Feasibility of Coplanar VMAT for Brain Metastases Using Halcyon With Knowledge-based Planning from Non-coplanar Plan

This study aimed to clarify the feasibility of creating coplanar single-isocenter volumetric modulated arc therapy (SI-VMAT) with a Halcyon using a knowledge-based planning system based on non-coplanar plans for treating multiple brain metastases.

A knowledge-based RapidPlan (RP) planning model was built using 32 TrueBeam SI-VMAT plans including one full arc and three non-coplanar partial arcs. To confirm its applicability across different beam geometries using the Halcyon system, the model was validated on the basis of the root-mean-square error (RSME), deviation rate, and absolute differences between estimated and actual dose-volume histograms (DVHs). Dosimetric performance of TrueBeam and Halcyon plans was then compared for 10 patients.

The average RSME values for the Halcyon and TrueBeam plans were 0.75±0.40 Gy and 0.46±0.28 Gy, respectively (p<0.05). Using the model for Halcyon plan creation, deviations from the estimated DVH boundaries were mainly observed in the low-dose region (<9 Gy), while the actual DVH closely aligned with the estimated DVH in the intermediate- and higher-dose regions. For comparison of dosimetric performance, statistically significant differences were found in the gradient index, conformity index, and normal brain volumes receiving ≥12 Gy, ≥18 Gy, and ≥27 Gy. However, these differences were deemed clinically negligible, with less than 1% disparity in the above doses to the normal brain.

The RP model can effectively predict doses for Halcyon, regardless of different beam configurations. Coplanar SI-VMAT with Halcyon can feasibly be used in clinical applications, such as the treatment of multiple brain metastases.