LAUSR

Introduction The aim of this work is to analyze the dosimetric impact of using Acuros algorithm (AXB) instead of Anisotropic Analytical Algorithm (AAA) in lung treatment for standard fractions and for hypofractionated stereotactic radiotherapy using VMAT and dynamical conformal arc (DCA). Methods Dosimetric impact was evaluated for VMAT treatment and for DCA. Plans were realized with the Eclipse TPS (v.13.7). Twelve VMAT patients were selected with the attempt to deliver 66 Gy (2 Gy/fraction) to the PTV median dose. Plans were calculated with AXB (v.13.7.14) keeping the same ballistic, optimization and number of monitor units (UM) as the AAA (v.13.7.14). Those plans were then optimized with AXB. For DCA treatments, two groups of five patients with a prescription of 3 × 18 Gy (resp. 5 ×11 Gy) to the 80% isodose to treat peripheric tumors (TP) (resp. central tumors (TC)) were chosen. Plans were calculated with AXB with the same ballistic and UM delivered with the AAA. AXB reports the absorbed dose in two modes: dose-to-medium (AXB-Dm) and dose-to-water (AXB-Dw). Both of them were studied. For each technic and each plan, relative dose differences in comparison to the AAA were calculated for ITV, PTV and organs at risk (lung, heart, spinal cord). Results In VMAT, calculation with AXB-Dm leads to a decrease of the mean PTV dose of 0.9% on the D95% and of 1% on the D50%. No significant dose differences was noted on lungs (mean doses and V20Gy) between the two algorithms. Using AXB, doses delivered to the heart and the spinal cord are reduced by 2–3% on average. In DCA, the PTV coverage calculated with AXB-Dm is on average decreased by 5.4% on the D95% for TP and by 3.8% for TC. The difference observed on the V20Gy of healthy lungs is decreased by −2.6% on average for TP while no differences are seen on TC. The maximum dose delivered to the spinal cord (+ 3 mm) and the dose to the heart (D1 cm3) are reduced for both types of tumors with AXB. For both techniques, calculations made with AXB-Dw display similar results. Conclusions The first results of this study seem to indicate that the transition to the AXB algorithm has a greater impact on DCA treatments than on VMAT: roles of the prescription mode and of the field size must be deeper analyzed. As expected, AAA overestimates the dose in low density heterogeneities and near air-tissue interfaces. The effects of the algorithm transition should also be studied on patient treatment plans quality assurance.