PET Signal Guidance In Biology-Guided Radiotherapy: Impact of Deblurring and Anchor Point Tracking on Dosimetry
Abstract
Purpose
To assess the effectiveness of a developmental anchor point tracking algorithm with PET deblurring in targeting tumors on the RefleXion X2 platform.
Methods
Planning using PET-based biology-guided radiotherapy (BgRT) on the RefleXion X2 removes target motion by deblurring the PET-avid region of the GTV. Deblurring collapses target motion into a single static frame for anchor point tumor tracking. Delivery accuracy is contingent on how well this frame of reference is preserved between the plan and delivery even with motion changes. An 18F-fluorodeoxyglucose fillable phantom was inserted into an ArcCHECK positioned on the CIRS motion platform. The phantom contained a 22 mm diameter spherical target. Studies were prepared at 10 Gy/fx using identical plan constraints with 8:1 target-to-background ratios. Six studies were performed using combinations of a static target and 15 mm cos4 4-second periodic motion, and step shifts. Study 1 used a static PET based plan and delivered to a static target (static-to-static baseline). Study 2 used motion at planning and motion at delivery (motion-to-motion baseline). Study 3 tested motion at planning but no motion at delivery. Study 4 used the static PET for planning but delivered with cos4 motion, measuring whether the system can track when no motion was used at planning. Studies 5 and 6 used motion at planning and a +/- 10 mm step shift of a static target at delivery.
Results
The ArcCHECK measured 3%/2mm gamma pass rates were 98.6%, 97.1%, 97.6%, 90.1%, 99.5% and 98.9% for Studies 1-6, respectively. When comparing Study 1 and Study 4 using film, tracking was preserved with dose shape differences within the PTV.
Conclusion
BgRT plans with the highest gamma passing rates were observed when maintaining or reducing target motion at delivery. The most extreme case, Study 4, exceeded 90% 3%/2mm which could still be clinically acceptable.