Author: Wesley S. Culberson, Albert Du, Ryan T. Flynn, Alonso N. Gutierrez, Patrick M Hill, Daniel E. Hyer, Kaustubh A. Patwardhan, Blake R. Smith, Nhan Vu, Karsten K. Wake 👨🔬
Affiliation: University of Wisconsin, Department of Medical Physics, School of Medicine and Public Health, University of Wisconsin - Madison, Miami Cancer Institute, Baptist Health South Florida, Department of Human Oncology, University of Wisconsin School of Medicine and Public Health, University of Iowa, Iowa Health Care 🌍
Purpose:
To perform end-to-end treatment planning and delivery verification of energy-specific collimated treatments optimized to achieve the maximum obtainable high-dose conformity using the dynamic collimation system (DCS) in pencil beam scanning (PBS) proton therapy.
Methods:
A comparison of 5 uncollimated and DCS-collimated treatments were performed using clinical datasets involving brain tumors located within 5 mm of the brainstem. DCS-collimated treatments were planned on an FDA-cleared treatment planning system, Astroid, that incorporates an enhanced partial voxel gradient descent trimmer localization algorithm that geometrically initializes and refines spot and trimmer positions to optimize the lateral penumbra near organs at risk. Beamlet weights were Pareto-optimized in Astroid to achieve optimal healthy tissue sparing while ensuring 95% target coverage and maximum dose of 107% of the prescription in nominal plans. Patient-specific quality assurance (PSQA) using a planar detector (res. 3.8mm) was performed on a subset of the optimized treatment fields to verify deliverability and benchmark clinical accuracy.
Results:
Uncollimated and DCS-collimated plans maintained equivalent target coverage within 0.6% while the DCS reduced the mean dose to the brainstem by 33.49% and the D2cc maximum dose by 18.77%, and the mean dose to the 10mm rind around the target by 18.92%. The 50% conformity index was 2.73 for uncollimated plans and was 2.20 for collimated plans. Deliverability of these highly modulated and collimated treatments was demonstrated with excellent agreement between the measured and calculated PSQA plans for the DCS-collimated and PBS plans (98.2% and 99.5%, respectively). DCS treatments showed no statistically significant deviation in their PSQA passing rates compared to the uncollimated treatment plans.
Conclusion:
Maximum conformity treatment planning has been achieved for the DCS technology applied to PBS treatments. While these treatments are highly modulated, deliverability and accuracy have been confirmed for these highly collimated treatments with the DCS.