Padmanaban Sriram, Warren Samantha, Walsh Anthony, Partridge Mike, Hawkins Maria A
Oxford Cancer Centre, Oxford University Hospitals, Oxford, OX3 7LE, UK.
CRUK/MRC Oxford Institute for Radiation Oncology, Gray Laboratories, University of Oxford, Oxford, OX3 7DQ, UK.
Radiat Oncol. 2014 Dec 23;9:286. doi: 10.1186/s13014-014-0286-3.
To investigate systematic changes in dose arising when treatment plans optimised using the Anisotropic Analytical Algorithm (AAA) are recalculated using Acuros XB (AXB) in patients treated with definitive chemoradiotherapy (dCRT) for locally advanced oesophageal cancers.
We have compared treatment plans created using AAA with those recalculated using AXB. Although the Anisotropic Analytical Algorithm (AAA) is currently more widely used in clinical routine, Acuros XB (AXB) has been shown to more accurately calculate the dose distribution, particularly in heterogeneous regions. Studies to predict clinical outcome should be based on modelling the dose delivered to the patient as accurately as possible.
CT datasets from ten patients were selected for this retrospective study. VMAT (Volumetric modulated arc therapy) plans with 2 arcs, collimator rotation ± 5-10° and dose prescription 50 Gy / 25 fractions were created using Varian Eclipse (v10.0). The initial dose calculation was performed with AAA, and AXB plans were created by re-calculating the dose distribution using the same number of monitor units (MU) and multileaf collimator (MLC) files as the original plan. The difference in calculated dose to organs at risk (OAR) was compared using dose-volume histogram (DVH) statistics and p values were calculated using the Wilcoxon signed rank test. The potential clinical effect of dosimetric differences in the gross tumour volume (GTV) was evaluated using three different TCP models from the literature.
PTV Median dose was apparently 0.9 Gy lower (range: 0.5 Gy - 1.3 Gy; p < 0.05) for VMAT AAA plans re-calculated with AXB and GTV mean dose was reduced by on average 1.0 Gy (0.3 Gy -1.5 Gy; p < 0.05). An apparent difference in TCP of between 1.2% and 3.1% was found depending on the choice of TCP model. OAR mean dose was lower in the AXB recalculated plan than the AAA plan (on average, dose reduction: lung 1.7%, heart 2.4%). Similar trends were seen for CRT plans.
Differences in dose distribution are observed with VMAT and CRT plans recalculated with AXB particularly within soft tissue at the tumour/lung interface, where AXB has been shown to more accurately represent the true dose distribution. AAA apparently overestimates dose, particularly the PTV median dose and GTV mean dose, which could result in a difference in TCP model parameters that reaches clinical significance.
探讨在局部晚期食管癌根治性放化疗（dCRT）患者中，当使用各向异性分析算法（AAA）优化的治疗计划改用Acuros XB（AXB）重新计算时剂量的系统性变化。
我们已将使用AAA创建的治疗计划与使用AXB重新计算的计划进行了比较。尽管各向异性分析算法（AAA）目前在临床常规中应用更为广泛，但Acuros XB（AXB）已被证明能更准确地计算剂量分布，尤其是在异质区域。预测临床结果的研究应基于尽可能精确地模拟给予患者的剂量。
本回顾性研究选取了10例患者的CT数据集。使用瓦里安医科达（Varian Eclipse，v10.0）创建了具有2个弧、准直器旋转±5 - 10°且剂量处方为50 Gy / 25次分割的容积调强弧形放疗（VMAT）计划。初始剂量计算采用AAA进行，通过使用与原始计划相同数量的监测单位（MU）和多叶准直器（MLC）文件重新计算剂量分布来创建AXB计划。使用剂量体积直方图（DVH）统计比较危及器官（OAR）计算剂量的差异，并使用Wilcoxon符号秩检验计算p值。使用文献中的三种不同肿瘤控制概率（TCP）模型评估大体肿瘤体积（GTV）剂量差异的潜在临床影响。
用AXB重新计算的VMAT AAA计划的计划靶体积（PTV）中位剂量明显低0.9 Gy（范围：0.5 Gy - 1.3 Gy；p < 0.05），GTV平均剂量平均降低1.0 Gy（0.3 Gy - 1.5 Gy；p < 0.05）。根据TCP模型的选择，发现TCP存在1.2%至3.1%的明显差异。AXB重新计算的计划中OAR平均剂量低于AAA计划（平均剂量降低：肺1.7%，心脏2.4%）。同步放化疗（CRT）计划也观察到类似趋势。
使用AXB重新计算VMAT和CRT计划时，观察到剂量分布存在差异，特别是在肿瘤/肺界面的软组织内，其中AXB已被证明能更准确地表示真实剂量分布。AAA明显高估了剂量，尤其是PTV中位剂量和GTV平均剂量，这可能导致TCP模型参数的差异达到临床意义。
