Room: AAPM ePoster Library
This study presents the first application of the bidirectional local distance (BLD) to planned and accumulated isodose surfaces as a tool to derive set-up margins.
Ten early-stage breast and three oropharyngeal cancer patients were retrospectively selected. The breast patients underwent accelerated partial breast irradiation, delivered using intensity modulated radiotherapy (27 Gy in 5 fractions) with daily cone-beam CT (CBCT) imaging. The oropharyngeal patients underwent volumetric modulated arc therapy (70 Gy in 33 fractions) with weekly CBCT imaging. For each patient, clinical treatment plans were re-optimized with 0 mm planning target volume (PTV) margins and normalized to achieve a minimum target coverage criterion (D98% = 95%). Deformable image registration of the planning CT to CBCTs enabled dose accumulation of these zero-margin plans. BLDs were computed between the planned and accumulated 95% isodose surfaces relative to each clinical target volume (CTV) contour point. An isotropic set-up margin was defined as the 95?? percentile of the BLDs in regions where the accumulated 95% isodose surface shifted towards the CTV. The treatment plans were re-optimized again with the derived set-up margin applied to the CTV and the resulting accumulated CTV coverage was evaluated using D98% = 95%.
The median (range) set-up margin derived using the BLD was 3 mm (2–8) for the breast patients and was 2 mm for all three oropharyngeal patients. When treatment plans were re-optimized with the derived set-up margin applied for each patient case, the median (range) accumulated CTV D98% was 96% (95–99). All 13 patients met the criteria for adequate CTV coverage.
Application of the BLD to planned and accumulated isodose surfaces for zero-margin treatment plans can be used to derive appropriate set-up margins. These results can be combined with other uncertainties in the external beam radiotherapy workflow to define evidence-based PTV margins.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the Natural Sciences and Engineering Research Council of Canada, and Alberta Innovates. There are no relevant financial disclosures or conflicts of interest to declare.