Abstract
BACKGROUND: Rotational error cannot be overlooked in single-isocenter multi-target (SIMT) stereotactic radiotherapy. This retrospective study aimed to evaluate the treatment accuracy of linear accelerator-based fractionated stereotactic radiotherapy (FSRT) using SIMT non-coplanar volumetric modulated arc therapy (VMAT) in patients with multiple brain metastases. We explored the impact of rotational error on planning target volume (PTV) margins, providing clinical evidence for the selection of appropriate PTV margin values. METHODS: A total of 161 patients with multiple brain metastases (733 treatments; actual clinical PTV margins ranged from 1~2 mm) were included. Theoretical PTV margins were calculated based on the Van Herk and Jenghwa Chang formulas. We analyzed the influence of the distance from each target to the treatment isocenter, rotational errors, and PTV margin on treatment outcomes. Additionally, individualized PTV margins for each patient were calculated using the Jenghwa Chang formula and patients were divided into subgroups according to a 2-mm threshold for further analysis. RESULTS: The mean residual translational setup errors ranged from -0.04~0.01 mm, and rotational setup errors ranged from 0.15°~0.49°, both within acceptable limits. According to the Van Herk formula, required margins in posterior-anterior, superior-inferior, and right-left directions were 1.44 mm, 1.68 mm, and 1.78 mm, respectively. By incorporating both translational and rotational errors using the Jenghwa Chang formula, the comprehensive margin ranged from 1.69~1.79 mm (calculated based on the 95% confidence interval of distances from targets to isocenter). Additionally, when the mean distance from all targets to their respective treatment isocenters was 30.62 mm, the required margin calculated solely for translational errors using the Jenghwa Chang formula was 1.23 mm; if rotational errors were neglected, target coverage probability would decrease from 95% to 73%. Further subgroup analysis showed that 25 patients whose individualized theoretical margins exceeded 2 mm tended to experience worse outcomes compared to others, including intracranial local failure (ILF, defined as lesion progression within the previously irradiated intracranial region during follow-up; 32.00% vs. 22.29%, P = 0.32), one-year local control (64.00% vs. 65.44%, P = 0.89), and one-year intracranial progression-free survival (iPFS, 44.00% vs. 51.45%, P = 0.85). However, these differences did not reach statistical significance. CONCLUSION: This study confirms that the SIMT non-coplanar VMAT technique ensures treatment accuracy for FSRT in multiple brain metastases. Rotational errors reduce dose coverage, and a minimum safety margin of 1.79 mm is recommended to ensure tumor coverage and reduce local failure, providing a basis for future treatment optimization.