Amgen Inc. will return to the drawing board with its drug for Parkinson's disease, GDNF, after further review of preclinical studies showed a risk of brain damage.

Friday's announcement came nearly five months after the Thousand Oaks, Calif.-based company halted Phase II trials in Parkinson's disease because of safety issues. Amgen ended patient access to GDNF (glial-cell-lined-derived neurotrophic factor) in September, and said further clinical and preclinical review indicated that additional research would be needed to understand the drug's toxicology.

Amgen reported in June that GDNF missed its primary endpoint in the Phase II trial, failing to show an improvement in symptoms as defined by the Unified Parkinson's Disease Rating Scale (UPDRS). That study divided 34 patients into two groups: those receiving a 15-mg daily dose - administered by continuous infusion into the putamen - and those receiving placebo. (See BioWorld Today, June 29, 2004.)

"The trial results showed no difference between the patients who were getting GDNF and the patients who were getting the placebo," said Amgen spokeswoman Andrea Rothschild, adding that, in fact, more patients in the placebo group were demonstrating benefits.

She said the company attempted to proceed with the study after the June findings "to see if there was anything we could discover by lengthening the study or increasing the dose with patients already exposed to the drug," but two safety problems arose. One was the appearance of neutralizing antibodies in some of the patients, but the more serious concern was the possibility of cerebellum lesions.

Rothschild said preclinical studies showed the lesions appeared to be caused by GDNF.

"We're talking about irreversible brain damage," she told BioWorld Today, adding that the lesions were seen in preclinical models only, and not in any patients. "At the end of the day, we decided we really cannot, ethically, continue to provide the drug to patients in the trial without further understanding the toxicology or looking at other ways to deliver the drug."

Overall, the GDNF study was "a huge disappointment, given the safety concerns coupled with the absence of medical benefit," Rothschild said, especially given the "very promising" observations in earlier small open-label studies.

The Phase II study was designed based on results from a UK open-label study involving five patients over three years. All five demonstrated an improvement in symptoms as defined by UPDRS.

"This is a very challenging disease," Rothschild said. "There's a very long issue of placebo effect, but right now, we really can't look at anything else until the toxicology is clearly elucidated."

GDNF, which Amgen gained through its 1994 acquisition of Boulder, Colo.-based Synergen Inc., is a recombinant protein, a duplicate of a naturally occurring GDNF found in the central nervous system. Amgen's drug is designed to work by regenerating dopaminergic neurons that die in Parkinson's patients, and by attacking the pathological process of the disease.

Despite recent setbacks, the company said it still is committed to investigating GDNF as a potential therapy for Parkinson's.

"For our next step, we have initiated discussions with patient advocate organizations, such as the Michael J. Fox Foundation, and other research-funding agencies," Rothschild said, adding that Amgen is working to make GDNF available to preclinical researchers and will publish Phase II and toxicology data.

In addition to GDNF, Amgen has about 40 products in development. In its quarterly earnings report, the company said product sales totaled $2.2 billion for the three months ending Dec. 31. At the end of 2004, the company had about $5.8 billion in cash and marketable securities.

Shares of Amgen (NASDAQ: AMGN) rose 64 cents Friday to close at $63.42.