Abstract
Purpose or Objective Stereotactic Body Radiotherapy (SBRT) has been shown to be a promising therapy for non-metastatic unresectable pancreatic tumors, which account for 40% of all pancreatic carcinomas. However, intrafraction motion, caused by respiratory motion and organ drift, is the main concern for efficient dose delivery in ungated upper abdominal
... read more
RT. The aim of this study is to characterize respiration and drift motion patterns of pancreatic tumors, across a clinical cohort of 13 patients that underwent MRgRT while wearing an abdominal corset. Materials and Methods We included 13 patients that underwent online adaptive MR-guided SBRT for malignancies in the pancreatic region (5x8Gy) on the MR-Linac. An abdominal corset was fitted, which has been shown to substantially reduce respiratory motion [1]. Interleaved coronal and sagittal Cine MRs of the tumor region were made at 2 Hz during the entire beam-on time of each fraction. We used deformable image registration to obtain GTV motion profiles, with position over time in the cranio-caudal (CCC, CCs: in both the coronal and sagittal plane), anterior-posterior (AP), and left-right (LR) direction. To isolate the motion caused by respiration and baseline drift, the motion profiles were respectively high-pass and low-pass filtered. From these decoupled motion profiles, the respiratory amplitude and maximum and minimum baseline drifts were extracted. Results The respiratory amplitudes for all patients and fractions, in each direction, are plotted in fig. 1. The minimum and maximum baseline drifts are plotted in fig. 2. The mean (SD) respiratory amplitudes were 4.2 (1.9) mm CCC, 3.9 (1.7) mm CCS, 2.3 (1.1) mm AP and 1.4 (0.6) mm LR, with low variability within patients. The mean (SD) maximum baseline drifts were 1.2 (1.1) mm CCC, 1.2 (1.0) mm CCS, 0.5 (0.4) mm AP and 0.5 (0.3) mm LR. The mean (SD) minimum baseline drifts were -0.7 (0.5) mm CCC, -0.7 (0.5) mm CCS, -0.6 (0.5) mm AP and -0.5 (0.4) mm LR. The wearing of the corset was associated with some inconveniences but none of the patients reported pain or other discomfort that urged loosening the corset. Conclusion In our cohort, respiratory amplitudes were small and stable within patients. Drift motions were small as well, with no large GTV displacements observed during the treatments. We conclude that ungated, high-precision MR-guided SBRT for pancreatic tumors is feasible for patients that wear an abdominal corset, potentially allowing dose escalation strategies. Motion mitigation using corsets could well be a simple and attractive alternative for more complex gating and tracking approaches when treating abdominal lesions.
show less