Washington Editor

GAITHERSBuRG, Md. - Federal advisers Wednesday were unconvinced that the FDA should approve Genentech's Avastin (bevacizumab) in combination with the chemotherapy agent paclitaxel as a first-line treatment for metastatic breast cancer.

In a 5 to 4 vote, the FDA's Oncologic Drugs Advisory Committee said there was not sufficient data to support such an approval. Wall Street took immediate notice, with shares of Genentech (NYSE:DNA) dropping $6.49, or 9 percent, to close at $66.29.

The FDA is not required to accept the recommendations of its advisory panels, but typically does.

Richard Pazdur, director of the FDA's Office of Oncology Drug Products, told reporters that the agency would consider the panel's overall discussion rather than just its vote on the matter before making a determination.

Regulators are set to make a decision about Avastin in breast cancer by Feb. 23, 2008, the Prescription Drug User Fee Act action date.

Avastin, which is approved in the U.S. as a therapy for metastatic colorectal and non-small-cell lung cancers, already is sold in Europe to treat advanced breast cancer.

More than 200,000 patients worldwide have received Avastin to treat cancer, said David Schenkein, Genentech's senior vice president of clinical hematology and oncology.

While there are many therapies available on the U.S. market to treat metastatic breast cancer, said Patricia Cortazar, medical officer in the FDA's Office of Oncology Drug Products, only two drugs have been approved under modern regulatory standards for the first-line treatment of the disease: Herceptin (trastuzumab), from Genentech Inc., approved in 1998, and Gemzar (gemcitabine), from Eli Lilly & Co., approved in 2004. Both drugs are approved for use in combination with paclitaxel, she noted.

South San Francisco-based Genentech is waging its bets for U.S. approval of the breast cancer indication on a Phase III study of 722 patients, known as E2100, that found that Avastin plus paclitaxel slowed the progression of tumors by a median of 11.3 months, which was 5.5 months longer than patients who had received paclitaxel alone.

However, there was no statistically significant improvement in overall survival in the combination arm, said Hong Lu, statistical reviewer in the FDA's Office of Biostatistics, and more than 70 percent of patients in that group experienced severe adverse events, including hypertension, proteinuria, arterial and venous thrombosis, congestive heart failure, bowel perforation and death, compared with 51 percent in the paclitaxel arm.

Genentech initially reported that none of the 363 deaths in the combination arm were related to Avastin.

But, said FDA medical officer Lee Pai-Scherf, drug regulators disputed Genentech's analysis and determined that at least six deaths were definitely or probably caused by toxicity related to Avastin.

All but one of those patients died within 30 days of the end of the study, she said. The sixth patient died of a heart attack seven weeks after the study. However, Pai-Scherf said, regulators determined that the heart attack was brought on by grade IV proteinuria related to Avastin use.

Genentech officials said Wednesday they agreed with the FDA's analysis of the six deaths.

The company argued that the improvement in progression-free survival (PFS) - the time from randomization to disease progression or death within 84 days of the last study treatment - in study E2100 outweighed the drug's toxic risks.

Throughout the meeting, panelists struggled with the issue of whether PFS alone without a demonstrated survival advantage should be considered a measure of direct clinical benefit in the initial treatment of metastatic breast cancer.

Data from studies using PFS are available sooner, whereas trials that use overall survival as the endpoint may take years to gather data.

The FDA's Cortazar noted that some drugs on the market to treat second- and third-line metastatic breast cancer, such as Bristol-Myers Squibb Co.'s Ixempra (ixabepilone), were approved based on trials with PFS as the primary endpoint.

However, she said, regulators heavily weighed the overall survival of patients in those studies and the drugs' safety as part of the approval.

If the FDA relied only on overall survival as the hurdle to jump in studies about metastatic cancers, Susan Hellmann, Genentech's president of product development, told reporters, it would take longer for new therapies to reach the market.

Panelist Gary Lyman, director of health services and outcomes research in oncology at Duke University Medical Center, argued that most oncologists and patients with metastatic breast cancer, an incurable disease, would accept PFS as an endpoint.

But Natalie Portis, the patient advocate on the committee, urged that the drug should not be approved simply for the reason of having another option on the market.

"It's a very painful reality that metastatic breast cancer is not curable," she said. But, Portis added, the toxic effects of Avastin are "too high of a price to pay."

Panel Chairwoman Maha Hussain, professor of medicine and urology at the University of Michigan in Ann Arbor, said she was concerned about the deficiencies in study E2100, noting that the trial was not initially designed to be used as the basis of a drug approval.

"There were too many uncertainties in the way the data were collected," she said. "A vote of 'yes' lowers the bar. I don't think we can sanction suboptimal conduct for trials."

Speaking during the public hearing portion of the advisory committee meeting, Robert Erwin, president of the Marti Nelson Cancer Foundation, said that Genentech's application should not be approved "if approval requires lowering the bar for approval."

The personal and clinical meaningfulness of PFS, he said, is going to vary among individual patients.

"It should not be taken lightly," Erwin said, when a patient hoping to obtain a significant improvement in PFS dies early because of an adverse event related to the treatment. However, he added, patients who "receive substantial benefit from the treatment should not be taken lightly either."

The FDA, said panelist Aman Buzdar, professor of medicine at the University of Texas, must decide whether approving Avastin for breast cancer is "enhancing the patient's choices" or "putting something of not substantial value" on the market.

The committee's mixed vote, Hellmann said, gives Genentech the opportunity to continue its discussions with the FDA about what is needed to show that Avastin is safe and effective.

The firm, Schenkein told reporters, has two ongoing Phase III studies of Avastin in metastatic breast cancer with data expected in 2008.

Both studies include about 700 patients and have PFS as the primary endpoint.

The FDA, Pazdur said, will continue to seek survival data from Genentech and other firms that use PFS as the primary endpoints in studies.