YouTube Facebook LinkedIn Google+ Twitter Xinginstagram rss  


By John Dodge

Special Issue 
· When Behold: Bio-IT Innovators
· When Only Brute Force Will Do
· Breaking Down Silos and Busting Bottlenecks
· Managing Innovation and Adversity
· Prescription for Success: Mix IT and Science
· The Power of Expression
· Taking Data Storage to Infinity — and Beyond
· Profiting from the Proteome
· Serving the National Health
· Powerhouse CRO (Slowly) Goes Electronic
· Talent Fuels Drug Pipeline in Swiss Time
January 13, 2003 | As ArQule Inc. struggles to make the transition from technology provider to drug discoverer, the company's mission remains the same: move new chemical entities (NCEs) into a better neighborhood.

Success means reducing the compound failure rate from 90 percent to 70 percent and halving the time it takes to discover a target and develop an investigational new drug. The strategy depends on a combination of compound design, high-throughput chemistry, and in silico predictive modeling.

But now ArQule must do this with 128 fewer people, a 31 percent reduction of its work force, as it focuses solely on drug discovery. The company announced this painful restructuring on Dec. 11 of last year, reflecting the dismal industrywide prospect of profitably selling informatics tools and services.

"[There is] a reluctance to provide investment for technology platforms. We remain confident that the application of predictive ADMET [absorption, distribution, metabolism, elimination, and toxicity] models has the potential to fundamentally improve drug discovery efficiency. [We will] integrate that within our drug discovery program," says ArQule CEO and President Stephen A. Hill. Hill, an Oxford-educated surgeon who spent 10 years at Hoffmann-LaRoche Inc., including three years as head of global drug development, joined ArQule in 1999.

With the elimination of ADMET technology development and marketing, the layoffs, along with the shuttering of two sites, will free up $100 million that ArQule can put toward drug discovery over five years, Hill says. At the end of September, ArQule had $85 million in cash and equivalents.

Hill says technology is one of four pillars that will support the company's eventual success. The others are managerial leadership, a portfolio of drug programs that outpaces attrition rates, and sufficient working capital. Perhaps more significantly, ArQule has a rich compound discovery and technology collaboration with Pfizer Inc. that could generate revenue as high as $345 million through 2008.

In Hill's eyes, being a "good, healthy" company means a $1-billion market cap (it is currently around $125 million), $250 million in the bank, a $40-million R&D budget, and six preclinical and one or two Phase I NCEs in the pipeline.

But with the cutbacks, ArQule is taking one step back to take two steps forward sometime in the future. The facilities slated for closure are Redwood City, Calif., where the bulk of the ADMET models and tools were developed, and Cambridge, England. According to Hill, negotiations with an unspecified partner for ArQule's ADMET platform had collapsed, which precipitated its sudden exit from the informatics business.

The frustration is apparent in Hill's voice when the subject of stock price is raised. "This is a company whose stock has been just under $40 and under $4 without any negative news while we've met all our goals and milestones. Share price rarely reflects valuation." Post-cutbacks, the stock seems fated to languish under $4 a share for a while.

Importance of Infrastructure 
But ArQule's technology infrastructure remains the company's major asset as it adjusts to the new realities. Founded in 1993, the company has pioneered novel ways to synthesize, analyze, and optimize small molecules, licensing the resulting compound libraries to large pharmas.

 Some Pain. Some Gain: ArQule CIO Alan Hillyard's database migration project became more challenging—and more important—following recent layoffs. The good news is that progress continues.
Before the cutbacks, the only compounds ArQule had disclosed publicly were one in women's therapeutics that Wyeth Pharmaceuticals nominated for development in October, and another that Solvay S.A. has targeted at irritable bowel syndrome. These partnerships also usually include technology transfer involving compound synthesis and predictive modeling.

But on the day the restructuring was disclosed, the company announced three internal programs in its year-old drug discovery effort. They include inhibitors against a panel of kinase targets; an ion channel program for treatment of neuropathic pain; and Catheptin S, a third program to treat tumors and inflammatory conditions that will move into animal testing within the next year.

The IT piece of the mission falls on the shoulders of CIO Alan L. Hillyard, formerly senior vice president of cheminformatics research at LION Bioscience AG. Hillyard says his job boils down to developing two things: technology that supports the synthesis of small molecules, and in silico tools that predict ADMET properties to create more druggable compounds.

"The more work we put in upfront — and by that we mean in silico work — the more desirable the drug," he says. "Rather than [just] a new chemical entity, we will have explored all the relevant chemical space around a compound." The result is what ArQule calls an optimal chemical entity or OCE, a term the company has trademarked. Finding OCEs is predicated on generating accurate information and acting on it. "We think about how can we make the compound more potent and the most druglike," Hillyard says. "We spend a lot of time on the computer before we make it in the laboratory."

But now he'll have to do it with less than half the people he had on Dec. 10 (see chart). Before that, a fifth of the company's then-400 employees worked in Hillyard's department, which consisted of traditional IT personnel, computational chemists, informaticians, ADMET modelers, and software engineers. Only the IT and informatics groups remain, with the computational chemists folded into the latter. The software engineers and ADMET modelers are gone.

"In the short term, we have less capacity than we had in early December," Hillyard says of the cutback. "One of the things to do is to sit back and see how we reassemble in the most effective fashion. [That's] going to happen in the next couple of weeks." He still has versatile IT subgroups with data enablers (traditional IT but now without software engineers), exploration and discovery (computational chemists less the ADMET modelers), and interpretation (informatics).

"We are still left with a group being on the forefront of exploring very tough problems," he says. "The company needs the applications and data integration, and that's what we will spend our time on. We were developing methodologies and applying them. Now, we'll just be applying them."

Informatics will continue to act as a bridge between the information management worlds and the scientists conducting experiments. As Hillyard explains: "They are responsible for getting the scientists' questions answered yesterday, not three months from now. They are the rapid response force and tend to have broad perspectives on biology and chemistry. They keep up on the latest gee-whiz stuff in visualization and analysis."

Meanwhile, the IT infrastructure team makes sure the hardware, software, and network all function smoothly. One of Hillyard's biggest projects at the moment is migrating ArQule's compound database — containing a million compounds — from SGI to Sun servers in order to stay current with underlying commercial components such as Oracle and MDL registration systems.

"We've been planning for it for a year," he says, adding that the project will be completed over the next few months.

Life Does Not Get Any Simpler 
Once that's done, his job is to keep ArQule's network running between its headquarters in Woburn, Mass., and its satellite facility in nearby Medford, Mass. "The network issues don't go away now that we have one facility six miles away instead of two 3,000 miles away," Hillyard says. That's why ArQule just put in six T1 lines between the two facilities, replacing a microwave link.

"The production of compounds is done both in Medford and Woburn, but our centralized database is in Woburn," he says. "It requires a lot of bandwidth to [remotely] tell a robot in real time to go left or go right." The Cambridge and Redwood City facilities solved smaller-scale problems, so each required a single T1 line.

Besides the downsizing, his other headache is pulling together vital information. "[It's] presentation and sharing of information for any drug discovery project. How do you make intelligent decisions on what to do next? If I'm a scientist, I'm worried about getting the most active compound, about good ADMET properties, and about being able to synthesize the compound. All those determinations may be made at different sites." He said that when ArQule had four sites. Now only two are left, so the problem of a farflung staff has been cut in half. But that's not to say life is any simpler for Hillyard.

So far, so good, says Joe Stocke, managing director of Stone Ridge Partners Inc., an investment firm that owns a block of ArQule's stock. "We think their combinatorial chemistry is valuable to a whole range of companies, and, by combining it with technology, they should speed up drug discovery. We're happy with the progress, but the stock market has been in [a funk]."*


For reprints and/or copyright permission, please contact Angela Parsons, 781.972.5467.