Daily Head and Neck Treatment Assessment for Optimal Proton Therapy Planning Robustness

SIMPLE SUMMARY: Proton therapy requires robust optimization to ensure patients receive target coverage and spare normal tissues. However, validation of setup variables has not been studied in patients receiving head and neck proton therapy. The aim of our retrospective study was to evaluate the absolute fractional deviation for dose and volume treatment metrics using 3 mm vs. 5 mm setup robustness in patients treated with head and neck proton therapy. We found that deviations from planning clinical target volumes were greater using 3 mm as compared to 5 mm setup robustness. In addition, variation was greater for patients with primary and secondary clinical targets as compared to primary targets alone. However, the average fractional deviation was less than 1% in primary targets and up to 2% in secondary targets using 3 mm setup uncertainty. We recommend robustness optimization using 3 mm setup uncertainty with daily CBCT for patients being treated with proton head and neck therapy. ABSTRACT: Robust optimization in proton therapy ensures adequate target coverage; however, validation of fractional plan quality and setup uncertainty in patients has not been performed. We aimed to assess plan robustness on delivered head and neck proton plans classified into two categories: (1) primary only (PO) and (2) primary and neck nodal (PNN) coverage. Registration at the machine was utilized for daily CBCT to generate a synthetic CT. The dose for the clinical target volume (CTV) and organs at risk (OAR) was compared to the expected robustness bands using 3.5% range uncertainty and 3 mm vs. 5 mm setup uncertainty. The fractional deviation was defined as D95% and V100% outside of uncertainty constraints. About 203 daily fractions from 6 patients were included for analysis. The percentage of fractions that exceeded robustness calculations was greater in 3 mm as compared to 5 mm setup uncertainty for both CTV and OAR volumes. PO plans had clinically insignificant average fractional deviation, less than 1%, in delivered D95% and V100%. In comparison, PNN plans had up to 2.2% average fractional deviation in delivered V100% using 3 mm robustness. Given the need to balance dose accuracy with OAR sparing, we recommend the utilization of 3 mm setup uncertainty as an acceptable simulation of the dose delivered.