Impact of two-dimensional dosimetric leaf gap on SRS-VMAT for single-isocenter multi-target brain metastases

To explore the effects of two-dimensional dosimetric leaf gap (DLG) on the gamma pass rate (GPR) for single-isocenter multi-target (SIMT) stereotactic radiosurgery for patients with brain metastases. Two-dimensional DLG data were measured by an electronic portal imaging device (EPID) with the sweeping gap methods provided by Varian. A total of 56 arc fields from 28 SIMT patients were analyzed, including 19 with multiple targets and 9 with single targets. The leaf gap in patients' plan were modified using an in-house program, which adjusted the multileaf collimator positions based on the DLG values from the two-dimensional DLG distribution. Dose measurements were performed with EPID for both the original and modified plans. Comparisons were made between the measurements and the predicted dose planes from the treatment planning system using 3%/2 mm gamma criteria. Additionally, factors influencing the improvements in measurement results were analyzed.Compared to the original plan, the GPR for sweep test plans with leaf gaps of 2 mm, 4 mm, 6 mm, 10 mm, 14 mm, 16 mm and 20 mm increased by 12.2%, 17.6%, 25.4%, 58.2%, 83.6%, 84.3%, and 82%, respectively, after modification. For the clinical VMAT fields, GPR improvements fluctuated significantly due to the variations of modulation complexity score, the monitor units, as well as the number and volume of targets. The maximum GPR improvements with 16% and 36.3% were observed for the single-isocenter single-target and SIMT patients respectively. A linear relationship was found between GPR improvement and both modulation complexity score and the number of targets, whereas no significant correlation was found with monitor units or target volume. The two-dimensional DLG significantly impacts treatment plans for SIMT and single-isocenter single-target patients, particularly for SIMT cases. The GPR improvement correlates linearly with increases in modulation complexity score and the number of targets, while the monitor units and target volume show minimal correlation with GPR improvement. This study underscores the importance of accurate beam modeling as a key factor influencing GPR improvement.