Contributed by: Alejandro Gutierrez

Most sequencing companies today would never pass up an opportunity to make some noises about the molecular diagnostics (MDx) market. Even if their product is only being sold to research customers. Even if they don’t yet have a product on the market. The logic goes something like this: the research market is somewhat limited and heavily dependent on government funding. Molecular diagnostics, on the other hand, is a large, fast growing market where next generation sequencing technologies have yet to materially penetrate.

So if the MDx opportunity is so attractive, why have next-generation sequencing firms focused their efforts to date on the research market? Why do start-ups begin their marketing push by targeting academic, non-profit and commercial research organizations?

If one looks back at other technologies that have established themselves in diagnostic labs, the pattern of research first, clinical labs later is well-established historically: new technologies fare better in the world of research, where they get tested, improved and ultimately accepted. Researchers are more willing to try new instruments as they seek to push the boundaries of knowledge and devise new applications. Clinical diagnostics, in contrast, is a “conservative” field subject to market rigor. Risk has to be carefully weighed against potential benefit – and new technologies, especially unproven ones, can be hard to justify.

Next-generation sequencing instrument manufacturers, like other technology providers before them, thus find themselves contemplating a market that has significantly different requirements than the research market. Researchers, especially in genome centers and large research institutions, will prize flexibility and throughput, whereas clinical laboratory directors will emphasize cost-effectiveness, reliability, and ease-of-use.

Beyond product design there are various other obstacles to enter the MDx market. First among these is the need to develop a “test menu”. Like software for computers, or games for gaming consoles, an instrument is not very useful without a test to run. The more tests an instrument can run, the more useful it becomes to potential buyers (it helps, of course, if it can run those tests with adequate quality and efficiency), and the more likely it is to sell. As it sells, and as its installed base grows, the instrument becomes more attractive to test developers.

Instrument manufacturers typically choose from among three strategies to develop tests: partnering, self-development, or a hybrid of these two. In the partnering model, the manufacturer sells or licenses its technology to test developers, who then use the platform to develop new tests (or adapt existing tests to the new platform). Test developers will choose platforms that provide the necessary technical features and performance to fit their test. If they are primarily focused on developing laboratory-developed tests (LDTs), they may not care as much about the technology’s installed base (except as a testament to general market acceptance and reliability). If the test developer is aiming to sell in-vitro diagnostic (IVD) kits, then installed base and regulatory certification come more prominently into play.

Manufacturers that have the necessary resources and seek more control over their fate (or that cannot convince others to develop tests for their platforms) may decide to become test developers. This can be done solo, or in some form of partnership, but in either case the company is entering a new business that requires a new set of technical and commercial capabilities. Furthermore, it begins life as a test developer with a potential competitive disadvantage: whereas independent test developers can choose the best platform for their tests, the manufacturer is constrained by the limitations, if any, of its own platform.

In the hybrid model manufacturers develop their own tests even as they seek others to do the same. The strategy is compelling because it gives the firm a measure of control even as it leverages its partners, but it presents one potential problem: in playing both games, the manufacturer could find itself competing with its partners or customers. One way around this issue is to stake out a territory (e.g., a therapeutic area) for self-development and let partners develop tests in other areas. Manufacturers with advantaged technologies, strong market positions, deep pockets, or all three, are more likely to succeed.

Two companies that enjoy strong market positions and have chosen hybrid strategies to pursue the MDx market are Illumina and Life Technologies. Both have a broad life sciences instrument portfolio and both are aggressive about bringing technologies from the research market into MDx. For example, Life Technologies has a range of FDA/CE-approved PCR instruments; it sells a self-developed HER2 assay, and has an agreement with Asuragen, a test developer, to distribute the latter’s tests (which are designed to run on Life Technologies’ PCR platforms). For its part, Illumina has its own CLIA lab, sells its FDA/CE-approved BeadXpress platform to test developers like EraGen, which is developing tests to run on that platform, and maintains internal efforts to develop and market IVD kits and LDTs based on its different technologies.

Luminex is an example of a company that began using a partnering model, but eventually adopted a hybrid approach. Luminex is not a sequencing firm; it has developed a bead-capture technology that is used for a variety of research and diagnostic applications. When it first launched, Luminex sought partnerships with other instrument manufacturers and test developers, leveraging the strength of its technology to build up a test portfolio. Its revenue was driven by royalties from tests, reagent sales, and instruments – the latter sold primarily via resellers and partners.

From 2000 to 2006, Luminex’s installed base grew ten-fold from less than 400 instruments. Its diagnostics partners increased from a handful to 17 in the same period. In 2006, the company announced a change of strategy and began to develop its own tests. Then, in 2007 Luminex amplified its efforts by acquiring test-developer and until-then partner Tm Biosciences.

So what are smaller sequencing firms that lack the market strength of Life Technologies and Illumina to do in pursuing the MDx market? Helicos and Pacific Biosciences (PacBio) have each adopted one of the first two strategies described earlier. Helicos has declared itself a molecular diagnostics company and is developing LDTs for its poorly-received HeliScope platform. PacBio, on the other hand, has sealed an exclusive co-development partnership with Gen-Probe, one of the leading instrument and test developers in the MDx field. Both are single-molecule platforms, which is considered advantageous in diagnostics applications because it eliminates the labor and the bias (and resulting inaccuracy) associated with the sample amplification required by the most widely-adopted sequencing platforms. Both platforms also rely on image capture to read bases. Beyond these similarities, however, the two technologies, and the two firms, are very different.

With the exception of its single-molecule approach, Helicos’ technology is very similar to that of the other second-generation platforms: it is characterized by short reads and time-consuming hybridization and wash cycles that result in extended run times (a standard run takes eight days). Although the HeliScope was the first single-molecule platform in the market, it has had minimal market success (thus the change in business strategy) and the firm is in a difficult financial position. Its decision to build its own CLIA laboratory and develop LDTs to run on its platform was probably the only viable option given its weak market and technological positions, but the firm’s current cash reserves will not be enough to see this effort through.

PacBio, although it has yet to enter the market, and although questions remain as to its platform’s accuracy, appears to have a much stronger single-molecule technology than Helicos, with significant performance differences such as long reads (possibly the longest of any technology available to date) and very rapid run times (as little as fifteen minutes). The company also enjoys a comfortable financial position, having raised close to $400 million since its inception (and now seeking an additional $200M from an IPO). A sizable portion, $50M, of that war chest comes from Gen-Probe as part of the firms’ exclusive deal. Though details of the agreement are sparse, the announcement indicates that the companies intend to develop a “system” that combines PacBio’s sequencing technology with Gen-Probe’s expertise in other areas of diagnostic instrumentation such as sample preparation, materials handling and automation. The press release does not discuss test development (though it does hint at therapeutic areas) – in other words, it is not clear if PacBio, Gen-Probe, or both jointly, intend to develop tests for their system. As long as it can retain some of the value, a potential strategy for PacBio would be to let Gen-Probe, or others, develop the tests. Embarking on a venture to develop its own tests (at least in the early stages) would fail to leverage its partner’s expertise and capabilities.

Both Helicos and PacBio are still a couple of years or more from making a mark in the MDx market. So, too, are the next-generation sequencing platforms from Illumina and Life Technologies. The few next-generation sequencing-based tests available today have been developed by test developers as LDTs. For example, GeneDx runs genetic panels for Long QT syndrome on Illumina’s GA platform, while Correlagen offers a multi-gene assay for familial cardiovascular diseases that runs on a HeliScope. Other diagnostic firms such as Sequenom and Ambry Genetics are developing genetic tests based on NGS, and are evaluating platforms from Illumina and Life Technologies.

Notwithstanding the current state, there is little doubt that next-generation sequencing will eventually play a large role in clinical diagnostics. Genomic Health, a pioneering developer of multi-gene expression tests (aka IVDMIAs), is betting heavily on next-generation sequencing. Gen-Probe, the market leader in NAT-based blood screening, has done the same with its investment in PacBio. Ion Torrent’s platform seems a safe bet (see this analysis by Jeremy Leipzig, for example) for fast-turnaround applications.

The more pertinent question for instrument manufacturers then becomes: which technologies will dominate the market, and what strategies will get them to that point?