Purpose: Stereotactic MR-guided online adaptive radiotherapy (SMART) is a rapidly emerging technique in which new treatment plans are generated based on a patient’s anatomy at each fraction. However, this process can also be time and resource intensive. The goal of this study is to determine which disease sites are most likely to meet criteria for the use of adaptive techniques.
Methods: A total of 230 patients were treated with 5-fraction (or fewer) SMART treatments on MR-guided linear accelerator at our institution (1145 total fractions) between October 2019 and February 2021. At each fraction, the patient’s anatomy was re-contoured by a physician and a treatment plan was re-calculated on the current anatomy to predict the dose that would have been delivered without adaptation. This prediction was compared to pre-specified prescription metrics, and adaptation was determined to be necessary if any critical organ at risk (OAR) metrics were exceeded, or if the target coverage was unacceptable. Patients were categorized by disease site (kidney/adrenal, liver, lymphoma, oligometastatic nodes, pancreas, prostate, spine and thoracic), and the timing of each adaptive fraction was recorded.
Results: A total of 1038 of 1145 fractions (90.7%) showed indication for adaptation, with 734 due to OAR metrics being exceeded (64.1%) and 304 due to suboptimal target coverage (26.6%). The sites with the highest proportion of fractions meeting adaptive criteria were prostate (85/85, 100%) and pancreas (310/323, 96.0%), whereas the sites that least frequently met the criteria were kidney/adrenal (71/94, 75.5%) and oligometastatic nodes (345/395, 87.3%). The median time for plan adaptation was 31 minutes (range 11-68).
Conclusion: There was a clinical benefit to plan adaptation for the majority of patients and fractions. However, adaptive need varied by disease site, with pancreas and prostate SBRT patients most likely to meet adaptive criteria.
Funding Support, Disclosures, and Conflict of Interest: CW, LS, RM, DC, EH have received research support from ViewRay Technologies outside the scope of this work.
Not Applicable / None Entered.