S183 ESTRO 33, 2014 EP-1548 Dosimetric comparison of AAA and Acuros XB algorithms for lung stereotactic radiotherapy O. Senkesen 1 , E. Goksel 1 , H. Kucucuk 1 , M. Yilmaz 1 , E. Tezcanli 2 , M. Garipagaoglu 2 , M. Sengoz 2 1 Acibadem Kozyatagi Hospital, Radiation Oncology Department, Kozyatagi - Istanbul, Turkey 2 Acibadem University, Radiation Oncology Department, Kozyatagi - Istanbul, Turkey Purpose/Objective: The recent developments led to improvement of dose calculation algorithms. Heterogeneity correction is especially important for lung cancer patients. The goal of this study was to compare the dose volume histograms (DVH) of volumetric modulated arc therapy (VMAT) plans for stereotactic lung cancer treatment plans calculated by Acuros XB (V11) and anisotropic analytical algorithms (AAA,V11). Materials and Methods: Ten lung cancer patients underwent stereotactic treatment for peripheral lung tumors were included in this study. Target volume contours were delineated on MIP (maximum intensity projection) images and ensured on the 4D CT scan, then they were merged to free- breathing CT scan images where the organ at risk (OAR) volumes were contoured. Mean target volume was 50 cc (11 - 100.9). VMAT treatment plans were calculated as 220°partial arcs. Comparison was done for two different constants, coverage and MU values. Initially, all treatment plans were planned with VMAT and calculated by both Acuros XB and AAA algorithms. All plans were normalized for 95% of the target volume covered by 100% of prescription dose. Subsequently, a new plan was generated for Acuros Algorithm keeping the fixed MU values obtained by AAA calculations.Conformity Index (CI), Gradient Index (GI) and Heterogeneity Index (HI) were measured for both algorithms. Mean GTV, mean PTV, lungV5, lungV10 and mean lung doses (MLD) were also compared. Paired-samples T test was used for statistics. Results: The comparison of MU values for fixed coverage plans showed a mean increase of 5% for Acuros algorithm. GI (p=0,018) and HI (p=0,007) were found to be significantly better for AAA when compared to Acuros algorithm. No significant difference was found between CI, Lung V5, V10, MLD and mean doses for GTV and PTV between AAA and Acuros algorithms. The plans generated with fixed MU values did not show any significant differences for analyzed parameters. Conclusions: Dosimetric differences were minor between calculations generated with Acuros and AAA algorithms for VMAT of stereotactic lung treatments. Although the AAA appeared to show a more homogenous dose distribution, Acuros reflected a more realistic calculation. The dosimetric accuracy of both algorithms and their relation to treatments outcomes should be further evaluated. EP-1549 Evaluation of a commercially available software for 3D dose and delivery verification of radiation treatment plans P. Sibolt 1 , U. Bjelkengren 1 , S.K. Buhl 1 , D. Sjöström 1 , C.F. Behrens 1 1 Herlev Hospital University of Copenhagen, Department of Oncology (R), Herlev, Denmark Purpose/Objective: Traditionally, patient specific radiation treatment plan verification is performed by independent monitor unit calculation or dose calculation in one point. The increasing complexity of treatment plans, however, indicates a need for more sophisticated ways of performing treatment plan verification. The purpose of this study was to explore a commercially available software (Mobius3D) utilizing an independent collapsed cone algorithm together with delivery log files (dynalogs) for 3D dose verification. Materials and Methods: The Mobius3D (M3D) software was commissioned, including e.g. spot check calculations for over 150 different collimator, MLC, SSD, depth, off-axis distance and energy combinations in a water phantom created in the Varian Eclipse treatment planning system (TPS). Moreover, a total of 108 clinical treatment plans (accepted for treatment by patient specific measurements) were recalculated in M3D as well as in a single point utilizing a clinically implemented software (IMSure) employing simple Clarkson integrals. Calculations included conventional static, IMRT as well as RapidArc plans. Deviations between mean target doses extracted from the TPS and M3D were calculated. These deviations were compared to corresponding deviations between point doses extracted from the TPS and IMSure. Evaluation of M3D was, furthermore, carried out by investigating the usefulness of the extra information provided by the 3D calculation carried out in the patient volume by M3D as well as the beta version of MobiusFX (FX) developed for analysis of dynalogs. Results: Commissioning and spot check calculations demonstrated dose discrepancies up to 5 % from the TPS in build up regions and in the penumbra as well as for small and large fields. For clinically relevant field sizes and depths deviations were within ±3 %. For both dynamic and static plans it was clear that point dose calculations result in a larger spread in the deviation from the TPS (Figure 1), implying that the usage of a system based on 3D dose calculations could decrease the risk of receiving false negatives. M3D provides the possibility to perform 3D gamma analysis and DVH verification, while FX gives valuable information about deliverability, including e.g. MLC performance. Conclusions: When implementing and utilizing independent dose calculation software in the clinic, one needs to be aware of any limitations as well as eventual fundamental differences from the TPS calculation algorithm. Utilizing M3D indicate decreased risk of detecting false negatives in comparison to point dose calculations. However, further investigation of the sensitivity and specificity is needed. M3D has the potential to complement standard plan checks, in the possibility to perform DVH verification, and FX has the potential to replace patient specific measurements, in the ability to utilize dynalogs for machine performance and delivery tests. EP-1550 Accuracy of contralateral breast doses using different treatment planning algorithms J. Heikkilä 1 , J. Seppälä 1 , T. Virén 1 , T. Lahtinen 1 1 Kuopio University Hospital, Cancer Center, Kuopio, Finland Purpose/Objective: New dynamic radiotherapy techniques have been developed and implemented in radiotherapy treatment planning systems. Simultaneously, new dose calculation algorithms have been developed into the treatment planning software. Since highly modulated treatment techniques increase the low-dose volume of healthy tissues, the accuracy of out-of-field dose calculations is an increasing concern. In this study the accuracy of calculated contralateral breast doses for photon beam breast radiotherapy were investigated using MOSFET measurements and Elekta's Monaco ® and Oncentra ® treatment planning software providing Pencil Beam (PB), Collapsed Cone (CC) and Monte Carlo (MC) algorithms. Materials and Methods: Tissue-equivalent humanoid phantom and MOSFET dosimeters were used to measure doses in the contralateral breast from left-sided breast irradiation using two different plans to PTV dose of 50 Gy. Four MOSFET dosimeters (TN 1002RD-H, Best medical Canada, Ottawa, Canada) were placed into the low dose region inside the right breast of the phantom (see Figure 1). Doses were calculated with PB and CC algorithms (Oncentra ® ) and MC algorithm (Monaco ® ). The Oncentra ® treatment plans consisted of two tangential fields and the Monaco plan one partial VMAT field from 300 to 180 degrees. The Oncentra ® plans were calculated using identical monitor units.