Kalorama Information, a New York-based market research firm, this month published a new report on the market for gene expression profiling platforms that forecasts double-digit growth for the segment leading to a market worth nearly $2.5 billion by 2012.
Kalorama’s study, Gene Expression Profiling Markets, discusses new trends in a market that array vendors routinely describe as mature. Specifically, it addresses the emergence of newer applications, such as splice variant arrays, comparative genomic hybridization, and miRNA expression profiling, along with new technologies like next-generation sequencing, both of which make the future of the market difficult to predict.
To follow up on some of the key themes discussed in the report, BioArray News spoke with Kalorama biotech analyst Justin Saeks last week. Prior to joining Kalorama, Saeks worked with various life science firms, including DiscoveRx, ExonHit Therapeutics, and MDS Sciex, and was a senior analyst at Frost & Sullivan. During the interview, Saeks addressed some of the larger issues facing the market, including the arrival of next-generation sequencing instruments and the appearance of array-based diagnostics in the clinic.
In the report, you characterize the expression market as both predictable and volatile. Can you give examples on why it is both?
The market is fragmenting. There are still new applications being introduced and then there’s next-generation sequencing, so things like that make the market unpredictable. The impact of these new applications and sequencers still has to shake out. The part that’s less volatile is the fact that people that are using arrays won’t suddenly switch to other technologies. They are locked in and pharma, especially, has been generating data using arrays for several years. It’s a priority for them to continue generating data they can compare to old data, not to mention they’ve invested in the equipment, personnel and what not. They are not going to switch over to other technologies or applications that quickly.
For a large segment of the market – arrays and RT-PCR – it’s foreseeable that it would not make dramatic changes in their labs for at least the near term, but the next-generation sequencing technologies introduce a wild card. The market could change its scientific dogma rapidly, for example a thought leader might publish a paper showing that sequencing technology is much better, causing people to switch from arrays to sequencing more quickly.
Still, I’d say the majority of the market right now is not going to have revolutionary changes in near term. But it is possible that next-gen sequencing could disrupt this market or slow the growth because it is more efficient, et cetera.
The report also notes that the share of gene-expression revenues from US customers has fallen in recent years, though the market continues to grow. Can you explain this phenomenon?
Basically, the early adopters were in the US, and in general that’s where the market developed. So now that arrays have become a more proven technology and usage has become more uniform around the world, they are seen as a standard tool. It’s still sort of cutting edge, but it’s shifting. It is more representative of life sciences research in general. This is something that probably happens with a lot of different technologies. Americans will be the first ones to take the risk and spend money on something that is new and more expensive.
There’s also the fact that when the companies first introduced microarray technology, it was pretty revolutionary as well as expensive. It was less likely for them to sell arrays abroad a decade ago because of their budgets and because the technology was less accepted. The sales people knew Americans, Europeans, and Japanese take risks and that’s why they invested here first.
Some firms have forecasted significant growth for newer applications like splice arrays and array CGH. What are some applications that you believe to be promising and may help grow the expression market?
Some of those applications would be technically considered different applications outside of gene expression analysis. But the technologies and companies overlap; for example, ChIP-on-chip and array CGH. One could argue that studying splice variants is a more accurate way to perform gene expression analysis because if you are not taking into account different splice variants, you are sort of ignoring the most detailed representations of RNA levels. It does seem that that splice arrays are high growth, certainly because it’s a smaller segment.
But it’s like what I said earlier about scientific dogma — some people think studying alternative splicing is important, some think it isn’t, so it’s a mix. In some labs, the question is more from a business angle – for example, does it help bring a drug to market? Over the next few years, I would expect that the application would be recognized as the more accurate representation because it is just a fact that most genes are spliced in different forms. Over the longer term that application could revolutionize gene expression analysis.
MicroRNA is another promising application; it could be considered a subset of gene expression analysis. I think that will be important because it’s information that in some cases will be valuable. There are other applications that are related, like comparative genomic hybridization and chromatin immunoprecipitation (ChIP)-on-chip. It does seem like these applications are small but they are growing fast and some people will find them to be valuable. Those same people will be looking at other types of microarrays because they already will have the equipment. I think that will expand the market because these applications are distinct and I think all of them are promising. They are responsible for some of the faster growth in the microarray market. Still, I think the market will continue to fragment to meet the needs of different people. Some will be used as diagnostics, like CGH is already being used for prenatal and newborn screening, so there’s a lot of interesting stuff happening in each one of those application areas.
Another factor is digital gene expression on next-gen sequencing instruments. How do you think this new application will impact the market?
It’s hard to say. It could take some revenues away from microarrays because some people will find they are more efficient with generating data, et cetera. If they can run a faster experiment with that then there’s no need for the microarrays.
DGE might actually slow the growth of the market if people can run experiments more efficiently with less consumables. But it is a very young market, and products are still being introduced. Even for the platforms that have been around for awhile, it remains to be seen how the end users work out the best way to fit them into their needs.
The reality is that in the case of most of the people that are buying a sequencer, they are not buying it for DGE, they are buying it for sequencing. They will do DGE because it’s another application they can do. So the question is how much of the work on that instrument will be done for DGE. So that’s a wild card.
It’s guaranteed that DGE will affect the market. It’s sort of a question of how much. I don’t think that will be clear for a year or two. It’s similar to the situation for people who have Affymetrix that I referenced before. They have invested a lot in array systems, personnel, bioinformatics, and infrastructure. They are just not going to switch over right away until they see a clear benefit. They will want to make sure they can compare the data with old data and not sacrifice too much to gain whatever it is they will gain from using DGE.
Finally, you note that much of the revenues in the expression-profiling market come from the research community. Why has this technology not yet made it into the diagnostics space and what kind of forecast can you provide about its future use in diagnostics?
It’s taken longer than the companies had hoped to reach the clinic. Diagnostics is a much larger market and Affy in its early days was hoping to bypass the research application market; they had their sights set on diagnostics from day one. But it’s taking a while and it has started to get moved into diagnostics – there are now a few microarrays that are being used in diagnostics, like Agendia’s MammaPrint test. There are only a few examples, but they have broken through that barrier.
It’s going to take time for arrays to reach the diagnostics market mainly because the medical community is not going to rush into anything until they know that there’s not much risk and there are actual benefits. Doctors are hesitant with good reason to place confidence on something until they are completely sure that it is safe and effective. Of course the health care industry is based on the economic savings of running the test versus not running it.
There’s also the aspect that the diagnostics industry has sort of been entrenched and slow to innovate. It’s been based around these large instruments that are in centralized labs in hospitals. There’s usually one analysis per test as opposed to multiple analytes on microarrays in that scenario. So in that way they are committed to their existing testing infrastructure and it’s going to take time for them to switch. It is also somewhat of a new paradigm or philosophy for them to base a decision on multiple-analyte profiles.
It seems like cancer will be the first area where this technology will be adopted and recognized as useful. That’s also because there are a lot of genetic changes happening in cancer that can be discovered at the RNA level. It is also the major disease in the population. That will probably be the one area where it gets initially accepted into the mainstream and it should probably grow from there.
The obstacle is that people are afraid to change at the basic level. But there’s also the economic side of things and people don’t want to stick their neck out. It has to take something to change that, like a dynamic paper, for example, the New England Journal of Medicine study (see BAN 1/22/2008). That will reach people in the mainstream that study autism. Arrays are now no longer just a technological curiosity as it is a way to provide valuable information.
So it will take time to reach that tipping point when it will become a widespread technology – maybe five to 10 years. But I think within three years you will see signs that it is headed in that direction. CGH is also opening peoples’ eyes to the fact that that’s a valid approach. It’s cost effective and produces better data than alternative techniques. It just has taken time to work its way through the system between regulatory issues, economic factors, scientific issues, et cetera.