Room: AAPM ePoster Library
Purpose: Respiratory motion management is important for safe delivery of pancreas SBRT. Fiducial markers are a reliable measure of internal motion but require an invasive insertion procedure. We evaluate the geometric correlation of external and internal surrogates, and explore the potential role of internal-external motion models.
Methods: High-risk pancreas patients are being recruited to an ethics approved, single-institution clinical trial delivering 40 Gy in 5 fractions on a standard linear accelerator. 1-4 fiducial markers were inserted in or near the gross tumor, facilitating online image-based assessment of superior-inferior (SI) internal surrogate motion. Anterior-posterior (AP) respiratory amplitude was assessed using the Varian Real-time Position Management (RPM) system. All patients are treated with exhale respiratory gating, delivered during breath hold (BH) or free-breathing (FB).
The external and internal surrogate motions were aligned online at the exhale phase and retrospectively compared on pre-treatment CBCT and kV intrafraction imaging. We report correlation using Pearson’s correlation coefficient, r, and associated statistical significance, p.
Results: 11 patients have been treated on this study. Analysis from two patients show that internal and external surrogate motions on CBCT exhibit strong correlation in BH (r(BH)=0.87, p(BH)<0.001) and FB (r(FB)=0.99, p(FB)<0.001). Fiducial motion is underestimated by RPM, with a root mean square (RMS) error for BH and FB on setup CBCTs of 0.09cm and 0.16cm, respectively. Intrafraction analysis shows the amplitude of motion for the external and internal surrogates agree within a maximum deviation of 0.15cm, although weaker correlation is observed (r(FB)=0.25, p(FB)=0.014).
Conclusion: Initial experience from the first 11 patients supports a requirement for fiducial-based motion management for safe delivery of pancreas SBRT. The respiratory gated motion management strategy has been validated using online imaging. External to internal motion correlation may enable use of motion models to reduce imaging dose and facilitate a more streamlined workflow.
Funding Support, Disclosures, and Conflict of Interest: This study is partially funded and supported by Varian Medical Systems.