Purpose: To evaluate the effect of plan complexity metrics on volumetric modulated arc therapy (VMAT) dosimetric accuracy and plan deliverability with the possibility of prediction of gamma passing rate (GPR).
Methods: A total of 103 clinical VMAT plans including 13 mediastinum, 25 head & neck (H&N), 40 intensity modulated radiosurgery (IMRS), and 25 prostate cases were created in RayStation (V.10) treatment planning system. 6MV-FFF of Varian Edge was used for all mediastinum and IMRS cases, where TrueBeam (6MV_FFF) for H&N and TrueBeam 10 MV beams were used for prostate cases. In-house scripts were developed in Iron Python to compute several Modulation Indices (MI) such as Plan Averaged beam area (PA), Plan averaged beam Irregularity (PI), total MU, Leaf Travel (LT/AL), mean Dose Rate variation (DR), and mean Gantry Speed variation (GS). Pretreatment verifications were performed on ArcCHECK phantom with SNC software (V8.2). Gamma was calculated with 3%2mm and 10% threshold according to TG-218. SPSS (V.27) was used for Pearson correlation and linear regression model.
Results: The following results were observed: a strong correlation between GPR and PA, PI, total MU and GS, (0.713, -0.682, -0.66, and 0.654 respectively), moderate correlation between GPR and LT (-0.382), and weak correlation between GPR and DR (-0.497). Based on linear regression model, PA, PI, and GS have most impact on GPR (GPR = 95.187+0.085PA-0.225PI+1.14GS).
Conclusion: We have studied the relationship between GPR and complexity of VMAT by analyzing different MIs. We found that PA, PI, GS & total MU has the greatest impact on GPR; however, because MU is already included in PA and PI calculations these metrics may not be orthogonal. Based on our results, the calculation of PA, PI, and GS may be useful in guiding VMAT plan evaluation and ultimately reducing uncertainties in planning and radiation delivery.