Adapt-On-Demand: A Novel Strategy for Personalized Adaptive Radiation Therapy for Locally Advanced Lung Cancer
Ruiqi Li,Tingliang Zhuang,Steven Montalvo,Kai Wang,David Parsons,Yuanyuan Zhang,Puneeth Iyengar,Jing Wang,Andrew Godley,Bin Cai,Mu-Han Lin,Kenneth Westover
DOI: https://doi.org/10.1016/j.prro.2024.02.007
IF: 3.439
2024-05-06
Practical Radiation Oncology
Abstract:Purpose Real-time adaptation of thoracic radiation plans is compelling because offline adaptive experiences show that tumor volumes and lung anatomy can change during therapy. We present and analyze a novel adaptive-on-demand (AOD) workflow combining online adaptive radiation therapy (o-ART) on the ETHOS system with image guided radiation therapy delivery on a Halcyon unit for conventional fractionated radiation therapy of locally advanced lung cancer (LALC). Methods and Materials We analyzed 26 patients with LALC treated with the AOD workflow, adapting weekly. We timed segments of the workflow to evaluate efficiency in a real-world clinic. Target coverage and organ at risk (OAR) doses were compared between adaptive plans (ADP) and nonadaptive scheduled plans (SCH). Planning robustness was evaluated by the frequency of preplanning goals achieved in ADP plans, stratified by tumor volume change. Results The AOD workflow was achievable within 30 minutes for most radiation fractions. Over the course of therapy, we observed an average 26.6% ± 23.3% reduction in internal target volume (ITV). Despite these changes, with o-ART, ITV and planning target volume (PTV) coverage (V100%) was 99.2% and 93.9% for all members of the cohort, respectively. This represented a 2.9% and 6.8% improvement over nonadaptive plans ( P 10%, V100% was 93.1% for o-ART and 76.4% for nonadaptive plans, representing a median 17.2% improvement in the PTV coverage ( P < .05). In these plans, critical OAR constraints were met 94.1% of the time, whereas in nonadaptive plans, this figure was 81.9%. This represented reductions of 1.32 Gy, 1.34 Gy, or 1.75 Gy in the heart, esophagus, and lung, respectively. The effect was larger when tumors had shrunk more than 10%. Regardless of tumor volume alterations, the PTV/ITV coverage was achieved for all adaptive plans. Exceptional cases, where dose constraints were not met, were due to large initial tumor volumes or tumor growth. Conclusions The AOD workflow is efficient and robust in responding to anatomic changes in LALC patients, providing dosimetric advantages over standard therapy. Weekly adaptation was adequate to keep pace with changes. This approach is a feasible alternative to conventional offline replanning workflows for managing anatomy changes in LALC radiation therapy.
oncology,radiology, nuclear medicine & medical imaging