It seemed to happen every quarter. Michael Becker, the CEO and president of Cytogen Corp. from 2001 to 2007, would chat with his wife at dinner about the state of his biotech company. He would lay out Cytogen's earnings, and she would pose the same simple question: How is it Cytogen doesn't make money despite being around for decades?

"The short answer is," Becker recalls, " 'Only in biotechnology.' "

The biotech industry today is a far cry from what it was in the 1980s when companies were flush with cash from investors excited by the promise and potential of the underlying science. But losses were large, investors became more discriminating and realistic, and survival became a much more difficult beast to tame.

"If you have a good idea now, you need to have revenue on sight," says Michael Ross, one of the first employees at Gen­entech Inc. in the early 1980s and now a managing partner at the San Francisco office of London-based venture capital firm SV Life Sciences. "The public market doesn't want to finance clinical and regulatory risks."

A few elite biotechs have been able to consistently churn out cash-generating products: Amgen Inc., Genzyme Corp., Biogen Idec Inc. and Genentech.

For instance, Genzyme and its orphan drug pipeline, which it began building in the mid-'80s, generated positive revenue for years before selling to Sanofi-Aventis SA in February for $20.1 billion. In 1982 Genentech won approval from the Food and Drug Administration for its synthetic insulin, which it licensed to Eli Lilly and Co. Over the next 27 years, Gen­entech had its financial ups and downs, but managed to commercialize a number of blockbuster biologic drugs. In 2009 Roche Holding Ltd., which had held a major stake in Genentech since 1990, acquired the rest for $46.8 billion.

Those winners are rarities. Among the many hundreds of biotechs formed since the '80s, many have failed, shut down or been sold off. More than a few, however, have continued to scrape by, a testament to persistence and technologies that continue to show promise. They have become the hardy survivors of biotechnology -- still chasing the dream after three decades of life.

For all the Genentechs and Genzymes, the biotech norm may well be a company like Cytogen, which for decades has had to get by without the kind of earnings generated by a major product. Cytogen's core technology -- a technology called monoclonal antibodies -- goes back to the earliest days of the industry. A number of companies were started early on to pursue much-ballyhooed antibodies. Two of them, Hybritech Inc. and Centocor Inc., were sold, to Eli Lilly and Johnson & Johnson, respectively in the '80s.

Monoclonal antibodies represent one of the technological pillars of modern biotechnology. Antibodies are used by the immune system to seek out and destroy bacteria and diseases, tracking down and binding to what are known as antigens -- small receptors on, say, a bacteria or virus -- thus either neutralizing the problem or notifying the immune system to do so. Different antibodies can target and bind to specific bacteria and diseases.

Monoclonal antibodies are created en masse by cloning immune cells that have been infected with a specific disease, thus producing specific antibodies. Though the idea of creating antibodies to target disease-ridden cells dates back to the beginning of the 20th century -- German chemist Paul Ehrlich coined the phrase "magic bullets" for antibodies -- scientists didn't accomplish the feat until the '70s.

By the '80s, everyone seemed to be making a play at researching the use of monoclonals, that is, antigen-specific antibodies, both as diagnostic tools and therapeutic treatments for some of the most serious diseases, including cancer. Using monoclonals to diagnose diseases was an early focus; scientists believed it might be possible to detect disease early, before it spread too deeply. And there was a commercial kicker: An antibody useful in diagnosing a disease might also be used to deliver a drug or toxin to kill the bacteria or cancer cell. Thus, in antibody research, diagnosis and treatment were tightly linked.

Still, the science behind monoclonal antibodies -- producing them, accurately targeting them, finding the right receptors or antigens for them -- proved more difficult and costly than most imagined. Additionally, scientists had trouble when the initial crop of monoclonal antibodies, made from mouse cells, interacted with the human immune system, although science was eventually able to develop "humanized" monoclonals.

While some researchers tackled the intricacies and pitfalls of antibodies in diagnostic development, others labored away on therapeutics and the effort to link some sort of drug or toxin to the distinctively Y-shaped molecule. Again, it took time and money before any kind of measurable success came from the work -- efforts made even more difficult because of the complexities of, say, cancer as a disease. Though nearly a dozen therapeutic monoclonal antibodies have been approved for cancer, as well as some for other diseases, the first were only approved in 1996. None have proved to be a panacea.

Many companies spent decades toiling without bringing their treatments to fruition. Others have had minimal success, and suffered years of losses.

Three of these antibody pioneers managed to survive decades of losses: Morris Plains, N.J.-based Immunomedics Inc.; Berkeley, Calif.-based Xoma Ltd.; and Cytogen, which was acquired in 2008 by Eusa Pharma Inc. How did they do it? Each has its own story, its own near-death moments and periods of euphoria, but the brute fact of survival in capital-intensive biotech is access to cash. The executives and boards of all three companies successfully raised and husbanded capital, despite downturns and disappointments. Executives came and went, strategies and models shifted, and while the three have yet to develop blockbusters, they did stay afloat with licensing agreements, R&D partnerships, sales of stock or products.

Here are their stories:

Immunomedics

Compare Immunomedics in 1982 with the company in 2011, and the differences don't seem very great.

Today the company is located in the same state, New Jersey, as it was in 1982, when it was founded. One of its founders, David Goldenberg, who spent years as an academic at the University of Kentucky before he co-developed Immunomedics' monoclonal antibodies, remains deeply involved, now that the chief executive's job he had held passed in 2001 to his wife, Cynthia Sullivan.

Immunomedics' research still targets cancer through monoclonal antibodies. And the biotech still loses money.

But looks can be deceiving. Immunomedics went public in 1983, raising $2.5 million. The company began life using diagnostics to test for autoimmune diseases and cancer, and saw some moderate success in the mid- to late '90s with two products, CEA-Scan and LeukoScan, both diagnostic-imaging products that use monoclonal antibodies to target cancer and infectious diseases, respectively. The two sold steadily, with revenues peaking in 1999 at $7.5 million before fading, but neither was enough to fund the company's development, and Immunomedics ran losses through the '80s and '90s as it continued to work on a family of antibodies.

By 2000 the company faced a strategic crossroads. Sullivan says the environment for diagnostic products was changing as regulators became increasingly strict on data they demanded from diagnostic clinical trials. But the company clearly needed another story to tell investors.

With net losses as far as the eye could see, high levels of paid-in capital and falling revenue, Immunomedics had to change. The company had raised $111 million in equity between 1982 and 1999. But it was burning cash. By 1999, Immunomedics was down to around $9 million from a cash safety net of about $30 million it had maintained throughout the '90s.

Biotechs, like their investors, have to play the cards they're dealt. "We hop into the markets if it's appropriate to do so," Sullivan says. "We have a clinical development plan. We are opportunistic."

Immunomedics had a backup plan that involved that link between antibody diagnostics and therapeutics. Using antibodies from its diagnostics, the company tried to shift in 2000 to therapeutics with a promising drug candidate for non-Hodgkin's lymphoma, epratuzumab. But while the upside in therapeutics was higher than in diagnostics, the costs were too. R&D required between $15 million and over $20 million a year. Where would the money come from?

Immunomedics turned to investors, who generally get more excited by therapeutics than diagnostics. In 2000 the company raised a further $50 million from stock sales. Despite the losses, the market cooperated; this was one of the last boom years for biotech, fueled by the excitement over the human genome project. The new capital helped move epratuzumab through initial clinical trials, though eventually Immunomedics licensed it to Amgen, a deal with an up-front payment of $18 million. Such licensing payments won't get a biotech over the top, but they provide some cushion and credibility.

That partnership with Amgen ended in 2004 without the drug's commercialization. Immunomedics then received back the rights. It retained rights to oncology indications, but eventually out-licensed the antibody's autoimmune indications again, this time to Brussels-based UCB SA in 2006 for as much as $38 million.

Two years later, a similar antibody, veltuzumab, also undergoing clinical trials for non-Hodgkin's lymphoma and autoimmune diseases, earned Immunomedics more cash: $40 million as an up-front payment in 2008, and $10 million in milestones in 2010.

Those payments have kept the company alive. In recent years, Immuno, as Sullivan calls it, has been able to sell debt to investors such as BlackRock Inc. And it has maintained longtime investors, including Fidelity Investments.

Sullivan points to Immunomedics' antibody R&D as a reason the company has been able to maintain those relationships, noting that certain investors are interested in work that may affect specific diseases. But she also credits recent successes. In November, the company announced positive results from a Phase 2b study of epratuzumab for treating lupus. "There's the potential," she says.

But the marathon is not yet over. Sullivan notes that today she sees Immunomedics not as a 30-year-old firm, but as a 12-year-old developer of therapeutics. "It's really not that old of a story," she says hopefully.

Xoma

Xoma is a story of change -- slow change. The biotech's penchant for adaptation to the demands of the financial markets stems from its skill with antibodies, which provided the raw material for a number of promising products, if not blockbusters, that attracted a steady stream of new money.

Xoma was founded in 1980, went public in 1986 with a $32 million offering and, of the antibody trio, has seen the greatest success, primarily through licensing agreements and its ability to raise capital. With all this success, however, failures frequently followed.

Xoma's high point may well be its 13-year collaboration with Genentech, which began in 1996. The two companies agreed to develop monoclonal antibody anti-CD11a, which was eventually used to treat rheumatoid arthritis. After years of clinical trials, rebranding of the drug with the name Raptiva and new and renewed financial agreements, Raptiva was eventually approved by the Food and Drug Administration and began generating sales.

Revenue from Raptiva was important to Xoma, which had been operating at a loss since the company's inception. The early years of any biotech are spent funding costly research and early-stage clinical trials. It becomes even more expensive as the drugs enter Phase 2 and 3 of clinical trials.

The early '90s were particularly difficult, as the company spent heavily on research and development and clinical trials to develop itself to full potential, incurring a loss of $47 million in 1992. Though that figure had dropped considerably by 1996, it still lost $29 million that year because of its costly operating budget. Xoma did have revenue during the time, but it came primarily from out-licensing products and technologies.

Xoma began receiving royalty payments for Raptiva in 2005 -- a mere quarter century after its founding -- from Genentech's U.S. sales and from its international partner, Serono SA, now Merck Serono SA. Those royalties continued through 2008, ranging from $6 million to $17 million a year.

The royalties also included revenue from another product Xoma created with Genentech, Lucentis, an injectable treatment for age-related macular degeneration, an eye disease. Xoma eventually sold out its royalty interest to Gen­entech in late 2009.

The time to revel in the revenue was cut short, however. Lawsuits related to Raptiva's clinical trials began surfacing, alleging health consequences from the drug, according to Securities and Exchange Commission filings.

Genentech issued a "Dear Healthcare Provider" warning about Raptiva in November 2008. Two cases of progressive multifocal leukoencephalopathy, a disorder that damages protective coverings of nerves in the brain, had been reported, the second of which resulted in the death of a 73-year-old who took Raptiva for four years, according to SEC documents.

The FDA slapped a warning on Raptiva in October 2008. The European Medicines Agency suspended the product's marketing authorization, as did Health Canada. By mid-2009, Raptiva was off the market. Lucentis, however, has proved to be a successful product, earning Genentech $1.5 billion a year. Xoma earned $5.1 million in 2009 and $8.8 million the year before.

That may have lent credence to Xoma for investors, who have pumped in $877 million in capital since its founding. That capital has continued to flow in recent years. According to SEC filings, Xoma raised $75 million in 2010.

Much of that investor interest may be related to the potential of Xoma 052, another monoclonal antibody focused on combating type 2 diabetes.

Xoma signed an agreement with Les Laboratoires Servier of Neuilly-sur-Seine, France, in December 2010 for worldwide distribution rights of Xoma 052, worth $20 million.

Yet another problem then befell the biotech. On March 22, Xoma released news that Xoma 052 did not meet its Phase 2b goal of lowering blood-glucose levels. The stock price tanked.

However, the treatment is being studied for other indications, such as Behcet's uveitis, an inflammatory eye disease, as well as cardiovascular disease. "Pending completion of the ongoing Phase 2a trial and analyses of both studies, we will be working with Xoma to determine the next steps in the Xoma 052 diabetes program," Isabelle Tupinon-Mathieu, head of therapeutic research and development for Servier, said in a statement.

Despite the hiccups, Xoma still has more than two dozen other licensing and development agreements with biotechs and pharmas, such as Pfizer Inc. and Novartis AG.

Pfizer is one of Xoma's oldest partners. The two entered into an R&D agreement in 1987 to produce a medication for sepsis. After the FDA rejected the product in the mid-'90s, requesting further clinical data, Pfizer eventually backed out of the agreement in 1997.

A 1998 filing with the Securities and Exchange Commission hints at Xoma's frustration: "FDA has substantial discretion in the product approval process and it is not possible to predict at what point, or whether, FDA will be satisfied with the Company's submissions or whether FDA will raise questions which may delay or preclude product approval."

Xoma didn't drop the sepsis drug. In 2006, the EMA approved Neuprex -- an injectable formulation of a protein the company developed in the mid-1990s -- as a potential orphan drug. The company last mentions Neuprex and sepsis in a 2008 SEC filing, when noting its interest in out-licensing the treatment.

In a 2009 filing, the company said it planned to focus its financial resources on Xoma 052. By then, the company was on its third CEO, Steve Engel, who took the job in 2007. Engel, an executive who had worked at La Jolla Pharmaceutical Co. of San Diego, Cygnus Therapeutic Systems and Micro Power Systems Inc., was unavailable for comment.

Despite various shortfalls, Xoma has been adept at securing investors and lining up partnerships to stay afloat. While net losses had been high for most of the 2000s, spiking to $45 million in 2008, 2009 proved to be Xoma's lone year of profitability, generating net income of $550,000. The company pulled in $98.4 million in revenue, including $29.1 million by selling rights to Lucentis to Genentech and a further $29 million for an expansion of an agreement with Osaka-based Takeda Pharmaceuticals Co. Ltd. for the discovery and development of monoclonal antibodies.

The company raised plenty of money in 2009, too. It pulled in $22 million by selling 22.2 million shares to institutional investors in May and June. In the third quarter of that year, it raised $26.4 million by selling 34.3 million shares to Azimuth Opportunity Ltd., an investment fund incorporated in the British Virgin Islands.

A source with knowledge of Xoma's history attributed the company's survival to its ability to attract new management and new executives when drugs fail, and to spark interest from investors by touting the company's deep product line, arrayed against various disease targets. "Every couple years there seems to be something to get excited about, and it never pans out," the source says. "They are masters of reinvention. Hope springs eternal."

Cytogen

For Cytogen, based in Princeton, N.J., the gamble was to escape large losses from the first 20 years of its life by essentially upping the bet.

Founded in 1981 on a premise similar to many other biotechs of that era -- more so on an idea than any tangible products -- Cytogen was initially led by Prince­ton venture capitalist Robert Johnston and Thomas McKearn, the scientist behind the company's antibody strategy. A rocky market and personnel problems kept Cytogen out of the public market after its founding until 1986. That year it raised $36 million and set out on its task of commercializing McKearn's idea of targeting antibodies for diagnostics and linking drugs to them for therapeutic purposes.

Once again the task proved more difficult than the theory. It was years before Cytogen's first marketable product, a diagnostic called OncoScint used to locate malignancies outside the liver for people with colorectal or ovarian cancer, generated revenue. And when it did in the mid-1990s, primarily through royalty revenue because the product was out-licensed, the sales were minimal, at less than $1 million.

Meanwhile, the company was spending millions of dollars annually, incurring net losses between $5 million and $30 million through the '90s, rising as high as $72 million in 1995. Part of the losses were due to being forced to reacquire OncoScint and Cytogen's other diagnostic, ProstaScint, after licensing partnerships failed multiple times throughout the '90s.

In 2001, Becker was hired to run Cytogen. Becker had a background as a biotech analyst and investment executive at Chicago-based Wayne Hummer Investments, and he was handed the task of moving the company from an antibody-based diagnostic producer to one focused on therapeutics, much like Immunomedics.

One difference between the two firms, which engaged in a legal battle in the early 2000s over patent issues that were eventually settled -- the Princeton biotech paid Immuno an undisclosed amount over allegations that Prosta­Scint infringed on a patent held by Immuno's Goldenberg -- is that Cytogen also became a company focused on commercializing products.

From its founding until his hiring, the company had burned through nearly half a billion dollars, Becker says, with only a few potentially tangible products: an imaging agent for prostate cancer and an antibody that could relieve the pain of bone cancer.

Becker targeted the bone cancer product, known as Quadramet, in his first move as CEO. The product had been licensed to another company, Berlex Laboratories, which was acquired by Leverkusen, Germany-based Bayer HealthCare Pharmaceuticals in 2006. Becker bought back the drug from Berlex for $8 million and created its first in-house commercialization and sales force to sell Quadramet and other products. Quadramet generated revenue, from $8.9 million in 1998 to more than $20 million in 2007. "We thought we could do a better job selling the product ourselves," Becker says. "The answer to whether you commercialize your products depends on whether you have the capital to do it."

Becker believed commercialization itself could be used to raise capital. During his six years at Cytogen, he raised some $130 million primarily by selling stock. Becker then used that money to build a sales force, in-license products developed by other companies and become "a full-fledged commercializa­tion organization."

It was a big gamble. "I'm going to do something completely different," Becker says he told potential investors. "We're going to take the company in a completely different direction. It's a great opportunity, but it comes with baggage of $500 million in [historical] net operating losses.

Cytogen's move into sales differentiated it from most other biotechs, which, fearful of the huge costs of a sales force, have increasingly focused on R&D while leaving development, clinical trials, or sales and marketing to others. Doing so was a conscious choice by Becker, who sought to convince investors that Cytogen was no longer a fledgling biotech spending millions of dollars on dreams, but instead a legitimate company with potential return on investment.

The strategy worked for a time, and Cytogen began in-licensing drugs and making deals, such as hormonal therapies and electrolyte solutions that were a far cry from its origins in biologics. By 2006, Cytogen in-licensed a product from Norway-based InPharma A/S with strong sales potential: Caphosol, a treatment for dry mouth that is often caused by chemotherapy.

But the economy began to crater just as Cytogen needed more capital to commercialize the drug. The company was running out of money. But investors under stress were more interested in companies with potential blockbuster drugs than in a firm like Cytogen, which was posting stable if unremarkable sales. "It's much easier to sell hope and dreams than it is to sell revenue streams," Becker says. "An investor is much more apt to write a check and help a company's valuation if there are dreams for the future."

Cytogen looked to sell.

In 2008, it inked a deal with privately held Oxford, U.K.-based Eusa Pharma to sell for $23 million, plus milestone payments on Caphosol, a drug Becker believed had blockbuster potential. Retrospectively, Becker says he would have liked a higher valuation, given that Eusa is now successfully selling Caphosol, Quadramet and ProstaScint. But Caphosol was still unproven, and the company was entering perilous waters.

So it was either sell or run out of money, says Becker, who now runs a business consultant firm, MD Becker Partners LLC out of Newtown, Pa. "It's very bittersweet," he says. "On the one hand, it validates that we were doing the right things. That we had identified a very attractive asset."

Building a successful biotech -- or, at least, a profitable one -- can be as dependent on chance, circumstance, serendipity and timing as it is on making good decisions on hiring, product development, R&D and commercial strategies. For a company like Genentech, success is clearly discernible, rising from the lab to the executive suite to shareholders. The lines are more blurred with the likes of Immunomedics or Xoma.

The same could be said for a biotech like Cytogen, which may have sold for too little or too early -- though who's to tell? While Becker would have liked to continue to invest in R&D, he believed doing so would be irresponsible to shareholders. Few disagreed at the time.

That's why when he talks to new biotechs today, he encourages them to pursue as long as possible the development of the dream of creating blockbusters. On the other hand, he adds, it's also important to be realistic. "Stay focused on hopes and dreams as long as you should," he says. " 'Should' being the operative word there."