Efficacy of Respiratory Motion Tracking for Cardiac Radioablation with the Cyberknife
Abstract
Purpose
To estimate the impact of heartbeat on the dose distribution in cardiac radioablation (CRA) patients treated with CyberKnife and compare the error in respiratory tracking to cancer patients with a similar protocol.
Methods
The blurring of the dose distribution caused by organ motion was modeled as a convolution of the planned dose distribution with a probability density function (PDF) retrospectively for 36 CRA patients. PDFs were constructed from the correlation error (CE), i.e. the difference between the predicted and actual position of the tracking fiducial (ICD lead), and from the mismatch in motion trace between tracking fiducial and target for a sub-cohort of 10 patients with 4DCT data. Treatment logs were obtained for 36 CRA and 10 cancer patients. CE periodicity with respect to breathing phase was analyzed with least-squares sinusoidal fitting and Fourier analysis.
Results
The spread in CE causes non-significant changes under 5% to dose statistics for CRA patients, e.g. increasing the V11.9Gy of the esophagus by 4.3% and decreasing the near-maximum dose to the stomach by 2.2% on average, without affecting adherence to dose constraints. When accounting for mismatch between the target and the tracking fiducial, mean target coverage is reduced from 99.0% to 95.5%. Frequency analysis of CE revealed no periodicity in all conditions except in the inferior-superior direction for CRA patients. The absolute CE was minimized in full inhale and exhale and maximized in dynamic breath phases. The average CE vector magnitude was higher in the CRA group.
Conclusion
This respiratory compensation system exhibits higher average CE for cardiac targets, most likely caused by heartbeat. Despite this, doses to target or organs are not significantly impacted, supporting the safety of this treatment. A decrease in target coverage when accounting for tracking mismatch supports the use of patient-specific safety margins in addition to respiratory tracking.