FDA and Avastin: Crossroads in an Era of Targeted Therapies

FDA, ODAC, and Accelerated Approval

Debu Tripathy

In early 2008, based on the results of its E2100 trial, which showed significant improvements in progression-free survival when combined with paclitaxel, Avastin (bevacizumab) gained an FDA accelerated approval. Median progression-free survival in the Avastin arm was 11.3 months compared with 5.8 months for paclitaxel alone (although overall survival in the two arms was similar). However, final approval would be dependent on subsequent trials showing similar degrees of benefit. When two additional trials were submitted for review, both showed significant improvements in progression-free survival, but again with no difference in overall survival. Subsequently, on July 20th of this year, the Oncologic Drugs Advisory Committee (ODAC) voted 12 to 1 not to recommend permanent approval of Avastin as first-line therapy in advanced breast cancer.

The issues surrounding the FDA’s process for the final decision on Avastin in this difficult clinical setting raises important questions about the standards of drug approval in this changing era of targeted therapy and personalized medicine. In the late 1990s, the FDA was urged to make sure that promising new drugs would be made available to patients as soon as possible; thus, the accelerated approval process was mandated by Congress and by Presidential decree.

Along with its regulatory work, one of the important roles of the FDA is to facilitate the development of drugs for very difficult to treat diseases, including cancers. So the general philosophy surrounding the accelerated approval process was that if the FDA received compelling results from early trials in a cancer in which no curative treatment existed- but these results did not carry the weight that would lead to a final approval-the agency would grant accelerated approval along with a statement of the conditions required to garner full approval.

Consider, for example, a drug that is being tested in a group of patients for whom there is no standard therapy; all you have are phase II data documenting a response in one-third of the patients and an apparently reasonable safety profile. These data would be grounds for submission for an accelerated approval. To gain permanent approval, the FDA would require a phase III trial either comparing the agent to the standard of care, or if there is no standard of care, then comparing it to best supportive care.

This general process has been followed for several oncology drugs. Regarding Avastin, the first question is, did the agent show preliminary activity that would have qualified it for accelerated approval. I would answer yes for two reasons; first, we do not have a biological agent that clearly extends survival when added to chemotherapy in HER-2 negative disease, and second, the randomized phase III E2100 trial did show almost a doubling of progression-free survival, a pretty significant effect, even though it didn’t show a difference in overall survival. The FDA said that they wanted to see confirmatory trials. There were two trials in progress at this time, and when their data were submitted to the FDA, they showed results similar to those of E2100; although the magnitude of the improvement in progression-free survival wasn’t as great, both trials showed statistically significant benefits.

The FDA never said that the confirmatory trials needed to show overall survival benefits; the agency simply implied that the trials needed to confirm the original findings of E2100, which in my mind they did. So why did ODAC vote to remove Avastin’s indication as a first-line treatment for metastatic breast cancer? I think that the committee felt that the initial E2100 trial and the confirmatory trials (AVADO and RIBBON-1) were qualitatively different. In other words, the weeks of progression-free survival in the confirmatory trial were less, there were more treatment-related deaths, and the overall survival was less-even though one cannot formally compare the degrees of benefit in these trials. But it’s also important to note that ODAC voted, albeit by a slim margin, against the original accelerated approval. So this riff between the FDA and ODAC raises questions about the process itself.

PFS vs OS: Did the FDA Change the Rules

As mentioned, there was no indication that the FDA wanted to see an overall survival advantage in Avastin. The regulatory agency doesn’t have ground rules regarding what benefit they’re looking for; they are simply charged in a more vague sense with determining if a drug is safe and effective. How they make that determination depends on the individual situation, evaluating each agent in its own clinical context. The unwritten rule, however, is that in first-line therapy, the FDA wants to see an overall survival benefit. The reason that Avastin was approved without showing overall survival was that the progression-free survival benefit was so great-almost a two-fold difference.

However, since none of the trials in Avastin were powered to show an overall survival benefit, I think that perhaps the rules were changed mid-stream-hence the ODAC vote to reject permanent approval. But this is a complicated decision-making process and ultimately if the progression-free survival in the confirmatory trials had been doubled, as it was in the first trial, perhaps we’d be looking at a different scenario. In the end, it’s all a matter of degree of benefit, and what the FDA is ultimately saying is that they don’t use any one metric, rather, they look to see whether the sum of the evidence shows that the benefit is not outweighed by the risk. In the Avastin approval process, in the agency’s mind, risk outweighed benefits.

A New Paradigm: Cost and Value

For the first time in the war against cancer, high-priced cancer drugs are coming under CMS scrutiny; could Avastin be a harbinger of things to come?

I think a lot of people are looking at the Avastin story as a bellwether that presages the future of expensive targeted therapies. By design, the FDA or any of its advisory boards do not filter costs into their regulatory and approval process. However, the issue of the economic cost of small incremental benefits has been raised, and some still wonder whether this factor somehow crept into the advisory board’s decision.

I think the pressure to consider the cost-vs-efficacy question is going to come from Medicare. The new head of CMS, Donald Berwick, has been moving very quietly but resolutely, looking at the way Medicare reimburses physicians. Part of the attempt to address the cost-vs-efficacy question is going to be a requirement that physicians demonstrate with outcomes data that they meet certain performance metrics.

Another part is going to be a payment cap on drugs and diagnostics to ensure that they meet a certain degree of benefit relative to their cost. While none of this has been codified yet, I think such moves are coming ecause studies indicate that many of the newer drugs and diagnostics have not been used in a discriminating manner. In other words, spending more healthcare dollars hasn’t necessarily equated to better outcomes.

I think we’ll see an incremental movement to assess cost against value when delivering healthcare. The Obama administration has devoted a lot of financial and intellectual capital to the comparative effectiveness research initiative, which in effect compares one drug or technology with another to determine which delivers the best cost for value. Ultimately, I think that not only Medicare, but also private insurers, are going to say, “I’ll cover this drug or procedure as long as it meets a particular benchmark that has been vetted by a comparative effectiveness entity.”

Moreover, researchers and venture capitalists in the drug development field will have to fashion their business to accommodate the cost vs value model. Just as no business person wouldbuild a $30,000 DVD player only because the resolution was slightly better, in a similar way, we’re moving into a new era of accountability in healthcare. Unfortunately, there will be some losers in this leaner business model, such as rare cancers, for which the incentives of risk vs reward will scare away potential investors. At the same time, I think it will encourage the pharmaceutical industry to develop drugs and diagnostics that are more cost effective, and that will have a positive effect on our healthcare industry. Finally, the public sector will need to strategically fill in the gaps for innovation and progress in these areas.