Manipulation of immunomagnetic targets in microfluidic channel flow

Collection of rare target cells in clinical samples, such as disseminated tumor cells in peripheral blood, poses difficult challenges to currently available cell separation methods. A novel method enabling highly efficient and specific concentration of rare target cells has been developed. The method proposes immunomagnetic separation within a two stage microchannel device in order to isolate target cells from milliliters of blood to nanoliter volumes. In the first stage, a magnetic gradient is applied to a shallow channel. Labeled cells move towards the channel surface where they are rolled by the viscous force of the flow. They are then siphoned out where they can be collected or transferred to the second stage. The second stage consists of a channel bordered by a thin ferromagnetic wire. Labeled cells align with the magnetic field gradient formed by the wire, and they can be analyzed or moved along the wire by increasing flow. Models were developed to describe the journey of a cell in an integrated device. Separate two-dimensional particle models predict cell trajectory in each channel while a three dimensional model predicts rolling speed for cells reaching the surface of the first channel. Experiments tracking the rolling velocity of cells and model cells show agreement with field flow fractionation (FFF) theory at high flow rates but deviate from this theory at lower flow rates. In general, the trajectory models accurately predict experimental capture percentage versus flow rates for model cells, particularly after simple modification to include hydrodynamic lift forces. Initial experiments using model cells spiked into mouse blood suggest that a reduction in capture efficiency in blood relative to buffer is predictable. The ability to predict and achieve nearly 100% efficient capture for clinically practical flow rates suggest that the proposed devices could outperform the current immunocytochemistry gold standard for selectivity of 1 : 10 6 (cancerous cells : white blood cells). The proposed technology could also facilitate integration with other microfluidic lab on a chip technologies. This could, in turn, result in sorely needed automation and standardization for disseminated tumor cell analysis. Also, there is a potential for integrated immunocytochemistry and genetic analysis. In addition, the theory and methods developed could be applied to other important separation problems including pathogen detection and manipulation of soluble bead based protein or DNA arrays.