Abstract
BACKGROUND AND PURPOSE: This study aimed to evaluate the dosimetric performance and robustness of volumetric modulated arc therapy (VMAT) planning techniques-planning target volume with internal margin (INT), auto flash (AF), and virtual bolus (VB)-under simulated geometric changes during whole-breast radiotherapy. METHODS: Nine patients with left-sided breast cancer were included. Three planning techniques were compared: INT with a 5-mm skin-sparing internal margin, AF with a 1-cm automatic skin flash margin, and VB with a 5-mm optimization bolus applied only during the planning. Respiratory motion was simulated by shifting the plan center (± 3 mm and ± 5 mm) and recalculating the dose distributions. RESULTS: Under static conditions, all plans provided adequate target coverage, with planning target volume (PTV) V95% values of 96.45% (INT), 97.35% (AF), and 98.19% (VB). Under breast swelling of up to 5 mm, AF maintained the most stable coverage (PTV V95% = 99.10%), outperforming VB (95.02%) and INT (92.31%) (p < 0.001). In simulated incomplete inspiration (5 mm), VB showed superior robustness, achieving a PTV V95% of 90.46% compared with AF (85.30%) and INT (85.08%) (p = 0.008). AF met the ideal criteria in all cases under swelling conditions, whereas the VB was more effective against suboptimal respiration. CONCLUSIONS: In VMAT breast radiation therapy (RT), the conventional INT approach was the least robust against variations, and additional techniques are required. AF effectively compensates for breast swelling, whereas VB preserves the target coverage under insufficient breath-hold conditions. Surface-guided radiation therapy (SGRT) and visual guidance are recommended to ensure accurate treatment.