Geometric and dosimetric consequences of intra-fractional movement in single isocenter non-coplanar stereotactic radiosurgery.

Purpose: To investigate the geometric and dosimetric impacts of intra-fractional movement for patients with single or multiple brain metastasis treated using Varian Hyperarc™ mono-isocentric radiosurgery.

Methods: A total of 50 single or hypo-fractionated Hyperarc™ treatment courses (118 lesions) were included in the analysis. Intra-fractional translational and rotational movements were quantified according to the post-treatment cone-beam CT (CBCT). Geometric displacements of all targets were calculated individually based on the assessed head movement in each treatment fraction and their relationships with treatment time and target-to-isocenter distances were studied. For dosimetric analysis, only single-fraction treatments (56 lesions) were included. Re-planning was performed with 0, 1, and 2 mm planning target volume (PTV) margins. Doses were then re-calculated on rotated CT images with isocenter shifted which emulate the change in patient treatment position. Target coverage, target and normal brain doses before and after intra-fractional movement were compared.

Results: The mean 3D target displacements was 0.6 ± 0.3 (SD) mm. Target shifts for patients treated within 10 min were significantly smaller than those treated in longer sessions. No correlation was found between target shift and target-to-isocenter distance as the origin of head rotation was not located at the isocenter. Loss of target coverage and minimum Gross Tumor Volume (GTV) dose due to intra-fractional movement were apparent only when no margin was used, leading to an extra 23% of the targets violating the dose acceptance criteria, in contrast, the effects on normal brain V were negligible regardless of the margin used. The use of 1 mm PTV margin can compensate clinically significant geographical miss caused by intra-fractional movements while limiting V to within dose criteria for 88% of the cases. The plan acceptance rate (fulfillment of both target and normal brain dose criteria) after intra-fractional movement was also the highest with the 1 mm margin.

Conclusion: Although intra-fractional movements during Hyperarc™ treatments were small, there were substantial dosimetric effects due to the sharp dose fall-off near target boundaries. These effects could be mitigated by using a 1 mm PTV margin and maintaining the effective treatment time to within 10 min.