OA191] Influence of the degree of the geometric uncertainty on the quality of single isocenter VMAT multiple target treatments

Abstract

Purpose Stereotactic radiotherapy with VMAT (SRT – VMAT) is widely used nowadays for multitarget single isocenter treatments, demonstrating high precision in target localization and dose distribution for both frame-based stereotactic and frameless image-guided target alignment. However, geometric uncertainties such as setup errors, internal organ motion, image fusion errors, geometric distortions in MR images, etc. [1] commonly appear in radiotherapy treatment chain and occasionally lead to compromised SRT efficacy. This study investigates the influence of the degree of the geometric uncertainty on the plan quality indices clinically used for plan evaluation and acceptance, for single isocenter VMAT multiple target treatments such as multiple metastases cases. Methods Multiple targets with different sizes (1 cm and 2 cm in diameter) were contoured on an anonymized patient CT dataset, in Monaco (Elekta) treatment planning system (TPS). Single isocenter SRT plans were generated for different target number and size using four non-coplanar arcs. In each plan, different degrees of systematic geometric offsets (0 – 2 mm with a 0.5 mm step) were deliberately introduced by changing the position of the plan isocenter in the TPS accordingly. Target coverage, as well as conformity and gradient indices clinically used for plan evaluation were determined for the different degrees of introduced geometric offsets and compared with the reference values which correspond to zero geometric uncertainty. Results Geometric uncertainties greater than 1 mm in any direction may influence significantly plan evaluation indices leading to target coverage and conformity index values deterioration (>10%) for the majority of plans/targets. The degree of deterioration depends on the target volume, target number and targets’ relative position and increases as a) target volume decreases, b) target number increases and c) the amount of overlap of lesions in the BEVs increases. Conclusions For limited number of targets (