3D Surface Imaging in Reducing Setup Error for Prone Whole Breast Radiotherapy

Abstract

For patients undergoing whole breast radiotherapy, treatment in the prone position allows for reduced dose to critical organs such as the heart and lungs. However, this treatment position comes with greater challenges in reproducibility and setup error given the unstable position. Our objective was to investigate the benefit of using 3D surface imaging to reduce daily setup errors for these patients. We performed a retrospective review of consecutive breast patients receiving prone breast radiotherapy at our institution. All patients were positioned initially using setup tattoos and twice a week, the positioning was adjusted using the aid of a motion tracking system. Then, a daily kilovoltage (kV) cone beam computed tomography (CBCT) image was acquired and positioning was adjusted to setup the breast and chest wall. Shifts in each translational direction were recorded and the three-dimensional vector displacement was calculated. For each patient the average displacements on days where a motion tracking system was used were compared to days whereas motion tracking system was not used. Patients were classified into significant benefit (more than 4 mm improvement), small benefit (0-4 mm improvement) or increased error (increased shifts on days a motion tracking system was used). Patient demographics were collected including stage, BMI, weight, heigh, age, ethnicity. Comparisons were made using t-tests. A total of 40 patients with stage T0-2N0 breast cancer were included. The median age was 64 and 60% of patients had left sided breast cancer. A total of 665 daily CBCT scans were analyzed, and the median daily vector displacement was 7.2 mm (1-21 mm). The median displacement on treatments where a motion tracking system was used (41% of treatments) was 6.1 mm versus 8.1 mm when not used (p<.0001). The use of a motion tracking system significantly reduced the occurrence of shifts more than 1 cm (12% vs. 33%, p < .0001) and 5 mm (65% vs. 83% p < .0001). For individual patients, 28% showed a significant improvement, 55% showed small improvement, 17% showed increased error. The median BMI was 27, weight was 71 kg, and neither was associated with an improvement in the use of Align RT (p > .05). However, patients in the small improvement group had a higher BMI than patients in the other two groups 29 vs. 25 (p = .01). A similar association was seen for weight (p < .05). No association was found for the benefit of a motion tracking system for height, age, stage or ethnicity. The use of 3D surface imaging for breast cancer patients receiving prone whole breast radiotherapy on average significantly reduced setup errors. For patients with higher BMI there was a consistent small reduction in setup error when compared to using setup marks alone. For patients with lower BMI, caution should be exercised as there was more variation with some patients demonstrating a large benefit and other patients having an increased setup error with the use of surface imaging. Further research is needed to investigate on the optimal use of this technology.