What the FDA giveth, the FDA can taketh away.
A few weeks ago, the FDA yanked Roche's right to market Avastin to treat breast cancer. Avastin was given an accelerated approval for the indication, which allows the drug to be marketed with less-than-thorough data. As part of an accelerated approval, drugmakers agree to run additional tests to confirm the initial findings.
Unfortunately, the confirmatory trial failed to confirm that Avastin helped the average patient, and the FDA had no choice but to yank the approval.
Roche is an obvious loser here, but it's not as horrible as when Merck (NYSE: MRK ) had to pull Vioxx off the market or GlaxoSmithKline lost Avandia. Avastin is still on the market for other types of cancer. The company doesn't break out the $6.2 billion in annual sales by individual indications, but it certainly won't lose all the sales. Avastin is still approved to treat colorectal cancer, lung cancer, kidney cancer, and a form of brain cancer called glioblastoma. Theoretically, it could be used off-label for breast cancer patients, although it might be hard to get insurance companies to pay for it.
The real losers here are companies that gained FDA approval through the priority approval process, as Genentech -- now owned by Roche -- did with Avastin. According to a Wall Street Journal article earlier this year, there were 13 cancer drugs that hadn't finished clinical studies to confirm their benefits, including Eli Lilly (NYSE: LLY ) and Bristol-Myers Squibb's (NYSE: BMY ) Erbitux, which was approved more than seven years ago. And there have been additional approvals since the article was published, such as Seattle Genetics' (Nasdaq: SGEN ) Adcetris.
Listening to FDA advisory panel meetings and reading the related documents, I get the feeling that the FDA has a love-hate relationship with the accelerated approval process. As scientists, they want definitive data showing that the drugs work. But there are diseases with unmet needs where early approvals could help many patients while waiting for that definitive data. Science isn't quick, so it's a balance between definitive proof and timely access.
The worry is that every failure -- about 10% of cancer drugs have failed to confirm initial results -- will push the FDA toward the "hate" spectrum of the love-hate relationship. The agency may become more demanding of drugmakers that haven't turned in the clinical trial results yet and could be more rigorous with those asking for accelerated approval.
Health insurers will benefit because they can justify not footing the bill for the $88,000 treatment, but that's a relatively short-lived benefit. As premiums reset, the potential cost will be removed from the complicated formula for determining premiums. Insurers are just a middleman with a markup, so the relative cost isn't all that important.
Patients might flock to Gemzar, which is approved to treat breast cancer, but its developer, Eli Lilly (NYSE: LLY ) , isn't going to benefit much; sales in the third quarter dropped 72% as generic competition kicked in. Teva Pharmaceuticals (Nasdaq: TEVA ) , Dr. Reddy's Laboratories, and Watson Pharmaceuticals all make generic versions of Gemzar, but it's a small contribution to the overall sales, so don't expect a breakout quarter because of the FDA's action on Avastin.
The big winners will be companies with breast cancer drugs in the pipeline that will see reduced completion with Avastin no longer marketed for that indication.
A class of drugs called PARP inhibitors seems like the best hope. By inhibiting PARP, the drugs make tumor cells more sensitive to chemotherapy because the PARP protein is involved in removing chemotherapy-induced platinum damage from DNA. Without the repair mechanism, the tumor cell is more likely to die.
Abbott Labs (NYSE: ABT ) has a PARP inhibitor called veliparib, as does Pfizer (NYSE: PFE ) . And Teva Pharmaceuticals (Nasdaq: TEVA ) has one as well through its purchase of Cephalon. Just don't bet the farm on the compounds; Sanofi's PARP inhibitor, iniparib, failed a phase 3 trial earlier this year. We'll have to see how the data plays out for the other before we know whether the mechanism of action doesn't work of whether iniparib just wasn't a strong enough inhibitor.