ESTRO 37 Abstract book

ESTRO 37

S566

Results Small but statistically significant fluctuations in random variation of less than 0.01 mm were observed between the three measurements (p<0.001). The average random variation observed was 0.075+/-0.005 mm. The strongest, but not significant, correlation in random leaf position variation measurements was observed between the 2nd and 3rd measurement (figure 1).

multi-institutional study for achieving similar plan quality results and for further benchmarking. Material and Methods 45 prostate SBRT plans from 9 centers were included. EL was ranked as 1 for 'no experience”, 2 for '<100 SBRT prostate cases planned” and 3 for '≥100 SBRT prostate cases planned”. Each center was provided with 5 anonymized CTs with pre-delineated volumes and was asked to create SBRT plans according to pre-set dose constraints. Dose prescription was 7Gyx5fr. The DVH text file and the RP_DICOM of each plan were used to extract dosimetric parameters, modulation indexes, and dynamic parameters from IMRT/VMAT plans. Results were summarized in a Spearman-correlation matrix and deemed correlated for R>0.50; statistical significance was set at p=0.01. In a 2 nd phase, planners were asked to re-plan the 5 th patient, based on the median DVHs of all partecipants, for improving target homogeneity and further sparing OAR doses. A Kruskal-Wallis rank test was performed to compare dosimetric and complexity parameters between the two series. Results In the 1 st phase, EL correlated with dosimetric parameters (see figure 1, with the correlogram showing significant positive (blue) and negative (red) correlations). The volume of rectum receiving > 32 Gy (V32 Gy Rectum, 9.1±4.4%) showed strong correlation with EL (p<0.0001): the higher the planner had experience level, the lower the doses to OARs were. When comparing only VMAT plans, the total Modulation Index (MItotal=0.70±0.21) was strongly correlated (p<0.001) with EL: the higher the experience level was, the lower the degree of plan complexity was reached. In the re-planning phase, the Kruskal-Wallis test showed significant differences in the doses to the rectum between both phases; no significant differences were found in the dynamic parameters of the plan and in the modulation indexes (see figure 2).

Between the 2nd and 3rd measurement a small but statistical significant drift was measured, see figure 2.

The outliers observed after 55 days are coursed by a positional displacement of one leaf of less than 0.5 mm during one of the sequences. After 55 days the width of the time dependent drift distribution was found to be 0.08+/-0.05 mm Conclusion The leaf positioning accuracy of the optically controlled MLC is stable over time with time dependent drift and random variation below 0.1 mm, which is small compared to the EPID pixel size. These results support the use of the Elekta Unity MR linac for high precision stereotactic treatments E. Villaggi 1 , M. Esposito 1 , C. Marino 1 , V. Hernandez 2 , J. Saez 3 , A. Bruschi 1 , G. Borzì 1 , C. Carbonini 1 , R. Consorti 1 , D. Fedele 1 , E. Moretti 1 , B. Nardiello 1 , S. Russo 1 , M. Stasi 1 , P. Mancosu 1 1 SBRT Working Group, AIFM - Italian Association of Medical Physics, Italy, Italy 2 Hospital Universitari Sant Joan de Reus, Medical Physics, Tarragona, Spain 3 Hospital Clinic de Barcelona, Radiation Oncology Department, Barcelona, Spain Purpose or Objective Dosimetric and plan complexity quality metrics have been demonstrated to be essential for compliance assessment. Improving homogeneity between centers in multi- institutional studies can be challenging due to different Experience Level (EL). The aim of this work was to preliminary evaluate the role of a re-planning phase in a PO-1012 Can re-planning improve standardization? A multi-institutional SBRT prostate comparison