Dosimetric accuracy and clinical quality of Acuros XB and AAA dose calculation algorithm for stereotactic and conventional lung volumetric modulated arc therapy plans

Radiation Oncology Dosimetric accuracy and clinical quality of Acuros XB and AAA dose calculation algorithm for stereotactic and conventional lung volumetric modulated arc therapy plans Petra S Kroon 0 Sandra Hol 1 Marion Essers 0 0 Department of Medical Physics, Institute Verbeeten , Brugstraat 10, 5042 SB Tilburg , the Netherlands 1 Department of Radiotherapy, Institute Verbeeten , Brugstraat 10, 5042 SB, Tilburg , the Netherlands Introduction: The main aim of the current study was to assess the dosimetric accuracy and clinical quality of volumetric modulated arc therapy (VMAT) plans for stereotactic (stage I) and conventional (stage III) lung cancer treatments planned with Eclipse version 10.0 Anisotropic Analytical Algorithm (AAA) and Acuros XB (AXB) algorithm. Methods: The dosimetric impact of using AAA instead of AXB, and grid size 2.5 mm instead of 1.0 mm for VMAT treatment plans was evaluated. The clinical plan quality of AXB VMAT was assessed using 45 stage I and 73 stage III patients, and was compared with published results, planned with VMAT and hybrid-VMAT techniques. Results: The dosimetric impact on near-minimum PTV dose (D98%) using AAA instead of AXB was large (underdose up to 12.3%) for stage I and very small (underdose up to 0.8%) for stage III lung treatments. There were no significant differences for dose volume histogram (DVH) values between grid sizes. The calculation time was significantly higher for AXB grid size 1.0 than 2.5 mm (p < 0.01). The clinical quality of the VMAT plans was at least comparable with clinical qualities given in literature of lung treatment plans with VMAT and hybrid-VMAT techniques. The average mean lung dose (MLD), lung V20Gy and V5Gy in this study were respectively 3.6 Gy, 4.1% and 15.7% for 45 stage I patients and 12.4 Gy, 19.3% and 46.6% for 73 stage III lung patients. The average contra-lateral lung dose V5Gy-cont was 35.6% for stage III patients. Conclusions: For stereotactic and conventional lung treatments, VMAT calculated with AXB grid size 2.5 mm resulted in accurate dose calculations. No hybrid technique was needed to obtain the dose constraints. AXB is recommended instead of AAA for avoiding serious overestimation of the minimum target doses compared to the actual delivered dose. AAA; Acuros XB; Plan quality; Stage I lung cancer; Stage III lung cancer; VMAT - Background Volumetric modulated arc therapy (VMAT) has been shown to be a powerful technique for irradiation of many treatment sites with obtaining higher dose conformity to the tumor while decreasing intra-fraction movements because of shorter delivery times [1-8]. Reliable and accurate dose delivery can be obtained using VMAT as shown by pre-treatment dosimetric plan validations [9]. VMAT could achieve at least comparable clinical plan qualities and skin dose levels than intensity modulated radiotherapy (IMRT) [10-12] and can successfully be used for stereotactic body radiotherapy (SBRT) for patients with stage I Non-Small-Cell Lung Cancer (NSCLC) [2,3,5]. In case of stage III large tumor lung cancers, it is difficult to limit doses to organs at risks (OARs) such as heart and lung. High doses are preferred since local control increased significantly (p = 0.02) when patients are treated with higher doses than 64 Gy [13]. Rengan et al. [13] stated that the median survival time for patients treated to 64 Gy or higher was 20 months versus 15 months for those treated to less than 64 Gy. Advanced planning techniques, IMRT and VMAT, have been shown to be able to increase the therapeutic dose with equal toxicity profiles compared to three-dimensional conformal radiotherapy (3DCFRT) [14]. Unfortunately, it is not always possible to deliver doses higher than 60 Gy to the planning target volume (PTV) using 3DCFRT, IMRT and VMAT because of dose limiting organs [14]. De Bree-Balk et al. [14] stated that possibly further improvements could be made by using hybrid techniques which combine two static fields with IMRT or VMAT, as also investigated by Verbakel et al. [15], who made a comparison between conventional static field plans, IMRT, hybrid-IMRT, VMAT and hybridVMAT; and by Chan et al. [16], who compared between 3DCFRT, VMAT and hybrid-VMAT. In both studies the VMAT plans consisted of at least 2 arcs and the hybrid plans of a combination of two static fields and IMRT or VMAT. They have concluded that hybrid techniques are superior in dosimetric outcomes for treating stage III lung tumours compared to the other techniques. The treatment planning for these studies has been performed using Varian Eclipse treatment planning version 8 or 10 with Anisotropic Analytical Algorithm (AAA). We have recently introduced the Varian Eclipse software version 10, with the AAA as well as the Acuros XB (AXB) algorithm for photon dose calculations in our institute. AXB solves the linear Boltzmann transport equation e.g., [17]. The dosimetric accuracy of AXB has already been investigated in several studies [18-22]. Fogliata et al. [19,21] have concluded that AXB gives acceptable characteristics in homogeneous media for small and large fields (range 0.80.8 to 40.040.0 cm2) using comparisons of AXB with AAA and measurements. In heterogeneous situations, the AXB algorithm has been shown to provide a valid and accurate alternative to Monte Carlo calculations for field sizes ranging from 2.52.5 to 30.030.0 cm2 [18,20,22]. Immediately after clinical introduction of the Varian Eclipse software, we also clinically introduced VMAT for lung SBRT stage I NSCLC as well as for lung stage III treatments. For this clinical introduction, we investigated the dosimetric accuracy and quality of stereotactic and conventional VMAT planning in Eclipse using AXB and AAA. Routinely, for all our patients, we perform pretreatment verification measurements using an ionization chamber in the isocentre, combined with film measurements in the isocentre plane. It was already shown by Gete et al. [23] that AAA calculations can slightly overestimate the minimum PTV dose relative to Monte Carlo calculations with BEAMnrc/ DOSXYZnrc for stage I lung tumors (PTV range between 3 19 to 62 cm ) with forward planning with multiple static non-coplanar conformal fields. It has also been shown by VMAT comparison studies that AXB leads to a slightly more accurate dose distribution than AAA [24,25]. For stage III lung tumors (average PTV 690 cm3), Fogliata et al. [26] have illustrated that AAA leads to a monitor unit underestimation of approximately 1-2% relative to AXB grid size 2.5 mm using a treatment planning comparison between 3DCFRT, IMRT and VMAT. Consequently, this suggests that AAA could overestimate the minimum target dose, which leads to lower target coverage than the prescribed dose, in case AXB represents the real dose distribution. Kan et al. [27] have illustrated that AXB was more accurate in predicting secondary build-up near and beyond air/tissue interfaces than AAA, using a comparison with measurem (...truncated)