PD-0044: Respiratory-induced motion of pancreatic cancer: time trends during the course of radiotherapy

Abstract

Purpose/Objective: Irradiation of pancreatic tumors requires large margins to account for respiratory-induced intrafractional tumor motion. These margins are derived from a single pre-treatment 4DCT scan. This respiratory-induced intrafractional motion can be visualized using implanted fiducial markers. The purpose of the current study is to quantify the respiratoryinduced motion of pancreatic cancer, both on the 4DCT scan as well as on daily cone beam CT (CBCT) scans during a 5 week course of radiation to reveal possible trends during the course of treatment. Materials and Methods: Eight patients (4 male / 4 female) with borderline resectable pancreatic cancer were treated with 25 x 2Gy plus weekly gemcitabine 300 mg/m. All patients received 2 to 4 fiducial markers by endoscopic ultrasound-guided implantation. The amplitude of respiratory-induced motion was measured on the pretreatment 4DCT scan and also on all available daily CBCT scans. Phase reconstructions of the CBCT scans were made by an automated selection of the 2D projection images corresponding to the peak inhale and exhale phases. These reconstructions were matched to the (reference) planning CT. The amplitudes (differences between the 2 extreme positions) were analyzed by matching on the markers (in three directions) and on the diaphragm (superior-inferior (SI) direction only). Rotations and deformations of the tissue were not analyzed. The amplitudes were linearly fit as a function of time (i.e. treatment day). The fits were tested for significance using a Student’s t-test and statistical significance was assumed at p≤0.05. Results: The amplitude of intrafractional motion of the markers was greatest in the SI direction (mean 5.9mm, SD: 2.7mm on CBCT). Differences up to 5.1mm (range: 0.3 5.1 mm) were found between the amplitude on the 4DCT scan and the mean amplitude of all CBCT scans of a single patient. The analysis of daily respiratory-induced motion of the markers revealed significant trends in the amplitude as a function of time (table and figure). The diaphragm did not always show the same trend as the markers. For the markers neither the amplitudes nor the trends were significant in the other 2 directions (data not shown). Conclusions: The 4DCT used for treatment planning is often not representative of the respiratory-induced motion amplitude during treatment. The CBCT projection images showed significant time trends in the amplitude of respiratory-induced motion during a five week course of radiation. These time trends in the amplitudes indicate the usefulness of adaptive treatment procedures. PD-0045 Time-resolved dose reconstruction of VMAT delivery to moving targets with and without dynamic MLC tracking T. Ravkilde, P.J. Keall, C. Grau, M. Hoyer, P.R. Poulsen Aarhus University Hospital, Department of Oncology, Aarhus, Denmark University of Sydney, Sydney Medical School, Sydney, Australia