Program Information
Impact of Intrafractional Prostate Motion On the Accuracy and Efficiency of Prostate SBRT Delivery: A Retrospective Analysis of Prostate Tracking Log Files
H Xiang1,2*, M Qureshi2 , R De Armas3 , A Hirsch1,2 , M Katz2 , B Nicholas2 , S Keohan2 , H Lu1 , J Efstathiou1 , A Zietman1 , J Willins1, 2 , L Kachnic1,2 , (1) Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114, (2) Boston Medical Center and Boston University School of Medicine, Boston, MA 02118, (3) Massachusetts Institute of Technology, Cambridge, MA 02139
Presentations
SU-E-J-150 Sunday 3:00PM - 6:00PM Room: Exhibit HallPurpose: To measure intrafractional prostate motion by time-based stereotactic x-ray imaging and investigate the impact on the accuracy and efficiency of prostate SBRT delivery.
Methods: Prostate tracking log files with 1,892 x-ray image registrations from 18 SBRT fractions for 6 patients were retrospectively analyzed. Patient setup and beam delivery sessions were reviewed to identify extended periods of large prostate motion that caused delays in setup or interruptions in beam delivery. The 6D prostate motions were compared to the clinically used PTV margin of 3-5 mm (3 mm posterior, 5 mm all other directions), a hypothetical PTV margin of 2-3 mm (2 mm posterior, 3 mm all other directions), and the rotation correction limits (roll ±2°, pitch ±5° and yaw ±3°) of CyberKnife to quantify beam delivery accuracy.
Results: Significant incidents of treatment start delay and beam delivery interruption were observed, mostly related to large pitch rotations of ≥±5°. Optimal setup time of 5-15 minutes was recorded in 61% of the fractions, and optimal beam delivery time of 30-40 minutes in 67% of the fractions. At a default imaging interval of 15 seconds, the percentage of prostate motion beyond PTV margin of 3-5 mm varied among patients, with a mean at 12.8% (range 0.0%-31.1%); and the percentage beyond PTV margin of 2-3 mm was at a mean of 36.0% (range 3.3%-83.1%). These timely detected offsets were all corrected real-time by the robotic manipulator or by operator intervention at the time of treatment interruptions.
Conclusion: The durations of patient setup and beam delivery were directly affected by the occurrence of large prostate motion. Frequent imaging of down to 15 second interval is necessary for certain patients. Techniques for reducing prostate motion, such as using endorectal balloon, can be considered to assure consistently higher accuracy and efficiency of prostate SBRT delivery.
Contact Email: