Shares of EPIX Pharmaceuticals Inc. slid more than 18 percent Friday after reporting that its oral 5-HT1A agonist, PRX-00023, failed to show statistical significance in a Phase III study in generalized anxiety disorder (GAD), but the company said it was encouraged by subset data in depression.

The double-blind, placebo-controlled trial enrolled 310 patients who were randomized to receive 40 mg once-daily for three days followed by 80 mg once-daily for the reminder of the eight-week study. Results showed that PRX-00023 suggested a positive trend in treated patients but missed the primary endpoint, which was defined as a change from baseline compared to placebo, as measured by the Hamilton Rating Scale for Anxiety.

To meet that endpoint, patients in the treated group had to show a two-point drop in HAM-A score over placebo, but top-line results showed a difference of only 1.3 points between the PRX-00023 arm and placebo.

That news cut 99 cents off the company's stock (NASDAQ:EPIX) Friday, and it closed at $4.32.

"We're disappointed in the results in GAD," EPIX CEO Michael Kauffman said in a conference call, adding that a higher than expected placebo response might be to blame.

However, he said, a subset analysis showing "encouraging results," of PRX-00023 in patients in the GAD trial suffering symptoms of depression convinced EPIX to move forward with the drug in major depression.

That subset - defined as patients with scores of 20 or higher on the Montgomery Asberg Depression Rating Scale, which comprised nearly half the enrolled patients - showed a change in baseline of 6.3 points in the PRX-00023 arm vs. 3.3 points in the placebo arm.

"We believe we have a clear path forward in depression," Kauffman said. The company's plan is to "continue analysis and then initiate a Phase II trial in major depressive disorder," anticipated to begin in the first half of 2007.

Although the depression space is "fairly crowded with various serotonin reuptake inhibitors and mixed reuptake inhibitors," the tolerability profile on PRX-00023 and its preliminary efficacy signal "suggest that it can compete in the depression market," Kauffman said.

He reported that there was a low rate of discontinuation due to adverse events. Safety and tolerability data from the Phase III study show no difference between the treatment and placebo arms in sexual function, sleep and appetite.

EPIX has anticipated partnering PRX-00023, and Kauffman said the recent Phase III results don't change that plan. He said the company will move forward on the Phase II trial in depression while seeking interest from potential collaborators.

PRX-00023 was one of several programs added to EPIX's pipeline through its August merger with Predix Pharmaceuticals Holdings Inc. That deal, announced in April, was valued at about $90 million, including EPIX's agreement to assume net debt at closing, along with a potential milestone of $35 million. (See BioWorld Today, April 4, 2006.)

The merger also brought on board PRX-08066, a selective, small-molecule 5-HT2B antagonist, in a Phase IIa study for pulmonary hypertension associated with chronic obstructive pulmonary disease, as well as PRX-03140, a small-molecule agonist of 5-HT4, which has completed Phase I studies in Alzheimer's disease.

Prior to the merger, EPIX's focus had been on imaging agents. Its lead product, Vasovist, a blood-pool imaging agent, is approved in Europe and Australia, and has been deemed approvable by the FDA, though the agency requested last month that additional trials be conducted. (See BioWorld Today, Aug. 29, 2006.)

Post-merger, the company said it had about 29.1 million shares outstanding, with a cash position of $114 million. Excluding new partnerships or milestones and assuming a cash payment to Predix shareholders, that amount should sustain operations through the end of 2007.