国际肿瘤学杂志

国际肿瘤学杂志 ›› 2025, Vol. 52 ›› Issue (7): 409-413.doi: 10.3760/cma.j.cn371439-20241231-00071

• 论著 • 上一篇    下一篇

代谢组学评估肺结节恶性风险的临床价值

李晓萱, 夏志鹏, 栾如梅, 万云焱, 姚周虹, 林鑫山, 林殿杰()   

  1. 山东第一医科大学附属省立医院呼吸与危重症医学科，济南 250021
  • 收稿日期:2024-12-31 修回日期:2025-04-22 出版日期:2025-07-08 发布日期:2025-07-23
  • 通讯作者: 林殿杰 E-mail:dianjielin@126.com

Clinical value of metabolomics in assessing the malignant risk of pulmonary nodules

Li Xiaoxuan, Xia Zhipeng, Luan Rumei, Wan Yunyan, Yao Zhouhong, Lin Xinshan, Lin Dianjie()   

  1. Department of Respiratory and Critical Care Medicine，Shandong Provincial Hospital Affiliated to Shandong First Medical University，Jinan 250021，China
  • Received:2024-12-31 Revised:2025-04-22 Online:2025-07-08 Published:2025-07-23
  • Contact: Lin Dianjie E-mail:dianjielin@126.com

RichHTML

6

PDF

3

摘要：

目的 评估代谢指纹图谱非靶向检测在肺结节中的诊断价值，并对多组学评估肺结节恶性风险的临床有效模型进行分析。方法 选取2021年11月至2024年10月在山东第一医科大学附属省立医院接受胸部CT检查并完成病理学诊断、代谢指纹图谱非靶向检测的73例患者作为研究对象，根据术后组织病理学诊断不同将患者分为肺恶性结节组（61例）和肺良性结节组（12例）。收集患者一般临床资料，包括性别、年龄、吸烟史、肿瘤家族史等，影像学资料包括结节密度、结节大小、结节位置、结节数量、特殊影像学表现（毛刺征、分叶征、空泡征、血管集束征等），以及患者代谢指纹图谱非靶向检测结果。比较两组患者的上述资料；绘制受试者操作特征（ROC）曲线评估各模型的预测价值。结果 肺恶性结节组和肺良性结节组患者的年龄（t=4.41，P<0.001）、结节大小（Z=2.67，P=0.008）、结节密度（χ2=4.64，P=0.031）、毛刺征（χ2=7.67，P=0.006）比较，差异均有统计学意义；性别、吸烟史、肺癌家族史、结节数量、结节位置、分叶征、空泡征、血管集束征、胸膜凹陷征、钙化征、支气管截断征、血管供应征、支气管充气征比较，差异均无统计学意义（均P>0.05）。肺恶性结节组代谢指纹图谱非靶向检测风险评估高风险患者（36例）明显高于肺良性结节组（0例）（χ2=13.97，P<0.001）。ROC曲线分析显示，Brock模型联合代谢指纹图谱非靶向检测的曲线下面积为0.930（95%CI为0.872~0.988），大于Brock模型及代谢指纹图谱非靶向检测分别预测的0.856（95%CI为0.769~0.942，Z=0.27，P=0.040）、0.768（95%CI为0.650~0.887，Z=0.30，P=0.004）。结论 代谢指纹图谱非靶向检测风险评估在肺结节诊断中可能作为辅助Brock模型的无创方式，具有较好的应用价值。将Brock模型与代谢图谱非靶向检测联合，能更准确地鉴别肺结节的良恶性。

关键词: 肺肿瘤, 早期诊断, 代谢组学

Abstract:

Objective To evaluate the diagnostic value of non-targeted detection of metabolic fingerprinting in pulmonary nodules and to analyze the clinical effective model of multi-omics for assessing the malignant risk of pulmonary nodules. Methods A total of 73 patients who underwent chest CT and completed pathological diagnosis and non-targeted detection of metabolic fingerprinting at Shandong Provincial Hospital Affiliated to Shandong First Medical University from November 2021 to October 2024 were selected as the research subjects. According to the postoperative histopathological diagnosis，the patients were divided into the lung malignant nodule group （61 cases） and the lung benign nodule group （12 cases）. General clinical data of the patients，including sex，age，smoking history，and family history of tumors，as well as imaging data，including nodule density，nodule size，nodule location，nodule number，and special imaging manifestations （spiculation，lobulation，vacuole sign，vascular convergence sign，etc.），and non-targeted detection of metabolic fingerprinting results were collected. The above data were compared between the two groups of patients，and the receiver operator characteristic （ROC） curve was drawn to evaluate the predictive value of each model. Results There were statistically significant differences in age （t=4.41，P<0.001），nodule size （Z=2.67，P=0.008），nodule density （χ2=4.64，P=0.031），and spiculation （χ2=7.67，P=0.006） between the lung malignant nodule group and the lung benign nodule group. There were no statistically significant differences in sex，smoking history，family history of lung cancer，nodule number，nodule location，lobulation，vacuole sign，vascular convergence sign，pleural indentation sign，calcification sign，bronchial truncation sign，vascular supply sign，and bronchial air sign （all P>0.05）. The number of non-targeted detection of metabolic fingerprinting high-risk patients in the lung malignant nodule group （36 cases） was significantly higher than that in the lung benign nodule group （0 case）（χ2=13.97，P<0.001）. ROC curve analysis showed that the area under the curve of the Brock model combined with non-targeted detection of metabolic fingerprinting was 0.930 （95%CI： 0.872-0.988），which was greater than that of the Brock model （0.856，95%CI： 0.769-0.942，Z=0.27，P=0.040） and non-targeted detection of metabolic fingerprinting （0.768，95%CI： 0.650-0.887，Z=0.30，P=0.004） alone. Conclusions Non-targeted detection of metabolic fingerprinting risk assessment may serve as a non-invasive method to assist the Brock model in the diagnosis of pulmonary nodules and has good application value. The combination of the Brock model and non-targeted detection of metabolic fingerprinting can more accurately distinguish the benign and malignant nature of pulmonary nodules.

Key words: Lung neoplasms, Early diagnosis, Metabolomics