FDA “Regulatory Flexibility” On Accelerated Approval Must Result In Some Withdrawals

If all drugs cleared for accelerated approval succeed in their confirmatory studies, then FDA is being overly conservative in its use of the expedited approval pathway, FDA Office of Hematology and Oncology Products Director Richard Pazdur says at the AACR annual meeting.

If FDA is exercising the appropriate regulatory flexibility in approving drugs under accelerated approval, some of those products inevitably will be withdrawn from the market for failure to confirm efficacy, the agency’s top oncology reviewer said at the American Association for Cancer Research annual meeting in Washington, D.C. April 7.

Office of Hematology and Oncology Products Director Richard Pazdur said that if none of the drugs receiving accelerated approval is ever withdrawn, then the agency is taking an overly conservative approach in its approval of products under the pathway. “If you are demonstrating the correct degree of regulatory flexibility, there are going to be successes and there are going to be failures,” Pazdur said. “If there are no failures, you … are being over-regulatory and over-conservative.”

The Avastin experience showed the difficulties of explaining to the general public why an indication was being withdrawn for efficacy, rather than safety, reasons.

Pazdur made his remarks while moderating a panel discussion among former chairpersons of the agency’s Oncologic Drugs Advisory Committee. A portion of the two-hour discussion focused on the agency’s landmark battle with Genentech Inc. to withdraw accelerated approval of the metastatic breast cancer claim for Avastin (bevacizumab) and the lasting impact of that regulatory action on the accelerated approval pathway.

The Avastin experience brought to bear a new difficulty for the agency – explaining to the public why an indication should be revoked for efficacy reasons, as opposed to safety concerns, Pazdur said. He also cited efforts under way in the agency to better reflect in labeling the conditional status of accelerated approval drugs.

Revisiting Avastin

The AACR panel discussion brought together six current and former leaders of ODAC to talk about their experiences serving on the advisory committee during the past 10 years.

Discussion inevitably turned to what could be considered the most significant development involving FDA’s accelerated approval process in the past decade – the agency’s decision to revoke Avastin’s approval for first-line use, in combination with paclitaxel, for metastatic breast cancer.

Genentech gained accelerated approval for the breast cancer claim in February 2008 on the strength of a 5.5-month median progression-free survival benefit in the E2100 trial, although there was no improvement in overall survival. The decision to grant accelerated approval followed a December 2007 review of the application by ODAC, which narrowly voted against approval, 5-4, due in part to concerns about the design and conduct of the E2100 study (Also see "ODAC Split On Avastin Use In First-Line Metastatic Breast Cancer" - Pink Sheet, 10 Dec, 2007.).

In granting accelerated approval, Pazdur overrode the recommendation of Patricia Keegan, who was director of the Division of Biologic Oncology Products. Keegan believed that Genentech had failed to adequately characterize the magnitude of treatment effect in first-line metastatic breast cancer and recommended a “complete response” action pending results from two confirmatory trials, AVADO and RIBBON1 (Also see "Avastin’s Breast Cancer Approval Followed CDER Rift On Efficacy" - Pink Sheet, 30 May, 2011.).

FDA brought Avastin back to ODAC in July 2010 with the results of the AVADO and RIBBON1 studies, which demonstrated a median PFS improvement ranging from approximately one to three months. The committee concluded bevacizumab did not delay disease progression long enough to be clinically meaningful and voted 12-1 in favor of removing the breast cancer indication from the label (Also see "Avastin Breast Cancer Approval Debate Pits Clinical Benefit Versus Statistics" - Pink Sheet, 26 Jul, 2010.).

In December 2010, the Center for Drug Evaluation and Research announced it was proposing to withdraw approval of the breast cancer claim because AVADO and RIBBON1 failed to confirm the magnitude of PFS benefit seen in the E2100 study. (Also see "Genentech May Not Get A Chance To Appeal FDA's Decision To Withdraw Avastin's Breast Cancer Approval" - Pink Sheet, 20 Dec, 2010.).

Genentech disputed FDA’s conclusions, refused to withdraw the indication and requested a hearing under the accelerated approval regulations. The first-of-its-kind public hearing was held over two days in June 2011, with ODAC playing the figurative role of “jury” and ultimately voting 6-0 in favor of the claim’s removal (Also see "Avastin’s Breast Cancer Claim: Will FDA’s Hamburg Take A Middle Road?" - Pink Sheet, 4 Jul, 2011.). In November 2011, Commissioner Margaret Hamburg issued her decision ordering the claim withdrawn (Also see "Avastin Loses Its Breast Cancer Claim; FDA’s Hamburg Opts For Withdrawal Over Restrictions" - Pink Sheet, 21 Nov, 2011.).

Neither Genentech nor CDER came out of the June 2011 hearing looking particularly good, and FDA officials have repeatedly pointed to the resource-draining experience as a reason why creation of new expedited approval mechanisms, or reforms to the existing accelerated approval process, should include pathways for easier withdrawal .

Pazdur reiterated this point in his comments at AACR. “If you take a look at the resources that went into that public hearing for Avastin, it would be impossible for the FDA to do that on a repeated basis,” he said. “I would just guesstimate that millions of dollars in resources, FTEs etc., lawyers, trying to do that hearing in a coherent fashion were spent, and probably even a greater amount by the company.”

Balancing Successes And Failures

Resource issues aside, one former ODAC chair questioned why FDA let the breast cancer claim get onto the label in the first place.

“What were you thinking when the FDA approved Avastin?” Maha Hussain, University of Michigan, asked Pazdur. Hussain chaired the 2007 ODAC at which the claim was first considered, and she was one of the five panelists who recommended against approval. “We obviously were right. The FDA should have listened to us,” she said.

“You have to make a decision at a particular time based on the evidence at hand,” Pazdur responded, offering a broader view of how the accelerated approval provisions should be interpreted and applied by the agency.

“I want to put forth a general principle here. If … the agency is using accelerated approval appropriately, there will be drugs that will have to come off the market,” he said. Accelerated approval may be granted based on a surrogate endpoint that is reasonably likely to predict clinical benefit, he noted. “That doesn’t mean that it’s a definitive demonstration.”

“You have to have a balance of drugs that make it and drugs that don’t if you are really going to be demonstrating the correct degree of regulatory flexibility. If all the drugs make it basically, why call it accelerated approval? They should be just full approvals.”

Wyndham Wilson, National Cancer Institute, was chairman at ODAC’s July 2010 Avastin meeting and was a member of the panel for the June 2011 hearing. He suggested Genentech failed to live up to its end of the accelerated approval bargain when its confirmatory trials failed to satisfy FDA.

Had the Avastin decision gone the other way, “it would have been a very dark day for accelerated approval and a very dark day for patients getting access to promising drugs early,” former ODAC Chairman Wilson says.

“If you’re going to approve drugs on limited data that are likely to show clinical benefit, you need to have confirmatory trials,” Wyndham said. “If in fact you don’t confirm, then … the drug company needs to graciously withdraw the drug. And that didn’t happen here.”

Nevertheless, the Avastin dispute and the ultimate outcome “were a good demonstration … of the system working,” Wilson said. “If the system hadn’t worked, I think it would have been a very dark day for accelerated approval and a very dark day for patients getting access to promising drugs early, because it really would have, I think, made the ODAC and made FDA raise the bar even higher.”

Echoing Pazdur’s remarks, Wilson continued: “The bar has got to be low enough that some of your drugs will not confirm, otherwise you’re not doing what the intent of accelerated approval is, which is to bring a drug that has promising activity forward so people have access to it before the final confirmatory trials are done.”

Rethinking Language And Labeling

In addition to battling with Genentech, FDA faced a fundamental public relations and educational challenge in explaining to the public why it believed the breast cancer claim should come off the Avastin label.

“When FDA usually takes a drug off the market, it’s usually for a safety reason, and people and John Q. Public can understand that,” Pazdur said. “It’s somewhat difficult to say, ‘Well, the drug worked in 2008, but it doesn’t work in 2009.’ Then you get the patients saying, ‘Gee I bought this drug or I mortgaged my house for this drug, and now the FDA is not saying it works.’ So there’s a fundamental issue here of taking a drug off for lack of efficacy versus our common scenario of taking a drug off the market for toxicity, and I think that’s something that we as a society have to do a better job in explaining the limitations of accelerated approval.”

Pazdur suggested there is room for improvement in both the regulatory language and product labeling.

The phrase accelerated approval is a “terrible term,” Pazdur said, adding that he prefers the term “conditional approval,” which is used in Europe.

FDA is working on labeling guidance that will emphasize the conditional nature of accelerated approval.

When a drug is cleared on the basis of accelerated approval, this conditional status is reflected in labeling with a statement that clinical benefit has not been demonstrated, Pazdur said. “What does that really mean to anybody? It really doesn’t give them the information that that drug was approved really under a condition that further studies be completed in a positive fashion. So we are trying to address this in new labeling and in a new guidance that we’re working on. But I think it’s very important that we really emphasize the conditional nature of this rather than the accelerated nature of the approval.”

Hussain said accelerated approval decisions should take into account multiple considerations that go beyond just the efficacy and safety data for a given drug.

“I would argue that somehow, somewhere the decision cannot be a one size fits all on the accelerated approval, that the condition of the disease, what’s available on the market, the strength of the evidence and other issues to be taken into consideration,” Hussain said, adding that the agency also should consider asking the sponsor to have an open access program while conducting confirmatory trials rather than granting them accelerated approval to market the drug.

Pazdur said such factors were actively debated within the agency in the case of Avastin. He also suggested that had the therapeutic area been one which lacked other treatments, the regulatory action would have been different. “In an area where you had no other therapies, I think there would be a much different decision there,” he said.