Validation of Lexicographic Optimization-Based Planning for Brain Metastases Radiosurgery with Coplanar Arcs

Abstract

<h3>Purpose/Objective(s)</h3> This study aimed to validate a not yet commercially available fully-automated lexicographic optimization planning for intracranial stereotactic radiosurgery (SRS) with coplanar arcs. <h3>Materials/Methods</h3> Twenty-one single-lesion SRS treatment plans (21 Gy/1 fx) delivered between November 2019 and December 2021 were retrospectively selected and automatically re-planned. An a-priori assigned priority list, a so-called Wish List (WL), was used to define the sequential multi-criterial optimization: four patient sets were necessary to achieve a robust WL. All plans were optimized with 2 coplanar 140°-arcs and calculated with the Monte Carlo algorithm (1 mm-dose grid, 0.5%-statistical uncertainty). The main criteria for planning approval was a brain volume receiving more than 12 Gy less than 10 cm³ (V_12Gy<10 cm³). A target coverage as high as possible was requested, with at least the 80% of the prescription dose covering the 99% of the planning target volume (PTV). Manual plans (MP) and automatic plans (AP) were compared in terms of dose-volume constraints and monitor units (MUs). Statistical significance was assessed performing the Wilcoxon-Mann-Whitney test with Bonferroni correction for multiple tests (alpha=0.05). Plan deliverability was verified by pre-treatment QA and gamma index analysis. <h3>Results</h3> The 21 AP re-planning took only 5 working days. The median D_98% of the gross tumor volume (GTV) and the PTV for MP and AP was 20.4 Gy [18.5-21.6] and 18.1 Gy [16.8-19.9], and 21.1 Gy [19.2-21.7] and 19.1 Gy [17.6-19.4], respectively. These increases were statistically significant (p<0.03). A significant improvement in PTV Paddick's conformity index has been registered: 0.3[0.0-0.6] for MP and 0.5[0.3-0.9] for AP with a p<0.001. The brain V_12Gy was 7.3 cm³ [4.4-12.7] and 7.6 cm³[4.2-13.3] for MP and AP, respectively (p>0.05). Other organs at risk were never significantly interested by clinically relevant doses. These AP results were obtained with a lower median number of MU (-11.6%) even if this difference was not statistically significant and plans registered a comparable gamma analysis (local 2%/2mm). <h3>Conclusion</h3> The novel autoplanning produced high-quality clinically acceptable radiosurgery plans with coplanar arcs, significantly reducing the overall planning time from about one working day for one MP to about 2 hours for one AP. While the OAR sparing was comparable between MP and AP, the target coverage was significantly increased, reducing the MU number and preserving the plan deliverability. The validation proved the capability to generate high-quality deliverable plans according to institutional-specific planning protocols.