Cypress Bioscience Inc. lost more than half its value Thursday after announcing that a pivotal Phase III study of milnacipran missed its primary endpoint as a treatment for pain associated with fibromyalgia.

The stock (NASDAQ:CYPB) toppled more than 56.5 percent, or $7.17, to close out the day at $5.53.

Milnacipran, a norepinephrine serotonin reuptake inhibitor, represents Cypress's lead clinical candidate. The company began developing it in August 2001, after licensing it from Pierre Fabre Medicament SA, a division of Paris-based bioMerieux Pierre Fabre. Used to treat depression in more than 3 million patients worldwide, milnacipran has been marketed outside the U.S. since 1997.

After completing a Phase II trial, Cypress started its first Phase III fibromyalgia study in the fourth quarter of 2003 and finished it in July. A second trial began in October 2004.

Results announced Thursday from the first trial included only top-line data showing that milnacipran did not achieve statistical significance in its primary endpoint. Nevertheless, Cypress and its partner, Forest Laboratories Inc., plan to continue with the ongoing second Phase III trial, and the initiation of an upcoming third Phase III study. The failed Phase III trial probably will delay the timing of a new drug application, originally slated for the end of 2006.

"The delay is likely to be a minimum of three months and may be as much as one year," said Jay Kranzler, chairman and CEO of New York-based Cypress.

In January 2004, Cypress signed over U.S. development and marketing rights for milnacipran to New York-based Forest in a deal worth up to $250 million. Under the agreement, Forest is responsible for sales and marketing, but Cypress has an option to co-promote milnacipran using its own sales force. (See BioWorld Today, Jan. 12, 2004.)

Kranzler said in a conference call Thursday that he expects Forest to remain committed to milnacipran and likely would cover any costs involved in amending the ongoing Phase III trial.

"Although the design of this trial is almost identical to the first one, we have the ability to modify this second trial in several critical ways based on what we have learned today," Kranzler said.

Cypress might decide to increase the size of the trial or to finalize a statistical analysis plan that will improve the validity of the responder analysis. The company was challenged in designing the first trial because the FDA wanted it to be a six-month study, and the responder analysis would not distinguish between different magnitudes or the timing of responses.

"We were concerned, as there was no precedent for drugs in fibromyalgia and related conditions lasting beyond 12 weeks in their efficacy," Kranzler said.

The completed pivotal trial, a randomized, double-blind, placebo-controlled study, included 888 patients. The primary endpoint was a composite response rate of an assessment of pain as measured by the Patient Experience Diary (PED), and an assessment of overall impression of patient well-being as measured by the Patient Global Impression of Change (PGIC). The "p" value for patients receiving 200 mg per day of milnacipran was 0.058 at three months using the Last Observation Carry-forward (LOCF) analysis.

"The trial certainly came close to achieving significance for the pain label," Kranzler said.

At the six-month mark, the "p" value was 0.053, and the results were statistically significant when the company used the Baseline Observation Carry-forward (BOCF) analysis (p=0.048). The FDA, however, might not accept the BOCF results in a registration study.

"This is not what we chose to use in the pre-determined analysis plan," Kranzler said.

Cypress also found that the magnitude of the treatment effect was maintained at the six-month analysis, indicating a durable response to milnacipran. Another endpoint of fibromyalgia syndrome, a composite of the primary endpoint and a physical function assessment, did not reach statistical significance compared to placebo at either the three- or six-month time frame.

Fibromyalgia syndrome is characterized by widespread pain and stiffness in the body, and is accompanied by severe fatigue, insomnia and mood symptoms. It is estimated to affect 6 million to 12 million people in the U.S. There are no marketed treatments.

Common prescriptions are "things like non-steroidal anti-inflammatories," said Sabrina Martucci Johnson, Cypress's chief financial officer. Patients also take antidepressants, muscle relaxants and sleep medications.

In terms of tolerability, milnacipran was generally well tolerated in the Phase III trial, with the most common adverse events being nausea, heart rate increase, headache and depression - all seen in previous trials.

The third Phase III trial, which could start in early 2006, would serve as a backup trial in case Cypress cannot submit an NDA solely based on data from the first two studies.

"We had always planned to do a third Phase III," Martucci Johnson told BioWorld Today. "It was up in the air as to whether we would need to complete it before we filed the NDA."

As of June 30, Cypress had cash, cash equivalents and short-term investments of about $111.5 million, and about 31.6 million shares outstanding. The company expects an operating burn for 2005 of between $12 million and $20 million.

Aside from milnacipran, Cypress is working on one other program, evaluating treatments for obstructive sleep apnea. The company entered three licensing agreements for that program: It licensed mirtazapine-related patents from Organon Ltd., of Dublin, Ireland, and patents from two other parties for combining mirtazapine with a second approved agent. That program is in pilot Phase IIa trials.

"The very earliest we could have data from these pilot trials that are ongoing is the first half of next year," Martucci Johnson said.