2016-08-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/650363摘要：研究背景:肺癌是世界癌症死因排名第一位的癌症。篩檢出早期肺癌可提供適當治療降低死亡率。發展非侵入性、高準確度之篩檢方法有其重要性。研究目的:發展以呼吸氣體偵測肺癌方法。研究原理:脂質過氧化在肺癌的致病機轉扮演重要角色。測量下呼吸道產生的肺泡氣體，可分析肺部病變所產生揮發性代謝產物。作業假說:肺癌會改變呼吸氣體揮發性有機化合物成份，脂質過氧化產生揮發性有機化合物可以電子鼻與氣相層析質譜儀分析，用於偵測肺癌。研究方法:我們設計三年期研究來驗證所提出的假說。第一年先以100 名肺癌手術病患進行先導性研究，驗證肺泡採樣技術與電子鼻的區辨能力。研究將由氣管插管進行肺泡氣體採樣以避免來自環境的汙染，以雙腔氣管導管分別蒐集患側與非患側肺泡氣體，再以電子鼻區辨患側與非患側揮發性有機化合物成份差異。第二年以病例對照研究設計，建立以電子鼻偵測肺癌預測模式。研究將納入150 名新診斷肺部腫瘤接受切除手術患者為病例組，以150 名年齡、性別配對之非腫瘤接受全身麻醉手術患者為對照組，以電子鼻分析肺泡氣體，將資料以2:1 比例分為訓練數據集與驗證數據集，分別用於肺癌預測模式建立、與肺癌預測模式準確性驗證。第三年將納入100 名個案組與100 名對照組，以氣相層析質譜儀分析肺泡氣體，尋找潛在肺癌揮發性分子生物偵測指標，並驗證以脂質過氧化原理偵測肺癌的假說。研究將分析肺泡氣體冷凝液脂質過氧化指標8-isoprostane 濃度與揮發性有機化合物相關性，找尋潛在揮發性生物偵測指標。以主成份分析辨識肺癌病患肺泡氣體揮發性有機化合物特徵，以建立肺癌預測模式，並以receiver operating characteristic curves，計算呼吸氣體診斷肺癌之準確性。預期結果:病例組呼吸氣體脂質過氧化相關揮發性有機化合物濃度高於對照組。病例組與對照組呼吸氣體揮發性有機化合物模式有差異。以呼吸氣體分析偵測肺癌方法有高準確性。<br> Abstract: Background: Lung cancer is the leading cause of cancer death in the world. Screening for lung cancer atearly stages is important to provide appropriate treatment and reduce mortality. Developing a non-invasiveand highly accurate screening method is warranted.Objectives: To develop a breath test for lung cancer.Rationale: Lipid peroxidation play an important role in the pathogenesis of lung cancer. Volatile metabolitesgenerated during pathological processes are best measured in the alveolar air because it originates from thelower respiratory airways.Working hypothesis: Lung cancer could alter the volatile organic compounds (VOCs) in the breath. TheVOCs that are related to the lipid peroxidation can be analyzed to detect lung cancer by using an electronicnose and gas chromatography/mass spectrometry (GC/MS).Experimental approach: We design a successive study of 3 years to test the proposed hypothesis. In thefirst year, we will conduct a pilot study among 100 lung cancer patients undergoing surgery to validate thetechnique of alveolar air sampling and the discrimination ability of an electronic nose. Study will collectalveolar air from endotracheal tube to prevent environmental contamination. We will separately collectalveolar air of diseased and non-diseased lungs from double-lumen endotracheal tube, and use an electronicnose to discriminate VOCs between diseased and non-diseased lungs. In the second year, a case-controlstudy is designed to develop the prediction model for lung cancer by an electronic nose. We will recruit 150subjects of newly diagnosed lung tumors that will receive tumor resection surgery as the case group and 150age and gender-matched non-neoplastic subjects who will receive surgery under general anesthesia as thecontrol group. Alveolar air will be analyzed by an electronic nose. Data will be split into a training set formodel building and a validation set to assess the predictive ability with a fraction of 2:1. In the third year,we will recruit 100 cases and 100 controls. We will analyze the alveolar air using GC/MS, identify potentialvolatile molecular biomarkers for lung cancer, and validate the hypothesis of lipid peroxidation in thedetection of lung cancer. We will analyze lipid peroxidation biomarker of 8-isoprostane in exhaled breathcondensate and measure the association with VOCs, identify potential volatile biomarkers, use principlecomponent analysis to identify distinct patterns of VOCs in the alveolar air of lung cancer patients, buildprediction models for lung cancer, and generate receiver operating characteristic curves to calculate theaccuracy of breath tests.Anticipated results: The concentrations of VOCs associated with lipid peroxidation will be higher in thecases than in the controls. Patterns of VOCs in breath will be different between cases and controls. The breathtests have good accuracy in the detection of lung cancer.電子鼻呼吸氣體測試肺癌揮發性有機化合物electronic nosebreath testlung cancervolatile organic compounds