Daily Head and Neck Treatment Assessment for Optimal Proton Planning Robustness

Abstract

Daily setup variation can affect the quality of treatment and robust optimization is used to ensure adequate target coverage. We aimed to assess quality assurance by evaluating fractional plan quality and its impact on plan robustness. Eligible head and neck plans with full field of view cone beam CT (CBCT) for >75% of daily fractions were classified into 2 categories 1) Primary only (PO) coverage and 2) Primary and neck nodal (PNN) coverage. Registration used at the machine was utilized for each daily CBCT to generate a synthetic CT (sCT), and the clinical target volumes (CTVs) and organs at risk (OARs) were manually edited by the reviewing physician. The daily dose for the CTVs and OARs were compared to the expected range from the robustness bands using 3.5% range uncertainty and 3mm versus 5mm setup uncertainty, fractional error was defined as volume and dose outside of uncertainty constraints. Statistical significance was calculated using students t-test. There was 203 eligible daily fractions from 6 patients were included for analysis. The average percentage of fractions that deviated from planning CTV volumes as well as average error in delivered dose using 3mm and 5mm robustness can be seen in (Table 1). In PO plans, there was minimal change in percentage of fractions that deviated from planned CTV volumes and overall low average error in delivered dose at both 3mm and 5mm robustness, 0.35% and 0.26%, respectively. In comparison, PNN plans had greater percentage of fractions that deviated from planning CTV and consequently greater average error in delivered dose, 1.00% versus 0.37% at 3mm and 5mm robustness, respectively. There was a significant difference in the percentage of fractional error with 3mm robustness between PO patients, 0.22%, and PNN patients, 0.98%, p<0.01. No statistical difference in percentage of fractional error was found with 5mm robustness for PO and PNN patients, 0.24% and 0.52%, respectively. All OARS maintained mean daily dose within or below 3mm robustness calculations. Based on our analysis, there was a difference in error of delivered dose between 3mm and 5mm robustness for head and neck proton plans, particularly for PNN coverage. Clinically, up to 1% error to CTV volumes is often not significantly meaningful and treating centers should consider utilizing 3mm robustness to maximize OAR sparing while maintaining proper target coverage. sCTs analysis is important for determining planning guidelines and may be able to identify patients needing an adaptive plan instead of currently utilized verification CTs.