Gasparri Roberto, Capuano Rosamaria, Guaglio Alessandra, Caminiti Valentina, Canini Federico, Catini Alexandro, Sedda Giulia, Paolesse Roberto, Di Natale Corrado, Spaggiari Lorenzo
Division of Thoracic Surgery, European Institute of Oncology, Milan, Italy.
Department of Electronic Engineering, University of Rome Tor Vergata, Roma, Italy.
J Breath Res. 2022 Sep 2;16(4). doi: 10.1088/1752-7163/ac88ec.
Currently, in clinical practice there is a pressing need for potential biomarkers that can identify lung cancer at early stage before becoming symptomatic or detectable by conventional means. Several researchers have independently pointed out that the volatile organic compounds (VOCs) profile can be considered as a lung cancer fingerprint useful for diagnosis. In particular, 16% of volatiles contributing to the human volatilome are found in urine, which is therefore an ideal sample medium. Its analysis through non-invasive, relatively low-cost and straightforward techniques could offer great potential for the early diagnosis of lung cancer. In this study, urinary VOCs were analysed with a gas chromatography-ion mobility spectrometer (GC-IMS) and an electronic nose (e-nose) made by a matrix of twelve quartz microbalances complemented by a photoionization detector. This clinical prospective study involved 127 individuals, divided into two groups: 46 with lung cancer stage I-II-III confirmed by computerized tomography or positron emission tomography-imaging techniques and histology (biopsy), and 81 healthy controls. Both instruments provided a multivariate signal which, after being analysed by a machine learning algorithm, identified eight VOCs that could distinguish lung cancer patients from healthy ones. The eight VOCs are 2-pentanone, 2-hexenal, 2-hexen-1-ol, hept-4-en-2-ol, 2-heptanone, 3-octen-2-one, 4-methylpentanol, 4-methyl-octane. Results show that GC-IMS identifies lung cancer with respect to the control group with a diagnostic accuracy of 88%. Sensitivity resulted as being 85%, and specificity was 90%-Area Under the Receiver Operating Characteristics: 0.91. The contribution made by the e-nose was also important, even though the results were slightly less sensitive with an accuracy of 71.6%. Moreover, of the eight VOCs identified as potential biomarkers, five VOCs had a high sensitivity (⩽ 0.06) for early stage (stage I) lung cancer.
目前，在临床实践中，迫切需要能够在肺癌出现症状或通过传统方法可检测到之前的早期阶段识别肺癌的潜在生物标志物。几位研究人员独立指出，挥发性有机化合物（VOCs）谱可被视为有助于诊断的肺癌指纹图谱。特别是，在尿液中发现了占人类挥发物总量16%的挥发性物质，因此尿液是一种理想的样本介质。通过非侵入性、成本相对较低且操作简单的技术对其进行分析，可为肺癌的早期诊断提供巨大潜力。在本研究中，使用气相色谱-离子迁移谱仪（GC-IMS）和由十二个石英微量天平组成的矩阵并辅以光离子化检测器制成的电子鼻（e-nose）对尿液中的挥发性有机化合物进行了分析。这项临床前瞻性研究涉及127名个体，分为两组：46名经计算机断层扫描或正电子发射断层扫描成像技术及组织学（活检）确诊为I-II-III期肺癌的患者，以及81名健康对照者。两种仪器都提供了一个多变量信号，在通过机器学习算法进行分析后，识别出了八种能够区分肺癌患者和健康人的挥发性有机化合物。这八种挥发性有机化合物分别是2-戊酮、2-己烯醛、2-己烯-1-醇、4-庚烯-2-醇、2-庚酮、3-辛烯-2-酮、4-甲基戊醇、4-甲基辛烷。结果表明，与对照组相比，GC-IMS识别肺癌的诊断准确率为88%。灵敏度为85%，特异性为90%——受试者工作特征曲线下面积：0.91。电子鼻的贡献也很重要，尽管结果的敏感性略低，准确率为71.6%。此外，在被确定为潜在生物标志物的八种挥发性有机化合物中，有五种对早期（I期）肺癌具有高灵敏度（⩽0.06）。
