Impact of spot positional errors in robustly optimized intensity-modulated proton therapy plan of craniospinal irradiation.

Abstract

We investigated the influence of random spot positioning errors (SPEs) on dosimetric outcomes of robustly optimized intensity-modulated proton therapy (RB-IMPT) plans in craniospinal irradiation (CSI). Six patients with CSI treated using the RB-IMPT technique were selected. An in-house MATLAB code was used to simulate a random SPE of 1mm in positive, negative, and both directions for 25%, 50%, and 75% of the total spot positions in the nominal plan. The percentage dose variation (ΔD%) in the six nominal and 54 error-introduced plans was evaluated using standard dose-volume indices, line dose difference, and 3D gamma analysis method. The introduction of a random SPE of 1mm resulted in a reduction in D99%, D98%, and D95% of both CTVs and PTVs by < 2% compared with the corresponding nominal plans. However, this leads to an increase in D1% of the lens by up to 16.9%. The line dose in the junction region showed ΔD% < 2% for the brain and upper spine and < 4% for the upper and lower spine. The 3D gamma values for 3% at 3mm and 2% at 2mm were above 99% and 95%, respectively, in all 54 error-introduced plans. The worst decrease in gamma values was observed for 1% at 1mm, with values ranging from 64 to 78% for all types of SPE. The RB-IMPT plan for CSI investigated in this study is robust enough for target coverage, even if there are random SPEs of 1mm. However, this leads to an increase in the dose to the critical organ located close to the target.