Pre-treatment verification of lung SBRT VMAT plans with delivery errors: Toward a better understanding of the gamma index analysis

Purpose: To study the sensitivity of the ArcCHECK in detecting delivery errors for lung stereotactic body radiotherapy (SBRT) using the Volumetric Modulated Arc Therapy (VMAT) technique and to evaluate the sensitivity of eight global and local gamma tolerances with different cut-off percentages. Methods: Baseline VMAT plans were generated for 15 lung SBRT patients. We delivered the smallest errors (gantry, collimator, and multileaf collimator MLC) which had >= +/- 2% dose difference in the modified treatment plans compared to the baseline plan (the clinical significance of those errors were assessed in our previous study. A total of 100 plan in which 15 baseline plans were measured using the ArcCheck detector along with ion chamber measurements. The sensitivity of the global and local gamma-index method using criteria of 1%/1 mm, 2%/1 mm, 2%/2 mm, and 3%/3 mm was investigated. Results: The gamma (gamma) pass rates for these plans exhibited considerable spread. The majority of simulated errors were not detected. Broadly similar detection levels were achieved with the different gamma criteria and cut-offs. Combining ion chamber measurements with ArcCHECK did not improve error detection. Conclusions: Commonly adopted gamma criteria are not sensitive enough to validate lung SBRT VMAT plans at the 2% dose difference level. The error detection levels are fairly consistent despite changes in gamma criteria and cut-offs. The choice of gamma criteria was not significant and there was no clear benefit in tightening the gamma criteria.