Results
Our protocol had a mean recovery efficiency of 69.5% and a limit of detection of at least four cells per millilitre. We developed an analysis method to reduce noise from magnetic beads used for CTC isolation. CTCs were isolated from CRC patients with a median of 37 CTCs (IQ 13.0-85.5) at baseline. CTCs from CRC patients were significantly (p < 0.0001) larger than cytokeratin (CK)-negative cells, and patients were stratified into two groups based on BRAFV600E and PD-L1 expression on CK-positive cells. The changes observed over time included not only the number of CTCs but also their distribution into four different subpopulations defined according to BRAFV600E and PD-L1 positivity. We developed a novel protocol for semi-automatic CTC isolation and phenotypic characterisation by combining two platforms. Assessment of CTCs from early CRC patients using our protocol allowed the identification of two clusters of patients with changing phenotypes over time.