Evaluation of rotational errors in treatment setup of stereotactic body radiation therapy of liver cancer.
Academic Article
Overview
abstract
PURPOSE: To evaluate the dosimetric impact of rotational setup errors in stereotactic body radiotherapy (SBRT) treatment of liver tumors and to investigate whether translational shifts can compensate for rotation. METHODS AND MATERIALS: The positioning accuracy in 20 patients with liver malignancies treated with SBRT was reevaluated offline by matching the patients' cone-beam computed tomography (CT) scans (n=75) to the planning CT scans and adjusting the 3 rotational angles (pitch, roll, and yaw). Systematic and random setup errors were calculated. The dosimetric changes caused by rotational setup errors were quantified for both simulated and observed patient rotations. Dose distributions recalculated on the rotated CT scans were compared with the original planned doses. Translational corrections were simulated based on manual translational registration of the rotated images to the original CT scans. The correction efficacy was evaluated by comparing the recalculated plans with the original plans. RESULTS: The systematic rotational setup errors were -0.06° ± 0.68°, -0.29° ± 0.62°, and -0.24° ± 0.61°; the random setup errors were 0.80°, 1.05°, and 0.61° for pitch, roll, and yaw, respectively. Analysis of CBCT images showed that 56.0%, 14.7%, and 1.3% of treated fractions had rotational errors of >1°, >2°, and >3°, respectively, in any one of the rotational axes. Rotational simulations demonstrated that the reduction of gross tumor volume (GTV) coverage was <2% when rotation was <3°. Recalculated plans using actual patient roll motions showed similar reduction (<2%) in GTV coverage. Translational corrections improved the GTV coverage to within 3% of the original values. For organs at risk (OAR), the dosimetric impact varied case by case. CONCLUSION: Actual rotational setup errors in SBRT for liver tumors are relatively small in magnitude and are unlikely to affect GTV coverage significantly. Translational corrections can be optimized to compensate for rotational setup errors. However, caution regarding possible dose increases to OAR needs to be exercised.
