Measurements and characterization of intra-fractional prostate motion during SBRT treatment of prostate cancer on cyberknife for optimizing image-guided dose delivery

Abstract

Prostate cancer is currently the most frequently diagnosed cancer among men in the United States. Due to the low a/P ratio (on the order of 1.5-2.0) compared to normal tissue, the therapeutic benefit of hypo-fractionation allows for higher cancer cell kill with a lower probability of late grade III-IV toxicity to surrounding healthy organ tissues. Such hypo-fractionated radiation treatment poses a more stringent requirement on the accuracy of target localization and dose delivery. CyberKnife is a robotic radiosurgery system that has been shown to be accurate for delivering stereotactic body radiation therapy for extra-cranial tumor targets. However, in the case of prostate SBRT, intrafractional prostate motion presents a unique challenge for the accurate delivery of treatment. The purpose of this work is to characterize and quantify the time dependence of intra-fractional prostate motion. Results from this study will be used for optimization of the treatment planning and improvement of SBRT delivery accuracy. Retrospective imaging guidance data were studied for prostate cancer patients treated with SBRT on the CyberKnife. At the time of treatment, prostate is localized and tracked by taking stereoscopic X-ray images of implanted fiducial gold markers. For smaller movements of translations and rotations, the beams can be corrected before beam-on dose delivery. For larger magnitudes of prostate movements that exceed the limits of beam corrections, treatment was automatically halted to allow re-establishing the target alignment before resuming the delivery. Treatments were planned and optimized to have each beam with MUs not exceeding 150~160 MU, thus to allow timely monitoring the prostate position between consecutive beams. All beam corrections were recorded in the system log files, and downloaded after all planned fractions of treatment were completed. This study analyzed both rotational and translational prostate movements at an interval of 10-15 seconds compared to previously published translation data only for an interval of ~40 seconds. For the case being analyzed, it was observed that under normal conditions, the translations of intra-fraction prostate motions are generally under 3-5 mm and the rotations are generally under 2°-5°, both are within the CyberKnife beam correction range. However, it was also well observed that the gas pocket formation and passing targeted region of interest are unpredictable during treatment delivery. They can cause much larger misalignments that are either beyond the PTV margins or beyond the beam angle correction range, leading to suspension of the treatment for extended time, typically 30-40 min per fraction. The study of time-dependence of intra-fraction prostate motion revealed significant uncertainties during treatment delivery. Timely imaging at shorter time intervals is effective for detecting such uncertainties, and helping to improve the accuracy of dose delivery. In addition, the results from this study also points to the need of finding better ways of managing intra-fraction prostate motion, such as using endorectal balloon to stabilize prostate position during treatment and release gas pockets from upstream before their reaching to the target region of interest. More patient data cases are needed to further support these important preliminary findings.