Sponsors appearing before FDA advisory committees should be prepared to take a public vow promising to withdraw an accelerated approval drug or indication if a subsequent confirmatory trial fails to verify the expected magnitude of benefit.

AstraZeneca PLC was the latest sponsor to face scrutiny from the Oncologic Drugs Advisory Committee over future withdrawal plans for a drug under consideration for accelerated approval with its ovarian cancer candidate olaparib.

The company is seeking accelerated approval of the first-in-class PARP inhibitor for maintenance treatment of ovarian cancer based on Phase II subgroup efficacy data. At their June 25 meeting, advisory committee members said consideration for marketing approval should await results from an ongoing, Phase III confirmatory trial, SOLO-2 ([A#00140630012]).

Committee members pressed AstraZeneca representatives several times about what the sponsor would do if olaparib were to receive accelerated approval but the SOLO-2 results subsequently fell short of the 7.1-month median improvement in progression-free survival seen in a Phase II trial subgroup.

Notably, AstraZeneca faced repeated questions about this issue in the public forum even thought it had already stated its willingness to withdraw the drug at a future time.

“AstraZeneca commits to work with the agency to review results from across the development program, including SOLO-2, and if these are not confirmatory in nature, will consequently abide by the results and take the necessary actions, which could include withdrawal of the NDA,” the company said in its briefing document for the meeting.

All Thanks To Avastin

The oncology advisory committee’s members sat as the proverbial jury during the June 2011 public hearing on Genentech Inc.’s challenge to the Center for Drug Evaluation and Research’s proposal to withdraw accelerated approval for Avastin (bevacizumab) in metastatic breast cancer. The committee unanimously voted to remove the claim – a recommendation adopted by FDA Commissioner Margaret Hamburg five months later.

Three years later, the memory of that bitter battle between FDA and a sponsor – and the desire to avoid a similar dispute in the future – is ever present in the minds of committee members and agency staff.

During a September 2013 review of Perjeta (pertuzumab) for accelerated approval in neoadjuvant breast cancer, agency staff and committee members pressed Genentech about its plans if confirmatory trial results fell short. A Genentech exec said the company would be willing to withdraw the neoadjuvant indication if the confirmatory trial results were “clearly negative,” although she left some room for how that could be interpreted Also see "“Not Another Avastin,” FDA Panel Warns Genentech At Perjeta Review" - Pink Sheet, 16 September, 2013..

We Promise, We Promise

Despite AstraZeneca’s statements in its briefing book, committee members sought greater assurance from the sponsor if the SOLO-2 results fail to replicate the findings from a subgroup of patients with germline BRCA mutation (gBRCAm) in Study 19, a Phase II trial. In that subgroup, olaparib was associated with an 83% improvement in PFS, which translated into a median advantage of 7.1 months over placebo.

The efficacy was less robust in the broader population, where olaparib was associated with a statistically significant 65% improvement in PFS that amounted to a median benefit of just 3.6 months.

University of Massachusetts oncologist James Liebmann asked AstraZeneca what the company believed would be a meaningful PFS improvement in SOLO-2.

“Presumably you don’t think 3.6 months is meaningful or otherwise you would have brought an indication for all patients with ovarian cancer in relapse to be treated with olaparib,” he said.

“We are looking to replicate the effects observed in Study 19 and to demonstrate consistent results in germline BRCA mutated patients, so we are going to be looking at approximately six months,” said AstraZeneca VP-Global Regulatory Affairs Oncology Hesham Abdullah. “Certainly we’re going to look to see if we can also replicate that in the context of the hazard ratio.”

However, Liebmann continued to press the issue. “My question with the confirmatory trial for olaparib is, if there is a statistically significant improvement in progression-free survival but because it’s so powered that that’s only one-and-a-half months, or two months or 3.6 months, would the sponsor be in agreement that the drug should in fact lose its approval if it gets accelerated approval?”

“We would certainly offer that and we would completely agree with you because these are not the treatment effects that we’re looking to replicate or demonstrate in SOLO-2,” Abdullah said.

After further discussion, committee chairman Mikkael Sekeres, a Cleveland Clinic oncologist, asked, “Do we feel as if we have a public acknowledgement from the sponsor that they’re going to be looking for the same degree of PFS in the confirmatory study and if they don’t get that they will voluntarily withdraw their drug from the market?”

“I feel that way,” Liebmann responded.

“We are certainly quite confident about replicating these results,” Abdullah said.

Nevertheless, AstraZeneca was pressed on this question again later by Antoinette Wozniak, Karmanos Cancer Institute.

“If this drug does get accelerated approval and the SOLO-2 trial does not show you what you want, so instead of a six-month progression-free survival it might be four-and-a-half months or whatever, what happens? Because once you sort of jump off the cliff of approval, it’s sort of hard to jump back on.”

“If the results of the SOLO-2 trial are not consistent with the results from Study 19, we would certainly offer to withdraw the drug from the market, including of course the proposed indication,” Abdullah responded.

Magnitude Of Benefit

A key issue in the Avastin dispute was FDA’s view that two confirmatory trials failed to verify the magnitude of benefit seen in the original E2100 study. In that initial study, bevacizumab demonstrated a 5.5-month median PFS benefit.

Although the PFS results from the AVADO and RIBBON-1 confirmatory trials were statistically significant, the median PFS benefit in the two trials was considerably lower, ranging from 0.8 months to 2.9 months.

Genentech, however, argued that CDER never established the magnitude of PFS benefit that the two confirmatory studies needed to achieve Also see "Avastin Breast Cancer Dispute Centers On “Magnitude Of Benefit” Questions" - Pink Sheet, 30 May, 2011.. The company and its experts also emphasized the hazard ratio as a more appropriate measure of a therapy’s efficacy than the median Also see "Avastin Debate Could Provide Clarity On Use Of Progression-Free Survival Endpoints" - Pink Sheet, 11 July, 2011..

Seeking to avoid a similar dispute in the olaparib setting, FDA asked the committee to comment on what would be an appropriate magnitude of treatment effect on PFS in terms of median improvement and hazard ratio that should be demonstrated in SOLO-2.

“I would remind people [three] years ago we had a different set up in this room,” Pazdur said, referring to the bevacizumab public hearing. “It was regarding the Avastin issue for breast cancer, which is the identical issue that we’re talking about. What is the clinical meaningfulness of a PFS endpoint in the context of not having an overall survival advantage? That is the essence of this application and why we brought it.”

However, the agency did not get much clarity from committee members, who struggled to come up with a number for a PFS treatment effect that should be demonstrated in SOLO-2. Suggestions ranged from six months to many years.

“I have a hard time also coming up with a set point” for PFS, Liebmann said. “I was actually frankly kind of interested that the sponsor was willing to concede that 3.6 months wasn’t good enough. I thought it was interesting that we have a bottom, we don’t necessarily have a top. I don’t know how you negotiate with that.”

Wozniak said what would be considered a good PFS result today might be different several years down the road after other treatments have been approved.

“I think that number is a little harder to determine,” Wozniak said. “Is six months good for ovarian cancer? Maybe, right now. Maybe it won’t be in a couple years.”

AstraZeneca Takes “Anti-Avastin” Pledge On Accelerated Approval Withdrawal

More from United States

More from North America