SILVER SPRING, Md. - A panel of cancer experts has ruled for a second time that Avastin, the best-selling cancer drug in the world, should no longer be used in breast cancer patients, clearing the way for the government to remove its endorsement from the drug.

The unprecedented vote Wednesday by the Food and Drug Administration advisory panel comes less than a year after the same panel reached the same conclusion.

The six members of the FDA oncology drug panel voted unanimously that Avastin is ineffective, unsafe and should have its approval for breast cancer withdrawn.

"I think we all wanted Avastin to succeed but the reality is that these studies did not bear out that hope," said Natalie Compagni-Portis, the lone patient representative on the panel.

The vote is not binding and FDA Commissioner Margaret Hamburg will make the final decision sometime after July 28. The drug is approved for multiple cancers and will still be available for breast cancer, though insurers are expected to drop coverage if it loses FDA approval.

The FDA began steps to remove Avastin's breast cancer approval in December, but Roche took the rare step of appealing that decision and lobbied the agency and Congress for a second hearing.

The dramatic, contentious tone of the two-day hearing underscored the difficulty of removing an option for cancer patients, even when backed by scientific evidence.

Immediately after the final vote, patients in the audience erupted in shouts against the FDA and its experts.

"What do you want us to take!? We have nothing else!" shouted Christi Turnage, of Madison, Miss. Turnage said her cancer has been undetectable for more than two years since starting therapy with Avastin.

A spokesman for the Abigail Alliance, which advocates for access to experimental medicine, said the vote should be overruled.

"This was a kangaroo court," said Steven Walker, the group's co-founder. "There wasn't one dissenting thought up there, let alone one dissenting vote."

Assuming the FDA follows through on the withdrawal, drugmaker Roche could lose up to $1 billion in revenue for its best-selling product, which generates over $6 billion per year. Avastin is FDA-approved for various types of colon, lung, kidney and brain cancer, which are not part of the debate. Doctors will still be allowed to prescribe Avastin for breast cancer, though insurers may not pay for it. When administration fees are included, a year's treatment of Avastin can cost $100,000.

Roche's Genentech unit argued the drug should remain available while it conducts more research on which patients benefit most from the injectable drug. The drug is approved for breast cancer that has spread, or metastasized, to other parts of the body. Such cancer is generally considered incurable.

"The data tell us it is better for women diagnosed with metastatic breast cancer to have Avastin as an approved treatment option," said Hal Barron, Roche executive vice president.

Wednesday's vote came after two days of hearings that often resembled a courtroom trial, complete with testimony, cross-examination and a final jury verdict. In a public comment period Tuesday, Avastin patients and their families took the role of witnesses against the FDA.

"Make no mistake, this hearing is a death trial, not of Avastin but of these women who rely on Avastin to say alive," said Terry Kelley, whose wife takes Avastin for breast cancer. "You are each personally responsible for the consequences of your own vote."

Kelley formed a group called Freedom of Access to Medicines to protest and lobby the FDA. Kelley says the group does not receive funding from Roche.

Panelists said Avastin's ability to slow tumor growth, measured through medical imaging scans, has not translated into meaningful benefit for breast cancer patients.

"I think as treating clinicians we have to ask ourselves: What are we doing in terms of helping patients? Simply delaying a change in a CT scan for a month or two is not significant unless it's accompanied by other improvements in how the patients are doing or overall survival improvement," said panelist Dr. Wyndam Wilson of the National Cancer Institute.

The FDA granted Avastin accelerated approval in 2008 based on one study in which it slowed growth of breast cancer tumors for more than five months when combined with chemotherapy.

But that delay shrunk to less than three months in follow-up studies when the drug was paired with other types of chemotherapy. Across all studies, patients taking Avastin did not live any longer and suffered side effects like infection, high blood pressure and blood clots.

Most cancer experts say the drug should remain available for patients who are already responding well, even if its approval is withdrawn.

"I think the FDA is doing the right thing since the drug has some serious complications," said Dr. Stephanie Bernik of Lenox Hill Hospital in New York. "However, there are definitely patients who are benefiting from the drug and if the FDA completely withdraws approval those patients may find it hard to get access."

One potential option to keep the drug available would be for Roche to pay for it when patients have no other option. The company already provides the drug for free to patients who meet certain financial criteria or don't have health coverage.

Roche has suggested that Avastin works differently depending on which chemotherapy drug it is paired with. The drugmaker essentially asked the FDA for time to repeat its initial study that had the strongest results, theorizing that Avastin works best with the chemotherapy paclitaxel. Such a study would likely not be completed before 2016.

But the FDA rejected that argument, saying there is no evidence Avastin interacts differently with various chemotherapies, and that continuing approval cannot be justified based on one study completed six years ago.

"No trial has shown that patients treated with Avastin lived longer than those not treated with Avastin," said FDA's director for new drugs, Dr. John Jenkins. "All clinical trials show an increase in serious adverse events."

The Avastin review will have broad repercussions for patients and the pharmaceutical industry.

Since the early 1990s the FDA has granted accelerated approval to dozens of drugs based on promising early results, on the condition that their effectiveness is confirmed in later studies. That policy has been praised by patients with HIV, cancer and other deadly diseases where access to experimental treatments can mean life or death.

But the flip side of the program means removing drugs from the market if their initial promise isn't confirmed by later studies. And until last year the agency had never removed a drug from the market because of incomplete or unconvincing follow-up data.

With the removal of that leukemia drug from Pfizer, and now the proceedings over Avastin, analysts say the FDA is poised to crack down on drugs whose effectiveness hasn't been confirmed in later studies.