Tucker Susan L, Liu H Helen, Wang Shulian, Wei Xiong, Liao Zhongxing, Komaki Ritsuko, Cox James D, Mohan Radhe
Department of Biostatistics and Applied Mathematics, The University of Texas M. D. Anderson Cancer Center, Houston, TX 77030, USA.
Int J Radiat Oncol Biol Phys. 2006 Nov 1;66(3):754-61. doi: 10.1016/j.ijrobp.2006.06.002. Epub 2006 Sep 11.
The aim of this study was to investigate the effect of radiation dose distribution in the lung on the risk of postoperative pulmonary complications among esophageal cancer patients.
We analyzed data from 110 patients with esophageal cancer treated with concurrent chemoradiotherapy followed by surgery at our institution from 1998 to 2003. The endpoint for analysis was postsurgical pneumonia or acute respiratory distress syndrome. Dose-volume histograms (DVHs) and dose-mass histograms (DMHs) for the whole lung were used to fit normal-tissue complication probability (NTCP) models, and the quality of fits were compared using bootstrap analysis.
Normal-tissue complication probability modeling identified that the risk of postoperative pulmonary complications was most significantly associated with small absolute volumes of lung spared from doses > or = 5 Gy (VS5), that is, exposed to doses < 5 Gy. However, bootstrap analysis found no significant difference between the quality of this model and fits based on other dosimetric parameters, including mean lung dose, effective dose, and relative volume of lung receiving > or = 5 Gy, probably because of correlations among these factors. The choice of DVH vs. DMH or the use of fractionation correction did not significantly affect the results of the NTCP modeling. The parameter values estimated for the Lyman NTCP model were as follows (with 95% confidence intervals in parentheses): n = 1.85 (0.04, infinity), m = 0.55 (0.22, 1.02), and D50 = 17.5 Gy (9.4 Gy, 102 Gy).
In this cohort of esophageal cancer patients, several dosimetric parameters including mean lung dose, effective dose, and absolute volume of lung receiving < 5 Gy provided similar descriptions of the risk of postoperative pulmonary complications as a function of the radiation dose distribution in the lung.
本研究旨在探讨肺癌放疗剂量分布对食管癌患者术后肺部并发症风险的影响。
我们分析了1998年至2003年在我院接受同步放化疗后行手术治疗的110例食管癌患者的数据。分析的终点是术后肺炎或急性呼吸窘迫综合征。使用全肺的剂量体积直方图（DVH）和剂量质量直方图（DMH）来拟合正常组织并发症概率（NTCP）模型，并通过自助分析比较拟合质量。
正常组织并发症概率模型确定，术后肺部并发症的风险与接受≥5 Gy剂量的肺的小绝对体积（VS5）最显著相关，即接受<5 Gy剂量的肺。然而，自助分析发现该模型与基于其他剂量学参数（包括平均肺剂量、有效剂量和接受≥5 Gy剂量的肺的相对体积）的拟合质量之间没有显著差异，可能是因为这些因素之间存在相关性。DVH与DMH的选择或分割校正的使用对NTCP建模结果没有显著影响。Lyman NTCP模型估计的参数值如下（括号内为95%置信区间）：n = 1.85（0.04，无穷大），m = 0.55（0.22，1.02），D50 = 17.5 Gy（9.4 Gy，102 Gy）。
在这组食管癌患者中，包括平均肺剂量、有效剂量和接受<5 Gy剂量的肺的绝对体积在内的几个剂量学参数，对术后肺部并发症风险的描述与肺内放疗剂量分布的函数关系相似。
