为了非侵入性地检测现场致癌过程，我们进行了呼气气溶胶微小RNA的早期技术开发和病例对照测试。在设计上，人类肺组织微小RNA-seq发现与TCGA和已发表的肿瘤识别微小RNA相协调，得出了一个由24个上调的微小RNA组成的面板。利用配对的EBC、上部和下部气道供体样本集，对呼出微小RNA的空气来源进行了地形“指纹”识别。通过定性RT-PCR对166例非小细胞肺癌病例和185例对照进行微小RNA面板的诊断性测试。数据采用逻辑回归(LR)和随机森林(RF)模型进行分析。进行了呼气微小RNA检测的可行性测试，包括优化的整体EBC提取，以及RT和定性PCR方法评估。在这种低模板设置下，基于内插剂的URT-PCR优于荧光探针PCR(TaqMan)。在应用中，调整后的逻辑回归模型确定了呼出miR-21、33b、212作为总体病例对照鉴别标志。联合临床+微小RNA模型的RF分析显示出了适度的额外鉴别能力(1.1-2.5%)，超出了单独临床模型：所有受试者为1.1%(p = 8.7e-04)；以前的吸烟者为2.5%(p = 3.6e-05)；早期阶段为1.2%(p = 9.0e-03)，使ROC AUC从0.74到0.83不等。我们得出结论，呼出微小RNA可以在定性上测量，部分反映下部气道特征；当这种技术得到进一步改进和量化时，可能有助于改善肺癌风险评估。©2023年，作者。

For detecting field carcinogenesis non-invasively, early technical development and case-control testing of exhaled breath condensate microRNAs was performed. In design, human lung tissue microRNA-seq discovery was reconciled with TCGA and published tumor-discriminant microRNAs, yielding a panel of 24 upregulated microRNAs. The airway origin of exhaled microRNAs was topographically "fingerprinted", using paired EBC, upper and lower airway donor sample sets. A clinic-based case-control study (166 NSCLC cases, 185 controls) was interrogated with the microRNA panel by qualitative RT-PCR. Data were analyzed by logistic regression (LR), and by random-forest (RF) models. Feasibility testing of exhaled microRNA detection, including optimized whole EBC extraction, and RT and qualitative PCR method evaluation, was performed. For sensitivity in this low template setting, intercalating dye-based URT-PCR was superior to fluorescent probe-based PCR (TaqMan). In application, adjusted logistic regression models identified exhaled miR-21, 33b, 212 as overall case-control discriminant. RF analysis of combined clinical + microRNA models showed modest added discrimination capacity (1.1-2.5%) beyond clinical models alone: all subjects 1.1% (p = 8.7e-04)); former smokers 2.5% (p = 3.6e-05); early stage 1.2% (p = 9.0e-03), yielding combined ROC AUC ranging from 0.74 to 0.83. We conclude that exhaled microRNAs are qualitatively measureable, reflect in part lower airway signatures; and when further refined/quantitated, can potentially help to improve lung cancer risk assessment.© 2023. The Author(s).