FDA's Approvable Problem

Nov 01, 2007

They scuttle product launches and send company stocks through the floor. They steal precious years from patents and require pricey new trials. They have the entire drug industry on edge, from struggling one-hit biotechs to struggling large-cap pharmas. They're FDA approvable letters—that odd regulatory response that isn't quite an approval and isn't quite a rejection—and in the past three or four years, they've become a major part of the FDA's arsenal for dealing with (or, as many say, not dealing with) new drug applications (NDAs).

As of October 1, the agency had already issued 23 approvable letters this year, including second letters for four drugs and a third for one. After hearing industry insiders grumble that a scared-safe FDA was upping the letter's use, Chris Milne of the Tufts Center for Drug Development looked into the issue in 2005 and counted 35 approvable letters for NDAs between 2001 and 2004, causing an average delay until approval of 20 months. FDA has issued 36 such letters since January 2006 alone.

So Approvability Gulch is a real place. The question is, what does the trend mean? Has FDA quietly raised its safety standards, as many people argue? Or is it using the letters to simply avoid approving drugs—waiting for the current political storms to settle? And if there is a new safety order, what are the potential risks for both the industry and the agency?

The Regulatory Creep of 'Basically Approvable'


John Jenkins, FDA
Here's how the Code of Federal Regulations defines approvable letters: "In selected circumstances, it is useful at the end of the review period for the Food and Drug Administration to indicate to the applicant that the application or abbreviated application is basically approvable, providing certain issues are resolved. An approvable letter may be issued in such circumstances.... As a practical matter, the approvable letter will serve in most instances as a mechanism for resolving outstanding issues on drugs that are about to be approved and marketed."

That may sound benign, but in the post-Vioxx real world, "resolving" an "issue" can lead a drugmaker off a cliff. Some "approvable" drugs are hardly on the verge of approval:

  • In 2005, Bristol-Myers Squibb killed its diabetes drug, Pargluva (muraglitazar), saying that it would take five years to gather data about cardio risks flagged by FDA. (The company may also have believed that additional data would sink Pargluva, a PPAR-antagonist like GlaxoSmithKline's Avandia, which was almost withdrawn this year because of its heart-attack risk.)
  • Sanofi-Aventis' novel obesity treatment, Acomplia (rimonabant), marketed in Europe, received an approvable letter from FDA in 2006. After submitting the FDA-requested data, the French firm got an unfavorable review from an FDA advisory committee in June and withdrew its NDA. Unlike BMS, Sanofi is standing by its much-vaunted drug and plans to submit a new NDA for diabetes. But the ETA is 2010, at best.
  • Novartis' Galvus (vildagliptin), its diabetes blockbuster-to-be, which was already on the market in Europe, received an approvable letter in February. The additional clinical trials are expected to postpone Galvus' launch until 2009, putting the DPP4-inhibitor hopelessly behind Merck's first-in-class Januvia, which started the race with a mere three months' advantage. This has cost Novartis at least $500 million in unmet projections and left its new U.S. rep army with little to sell.

Novartis CEO Dan Vasella, virtually the only pharma CEO willing to speak openly on the subject, has his take on the trend. "The FDA has become subject to politics," he told the Financial Times in September.