SAN DIEGO (TheStreet) -- A few weeks ago, I called La Jolla Pharmaceuticals (LJPC - Get Report) a "hedge fund roach motel" and expressed doubts about the company's galectin inhibitor GCS-100, in part because the dodgy history of the drug and the target.
Eh, I probably was too harsh with the "roach motel" line, but there are still a lot of unanswered questions about GCS-100 stemming from Monday night's release of results from a phase II study in chronic kidney disease patients.
La Jolla shares are up 74% to $18.97 Tuesday on what the La Jolla CEO George Tidmarsh described as "positive," "remarkable," "consistent," "predictable" and "important" GCS-100 data. That's a lot of adjectives!
The phase II study enrolled 121 patients with chronic kidney disease and randomized them to placebo or two different doses of GCS-100: 1.5 mg or 30 mg. After 8 weeks of treatment, La Jolla says patients treated with 1.5 mg of GCS-100 demonstrated a statistically significant improvement in kidney function and reduced blood levels of galactin-3, the target of the drug.
What happened to patients treated with the higher 30 mg dose of GCS-100. Basically nothing. These patients had almost no change in kidney function and a large increase in blood levels of galectin-3. The low dose of GCS-100 works but the higher does was totally ineffective. That's called an inverse dose response and it's uncommon and strange. GCS-100 is designed to knock down the galactin-3 protein and thereby improve kidney function, but a high dose of the drug did exactly the opposite.
These are what the primary endpoint data on GCS-100 look like graphed:
And these are the data on changes to galectin-3 levels.
La Jolla's Tidmarsh claims the high dose was never supposed to work because of a "rebound" or "biological feedback loop."
Hmm. Okay, but what exactly is this "biological feedback loop" which causes a higher dose of GCS-100 to not work?
"We haven't gotten to the bottom of it yet... We are currently sorting that phenomenon out," said Tidmarsh on a conference call.
Here's a possible explanation: GCS-100 doesn't really work at all. This seems as plausible as La Jolla's claim that a really low dose of GCS-100 is effective but a higher dose is not.
One more interesting interpretation of the GCS-100 data to share. This is not from La Jolla but was produced by a helpful Twitter follower last night:
The vertical lines represent the standard deviation of the kidney function data released by La Jolla. The standard deviation measures the variability of data collected. A large standard deviation means there's more variability in the data collected i.e. more patients exhibiting extreme measurements further from the mean, or average.
Look at the chart: The vertical lines represent the standard deviation for each group. When the standard deviations overlap greatly, as they do here between the placebo, GCS-100 1.5 mg and GCS-100 30 mg patients, it means a large percentage of these patients are all acting the same.
What La Jolla calls "positive" data on GCS-100 is really just statistical noise.
But in a biotech stock bubble, investors don't really care about noise, as long as someone calls study results positive.
Follow Adam Feuerstein on Twitter.