A Star Drug Is Born

In June, 1988, Steven Shak was bleary-eyed from searching through old scientific texts. The Genentech Inc. researcher was convinced he could find an enzyme to attack the thick mucus that plagues cystic fibrosis sufferers, the residue of infections that eventually destroy their lungs. Shak imagined spraying such an enzyme into the air passages of patients to help them cough up the sputum and breathe more easily. His quest led him to a cow enzyme called deoxyribonuclease. But when the first samples of the stuff arrived, Shak's heart sank. "Do not shake," said the warning label. If the enzyme was that fragile, it would never work as a spray.

The young scientist persisted, however. He soon discovered that the enzyme was more robust than he thought. And on Aug. 9, his work paid off: A Food & Drug Administration advisory committee recommended a human cousin of the enzyme, a drug called DNase, as a treatment for the nation's 20,000 CF victims. That victory after only a single minor snag concluded one of the most remarkable efforts in the recent history of drug development. DNase is just a treatment, not a cure. Still, coming as it does after several stagnant years for Genentech, it will go a long way toward restoring the company's reputation as biotechnology's pacesetter. DNase is "a true clinical breakthrough," says Harri V. Taranto, managing director of the Wilkerson Group Inc., a medical products consulting firm. Since DNase promises to cut treatment costs, he adds, "it's the prototype of what the industry needs to put out for the rest of the 1990s."

It's a prototype, too, for how creativity and determination can defy the odds of drug development. Pharmaceutical executives, in an effort to defend price increases, often lament that it takes some $359 million and 12 years on average to develop a drug. By comparison, DNase looks like a bargain. While the company won't reveal its investment, knowledgeable sources put the development tab so far at a relatively modest $100 million. In part, that's because Genentech took unusual risks to accelerate R&D--and delivered DNase in just five years.

SHOW TIME. Just as notable, much of this was accomplished by a team of young scientists, some of whom, like Shak, had no experience in drug development. A pulmonary physician by training, he left academe for Genentech in 1986, driven by the desire to find treatments using basic science. He came along just as the company was launching t-PA, its best-known drug. That gene-spliced enzyme, which breaks apart the blood clots that cause heart attacks, gave Shak the idea that an enzyme might do the same for CF.

Once Shak proved in mid-1988 that his enzyme could survive in a spray, his bosses turned him loose to gene-splice a human variety. In theory, it would chew up excess DNA--the main ingredient of sputum lining the lungs of CF patients. After four months of 100-hour weeks in the lab, Shak had a beaker's worth of his potion and was ready for the test that shook Genentech. He took mucus from a CF patient, cut it in half with a razor, and divided it between two test tubes. He poured saline solution in one, DNase in the other, then turned them upside down. Almost instantly, the sputum in the DNase tube turned to liquid and ran down the side; the gunk in the other stayed put. Shak showed everyone his "thing in a test tube," as Genentechers dubbed it. The results were so dramatic "it didn't take a lot to get the go-ahead to drop everything else I was doing," he recalls.

Arthur D. Levinson, senior vice-president for research and development, immediately put the chance that DNase would become a drug at 70%, an unheard-of level at such an early stage even for Genentech's can-do gene jocks. By the fall of 1988, Shak had formed a 15-person team with staff members from marketing, manufacturing, and R&D. The ensuing progress was so speedy that "I'm not sure we could have saved a week in the whole process," Levinson says, looking back.

Shak's team shifted into overdrive, however, just as Genentech stumbled. By mid-1989, then-President and now Chief Executive G. Kirk Raab began to worry that funding for projects such as DNase was in jeopardy. After a decade of barrier-breaking successes, the company had wildly overestimated the market for t-PA at $1 billion annually, when it was looking more like a $200 million drug. That meant Genentech needed other sources of funds to finance its R&D. In late 1989, when Raab asked Shak to give a presentation on DNase to a visitor from the Swiss drug giant Roche Holding Ltd., Shak had no idea how important his talk would be.

In early 1990, Genentech agreed to sell 60% of itself to Roche for $2.1 billion. Today, that visitor, Jergen Drewes, president of international R&D for Roche, says Shak's enthusiasm for DNase helped bring about the deal. "If they can [be this creative]," Drewes recalls thinking, "they are going to make a difference."

It wasn't just Shak's bright idea that Drewes liked--Genentech made and tested DNase in a winning way. Developing drugs is typically a serial process, since biological experimentation is prone to surprises. Management's confidence in DNase, however, prompted Genentech to perform some key tasks in parallel--much like concurrent engineering in the auto and electronic industries.

At Genentech, a decision to leapfrog conventional lab procedure made concurrent development possible. Biotech drugs are made using genetically engineered cells to churn out large amounts of a desired protein--such as DNase. The "cell lines" that do this can be made in just a few weeks. But lines created that quickly often aren't very productive: They make only enough material for a small trial, with higher-yielding cell lines added later. On DNase, Levinson skipped the first step. He had his lab go straight to a high-yield line, a nine-month effort.

That decision produced large amounts of DNase for testing earlier than Genentech would have gotten otherwise. And it involved the company's manufacturing engineers in the final process earlier than usual. That earned a big payoff. If the process for making a biotech drug is changed after the drug is tested in people, the FDA requires special studies to prove that the drug doesn't change with the method. By developing the high-yield line early, says Craig W. Rice, associate director of cell culture production and manufacturing, his team kept production consistent into commercial runs and avoided extra studies. That, he thinks, shaved as many as four years off the approval process.

In another calculated bet, Genentech took another shortcut. It began building its $40 million DNase production facility in February, 1992--well before finishing tests of the drug on humans. That put the pressure on Henry J. Fuchs, the company's associate director of clinical research. A month earlier, Fuchs had started a 968-patient trial, the most extensive ever on CF sufferers. The 35-year-old physician, who had come to Genentech in 1988 from academe, tapped into the network of CF care-givers, parents, patients, and researchers, and ran trials in 51 of the Cystic Fibrosis Foundation's 114 centers. Patients reported quick relief from the drug. "I noticed the difference immediately," says Don West, 24, of Garrettville, Ohio. "It increased my energy, and my lungs stayed clearer."

HELPFUL FEDS. Once DNase gets final FDA approval, which is expected shortly, doctors will likely prescribe it as a daily medication, inhaled through an atomizer. Although not a cure, the drug can cut lung infections and the use of antibiotics by about 30%, the trials showed. So far, the primary side effect of the drug seems to be hoarseness. The tests also found that DNase can boost lung function 5% to 8% and cut average hospital stays by 1.4 days, or 18%, over a six-month period. Genentech can thank the FDA for some of that data: Initially, the company intended to measure just lung function. But regulators, who seemed predisposed to facilitate approval, suggested measuring the other clinical signs.

Those statistics should help the company defend the drug's price, which analysts expect to be $10,000 to $15,000 per year. That sounds hefty, but some doctors say CF patients spend up to $50,000 a year on antibiotics and hospitalization. DNase, they add, could eliminate some of those costs. There's one more plus. While doctors expect most patients to continue the daily back and chest pounding called "percussion therapy" that loosens the mucus so it can be expelled, some--including West--can clear their lungs using just DNase. That's an improvement in quality of life that's hard to quantify.

Analysts estimate that DNase is likely to generate some $300 million in annual revenues by 1996. Eventually it could produce up to $500 million in yearly revenues worldwide if it is used on other pulmonary maladies--it's now being tested on chronic bronchitis. Roche will help by co-marketing the drug in Europe. Strong sales of DNase-combined with a brightening outlook at last for t-PA and a handful of promising drugs in the pipeline for fighting cancer and inflammatory diseases--could end five years of erratic profits at Genentech (chart). In just the past few months, in fact, the company's stock has risen to 44, a five-year high. That can only help reinforce Drewes' pledge to "protect Genentech from a stifling embrace."

Other CF drugs still in testing could one day dim Genentech's rosier outlook. And an exciting new approach called gene therapy has the potential to cure CF, making DNase obsolete. That could be even a bigger market, since a cure for CF would command a premium over drugs that just treat some of its symptoms. Genentech has tried to buy insurance for that day by investing in GenVec, a gene therapy startup (box).

It's true that not every drug can sail though development as smoothly as DNase, whose potential was extraordinarily clear from the start. Many treatments for complex conditions such as cancer are not nearly so straightforward, so a more cautious approach is warranted. Still, Levinson thinks DNase's success will have a residual benefit for Genentech--that it will inspire the company's researchers to aim higher, even if not every mission succeeds. Shak, for one, is glad he enlisted: "If I'd had this idea in academia," he smiles, "nothing would have happened."

Before it's here, it's on the Bloomberg Terminal.