Reproducibility of the lung anatomy using Active Breathing Control: Dosimetric consequences for scanned proton treatments

Purpose/Objective
The treatment of moving targets with scanning proton beams is challenging. By controlling lung volumes, Active Breathing Control (ABC) assists breath-holding for motion mitigation. The delivery of proton treatment fractions often exceeds feasible breath-hold durations, requiring high breath-hold reproducibility. Therefore, we investigated dosimetric consequences of anatomical reproducibility uncertainties in the lung under ABC, evaluating robustness of scanned proton treatments during breath-hold.

Material/Methods
T1-weighted MRIs of five volunteers were acquired during ABC, simulating image acquisition during four subsequent breath-holds within one treatment fraction. Deformation vector fields obtained from these MRIs were used to deform 95% inspiration phase CTs of 3 randomly selected non-small-cell lung cancer patients (Figure 1). Per patient, an intensity-modulated proton plan was recalculated on the 3 deformed CTs, to assess the dosimetric influence of anatomical breath-hold inconsistencies.

Results
Dosimetric consequences were negligible for patient 1 and 2 (Figure 1). Patient 3 showed a decreased volume (95.2%) receiving 95% of the prescribed dose for one deformed CT. The volume receiving 105% of the prescribed dose increased from 0.0% to 9.9%. Furthermore, the heart volume receiving 5 Gy varied by 2.3%. Figure 2 shows dose volume histograms for all relevant structures in patient 3.

Conclusion
Based on the studied patients, our findings suggest that variations in breath-hold have limited effect on the dose distribution for most lung patients. However, for one patient, a significant decrease in target coverage was found for one of the deformed CTs. Therefore, further investigation of dosimetric consequences from intra-fractional breath-hold uncertainties in the lung under ABC is needed.