Verification and dosimetric impact of Acuros XB algorithm on intensity modulated stereotactic radiotherapy for locally persistent nasopharyngeal carcinoma

Purpose: The main aim of the current study was to assess the dosimetric impact on intensity modulated stereotactic radiotherapy (IMSRT) for locally persistent nasopharyngeal carcinoma (NPC) due to the recalculation from the Anisotropic Analytical Algorithm (AAA) to the recently released Acuros XB (AXB) algorithm. The dosimetric accuracy of using AXB in predicting air/tissue interface doses from an open single small field in a simple geometric phantom and intensity modulated small fields in an anthropomorphic phantom was also investigated. Methods: The central axis percentage depth doses (PDD) of a rectangular phantom containing an air cavity were calculated by both AAA and AXB from 6 MV beam with small field sizes (2 x 2 to 5 x 5 cm(2)). These data were compared to PDD measured by thin thermoluminescent dosimeters (TLDs) and Monte Carlo simulations. The doses predicted by AAA and AXB near air/tissue interfaces from five different IMSRT plans were compared to the TLD measured doses in an anthropomorphic phantom. The PTV coverage, conformity and doses to organs at risk (OARs) calculated by AAA and AXB were compared for 12 patients, using identical beam setup, leaves movement and monitor units. Results: Testing using the simple rectangular phantom demonstrated that AAA and AXB overestimated the PDD at the air/tissue interfaces by up to 41% and 6%, respectively, from a 2 x 2 cm(2) field. The secondary build-up curves predicted by AXB caught up well with the measured data at around 2 mm beyond the air cavity. Testing using the anthropomorphic phantom showed that AAA overestimated the doses by up to 10%, while the measured doses matched those of the AXB to within 3%. Using AAA, the planning target coverage represented by 100% of the reference dose was estimated to be 4% higher than using AXB. The averaged minimum dose to the PTV predicted by AAA was about 4% higher and OARs doses 3% to 6% higher compared to AXB. Conclusions: AXB should be used whenever possible as the standard reference for IMSRT boost of NPC cases. The more accurate AXB indicating lower target coverage and lower minimum target dose compared to AAA should be noted. (C) 2012 American Association of Physicists in Medicine. [http://dx.doi.org/10.1118/1.4736819]