Online Extra: Drug Development, Genentech Style

Its promising colon cancer fighter Avastin is a prime example of how the biotech keeps following the science

When Earl Woodard was diagnosed with advanced colon cancer in November, 2000, he didn't have much hope. Surgeons were able to remove some of his tumor, but the cancer had spread to his lymph nodes, making it unlikely he would survive more than a few months. Then Woodard went to Duke University and enrolled in a trial of an experimental cancer drug, Avastin, which was developed by Genentech (DNA ) Today, he's cancer-free. "It's a miracle," says Woodard, a 54-year-old pilot from Carthage, N.C. "I'm really happy Genentech stuck with this drug."

Some people didn't believe the South San Francisco (Calif.) biotech should stick with Avastin. After the drug failed a breast-cancer trial in 2002, Genentech CEO Arthur Levinson was pelted with criticism from Wall Street analysts, who felt he was spending too much on a drug that had never shown much promise. But Levinson was encouraged by Woodard and other patients in the colon cancer trials who seemed to be responding to the drug.

His patience paid off earlier this year, when trial data showed that the drug, when taken with chemotherapy, extended the lives of colon cancer patients by an average of five months. Avastin, which will be submitted for Food & Drug Administration approval this month, has been granted fast-track status, meaning Genentech will get an answer within six months. High expectations for the drug have pushed Genentech's shares up 164% this year, to $87.

NEW APPROACH. Avastin was born from a highly coveted Genentech tradition of allowing scientists to pursue their personal research passions. Levinson has always believed the greatest discoveries don't necessarily happen by design. In the case of Avastin, he was right. It was the pet project of Napolean Ferrara, a gynecologist who was originally hired to develop a drug to induce labor.

"It didn't work very well," Ferrara admits. So he started spending more time investigating the idea that blocking the formation of blood vessels that feed tumors -- a process called anti-angiogenesis -- might slow or stop some cancers.

Although the notion wasn't new, Ferrara's approach to it was. Researchers had tried for more than 30 years to make anti-angiogenesis drugs, only to fail time and again. Ferrara set out to find a protein that spurred blood-vessel growth. Then he developed an antibody to block it. He discovered the protein, which he called vascular endothelial growth factor (VEGF), identified the gene that encoded it, and designed an antibody that would shut down the whole process.

"A BREAKTHROUGH." Genentech's scientists don't know why Avastin didn't work in breast cancer. But Levinson, a renowned cancer researcher in his own right, knew that because all cancers are different, he couldn't give up on the drug. "We know a lot about the molecular basis of cancer now," he says. "Cancer is a heterogeneous disease. Just because it didn't work in one cancer doesn't mean it won't work in others."

In fact, Genentech has plans for Avastin beyond colon cancer. Early-stage trials in kidney cancer have been promising, and it's also testing the drug in lung cancer.

Oncologists have high hopes for Avastin. "There aren't many cancer trials that have showed an improvement in survival over traditional chemotherapy," says Mark Kris, chief of thoracic oncology at Memorial Sloane Kettering in New York. "This is a breakthrough."

PREPARING NOW. Still, Genentech faces some high hurdles. If Avastin surpasses peak sales estimates of $1 billion a year, Genentech could confront a manufacturing crunch. Building enough production capacity is notoriously tricky in this industry because biotech plants take five years to assemble. Projecting demand that far out -- especially for a drug that suffered so many setbacks -- has been virtually impossible.

So Levinson erred on the side of caution. Genentech has two plants, and it's outfitting a third that it bought from GlaxoSmithKline (GSK ). "It's worse not to be able to supply the market than it is to overbuild," Levinson says.

And if Avastin is a success, Levinson will be faced with what he calls a "wonderful problem": deciding how best to spend the profits. Genentech pours about 20% of its revenues back into research and development each year -- aggressive by pharmaceutical-industry standards. But Levinson says he plans to continue focusing his resources on building the drug pipeline, to ensure that Genentech will have life after Avastin.

"We've got to keep following the science here," Levinson says. Only then will Genentech be able to live up to Wall Street's lofty expectations.

By Arlene Weintraub in Los Angeles

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