Avastin

On Friday, we posted a press release from Breast Cancer Action (Breast Cancer Action: Patients Lose, Genentech Wins with FDA’s Avastin Ruling) criticizing the FDA’s approval of Genentech’s [NYSE:DNA] Avastin for treatment of metastatic breast cancer. We noted that it was odd that the FDA approved this drug, given that the Oncologic Drugs Advisory Committee (ODAC) had recommended. Applications for drugs to be approved for new uses are first sent to Advisory Committees comprised of physicians with expertise in that speciality (here, oncology). If the Advisory Committee recommends that the drug not be approved, the FDA usually approves it — and if it advises that it not be approved, then the FDA usually doesn’t. As an article in BioWorld Today (FDA Splits with ODAC, OKs Avastin Use in Breast Cancer) describes, the vote of the ODAC was close, but serious concerns were still raised by it:

The approval came as somewhat of a surprise, considering that the FDA usually follows the advice of its advisory panels. Although the ODAC vote on Avastin in breast cancer had been close, at 5 to 4, the panel had raised concerns about trial design, toxicity and survival data.

Breast Cancer Action raised some serious concerns about the FDA’s accelerated approval of Avastin for metastatic breast cancer in its press release which we posted on Friday. Barbara Brenner, Breast Cancer Action’s Executive Director, raises some additional concerns about the approval — not about the drug itself, but about how the drug was approved, and based on what type of clinical trial. In an email to advocates (reprinted here with her permission), she describes the danger:

I want to give folks some context for what happened here, particularly in light of the recommendation of the ODAC some weeks ago that the drug not be approved at this time. BCA encouraged that outcome. See the first story on the home page at http://www.bcaction.org/ for the history.

What happened today is that the Commissioner of the FDA, Andrew Von Eschenbach caved to enormous pressure from industry. Some of you have seen the Wall Street Journal editorial yesterday saying that it would be a “moral tragedy” if the drug weren’t approved because lives were at stake, even though there is no evidence that the drug improves survival. What we know is that the head of Genentech met with the WSJ editorial board last week. And they’ve been dribbling out little bits of data about other trials that also don’t have survival data.

The FDA Commissioner, as you know, is never bound by an ODAC recommendation. In this case, the Commissioner Von Eschenbach is as tied to industry as the rest of the Bush administration is.

As you can see from Genentech’s press release , we are very sad that that industry’s interests have trumped those of patients in the case of this particular drug, and that they have evidently acceded to a lower standard of drug approval — PFS (progression free survival) — instead of overall survival.

But the standard has been changed in another critically important way as a result of today’s decision. Before today, a drug could not be approved for marketing unless it had been subjected to a drug registration trial with standards set by the FDA. This standard is higher than the standard for classical clinical trials, largely because routine clinical trials often mask the true side effects of drugs or are done in populations that are likely to result in overstated benefit.

The decision made today was on the basis not of a registration trial, but of an NCI-sponsored clinical trial, not designed for drug registration. This means that, in the future, other drugs companies will be able to argue to the FDA (probably successfully) that the FDA drug registration standards don’t have to be met, basing approval on clinical trials will be just fine.

As I sat last week trying to figure out why Genentech had not even mentioned the AVADO trial (the one that was designed as registration trial, the results of which were expected in the first three months of this, and were in fact partially released last week, one week before the decision today) in their Avastin presentation to the ODAC in December, I came to a stunning realization: what Genentech was hoping for was a decision that would permit drugs companies to avoid the rigors of registration trials in the future, and rely on clinical trials that are often controlled by industry. And that is what they got with today’s decision.

The bar has been lowered in more ways than we can count. A very sad day for patients.

The BioWorld Today article described above gives some of the history:

In May 2006, Genentech submitted a supplemental biologics license application to the FDA seeking approval of Avastin in breast cancer. But the FDA raised concerns about the underlying data, which had not come from a traditional double-blind, placebo-controlled, company-sponsored Phase III trial. Instead, the sBLA was based on an open-label study, known as E2100, conducted by the National Institutes of Health-affiliated Eastern Cooperative Oncology Group. In September 2006, the agency delivered a complete response letter asking Genentech to audit the E2100 data in the same manner expected of a company-sponsored trial. (See BioWorld Today, Sept. 12, 2006.)

Genentech did as the FDA asked, resubmitting the revised sBLA in August 2007. But a second setback occurred in December, when ODAC narrowly voted against approval. (See BioWorld Today, Dec. 6, 2007.)

E2100 had randomized 722 previously untreated breast cancer patients to receive Avastin plus paclitaxel chemotherapy or paclitaxel alone. The study met its primary endpoint of improving progression-free survival (PFS): patients receiving Avastin achieved a median PFS of 11.3 months compared to 5.8 months for the control arm. However, there was no statistically significant difference in median overall survival, with the Avastin group surviving 26.5 months and the control group surviving 24.8 months, leading ODAC to debate the clinical merits of PFS.

The panel also raised questions about the design of E2100 once again and voiced concerns about adverse events. Six deaths were found to be ‘definitely or probably’ caused by toxicity related to Avastin, and 71.1 percent of patients in the Avastin arm experienced severe adverse events, compared to 51 percent in the control arm.

But between ODAC’s negative decision and the FDA’s positive one, data from a new Phase III trial became available. Known as AVADO, the trial was sponsored by Roche and properly designed. The randomized, double-blind, placebo-controlled Phase III study compared Avastin plus docetaxel chemotherapy to placebo plus docetaxel in the first-line treatment of 736 patients with locally recurrent or metastatic HER2-negative breast cancer. Avastin was administered at 15 mg/kg or 7.5 mg/kg every three weeks, and both doses resulted in a statistically significant improvement in PFS. [PAL Ed.: But not in overall survival, as Breast Cancer Action also pointed out in the trial upon which the approval was formally based] (See BioWorld Today, Feb. 14, 2008.)

Although the AVADO data were not officially included in the Avastin sBLA, Genentech submitted them to the FDA for consideration. That may have helped to sway the agency’s decision to go ahead and grant accelerated approval.

In its news release, Genentech said that converting its accelerated approval into a full approval will be dependent on an FDA review of the full AVADO data as well as data from a Genentech-sponsored Phase III trial known as RIBBON I. Initial RIBBON I data are expected later this year, while overall survival data from AVADO are expected around mid-2009.Genentech also plans to provide the FDA with data from three additional randomized trials that are either ongoing or planned.

So the question remains: Will FDA’s approval of a new indication based on a trial that was a not traditional double-blind, placebo-controlled, company-sponsored Phase III trial pave the way for the loosening of standards on what data the FDA will accept for approvals? Does the FDA’s overruling of its own advisory committee make it likely that the FDA will ignore its advisory committee recommendations more often? And what does the FDA’s approval of a drug based on a secondary endpoint like progression-free survival when the drug failed to meet the primary endpoint of overall survival bode for the future?

Got thoughts on these questions? Post ’em in the comments.