(Business 2.0) – One day last February, a 60-year-old man with sandy-gray hair checked into the Ritz-Carlton in Half Moon Bay, Calif., under the assumed name Fred Drake. Before entering a windowless conference room, he turned off his cell phone. If he needed to make an outside call, even to his wife, he would have to go through an attorney.

The need for secrecy was obvious to anyone familiar with insider trading regulations. "Drake" was about to learn the outcome of a project that might yield billions of dollars in sales around the world. But no one understood better than he that the project was about far more than money. His real identity was Don Francis--scientist, entrepreneur, and media-drenched AIDS crusader--and the project in question was the first large-scale human trial of an AIDS vaccine. It was an effort that had consumed 10 years of his life. He had formed his own company, VaxGen, to keep the vaccine alive after the government declined to fund it, and he'd spent $120 million of investors' money on clinical trials to prove its efficacy. Now his quest--not to mention the fate of his company and the lives of millions of people in the path of the epidemic--hung on this moment: the "unblinding" of the first trial data.

That night in the cluttered hotel room, amid stacks of printouts and half-eaten trays of food, the dream melted. As the VaxGen scientists and outside investigators pored over data, it became clear that the vaccine had failed. Francis now makes light of his reaction, but other VaxGen veterans describe a disappointment so profound as to be almost physical. "It was as if the floor dropped away," one says. Another exec's mother captured the sense of despair in a pungent e-mail she sent after the results were made public. It contained just two words: "Oh, shit."

Overnight, VaxGen became a poster child for the downside of risk: The stock collapsed, cash continued a dangerous countdown to running out, and in the annual report, released just weeks after the trial results, the company disclosed that its accountants had questioned whether it could continue as a going concern. To survive, VaxGen has had to turn its back on its original mission and seize on a secondary project, a government contract to develop an improved anthrax vaccine. Stigmatized by failure, the company has struggled to convince investors that it's a legitimate enterprise. And it has had to do all this under the klieg lights of scrutiny directed at almost anything having to do with AIDS.

VaxGen's story raises some inevitable questions. Had the company diversified beyond the AIDS vaccine sooner than it did--had it seen itself less as a crusade and more as a business--it may have avoided its current crisis. But business decisions are always easier in hindsight. The expense and risk of drug development means that small companies often can't afford more than one project at a time. VaxGen says that had it tried to pursue multiple projects early on, it would have been much harder to get the vaccine trial off the ground. The fact is, in biotech, you often need to gamble all on a single, compelling dream; and in a business in which 80 percent of drugs that enter clinical trials don't make it to market, you have to be prepared for failure. "We have a win-big-or-die strategy," says Stanley Crooke, CEO of Isis Pharmaceuticals, who has steered his 14-year-old company through its own wild rides.

Six months before VaxGen's launch, Francis told the New York Times Magazine, "This is the story of learning not to be afraid to fail." VaxGen is still finding out how expensive that lesson will be.

Of all the benefits that biotechnology might ever bestow on the human race, an AIDS vaccine has to be one of the worthiest--and one of the most urgent. In its more than 20-year rampage, the epidemic has killed an estimated 22 million people. The virus that causes the disease, HIV, currently resides in an estimated 42 million people and infects nearly 14,000 more every day. In the developed world, expensive drug treatments have made infection mostly manageable. In Africa and Asia, where the epidemic rages most fiercely, it is almost always a death sentence.

One reason a vaccine remains elusive is the sheer scientific difficulty of creating one. HIV is an ingenious adversary, mutating so rapidly that the immune system can't mount an effective response. Even now, scientists don't know exactly what kind of approach is needed to induce protection.

A close second to the scientific hurdles, according to Chris Collins, executive director of the AIDS Vaccine Advocacy Coalition, are the economic disincentives. As opposed to pills that might be taken daily for many years, vaccines are ideally administered only once. Indeed, revenues for all vaccines in the world barely add up to what consumers spend just on the cholesterol drug Lipitor. What's more, there are pressures to keep vaccine prices low, especially for a disease like AIDS, which is prevalent in countries that can barely afford condoms. "When you announce that you're going to work on HIV vaccines, investors run the other way," says John Donnelly, a lead scientist at Chiron, a biotech in Emeryville, Calif.

Despite these major obstacles, researchers had been working in parallel at Chiron and Genentech on a promising potential vaccine, a synthetic protein known as gp120. The protein occurs naturally on the surface of the HIV virus, and exposing people to its synthetic twin could induce the production of antibodies that, in theory, would prevent the virus from invading healthy immune cells. Intriguingly, the scientists had demonstrated in 1989 that gp120 protected chimpanzees from infection.

Development of the vaccine proceeded with painful slowness, however, for both scientific and economic reasons. Success in chimps doesn't necessarily translate into success in humans. That could be proved to the FDA's satisfaction only through human clinical trials, the last stage of which, Phase III, could be expected to cost between $50 million and $150 million. In June 1994 the government's National Institutes of Health decided not to pick up the tab for Phase III testing. Chiron dropped development in favor of other vaccine candidates. Genentech put the brakes on as well. Francis, then a Genentech scientist, was adamant that the vaccine should be tested. He set out to form an independent company to see gp120 through Phase III trials and then manufacture it under license from Genentech. VaxGen was born, in other words, because it seemed the only way to keep the vaccine alive.

While Francis had little experience as a businessman, he brought considerable star power to VaxGen's fund-raising. As a young scientist at the Centers for Disease Control and Prevention in the 1980s, he had tirelessly exhorted reluctant bureaucrats to focus attention on the growing AIDS epidemic, a stand that was lionized in the book And the Band Played On. (Matthew Modine portrayed Francis in the movie.) With a youthful charisma and a flair for dramatic gestures, he proved irresistible to the media. Scientists began to call him "St. Francis," both because of his evangelizing and because of a certain blindness to the possibility that he could be wrong. "Don's friends think he's a passionate crusader," says Patricia Thomas, author of Big Shot, a book chronicling the search for an AIDS vaccine. "His enemies think he's a zealot and nut."

Teamed up with a biotech entrepreneur named Robert Nowinski, Francis went on the road to attract investors to VaxGen. Ever ebullient about gp120's promise ("I know the vaccine will work," he told the New York Times Magazine), Francis played up the chimp study and the vaccine's market potential. "People said that vaccines are bad business. We said there's a niche there," he recalls. "If we make a product, they will come." In the space of four months starting in December 1996, they raised $27 million and eventually set up shop in Brisbane, Calif., with Francis as president and Nowinski as CEO. They hired talent from Genentech, including John Curd, well-known for having shepherded a breast cancer drug through trials, and Phil Berman, one of the original developers of gp120.

Although gp120 had gone through preliminary testing, one of the most challenging tasks lay ahead: running two groundbreaking human trials, one in North America and the Netherlands, the other in Thailand. Given the virulence of AIDS and the stigma attached to it, many experts doubted whether a Phase III trial of a vaccine was even possible. At the heart of the trials was an ethical dilemma: The company was duty-bound to warn its subjects away from the same risky behavior that it needed them to engage in to prove the vaccine's effectiveness. VaxGen employees crisscrossed the United States, Europe, and Asia, fighting rumors that the vaccine could cause AIDS, contending with multiple medical bureaucracies, and deflecting a constant barrage of usually hostile second-guessing from the media.

What kept the company going during these times was a sense of mission. Francis's devotion to the cause resonated with the staff, a number of whom had lost friends or family members to AIDS. "Employees would have followed him off a cliff," Thomas says. The company kept up morale with plentiful staff parties--called "vax-scenes"--to celebrate milestones such as enrolling a certain number of people in the trial. The relentless negative press coverage, from both the AIDS community and doubting scientists, also helped draw the staff together in a kind of siege mentality. "There was a sense that we were this big underdog and the world was against us," says Patti Orozco-Cronin, a former project manager. "We needed to shut out the world beyond VaxGen."

Nevertheless, after the North American and European trial became fully enrolled, in October 1999, some started questioning VaxGen's ultimate mission. What if Aidsvax, VaxGen's name for gp120, failed the trial? Prudence dictated that VaxGen line up other business, but mere talk of diversification triggered heated debate. "Many people were focused on the AIDS piece--their life's work was around AIDS," says one former manager, who thought diversification was overdue. "They weren't drug developers."

Eventually the diversifiers won, and in the fall of 2001, the company replaced Nowinski with a new CEO, Lance Gordon, a veteran of drug development at Oravax and North American Vaccines. Within weeks of his arrival, Gordon called the executive committee together and asked, "Do we want to be a project or a company?" Gordon immediately began looking into other lines of business. The 9/11 attacks, by chance, provided a quick opportunity: a $16.2 million contract with the government to test an improved anthrax vaccine that had been developed by the Army. Meanwhile, work on the Aidsvax trials continued.

The company considered an Aidsvax failure as a scenario, but employees did their best not to dwell on it. Outside VaxGen it was a different story. Scientists had had reservations about gp120 since the 1990s, and many started embracing a strategy to combat HIV after infection. One critic went so far as to deride Aidsvax as "dishwater." But as the unveiling of the data drew near, a VaxGen veteran recalls, the mood was one of confidence: "The question was not 'Would we be successful?' but 'How successful would we be?'"

Late on the Sunday night after the meeting at the Ritz-Carlton, the news about Aidsvax leaked. The next day, the staff convened in "the classroom," a nondescript spot off the company kitchen. By then secrecy was unnecessary. The New York Times had just run the story on its front page: "Large Trial Finds AIDS Vaccine Fails to Stop Infection."

A number of employees in the room had been working through the weekend, analyzing data and preparing for the onslaught of press. They now sat on the floor or on cabinets, exhausted and dazed by disappointment. Company execs say they served champagne, though in the gloom of the moment, some staffers didn't even notice. "I can't describe what the feeling was," one employee says. "'Zombie' might be a good word." Francis, incongruously upbeat, stood before the staff and put up slides of trial data, praising the "beautiful study" they had pulled off. As if programmed by his DNA, he went into rallying mode, speaking of the possibility that in the data lay clues to an eventual vaccine.

He had seized on what many viewed as a statistical glitch: Results from the trial's small sample of minority participants suggested that the vaccine might have helped them. Francis and others drew public attention to the evidence, slender though it was. The ambiguous data triggered yet another internal battle for VaxGen's soul--and still more occasion for the AIDS community to pound on the company. Some accused VaxGen of offering false hope to minorities (not to mention investors). Meanwhile, Francis passionately argued that VaxGen couldn't turn its back on any sign that Aidsvax was effective. Piers Whitehead, the product development chief, took it upon himself to be, as he puts it now, "the coldhearted businessperson." Even if they could agree on the significance of the minority results, he pointed out, there were other issues. "You have to factor in financial considerations, PR and ethical issues, investor expectations," he said. "It turns into quite a complex algorithm."

In the end, money--or lack thereof--talked. As of the end of March 2003, VaxGen had about $10 million in the bank, enough to carry it only through November. It had already lost a total of $132 million. To add to VaxGen's headaches, the company had become the target of class-action lawsuits alleging that its executives knew in advance that the drug had failed. The company appealed for funding from the government or foundations to continue work on Aidsvax. But after wrapping up the Thai trial, VaxGen would spend no more of its shareholders' money.

The company quickly shifted its emphasis to anthrax--and now sounds like a die-hard believer in diversification (a little late in the game, some company veterans note ruefully). "If you don't start thinking about setbacks until they happen, you're dead," says CEO Gordon. VaxGen laid off about 10 percent of its workers and transferred others from Aidsvax to anthrax. As a business proposition, an anthrax vaccine is far less risky than Aidsvax. For starters, the vaccine needs no Phase III trials, since the disease rarely occurs naturally in people and ethics would preclude exposing patients to the disease. Most important, the federal government is funding the development work and is likely to buy the vaccine if it's a success.

VaxGen has made some progress in convincing the world that it now has a lower risk profile. In the spring, the company raised $12 million in private equity placements. And when VaxGen got the nod to conduct a Phase I trial, the stock leaped 68 percent to $5.62. Gordon even speaks of VaxGen's becoming profitable next year if it wins contracts to test and supply the vaccine for government stockpiles. Meanwhile, VaxGen is looking to license new drugs for development and is pursuing the potentially lucrative niche of manufacturing drugs for other firms. In partnership with overseas investors, VaxGen just finished construction of a biologics manufacturing facility in South San Francisco and broke ground on a bigger one in Korea.

It all adds up to little more than a toehold on survival, but if it works out, VaxGen would not be the first biotech to come back from the brink. (See "Comeback Kids," at left.) Buzz Burlock, who runs a health-care hedge fund in San Francisco, argues that brushes with extinction are the norm rather than the exception. "Man, if you're not taking risk, you're not doing anything in biotech," he says. But skeptics point out that the company can't survive forever on the low-margin biodefense business and that no manufacturing customers or licensing partners have so far appeared. Says Ron Garren, chief biotech strategist with InvestBio of New York, "Why would anyone give them money? They'd be buying a prayer."

As for Francis, he is seen less often in the hallways these days. He has made an effort to redirect his crusading verve to anthrax. "If some asshole sprays anthrax in an airport," he says, "we'll be damn glad to have a vaccine."

Yet it's difficult to shake the feeling that for Francis the company's mission has never changed. He observes sardonically that anthrax is hardly a mass killer, and he still hopes the government will fund further study of Aidsvax or that the Thai trial, the results of which should arrive late this year, will vindicate the vaccine. "I probably never think about the business reality," he says, laughing. Then he turns serious. "The business ultimately is a vaccine for AIDS. The reality is, How are we going to pay for it in the meantime?"

Susan Orenstein (sorenstein@business2.com) is a senior writer at Business 2.0. Additional reporting by Bob Porterfield.