NEW YORK, Nov 1 - When Affymetrix entered the uncharted waters of DNA chips in 1991, microarrays were highly specialized and complex tools that few companies were ready to market or use.

But over the past few years, many companies have begun to understand the underlying technology and find ways to circumvent Affy’s host of over 300 protective patents, pointing to the possibility that these tools could become commodity items.

While Affy currently commands $180 million of the existing $220 million market, companies such as Agilent, Corning, Mitsubishi, Motorola, and 3M have all recently announced they are entering the chip race.

The reason for this sudden explosion in chip development is simple: The market is growing at a rate of 50% per year, according to Paul Knight, financial analyst at Thomas Weisel Partners, and many companies are betting it will soon extend beyond the laboratory and into doctor’s offices and agriculture.

“The industry is going to become much more competitive. You wouldn’t see these players getting involved if the potential market wasn’t huge,” said Rob Olan, a biotech analyst at Chase

Academic research is already demonstrating that gene expression profiling using microarrays can be an effective and precise diagnostic tool for numerous types of disease, from cancer to lung disease, according to Mark Geraci, a pulmonary disease specialist and director of the University of Colorado’s gene expression core facility, the largest academic user of Affy’s chips. Pharmaceutical companies’ use of microarrays in drug toxicity testing can shorten the drug pipeline by as much as 7 years, Geraci added.

But anyone who wants to capture this burgeoning market will have to conquer some serious obstacles, not the least of which is Affy’s own head start in the race.

Affymetrix customers have published over 300 peer-reviewed papers using its technology. In addition, the company is already supplying low-density custom chips to over 550 customers.

Affy’s chips offer the advantage of redundancy, with 32 different features on each chip to test the expression of a gene. They also offer different high-density arrays for human genes, murine genes, E. coli genes, and even custom arrays.

“Many of us are going to be watching very closely how these other companies’ products compare [with Affy’s] in terms of variety of product, validity of the data, ease of use and cost,” Geraci said.

The Cheaper Chip Challenge

Cost may be the most important factor in determining which chips become the most popular, especially in the academic market. Affymetrix recently lowered the academic price to $185 per chip, but the increasing number of market competitors may drive down cost much further and make the chips more attractive to a wider variety of researchers.

“Affymetrix’s chips are expensive so a lot of hospitals and universities will not use them, or can only use them” in limited amounts, said David Ghosh of the Institute for Transcriptional Informatics in Pittsburgh. “There’s lots of room for someone to develop a cheaper chip.”

Currently, many researchers make their own chips, but this may change if chips become cheap commodity items. “At $25 to $50 a chip you’re beating the cost for a researcher to make his own chips in house,” said Rudolph Spangler, senior research associate at the Rockefeller University.

Business behemoths like Motorola, Mitsubishi, Corning, and 3M are hoping to meet this price challenge by leveraging their manufacturing know-how into mass production systems that drive down the cost per chip.

“At the end of the day it’s the ones who can effectively industrialize the process that will survive,” said Greg McPherson, project manager at 3M Bioanalytical Technologies

Corning says it will be able to produce chips at a rate of one chip per minute, or in its estimation, 10 to 20 times faster than today’s conventional processes, leading to lower cost chips than any on the market.

The fiber-optic and cable producer and other manufacturing giants, however, still have to figure out the life sciences side of microarrays.

Partnering and Patenting

To address this obvious weakness, Corning has just entered into a partnership with MIT’s Whitehead Institute for technology to develop DNA content for its chips.   It says its microarrays will be on the market by the first quarter of 2001.

Meanwhile, Motorola has licensed Incyte’s gene patent portfolio and databases for bioarray commercialization. Motorola plans to begin marketing its high-density BioChips by the end of 2000, and will commercialize its low-density clinical microsensors and protein chips by later next year, according to company statements.

Motorola’s chips, which are being printed with Packard Bioscience’s Piezo array printing technology, will offer the additional feature of click-through support by Incyte’s website. Chip users will be able to click on their data to get gene identification, analytical services, and reagent offerings from Incyte.

Mitsubishi Rayon and 3M are also developing novel approaches that they hope will revolutionize attempts to industrialize DNA chip manufacturing, and completely avoid Affymetrix’s phalanx of patents for its high-density silicon microarrays.

Mitsubishi plans to market “fibrous” DNA chips that would not require a license from Affymetrix, and has filed 30 patent applications in over 40 countries for these chips.  

Mitsubishi’s pilot manufacturing facility for these chips is scheduled to open in the spring of 2001 and various “strategic tie-ups and marketing schedules are currently under consideration,” the company said in a statement. Officials from the company were unavailable to comment.

3M is still secretive about its program but describes its technology as DNA spotted onto heat shrinkable film. After spotting, the film is heated to shrink it by a factor of 25, according to McPherson.

Packard Bioscience is pursuing a different strategy altogether—collaborative chip development. Packard would like to see the majority of chip manufacturers follow Motorola’s lead and use PiezoTipnology array printing technology to print their chips, said Jocelyn Burke, vice president and general manager of Packard’s BioChip Ventures Division.

Agilent technologies has also forged its own way through the crowded microarray market, opting for the small-scale custom route rather than mass production. Agilent is betting its customers would prefer its customized arrays to the assembly-line variety Affy makes.

Agilent recently announced early access to its chips and chip scanners for Exelixis and Paradigm Genetics, both highly skilled companies capable of determining the custom content they want on their chips and extracting value from the technology while Agilent’s developing program matures.

Agilent, like its competitors, is touting its program as if DNA chips were already the dependable commodities that they promise to become. Wilson Woo, marketing manager for life science products at Agilent, said his company could produce thousands of custom chips within a couple days—once the content has been decided on.

These new competitors pose a significant threat to Affymetrix, as do a host of legal suits brought by companies such as Oxford Gene Technologies, Incyte, and Hyseq. Financial analysts say the company will certainly lose market share and may be forced to pay royalties and/or cross license its patents.

But Affy’s chip sales will still continue to grow at a rapid clip, analysts believe, and may not be seriously threatened unless another company can offer a product that is not only cheaper, but better.

“If the price goes down in any one sector, people will gravitate toward that product,” said Geraci. “But they have to prove themselves in terms of quality of their products, the density of arrays and the reproducibility of data.”

-- with additional reporting by Marian Moser Jones