ZEROING IN Pioneer research firms are putting their resources into production. AVENTIS PHOTO

Exelixis has changed as well, according to Geoffrey Duyk, president of research and development and chief scientific officer. "In 2000, we were a functional genomics company with aspirations of doing drug discovery," Duyk says. "Now we are a drug discovery company aspiring to be a commercial operation."

More significantly, Exelixis now has a product pipeline. The investment community is no longer interested in the technology platforms that drove investment in the sector in the 1990s, according to Duyk and other sources throughout the industry.

The same can be said for pharmaceutical companies looking to partner with drug discovery firms. The market now focuses on products or promising drug candidates in late-stage clinical trials in assessing the value of biotech companies. As life comes back to the life sciences sector on Wall Street, drug discovery firms like Exelixis, which netted $75 million in an equity offering in June, will need to show that their pipelines have developed and that their business models have made the appropriate adjustments.

Exelixis, Array Biopharma, Human Genome Sciences, Inpharmatica, and other companies claim they planned from the start to become drug companies. Others, such as Celera and Incyte, two genomics stars in the 1990s, have dropped strict technology platform and service models, adding proprietary research in the past two years.

Millennium Pharmaceuticals, Vertex, and CuraGen, three firms well along in advancing drug candidates to market, reorganized this year, significantly cutting early-stage discovery staff and shifting resources into the far more expensive development of drugs in their pipelines. Duyk says Exelixis has already cut discovery research staff and may eventually have to do so again as its pipeline advances.

The shift away from pure science to applied science as companies refocus their efforts on products has also led to a round of management turnover. Pharmaceutical industry research and business directors, experienced at bringing drugs to market, have moved into the top spots at drug discovery firms since 2001, replacing the researcher-entrepreneurs who sequenced the genome.

Regardless of whether firms view themselves as progressing on their original path or as fundamentally new organizations, there is no denying that the business of strictly supplying tools and services to the pharmaceutical industry is played out. Providing services remains an important means of generating revenue for nearly every drug discovery company. But a pure "service model" is no longer considered viable. Profit margins for service firms are too slim, and financing has shifted almost entirely to companies pursuing a proprietary pharmaceutical business model--companies taking risks on rapid-growth strategies.

Some industry watchers worry about the shift in emphasis away from development of novel discovery techniques. Others see a blurring of the line between drug discovery and development as new methods emerge for assessing a compound's full efficacy as a drug at the earliest stages of discovery (see page 33). Two years after the sequencing of the human genome, they say, drug discovery methodology and information technology are being deployed in new ways to boost the efficiency of discovery and development and the quality of compounds moving through the pipeline.

The pharmaceutical industry's interest in drug discovery has clearly shifted from services to products. Technology "tool" providers are playing in a shrinking market, says Varavani Dwarki, director of technology licensing and alliances at Aventis. Research spending is increasing and drug companies will access any technology they need to maximize productivity, "but you don't need to have the same technology in 10 different formats," Dwarki says.

On the other hand, drug companies will be increasingly on the lookout for compounds and drug candidates with which to fill their pipelines.

Millennium, which has two commercial products, can be seen as a template for the evolution of a drug discovery company. The firm, founded in 1993 to develop genome screening, moved downstream almost immediately to explore ways to validate targets, increasingly working through partnerships that created a crucial revenue source, according to Marsha Fanucci, vice president for finance and corporate strategy. In 1997, the company made its first acquisition, ChemGenics Pharmaceuticals, to work on library screening and lead optimization.

THIS APPROACH was enforced with more acquisitions--LeukoSite was acquired in 1999 for its clinical development capabilities and its pipeline. "That took us downstream and changed the focus of the company to really being able to take products forward into the clinic," Fanucci says. With the acquisition of Cambridge Discovery Technology in 2000, more chemical skills were added to enhance target screening. In 2001, the acquisition of Cor Therapeutics put the company in the commercial arena with a sales and marketing infrastructure and a product.

"In order to create a company, you really require these external transforming events to keep people focused on the more downstream side of development," Fanucci says, "and in some instances to bring in a pipeline."

Millennium's commercial products, Integrilin, an acute coronary syndrome therapy acquired with Cor, and Velcade, a multiple-myeloma drug approved by the Food & Drug Administration in May, are still in development for new applications, and Millennium has 11 other compounds in clinical studies. The company has close to $1 billion in the bank and says it plans to be profitable by 2006.

This year's restructuring initiative was designed to focus Millennium's resources on drug development and commercialization. Millennium shuttered R&D in South San Francisco and Cambridge, England, and said it would cut staff by 26% by the end of next year.

Earlier this month, the company announced changes in executive management. Kenneth M. Bate, chief financial officer, took the post of executive vice president of commercial operations with responsibility for business strategy. Fannuci was promoted from vice president for mergers and acquisitions to her current, newly created position.

She says drug discovery remains a distinct, though downsized, part of Millennium's operations as the company turns its attention to development and clinical trials. "We can turn up the jets of the discovery engine again later," Fanucci says, emphasizing that Millennium's plan from the start was to become a drug company.

	Haseltine HUMAN GENOME SCIENCES PHOTO

Similar claims are made by other firms that have gone through big changes in taking genomics from the laboratory to the marketplace. "Our model did not evolve," says William A. Haseltine, chief executive officer of Human Genome Sciences (HGS). "It unfolded. What we are doing now was described accurately in our original 10K filing with the SEC [Securities & Exchange Commission]. We said we wanted to be a company that discovers, develops, manufactures, and sells protein antibodies based on rapid discovery of human genes."

Haseltine says the company, also founded in 1993, supported itself in the early years by offering discovery technology to partners but that it never intended to stick to a service-oriented business model. He notes, however, that HGS continues to make money through its partnerships. For example, it currently has rights to about 30% of the revenue GlaxoSmithKline derives from products it discovers using HGS technology.

HGS spent its first years on human gene discovery. Haseltine says the firm had discovered virtually all the human genes in their cDNA form by 1995 and began converting its knowledge into discovering and developing drug candidates by adding skills in protein production, high-throughput screening, antibody technologies, and pharmacology. It then added expertise in clinical and preclinical manufacturing. The company now employs about 1,100 people, 800 of whom work in drug development.

HGS is conducting clinical trials on 10 compounds to treat cancer, autoimmune diseases, hepatitis C, growth-hormone deficiency, chronic venous ulcers, and immunodeficiencies. Its Albuferon-alpha (albumin-interferon alpha), a novel, long-acting form of interferon alpha, was approved for clinical trials in 2001 for the treatment of hepatitis B and C and a broad range of cancers.

Haseltine suggests that the financial community has to catch up with developments in the sector to identify companies that create the greatest value. "One of the fundamental confusions of the early days was thinking that the service model was a pharmaceutical model," he says. "The market has to learn the distinction between service companies, instrument companies, and pharmaceutical companies. They have very different structures, developmental time frames, expense structures, and report structures. They are different beasts altogether."

ANOTHER COMPANY that was launched in 1993, CuraGen, announced a reorganization similar to Millennium's last month. CuraGen filed its first Investigational New Drug Application (INDA) this year for CG53135, a protein the company identified four years ago as a possible treatment for oral mucositis, an inflammation of the mucous membranes that is a side effect of chemotherapy and radiation treatment. Next year, it plans to file an INDA on the same molecule for inflammatory bowel disease and another INDA for CR002, a fully human monoclonal antibody against platelet-derived growth factor D in kidney inflammation.

Intellectual property is a key asset in drug discovery, according to Fred Aslan, manager of corporate strategy and investor relations at CuraGen. "CuraGen was one of the first genomics companies out there, which gives us a little bit of a first-mover advantage," he says. "We spent a substantial portion of our initial assets identifying novel genes. We have a nice portfolio of intellectual assets, and we're in a position to advance these into the clinic. If another company were to start on this today, it just wouldn't be able to raise the money to do it. There aren't enough resources around today to do the legwork."

Other companies also spent the 1990s developing intellectual property in genomics. Celera, for example, was founded to work on sequencing the genome and developing a database. Adding proprietary drug discovery was a relatively recent change in course. The company acquired Axys Pharmaceuticals in 2001, which added medicinal chemistry and X-ray crystallography to its nascent drug discovery platform. Axys also brought with it a preclinical small-molecule pipeline.

	Ordoñez CELERA PHOTO

In 2002, Celera made headlines with changes in management that underpin the company's new commitment to developing drugs. Most notable was the replacement of CEO J. Craig Venter, whose high-profile face-off with the public Human Genome Project in the 1990s established him as the archetypical genomics entrepreneur (C&EN, Aug. 19, 2002, page 45). He was replaced by Kathy Ordoñez, the former president of Celera Diagnostics. Ordoñez previously spent 15 years at Hoffmann-La Roche, most recently as CEO of Roche Molecular Systems.

According to Robert M. Bennett, director of investor relations at Celera, the company saw its original business, an online database called Celera Discovery System, losing value over time, especially with the increased availability of public databases. Database sales and marketing was shifted to Applied Biosystems, a sister company owned by Celera's parent, Applera. Celera maintains the revenue stream and access to intellectual property, Bennett notes.

"Our hope is to get products into clinical trials in the coming year, and we're building an infrastructure to get them through Phase II trials," Bennett says. Axys had little drug development activity, he adds, and Celera has been enhancing development largely through hiring.

Robert F. G. Booth, Celera's newly appointed chief science officer, also from Roche, has built the company's development organization to about 30 scientists. Bennett says the first candidates moving into the clinic will be small-molecule compounds from Axys or candidates that Celera develops in its partnerships with pharmaceutical companies including Aventis and Merck.

Despite the shift in business focus and management at Celera, Ordoñez says she does not see pure and applied science in shades of black and white. "I think you need pure researchers who are interested in models, biology, and medicine at their purest level. But they need to be able to interface fluently with applied scientists who are concerned about taking things further downstream," she says. "What you need is leadership--people like Robert Booth who understand the importance of both research and development."

Friedman INCYTE PHOTO

INCYTE'S RECENT HISTORY is nearly identical to Celera's. The company's original database services are now run by a separate management team in Palo Alto, Calif., with top management--in this case newcomers from DuPont Pharmaceuticals--moving this month from leased space in Newark, Del., to a permanent facility at the DuPont Experimental Station in Wilmington, Del. There, Incyte will focus on drug development.

"It's the same set of laboratories where many of us worked when we were at DuPont," says CEO Paul A. Friedman, the former president of R&D at DuPont Pharmaceuticals who took the position at Incyte about a month after the sale of DuPont's business to Bristol-Myers Squibb in 2001. "A majority of scientists here are very excellent people that I was able to recruit out of Bristol-Myers Squibb," Friedman says. "They had formerly been DuPont Pharmaceuticals biologists and medicinal chemists."

After 14 months putting together a development infrastructure, Incyte recently picked its first compound to enter clinical toxicology development: a CCR2 chemokine receptor antagonist for chronic inflammatory diseases. The company believes the compound will lead to small-molecule therapies for rheumatoid arthritis, multiple sclerosis, and inflammatory bowel disease. Earlier this year, the company acquired Maxia Pharmaceuticals, which has a novel insulin sensitizer in late preclinical testing.

While few drug discovery firms have made changes on the scale of Celera's and Incyte's, most have experienced some level of reorganization, divestiture, downsizing, and false starts as they pushed compounds toward the clinic. ArQule acquired Camitro, a computational technologies firm, in 2001. But in December, the firm announced that it was restructuring in order to focus on drug discovery and proprietary research. It recently closed a Camitro laboratory in Redwood, Calif., and sold another one in Cambridge, England, to Inpharmatica.

ArQule CEO Stephen A. Hill says Camitro's specialization in predictive models to assess absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties of chemical compounds was seen as a good fit with the company's strategy of vetting compounds for therapeutic performance early in discovery. "But despite our efforts, we were unable to commercialize the technology," he says. Instead, ADMET screening is now used internally as ArQule works on balancing its revenue-generating partnerships with proprietary research.

It's a hard balance to find, Hill says, given what he sees as a near crisis in biotechnology that is centered on early research. "The productivity of drug discovery is deteriorating," Hill says. "If you look at the total number of dollars put into drug discovery and look at what comes out at the end, it is not a pretty picture. There is a real problem there for the future of the industry in terms of making drug discovery productive."

COVERING THE BASES An Incyte scientist works in the company's growing drug development effort. Incyte began as a genomics database service.
INCYTE PHOTO

IT IS ALSO HARD to make it pay. "The cost of doing target ID and validation, screening, lead optimization, and preclinical trials far exceeds the value of out-licensing a compound," he says. "If you stop before preclinical development and out-license your compound, you can make a business of it. But to succeed with proprietary R&D, you need to go as far as Phase II." So, Hill says, it's more cost-effective to license outside technology, or in-licence it, at Phase I.

Essentially, pharmaceutical and large biotech companies can in-license lower cost, higher quality programs than they can produce from their own research efforts, Hill says. "So at the moment, innovation is being done in one place, but the value is captured someplace else," he says. This could lead to a dearth of licensable technologies five to 10 years from now.

In an ideal situation, Hill says, companies will maintain in-house discovery efforts to ensure they will have an alternative to licensing outside technology in the years ahead. ArQule enhanced its ability to do so this month with the acquisition of Cyclis Pharmaceuticals, a firm with a strong biology focus that will add target identification and validation to ArQule's chemistry-intensive lead generation and optimization know-how. Cyclis has filed an INDA for an activated-checkpoint therapy compound for cancer that Hill says will go into Phase I clinical trials later this year.

Coming up next at ArQule is a p38 mitogen-activated protein (MAP) kinase program for arthritis in which the firm employed chemotype-shifting technology and predictive ADMET models in identifying eight leads from its library of kinase targets. The company, which Hill says has $80 million in the bank and a "fairly modest burn rate," is also advancing an ion-channel program focusing on two targets for the treatment of neuropathic pain.

Major partnerships at ArQule include a collaboration with Bayer to design a library of compounds for screening against Bayer's proprietary targets and a deal with Pfizer in which ArQule will provide expanded lead generation capabilities and share its library design and informatics platforms.

Conway ARRAY BIOPHARMA PHOTO

Like ArQule, Array BioPharma incorporates collaboration and proprietary research in its business model. "The collaboration has given us the cash flow to build out the pipeline," CEO Robert E. Conway says. The company is working on a total of 26 targets in partnerships with Icos, Eli Lilly, Roche, Takeda, Japan Tobacco, and others. Array's proprietary research uses the same technology it brings to collaborations, including structural biology, high-throughput screening, lead generation and optimization, drug-metabolism testing, and current Good Manufacturing Practice-certified manufacturing. Most advanced is an MEK inhibitor in toxicology testing and a p38 anti-inflammatory.

Array has raised $100 million in public and private financing since going public in November 2000. "We have $40 million in cash, which gives us several years of runway," Conway says. The firm is giving no guidance on when it will achieve a profit. "The goal now is to keep the burn rate manageable and create proprietary drug candidates."

The strategy is similar at Inpharmatica, which started as a bioinformatics company, building a chemical genomics predictive modeling platform called PharmaCarta and an online database called Biopendium. The company, which gained a lab with ADMET and medicinal chemistry capabilities with the acquisition of ArQule's U.K. operation, had partnerships and proprietary research in its original business plan, according to Inpharmatica CEO Malcolm Weir.

"Our business model has always been the hybrid one--maintain service revenue from the platform to support development activities," Weir says. "But the true value in discovery activities is only ever really unlocked by participating directly." Service and technology platform activities are hard to maintain as a growth business, he adds.

	Weir INPHARMATICA PHOTO

Inpharmatica recently received its first milestone payments in a partnership with Serono in which Inpharmatica is developing a novel secreted protein it identified using PharmaCarta. In its own lab, the company is working on various compounds based on nuclear receptors and cytochrome P450 enzymes, Weir says.

Drug discovery takes different routes depending on technology, but business models all gravitate toward a mix of partnerships and proprietary research. Development follows a low-cost, high-volume path.

"We learned that the cost of discovering an antisense inhibitor is relatively low compared to the overall cost of drug discovery," says C. Frank Bennett, vice president of antisense research at Isis Pharmaceuticals. "Therefore, rather than trying to decide if we want target X, Y, or Z, we develop inhibitors to all the targets, compare them head-to-head, and let the data help us make our decision on which ones to focus on downstream. The antisense inhibitors we use are also our drug products."

The firm has one commercial product, Vitravene: a cytomegalovirus retinitis drug marketed by Novartis Ophthalmics. Isis has 13 drugs in development, two of which are in Phase III clinical trials. It has collaborative partnerships with a dozen companies, including Lilly, Amgen, and Chiron.

As time pressure and cost of operation increase, technology becomes less of a distinguishing factor, and speed becomes a primary competitive advantage, according to Michael M. Morrissey, senior vice president of discovery research with Exelixis. "Over the past two or three years, we built a strong discovery engine that spans target discovery, validation, drug discovery, preclinical development, all the way through to clinical work," he says. "The mind-set is on great speed and high quality."

Morrissey acknowledges that the competition from big pharma as well as other biotechnology companies is intense. "We have to be very agile in how we do our science and in how we reach go/no-go decisions in order to put compounds into the clinic as rapidly as possible." This, he says, leads to greater integration of discovery and development efforts. "We will always maintain a significant research effort because that's where the targets and our novel insights come from. We aren't losing that capability, but are building on it as the foundation of a clinical research organization," Morrissey says.

In fact, he views Phase I and early-Phase II studies as extensions of discovery, and the earliest research is done with an eye on questions that need to be answered downstream. "We are very aggressively focused on the issues that revolve around having a drug be effective in the clinic," Morrissey says. "You want to ask the questions about target relevance and compound activity in the appropriate model, which is a human disease setting in which you gain clear insight into the role your compound or target has in a given disease pathology." ADMET and pharmacokinetic data are gathered from the start, he says.

Exelixis currently has a rebeccamycin analog licensed from Bristol-Myers Squibb moving into Phase III clinical trials. Other compounds in the pipeline were developed in-house. An INDA for XL784, a potential cancer drug, was filed in March, and Exelixis has started Phase I safety testing of the compound. Several other compounds are in preclinical testing.

Partnerships continue to play a large role at Exelixis. Last October, the firm announced a deal with GlaxoSmithKline under which it plans to discover, develop, and commercialize novel therapeutics in the areas of vascular biology, inflammatory disease, and oncology.

Aventis' Dwarki does not suggest that service-based partnerships have completely dried up in the pharmaceutical industry. There is still room to supply basic tools. The biotech sector, he says, needs to do some work on discovery and development technologies, and any supplier with a breakthrough may get some business. "It's easy for companies to come up with early gene identification," he says. "But toxicogenomics technologies are not in place right now. There is a big gap in overall understanding."

PHARMACOGENOMIC and pharmacogenetic technology are developing at a very fast pace, but they are not yet feasible in large applications, Dwarki says. "A company might say it only costs a dollar to do something on one patient, but it's mind-boggling to think of using the technology on your whole clinical trial practice. It becomes extremely expensive."

Dwarki says Aventis is accelerating work in its own lab by merging discovery and development. "Aventis is probably the only company that doesn't call what we do R&D. We call it drug innovation and approval," he says. "Lead generation and lead optimization activities are done in parallel."

Many changes in drug discovery are endemic to a new branch of science establishing itself in a business context. As the science of genomics matures, interest naturally shifts from wide-open technology to the fruits of that technology: new drugs. And industry watchers agree that as drug discovery evolves, the cyclical nature of business will result in a serious look back upstream for new science before too long.

"People have short memory spans," ArQule's Hill says. "At the moment, everybody is interested in products. But there will come a time when people will think that drug discovery is just not efficient enough. Different technologies that improve efficiency will become flavor of the month again."

With business management in place and proprietary research moving into clinical trials and commercial production, strategic realignment is not likely to be a major feature of future cycles. "It's unwise to change your business model based on short-term swings in sentiment," Hill says.

Chemical & Engineering News
Copyright © 2003 American Chemical Society