Dosimetric Evaluation and Comparison of Collapsed Cone Algorithm with AAA and AcurosXB Algorithms for Lung Cancer Treatment Planning on the Varian Halcyon Accelerator

Abstract

Abstract Objective: This study aims to compare the dosimetric calculationsin standard non-small cell lung cancer (NSCLC) radiotherapy planningusing the Collapsed Cone (CC) algorithm of the RayStation planningsystem on the Varian Halcyon accelerator, with the Analytical AnisotropicAlgorithm (AAA) and Acuros XB algorithms on the Eclipse accelerator. Methods: The RayStation planning system was employed to designthe RapidArc radiotherapy plans for 15 lung cancer patients on theHalcyon accelerator, using the CC algorithm for dose calculation.The plans were then transferred to the Eclipse planning system. Withoutmaking any changes to these plans, doses were computed using bothAAA and Acuros XB algorithms. The differences in dose-volume indicesfor the clinical target volume (CTV), planned target volume (PTV),and other organs at risk were evaluated. Results:Compared with the CC algorithm, the AAA algorithmshowed slightly lower dose-volume indices for both the CTV (D2%,D50%, D95%, D98%, and mean dose) and the PTV (D2%, D50%, andmean dose), with differences within 1%. Specifically, the PTV's D95%and D98% were respectively 1.7% and 2.4% lower, the heart's meandose was 3.2% lower, the lung's V20Gy was 0.2% lower, the averagedose was 0.2% higher, and the spinal cord's maximum dose (representedby D0.1cc) was 0.9% lower. In contrast to the CC algorithm, the AcurosXB algorithm also showed slightly lower dose-volume indices for theCTV and the PTV, with differences within 1.5%. Specifically, thePTV's D95% and D98% were respectively 1.6% and 2.3% lower, theheart's mean dose was 4.4% lower, the lung's V20Gy was 0.3% lower,the average dose was 1.4% lower, and the spinal cord's maximum dosewas 2.3% lower. Paired t-tests indicated that all these results hadsignificant differences (p<0.05). Conclusion: This study aimed to quantify the impact of usingRayStation system and CC algorithm on Halcyon accelerator for clinicaldose evaluations, by comparing it with the AAA and Acuros XB algorithmsof the Varian Eclipse planning system. For lung cancer radiotherapyplanning, dose calculations obtained from the AAA and Acuros XB algorithmsunder the Eclipse system were slightly lower compared to the RayStationCC algorithm across various clinical indices. Notably, the differencesin the PTV's D95% and D98% indices were over 1.5%, while dose indicesfor various organs at risk were approximately 2∼3% lower. The results of this study can serve as a reference for institutionsthat are preparing to use the RayStation planning system for the designand clinical application of plans on the Varian Halcyon accelerator.