A negative selection methodology using a microfluidic platform for the isolation and enumeration of circulating tumor cells

Abstract

Circulating tumor cells (CTCs) exist in the peripheral blood stream of metastatic cancer patients at rates of approximately 1 CTC per billion background cells. In order to capture and analyze this rare cell population, various techniques exist that range from antibody-based surface marker positive selection to methods that use physical properties of CTCs to negatively exclude background cells from a CTC population. However, methods to capture cells for functional downstream analyses are limited due to inaccessibility of the captured sample or labeling techniques that may be prohibitive to cell function. Here, we present a negative selection method that leverages a Microfluidic Cell Concentrator (MCC) to allow collection and analysis of this rare cell population without needing cell adhesion or other labeling techniques to keep the cells within the chamber. Because the MCC is designed to allow collection and analysis of non-adherent cell populations, multiple staining steps can be applied in parallel to a given CTC population without losing any of the population. The ability of the MCC for patient sample processing of CTCs for enumeration was demonstrated with five patient samples, revealing an average of 0.31 CTCs/mL. The technique was compared to a previously published method - the ELISPOT - that showed similar CTC levels among the five patient samples tested. Because the MCC method does not use positive selection, the method can be applied across a variety of tumor types with no changes to the process.

Keywords: Circulating tumor cells; ELISPOT; Microfluidics; Negative selection.

Figure 1

A) Summary of blood processing method. Whole blood was split into two parallel processed samples, one containing spiked LNCaPs for efficiency studies and the other remained non-spiked for CTC enumeration. B) Top view schematic of the MCC showing the isometric flow pattern created when pipetting sample to the center chamber extending radially to the output ring. The passive pumping pressure created between the input and output droplets is evenly distributed through the transport channels, allowing for an increased throughput for sample processing. C) Device loading and postprocessing methodology is shown for the EpCAM primary and secondary staining within the MCC.

Figure 2

A) Top-view schematic of the MCC with the stitched fluorescent image shown expanded (right). Below, an image of spiked GFP-fluorescent LNCaPs (white arrows, green) that were stained for EpCAM (red) shown with and without the internal green marker to demonstrate the clarity of the surface EpCAM stain. Blue arrows show beads auto-fluorescing. B) Count concordance shown between EpCAM cells over counted GFP cells to demonstrate robust staining and enumeration using spiked LNCaPs into a blood sample. Cells were spiked and counted using the red channel for EpCAM and compared to counts using the green channel for the LNCaP internal fluorescence. C) LNCaPs were spiked into whole blood samples to evaluate the efficiency of the capture. Average efficiency of all processed samples was 60.4 ± 24.1%, however two of the samples (*) were confounded due to a device inconsistency (see methods – efficiency analysis).

Figure 3

A) Individual spiked CTCs imaged within the MCC shown (arrows). B) Prostate Cancer CTCs per mL of whole blood calculated based on enumerated cells within the concentrator. CTCs were detected and enumerated in each patient except for CTC015. C) Breast Cancer CTCs per mL of whole blood calculated based on enumerated cells within concentrator. No CTCs were detected in either patients 14 or 20.

Figure 4

Validation of Concentrator CTC enumeration via comparison with ELISPOT methodology. Blood was split upon collection and processed in parallel. Data representing CTC enumeration shown (below)