monte Carlo EGS4/MCDOSE user code is used in the treatment planning dose calculation, and the patient CT data are converted into respective patient geometry files for Monte Carlo dose calculation. The model interpolates CT images at different phases of the breathing cycle from patient CT scans taken at end inspiration and end expiration phases and the chest wall position. Correlation between the voxels in a reference CT dataset and the voxels in the interpolated CT datasets at any breathing phases is established so that the dose to a voxel can be accumulated through the entire breathing cycle. Simulated lung tumors at different locations are used to demonstrate our model in 3DCRT for lung cancer treatments. We demonstrated the use of a 4D treatment planning tool in evaluating the breathing motion effect on delivered dose for different planning margins. Further studies are being conducted to use this tool to study the lung motion effect through large-scale analysis and to implement this useful tool for treatment planning dose calculation and plan evaluation for 4D radiotherapy.

Purpose: To investigate the respiratory motion effect on lung tumor radiotherapy using 4D Monte Carlo treatment planning and 4D CT. Method and Materials: 4D CT images for four lung patients (two upper lung tumors and two lower ones, with different volumes) were acquired by using a GE LightSpeed‐QX/I scanner. Ten phase bins were used in the 4D‐data acquisition. A 4D Monte Carlo treatment planning system based on the EGS4/MCDOSE code was developed to calculate the 3D dose and map the 3D dose of the CT at each phase to the inhale CT (as reference). CT images at different phases were registered with the inhale CT image using a BSpline deformable registration model. Isodose lines and the DVHs of tumor and normal lung were used to compare the 4D plan (3D dose mapped from the CT at 10 selected phases to the inhale CT) and the 3D reference plan (3D dose for the inhale CT) for each patient. Respiratory motion effect was investigated for the different tumor volumes and locations. Results: In our study, for the upper lung tumors, the respiratory motion effect on target dose coverage was insignificant (<2% difference between the 4D plan and the 3D reference plan). However, for the lower lung tumors, the motion effect was clinically significant (>3% difference between the 4D and 3D plans). For the same PTV margin, less motion effect was observed for larger tumor volumes. The motion effect for the normal lung volume receiving dose was not correlated with the tumor volume or the location. Conclusion: Respiration motion may significantly influence the tumor dose in lung radiotherapy and 4D dose calculation may be necessary when treating tumors in the lower lung or when the tumor volume is small.

Purpose: We analyzed the accuracy of stereoscopic image‐guided radiotherapy (S‐IGRT) with the ExacTrac® system (BrainLAB AG, Heimstetten, Germany), and demonstrated the dosimetric advantage for lung cancer treatment. Methods: The accuracy of target localization using the ExacTrac system was analyzed by re‐scanning ten patients immobilized in the Vac‐loc bag with BB tags put on the isocenter marks determined by the ExacTrac system. The re‐scanned CT data from each patient was fused and compared to the isocenter locations on the original CT used for the treatment planning. The Active Breathing Coordinator™ (Elekta, Norcross, GA) was employed to minimize the breathing motion effect on S‐IGRT. The dosimetric advantage of S‐IGRT was demonstrated by comparing the normal lung volume receiving 15%, 30% and 50% of the prescription dose (V15%, V30% and V50%) between the plans with smaller margins in the S‐IGRT and those with normal margins in the conventional radiotherapy for the ten patients. Results: The average isocenter shifts using S‐IGRT were within 3.4 ±1.7mm in the lateral, 3.6±1.9mm in the anterior/posterior, and 2.2±3.2mm in the superior/inferior directions. The added margins around the CTV to create the PTV were chosen to be 10mm in the superior/inferior and 5mm in radial direction for the lung cancer treatment planning in S‐IGRT. For the ten patients with CTV volume from 8.3cm3 to 43.4cm3 and lung volume from 2736cm3 to 3640cm3, the averages of V15%, V30% and V50% were 17.4%, 9.8%, 5.9% for the S‐IGRT plans, and 33.4%, 18.2%, 12.8% for the conventional radiotherapy plans with 2cm margins in the all directions. Conclusion: The S‐IGRT with the ExacTrac system provided highly accurate tumor localization. The margins from CTV to PTV in S‐IGRT treatment planning could be reduced significantly due to the accurate target localization which will reduce the lung volume receiving doses in the medium and low ranges.

Purpose: To investigate dosimetric differences among three-dimensional conformal (3D-CRT), dynamic conformal arc therapy (DCAT) and intensity modulated radiotherapy (IMRT) for brain tumor treatment for a broad range of brain tumor volumes and shapes in an effort to determine whether a preferred method can be identified based upon pre-treatment characteristics. Method and Materials: Fifteen patients treated with Novalis were selected. We performed 3D-CRT, DCAT and IMRT plans for all the cases. The beam numbers in 3D-CRT or IMRT plans were the same as the arc numbers in the DCAT plans, and the gantry angle of each beam in 3D-CRT or IMRT plans was the middle angle of each arc in the DCAT plans. The PTV margin was re-chosen as 1mm, and the specific prescription dose was re-set to 90% for all the plans. The target coverage at prescription dose (TV90%), conformity index (CI) and heterogeneity index (HI) were used to compare the different plans. V50% and V80% of the organs at risk (OAR) were also calculated. Results: For small brain tumors (PTV⩽2cc), three dosimetric parameters had approximate values for both 3D-CRT and DCAT plans . The CI for IMRT plans was high . For medium brain tumors (2cc

Purpose: Stereotactic body radiation therapy (SBRT) for non‐small cell lung cancer has been shown to limit toxicity. Heterogeneity correction on lung cancer radiotherapy has not been recommended by RTOG. In this study, dosimetric difference between the SBRT plans with/without heterogeneity correction is analyzed. Method and Materials: Nine lung cancer patients treated with SBRT techniques using a 6 MV Novalis system were selected. Using the path length algorithm in BrainLAB treatment planning system, all the treatment plans were applied the heterogeneity correction. With same beam parameters, we performed the plans without the heterogeneity correction, and compared the dosimetric difference to the treatment plans. The heterogeneity correction factors (Kc) at iso‐center, target coverage, heterogeneity index (HI), and conformity index (CI) were used in the comparison. Results: The average of Kc at isocenter for 14 planning target volumes (PTV) was 1.002±0.02, only three Kc values (1.07, 0.985, 0.981) were relatively off to the average. Except one case, the other 13 target coverage values of the plans with the heterogeneity correction were better than those without the correction. The maximum difference of the target coverage was ∼7%. All the HI values of the plans with the heterogeneity correction were better than those without the correction. The maximum difference of the HI reached 300%. The difference of the CI between the compared plans was within ∼10%. The CIs of the plans without correction were better than those corrected. For one case — the tumor located in the lung base, the impact of heterogeneity correction was significant. Conclusion: The impact of heterogeneity correction for tumor dosimetry on SBRT for lung cancer is case depended. For most primary lung tumors the difference between the plans with/without heterogeneity correction is clinically insignificant. For the tumors near the interface of different mass density, the heterogeneity correction is necessary.

The purpose of this study is to evaluate the dosimetric accuracy of MRI-based treatment planning for prostate cancer using a commercial radiotherapy treatment planning system. Three-dimensional conformal plans for 15 prostate patients were generated using the AcQPlan system. For each patient, dose distributions were calculated using patient CT data with and without heterogeneity correction, and using patient MRI data without heterogeneity correction. MR images were post-processed using the gradient distortion correction (GDC) software. The distortion corrected MR images were fused to the corresponding CT for each patient for target and structure delineation. The femoral heads were delineated based on CT. Other anatomic structures relevant to the treatment (i.e., prostate, seminal vesicles, lymph notes, rectum and bladder) were delineated based on MRI. The external contours were drawn separately on CT and MRI. The same internal contours were used in the dose calculation using CT- and MRI-based geometries by directly transferring them between MRI and CT as needed. Treatment plans were evaluated based on maximum dose, isodose distributions and dose-volume histograms. The results confirm previous investigations that there is no clinically significant dose difference between CT-based prostate plans with and without heterogeneity correction. The difference in the target dose between CT- and MRI-based plans using homogeneous geometry was within 2.5%. Our results suggest that MRI-based treatment planning is suitable for radiotherapy of prostate cancer.

The goal of lung cancer radiotherapy is to improve tumor control without increasing toxicity and compromising quality of patient life. In this study, we used an analytic model accounting for breathing motion to evaluate the motion effect on delivered dose for lung cancer treatments with three-dimensional conformal radiotherapy (3D-CRT) and intensity modulated radiotherapy (IMRT). The gated radiotherapy in 3D-CRT and the dynamic multileaf collimator technique (where the delivered beam position changes synchronously with respect to target motion) in IMRT were investigated to optimize the treatments. Taking into account the thoracic motion and lung tumor shift independently, an analytical model has been developed to reconstruct a patient geometry during treatment based on two initial CT data taken at the inspiration and expiration phases and the chest wall motion measured using an optical motion detector. The 3D dose data for any patient geometry can be obtained from Monte Carlo simulations with EGS4/MCDOSE code. Correlation between the voxels in the inspiration geometry and the voxels in the reconstructed geometry at any point of a breathing cycle is established so that the dose to a voxel can be accumulated accurately during a treatment. For lung cancer treatments with 3D-CRT and IMRT, breathing motion effect on delivered dose to the targets and the critical structures were analyzed for different target margins. Gated radiotherapy with different thresholds and the dynamic MLC technique with different time delay of leaf movement were studied for individual patients based upon considerations of anatomy and respiratory motion. In this study, a 0.5cm chest wall movement resulted in ∼1cm change for a tumor in the top portion of the lung, and ∼1.5cm shift in the lower portion of the lung, in the superior-inferior direction. Our results show that if 1cm target margins were chosen in the 3D-CRT, the breathing motion would cause cold spots in target dose coverage for these tumors. If 2cm margins were applied, target dose coverage would not be affected by the breathing motion, however, the lung dose would increase by up to ∼60%. Gated radiotherapy applied in 3D-CRT could reduce the motion effect with suitable thresholds. However, the balance between elimination of motion effect and the treatment time should be considered in the clinic. We found that IMRT plans could improve the dose conformity to the target and reduce the lung volume (∼30%) that received a high dose compared with the 3D-CRT plans with the same gantry angles and the same number of beams. Furthermore, if we use the dynamic MLC technique including breathing motion into the MLC leaf sequencing, the dose to the adjacent normal tissues and critical structures could be greatly reduced by applying smaller treatment margins. If there was 0.1 second time delay between the breathing motion tracking system and the movement of MLC, the discrepancy would be about 2% in the target dose coverage, which was clinically acceptable. In lung cancer radiotherapy, the target dose can be improved by increasing the treatment margins, but this will result in more normal tissues inside the treatment beams. Therefore, the conventional margin increasing method is not a good solution for solving the uncertainties due to breathing motion in lung cancer treatment. Techniques to compensate or correct for organ motion should be used for lung cancer treatment. Our preliminary results suggest that dynamic beam delivery can provide both target dose conformity and normal tissue sparing with smaller treatment margins

Recently, intensity-modulated radiation therapy and modulated electron radiotherapy have gathered a growing interest for the treatment of breast and head and neck tumours. In this work, we carried out a study to combine electron and photon beams to achieve differential dose distributions for multiple target volumes simultaneously. A Monte Carlo based treatment planning system was investigated, which consists of a set of software tools to perform accurate dose calculation, treatment optimization, leaf sequencing and plan analysis. We compared breast treatment plans generated using this home-grown optimization and dose calculation software for different treatment techniques. Five different planning techniques have been developed for this study based on a standard photon beam whole breast treatment and an electron beam tumour bed cone down. Technique 1 includes two 6 MV tangential wedged photon beams followed by an anterior boost electron field. Technique 2 includes two 6 MV tangential intensity-modulated photon beams and the same boost electron field. Technique 3 optimizes two intensity-modulated photon beams based on a boost electron field. Technique 4 optimizes two intensity-modulated photon beams and the weight of the boost electron field. Technique 5 combines two intensity-modulated photon beams with an intensity-modulated electron field. Our results show that technique 2 can reduce hot spots both in the breast and the tumour bed compared to technique 1 (dose inhomogeneity is reduced from 34% to 28% for the target). Techniques 3, 4 and 5 can deliver a more homogeneous dose distribution to the target (with dose inhomogeneities for the target of 22%, 20% and 9%, respectively). In many cases techniques 3, 4 and 5 can reduce the dose to the lung and heart. It is concluded that combined photon and electron beam therapy may be advantageous for treating breast cancer compared to conventional treatment techniques using tangential wedged photon beams followed by a boost electron field.

Purpose/Objective: To date, we have treated approximately 800 patients with IMRT at this institution using the Siemens Primus or Primart linear accelerators and the SMLC delivery technique. Prior to the initial patient treatment, each plan is evaluated and a series of QA tests are performed including absolute dose verification. This verification is a comparison between the dose calculated in phantom by the treatment planning system and the dose measured in phantom during irradiation. Our clinical acceptance criteria dictate that the values should agree to within ± 3% for treatment to begin. Values exceeding this limit trigger additional analysis, measurement and double-checking prior to patient treatment. An example is given where the mean deviation value was −6.5% with a maximum deviation measured of −8.8%. The focus of this work is to analyze the accuracy of our dosimetry QA procedure and address discrepancies that are significantly greater than ± 3%.Materials/Methods: All IMRT treatment plans are generated using the Corvus treatment planning system. Once a satisfactory plan is obtained a hybrid plan is generated. This plan utilizes the leaf sequences and associated MU designed for the patient plan and calculates dose on a virtual phantom. The physical phantom is then irradiated under the same conditions using the same leaf sequences. An absolute dose comparison is made by comparing dose in the physical phantom as measured using a 0.125cc PTW ionization chamber with dose predicted at the same point(s) in the virtual phantom by the treatment planning system. Agreement to within ± 3% indicates that the plan is appropriately deliverable as far as absolute dose is concerned. Values exceeding this limit prompt additional investigation. This investigation includes reassessing the treatment plan and hybrid plan for errors in data extraction, investigation of input data into the R&V system for accuracy and completeness, and investigation of deliverability with respect to beam placement. The phantom irradiation is then repeated to eliminate errors due to improper setup. Discrepancies outside a ± 4% window are verified using Monte Carlo simulations of both the dose in the virtual phantom and dose in the original patient CT data set. With agreement between the original Corvus calculated relative dose distributions and the distributions generated in film and Monte Carlo simulations, the MU can be scaled appropriately to bring the absolute dose into our acceptable range. When agreement is not present the individual patient plans are regenerated with changes in input parameters and/or beam directions.Results: Upon evaluation of our absolute dose data it was found that the agreement between measured dose and dose predicted by the planning system in phantom was within ± 1% for 37.9% of the cases, ± 2% for 68.6% and ± 3% for 93.6% of the cases. The discrepancies were more than 3% and more than 4% in only 6.4% and less than 1% of all cases, respectively. Monte Carlo calculation of absolute dose compares to within 2% of dose in phantom or the patient CT data set in the non-gradient portion of the dose distribution associated with the target as tested for 20 prostate cases.Conclusions: There appear to be some intensity maps that are not deliverable to within our clinical acceptance criteria even though the planning system generated a leaf sequence and associated MU set. Verification of absolute dose as well as the relative dose distribution is essential especially in these cases. While MU scaling may result in acceptable delivery, verification of the resultant dose distribution is needed since the Siemens linacs allow for integer MU delivery only. Monte Carlo can serve as an independent dose verification tool in the QA process. Evaluation of different leaf sequencing options is made for these cases and comparisons made using both measurement and Monte Carlo simulations. Purpose/Objective: To date, we have treated approximately 800 patients with IMRT at this institution using the Siemens Primus or Primart linear accelerators and the SMLC delivery technique. Prior to the initial patient treatment, each plan is evaluated and a series of QA tests are performed including absolute dose verification. This verification is a comparison between the dose calculated in phantom by the treatment planning system and the dose measured in phantom during irradiation. Our clinical acceptance criteria dictate that the values should agree to within ± 3% for treatment to begin. Values exceeding this limit trigger additional analysis, measurement and double-checking prior to patient treatment. An example is given where the mean deviation value was −6.5% with a maximum deviation measured of −8.8%. The focus of this work is to analyze the accuracy of our dosimetry QA procedure and address discrepancies that are significantly greater than ± 3%. Materials/Methods: All IMRT treatment plans are generated using the Corvus treatment planning system. Once a satisfactory plan is obtained a hybrid plan is generated. This plan utilizes the leaf sequences and associated MU designed for the patient plan and calculates dose on a virtual phantom. The physical phantom is then irradiated under the same conditions using the same leaf sequences. An absolute dose comparison is made by comparing dose in the physical phantom as measured using a 0.125cc PTW ionization chamber with dose predicted at the same point(s) in the virtual phantom by the treatment planning system. Agreement to within ± 3% indicates that the plan is appropriately deliverable as far as absolute dose is concerned. Values exceeding this limit prompt additional investigation. This investigation includes reassessing the treatment plan and hybrid plan for errors in data extraction, investigation of input data into the R&V system for accuracy and completeness, and investigation of deliverability with respect to beam placement. The phantom irradiation is then repeated to eliminate errors due to improper setup. Discrepancies outside a ± 4% window are verified using Monte Carlo simulations of both the dose in the virtual phantom and dose in the original patient CT data set. With agreement between the original Corvus calculated relative dose distributions and the distributions generated in film and Monte Carlo simulations, the MU can be scaled appropriately to bring the absolute dose into our acceptable range. When agreement is not present the individual patient plans are regenerated with changes in input parameters and/or beam directions. Results: Upon evaluation of our absolute dose data it was found that the agreement between measured dose and dose predicted by the planning system in phantom was within ± 1% for 37.9% of the cases, ± 2% for 68.6% and ± 3% for 93.6% of the cases. The discrepancies were more than 3% and more than 4% in only 6.4% and less than 1% of all cases, respectively. Monte Carlo calculation of absolute dose compares to within 2% of dose in phantom or the patient CT data set in the non-gradient portion of the dose distribution associated with the target as tested for 20 prostate cases. Conclusions: There appear to be some intensity maps that are not deliverable to within our clinical acceptance criteria even though the planning system generated a leaf sequence and associated MU set. Verification of absolute dose as well as the relative dose distribution is essential especially in these cases. While MU scaling may result in acceptable delivery, verification of the resultant dose distribution is needed since the Siemens linacs allow for integer MU delivery only. Monte Carlo can serve as an independent dose verification tool in the QA process. Evaluation of different leaf sequencing options is made for these cases and comparisons made using both measurement and Monte Carlo simulations.