ESTRO 36 Abstract Book

S90 ESTRO 36 2017 _______________________________________________________________________________________________

Material and Methods Anthropomorphic plastic phantoms were made with each having a simulated tumor bed that can be visualized using both ultrasound and CT. In the control, arm, the tumor is identified using ultrasound and inserted under ultrasound guidance. A tissue-locking needle and US probe are equipped with a real-time EM tracker. Under US guidance, the localization needle is placed within the tumor bed, which provides a rigid reference. The cavity is then contoured on US, creating a model in a virtual view. An EM tracked needle guide is pointed at the tumor bed and the catheter needle is inserted through the guide into the tissue. Additional parallel catheters are planned on the virtual view based on the first insertion and implanted in the target. The guidance software is built on the 3D Slicer (www.slicer.org) and SlicerIGT (www.slicerigt.org) open source platforms. In these experiments, a total of 10-15 catheters were inserted in each of the six phantoms. The goal was to place each catheter within the tumor bed. Three phantoms had catheter needles inserted with ultrasound only, while the other three had catheters inserted with combined EM tracking and US guidance. All six insertions were conducted by the same operator and the placement of the catheters was determined with CT. Results Under US guidance only in the three phantoms, 17 out of 26 catheters passed through the tumor bed. The average mean spacing was 0.86 cm +/- 0.33 cm. Under combined EM tracking and US guidance, 35 out of 40 catheters passed through the tumor bed. The average mean spacing was 1.05 +/- 0.19 cm. Conclusion These phantom experiments verify that EM tracking can be used to target catheter needles to the tumor bed. Additional research is currently being performed to translate this technique to patient trials. OC-0179 Dosimetric impact of errors in HDR-iBT of the breast using a catheter tracking method M. Kellermeier 1 , B. Hofmann 1 , V. Strnad 1 , C. Bert 1 1 Universitätsklinikum Erlangen- Friedrich-Alexander- Universität Erlangen-Nürnberg, Department of Radiation Oncology, Erlangen, Germany Purpose or Objective Electromagnetic tracking (EMT) was used to measure the implant geometry in fractioned HDR interstitial brachytherapy (iBT) of the breast. Based on the tracking data the dosimetric impact of common clinical errors, e.g. as reported in the United States by the Nuclear Regulatory Commission, were assessed using treatment planning For tracking of implant catheters, 28 patients were accrued within an institutional review board-approved study. The geometry of interstitial single-leader catheters (median: 18 pcs) was tracked on the HDR treatment table directly after each of the treatment fraction (up to nine during five days). Tracking has been performed by manual insertion of a small EMT sensor into each of the catheters. The breathing motion was compensated by computing the center of mass from three additional EMT sensors on the breast. Taking the tracking-based catheter data, different errors (swaps and shifts of catheters, changing the tracking direction of catheters, i.e. tip-end swap) were simulated. For dose calculation, the dwell positions (DPs) were determined along the catheter traces and the dwell times were taken from the approved treatment plan. Common contour-independent QC like the dose non-uniformity ratio (DNR) were analyzed. For investigation of contour- dependent QC, like the coverage index (CI) of the PTV, the corresponding EMT-derived DPs were registered to the CT- quality criteria (QC). Material and Methods

derived DPs from treatment planning. In addition, the maximal dose to the skin was determined. QC of EMT- based dose distributions were normalized to the corresponding values from treatment planning, so the relative changes are reported. Results Without simulated errors, the maxim um dosimetric deviations to the treatment plan were found on the 2 nd treatment day in median -6.2% for the DNR and -4.3% for the CI of the PTV. For error simulation, 15,107 pairwise swaps of catheters were analyzed. The reconstructed dose distributions resulted in DNR changes form -22.7% to 38.9% (mean: 0.6%, SD: 5.5%) and CI changes from -63.5% to 11.4% (mean: -7.4%, SD: 7.8%). For each shift of single catheters, 2,264 combinations of dose distributions were calculated. Relative dosimetric changes for DNR ranged from -4.1% to 3.5%, from -6.8% to 6.2% and from -8.8% to 8.1% for catheter shifts of 5, 10 and 15 mm, respectively at mean values between 0.0% and -0.3%. The CI for the PTV showed a mean change of -0.3%, -1.3% and -2.8%, respectively. Increased catheter shifts correlated with a higher local dose at the skin (see figures). In addition, each 3D dose distribution was analyzed to identify individual local dose deviations.

Conclusion Statistically, the maximum dose deviation was found on the 2 nd day, what might impact boost treatments with two fractions only. Based on EMT-determined dose calculations adaptive treatment protocols and tests for possible treatment delivery errors should be implemented. Further work is required for the registration method. OC-0180 Prospective study of APBI With Multicatheter Brachytherapy in Local Relapses of Breast Cancer E. Villafranca Iturre 1 , L. Rubi 2 , M. Barrado 1 , A. Sola 1 , P. Navarrete 1 , A. Manterola 1 , M. Dominguez 1 , G. Asin 1 , M. Campo 1 , I. Visus 1 , G. Martinez 1 1 Hospital of Navarra, Radiation Oncology, Pamplona, Spain 2 Hospital Juan Ramon Jimenez, Radiation Oncology, Huelva, Spain

Purpose or Objective