Combined with modern radiation treatment techniques, such as 3DCRT or IMRT, reduction of PTV margins in prostate cancer external beam radiotherapy [EBRT] reduces dose to nearby normal tissues and could potentially allow an increased radiation dose to the prostate. However, tight margins could lead to geographical misses due to daily prostate displacement. The purpose of this study was to compare 2 distinctly different image-guided verification methods for prostate alignment during daily radiotherapy. A prospective clinical trial was conducted whereas prostate displacements were measured in 40 patients undergoing daily EBRT for prostate cancer. Comparison was made between a two-dimensional trans-abdominal ultrasound-based system (BAT System, NOMOS, Pennsylvania) [BAT] and a three-dimensional trans-abdominal ultrasound-based system (RESTITU System, Resonant Medical, Montreal) [RES]. A total of 217 BAT and 217 RES displacements measurements were collected within a minute of each other with all patients in the supine position. For prostate alignment with the BAT, two nearly orthogonal 2D ultrasound images obtained immediately prior to treatment were compared to the position of the CT-designed treatment planning volumes (i.e., cross-modality calculation method, comparing ultrasound to CT data). The RES system compared 3D ultrasound images acquired immediately prior to the planning CT scan to those acquired immediately prior to the treatment (i.e., intra-modality calculation method, comparing ultrasound to ultrasound data). Since the RES system resides in the CT-simulation room as well, prostate displacements were also compared to those measured by CT scan, which is the gold standard since it is used for treatment planning. This comparison was done in 10 patients, who were rescanned 3 additional times, at bi-weekly intervals, during their treatment course. The prostate displacement between the planning CT and the subsequent CT scans was calculated with the AcQSim CT-Sim Software (Phillips Medical Systems, Massachusetts) and compared to the 3D-US data obtained just prior each of the CT scans. Analysis performed on the difference in the paired 217 BAT and RES prostate displacements shows a mean difference of 0.9 (95% C.I. 0.5, 1.3) mm in the lateral [LAT] direction (p < 0.0001), of 0.1 (95% C.I. -0.5, 0.7) mm in the anteroposterior [A/P] direction (p = 0.604), and 6.0 (95% C.I. 5.4, 6.6) mm in the superoinferior [S/I] direction (p < 0.0001). These trends in the difference results are similar to those determined by Langen et. al.1, who compared the BAT system and implanted radio-opaque markers. Analysis of the thirty CT-scan and RES measured prostate displacements shows mean values of 0.2 (95% C.I. -0.4, 0.8) mm in the LAT direction (p = 0.314), - 0.3 (95% C.I. -0.9, 0.3) mm in the A/P direction (p = 0.405), and 0.1 (95% C.I. -0.5, 0.7) mm in the S/I directions (p = 0.727). Our study indicates that a significant systematic difference exists between cross-modality (BAT) and intra-modality (RES) methods when assessing prostate alignment during daily EBRT. Since RES displacements are consistent with the CT displacements, a more accurate prostate alignment is obtained when the RES method is used, compared to the BAT method.
Read full abstract