You Can't Fight The FDA

This morning, shares in GPC Biotech and Spectrum Pharmaceuticals are plummeting after a panel of experts advised the U.S. Food and Drug Administration to delay approving their prostate cancer pill, satraplatin.

Shares in GPC, the German biotech that would have sold the drug, dropped 33% to $13.54. Spectrum, which would have received a royalty on sales, dropped 12% to $4.19. (For previous doubts on satraplatin, see "Cancer Drug Winners And Losers.")

There were big hopes for the pill because a clinical trial showed that it reduced the nearly constant pain of patients whose prostate cancer had spread to the bone and had failed other treatments. Better yet, it slowed the cancer's spread, according to a measure devised by the company. Those results were highly statistically significant.

But both the way pain was measured in the trial and the way that progression of the cancer was gauged were deemed unreliable by the 12-member FDA panel. There was a spectacle at the event, watched via a Webcast. It basically came down to a debate between the company and the FDA in which the FDA insisted, fairly strenuously, that it had let the biotech know that its measures of disease progression and pain were not valid.

Biotech companies are run by brilliant, somewhat stubborn people who have great faith in their opinions. Somehow, it sometimes seems, executives can forget that, whatever happens, you can't fight the FDA. If you come up with a brilliant way of proving your drug works but you can't convince the FDA that it's brilliant, you need to come up with another approach.

The meeting was vaguely reminiscent of the story earlier this year of Encysive Pharmaceuticals (nasdaq: ENCY - news - people ), which saw its drug for a rare lung disease rejected even as a rival from Gilead was approved. On a conference call, Chief Executive Bruce Given basically blamed the FDA for his troubles, saying he had thought that the agency had been convinced by existing data after several go-arounds. Encysive's drug might well have made it if Encysive had run another clinical trial earlier in the process. Given has since left the company.

GPC's case is less onerous. But either the FDA failed to give good guidance or GPC failed to listen. If it had used a more validated measure of patients' pain, the panel might very well have voted to approve its drug. Instead, Otis Brawley, the Emory University oncologist who is the incoming chief medical officer of the American Cancer Society, answered the question of whether satraplatin had been proved to reduce pain with one of the most painful words in biotech: "almost."

Still, one has to wonder if the stock market is overdoing its punishment of GPC today. Certainly, the company won't win any prizes for wisdom or foresight. And at the meeting, it revealed that it would take much longer for final data, which it hopes will prove satraplatin increases survival, to become available. Patients in its study are simply living longer than expected, whether they got satraplatin or a placebo.

Now, those data won't be available until next year; they were expected at the end of this year. Megan Murphy, an analyst at Lazard Capital Markets, who covers Spectrum, has removed all satraplatin royalties from her model of that company, which she still rates a "buy." And there is also the problem that it's not clear whether the study will really prove to oncologists that they need to use this drug. It could get approved, and then not sell.

But this appears to be a medicine that works, although GPC hasn't proved that yet, and the proof could come next year. Not everyone is so confident. In a note to investors, Matthew Osborne, a biotech analyst at Lazard Capital Markets, said the odds of a survival benefit were only 50/50.

Given all the lousy bets out there in biotech, investors might want to watch the stock closely as a potential value play.