摘要
目的建立^(18)F-FDG PET/CT肿瘤代谢异质性多参数联合模型, 探讨其对肺部病变良恶性的鉴别价值。方法回顾性分析2017年2月至2024年2月间于南京大学医学院附属鼓楼医院经^(18)F-FDG PET/CT诊断为肺恶性病变且明确病理的251例患者(男157例、女94例, 年龄15~88岁), 分析其临床资料、原发灶^(18)F-FDG PET/CT传统特征[SUV_(max)、肿瘤代谢体积(MTV)、病灶糖酵解总量(TLG)]及原发灶PET肿瘤内代谢异质性指数[HI;如累积SUV体积直方图AUC(AUC-CSH)、线性回归斜率、CV]。分别采用SUV 2.5和_(40%)SUV_(max)计算AUC-CSH、CV。采用logistic单因素及多因素回归分析提取临床特征、PET/CT参数在肺部病变良恶性鉴别诊断中的独立预测因子, 并建立多参数联合模型, 采用ROC曲线验证模型的诊断效能。结果 251例患者中, 良性101例、恶性150例。单因素分析中, 性别、年龄、肿瘤标志物、毛刺征、分叶征、血管集束征、空泡征、长径、短径、SUV_(max)、AUC-CSH2.5、AUC-CSH_(40%)、CV_(2.5)、CV_(40%)[比值比(OR):0.57~17.39, 均P<0.05]为良恶性肿瘤诊断的预测因子。多因素分析中, 性别、年龄、肿瘤标志物、分叶征、血管集束征、SUV_(max)、AUC-CSH_(40%)、CV_(40%)为良恶性肿瘤诊断的独立预测因子(OR:2.30~13.18, 均P<0.05)。联合上述独立预测因子建立的多参数联合模型的AUC、灵敏度、特异性和准确性分别为0.89、77.33%(116/150)、84.16%(85/101)和80.08%(201/251)。结论 ^(18)F-FDG PET/CT多参数联合模型在肺部病变良恶性鉴别中具有较高的价值。
Objective:To establish a combined model of tumor heterogeneity metabolic parameters using ^(18)F-FDG PET/CT and explore its value in differentiating benign from malignant pulmonary lesions.Methods:A total of 251 patients(157 males,94 females;age 15-88 years)who were diagnosed with malignant lung lesions by ^(18)F-FDG PET/CT and with definitive pathological results at Nanjing Drum Tower Hospital,the Affiliated Hospital of Nanjing University Medical School from February 2017 to February 2024 were retrospectively enrolled.Analysis was conducted on clinical data,traditional parameters(SUV_(max),metabolic tumor volume(MTV),total lesion glycolysis(TLG))of primary lesions on ^(18)F-FDG PET/CT,and intra-tumoral metabolic heterogeneity index(HI;such as cumulative SUV volume histogram AUC(AUC-CSH),linear regression slope,CV).AUC-CSH and CV were calculated using SUV thresholds of 2.5 and _(40%)SUV_(max).Logistic univariate and multivariate regression analyses were used to extract independent predictors in clinical features and PET/CT parameters for the differential diagnosis of pulmonary lesions.A multi-parameter combined model was established through logistic regression and validated for diagnostic efficacy using ROC curve analysis.Results:Among 251 patients,101 were benign and 150 were malignant.In univariate analysis,gender,age,tumor markers,spiculation sign,lobulation sign,vessel convergence sign,air bronchogram,long diameter,short diameter,SUV_(max),AUC-CSH 2.5,AUC-CSH _(40%),CV_(2.5),and CV_(40%)were predictive factors for the diagnosis of benign and malignant tumors(odds ratio(OR):0.57-17.39,all P<0.05).In multivariate analysis,gender,age,tumor markers,lobulation sign,vessel convergence sign,SUV_(max),AUC-CSH _(40%),and CV_(40%)were independent predictors for the diagnosis of benign and malignant tumors(OR:2.30-13.18,all P<0.05).The AUC,sensitivity,specificity,and accuracy of the multi-parameter combined model established with the above independent predictors were 0.89,77.33%(116/150),84.16%(85/101),80.08%(201/251),respectively.Conclusion:^(18)F-FDG PET/CT multi-parameter combined model has high value in the differentiation of benign and malignant pulmonary lesions.
作者
来瑞鹤
耿羽智
何健
盛丹丹
Lai Ruihe;Geng Yuzhi;He Jian;Sheng Dandan(Department of Nuclear Medicine,Nanjing Drum Tower Hospital,the Affiliated Hospital of Nanjing University Medical School,Nanjing 210008,China;Department of Nuclear Medicine,the Second Affiliated Hospital of Nanjing Medical University,Nanjing 210011,China)
出处
《中华核医学与分子影像杂志》
北大核心
2025年第9期525-529,共5页
Chinese Journal of Nuclear Medicine and Molecular Imaging
基金
南京市卫生科技发展专项资金(YKK24090)。
