微流碟片平台分離及偵測血液中稀少細胞之研究

Abstract

血液中某些稀少細胞的檢測，往往能提供作為臨床疾病診斷之重要指標。例如，轉移性癌症是腫瘤釋放循環腫瘤細胞 (circulating tumor cells(CTC)) 到病人的血液或淋巴管中，經由血液或淋巴循環系統散布到身體其他器官，並在健康器官中形成新的腫瘤，而目前的醫學影像和血液中的癌症相關抗體診斷在初期的轉移癌症中是難以檢測。所以在臨床診斷中，能提供醫生病人血液中的CTC轉移的即時情況極為重要，能作為治療效果的評估和治療過程的監視。然而CTC在血液中的數量極為稀少，每十億血球細胞中只有數個CTC，因此快速且廉價地分離和偵測血液中的CTC為工程和醫學上的重要的挑戰。 本研究分別針對免疫磁珠正負篩選之分離方法，提供兩種不同設計之微流體碟片平台及建立相關系統，並於免疫磁珠分離細胞後，以多個免疫螢光標定來判斷抓取到的目標細胞，予以定量。負篩選分離實驗中，為了解設計碟盤的特性與執行率等性質，我們以人類血癌細胞株(Jurkat)模擬白血球，而人類乳腺癌細胞株 (MCF7) 模擬病人血液中的稀少CTC進行初步測試，而後直接用血液分離出的周邊血液單核球(MNC)與MCF7進行實驗。負篩選實驗為Jurkat/MNC標定免疫磁珠Anti-CD45-PE和 Anti-PE BD magnetic beads，並由磁鐵抓取這些非腫瘤細胞後，於碟片內特製的觀察區域偵測具Anti-cytokeratin-FITC信號之目標細胞(MCF7)。而在正篩選分離之實驗中，我們仍以MCF7模擬病人血液中的稀少CTC，並用健康人的全血來模擬病人血液的背景細胞及與台大醫院合作實際偵測乳癌病人之CTC數。MCF7與Anti-EpCam-PE和 Anti-PE BD magnetic beads進行結合，使其能被碟盤上的磁鐵陣列抓取，抓取後在碟盤上進行Anti-cytokeratin-FITC 和 Hoechst33342兩個螢光抗體標定，作為螢光顯微鏡下判斷CTC的螢光信號。為了能在碟盤上進行多個螢光標定，設計出氣孔閥 (Vent valve) 作為多個螢光染劑和清洗溶液的序列式開關，並利用氣孔控制盤 (Vents control plate, VCP) 來同步控制各個對應的氣孔閥門，且此閥的操作在未來是有可能完全自動化並整合在微流碟盤中，減少人為操作的實驗誤差。 實驗結果顯示，在１毫升的健康人全血中混入50~300個 MCF7的實驗中，均有50~60%的回收率，而一片碟盤一次的處理速率為每小時２毫升血液甚至可以再增量，且此技術的靈敏度可達大於10-7。因此系統有高的血液處理速率、操作簡單、低成本的材料和加工、細胞損失率在可接受範圍並且有機會自動化減少人為操作成本和誤差。可望未來能經由病人的臨床測設結果建立起此系統的標準化並提供醫生良好的診斷工具。

Rare cells in blood often possess high clinical significance. However, cyto-analysis of rare cells often requires separation and detection with either procedure of substantial challenge. Circulating tumor cells (CTCs) in the peripheral blood of metastatic cancer patients represent a potential alternative to invasive biopsies as a source of tumor tissue for detection, characterization, and monitoring of non-haematologic cancer. This thesis outlines a novel disk-based microfluidic device to capture and detect rare cells from blood sample. Immunomagnetic negative selection and positive selection approaches were demonstrated in our disk platforms. For negative selection approach, the microfluidic platform’s unique features include multistage magnetic gradient to trap labeled cells in double trapping areas, drainage of fluid to substantially shorten detection time, and bin-like regions to capture target cells to facilitate seamless enumeration process. Proof-of-concept was conducted using wide range of MCF7 as target rare cells (stained with anti-cytokeratin-FITC antibodies) and spiked into Jurkat Clone E6-1 non-target cells (labeled with anti-CD45-PE and anti-PE BD magnetic beads). Then, mononuclear cells (MNC) from healthy blood donors were mixed with MCF7s, modeling rare cells, and tested in the disk. Results show the average yield of detected MCF7 is near-constant 60±10% over a wide range of rarity from 10-3 to 10-6 and this yield also holds for MCF7/MNC complex mixture. Comparison with autoMACS and BD IMagnet separators revealed the average yield from the disk (60%) is superior to that of autoMACS (37.3%) and BD IMagnet (48.3%). For positive selection testing, as proof-of-concept, experiments were conducted where MCF7 was used to simulate CTCs and healthy whole blood was used for background peripheral blood. A continual flow process via a centrifugal microfluidic disk platform to capture MCF7 in blood immunomagnetically and enumerate them on-disk with a complete batch process of multi-fluorescence labeling is presented. The MCF7 are labeled with anti-EpCAM-PE and anti-PE magnetic beads for magnetic force capturing and with anti-cytokeratin-FITC antibodies and Hoechst33342 for detection. In order to allow precise timing in liquid delivery during the multi-fluorescence labeling processes, on-disk deterministic vent valves were designed. To characterize the disk performance of target cell capturing and fluorescence labeling, three different labeling procedures were used. Results show that the cell-capture yield of the disk was about 65% and the throughput was 2ml/hr or more. After staining two-label fluorescence on the disk, the yield was around 50%. The sensitivity of the technique in enriching rare cells from whole blood (>1ml) is up to 10-7. Direct fluorescence labeling on the disk without sample transfer and manual operation greatly helped to reduce cell loss. The total procedure, from magnetic bead labeling to completing two-label fluorescence staining, takes place within 1.5 hours. In order to determine the efficiency of the disk in enumerating CTC from patient with epithelial cancers, the breast cancer data of patients were collected. Advantages of the present platform include simple operation, high throughput, an acceptable level of cell loss, and a potentially low system cost, which should substantially ease the effect in cyto-analysis of rare cells.