Analysis of prostate bed motion using daily cone-beam computed tomography during postprostatectomy radiotherapy.

Abstract

To report on the interfraction total positioning error of the postoperative prostate bed and to quantify its components (bony misalignment [BM] and prostate bed motion [PBM]) using daily kilovoltage cone-beam computed tomography (CBCT). The role of an adaptive radiotherapy schedule (ART) was investigated. A total of 547 daily CBCT images from 15 consecutive patients who had been treated with prostate bed radiotherapy were retrospectively analyzed. The positioning error was measured by rigid co-registration of the daily CBCT with pretreatment CT planning scan. The total positioning error was quantified by co-registration of the CBCT with the CT planning scan to match the anterior rectal wall. Automatic bony pelvis co-registration was performed to separate BM and PBM. The ART was determined by the average total positioning error from the first 5 CBCT images. The systematic error for the total positioning error in the left-right, superoinferior, and anteroposterior direction was 2.69, 2.00, and 2.65 mm with a random error of 1.99, 1.49, and 2.25 mm, resulting in a planning target volume margin of 8, 6, and 8 mm, respectively. ART reduced the margin by 54%, 44%, and 40%, respectively. Systematic errors in the left-right, superoinferior, and anteroposterior direction for BM was 2.66, 1.83, and 2.60 mm and for PBM was 0.44, 0.92, and 2.50 mm with a random error of 1.88, 1.24, and 1.77 mm for BM and 0.99, 1.38, and 2.32 mm for PBM, respectively. Without treatment verifications, 6-8-mm planning target volume margins are required because of PBM and BM. The anteroposterior PBM was significant. An ART protocol can reduce these planning target volume margins.