December 6, 2007
Advisory Panel Rejects New Use for Cancer Drug


By ANDREW POLLACK
A federal advisory committee voted yesterday that Genentech’s drug Avastin should not be approved as a treatment for breast cancer.

By a 5-4 vote, the committee decided that Avastin’s ability to delay the worsening of cancer did not outweigh the drug’s toxic side effects, especially since women getting Avastin did not live significantly longer in the end.

“Absolutely it’s a very painful reality that metastatic breast cancer is not curable,” said Natalie Compagni Portis, a patient representative on the committee, who voted against an approval. But, she added, “I don’t think that means that we should just say, ‘Well, here, try this,’ if there isn’t meaningful data to support it.”

Avastin is already on the market as a treatment for colon and lung cancer. The ultimate decision on whether to add an approval for breast cancer will be made by the Food and Drug Administration. But the agency usually adheres to the advice of its advisory committees.

The vote was a rare setback for Genentech, widely considered the world’s most successful biotechnology company. The company’s stock plunged by 8 percent Wednesday, closing at $66.64.

Although the F.D.A.’s staff had been critical of the drug in an analysis released on Monday, many Wall Street analysts had nevertheless thought the committee, made up mainly of cancer doctors, would vote in favor of approval.

Analysts were expecting an approval in breast cancer to add $1 billion or more to annual sales of the drug. Genentech’s stock has been steadily declining for the last two years because its once meteoric growth appears to be slowing.

Avastin is already one of the world’s best-selling cancer drugs, with United States sales alone of $1.7 billion in the first nine months of this year.

Roche, which owns a majority of Genentech, sells the drug overseas. In Europe Avastin was approved for breast cancer based on the same clinical trial that failed to pass muster with the F.D.A. committee.

The debate by the committee, which met in Gaithersburg, Md., reflected difficult questions that often surround cancer drugs, over how much benefit a drug needs to show before it can be approved.

The drug was being considered for approval as an initial treatment for women whose breast cancer had recurred or spread to other parts of the body. In the main clinical trial, women who got Avastin plus the generic cancer drug paclitaxel went a median of 11.3 months before their tumors worsened, compared with 5.8 months for women who received only paclitaxel.

But the women who received Avastin did not live meaningfully longer. And they suffered more severe side effects, including five or six deaths out of 363 patients from the drug itself.

A big question for the committee was whether the delay in tumor worsening — a measure called progression-free survival — was really meaningful to patients, or whether it would be necessary to show that a drug lengthens life.

Some oncologists on the panel argued that a delay of 5.5 months in progression would be meaningful because it might delay the onset of symptoms and relieve patients of anxiety.

They also said the progression-free survival was a more practical goal for a breast cancer trial. The effect of a drug on survival can be difficult to measure because patients take other drugs after the trial is over.

“Many of my patients after their first-line approach go through five, six, seven, additional regimens,” said Dr. Gary H. Lyman, an oncologist at Duke University Medical Center who voted in favor of approval. “One has to wonder if the survival differences aren’t being clouded.”

But others argued that progression-free survival is hard to measure. The radiologists who looked at X-rays to judge whether the cancer had worsened disagreed among themselves in about one-third of cases.

“You’ve not shown these patients are living better, and certainly they are not living longer,” said Dr. Maha Hussain, the committee’s chairwoman.

Dr. Hussain, a prostate cancer specialist at the University of Michigan, agreed that “patients are a nervous wreck when their disease is progressing, and that you have no tool to measure.” But she also said that women with breast cancer have many other drugs to choose from.

She voted against Avastin.

Dr. Susan Desmond-Hellmann, Genentech’s president for product development, said the company might ask the F.D.A. for conditional approval on the promise of later providing survival data from two other clinical trials that are now under way.