The past two decades have seen a phenomenal investment in microtechnology in the biological sciences. But it is often difficult to discern what impact, if any, this diverse technology has had on the clinical laboratory. This difficulty is, in part, due to the almost transparent incorporation of microtechnology in large devices used in the clinical laboratory, or the lack of appreciation of the microtechnological features in many smaller devices.
The list of advantages of microtechnology is long and varied, but critics also point out the failure to bring about significant changes in practice and laboratory costs.
Five leading experts in the field of microtechnology drawn from academia and industry have responded to key questions about the potential scope of the technology and how it can impact the clinical laboratory. They have also provided their assessments of how point-of-care (POC)2 devices will change on account of this technology and what effect well-funded research in biowarfare may have on clinical laboratory practice.
What is the scope of microtechnology and what are its advantages?
Sabeth Verpoorte, as an academic with a well-established group working in this field, how do you respond to this question?
Sabeth Verpoorte3 : The scope of microtechnology with respect to fluid handling in clinical chemistry is extremely broad—and growing. Microchannels are uniquein that they are characterized by large surface-to-volume ratios and extremely well-defined and predictable solution flows. These systems enable exquisite control of fluid transport, and thus vastly improved (bio)chemical processing of samples. Advantages include reduced volumes, increased speed of analysis, and integration of multiple sample-handling functions into a single device for automated, hands-off operation. Microfluidics development has recently expanded into the biological realm, as this technology facilitates improved in vitro technologies for cell culture and analysis.
Allen Northrup4 : Microtechnology is a composite, single word that potentially covers a very broad field. If you take it literally, it’s a “technology” consisting of micron-sized parts, volumes, and physical dimensions. In clinical chemistry, many procedures utilize forms of microtechnology for fluidic handling and analysis. In general, “microtechnology” is perceived as including etched single-crystal silicon, glass, and micromolded or other micromachined components. However, for a variety of economic and practicality-based reasons, most clinical analyzers use “traditional” micron-sized components (e.g., extruded polymer tubing), but the scope and number of micromachined sensors and detectors is significant.
There is also a “Handheld Myth.” Handheld is great if you can commit to the use of your hands to operate the device, while not doing 1000 other things that need to be done with your hands. Lets face it, users are going to set them down somewhere. Also, other than costs (of course important) and an occasional benefit in detection, submicroliter vs microliter volumes present more technical issues (garden hose effect) than the advantages they may bring. So a cell phone–sized analytical device performing traditional analyses will be inherently much more difficult to produce cost-effectively than one the size of a computer. In any case, the potential return on investment (ROI) has to be taken into account. However, new markets can be opened up due to its size coupled with new capabilities (iPhone effect). Therefore making a device small and portable has appeal, but not at the expense of price and performance.
Paul Yager5 : Microtechnology today is a set of technologies that uses some form of microfabrication to create small parts. Originating in the semiconductor-based silicon microfabrication world, it grew in the last 20 years to encompass a set of technologies for very different purposes and fabricated in many different materials. One pole of the work is micro-electro-mechanical systems (MEMS), which focuses on mechanical and electromechanical devices that have features smaller than can be manufactured by conventional machining methods. The other pole is microfluidics, which enables the movement and control of small volumes of fluids, occasionally in integrated systems. The primary advantage that encompasses all of these technologies is operating at small sizes and working with small volumes. The physics at these small scales often allows operation based on principles that are less important at large size scales, so there are some unique capabilities of micro devices. For example, in the case of microfluidics, one can create novel devices based on transport by diffusion that have no equivalent at the macro scale. The one great success story so far has been the accelerometer, which is now ubiquitous, from cars to iPhones. There the advantages over all rival technologies were extreme, and the advantages of adding the capabilities of on-board accelerometers were very well exploited. This has not yet been seen in microfluidics.
Stephen Quake6 : I think that microtechnology offers some key advantages. First, the ability to integrate many steps or many assays on a single device is enabling a phenomenal degree of automation. Much like the first integrated circuits replaced rooms full of vacuum tubes, so too is biological automation being transformed by microtechnology. Second, microtechnology offers the ability to take advantage of unique physics and chemistry that is only available at the nanoliter scale; this has been demonstrated in countless ways in the academic literature, and has led to commercial products in both protein crystallization and DNA analysis.
James Landers7 : The scope of microtechnology is as broad as any of the subdisciplines of chemistry, physics, and engineering involved in developing it. It can be applied to any sector where there is a discrete advantage in scaling down size and weight while enhancing some aspect of the performance in a more automated format. This is not unidimensional, and can involve any number of parameters from reaction rates, mass transfer, and separation power, to analyte detectability, mass of sample required, volume of reagents needed, or mass of product desired. With the ability to seamlessly stitch both sample preparative and analysis processes on a single integrated fluidic device, the benefits should include the need for less sample, less reagents (therefore, lower cost), automation, smaller size, possible portability, and the potential for POC testing. One of the stealth advantages of microfluidic devices capable of sample in–answer out is that it is a “closed system.” This is ideal for clinical analysis where the technologist needs to be isolated from the potentially hazardous (e.g., HIV) sample, but also a necessity in forensic DNA analysis where the sample should be isolated from the analyst to avoid contamination. It is these characteristic advantages that will lead to application not only in forensics and diagnostics, but also with prognostics, food safety, national security, environmental analysis, and agricultural biotechnology.
One of the unspoken expectations with microfluidic technologies is that “cost of analysis” will decrease. This is not an unreasonable expectation, primarily because fluidic integration and automation mean less technologist labor, and tech time is money. However, it also assumes a reasonable cost to the user for the microfluidic device. As the development effort for microfluidic systems has ramped up, we have witnessed two orthogonal approaches: simple, cost-conscious design/fabrication vs the development of complex, powerful, and elegant micro devices that fully exploit microelectronics industry tools. History has shown that the notion of “complexity first, then reduce cost” is risky and rarely provides a viable path. In fact, engineering schools expressly teach the opposite, as do the best business schools in the country. It will be interesting to see how this plays out.
What analytical problems could microtechnology solve?
Paul Yager, you and your team have been prolific in publishing papers showing new structures and microfluidic platforms. What current problems will this technology solve?
Paul Yager: Microfluidic devices are, by definition, particularly good at operating with very small volumes of fluids. As such, they are well suited to the analysis of small samples, and controlling small volumes of reagents. This is particularly advantageous for condensing many parallel processing elements into a small volume to enable highly parallel chemical or biochemical processes like high-throughput DNA analysis or drug screening. It is advantageous for POC instrumentation, as long as the cost of disposables can be kept low. The inherent ruggedness conferred by reducing the inertial mass of the microdevices allows instruments to be fabricated that can survive and even operate in environmental conditions that would destroy conventional macroinstrumentation. This is particularly true if the fabrication materials are neither massive nor brittle. In addition, an important feature of microdevices has been conferred by their ability to rapidly remove heat from reacting solutions, thereby allowing the devices to safely support chemical reactions so exothermic that they would cause explosions in larger systems.
Sabeth Verpoorte: The small volumes available in microchannel networks mean improved performance for many established analytical approaches. One now-classic example of this is microchip electrophoresis, in which chip technology has allowed the reduction of separation times from minutes to seconds and even milliseconds. The peripheral system required to execute separation on the chip can also benefit from microtechnology, with development of handheld instruments as a result. One intriguing consequence of this is the application of microchip electrophoresis to new problems, such as POC analysis of Li+ in blood (www.medimate.nl).
Other analytical problems that microtechnology will help to solve are those requiring high-sensitivity analysis, such as the detection of low-abundance proteins in biomarker discovery. Integration of all sample processing steps into a single device means no loss of analyte due to transfer between containers.
Allen Northrup: The most likely problem that will be solved will be associated with the detection of the sample, inasmuch as the analytical volume benefits from being smaller. It will not solve the macro, real-world to microvolume interface challenges. If low cost and performance are maintained or improved, then it offers the solution for further distribution of analytical systems—doctors’ offices, local clinics, homes. This, in turn, opens up opportunities for personalized medicine and field-based analyses, which potentially increase the benefits of diagnostics and therapeutics.
Stephen Quake, you have been successful in transferring technology from an academic setting into a biotech company. What are some of the examples of this success?
Stephen Quake: I see many of these through the research of customers of Fluidigm, a company I cofounded (full disclosure: I have a financial interest). There is now a string of papers in Science, Nature, Cell, and several other journals describing protein structures in which this microfluidic technology played a key role in discovering crystals. There is also a growing literature around the use of our company’s PCR arrays for the use of single-cell analysis, in areas ranging from cancer to developmental biology to environmental microbiology.
James Landers: With the exception of clinical chemistry analyzers, many of the analytical processes in clinical chemistry laboratories are time-consuming and labor-intensive. Serum protein analysis still involves agarose gel electrophoresis and gel-based immunotyping for gammopathies. For molecular diagnostics, the orthogonal nature of DNA extraction, PCR amplification, and size-based DNA fragment sieving for separation and detection creates significant challenges for seamlessly integrating these processes into a single platform. While robotic systems have evolved for the efficient batching and high-throughput modularization of the more macroscale processes both in clinical and in forensic labs, a closed system, microfluidic (not macrofluidic), sample in–answer out solution has evaded us. With most of the fundamental aspects of the individual analytical processes worked out at the microscale, integrated microfluidic systems are beginning to surface—for ions, metabolites, infectious agents, and human identification—and these begin to deliver what microfluidics promised almost two decades ago. The advantages of such systems are obvious. Microscale volumes should require less sample and reagents—this should, in turn, decrease the cost of the test. This assumes a cost-effective microanalytical system that is single-use and disposable. While one can imagine the positive impact more cost-effective testing would have on national healthcare, it is questionable whether this would be reaped in an HMO-driven system. Cleverly engineered instrumentation to drive the miniaturized analytical systems would give way to more portable instrumentation that, upon further reduction, could become handheld and field-portable for almost any application. Reduced size and portability spells POC testing in clinical settings and opens the discussion on testing in physician offices or the ER rather than in central laboratories.
Why has commercialization of microtechnology been so slow?
Sabeth Verpoorte: A number of factors have played a role, some of which are purely external. For instance, the downturn of the stock market in the early 2000s leading to the burst of the technology bubble adversely affected microfluidics companies at the time. Another factor that may have played a role is the wait-and-see attitude that many larger companies in the analytical sector, but also pharmaceutical and diagnostic sectors, have adopted. Commercial successes like the Agilent Bioanalyzer 2100 (DNA/RNA/protein analysis) have generally depended on a strong joint venture between a chip application provider and a large end user with an established market of customers. End-user involvement from the early stages of an application development is thus perhaps not an absolute guarantee of success, but it can certainly help avoid problems that smaller companies may have in understanding potential customer needs.
Allen Northrup: For microtechnology to become a commercial success in any application, it must show significant improvements (performance or cost) compared to the current technology in use and the complexity of manufacturing it. If microtechnology is simply a result of shrinking components, then it could be argued that it is already very commercially viable; but as a goal unto itself, it will not be successful unless it solves a critical need and is cost effective. There are plenty of microtechnology-manufacturing-based components in the form of sensors. For commercially viable fluidic devices to derive from microtechnology, it will depend on the cost/performance parameters.
Paul Yager: So far nobody has found a set of devices that are of critical importance to large groups of customers that absolutely must be performed using microfluidic devices or systems. In research, the entrenchment of the 96-well plate has caused a huge investment in that platform, and that is difficult to dislodge. In diagnostics, the low-cost lateral flow technology is advancing rapidly, and is fully entrenched. In addition, the lack of common standards and interchangeability, and a large number of patent fights, have not helped matters. The largest disappointment was certainly the fact that, while researchers showed remarkably clever microdevices, they almost always required extensive and very “macro” support systems of tubing, pumps, and power supplies. There is progress today, and many microfluidic subsystems are finding their way into commercial and precommercial systems. One can argue that all new technologies go through a boom-bust cycle, and microtechnologies are just about to hit the rebound phase. When much of the laboratory research work is done using microfluidic devices and systems, the commercial applications will follow.
Stephen Quake: I don’t think it has been that slow—people are just impatient! It’s been a solid two decades of commercial development and one now sees a number of clever products on the market, many of which are selling well. Remember that the idea of a PDA was introduced in a Star Trek episode in 1966, and it took a good four decades for us to get Blackberries and iPhones in widespread use.
James Landers: If we assume the very early 1990s spawned the field of microfluidics, this field is approaching the end of its second decade. The basis for the “slow” development and commercialization of this technology is multifaceted, but over-hype, inadequate venture capitalization, and late focus on system integration are all likely contributors. This latter point is an interesting one in light of the fact that early work from Widmer’s group at CIBA in Basel in 1990 coined the term “total chemical analysis system,” where “total chemical analysis” immediately infers system integration. The analysis speed advantages of carrying out separations in the microfluidic format were obvious, and while the 1990s saw analytical separation and detection (primarily of DNA or proteins) advance tremendously during this time, it became clear that expediting the separation only brought incremental improvement to the overall analytical process in real-world applications. Sample preparation was the bottleneck; we were slow to engage the true power of microfluidic systems by integrating all functionalities in a single device. Efforts to microminiaturize sample preparation processes increased and fledgling attempts to integrate multiple processes were seen. However, it wasn’t until the early 2000s that microfluidic flow control was addressed—an “engineering” problem at heart—and valving approaches emerged that allowed the community to begin to respond with a bona fide effort on creating “integrated systems.” It is not clear whether it was immiscibility at the interface of analytical chemistry and systems engineering, or a lack of clarity that integration problems were truly rooted in macro-to-micro interfacing and flow control. Either way, delays in addressing microfluidic systems as “total analysis systems,” as Widmer and colleagues envisioned it almost 20 years ago, has likely delayed the progress in this field. As for those in the application trenches, my sense is that they must see the total integrated microfluidics solution that can be plugged into (or replace) their workflow—absent that, enthusiasm will be lukewarm at best.
In a completely orthogonal thought-stream, one could ask, “Has it really been slow?” And, “If so, relative to what?” The microelectronics industry evolved over three labor- and funding-intensive decades to arrive at a fully matured technology that has revolutionized almost every aspect of life as we know it today. Chronologically, microfluidics is only halfway through that time course and still evolving. One could argue that the breakthroughs we saw with valving in 2002 were the linchpin for creating truly integrated fluidic systems, and in that sense, “we have just begun.” In addition, microelectronics saw trillions of dollars invested into it or supporting technologies between 1980 and 1995; if microfluidics were to see a miniscule fraction of this, “slow” would no longer be part of the lexicon, and we would have already witnessed the development of a few key integrated systems (killer apps) that would have opened the floodgates to broader development, by illustrating to the end users that such microsystems hold clear-cut advantages.
Will the microtechnology applications in biowarfare detection be the stimulus to clinical applications?
Allen Northrup, you played a pioneering role in developing microtechnology and commercial and government systems in the fields of biowarfare detection and clinical diagnostics. How do you view this question?
Allen Northrup: In general, biowarfare applications are less sensitive to cost issues, allowing new unique applications to be investigated. This could be a stimulus to clinical applications, if: the technology significantly solves an analytical problem, it is extensible to a clinical solution, and it does not increase the cost of the analysis. Clinical applications are under more cost pressures (by orders of magnitude) than are biowarfare applications. Typically one of the first things an investigator will ask regarding the viability of a technology is how it will compete with current products in terms of cost. Insurance reimbursements vs clinical utility vs cost will also have to be considered.
Sabeth Verpoorte: In part, sure. Certainly, chip-PCR detection of pathogens and similar applications have enjoyed substantial funding from defense agencies in the US, and there have been some wonderful examples of (portable) instrumentation that have come out of these efforts. However, there are an increasing number of groups worldwide considering the combination of microtechnology and molecular biological techniques for the study of disease mechanisms and other health issues. This research, which is generally not funded by the military, will result in new diagnostic and therapeutic applications of microtechnology.
Paul Yager: Fifteen years ago, many of us thought that this would be the driver to widespread commercialization of this technology. This has proven not to be the case, in large part because laboratory robotics was just becoming a mature and entrenched technology at that time. Today the biowarfare scare has diminished, and there is less funding in this area. When the Department of Defense sees that it has come to rely on microfluidics, then the technology’s future will be assured, but that hasn’t happened yet.
James Landers: If history has any bearing, quite possibly, as national security applications are powerful drivers of technology development. It is important to realize that the birth of microelectronics was largely driven by the race-to-space (NASA) and by military/civil needs. It was during World War II that the need for reduced size, weight, and power of military electronic systems became apparent. With more complex military electronic systems, the increase in size and mass eventually became impractical, especially in aircraft, and this drove the development (and forced the funding) of smaller, lighter, and more efficient electronic circuits. Layered on top of this was NASA’s need to have smaller, lightweight, and more powerful circuits for the space-limited spacecraft that were evolving. Computer science and computer engineering were shaped by military funding of the early stages of digital computing, where most of the basic technologies for digital computing were developed through a program focused on the development of an automated radar shield. We reap the benefits of this R&D now in our daily lives, with access to instant communication, GPS systems that guide navigation with pinpoint accuracy, and complete personal musical libraries at our fingertips. Two other related issues factor into the “clinical vs biowarfare detection” question: impediments to acceptance and funding. While clinical applications appeared to be the first to reap the benefits of microtechnology, FDA 510(k) approval and other barriers to adoption may hamper it. In contrast, biowarfare detection is narrower in scope and has focused application and unequivocal (less stringent) parameters for adoption. With funding, venture capitalists may have been hesitant to come to the table with clinical application of microfluidic technologies, partially because this sector is heavily regulated, dominated by a few behemoths, and represents a healthcare market that is cost-conscious. On the other hand, biowarfare detection saw serious capital investment come into play with the advent of 9/11.
Stephen Quake: I think the basic research funding from agencies like the Defense Advanced Research Projects Agency (DARPA) has been a tremendous stimulus toward the development of clinical applications, although in the commercial sector the connection is more tenuous.
Will microtechnology lead to an expansion of point-of-care and a reduction in central laboratory–based testing?
James Landers, your research and academic involvement spans engineering and pathology. What perspective do you have on this issue?
James Landers: The potential benefits of “decentralization of central reference labs” have been debated for some time. Whether or not there is evidence that this is occurring is not clear. It is not difficult to see how a new, turnkey clinical diagnostic technology might induce some shift from central laboratories to smaller testing laboratories (or even to physicians’ offices), as the direct benefit would be decreased turnaround time. However, a number of variables are involved in predicting the extent to which this would happen in the short term, where the end user is the physician. The move in this direction largely depends on the willingness of physicians to adopt a new technology that will likely provide the same result, but more rapidly and, perhaps, in a more information-rich manner. There is a definite “activation energy barrier” to acceptance of the same data in a different format; the fact that the results from capillary electrophoresis of serum protein (an HPLC-like trace) had to be back-converted into a rendering of a stained agarose gel is proof that change is difficult. When the physicians see the benefit of new microfluidic technology, and call for it to be readily available, there will be little barrier to its adoption in central or decentralized laboratories. In terms of microfluidic technology finding a place in physicians’ offices, the jury is still out, for as many arguments exist for and against this. The discussion becomes even more interesting when the end user is no longer the physician, but the patient.
In an interesting twist, the forensic sector may be the forerunners on this front. There is clear-cut interest in taking DNA analysis out of purely forensic laboratories and placing it at the crime scene or at the booking station. While this could only be done with very rigorous quality control and is fraught with practical/policy challenges, one can’t but wonder whether this foreshadows the trends in clinical diagnostics. The twist would be that the clinical diagnostic sector would be following a trail forged by forensics.
Allen Northrup: In order for medical care to extend into the personalized format (i.e., at home or a physician’s office) it must be small, fast, simple, foolproof, and able to fit within the economic constraints imposed by the insurance providers. If this can be achieved it could allow for more pharmacogenomic-oriented tests and therapies, which can improve treatments and outcomes. If microtechnology contributes to this end, then it would help expand medical care into the local, at-home or physician’s office environment.
Sabeth Verpoorte: It is already doing so. Just consider the i-Stat portable clinical analyzer, which has made inroads in the POC area in many hospitals worldwide. Other developments in areas such as glucose monitoring (work being carried out in my laboratory and others) and microchip electrophoresis for Li+ analysis in blood (mentioned above) are emerging in increasing numbers in recent years, and will have an impact on the POC market. If microtechnology can promise lower costs per test, this trend is assured. One area in which microtechnology has attracted particular attention lately is POC diagnostics for the Third World. In this case, the issue is not a reduction in central laboratory testing, but just providing basic diagnostic testing for patients who have not ever had this privilege in any form.
Stephen Quake: Difficult to say, I can argue it both ways and perhaps we will see both. In other words, perhaps POC testing will expand for some areas, but with all the new measurement assays being developed it doesn’t necessarily have to be at the expense of central laboratory testing.
There is a good economy of scale for central laboratories and I suspect that as more complicated assays get put into practice there will still be a growing market for central laboratories.
Paul Yager: Yes, absolutely, although it may not be the first-generation microfluidic devices that win the race. For example, paper is a wonderful analytical material, and capillary “pumping” has the ability to avoid much of the support equipment needed to push fluids through tubing.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors’ Disclosures of Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: J.P. Landers, MicroLab Diagnostics.
Consultant or Advisory Role: None declared.
Stock Ownership: J.P. Landers, MicroLab Diagnostics; P. Yager, Micronics, Inc. (stock options only).
Honoraria: None declared.
Research Funding: J.P. Landers, National Science Foundation, NIH; P. Yager, the Bill and Melinda Gates Foundation, NIH.
Expert Testimony: None declared.
Other Remuneration: J.P. Landers, Society of Analytical Chemists of Pittsburgh.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
1 Peter Wilding, Professor Emeritus, Department of Pathology and Laboratory Medicine, University of Pennsylvania Medical Center, Philadelphia, PA.
↵2 Nonstandard abbreviations: POC, point of care, ROI, return on investment; MEMS, micro- electro-mechanical systems; DARPA, Defense Advanced Research Projects Agency.
↵3 Sabeth Verpoorte, Chair of Analytical Chemistry and Pharmaceutical Analysis, Groningen Research Institute of Pharmacy, University of Groningen, the Netherlands.
↵4 M. Allen Northrup, CEO and Chief Technology Officer, MicroFluidic Systems, Fremont, CA.
↵5 Paul Yager, The Hunter and Dorothy Simpson Endowed Chair of the Department of Bioengineering, University of Washington, Seattle, WA.
↵6 Stephen Quake, Lee Otterson Professor of Bioengineering and Applied Physics, Stanford University, Stanford, CA.
↵7 James Landers, Professor of Chemistry, Professor of Mechanical Engineering and Associate Professor of Pathology, University of Virginia, Charlottesville, VA.
- © 2010 The American Association for Clinical Chemistry