Patient specific IMRT quality assurance with film, ionization chamber and detector arrays: Our institutional experience

Purpose: This study was conducted to review patient specific IMRT QA delivered at our institution using Varian LINACs and TomoTherapy Hi-Art system and categorized according to the anatomical site, type of treatment machine, and treatment planning systems (TPS). Material and methods: Three thousand and three hundred seven patient data were evaluated for a time ranging from 2006 to 2011; these data were gauged using several methods used in the QA process. Ion chambers and film were used in 1738 patient plan QA in the earlier years followed by ion chamber arrays in 1569 patient plan QA in the latter years. Patients were grouped according to several parameters including TPS, site of treatment, and type of treatment machine in comparing the measured versus computed dose differences. From 2006 through early 2009, 736 TomoTherapy plans, 651 Pinnacle³ plans, and 351 Corvus plans were evaluated using ion chambers and films. The pass criterion at the institution at the time of these measurements was 3% dose difference and 3mm distance to agreement. For the years ranging from 2009 to 2011, 1569 patient IMRT QAs were performed and evaluated on the institution's pass criteria of 90% γ value on Varian linacs with Millennium 80, 120 and High-Definition 120 multileaf collimators. Results: Average point dose difference between measured and calculated plans for Pinnacle³, Hi-ART TomoTherapy, and Corvus TPS were 0.1205%, -0.0042%, and -0.0178%. Among the QA plans measured using a 2D array, average gamma values for brain, head and neck, thorax, abdomen, and pelvis were 97.2%, 95.7%, 96.2%, 97.0%, and 96.2%, respectively. Average gamma values based on 80, 120, HD 120 and TomoTherapy MLC configurations were 96.5%, 96.2%, 96.3%, and 97%, respectively. A 2-tailed paired Student's T-test did not reveal the presence of statistically significant differences based on either TPS, anatomical sites, number of beams or arcs, number of control points, or the MLC configuration (p-value>0.05). Conclusions: Among the TPSs, Tomoplan and Corvus had the best agreement with the point dose measurement. Based on anatomical location of treatment site, head and neck cancers had the lowest gamma value for the patients treated and brain sites had the highest gamma value. However, the values are not significantly different. TomoTherapy machines continue to have the best overall gamma values as compared to CLINAC machines.