Fig.5 . E.coli separation from diluted blood. (A) Qualitative analysis showing fluorescently labelled (using bacterial viability kit) E.coli migrating to the inner wall and collected in O1 while blood cells remain focused at the outer wall and collected in O2. Scale bar : 100 µm. (B) Quantification of E.coli (1000 CFU/mL) spiked into three diluted blood samples (1:10, 1:5, and 1:2). Agar plating based quantification show E.coli separation efficiency between 90% to 82% depending on blood dilution.
For sepsis applications, ideally large volumes (2-7 mL whole blood) will be necessary for any sample preparation method. Microfluidics, defined as the science and technology of systems that can handle small volumes of fluid in channels of ten to a few hundred microns in size, has proven to be well suited for applications in clinical diagnostics. However, while impressive progress has been made in high throughput rare circulating tumor cell isolation, there has been limited success for bacteria isolation from blood using microfluidics. In general, there is a trade-of between separation efficiency and throughput for any given technology. To this end, in comparison to the current state of the art, our method is well suited for high separation efficiency without compromising throughput much. In Fig.6, we compare work published using both active (magnetophoresis 38, acoustophoresis32, 33 and dielectrophoresis 35, 36) and passive (size-based separation 28, 31, 34, 37) to our current work for bacteria isolation from blood. While the experimental conditions (such as dilution ratio and bacteria concentration) vary between the studies, making the comparison of performance difficult, Fig 6 shows elegantly shows the compromise brtween throughput and separation efficiency. While a more fair comparison would be to use similar experimental conditions, e.i same bacteria concentration, we found that ost of the work (∼ 70% of the studies) used high bacteria concentration of ≥ 104CFU/mL28, 31, 32, 34, 35, 36,38 and most of the work in inertial microfluidics that show good separation efficiency have lower throughput 28-36 while studies that highlight higher throughput show lower separation efficiency37, 38. The separation efficiency corresponding to the three different blood dilutions in the current work is highlighted in red in Fig.6. Using a single spiral, we demonstrated processing of 1 mL of blood at extremely high throughput with separation efficiency >80% for all the blood dilutions tested. To our best knowledge, this is the highest throughput reported for a single passage using a single chip at the given high separation efficiency. The extremely high separation efficiency is attributed to the fact that pre-positioned blood cells at the outer wall remain fully focused throughout the length of the spiral channel while bacteria migrate. In this work, we have focused on separating bacteria from RBCs and nucleated cells, but given the size of platelets (2 to 4 µm) we anticipate majority of platelets should also be separated based on the size difference.