Author: Wen C. Hsi, Tae Kyu Lee, Biniam Yo Tesfamicael 👨🔬
Affiliation: Allina Health, Department of Radiation Oncology, University of Arkansas for Medical Sciences (UAMS), Oklahoma Proton Center 🌍
Purpose: When minimal dose-variations induced by inter-fractional anatomical-changes and positioning-deviation were found for having limited impact on clinic-outcomes of oropharyngeal head-and-neck (HN) cancer for patients treated with pencil-beam-scanning PBS proton-beam-therapy (PBT), this study focused on the biological TCP/NTCP modelling for oropharyngeal cancer of patients treated with PBS and uniform-scanning US-PBT in two institutions.
Methods: We selected 8 oropharyngeal of 55 HN-cancer patients treated PBS-PBT with simultaneous-integrated-boost SIB approach in one-institute. We also selected 10 oropharyngeal-cancer treated US-PBT with sequential-boost SB approach in another-institute. Similar image-guided-radiation-therapy (IGRT) of patient-positioning was utilized in both institutions. The positioning-deviation and dose-variation of anatomical-change were investigated in PBS-PBT institute. The treatment plan of PBS-PBT was constructed with a 3mm-3% robustness-optimization for having 100% prescription-dose covering 97-99% volume of clinical-treatment-volume (CTV) on a planning CT (PCT). To investigate dose-variation, the planned beam-set on PCT was rigidly mapped to recalculate on repeated-CT QACT then was deformably transported back to the PCT. Although the fractional-dose to organ-at-risk OAR can varied significantly in SB approach between initial and boost plan, we used the mean fractional-dose to OARs for our TCP/NTCP modelling by the concept of QUANTEC constraints.
Results: Although the deviation of couch-position over the course of each patient has the 2.5/2.0 mm mean-value on longitudinal maximum-standard-deviation for brain/HN respectively, the deviation between treatment and planned-isocenter was reduced within 1.0mm standard-deviation on all 3 directions for brain/HN. The minimal position-deviation indicated the potential dose-variation mainly caused by anatomical-changes. Also, dose-variations of the anatomical changes over the 3mm-3% robustness-planning strategy has small impact. Without considerable dose-variation, our initial TCP/NCTP models used the mean fractional-dose over total-fractions to predict clinic-outcomes.
Conclusion: We built biological TCP/NCTP models using mean fractional dose for SIB approach. However, the models need to include considerable shift on fractional-dose between initial and boost plans for SB approach.