Abstract 1531: Isolation of both EpCAM+ and EpCAM- tumor cells in microfluidic devices

Abstract

Abstract Isolation and enumeration of circulating tumor cells (CTCs) in peripheral blood have been used for cancer diagnosis, prognosis, and theragnosis. The majority of the CTC isolation methods, including the FDA-approved CellSearch®, employ antibodies against epithelial cell adhesion molecule (EpCAM). However, it is known that some CTCs express little or no EpCAM, partially due to epithelial-to-mesenchymal transition (EMT). As a result, an ideal method should be able to isolate all types of tumor cells including epithelial tumor cells (EpCAM positive), mesenchymal tumor cells (EpCAM negative), and those cells in the transition. We have investigated in using fibrin-immobilized microfluidic devices to isolate both EpCAM+ and EpCAM- cells, since it has been found that fibrin deposits on the surfaces of carcinoma and sarcoma cells where normal cells and tissue showed no deposits of fibrin during tumor progression/metastasis [1]. The microfluidic devices we used [2] are in the size of a microscope slide and each of them consists of an inlet, an outlet, and eight parallel channels connected via consecutive bifurcation between the inlet and outlet. There are micromixers inside the channels for enhanced interactions. The channels were first exposed with thrombin and then fibrinogen, followed by a buffer to stop the thrombin-activated fibrinogen's polymerization. At the end, a thin layer of fibrin polymer was formed on the surface of microchannels and the device was then used to isolate tumor cells. We tested the device for isolation of Pan-1 (EpCAM+ pancreatic cancer cells), CCRF-CEM (EpCAM- leukemia cells), and MIAPaCa-2 cells (EpCAM- pancreatic cancer cells) in PBS (phosphate buffered saline). All of them showed capture efficiency of 80% to 90% at a flow rate of 1.0 μL/s. The capture efficiency, calculated by dividing the number of cells captured by the number of cells introduced into the device, decreased as the flow rate increased from 0.5 μL/s to 3 μL/s, because there were less interaction time between cells and fibrin at a higher flow rate. Capture efficiency of 86+4% was obtained when CCRF-CEM cells were spiked in human whole blood at a concentration of 1 thousand to 1 million cells/mL. We also studied the effects of combining aptamers with fibrin on the channel surfaces and found an enhanced interaction between aptamer-specific tumor cells and the aptamer-fibrin surface. In summary, microfluidic devices with fibrin-immobilized channel surfaces were able to capture both EpCAM+ and EpCAM- cells, suggesting a possibility to isolate both epithelial and mesenchymal tumor cells. Moreover, the fibrin surface could be enhanced by tumor-cell-specific capture agents such as aptamers or antibodies.