Cyberonics Still Misses FDA's Point

Thursday night, Cyberonics'(CYBX) CEO Skip Cummins held his 463rd investor conference call since August (OK, I'm exaggerating a little), when the Food and Drug Administration rejected the company's application to market its vagus nerve stimulation (VNS) device as a treatment for severe depression.

Cummins can't seem to go more than a few days without getting on the box to tell us all how misguided the FDA decision was, and how Cyberonics is determined to convince regulators to change their minds. And once again, Cummins was in Washington, D.C., for the call, not at Cyberonics' Houston headquarters, which makes me wonder if he's pitched a tent and is now living in the FDA parking lot.

But all joking aside, when you get beyond Cummins' persistence, there isn't much substance to his soliloquies. And there's certainly nothing to warrant confidence that Cyberonics is any closer to getting the VNS device approved for depression.

This latest conference call was convened so Cummins could explain that Cyberonics had submitted additional data on the VNS device to the FDA. Cummins said the new information, which included two-year safety and efficacy data from depressed patients implanted with the VNS device, was intended to answer the issues raised by the FDA in August when it sent the company a "not approvable" letter.

Cummins says the FDA has up to 180 days to respond to the new data, but that he expects the agency to have an answer in 120 days, or the end of January 2005.

That's fine, but none of this new data addresses the FDA's central concern: a lack of randomized, controlled data to prove the efficacy of the VNS device as a depression treatment. This is the issue that tripped up Cyberonics at the FDA, and nothing that the company has provided the agency so far, including Thursday, changes that fact.

It's something Cummins can't seem to come to grips with. In response to several questions during Thursday's conference call, Cummins suggested that randomized, controlled data wasn't necessary or relevant. He insisted that the two-year, non-randomized data submitted obviated the need for randomized, controlled data (stronger, more convincing) because any placebo effects that may have cropped up in the studies would disappear after such a long period of time.

Huh? By Cummins' logic, this implies that no drug company would ever need to run randomized controlled studies for anything. All they would have to do is give patients a drug and follow their progress for long enough to get the positive effects they're looking for.

That's nonsense. You can run a non-randomized study for as long as you want, but that doesn't mean potential biases disappear. There's a reason why rigorous, randomized and controlled clinical trials are the cornerstone of every good -- and approvable -- FDA application.

Besides, it doesn't really matter what Cummins thinks, because the FDA told Cyberonics in the not approvable letter that it wants the company to run another randomized controlled study. It took nearly the entire conference call for Cummins to admit that fact, but he finally did. Of course, Cummins then added that doctors and depression experts have advised the company that such a trial was impossible to run. Well, pardon me for being skeptical about the opinion of Cyberonics' paid consultants; I'd rather believe the FDA.

If the FDA was really on the fence about approving the VNS device for depression back in August, the agency could have sent Cyberonics an approvable letter. That's like a conditional approval, contingent on the company submitting some additional data or information to clarify some outstanding issues. But that's not what happened. The FDA rejected the VNS device outright. That's serious. It's not something that happens very often.

For Cyberonics to think that it can throw together some data, especially data that are not very new or relevant, in six weeks' time and somehow use that to convince the FDA to reverse course seems like a pipe dream to me. More likely, Cyberonics will still have to conduct another clinical trial -- randomized and controlled -- if it ever wants to get the VNS device approved for depression. That's going to take a few years and more than a few dollars.