Fig. 5. Capture efficiencies for two uses of the MagWIRE: continuous versus single-pass flow.

In continuous flow, there is no significant difference in capture efficiency with respect to whether the cells are prelabelled with MPs or labelled in flow (P = 0.20). Furthermore, the MagWIRE performs equally well in the more viscous environment of blood (prelabel: P = 0.34; postlabel: P = 0.20). MagWIRE capture rates do not change significantly when prelabelled cells are only allowed to pass by the MagWIRE once (buffer: P = 0.52, blood: P = 0.20). However, the shortened labelling time (~10 s versus 10 min) causes postlabelled capture efficiencies to drop in single-pass flow (buffer: P = 0.025; blood: P = 0.004). While higher average capture efficiencies are achieved with continuous flow, the single-pass case allows rapid capture of large biomarkers with short circulating half-lives, and minimizes systemic exposure to MPs. Data were averaged across experiments with cell concentrations of 10, 100, and 1,000 cells ml⁻¹ (mean ± s.e.m., n = 6 independent experiments, see section ‘Statistics and reproducibility’). NS, P > 0.05, *P < 0.05, calculated by Mann–Whitney test. The 95% confidence intervals for each bar (from left to right) for continuous flow are as follows: 23.1–51.2%, 33.5–54.6%, 35.6–73.6% and 33.6–41.9%. The 95% confidence intervals for each bar (from left to right) for single-pass flow are as follows: 18.4–46.1%, 41.4–70.9%, 10.3–20.5% and 6.2–14.3%.