Many people might not have noticed Rigel Pharmaceuticals until their stock fell down the stairs today. We'll pause a moment to consider the fate of this upbeat analyst, whose clients were long-and-wrong if they followed his recommendation to buy the stock a couple of weeks back. (There's a case to be made for buying the stock now that it's been hammered, but that's a topic for another day).

Rigel had taken an inhibitor (R112) of a kinase enzyme called Syk into the clinic for nasal allergies, and they missed their endpoints by a mile. They ran about 400 patients through a seven-day course of either their inhaled drug, a placebo, or Beconase, an inhaled steroid from GSK. Beconase was significant against placebo, but R112 wasn't.

But what must have been especially hard to take was that they'd already run an earlier trial with over 300 patients, where R112 seemed to work pretty well versus placebo. These results were a big part of why Rigel's stock was as high as it was. What went wrong?

We can start with the clinical trials themselves. I'm going to stipulate that today's result was correct, and that the compound did indeed fail. I think the two trials were measuring similar endpoints, so one possibility is that the earlier one was just run incompetently. I'm not in a position to say, and you'd like to rule that explanation out, but it's a possibility. What looks like a good P value can cover a multitude of sins.

A second level of explanation is the drug's mechanism. It's hard to believe that the underlying biology is completely mistaken, though. Syk is a reasonable target for allergy. It's involved in the signaling inside mast cells (and some other members of the immune system) after they're stimulated by an antigen, and blocking it would seem to be just the ticket.

I'm going to go with a hybrid theory. The first efficacy trial was only a two-day evaluation. Perhaps there's a compensatory mechanism that kicks in and cancels out the effect of the drug on repeated dosing? That would make this failure a team effort between the biology and the design of the first trial, and that's what seems most likely to me in the absence of other data. Fortunately for Rigel, this wasn't a particularly long or expensive trial - they announced only back in August that they'd enrolled their first patients. It's a nasty failure, but they got out of it relatively cheaply.

And they have deals with a range of other companies for other kinase inhibitor projects. One of them, though, is of particular interest after today's news. Pfizer is a partner of theirs for their next-generation Syk kinase inhibitors, this time aimed at the asthma market. Is my compensatory mechanism guess correct, and does it apply to that therapeutic area, too? Pfizer will be paying to find out. . .