Changes in the radiological depth correlate with dosimetric deterioration in particle therapy for stage I NSCLC patients under high frequency jet ventilation.

BACKGROUND: Particle dose distributions are highly sensitive to anatomy changes in the beam path, which may lead to substantial dosimetric deviations. Robust planning and dedicated image guidance together with strategies for online decision making to counteract dosimetric deterioration are thus mandatory. We aimed to develop methods to quantify anatomical discrepancies as depicted by repeated computed tomography (CT) imaging and to test whether they can predict deviations in target coverage. MATERIAL AND METHODS: Dedicated software tools allowed for voxel-based calculations of changes in the water equivalent path length (WEPL) in beam directions. We prepared proton and carbon ion plans with different coplanar beam angle settings on a series of lung cancer patients, for which planning and localization CT scans under high frequency jet ventilation (HFJV) for tumor fixation were performed. We investigated the reproducibility of target coverage between the optimized and recalculated treatment plans. We then studied how different raster scan and planning settings influence the robustness. Finally, we carried out a systematic analysis of the variations in the WEPL along different coplanar beam angles to find beam directions, which could minimize such variations. RESULTS: The Spearman's correlations for the GTV ΔV95 and ΔV98 with the ΔWEPL for the proton plans with a 0° and -45° two-field configuration were 0.701 (p = 0.02) and 0.719 (p = 0.08), respectively. For beam configurations 0° and -90°, or 0° and + 45°, with lower ΔWEPL, the correlations were no significant. The same trends were observed for the carbon ion plans. Increased beam spot overlap reduced dosimetric deterioration in case of large ΔWEPL. CONCLUSION: Software tools for fast online analysis of WEPL changes might help supporting clinical decision making of image guidance. Raster scan and treatment planning settings can help to compensate for anatomical deviations.