William A. Haseltine had a 55-acre, half-billion-dollar compound and a manufacturing plant built for his former biotech company, Human Genome Sciences, despite its having never brought a single product to market.
With a facade made entirely of reflective glass, the six-story corporate headquarters juts into the sky almost invisibly. Inside, an atrium-style lobby has a lush garden and a 100-foot-long sculpture that depicts the birth of a protein. The floor is made of fossilized stone tile, and light filters through panels in the ceiling.
But Haseltine's vision for a new company just barely includes four walls.
In a recent Ernst & Young global biotechnology report, Haseltine, who retired as chairman and chief executive officer of Human Genome in 2004, predicted that the biotech business model of the future would be all about outsourcing. Management would keep office space minimal and contract "almost all of their research, process development, manufacturing and clinical trials to specialty services companies."
In science circles, the concept has come to be known as the "virtual company" in which costs are kept low by maintaining only a handful of in-house executives to oversee development projects distributed across the globe.
"It used to be much more of a reality in the IT space where it seems like outsourcing is an easier thing to do, but [over the past few years] it has gravitated over to the biotech space as more and more good labs are created in China, Japan and Korea," said Mark Heesen, president of the National Venture Capital Association based in Virginia. IT is the industry term for information technology.
Ruxton Pharmaceuticals in Baltimore County has only one employee watching over its development of a treatment for ALS disease. RegeneRx Biopharmaceuticals in Bethesda has eight on staff, most of whom manage up to six contracts each with outsourcing companies. And Neuralstem Inc., a Rockville company using fetal stem cells to develop central nervous system treatments, is made up of a two-man team.
"We outsourced everything that could be outsourced," said Richard Garr, Neuralstem's co-founder, chief executive and chief financial officer. The company is a decade old and at one time had as many as 50 employees, but funding setbacks a few years ago led Garr and his co-founder and chief scientific officer, Karl Johe, to lay off the staff and reorganize in 2004 to its much-leaner form.
The explosion of "contract research organizations" capable of doing quality biotech work has led many companies to outsource at least some of their operations, but the extreme virtual version is relatively new, having begun about a decade ago.
At the time, it was largely frowned upon by investors, who were used to counting offices and factories as assets. But that perception has changed as it has become more expensive to bring a drug to market, and venture capitalists such as Heesen have learned to appreciate the idea of spending less on infrastructure and more on product development. CEOs say they like the flexibility of being able to change course on a dime if something goes wrong -- without having to lay off hundreds of employees.
"You're betting on science, and it's very difficult to schedule and predict what's going to happen," said Bruce Robertson, who lives in Maryland but splits his time between here and Atlanta, where he is managing director of the investment firm H.I.G. Ventures.
"The question is when you're wrong -- and you know you probably will be -- then what happens? What are the cons of that? If you've got a big in-house [operation] with a high burn rate, the consequences can be more severe. In general, it's more capital-efficient, you can use your capital more efficiently on an outsourced basis, particularly in the early days."
It is a stark contrast to the philosophy of just a few years ago, when investors were caught up in the excitement of a much-publicized race to map the human genome, reminiscent of the fervor for dot-coms. They gave hundreds of millions of dollars to developing biotechs, and the companies often turned around and used the money to elaborately outfit offices and construct manufacturing plants for treatments that frequently failed to materialize.
After it became clear that drugs based on a person's genetic code were years away from realization, much of the money dried up, and many companies were left with facilities and employees they could not afford.
That was the case for Rockville's Celera Genomics, whose founder is credited, along with the National Institutes of Health, as being the first to sequence human DNA.
"Celera and Human Genome are in the top 10 for the category of 'build it first and then think about it later,'" said David S. Block, a former Celera vice president who has started a, for now, one-man company in Baltimore County. His business, Ruxton Pharmaceuticals, is developing drugs for neurodegenerative diseases.
"Everything that was built at Celera was subsequently taken apart," he said. The 300 gene sequencers went to the refuse bins. The plan to develop drugs was trashed as well: The company announced this year that it was abandoning those efforts.
"Celera burned through probably a billion dollars and decided to get out of the drug game because it's too hard," Block said.
Experts estimate that it takes a dozen years and a billion dollars to commercialize a single drug, in part because of increased regulatory requirements, but also because of payroll and construction expenses. Human Genome Sciences, under new leadership, recently announced plans to sell its headquarters and manufacturing facility and lease them back in order to bankroll the company's drug development.
Block has brought in $5 million in venture investment thus far and has half of it left after two years, he said, largely because he does not have to pay staff other than himself and has only one office. The bulk of the work -- compound screening, animal model testing -- is being done through contractual relationships.
He is looking to hire other help, but most of the responsibility will still fall on his shoulders.
"It's stressful, you know, it just is," he said of working alone. "But it allows you to do a very important thing: to not churn through the money. ... The bottom line is you're trying to manage your expenses as carefully as possible while moving things forward without losing any time."
According to a January report from the Tufts Center for the Study of Drug Development, those who use contract companies in developing their drugs get them to market faster than those who do not. And clinical trials are completed 30 percent more quickly by outsiders than insiders, according to KMR Group data analysts. That could translate to a time savings of five months and additional revenue potential of $150 million based on average sales.
In the six years since re-inventing itself as a virtual company, RegeneRx Biopharmaceuticals has begun three clinical trials testing the skin-healing properties of a molecule that the founder discovered, and it has plans to initiate two more. But CEO J.J. Finkelstein cringes at the "virtual" label.
"It sounds like it's not a real company, like it's just make-believe," he said. "I prefer 'outsourcing company.'"
From 2,300 square feet of Bethesda office space, RegeneRx's staff of eight full-time employees watches over contractors and universities -- through research partnerships -- conducting the company's business, including its clinical trials and drug development. Finkelstein, one of the company's founders in 1982, said this was the plan he wanted for the business all along.
"In the '80s, I tried to do it with this company, but things were very different at the time," he said. There were very few vendors available to do the work, and Wall Street wasn't ready for the virtual company. Instead, RegeneRx took the usual route, with research and manufacturing facilities set up in California and a good deal of office space.
Finkelstein left the company in 1989, before a setback in the mid-1990s forced RegeneRx to get rid of its California facilities, reduce its staff to three from 29, cease research and development and move to smaller offices.
"They had essentially a lame-duck president, in a single office trying to pay the bills and figure out what to do. .... That's when we came in," said Finkelstein, who was lured back in 1999 and organized a core group of investors with the caveat that the business plan would be outsourcing.
After spending the first three or so years reorganizing and working on a game plan, Finkelstein hired his first employees in 2002 and now hopes to move a drug to the market within the next two years.
Going virtual is not for everyone, though, cautions Hal Broderson, managing director of Rock Hill Ventures near Philadelphia. He has been in the life sciences business for two decades, and his venture capital firm has put together a portfolio of about a dozen virtual companies since 1996, back when, he says, his crew was "a little bit of pioneers with arrows in our backs." Since then, he has given talks on the topic at two of the Biotechnology Industry Organization's annual conferences during the past decade.
"When GlaxoSmithKline comes in to look at your company, it wants to kick tires and see beakers and Bunsen burners and a lab. We don't have that because we rely on contractors to do that for us," Broderson said. "That isn't what they're used to seeing."
At the same time, most bigger companies will not pay for those things, he added. When Johnson & Johnson bought one of Broderson's companies, it did not need the manufacturing facility or human resources and accounting staff that came with it. J&J; just wanted the product. "They weren't going to give us a return on that infrastructure," Broderson said. "So we sort of scratched our heads and said, 'Why build this infrastructure if we're not going to get credit for it?'"
His virtual companies require far less capital to sustain, but that, too, has been a problem. Venture capitalists often want to invest large amounts -- upward of $25 million -- in the hopes it will beget large returns. Virtual companies often need less than $10 million over several years.
The venture capital industry plans to invest $40 billion over the next couple of years, according to analysts who put together the MoneyTree Report, which chronicles such spending. About 20 percent of that -- or $8 billion -- will go toward biotech-type businesses.
"The problem is venture capitalists have so much more money under management now that the model doesn't make sense from their point of view," Broderson said. In other words, the virtual investment is just too small to make it worth many investors' time.
And others just think it is a bad idea.
"I don't see a lot of companies becoming virtual," said Roger Longman, a managing partner at Windhover Information Inc. of Connecticut. His company provides analysis and commentary on the drug-development world.
"A lot of these processes have been deemed too important to outsource. It's much more important to have those people working for you so you can keep the quality to a certain level, or at least keep yourself believing that you can keep the quality at a certain level," he said.
Morrie Ruffin, vice president of business development for the Biotechnology Industry Organization based in Washington, said virtual or fully built-out companies are not the point when it comes to assessing a business, but intellectual property is.
"Investors aren't going to buy the company just because they're outsourcing. They're investing in the technology," Ruffin said. He acknowledged that the virtual trend has been developing over the past few years but added: "I would hesitate to draw any conclusions at this point."