107例NSCLC患者放疗后≥2级放射性肺炎预测模型的建立和分析  

Establishment and analysis of a disease forecast model of ≥2 grade radiation pneumonitis based on 107 cases of post-radiotherapeutic NSCLC

作　　者：刘志坤[1] 苏景伟[1] 张魏丽[2] 崔彦莉[3] 李娟[1] 祝淑钗[1] Liu Zhikun Su Jingwei Zhang Weili et al(The Third Ward of Radiotherapy Department, Fourth Hospital, Hebei Medical University, Shijiazhuang 050011, China School of Public Health, Hebei Medical University, Shijiazhuang 050017, China)

机构地区：[1]河北医科大学第四医院放疗科三病区,河北石家庄050011 [2]河北医科大学公共卫生学院,河北石家庄050017 [3]邢台市人民医院放疗科,河北邢台054000

出　　处：《实用肿瘤杂志》2017年第3期239-243,共5页Journal of Practical Oncology

摘　　要：目的探讨非小细胞肺癌三维适形放疗后放射性肺炎发生的相关因素并建立数学预测模型。方法收集行三维适形放疗的非小细胞肺癌患者107例。全组患者均为根治性放疗,剂量采用常规放疗,分割方式为2Gy/f,处方剂量60~78 Gy,中位剂量66 Gy。不同组别患者放射性肺炎的发生情况单因素分析采用χ2检验。Logistic回归分析筛选影响放射性肺炎发生的独立预后因素。受试者工作特征(receiver operating characteristic,ROC)曲线分析评价其临床诊断性能。结果本组患者放射性肺炎发生率为62.6%,≥2级放射性肺炎的发生率为38.3%,其中2级23例(21.5%),3级14例(13.1%),4级4例(3.7%)。单因素分析结果显示,放射性肺炎的发生在慢性阻塞性肺疾病、T分期、射野数目、临床靶区(clinical target volume,CTV)的体积、CTV的平均剂量、计划靶区(planning target volume,PTV)体积、PTV的平均剂量、双肺体积和双肺Dmean、V5、V10、V15、V20、V25、V30、V35、V40方面差异均具有统计学意义(均P<0.05)。多因素分析显示,T分期、双肺Dmean、V20、V40为影响≥2级放射性肺炎发生的独立因素(均P<0.05)。在此基础之上,建立放射性肺炎的预测模型为Y=ex/(1+ex),其中x=-5.797-0.986×T分期+1.193×肺平均剂量+1.259×V20+1.329×V40。结论 T分期、双肺Dmean、V20和V40为影响接受三维适形放疗的非小细胞肺癌患者发生≥2级放射性肺炎发生的独立因素,建立的数学预测模型对这类患者≥2级放射性肺炎的发生有较好的预测价值。Objective To explore the related factors for the development of radiation pneumonitis in non-small-ceil lung cancer （NSCLC） patients after three-dimensional conformal radiotherapy （3D-CRT） treatment, and establish a mathe- matical prediction model. Methods One-hundred-and-seven NSCLC patients who have received 3D-CRT were enrolled. All patients received the biological effective dose （BED） between 60 - 78 Gy （ median dose = 66 Gy） and 2 Gy/F. Univa- riate factor analysis was examined byx^2 test for radiation pneumonitis patients of different groups. Independent forecast fac- tors for radiation pneumonitis were analyzed by logistic regression model. The clinic efficacy was analyzed by receiver oper- ating characteristic （ROC） curve. Results In the 107 patients, the rate of radiation pneumonitis was 62.6% , and the rate of 〉12 grade radiation pneumonitis was 38.3%. Among them, 23 cases （21.5%） were grade 2 radiation pneumoni- tis, 14 cases （ 13.1% ） were grade 3, and 4 cases （3.7%） were grade 4. Univariate analysis showed multiple significant influence factors for radiation pneumonitis, including chronic obstructive pulmonary disease （COPD） , T stage, number of radiation fields, clinical target volume （CTV） and its mean dose, planning target volmne （PTV） and its mean dose, bilat- eral lung volume and its mean dose, lung V5 - V40 （ P 〈 0.05 ）. Moreover, nmltivariate analysis suggested T stage, the mean dose of bilateral lungs, V20 and V40 as independent factors of radiation pneumonitis （ P 〈 0.05 ）. According to these data, we established the forecast model for radiation pneumonitis [ Y = e^x/（ 1 ＋ e^x） ,x = -5. 797 - 0. 986 × T stage ＋ 1. 193 × the mean dose of bilateral lungs ＋ 1. 259 × V20 ＋ 1. 329 × V401- Conclusion T stage, the mean dose of bilateral lungs, V20 and V40 are the independent factors of ≥2 grade radiation pneumonitis in NSCLC patients after 3D-CRT treatment. The forecast model has good prediction value for the occurrence of

关 键 词：癌 非小细胞肺/放射疗法 放射疗法 适形 肺肿瘤/放射疗法 辐射性肺炎/病因学 LOGISTIC模型 R0c曲线 

分 类 号：R730.55[医药卫生—肿瘤] R734.2[医药卫生—临床医学]