ESTRO 2020 Abstract Book

S29 ESTRO 2020

Purpose or Objective With the introduction of Magnetic Resonance-guided Radiotherapy (MRgRT), non-invasive continuous and direct monitoring of targets and organs at risk (OAR) during treatment delivery has become available. The gating strategy adopted in low tesla MRgRT systems consists in acquiring 4 MR images/second on a sagittal plan and stopping the radiation beam each time that the target structure moves beyond a boundary region pre-defined by the user. In clinical practice, the GTV is generally considered as target structure and PTV as boundary. This approach can lead to different criticalities, especially when treatment plans with inhomogeneous doses and low conformity are delivered. Aim of this study is to investigate the feasibility of using the prescription isodose volume (PIV) as boundary to optimise treatment delivery during MRgRT. Material and Methods A total of 31 patients (8 Lung, 15 Liver, 8 Pancreas) treated with MRgRT were retrospectively selected. For lesions larger than 2 cc, the Paddick conformity index (PCI) of the dose distribution and 2 boundaries were considered : a geometric one equal to PTV and a dosimetric one equal to PIV. The areas of the 2 boundaries (A GEO and A DOS ) were measured on the sagittal MR image passing through the GTV centre of mass. The comparison between the 2 boundaries was carried in terms of difference of boundary area and spatial displacement. The variation in terms of boundary area was measured as A DOS - A GEO . The spatial displacement between the 2 boundaries was evaluated as the distance between the centres of mass of the two areas (D com ) in anteroposterior (AP) and craniocaudal (CC) directions. Results The treatment plans of lung and liver lesions showed higher conformity than those related to pancreatic lesions (Mean PCI was 0.76 ±0.07 for Lung, 0.77 ±0.07 for Liver and 0.69 ±0.06 for pancreas). Figure 1 shows the spatial displacement of the two boundaries for all the patients analysed. For treatment plans with high dose conformity (lung and liver), no significant spatial displacement was observed between the two boundaries (D com <1mm for 91% of cases). Using the dosimetric boundary in such cases leads to an increase of boundary area of +22%, speeding up the treatment. For treatments with lower dose conformity (pancreas), a limited increase of boundary area was found (+2%). However, a significant displacement between the two boundary areas was observed (mean D com = 2.3 ± 0.7 mm), mainly due to the presence of surrounding OARs, such as stomach or duodenum (fig.2). The displacement is mostly located in cranial (+1.1± 1.3 mm) and posterior direction (-1.4 ± 1.1 mm). Fig.1

acquisition to assess target volumes and organs at risk, and (iii) repeat 2DkV pair to verify translation and rotation corrections before delivering treatment. Imaging data from fractions 1-5 and then weekly [bi- weekly for head and neck treatment sites] was collated to evaluate (i) initial 2DkV imaging dose using estimated delivered dose (µGy), (ii) elapsed time between 2DkV and CBCT acquisition, and (iii) concordance of online matched values for the 2DkV and CBCT image registrations, evaluated by registration on bony anatomy and soft tissue, using ARIA OIS (v13.7, Varian Medical Systems, USA). To assess the correlation between 2DkV and CBCT image registration, data was analysed using Pearson’s Correlation Coefficient and Bland-Altman analysis.

Results 229 fractions were evaluated (per patient: range 8-19, median 10). 19 (8.3%) fractions required patient repositioning following the initial 2DkV. Using a 2-step imaging process would reduce the imaging dose by 3.4mGy on average for all patients over a whole treatment course. The use of the 3-step process required a mean additional time of 5.1 minutes (range: 3.3 to 9.9) compared to the 2- step process. Overall image results indicated that correspondence between the mean displacements from the initial 2DkV and CBCT images for all treatment sites was high, with R=0.94, 0.94 and 0.80 in the anterior-posterior, superior-inferior and right-left directions respectively. Bland-Altman analysis showed there was very little bias and narrow limits of agreement. However, discrepancies in rotation correction were observed: although there was only small bias, there were relatively wide limits of agreement (no less than +/-0.6 o ). Conclusion The results presented confirmed only a minority of fractions required patient repositioning after the initial 2DkV. Removing this from the standard 3-step IGPBT process and streamlining to a 2-step workflow (commencing with CBCT) would reduce imaging dose and treatment times, thus improving efficiency and overall service capacity. The 2-step IGPBT workflow has now been implemented at our PBT Centre and is standard verification protocol. For challenging cases (e.g. paediatric patients under GA) further investigations are required before the 3-step workflow can be modified. PD-0069 Use of prescription isodose boundary to optimise treatment delivery during MR-guided Radiotherapy C. Votta 1 , D. Cusumano 1 , L. Boldrini 1 , V. Pollutri 1 , L. Placidi 1 , G. Chiloiro 1 , A. Romano 1 , M.V. Antonelli 1 , V. Valentini 1 1 Fondazione Policlinico Universitario "A. Gemelli" IRCCS, Dipartimento di Diagnostica per Immagini- Radioterapia Oncologica ed Ematologia, Roma, Italy