以質譜法開發偵測癌症之生物標誌與定量人體血漿中之單株抗體藥物

Abstract

癌症一直是全球人類死亡的主要原因之一，許多的研究，例如癌症風險因子評估、癌症的偵測方法、以及改善癌症的治療都是為減少其致死率。透過生物標誌的尋找，可以幫助病人早期發現疾病，達到早期診斷早期治療的效果。此外，由於單株抗體類的藥物常被廣泛的應用在癌症的治療上，而目前許多單株抗體藥物類的藥物動力學與藥效學還在進行研究，因此提供一個準確、精確的定量方法將有助於這些研究的進行。本研究使用氣相層析質譜儀與液相層析質譜儀來開發生醫分析方法應用於癌症偵測以及單株抗體藥物治療，期望可以改善癌症的治療，達到促進個人化醫療的契機。 本論文分成兩個部分，第一部分我們使用標的代謝體學方法在氣相層析質譜儀平台上尋找具有潛力的脂肪酸做為乳癌的生物標誌。早期偵測乳癌不但可以減少致死率，同時也可以提升治癒率。為了尋找具有潛力的脂肪酸作為生物標誌，我們利用氣相層析質譜儀開發代謝體學的脂肪酸分析方法，測定血漿樣品中的脂肪酸組成。血漿樣品經由乙醯氯進行衍生化反應，以氣相層析質譜儀進行分析，而確效的結果也顯示大於90%的定量結果誤差小於15%，在日內以及異日間的再現性評估，超過90%的結果其相對標準偏差都落在15%以內。我們進一步分析了健康受試者以及乳癌病患的血液中脂肪酸的含量。考量停經可能會影響脂肪酸的組成，我們將樣品分為停經前後的組別。本研究觀察到在停經後組別中，乳癌病患的血漿樣品內，C18:2n6、C22:0、C24:0這三個脂肪酸的含量顯著低於健康受試者，而在停經前的組別中，除了上述三個脂肪酸外，還有C16:0、C18:0、C18:1n9c、C20:0、C20:4n6、C22:6n3以及C24:1n9在乳癌病患血漿樣品中的含量也都顯著低於健康受試者。我們也進一步利用這些具有潛力的乳癌生物標誌建立乳癌的預測模型。最佳化後的預測模型如下: 在停經後以C18:2n6、C22:0、C24:0建立的預測模型，其受試者操作特徵曲線下部分面積數值為0.72，靈敏度與專一性數值分別為70.6%以及70.4%。而在停經前的組別當中，我們分別以C22:0、C24:1n9建立的預測模型，其受試者操作特徵曲線下部分面積數值為0.78，靈敏度與專一性數值分別為73.3%以及70.8%。 第二部分使用液相層析質譜儀開發兩個單株抗體藥物的血中濃度分析方法，我們先建立了一個以Protein G 磁珠樣品前處理的方法，從複雜的血液樣品基質當中，純化目標分析物，單株抗體藥物－Bevacizumab，並利用另一個immunoglobulin G (IgG)類的單株抗體藥物－Tocilizumab做為內標，校正在樣品配置過程中，可能產生之誤差。分析方法確效後，我們將本方法應用於實際臨床檢體之測量，定量病患體內的Bevacizumab濃度。結果顯示，儘管給予的Bevacizumab劑量相同，採血的時間點也相同，病人個體之間的藥物血中濃度仍然會具有達到兩倍多的差異。本方法可以應用於Bevacizumab的藥物動力學以及藥效學的研究，並提供一個監測Bevacizumab藥物血中濃度的方法。 有鑑於免疫球蛋白G (IgG) 類的藥物在目前大部分經由美國食品藥物管理局核准的單株抗體物中佔有很高的比例，為了促進這類型藥物的藥物動力學以及藥效學的研究，並提供藥物濃度監測的方法，我們另外開發了一個具有廣用性的液相層析質譜儀分析平台，定量血液樣品中IgG類的單株抗體藥物濃度，並以三個IgG類的藥物:Pembrolizumab、Nivolumab、Bevacizumab做為測試藥物，來證明平台的應用性。我們選擇了protein G作為我們的樣品純化方法，並發展雙內標校正法，使用一個IgG類的藥物tocilizumab做為樣品配置時的內標，搭配柱後注入內部標準品方法。研究結果顯示，三個測試單株抗體藥物的定量準確度的數值都落在100 ± 15%之間，在日內以及日間的再現性評估，相對標準偏差也都落在15%以內，顯示了雙內標校正法可以有效改善定量準確度。我們將分析方法應用於臨床血液檢體的波谷濃度點，定量結果顯示本方法具有足夠的靈敏度可以定量三個測試藥物的血中濃度。 本論文透過標的代謝體學的方法，尋找具有潛力的脂肪酸做為乳癌生物標誌，並開發單株抗體藥物血中濃度的測量方法。這些方法都經過分析方法確效，並且實際應用在臨床檢體的分析以確認方法的應用性。未來也期望這些分析方法可以更廣泛的使用於臨床，達到個人化醫療的最終目標。

Cancer is one of the leading causes of death globally. To fight this disease, the researches such as the risk factors of cancer, cancer detection methods, and cancer treatment are all growing rapidly in recent years. Biomarker discovery was one of the potential strategies for early detection of cancer. In addition, due to the promising efficacy of therapeutic monoclonal antibodies (mAbs), these types of drugs are frequently used in the cancer treatment. Many studies are still investigating the pharmacokinetic (PK) and pharmacodynamic (PD) properties of therapeutic mAbs. Therefore, an accurate and precise quantification method would facilitate these studies. This study used both gas-chromatography mass spectrometry (GC-MS) and liquid- chromatography mass spectrometry (LC-MS) platforms to develop bio-analytical methods for cancer detection and quantification of therapeutic monoclonal antibodies (mAbs) drugs, and hopefully to improve cancer treatment and also to facilitate precision medicine. This thesis is divided into two parts. In the first part, we used a targeted metabolomics approach to investigate the potential fatty acids as the biomarkers for breast cancer (BC) detection on GC-MS. Detect BC at early stage could not only reduce the cancer mortality but also improve the treatment outcome. To investigate potential fatty acid biomarkers, we developed a fatty acid profiling method for plasma metabolomic studies by GC-MS. The plasma samples were derivatized by acetyl chloride, and then analyzed by GC-MS. This analytical method was validated with accuracy within 100 ± 15%, and the intraday and interday precision were lower than 15% RSD for over 90% analytes. We further analyzed fatty acid profiles in both healthy volunteers and BC patients. To consider that the fatty acid profiles would be affected by the menopausal status, we divided the samples into pre- menopausal and post- menopausal. In post-menopausal group, the concentrations of three fatty acids including C22:0, C24:0, C18:2n6 were significantly lower in BC patients. In addition to C22:0, C24:0, C18:2n6, seven fatty acids, including C16:0, C18:0, C18:1n9c, C20:0, C20:4n6, C22:6n3, and C24:1n9, showed significant differences in pre-menopausal patients. We further applied these potential BC biomarkers to build the prediction model. The optimal prediction model for post-menopausal group was built by C18:2n6, C22:0 and C24:0 with the area under the curve (AUC) value of 0.72 (70.6% sensitivity and 70.4% specificity). In pre-menopausal group, the prediction model was built by C22:0 and C24:1n9, with the AUC value of 0.78 (73.3% sensitivity and 70.8% specificity). In the second part of this thesis, we developed two analytical methods for therapeutic mAb quantification in human plasma samples. In the first study, we developed a protein G purification method to trap the target analyte, bevacizumab, from human plasma, and applied another immunoglobulin G (IgG) based drug tocilizumab as the internal standard to correct the potential loss during sample preparation procedure. The method was validated, and then used to quantify the bevacizumab concentrations in patients’ plasma samples. The quantification results revealed that bevacizumab concentrations fluctuated significantly between individuals even though the dosage of bevacizumab and the collection time point were both the same for the tested individuals. This method could be applied to investigate the PK and PD properties of bevacizumab, and also for conducting therapeutic drug monitoring (TDM) of bevacizumab. IgG represents a high percentage of mAb drugs that have been approved by the Food and Drug Administration (FDA). To facilitate therapeutic drug monitoring and PK/PD studies, we developed a general LC-MS/MS method to quantify the concentration of IgG-based mAbs in human plasma. Three IgG-based drugs (bevacizumab, nivolumab and pembrolizumab) were selected to demonstrate our method. Protein G beads were used for sample pretreatment, combined with a two internal standard (IS) calibration method that includes the IgG-based drug-IS tocilizumab and post-column infused IS. The results indicated that the accuracy for three demonstration drugs were all within 100 ±15%. The intraday and interday precision were lower than 15% RSD. Using two internal standards was found to effectively improve quantification accuracy. The successful application of the method to clinical samples at trough levels demonstrated its’ applicability in clinical analysis This thesis used a targeted metabolomic approach to investigate the potential fatty acids as the BC detection biomarkers, and developed LC-MS/MS methods for quantifying therapeutic mAbs concentrations in human plasma. All of these methods were validated, and applied to clinical samples to investigate the method applicability. It is anticipated that these methods could be applied to clinical practice, and to achieve the goals of personalized medicine.