Nanogen Stakes A Shape-Shifting Market

Newly abundant genomic data is driving formation of all sorts of technology-based businesses. But markets for new tools are already chaotic and competitive, in part because customers know they've got many choices. Nanogen is in the thick of it. Pharmaceutical researchers weren't so interested in the firm's low-density gene chips, so the firm is now focusing on the clinical diagnostics market where managers believe flexible, accurate NanoChips will be better appreciated. The trouble is, the long-foretold market for molecular diagnostics still barely exists. On one hand, market immaturity spells opportunity for Nanogen as an early entrant, but it also means the company has to bushwhack a new path for its technology. It's not easy. For now, Nanogen is marketing its system to researchers in clinical diagnostic labs, university hospitals and government institutions-scientists at the cutting edge, who may become key content developers. The firm is also working to better serve drug makers. The company's customers display little loyalty yet: they're eager to try other new technologies too. Nanogen is betting that the superiority of its system will win hearts and minds as the market for molecular diagnostics takes shape.

Nanogen has spent years developing gene-chip technology. Now it's trying to launch a system in a still-inchoate market that is chaotic, competitive and alluring.

Deborah Erickson

• Many companies would like to follow the pioneering path of Affymetrix, and become key suppliers of tools that enable molecular biological research.

• Newly abundant genomic data is driving formation of all sorts of technology-based businesses. But markets for new tools are already chaotic and competitive, in part because customers know they've got many choices—at hand and on the way.

• Nanogen is in the thick of it. Pharmaceutical researchers weren't so interested in the firm's low-density gene chips, so the firm is now focusing on the clinical diagnostics market where managers believe flexible, accurate NanoChipswill be better appreciated.

• The trouble is, the long-foretold market for molecular diagnostics still barely exists. On one hand, market immaturity spells opportunity for Nanogen as an early entrant, but it also means the company has to bushwhack a new path for its technology. It's not easy.

• For now, Nanogen is marketing its system to researchers in clinical diagnostic labs, university hospitals and government institutions—scientists at the cutting edge, who may become key content developers. The firm is also working to better serve drug makers.

• The company's customers display little loyalty yet: they're eager to try other new technologies too. Nanogen is betting that the superiority of its system will win hearts and minds as the market for molecular diagnostics takes shape.

The promise of genomics is still largely that—promise. Scientists are only just starting to draw connections between particular genes and mutations in them, and clinical realities such as the presence, absence and progression of disease, and patients' sometimes quite varied responses to drug therapies. Now that researchers can for the first time access a full draft of the human genome, studies of genomic data should, theoretically, become easier and more fruitful.

Dozens of companies have rushed in, and continue to rush in, to develop the picks and shovels that will help researchers mine the riches of the genome. The biggest toolmaker to date is Affymetrix Inc. It was the first firm to adapt computer-industry techniques to create so-called "gene chips," arrays of DNA on silicon chips. Affymetrix has focused on improving the density of its chips to handle more and more genetic material. Now lots of firms make DNA arrays and chips, while others are devising entirely new ways of studying the genome, the proteome, cells, databases—all sorts of avenues that may lead to new understanding of the human body.

Given the abundance of material to study, and of tools for doing so, there really is no way of knowing which research methods will come to be the preferred ones—let alone which may go on to have impact beyond the realms of research. Indeed, it's possible that Affymetrix, one of Alejandro Zaffaroni's first start-ups, could be obsoleted by another one testing a far broader set of biomarkers—SurroMed Inc. , founded in 1998, but just now beginning to show itself. The swirl of options and uncertainties is complicating matters for all toolmakers.

No one understands this better than Nanogen Inc. , a research instrument-maker founded in 1993 when automated DNA sequencers were not commonly available, public stores of genes were low, and the mania for SNPs (single nucleotide polymorphisms) had yet to take over researchers' consciousness. The company was founded to commercialize the concept that electronics could enhance molecular biology. Nanogen spent years developing technology, eventually devising a chip that researchers could customize themselves, to study small numbers of genes and mutations. The chip and a system for reading it were launched in mid-2000, smack into a world where most researchers are demanding tools to deal with ever-greater volumes of DNA.

Nanogen believes scientists now crazed for quantity will want to hone in on small numbers of genes eventually, and that its NanoChipswill become widely appreciated—including by pharmaceutical researchers, who now show little interest. In the meantime, the firm is marketing to researchers whose tasks make them more likely to appreciate accuracy and flexibility over volume. People engaged in clinical diagnosis of gene-based disorders don't need to look at lots and lots of genes. They do need precise tools to look at specific mutations known to correlate with diseases such as cystic fibrosis and sickle-cell anemia. Nanogen thinks it has what they need.

Today, gene-related disorders are diagnosed via well-established but tedious lab methods such as DNA sequencing and polymerase chain reaction (PCR), so new methods like Nanogen's that promise to speed testing are at least being received with interest by clinical diagnostic researchers. That's the advantage of this sector. The disadvantage is that very few gene-based tests actually exist today, because scientists simply haven't made many connections strong enough to support a test. This state of affairs is changing as molecular biological research advances; indeed many industry observers say the entire field of clinical diagnostics will soon be revolutionized. Nanogen's managers hope their system will be one of the tools that forms a bridge between clinical research and new ways of practicing clinical medicine.

But no one can predict when clinical research will start turning up genetic data compelling enough to be the basis of new diagnostic tests. Even the form of that data is in question—will the new diagnostics that find market acceptance be based on DNA, RNA, proteins or chemical markers? Perhaps new algorithms will assess many such variables simultaneously, before predicting disease or suggesting a certain type of therapy or drug dosage over another. Growing numbers of firms are promoting new diagnostic technologies and wooing researchers who may be key content developers as well as gatekeepers to the clinic. Nanogen is putting a commercial stake in the shifting ground, insisting that its technology is flexible enough to evolve as the market does, however it does.

Insight into Electronics

Nanogen was founded to pursue the vision that electronics could enhance molecular biology. At first, the company's goal was to create a new research instrument to facilitate various experiments involving proteins and enzymes, and DNA hybridization. The idea drew on a revelation that had come to Mike Heller, PhD, Nanogen's founding scientist, as well as to other researchers in the field in the early 90s. They saw the principles that underlie well-known electrophoretic research methods such as chromatography in a slightly different way, and this sparked an insight. "Instead of moving fluid, as in chromatography, it was realized that you could move molecules," explains Bruce Wallace, PhD, Nanogen's VP molecular genetics. It wasn't easy to design the NanoChip, he says, "but once we understood the concept—that we could bring molecules where we wanted and let them interact there, it all started to fall into place."

The company began in 1993 to develop a microchip like those common in the computer industry for biological purposes. Each chip would have multiple test sites in the form of individually wired electrodes that would function something like separate little test tubes. The set-up would let the company switch on current where and when doing so would assist a step in the diagnostic process. The system takes advantage of the naturally occurring positive and negative charges of DNA and RNA. Switching current on helps attract DNA to the test site and draw it tighter to the surface, to enhance binding of whatever is intended to bind. Reversing the current encourages unbound DNA to move away from the area.

Nanogen's chips are distinguished not just by electronics but by specific chemistry that takes advantage of the wiring, Wallace declares. The on-chip environment in which sample meets probe is deliberately designed to favor hybridization only where current is. Consequently, he claims that many of the undesirable interactions that frequently occur in passive DNA hybridization assays—such as the intra-interaction of sequences within a DNA molecule—are minimized. In its system, Nanogen utilizes a low-conductance buffer, histadine, which has no net charge and thus doesn't support interactions between DNA strands. But when the electronic condition near a binding site is changed, and histadine takes on a positive charge, then the chemistry does favor hybridization of matching DNA strands—but only near the electrode. "We provide the charge that allows the zipper to close," Wallace asserts.

Given that it started up in 1993, some industry observers question just why it took the company so many years to develop NanoChipsand the reader system for them. Company founder, chairman and CEO Howard Birndorf admits that the small firm "may have gotten a bit a carried away, over-engineering the system. And we could have made things cheaper." If it had been more experienced, it would have moved faster, perhaps not gotten sidetracked into developing different sorts of technology. But the company has learned valuable lessons along the way, he contends—including how to seek out help.

Now Nanogen has the corporate understanding developed in the course of creating several generations of its system from scratch—and it's also got a powerful partner to help it confront and move past old problems. In January 2000, Nanogen signed an agreement to work with Hitachi Ltd. [See Deal], and now Birndorf says his company is "positioned to engineer out costs and engineer in quality." The agreement was broadened in July of that year, and now Hitachi has ten years of non-exclusive rights to manufacture all the components of the workstation, as well as rights to exclusively market in Japan and other Asian countries the workstations it manufactures. Hitachi also has non-exclusive rights to distribute NanoChip cartridges in these markets.

Now that it's working with Hitachi, a company that makes thousands of different instruments and which has collaborated previously with powerhouse diagnostics players Roche and its subsidiary Boehringer Mannheim Corp. , Nanogen believes it's definitely on the right track. "They've built all kinds of research machines; they're experts," Birndorf points out. Because of this partnership he believes that Nanogen will have iterations of its system that offer more functionality for less money—not overnight, but over time. Already, management says the firm has realized some cost reductions, and is gaining understanding of customer needs and how to meet them.

In June 2000, Nanogen launched its NanoChipMolecular Biology Workstation, an automated system for evaluating SNPs and short tandem repeats (STRs) in samples of patient DNA. The workstation consists of four substations: a loader that handles fluid and electronically addresses probes on the chips/cartridges; an array scanner that uses a laser to read test results in under two minutes per chip; a so-called fluidics station that automates movement of test samples and reagents into the cartridges; and software and hardware that control the system. The chips ship without any DNA attached. That way, customers can put down whatever they want to study, in small quantities. They can, in theory, test 96 patient samples for one genetic mutation, or put down 96 mutations and test one sample against that. But duplications and controls reduce the number of viable sites, and are recommended by Nanogen. So the chips as they exist and are utilized now actually viably test some 10 to 20 genes or SNPs.

The format is markedly different from Affymetrix chips, which are shipped to customers pre-spotted with massive arrays of DNA—up to 60,000 spots per chip these days. Researchers apply samples, for instance of diseased tissue, to the chip's surface and then the entire thing is reacted with heat or chemicals to see what has bound where. Nanogen's chips can be tested sector by sector, so no DNA is "wasted."

Nanogen's managers thought that all sorts of scientists working with DNA would like the fact that NanoChipscould give them more flexibility, more focus and a higher degree of accuracy than they could get from Affymetrix's expensive, pre-fabricated chips. They thought researchers would want low-density chips to do in-depth studies of genes and mutations they'd already identified with far denser arrays. But Nanogen soon found the Big Pharma outfits it called on didn't much appreciate its low-density chips, geared for processing at low volumes. Drug makers are still busy using massive arrays to look more broadly at which genes are present at a given stage of disease. The rage for high-throughput systems hasn't subsided yet; it may not ever.

Eventually, Nanogen's president and CFO Kieran Gallahue thinks pharmaceutical researchers will want to hone in on fewer genes. But instead of trying to convince drug makers to do that before they feel ready, the firm is hustling to give the market another version of what it's already comfortable using. By the end of 2001, Nanogen hopes to introduce a gene expression chip that will let potential customers do the sorts of larger-scale experiments they've learned to do with Affymetrix chips on a NanoChipworkstation. If the company can sell or at least place its devices in pharmaceutical research labs for that purpose, proximity may encourage researchers to try the sorts of focused tests Nanogen developed initially. Work on the new, expanded chip is moving ahead, but in the meantime, the company has turned its attentions elsewhere.

Now Nanogen is pitching its chips to clinical diagnostic researchers as the best means of studying new genetic markers, as well as the ideal platform for converting those research findings into diagnostic tests that will have a role in clinical care. Gallahue thinks companies like Affymetrix, Applera Corp. 's Celera Genomics Group , Sequenom Inc. and Orchid BioSciences Inc. have approached genomics too much as an industrial process. Their high-density chips may be suited to large-scale discovery efforts where broad brush strokes can satisfy scientists, but he contends that they won't be able to shift from the lab to the clinic, as he expects NanoChips will.

"Clinical research is where genomics gets personal, and attributes like accuracy become more important," Gallahue says, noting that he doesn't think other chip-based systems on the market today can offer the kind of accuracy and flexibility that Nanogen can. That's not so surprising, given that the market for chips has so far been driven by the demands of pharmaceutical research who have prized other traits. "There are lots of technologies out there being used for different research purposes, where having accuracy of 80% or 90% is fine," Gallahue says. But that will never be good enough in a clinical lab, where false results are a potential legal liability.

Where's the Revolution, Already?

Nanogen is far from the only company championing the notion of a diagnostics revolution. For two decades and more, futurists and physicians alike have predicted that all sorts of diseases and susceptibilities and tendencies will be diagnosed on the basis of individuals' genetic material. The reality has lagged sorely behind the vision, however, primarily for lack of markers. Today there are still precious few tests for human genetic mutations, such as those associated with cystic fibrosis or sickle-cell anemia, and the current gold standard method of detecting these tell-tale variations is good old-fashioned DNA sequencing.

The last decade or so has seen only a handful of new gene-based tests come into common use: some identify bacteria such as Chlamydia and Gonorrhea and others viruses such as hepatitis B, hepatitis C, HIV and cytomegalovirus (CMV). Only a few new cancer markers with therapeutic impact have come along: there's Brca 1 and 2, and recently a test for Her2neu, although apparently that mutation was known 20 years ago. Demand for these tests has increased as the information they provide has come to be linked with treatment options: before antivirals capable of controlling viruses were invented, it didn't much matter what kind of a viral load a patient carried. Still, aside from these tests and a few more for clotting disorders, or the rare genetic condition known as Fragile X, there simply aren't many molecular diagnostic tests in current use.

Nanogen's Birndorf insists the tide has turned and that new genetic markers of disease, now being identified in labs all over the world, will at last be making their way into diagnostic panels. "I've been in the diagnostic field for 20 years, and all along we've thought molecular diagnostics would come. Well, the time is finally here," Birndorf declares. Nanogen doesn't have any new markers itself, however, nor does it have in-house capacity to get them. It has deliberately concentrated on its platform instead of chasing content.

To get access to new markers, Nanogen has recently signed development agreements with researchers at academic institutions and government labs. The firm's collaborators are evaluating genetic mutations in association with cardiovascular disease, immune function, cancer, drug response and infectious diseases. If and as the work bears out, Nanogen hopes to license rights to the markers and create kits to run on its workstation. In the meantime, the firm will be happy to see researchers using the NanoChipworkstation to investigate mutations as new possibilities for testing. At least they'll be consuming reagents, and getting to know and hopefully like the system.

Seeing Researchers as Content Developers

"Clinical researchers are the gatekeepers of the clinical diagnostics market," Gallahue declares. These people are working at the cutting edge of science to establish connections between genes and disease, processing patient samples at low to medium volumes that the NanoChipworkstation can handle. "We want them to be our content developers," Gallahue says. Some researchers are saying maybe to that proposition.

By forming relationships with researchers, Nanogen hopes it is creating paths to "content," the sine qua nonof new diagnostics. The term content refers to the physical evidence of connection between disease and DNA, RNA and proteins—really any testable substance that researchers can get a handle on.

For now, one of the main ways Nanogen's research collaborators are using the NanoChipsystem is to investigate how particular SNPs are associated with a population of individuals that have a common condition. A research group at the Mayo Clinic in Rochester, MN, for instance, has been using the company's chips to look at SNPs to assess the way children with leukemia metabolize a particular drug.

Mayo researchers, following the lead of peers at St. Jude Medical Inc. , have observed that certain patients respond well to a basic dose of drug, while some seem to need more and others taking the same dose are dying from adverse drug reaction. They've found that difference in response pattern seems to be linked to variations in the gene for an enzyme known as TPMT (thiopurine methyltransferase).

Clinical investigators do different sorts of experiments than scientists working within drug companies, Bruce Wallace notes. For instance, people in university labs tend to know key information about the patients from whom they've gathered samples, and so they can sort it instead of simply looking for random mutations. In the study at Mayo Clinic, researchers could put down on NanoChipsDNA from all non-responders, or cluster samples from children that were low metabolizers separately from those who were high metabolizers. Nanogen's detection system, like many others, uses fluorescent to tags to reveal hybridization of target and probe. Since Mayo sought only two SNP varieties, the diagnostic glowed red or green depending which variation a patient had, and that visual system, interpreted by Nanogen's reader device, helped researchers make the connections between phenotype and genotype. This particular study was ideal for Nanogen's system, Wallace points out, because it allowed the researchers to customize their array experiments to ask very specific questions.

Although it is conceived and packaged as a system, the NanoChipworkstation is really just a technology, another way of doing research. This reality could prove to be a saving grace for Nanogen, at least in the near term, explains Susan Morgensztern, a Seattle-based consultant to the diagnostic and biotechnology industries. As long as researchers are putting down their own DNA on a chip, Nanogen can sell chips and instruments to researchers, and even describe protocols to test for specific genes or mutations, long before needing FDA approval. "But as the content becomes clinically relevant, easy-to-use systems, reimbursement and FDA approval will be necessary to penetrate the widespread laboratory market," Morgensztern declares.

Leveraging Scientific Credibility

As Nanogen markets its low-density, low-volume system to researchers at universities, government laboratories and teaching hospitals, the company intends to leverage scientific credibility it has worked hard to build. Nanogen started publishing scientific papers several years ago, and they seem to have had the desired effect of engendering respect.

The Children's Hospital of Philadelphia became the first institution to join Nanogen's Development Site Program in October 2000, agreeing to provide expertise in exchange for early access to the firm's technology. By that time, researchers at the university had been aware of Nanogen's technology for some time. Paolo Fortina, MD, a researcher at the hospital, says he and his colleagues in a journal club were all "impressed" when the respected journal Nature Biotechnology featured articles written by Nanogen researchers on its cover in June 1998 and again in April 1999. The earlier article described how the firm's microfabricated bioelectronic chips had been used to prepare, hybridize and analyze DNA and RNA from E. coli. The later article discussed the chips' role in SNP-detecting experiments.

Fortina says he appreciates that Nanogen's technology lets him customize chips and perform parallel analyses. In the same unit time it previously took to do one test, he can test many samples for, say, a genetic mutation that causes deafness. "That time savings is a major advantage," he declares.

Fortina and Eleanor Pollak, MD are currently using the NanoChipsystem to devise new diagnostic tests for genes which carry mutations already shown to be associated with thrombosis: Factor V, Factor VII, MTHFR, and protein C. All of these genes carry SNPs shown to correlate with increased or decreased risk of a cardiovascular incident. Over the past three or four months, the investigators have been putting down on NanoChips samples from many different individuals, and control sequences, and testing them for SNPs or other mutations in relevant genes.

Where does the work stand today? Fortina says they've been running tests on DNA of individuals who've already been genotyped, to confirm that results generated through the NanoChipinstrument are as accurate as those derived through the gold standard method of genetic diagnosis. "DNA sequence analysis is and will be, for a long time, the standard to which any instrument is compared," he declares. So far, he says Nanogen's system has been showing 99.9% accuracy, but emphasizes that is not a final statement. "It's too early to make a final conclusion. We haven't been testing thousands of individuals—still only in the hundreds of patients. On other hand," he says, "I don't see on the market a chip-based platform—today—that is as good. But this is a fast-moving field, and there are plenty of rumors that other companies are developing exciting high-throughput instrumentation."

Test ‘Em and See

Nanogen continues aggressively asserting its technological superiority, challenging researchers to send the firm its toughest samples, and see for themselves what the technology can do. "We know it's risky, but we've got to prove we're the best," Gallahue declares, adding that 99% accuracy isn't going to be good enough for tests whose results, like the one for drug metabolism in children with leukemia, could determine the difference between life and death.

"Data takes the bull out of any discussion," Birndorf asserts, so that's what he intends to gather, confident that NanoChipwill come out looking good. He says many firms can do a reliable test for clotting protein Factor V, but not many have the technological chops to test for Factor VII deletion—another protein that clinicians want to look at in association with coagulation. Certainly, Birndorf says, not many technologies allow researchers to test for both sorts of mutations on a single platform. Most labs have three or four machines to do various sorts of genetic testing, he notes, adding that he'd like to see them have just one—the NanoChip system.

Management hopes that positive buzz about the workstation will spread as more people get their hands on it, see that it's robust and accurate, and start talking to other researchers about their experiences. Now that Nanogen is marketing mostly to academic researchers who publish and discuss their work, rather than pharmaceutical researchers jealously guarding drug leads and techniques for discovering them, Gallahue figures the scenario is more likely to develop as Nanogen wishes.

If and as acclaim builds in the research community, the company hopes content developers will come to see its platform as the ideal one to work with—in the near term for research, but more importantly later, when the time comes to develop commercializable tests. "Nobody knows where killer content is going to come from, but there's clearly going to be more than one source of it, and we can't do everything ourselves," declares Kieran Gallahue. The company anticipates tapping into some content for free, via the huge public troves of genes.

Birndorf figures there will be plenty more opportunities for Nanogen to partner as content is discovered—with its 13 current collaborators and others too. He's well aware that his firm's competitors are also approaching clinical researchers. Savvy academic scientists are keeping their options open these days, since every firm with an interest in the area is angling for a hook into viable genes and mutations. No one wants to sign away a potentially big discovery for too little reward, or to a player that can't commercialize a finding successfully.

Birndorf thinks the flexibility of Nanogen's platform gives it a marketing edge that help the firm catch hold of content too. "Just thinking that a gene mutation is going to be interesting to the market doesn't make it so," he asserts, pointing out that all the gene mutations researchers discover will have to be validated by some means. Since mutations occur in different forms, such as insertions, deletions and microsatellite DNA, Nanogen is betting that researchers will come to appreciate the ease of doing various sorts of validation experiments on one simple system.

But it isn't easy selling tools into a market as amorphous as the one for molecular diagnostics is now. Although Nanogen has made what it considers a "commercial" launch of its workstation, no research group has sworn allegiance to it and no tests specifically for it have been developed, let alone approved by the FDA.

Nanogen's system is going up against entrenched methods such as straight sequencing, RFLP analysis, PCR and other methods of amplifying DNA. It's up against other low-density chips made by firms such as Genometrix Inc. and BP Amoco PLC 's Vysis Inc. , and must also compete against Orchid's SNP-IT biochemistry and Cepheid 's DNA amplification and detection system, to name just a few. Scientists are interested in trying out all of these new methods.

Affymetrix has lately begun making the case for using its high-density chips for disease diagnosis, clinical researchers note. The chip pioneer has been explaining to a new market segment what sorts of experiments researchers might want to try. And it has been outfitting itself too. In early April, Affymetrix announced an agreement with GeneData AG , a private Swiss biotechnology company, to co-market software meant to enable enterprise-wide management and analysis of gene expression data from Affymetrix's chips [See Deal]. Just because Affymetrix hasn't done much in the clinical diagnostics market yet, doesn't mean it won't eventually. The same can probably be said for other firms who've been selling gene-based technologies to pharmaceutical researchers, but could come seeking leverage through diagnostics.

Reference Labs Do Applied Research

While Paolo Fortina and other clinical researchers work at university hospitals to develop new sorts of diagnostic tests, scientists at clinical reference labs are trying to apply new technologies to come up with better, faster versions of tests they already do—or would like to do, but currently send out to others with more expertise. Thus, reference labs are another customer group that Nanogen thinks it can interest in its system in the near-term.

Randy White, PhD, EVP, technical operations for American Medical Laboratories Inc. (AML), the third largest clinical lab in the US, with over $300 million in annual sales, explains that reference laboratories like his are in a certain sense at the top of the food chain for diagnostic testing. Requests for tests boil up from various strata of the health care system, from around the nation, adding up to high volumes. "The reason those tests are referred is because they're difficult to do, they're not automated, and it takes a very special group of talented folks to do them," he says.

If something like a Nanogen system can make a difficult test simple to do, demand for that test will jump up—but not necessarily to the benefit of the reference labs. Medium-sized labs and hospital labs that used to send the work out to a reference lab—or even discourage testing since every order would represent a cost—will eventually start doing those tests themselves.

"If I don't move forward, and stay at the competitive edge, my business shrinks," White declares. Somebody has to be involved in the commercialization of new tests, and White says he certainly will be. That's why he's been checking out Nanogen's system.

"If you're going to be at the top of the food chain, then you must be involved with the new technology. Today only specialized labs like us can test for Factor V Leiden, but tomorrow when some mid-tier lab can use Nanogen's machine to do that test, we'll have to be doing something else," White declares. Factor V Leiden is a gene relatively recently discovered to be associated with predisposition to phlebitis, he explains. White's group has been working to develop an easy-to-do version of the currently arduous test, and he says, "We know it works."

AML is also working to develop a test for a genetic disorder called hemochromatosis, an iron-storage disorder that causes the substance to crystallize in the joints and the liver, instead of being absorbed by the body. White says that nowadays, if a patient comes into a hospital with jaundice, joint pain and high iron values, and he really, really wants to know if he's got this disorder, then that hospital or mid-sized lab will refer that test out, as service. But that lab isn't going to be actively encouraging doctors to recommend genetic testing for patients that present with those symptoms, because it represents a cost.

If labs could do the work in-house, demand for hemochromatosis testing would certainly rise—and White says it should, since 4-5% of the population carries the gene for the disease. It's a painful, debilitating condition that can lead to patients needing liver transplants, which cost about $180,000 each. But people that are diagnosed early can be treated with phlebotomy (therapeutic bleeding) every three weeks and are basically fine, White says. To him, this is clearly a situation where genetic screening could save suffering and costs.

Like the deals Nanogen has with academic and government collaborators, its relationship with AML is a loose one that leaves room for competition. The giant reference lab has many collaborations ongoing with developers of new diagnostic technologies, so it's not tying itself down. If AML does come up with a protocol that works with the NanoChipsystem, then the lab would like to work out a deal where Nanogen would be able to sell the protocol to other labs in the arena, in exchange for some kind of a break to AML—perhaps on price or in equipment. White knows that other labs could be the first to come up with a workable NanoChip protocol, and get the break. Likewise, Nanogen knows that labs might share their protocols with other companies promoting other systems.

Big Competition, Up Close

Nanogen is not the only firm promoting an automated system based on electrical detection of DNA. Motorola Inc. is running close behind it. The huge firm, known for systematically reinventing itself, is intent on becoming a major player in molecular diagnostics. To that end, it has invested $500 million over the past few years—including $280 million to acquire Clinical Micro Sensors Inc. (CMS) [See Deal] and its eSensor electrical detection technology. (See "Motorola: Paging Diagnostics," IN VIVO, March 2001 [A#2001800065.)

Birndorf says he's not particularly worried about Motorola. At present, the big firm has only an electronic detection system, which can be no better than the underlying system, he asserts. Passive hybridization isn't as fast, accurate or powerful as the active hybridization that occurs when Nanogen applies electronics to facilitate the assay itself. Current enhances the specificity and sensitivity of binding, which in Nanogen's system is then detected by fluorescence, he explains. Stepping back from technical details, Birndorf takes a larger swipe at Motorola's capacities. "I don't know how many times large companies like Motorola have tried to get into the clinical diagnostic market. None of them has managed it yet," he observes. Hewlett-Packard Co. , International Business Machines (IBM) Corp. and Texas Instruments have all tried at various times through the years to enter the space, "but it's not their corporate knowledge, not their know-how," Birndorf declares.

Despite Nanogen's brave talk, Motorola could clearly be a brutal competitor for the small firm—tougher even in the markets than the courts, where the two are headed. The big company can anticipate economies of scale from its existing infrastructure and years of manufacturing expertise. Motorola's pockets are deep enough to let it withstand long sell cycles native to an early-stage market. If it needs to, Motorola can take early losses for longer-term strategic gain. Nanogen, as an eight-year old biotech company that has only just begun generating revenues is under more financial pressure. Even though it has some $90 million cash in the bank, and a valuation of about $125 million, Nanogen can't afford to go a whole lot farther without revenues. Although analysts anticipate that Nanogen may launch another non-biochip-based product line by year-end, the NanoChipsystem is clearly the company's core technology, and it must start paying off soon.

Perhaps that's why the smaller firm has thrown the first punch in what could be a nasty fight. It filed an action in May 2000, concerning a patent for detection of electronic hybridization, issued to inventor Mark Hollis and co-inventors. The patent, which also contained two claims that relate to electronic hybridization directly, was assigned to a consortium comprised of Massachusetts Institute of Technology (MIT), Houston Advanced Research Center (HARC), and Baylor University 's Baylor College of Medicine . These institutions issued a license to the patent to Beckman Coulter Inc. in certain fields of use, and they then granted a second license to Genometrix, which later assigned the rights to Motorola. Nanogen tried to obtain a license as well, but was turned down. That's when it filed a complaint asking for declaratory judgment that it was entitled to a license, because the work claimed in the original patent was funded with US government money. Nanogen also requested that the patent be declared invalid, and that it be found not to infringe the patent.

Nanogen settled with Beckman Coulter after just a month, obtaining the same license that MIT, HARC and Baylor had given to Beckman Coulter. The firm dismissed Beckman from the lawsuit without prejudice, but later added Genometrix as a defendant, explains Nanogen general counsel and VP, Vera Pardee. Motorola then filed a counter-claim against Nanogen, asking for declaratory judgment of its own: they believe that Nanogen does infringe, and have asked for an injunction.

The suit is ongoing, but fact discovery is substantially complete, Pardee says. The crux of the dispute with Motorola remains whether the field of use in Nanogen's license from Beckman covers Nanogen's activities, whether its product infringes those claims and whether the Hollis patent is valid.

In the face of so many uncertainties, Nanogen has little choice but to put its head down and work to get its system placed and embraced in the market. As hard as the going promises to be—as it is for any firm introducing a new technology for a new application—Nanogen has some blessings worth counting. By being first out with an electronic-hybridization system that includes chips and reading devices, Nanogen has full freedom to position its technology in the most advantageous light. It can set the terms of the debate in its promotional efforts—emphasizing accuracy and flexibility—instead of reacting to the claims of other firms, who might stress that their technologies are capable of much higher-throughput testing than Nanogen's system. Nanogen isn't as deep-pocketed as Motorola, but management feels the firm has enough to do effective marketing.

Motorola's second-mover position has its charms, too, especially in a technology field—different than a therapeutic market where the first compound to solve a problem usually grabs majority share. If the courts decide Motorola is permitted to do electronic hybridization, the firm will clearly be looking to learn from what Nanogen does right and what it does wrong. The bigger outfit will be able to send out lots of representatives to listen to customers' complaints and wishes, and possibly find an edge to take away the market. It might use some of the tactics that helped Abbott Laboratories Inc. stomp Beckman Coulter's Hybritech Inc. , the pioneer of the monoclonal antibody-based PSA test for prostate cancer and a company also founded by Howard Birndorf.

Subject to Scrutiny

Responsive positioning is a time-honored marketing game that Nanogen itself has been playing: describing Affymetrix as manufacturing-centric, rigid and expensive, and itself as customer-centric, flexible and low-cost. But if the company's system doesn't do what customers want—or can be convinced to want—a clever marketing pitch won't compensate for a technological shortfall. At this early stage of the game, Nanogen's customers are sizing up the company dispassionately. The scrutiny will continue for a while yet, and will extend to others too.

Paolo Fortina, for instance, says that as head of a clinical diagnostics lab, he strongly believes that all instruments coming into the market, especially one seeking to be number one, will have to be thoroughly validated. In an ideal world, he muses, he'd pick ten companies, each with its own instrument, then take 1,000 samples, previously genotyped by sequencing, and run those samples on each instrument for the same test. "You need instrumentation that is 100% accurate," Fortina declares, adding, "I'm using Nanogen, but I'd love to have four or five other instruments and do exactly the same tests, on the DNA of the same individual, "and see who is giving me what I expect."

What would Nanogen have to do to win Fortina's loyalty, so he could be counted on as a customer that likes this brand and buys it? He declares, "There is not going to be one platform that will solve any and all kinds of biological questions that we have today. What [sorts of tests] you do and which kind of instrument you choose depends on what kind of questions you need answered and how much money you have." Fortina figures there won't be one single dominant player in the market; rather he foresees a limited number of specialized devices holding sway.

Fortina says he thinks Nanogen's system, as it is configured today, fits well in the diagnostic labs of academic institutions and hospitals, because it is intended to do mutational analysis and SNP detection on a limited number of samples.

One of the biggest questions facing Nanogen is whether its low-density chips will continue to be diagnostically relevant as research advances. One of the firm's earliest collaborators, Fortina notes that his early research into thrombosis inspires him to envision a diagnostic chip that holds known genes, and known mutations related to cardiovascular disease, so they can be tested for simultaneously. "You could have that information back in a day," he says, musing that at first the results of such a diagnostic panel might only be statistically significant, perhaps correlating to increased risk of heart attack at an early age. "But in the future we'll have the ability to say with certainty, if you have that mutation you are at high risk of a cardiovascular incident. We could start taking preventive steps right away," he declares.

Will Nanogen's chips be capable of the kind of full-genome testing Fortina and other researchers aspire to do? Loose agreements of the sort this researcher and others are signing raises the very real possibility that an early collaborator could go off to work with another firm that provides a technology he finds more appealing. But even greater uncertainty exists outside the collaborations and alliances that companies form. Agreements have at least some defined components; they tilt in predetermined directions. It's the realm of research that really ups the ante for companies that have placed their bets on particular technologies.

How Many SNPs Will Suffice?

There's no way around the risk that new technologies and new players may leapfrog early entrants to a new market. Companies have got to commit to commercialization at some point. They've got to define and refine what they have, even if it's a method rather than a product, and carry that forward to customers. Nanogen is just at this stage. For better or worse, the NanoChipworkstation is currently engineered to be a flexible but very directed system for detecting a few SNPs or a few STRs in a low volumes of samples. University labs and hospitals, and reference labs hoping to offer new high-value tests seem to like the NanoChip system, and respect the technology underlying it.

Increasingly it is the direction of cutting-edge research that is pressuring Nanogen—and not just Nanogen, but many, many technology-based life science companies that established themselves a few or more years ago, when science was inspired by the increasing availability of genomic data but not overwhelmed by it. These days, researchers everywhere are thinking more broadly than they did or could just a few years ago. Instead of concentrating on single genes, and single point mutations in those genes, researchers are increasingly wondering what patterns they may be able to recognize in masses of genomic data, and what technological approaches might best help them do that.

Scientists the world over are involving themselves in the identification and study of SNPs that spell differences in disease susceptibility and outcome. Did the term even exist five years ago? It certainly wasn't in common usage in the biotechnology industry: the talk then was all about genes, sequences, perhaps some whispers of functional genomics. Now, some experts say that ten new SNPs are being discovered each day. It's not at all certain how many of these will actually tie into disease. In fact, it's a matter of great debate in boardrooms and on research campuses—and it's got big implications for companies developing and selling new research tools.

"To discover a SNP is one thing, to have it correlate with disease is another," says AML's Randy White. He thinks that a few thousand SNPs will be identified over the next three to five years, and that perhaps ten percent of them will be clinically correlated with disease. He reckons, "in a few years, we'll end up with 150-200 SNP-based assays." Other researchers have a much different view of how the world will turn.

Steve Chanock, MD, of the National Cancer Institute of the National Institutes of Health , thinks that within a few years, researchers and clinicians are going to be basing their diagnoses on SNPs—lots and lots of them by today's standards: not three or four or even 15, but more like 50 or 100 at a pop. "The idea of testing for a gene is really not going to be so helpful three or five years from now," Chanock declares, adding, "There is no such thing as a monogenic disease any more." To support his point, he explains that in studies of patients with cystic fibrosis—once the classic example of the single-gene disorder—variations in the gene for mannose-binding lectin have come to be associated with increased disease severity and a higher risk of infection. Variant MBL alleles have also been shown to correlate with susceptibility to HIV infection and progression to AIDS; as well as with lupus and chronic hepatitis B infection.

Chanock thinks the connections between SNPs and disease will only continue accumulating, and will need to be looked at in quantity. In his scenario, 100 different SNPs will be discovered to play a role in breast cancer. A patient that has mutations in 30 of them will have her estrogen therapy adjusted differently than someone with mutations in 70, or receive other treatment entirely. In stroke, he thinks researchers may come to see that 50 SNPs are involved, "and they'll ask themselves, what does that mean if a patient has 10 protective ones…can those ten trump the other 40? Those are the sorts of questions we'll be facing in the SNP world, not questions about the presence of one or two," Chanock asserts.

If Chanock's predictions are right, then Nanogen's current low to medium-volume system for detecting limited numbers of SNPs will have to evolve into something quite different than it is today, if it is to compete in the clinical diagnostic marketplace of the future. "Nanogen has positioned its technology so that as new SNPs come off press, they can look at them in a very directed manner. The question is, how well can they move to that level of higher throughput? The field has moved there already, and they're scrambling to get there," he asserts.

If Nanogen is the designer approach, then Affymetrix is the bulk approach. Chanock says he'd like a hybrid of both, a system that offers a versatile and high throughput way of looking at SNPs. Somebody is going to provide that, he reckons, because there's reason to do so. Chanock works with molecular epidemiologists, and notes that already, about 1.8 million individuals have banked DNA—all of them through approved, funded studies sanctioned by institutional review boards. In ten years, he thinks a good portion of the American population could have their whole genome sequenced—and if that happens, "Nanogen and Affymetrix could both be out of business."

Chanock acknowledges that his comments will sound inflammatory to some ears, but argues his thinking is right in line with what growing numbers of scientists and business people are contemplating now. Nanotechnologies that hugely simplify sequencing are all the rage in the academic world, he says, adding that "Applied Biosystems and Perkin Elmer [Applera] are all over this."

Welcome to the amorphous market—chaotic, competitive, rife with uncertainty, as exciting as it is challenging. Nanogen's CEO Kieran Gallahue insists his firm is poised to thrive here and now, despite the tumult and because of it. Nanogen's technology is highly flexible, he says. It may be able to be scaled up to accommodate the sorts of multiplexed assays Chanock believes will become vital to clinical diagnosis. The technology promises to be amenable to further miniaturization, in which case it could lead the company towards creation of point of care diagnostics. Nanogen's electronic hybridization method may even be adapted to detect proteins, another new research area stirring passions these days.

Nanogen is looking into these possibilities and weighing its options for the future now, Gallahue says. But the firm has got to live in the present. Yes, researchers all over the world are trying to utilize massive amounts of genetic information. But scientists are not going to simply jettison the gene-based studies that have occupied them for years. They'd be foolish to do so.

Nanogen believes its system will help create a bridge between research as it is performed now into clinical practice of the not-too-distant future. The company's objective now is to get the technology validated, and that process has at last begun. Management thinks they've got enough cash in the bank to see them through. The company has a powerful partner with relevant expertise and a personal interest in applying it to help Nanogen. The firm has connections, collaborators, enough scientific credibility to get it a hearing. It's got a lawsuit looming, too. And eventually it will need to worry about getting tests and devices approved by the FDA.

If all goes well, Nanogen will move along the value chain, positioning the NanoChipsystem first as a research tool, then as a tool with applications in new areas. Then it will license in content, and develop kits to run on the machine. It's also possible that the rapid change will foil Nanogen's plans. It could be that the assumptions the company has built its business on are no longer accurate given the state of technological development today. The clinical researchers the firm views as gatekeepers may be removed from that position, if some other firm or constellation of companies matches DNA or SNPs to format in a novel way. There is so much unknown, and unknowable.

There is only this certainty: in a market heaving with change, Nanogen has to define itself on its own terms. It must trust its vision, make its case, and hope for the grace to evolve along with the market for clinical diagnostics, however it shapes up.