Case Comprehensive Cancer Center, Case Western Reserve University Cleveland, OH
E. Jaye1, A. Rezaei1, T. Arsenault2, T. Baig1, R. Woods2, P. Mohindra1, and K. Nieto1; 1University Hospitals Seidman Cancer Center, Cleveland, OH, 2University Hospitals, Seidman Cancer Center, Case Western Reserve University, Cleveland, OH
Purpose/Objective(s): A typical workflow for complex cervical HDR brachytherapy planning at most institutions includes delineation of Organ at Risk (OAR) contours on a CT acquired after device placement. Additionally, at our institution, an MRI is obtained for HR CTV delineation. Then the plan is finalized and the treatment is delivered. Due to limited MRI slots, the completion of the MRI can take place up to six hours after the CT. This study aims to assess the potential deviations in OAR delineation and resulting dose statistics caused by variations in the time between CT and MRI scans. Materials/
Methods: Data were acquired through an IRB-approved retrospective review of cervical cancer patients treated with HDR brachytherapy, utilizing both CT and MRI scans for treatment planning. OARs were delineated by our institution’s radiation oncologists for 17 cases. Contours for rectum, bladder, bowel, and sigmoid were delineated on both CT and MRI. The MRI was fused to the CT based on the HR CTV and applicator and approved by the physician. The RT doses from the delivered clinical plans were overlaid to assess dose statistics for the two contour sets. Analysis was employed to determine differences in OAR doses between MRI and CT-based contours. Results: The average time between CT and MRI scans was 4.34±1.88 hrs (0.46-6.87 hrs). The mean ± STDev of D2cc for each OAR on CT and MRI, p value, and mean ± STDev of the percent difference of the two were determined for a single fraction. Rectum D2cc on CT and MRI was 3.40 ±1.54 Gy and 3.64±1.24 Gy respectively with p=0.39 and average D2cc 34.5±124% higher on MRI. This large STDev is attributed to 2 cases where the rectum moved significantly closer to the HR CTV. Without these 2 outliers, the average percent difference in D2cc was 0.68±9.65%. Bladder D2cc on CT and MRI was 5.10 ±1.35 Gy and 4.62±1.37 Gy respectively with p=0.02 and average D2cc 9.46±16.3% lower on MRI. Sigmoid D2cc on CT and MRI was 2.68 ±1.20 Gy and 2.70±1.22 Gy respectively with p=0.84 and average D2cc 3.17±27.35% lower on MRI. D2cc for bowel was not evaluable for 6 cases (contours were outside of the 10x10cm dose grid). Bowel D2cc on CT and MRI was 3.63 ±1.36 Gy and 2.90±1.47 Gy respectively with p=0.03 and average D2cc 20.2±27.4% lower on MRI. The data show a significant decrease in D2cc to bladder and bowel when contoured on MRI as opposed to CT. There was no correlation between scan time difference and OAR D2cc percent difference between scans. Conclusion: The time difference between CT and MRI scans does not significantly impact dose to OAR contours in HDR brachytherapy treatment planning for cervical cancer. However, there was a small but significant decrease in D2cc to bowel and bladder when contoured on CT versus MRI independent of scan time difference. This may be due to improved delineation of these OARs on MRI. Contouring OARs on CT appears to have a conservative impact, implying it is safe to continue to contour OARs on the CT and the HR CTV on MRI.
