Investigation of Linac-based Image-guided Hypofractionated Prostate Treatment

Abstract

Purpose/Objective: A hypofractionation protocol for prostate treatment was initiated in our department on December 2003. This short course high dose per fraction treatment regimen requires special attention to any potential geographical miss during treatment, such as inter- and/or intra-fraction prostate motion. The CyberKnife has been used as the radiation delivery device for this protocol. The choice of the CyberKnife is due to its incorporation of near real-time imaging and localization of the prostate based on three gold fiducial seeds placed in the prostate prior to CT. The purpose of this work is to investigate the use of IMRT and the Varian Trilogy Accelerator with on-board (kV) imaging for treatment of our hypofractionated prostate patients. Materials/Methods: Patients eligible for this protocol are those with newly diagnosed low risk prostate cancer (T1c T2a, iPSA ≤ 10 ng/ml, bGS ≤ 3+3). A total dose of 36.25Gy is delivered with 7.25Gy per fraction for 5 fractions over 5 days. An anthropomorphic pelvis phantom was used as the test patient. Three gold fiducial seeds were placed in the phantom in the region of the prostate. A 5 field IMRT plan was created for a typical prostate volume using the Varian Eclipse planning system. The on-board imager (OBI) of the Trilogy Accelerator is mounted on the treatment machine gantry via robotically controlled arms which operate along three axes of motion. A 150 kV X-ray tube is opposed to the amorphous silicon flat-panel detector and the source-detector pair are oriented orthogonal to the direction of the MV treatment beam. The IMRT plan was imaged and delivered on a 23EX Trilogy Accelerator in our department. Orthogonal image sets (one MV image and one kV image) were obtained prior to beam-on at each treatment gantry angle. This was to simulate determination and correction for inter-fraction setup errors (first gantry angle) and intra-fraction prostate motion (subsequent gantry angles). The dose distribution and DVHs for a patient previously treated on the CyberKnife was compared to an IMRT plan. Results: Based on our study, we estimate the time for patient setup and imaging of the first field is 4 minutes. This is essentially the time it takes to get the patient in treatment position and take images to correct for any inter-fraction setup error. Orthogonal MV-kV imaging at each subsequent treatment field beam angle requires about 1 minute per field. Moving the table to correct for any detected intra-fraction prostate motion at this step adds approximately 15 to 30 seconds. The total time to deliver each treatment field including gantry motion, mode up, beam-on, and leaf movement for IMRT delivery was about 1.4 minutes per field. Total treatment time from patient on the table to patient leaving the room would be approximately 16 to 22 minutes for a 5 to 7 field IMRT plan. Time between successive images for intra-fraction prostate localization is 40 to 90 seconds. These times are in contrast to the total treatment time per fraction on the CyberKnife of 40 to 45 minutes with imaging for prostate localization every 30 to 75 seconds. A 7 field Eclipse IMRT dose distribution compared favorably to that produced by the CyberKnife. After both plans were normalized to deliver 36.25Gy to 95% of the PTV, the maximum dose in the PTV was 39.2Gy for the IMRT plan and 39.6Gy for the CyberKnife plan. Maximum dose to the rectum was 2% greater in the CyberKnife plan but the dose to 20% of the rectum was 27.7Gy in the IMRT plan and 30.7Gy in the CyberKnife plan. Conclusions: Orthogonal MV and kV image sets provide a time savings when acquiring multiple sets of images during a treatment fraction. The Varian Trilogy Accelerator with on-board kV imaging is well suited for image guided prostate treatments.