ESTRO 36 Abstract Book

S834 ESTRO 36 _______________________________________________________________________________________________

46%, and 74% reduction in number of beams by 74% with the MLC-based plans (all p < 0.001). Conclusion Compared to the Iris, the InCise MLC produced comparable target coverage but was significantly better in dosimety with significant improved delivery efficiency. EP-1550 Investigating the advantages of CyberKnife M6 MLC over Iris collimator for Liver SBRT plans R. Doro 1 , L. Masi 1 , V. Di Cataldo 2 , S. Cipressi 2 , I. Bonucci 2 , M. Loi 3 , L. Livi 4 1 IFCA, Medical Physics, Firenze, Italy 2 IFCA, Radiation Oncology, Firenze, Italy 3 University of Florence, Department of Clinical and Experimental Biomedical Sciences "Mario Serio", Firenze, Italy 4 Azienda Ospedaliera Universitaria Careggi, Radiotherapy Unit, Firenze, Italy Purpose or Objective The purpose of this study is to evaluate the performance of the CyberKnife M6 systems equipped with MLC for Liver SBRT plans. To this aim, MLC plans were compared to clinical plans generated using circular apertures. Material and Methods 21 clinical treatment plans for Liver SBRT created with IRIS variable aperture collimator were optimized again on Multiplan 5.3 TPS using MLC. Plans were created both for first and second treatment cases and were prescribed either in 3 or 5 fractions with prescription doses ranging from 30 Gy to 45 Gy. PTV dimensions ranged from 25.7 cm 3 to 233 cm 3 . The same OAR constraints were applied both for IRIS and MLC plans. Evaluation parameters of each plan included PTV coverage, Paddick's new CI (nCI), homogeneity index (HI), gradient index (GI) and prescription isodose. OAR (duodenum, stomach, bowel, hearth) dose sparing was analyzed using the maximum and mean doses (Dmean). Liver dose sparing was analyzed using mean dose and the volume either inside 15 Gy (3 fractions) or 21 Gy (5 fractions) isodose. The dose delivery efficiency was evaluated on the basis of planned monitor units (MUs) and the reported treatment time per fraction. The dose to the PTV was also summarized by the generalized equivalent uniform dose (gEUD), using a=-20. The mean values, standard deviation and p-values (two tailed Student's t test) were computed between the two comparison groups and statistical significance set at p< 0.05. Results The evaluation parameters for the MLC and IRIS plans are shown in table 1. MLC plans achieved equivalent PTV coverage and conformity when compared to IRIS plans and minimized the low dose extension improving significantly (p<0.001) the dose fall-off gradient with GI increasing from 2.65 (MLC) to 3.13 (IRIS). Plans created using MLC were generally prescribed to higher isodose levels (73% vs 70.5%), which resulted in significantly more homogeneous dose inside the PTV (HI=1.37 vs HI=1.42, p=0.02). This, however, did not affect significantly the PTV gEUD which was equivalent between IRIS and MLC. No significant difference was observed for OAR dose sparing between the two groups of plans, with the exception of Bowel mean dose which was significantly lower for MLC. Average treatment time was significantly (p=0.01) reduced from 34.7 min. to 29.2 min when using MLC. MLC MU mean value was lower than IRIS MU, but statistical

Conclusion Non-coplanar technique can reduce MLD, lung V20 and spinal cord dose in both FB and DIBH. While these reductions were relatively small in our patient group as a whole - and compared to reductions possible by DIBH alone - they were substantial in some patients. Therefore, the NC approach should be exploited in patients not compatible with DIBH for OAR dose reduction. EP-1549 Cyberknife Iris based versus InCise based plans for 20 cases of prostate oligonodal metastases C.L. Chaw 1 , N.J. VanAs 1 , V.S. Khoo 1 1 Royal Marsden Hospital Trust & Institute of Cancer Research, Academic Uro-oncology, London, United Kingdom Purpose or Objective To compare dosimetry and delivery efficiency between Iris collimator and InCise multileaf collimator (MLC) (Accuray Inc.Sunnyvale, CA) for patients with oligometastatic nodal disease from prostate cancer. Material and Methods Treatment plans for 20 patients were performed on Multiplan TM 5.1.3 treatment planning system utilizing MLC and Iris for 30Gy in 3 fractions. To minimize variation between cases, nodal metastases located in pelvis and abdomen with a distance of 0-10mm to organs at risk (OAR; rectum, small bowel and duodenum) were chosen. The clinical target volume (CTV) to planning target volume (PTV) margin is 3mm. Dosimetric evaluation included PTV coverage, CTV coverage, conformity index (CI), Paddick’s new CI (nCI), homogeneity index, and gradient index. Treatment delivery efficiency is measured by beam delivery time (start of first beam to end of final beam, including beam-on-time, robot motion, and intra-fraction imaging), number of monitor units and number of beams used. OAR dose sparing were analysed by D max small bowel dose constrained at D0.5cc: 25.2Gy, D5cc: 17.7Gy, D max rectal dose constrained at D0.5cc:28.2Gy and D max duodenum constrained at D0.5cc: 22.2Gy; D5cc:16.5Gy and D10cc:11.4Gy. Statistical significance was tested using Wilcoxon signed rank test. Results There were no statistically significant differences in conformity indices or target coverage, but MLC plans were more homogenous with small but significantly lower mean target dose than Iris (2% difference in PTV mean dose; 4.8% difference in CTV mean dose; all P < 0.001). Gradient index was also improved by 13% using MLC plans (P < 0.001). All OAR constraints were satisfied by both devices. The small bowel mean dose was significantly lower by 52% using MLC (p < 0.001). There was a significant reduction in delivery time by 47% (mean 19.7 mins [range:13-30 mins] vs 37.0 mins [24-56], total monitor units used by