SU‐E‐J‐27: Residual Interfraction Errors for Prostate SBRT with 6 Dimensional Couch Shifts Using Stereoscopic X‐Ray and Cone Beam CT

Abstract

Purpose: To evaluate residual interfraction alignment errors following prostate SBRT patient setup with stereoscopic kV‐X‐ray (Exactrac), kV‐CBCT, and a 6‐D robotic treatment couch (RTC). Methods: Data was reviewed from seven prostate cancer patients undergoing five‐fraction SBRT with pre‐and daily‐treatment endorectal balloon placement. Setup protocol, using stereoscopic X‐ray images, consisted of an initial RTC displacement based on bony registration (3 shifts, 3 rotations) followed by a RTC displacement based on implanted gold seeds (3 shifts only). Final image verification of the prostate position was performed with kV‐CBCT, upon which additional shifts prior to treatment delivery were permitted. The prostate (CTV), rectum wall and bladder were delineated on each daily CBCT and planning CT by the same physician. Shifts recorded at the time of treatment (X‐ray and X‐ray plus CBCT) were applied. Planning CT and CBCT contours were then superimposed and evaluated with dice similarity coefficient (DSC). Daily CTVs were also compared with the PTV (simulation CTV + 3mm). Paired student t test was used for statistical comparison. ResultsThe average of 96 stereoscopic X‐ray shifts were 0.96±4.08cm, 0.72±3.42cm and 0.58±3.64cm in the lateral, vertical and longitudinal directions, and 0.78±2.42°, 0.55±1.60° and 0.21±1.09° for the pitch, roll and yaw angles, respectively. CBCT‐guided shifts were performed in 25.7% of the fractions, with an average of 0.01±0.03cm, 0.23±0.21cm, and 0.14±0.26cm in the lateral, vertical, and longitudinal directions, respectively. With X‐ray only and X‐ray plus CBCT shifts, 95.1±3.7% and 96.2±3.2% (p=0.01) of daily CTVs were covered by PTV, respectively. The DSC were 77.1±5.8% and 79.2±4.6% for CTV (p<0.01), 40.1±12.8% and 42.1±12.7% for rectal wall (p=0.03), and 65.2±17.5% and 64.5±18.0% for bladder (p=0.34), respectively. Conclusion Interfraction error following Exactrac‐guided RTC shifts was <3mm, resulting in protocol‐appropriate CTV coverage at treatment based on a 3mm PTV margin. Further CBCT‐based shifts improve prostate alignment, but not significantly.