Fluid state

Microfluidics – the science of manipulating fluids at the micron scale – is still a relatively young field. Researchers are excited about the technology’s potential application in medical diagnostics and treatment, but there are hurdles to be overcome before it becomes truly commercially viable.

The great advantage of microfluidics in medicine is that it can replace a traditional diagnostic laboratory. Each microfluidic device uses a small chip engraved with tiny, hair-width channels to house fluids such as blood or saliva samples. Just as in a lab, the interaction between the fluids and a chemical reagent is used to detect biomarkers, but the tiny scale makes diagnosis quicker and easier. Only very small amounts of fluid – a pinprick of blood, for example – are needed, and the quantities of reagent required are similarly minute. It’s also possible to create integrated microfluidic devices where a single substrate chip is used to perform several different functions. In short, microfluidics appears to solve problems of portability, cost and ease of use in a single stroke.

Complex liquids

Behind such simplicity, of course, lies complexity. At this small scale, the science of how liquids behaves is very different, says Hsueh-Chia Chang, Bayer professor of engineering, and director of the Center for Microfluidics and Medical Diagnostics, at the University of Notre Dame. “In the past 20 years, we have studied the science of microfluidics in detail,” he says. The science is now well understood, but the chips used by researchers to test the scientific principles are expensive to make. There is a big leap from those chips to the type of mass-produced chip needed to make the use of microfluidics in diagnostics commercially viable.

Chang says that the most profitable microfluidic products currently on the market use digital polymerase chain reaction (dPCR) to amplify segments of DNA and RNA for the diagnosis of diseases such as cancer or Ebola. This digital technology is more precise and accurate than older polymerase chain reaction technologies. The technology is mostly sold by large companies such as Bio-Rad and Dolomite, and the market is expected to grow by 10% every year for the next five years.

A number of start-up companies are also developing microfluidic products. More than 100 have launched products based on technology that can remove circulating tumour cells from blood samples. “These are cancer cells that have been released by cancer tumours into the blood stream, and they are circulating with the blood, so if one could detect these cancer cells and identify where they come from – whether it’s the lung or the liver, for example – it could lead to an early diagnostic device,” says Chang. This is something that is difficult to do at the macro scale, but the use of microfluidics enables a greater degree of precision and efficiency.

Nanofluidics potential

The complementary technology of nanofluidics may start to reach commercial potential soon. While microfluidics works at the cellular level, with materials ranging from 1–100μ in size, nanofluidics deals with materials smaller than one micron – mostly individual molecules. A nanopore is a pore of nanometre size, created by a pore-forming protein or as a hole in synthetic materials. When a nanopore is present in an electrically insulating membrane, it can be used as a singlemolecule detector.

The main company exploiting nanofluidic technology to commercialise nanopore arrays is the UK-based Oxford Nanopore, which uses a ‘soft’ nanopore to identify proteins by analysing the signals produced when molecules pass through it. Chang explains, “The electrical current that goes through a nanopore would be blocked by the DNA molecule, and it’s so small that how much current is blocked can tell you the different bases on the DNA – the four bases, A, T, C and G will produce different kinds of current signals.” But there is one drawback to the Oxford Nanopore product, Chang says. The throughput is very low. “It takes about one minute for a molecule to go through, so you cannot count too many molecules. An array enables them to have more pores, but this still limits them to a few hundred molecules and that’s too small in number for most samples.”

The alternatives to soft nanopores are solid state nanopores, which are drilled out of polymeric materials or semiconductors. They offer a higher throughput but there’s a catch. “The trade off is that you cannot tell which base it is, and sometimes you cannot even tell when one molecule has gone through,” Chang says.

Although it’s possible that microfluidics and nanofluidics could be used in developed countries in the long term, the most likely immediate use is for point-ofcare diagnostics in developing countries. “We can obviously go into any clinic in the US or UK and do a tissue biopsy for breast cancer screening, for example, but in the developing world they don’t have technicians who can extract the tissue and they don’t have people who can read the slides after the tissue has been extracted,” Chang says. “So you’re looking for alternatives. [Microfluidic devices] may not be as accurate as a tissue biopsy by a skilled technician, but it’s much better than nothing at all.”

Microfluidic chips

Chang’s own team at Notre Dame is working on developing polymerbased microfluidic chips containing embedded microelectrodes that can detect peptide biomarkers, and DNA and RNA molecules. This technology could have particular value in controlling epidemics such as AIDS, Zika or dengue fever. In the case of dengue, there are four kinds of virus, and humans only suffer severe symptoms if infected for the second time by the same virus. It’s important to know what kind of virus the mosquitoes are carrying in each area, so that clinicians can prioritise treatment of patients who have been bitten by the mosquito with the same virus a second time.

“We try to isolate cells with microfluidics, and then try to isolate and identify molecules with nanofluidics. So with the dengue virus, what one would do is take saliva samples or blood samples from the patient, and then try to isolate a virus,” says Chang. “Once that’s done, we try to isolate the RNA segments from the virus and then we would decipher these sequences on the RNA to discover which of the four kinds of dengue virus it is.”

Chang has licensed technologies to two start-ups, FCubed and AgnDx. FCubed is producing DNA-based molecular devices and accompanying test kits. The devices are going through clinical trials, because they have to go through regulatory hurdles in different countries before they can be approved for use. However, the technology is already being used for other purposes such as food safety, and environmental and water detection. AgnDx is creating a product to diagnose pancreatic cancer that will be aimed at the developing world, particularly Asia, where the disease is prevalent.

Despite the huge potential offered by microfluidics and nanofludics, there are obstacles to overcome before the technologies can be widely used in medicine. “The microfluidic and nanofluidic technologies are chips – very much like electronic chips,” says Chang. “These chips will be inserted into portable instruments. These are obviously reusable, so each one may cost a few hundred dollars or even a thousand dollars, but that’s not an issue. The issue is with the chips – we’ll put blood samples or urine samples or even saliva samples into the chips, so they cannot be reused; they have to be disposable. The companies selling this technology will make money by selling the chips, not the instrument. So it comes down to making these chips in large quantities at low cost.”

Currently, the main material used to make the chips is glass, though there have been efforts to make chips out of ceramics or metal. Chang believes that polymer is the best material to create chips that could be manufactured at low cost in large volumes, and is now working on ways to achieve this. Much of his team’s effort, he says, is focused on finding a way to make polymer chips that can be manufactured by injection moulding, a large-scale production technology. At the moment, however, manufacturing of polymer chips is still expensive. “It’s almost like making an electronic chip, but while an electronic chip will last forever, this chip can only be used once,” he says.

Although polymer is a much more suitable material than glass or ceramics, it presents its own difficulties. “The number one challenge is that you have to cut very small channels, micron size. How do you cut such small channels in a polymer material? Then there’s a question of how you bond them together to form a chip that will not leak, so they can actually withstand high pressure. There are also considerations like optics – would these polymer chips be transparent?” Chang says. Nevertheless, he adds, “The manufacturing technology for polymer chips, for nanofluidic and microfluidic diagnostic devices, has now been developed to the extent that we can start manufacturing them right away.”

Home use potential

In future, the technology may develop to the extent that it can be used to create diagnostic products for use in patients’ own homes. “My own opinion is that will be the last thing to happen,” says Chang. “Obviously, when that happens it becomes a consumer product and a huge trilliondollar market. We are all hoping that will happen, but I don’t think it will be the first thing to be realised. I think what will happen is there will be certain screening tests – let’s say one of these point-of-care devices can screen for lung cancer, so when you go into the hospital, even in the developed world, and do a lab test, it could be included in the lab test as a low-cost screening test for lung cancer. So if you charge $100 for a lung cancer screening test, that would bankrupt the healthcare system or the patients would not be able to afford to pay, but if you charge $10, that becomes viable.”

Once the low-cost screening technologies mature, says Chang, it may then become possible to screen for multiple cancers. “Once that becomes a test of its own, we can start a company for early screening of certain diseases,” he says. “And when that becomes mature enough, then perhaps that company will sell an instrument to the patients so they can do the screening themselves. But you don’t need to use a device very often, so screening [happens] once or twice a year maybe. I think we have to think very carefully about consumer home-use diagnostic devices.”

But for now, the key issue is to solve the problem of manufacturing. “Chip manufacturing is the most important part,” Chang says. “And to make it down to the very small scale, micro and nano, that’s the biggest challenge. Research labs can afford to use a few very expensive chips, but to become a commercial product, low-cost manufacturing is the key. And we haven’t come up with really robust technologies that can make them reproducible on such small scales.”