Microfluidics in the clinical lab

Nov. 21, 2016

There has been a great deal of hype for the last 20 years that microfluidics will revolutionize the use of the clinical lab by spawning hand-held, sensitive, and lower-cost diagnostics systems. In fact, the emergence of such systems has been relatively sparse, and the benefit of microfluidics is being seen in the core of larger systems that still intend to be in the clinical lab. Albeit micro in the fluidics, there must be a macro system around it, such as a cartridge of reagents, a manifold for multiple fluid processing steps, and larger still optical detection modules to measure the physics going on at the micro level. Where the real advantage lies is in the key microfluidic principles that govern the behaviour of the small, and it is these principles that are fundamental to the new diagnostic systems that can be used in the clinic today.

For example, “liquid biopsy” is a term for sampling blood for oncology diagnostics, with an aim to capture cells to detect or monitor cancers. There is an inherent difficulty in that such cells can be a needle in a haystack, occurring in parts per million or less of blood samples. One currently available system achieves label-free enrichment heterogeneous populations of cancer cells from 7.5ml of patient’s blood. It uses a microfluidic principle called Dean Flow, where liquid flowing in a spiral channel at just the right speed will cause target circulating tumor cells to gather at one side of the channel. By forking the channel, the target cells can be diverted to an output of enriched and purified material that can be detected in laboratory cytology systems.

Without such enrichment, cytology systems would have to trawl through blood samples dominated by normal leukocyctes to find the rare cells, taking more time and therefore adding cost. The principle here is that in a spiral channel, liquid flow builds a recirculation flow across the channel, much like a secondary flow makes tea leaves collect in the center of a stirred tea cup. With the knowledge that larger cells are likely to be the circulating tumor cells (CTCs), and that larger cells have a more difficult time recirculating across the spiral channel, they can be separated from smaller, more common cells.

Another principle fundamental to microfluidics is the rapid rate of heat transfer possible when the surface area-to-volume ratio becomes large. Analogous to small droplets of fine spray cooling the skin faster than a single splash, it is possible to change the temperature in a small amount of liquid much more rapidly than a large volume of liquid. This effect is key to PCR reactions inside some clinical lab consumables to amplify DNA. The tiny volumes of sample are heated and cooled for dozens of cycles over 45°C swings so tests can be completed in merely a couple of hours, such as for one commercially available TB test. Inside a microfluidic consumable, the sample is flowed through flat chambers that are placed against the heat source (or sink). The flat shape maintains a high surface area-to-volume ratio, and the tiny thickness allows rapid heat transfer, according to the diffusivity law that scales as D~L2/T, so the time T to diffuse can be driven down with the square of the thickness L.

Flowing a liquid through a porous medium for the purpose of wetting lots of surface area is a principle familiar to anyone making coffee. Small particles have a great deal of surface area to react with the liquid, and chemical processes can be greatly accelerated. The liquid flows among particles in the little voids and follows a sinuous path to the exit. There are many different routes to the exit, so the fluid effectively forms many parallel paths, each with a small flow, and it takes a great deal of pressure to maintain the flow.

A striking example in the clinical lab is the pressure required to squeeze liquid through a long column packed with beads in an HPLC (high pressure liquid chromatography) machine. In liquid chromatography, the molecules in a sample react differently to the surfaces of solid particles in the column, due to chemical interactions between the polar molecules in the solvent and the material in the column. Some molecules in the sample are attracted more than others, and spend longer in the column, so by the time the sample has flowed through, the molecules are spread out. This effect is heightened when the beads in the packed column are small, such as a few microns across, so the surface area for the chemical interactions is maximized. The consequence is that the pressure to drive the liquid sample through the column becomes huge.

In HPLC, the sample is driven through the column at hundreds of atmospheres of pressure. The different molecules exit the HPLC column spread out over time, and can be measured separately to generate a chromatography trace in the clinical lab. To increase sensitivity and specificity, one can imagine using smaller and smaller beads, and going to higher and higher pressures. Ultra-high pressure liquid chromatography, or UHPLC, can use particles as small as two microns, requiring pressures of 15,000 psi (more than 1,000 atmospheres) of pressure in a packed bed, which requires using advances in materials and bonding techniques for containing the pressure.

Finally, we are all familiar with an emulsion—for instance, oil and vinegar salad dressing—that clearly shows the refusal of certain liquids to mix. Taking two immiscible liquids down to microfluidic scales, tiny spherical water droplets known as picodroplets remain stable within a channel of oil. These water-in-oil systems can permit the generation of thousands of separate aqueous reaction vessels for rapid drug screening and cellular testing. At small scales, these water droplets are dominated by surface tension and form nearly spherical shapes that can be controlled. In a picodroplet generator, water and oil meet at a microscale crossroads, and the water in the middle pinches off into separate droplets that remain distinct from each other. It is possible to fill these thousands of picodroplets with different drug titrations, different cells, or individual DNA base pairs, which makes each picodroplet like a separate reaction vessel.  This way, it is possible to massively increase the throughput of drug discovery, cell analysis, or DNA sequencing.

The hype has not so far been realized, but the fundamental principles of microfluidics will underlie key technologies that will enable the future clinical laboratory.

Richard Day, PhD, serves as Associate Director, Medical Technology Division, for Cambridge Consultants in Cambridge, UK.