Room: AAPM ePoster Library
Purpose: To compare daily plan adaptation strategies in Locally Advanced Pancreatic Cancer (LAPC) patients treated with SBRT using a non-coplanar beam geometry and real-time tumor tracking.
Methods: A total of 37 CT scans were collected from 10 LAPC patients, consisting of 10 planning CT (pCT) and 27 daily pre-treatment scans (preCT). All CTs were manually delineated and their treatment plans were computed using VOLO?? (Accuray Inc., Sunnyvale, USA) until satisfying organs-at-risk (OAR) (stomach, duodenum and bowel) dose-constraints (V35Gy<0.5cc). Next, planned dose distributions were rigidly transferred to preCT to evaluate daily doses. Plans exceeding dose-constraints in daily organs were used to compare two adaptive strategies: 1) Full replanning without further optimization; 2) Target realignment optimization to achieve OAR constraints. Shifts were optimized by minimizing V35Gy dose-volumes while maximizing PTV coverage within a limited geometrical search space from -5 to 5mm in all directions. Resulting dose distributions were dosimetrically compared by means of PTV coverage and V35Gy volumes to determine the added value of each strategy.
Results: Transferred doses resulted in dose-constraint violations for 20/27 daily scans, with a PTV coverage and OAR V35Gy of 82 (76-84)% and 1.2 (0.9-1.9)cc for corresponding median and interquartile ranges. After full replanning vs. target realignment optimization, PTV coverage was 80 (76-84)% vs. 78 (73-81)%, and V35Gy, 0.04 (0-0.2)cc vs. 0.5 (0.4-0.8)cc, respectively. Optimal shifts resulted in displacements of 3±2mm. Adapted plans could not achieve clinically acceptable OAR dose-volumes in 3/20 and 6/20 daily scans after replanning vs. realignment.
Conclusion: Full replanning resulted in the best strategy to adapt daily plans in our cohort of LAPC patients. However, target realignment optimization achieved clinical acceptable plans in most of fractions (14/20). Based on these results, we believe target realignment optimization can be a potential alternative for pancreatic plan adaptation, requiring less workload, time and clinical resources.
Funding Support, Disclosures, and Conflict of Interest: This work was in part funded by a research grant of Accuray Inc., Sunnyvale, USA. Erasmus MC Cancer Institute also has a research collaboration with Elekta AB, Stockholm, Sweden.
Organ Motion, Target Localization, Optimization