Technical validation of a new microfluidic device for enrichment of CTCs from large volumes of blood by using buffy coats to mimic diagnostic leukapheresis products

Abstract

Diagnostic leukapheresis (DLA) enables to sample larger blood volumes and increases the detection of circulating tumor cells (CTC) significantly. Nevertheless, the high excess of white blood cells (WBC) of DLA products remains a major challenge for further downstream CTC enrichment and detection. To address this problem, we tested the performance of two label-free CTC technologies for processing DLA products. For the testing purposes, we established ficollized buffy coats (BC) with a WBC composition similar to patient-derived DLA products. The mimicking-DLA samples (with up to 400 × 10⁶ WBCs) were spiked with three different tumor cell lines and processed with two versions of a spiral microfluidic chip for label-free CTC enrichment: the commercially available ClearCell FR1 biochip and a customized DLA biochip based on a similar enrichment principle, but designed for higher throughput of cells. While the samples processed with FR1 chip displayed with increasing cell load significantly higher WBC backgrounds and decreasing cell recovery, the recovery rates of the customized DLA chip were stable, even if challenged with up to 400 × 10⁶ WBCs (corresponding to around 120 mL peripheral blood or 10% of a DLA product). These results indicate that the further up-scalable DLA biochip has potential to process complete DLA products from 2.5 L of peripheral blood in an affordable way to enable high-volume CTC-based liquid biopsies.

Figure 1

Schematic representation of DLA-enrichment workflow using the DLA biochip. 0.5, 2 and 4 mL of DLA samples with a concentration of 100 × 10⁶ WBCs/mL were diluted to 20 mL before the enrichment. The upper panel shows the first enrichment cycle through the DLA biochip which lead to a first enriched fraction (sample1), which is recovered, diluted and re-loaded through the same DLA biochip for a second enrichment cycle (lower panel).

Figure 2

Cross-sectional views and Dean vortices generated in curvilinear microchannel for cell separation. (a) Adapted from. Copyright 2013 by Creative Commons CC-BY-NC-ND. (Left panel) schematic design of CTChip FR1 showing the input slots for buffer and sample, output slots for CTCs enriched fraction and WBCs-rich waste. (Right panel) Schematic cross section showing the hematogenous cells (in white and red) and CTCs (in green) at three different positions (X, Y, Z as in the left panel). (b) (Left panel) schematic design of the DLA biochip showing the input slot for the pre-diluted sample and the output slots for CTCs enriched fraction and WBCs-rich waste. (Right panel) Schematic cross section showing the hematogenous cells (in grey and red) and CTCs (in green) at two different positions (1 and 2 as in the left panel).

Figure 3

Composition of DLA samples, unprocessed and processed buffy coats. (a) Relative distribution of RBC, platelets and WBCs with (b) indication of the fraction of the different leucocyte subpopulations determined with CELL-DYN Ruby hematology analyzer and representation of the median values with interquartile range. Whiskers represent total range. The difference between the median of the percentage of five individual populations of WBCs in the three different highly concentrated blood products as in (a) was compared. Statistical analysis (Mann–Whitney U test): *P < .05; **P < .01; ***P < .001; ****P < .0001.

Figure 4

Spiked-in cells recovery rates after enrichment with the CTChipFR1 and DLA biochip. (a) 1000, (b) 4000 and (c) 8000 cancer cells from each of three differentially pre-labelled cell lines were spiked with flow cytometry respectively into (a) 50 × 10⁶, (b) 200 × 10⁶ and (c) 400 × 10⁶ WBCs from mimicking-DLA products. Enriched cells were enumerated again by flow cytometry and compared with the number of initially spiked cells to determine recovery rates. The difference between median recovery rates in percentage were compared after enrichment with the ClearCellFX System (green boxes) and DLA biochip (black boxes). Whiskers represent total range. Statistical analysis (Mann–Whitney U test): *P < 0.05; **P < 0.01.

Figure 5

Evaluation of WBC contamination after one and two cycles of enrichment with the DLA biochip. mDLA products containing 50 × 10⁶, 200 × 10⁶ and 400 × 10⁶ WBCs were enriched with the DLA biochip and the (a) Mean number of WBCs in the enriched fraction after one (S1, analyzed with CELL-DYN Ruby hematology analyzer), and two enrichment cycles (S2, analyzed with flowcytometry). Lines represent standard deviation. (b) Mean percentages with standard deviation of WBCs in mDLA products containing 400 × 10⁶ WBCs and enriched with the DLA biochip were determined in the S2, in the waste of the first enrichment cycle (W1) and in the waste of the second enrichment cycle (W2). (c) Representative snapshot of the WBCs flow during the first and the second enrichment cycle, tracked using the Photron Fastcam SA3 (MEC, Indiana) connected to the Olympus IX71 inverted microscope (Olympus, US).