- 學位論文
以氣體分析技術發展肺癌生物指標並探討肺癌與空氣汙染相關性研究
Use of gas analysis technology to develop volatile biomarkers for lung cancer and explore the association between lung cancer and air pollution
摘要
肺癌為癌症死因第一位，於不吸菸族群肺癌之發生率仍持續上升，推測與空氣汙染所致。肺癌致病機轉造成氧化壓力上升，產生揮發性有機化合物 (Volatile Organic Compounds, VOCs)並擴散至毛細微血管中。於先前研究用於新篩檢工具開發，發現以電子感應器陣列分析肺癌與對照組患者氣體，並利用機器學習建立預測肺癌模型驗證準確度為85.7%。然而呼氣分析仍可能含有其他如上皮、口腔細胞等其他VOCs之干擾，研究將分為二個目標: (1)分析肋膜積液之揮發性代謝物作為肺癌患者的生物標誌物,(2)使用肋膜積液作為暴露指標。以肺癌與非癌症病患的橫斷性研究設計，納入診斷為肺癌並接受胸水引流術之患者，與非惡性肋膜積液接受胸水引流術之患者，收集引流瓶中的肋膜積液，進行頂空分析，並以固相微萃取做為樣本前處理方法。分析內源性小分子VOCs選用CAR/PDMS纖維進行50℃頂空加熱，轉速800rpm，萃取10分鐘，再進行氣相層析質譜儀分析。共發現213個代謝物，基於偏最小平方判別分析（Partial least squares Discriminant Analysis, PLS-DA）中VIP(variable importance in the projection)值 > 1之代謝物，我們發現了78個代謝物，並使用這些高貢獻率的代謝物來構建PLS-DA，表現出很明顯的區分，置換測試顯示R2 = 0.79和Q2 = 0.65。此外，進一步進行了1000次重複的t檢驗。結果顯示，共49種代謝物的數據具有統計學意義（P <.05）。 PLS-DA也表現出明顯的區分，置換測試顯示R2 = 0.73和Q2 = 0.53。其中與肺癌機轉脂質過氧化相關之酮類包含methyl vinyl ketone, 4-amino-4-methyl-2-pentanone，甲基烷類包含1-methyl-2-propyl- cyclohexane, 1,1,2,2-tetramethyl-cyclopropane可能可視為生物指標。而空氣汙染相關之苯、甲苯、乙苯、二甲基苯以及多環芳香烴為待測物，並依據患者住家距離交通主要幹道來做族群分類。經固相微萃取最佳化條件後，使用PDMS/DVB纖維進行80℃加熱，轉速240rpm，萃取60分鐘，再進行氣相層析質譜儀分析。結果發現住家距離交通主要幹道小於300公尺的族群，Naphthalene與Fluoranthene濃度分別為0.57與0.30 ng/ml且顯著高於住家距離交通主要幹道大於300公尺的族群，由此推估肋膜積液可能可以成為空氣汙染暴露指標。將來，必須設計世代研究和確定暴露來源，以進一步探索暴露的空氣污染物與肺癌之間的相關性。
並列摘要
Lung cancer is the leading cause of cancer death, and the incidence of lung cancer in non-smoking populations is still rising. It was suspended to air pollution. Lung cancer pathogens cause an increase in oxidative stress, producing volatile organic compounds (VOCs). In the previous study was used sensor array to analysis exhaled air and building machine learning to early screening. The prediction accuracy of the predicted lung cancer model is 85.7%. However, breath analysis may still contain interference from other VOCs such as epithelial cells and oral cells. The study was be divided into two objectives: (1) To develop a screening test for lung cancer using a sensor array, (2) Use residential distance to major roadways to explore air pollution exposure. Designed for cross-sectional study of lung cancer and non-cancer patients receiving thoracentesis, collecting pleural effusion in drainage bottle and analysis by its head space. The analysis of solid phase microextraction (SPME) was used as a sample pretreatment method. The endogenous small molecule VOCs were analyzed by CAR/PDMS fiber for 50 °C headspace heating at 800 rpm for 10 min and analyzed. Based on the VIP score of the PLSDA model, we found 78 metabolites whose VIP scores were greater than 1. We used these metabolites that distributed were high contribution ratios to build PLS-DA. The PLS-DA also showed great discriminant and the permutation test showed the R2= 0.79 and Q2=0.65. Moreover, we conducted bootstrapping student's t-test with 1000 replications on these metabolites. The result showed that compared to the non-malignant patients, the data of 49 metabolites from the lung cancer patients were statistically significant (P < 0.05). The PLS-DA also showed great discriminant and the permutation test showed the R2= 0.73 and Q2=0.53. The ketones associated with lipid peroxidation of lung cancer include methyl vinyl ketone, 4-amino-4-methyl-2-pentanone, and methyl alkane contains 1-methyl-2-propyl-cyclohexane may be considered biomarkers. The air pollution-related benzene, toluene, ethylbenzene, dimethyl benzene and polycyclic aromatic hydrocarbons are the substances. After the SPME is optimized for extraction, the PDMS/DVB fiber is used for heating at 80℃, and the rotation speed is 240 rpm. , extraction for 60 min, and then analysis by its headspace. According to the residential distance from the main roadways to do group classification. The results showed that the concentration of naphthalene and fluoranthene was 0.57 and 0.30 ng/ml, respectively, which was significantly higher than that of the main road with residential distance of more than 300 meters. It is estimated that pleural effusion may be an indicator of air pollution exposure. In the future, the source of exposure and design cohort studies must be identified to further explore the correlation between exposed air pollutants and lung cancer.