摘要
目的探索肺腺癌(ADC)患者原发灶葡萄糖代谢半定量参数以及一系列其他临床指标与表皮生长因子受体(EGFR)基因突变状态的关联性,建立非侵袭地评估EGFR突变状态的预测模型。方法收集2011年12月至2018年4月在北京大学第一医院核医学科PET/CT中心进行18F-氟代脱氧葡萄糖(18F-FDG)PET/CT全身显像的初发ADC患者192例为研究对象。进行术前18F-FDG PET/CT、胸部薄层CT、术前血清肿瘤标志物检测及EGFR基因分析。采用SPSS 25.0软件进行t检验、Mann-Whitney检验和χ^2检验。采用R 3.5.0软件通过分类决策树(CART)构建EGFR突变状态的预测模型,采用Bootstrap法1000次置换检验验证模型预测效能。结果初发肺腺癌患者的EGFR突变率为33.3%(64/192)。CART分析结果显示,年龄、吸烟状态、平均标准化摄取值(SUVmean)、原发灶肿瘤代谢体积(pMTV)、原发灶总糖酵解量(pTLG)、癌胚抗原、鳞状细胞癌抗原、神经元特异性烯醇化酶、组织多肽抗原和胃泌素释放肽前体是EGFR突变的独立显著预测因子;曲线下面积为0.785(95%CI:0.743~0.827),敏感性和特异性分别为85.5%和71.4%。结论原发灶高水平葡萄糖代谢半定量参数SUVmean、pMTV和pTLG对ADC患者EGFR突变具有独立的预测价值,联合其他临床参数构建的预测模型能有效、非侵袭性地对EGFR突变进行预估判断。
Objective To explore the relationship between epidermal growth factor receptor(EGFR)gene mutation status and semi quantitative parameters of glucose metabolism in primary focus or other clinical indexes in lung adenocarcinoma(ADC)patients,and to establish the noninvasive evaluating prediction model of EGFR gene mutation status.Methods From December of 2011 to April of 2018,192 patients with primary ADC undergoing 18F-fluorodeoxyglucose(18F-FDG)PET/CT whole body imaging in nuclear medicine PET/CT center of the Peking University First Hospital were selected as the subjects.The 18F-FDG PET/CT,thin slice CT of chest,detection of serum tumor markers and EGFR gene analysis before operation were performed.The t test,Mann-Whitney test andχ^2 test were used to analyze the data,the used software was SPSS 25.0.The CART analysis was used to construct the prediction model of EGFR gene mutation status,1000-time Bootstrap method was used to validate the prediction efficiency of model,the used software was R 3.5.0.Results The EGFR mutation rate in patients with primary ADC was 33.3%(64/192).CART analysis showed that age,smoking status,average of standardized uptake valuer(SUVmean),primary metabolic tumor volume(pMTV),primary total lesion glycolysis(pTLG),carcinoembryonic antigen,squamous cell carcinoma antigen,neuron-specific enolas,tissue polypeptide antigen and precursor of gastrin releasing peptide could be independent and significant predication factors of EGFR mutation;the area under the curve was 0.785(95%CI:0.743-0.827),the sensitivity and specificity were 85.5%and 71.4%,respectively.Conclusion The semi quantitative parameters(SUVmean,pMTV and pTLG)have the independent prediction value for EGFR mutation in ADC patients,which combine with other clinical parameters could construct the prediction model for effective and noninvasive prediction judgement of EGFR mutation.
作者
廖栩鹤
陈思
王荣福
刘萌
陈雪祺
殷雷
陈大方
LIAO Xu-he;CHEN Si;WANG Rong-fu;LIU Meng;CHEN Xue-qi;YIN Lei;CHEN Da-fang(Department of Nuclear Medicine,Peking University First Hospital,Beijing 100034,China)
出处
《中国慢性病预防与控制》
CAS
CSCD
北大核心
2020年第3期176-181,共6页
Chinese Journal of Prevention and Control of Chronic Diseases
基金
北京大学第一医院自然科学基金项目(2017QN14)。
关键词
表皮生长因子受体
最大标准化摄取值
平均标准化摄取值
肿瘤代谢体积
总糖酵解量
分类决策树
肺腺癌
Epidermal growth factor receptor
Maximum standardized uptake value
Average of standardized uptake value
Metabolic tumor volume
Total lesion glycolysis
Classification decision tree
Lung adenocarcinoma
