( BROOMFIELD Colo. May 01 2007 /PRNew...)Accera, Inc. Announces Results of Phase II Study in Alzheimer's,Disease at American Academy of Neurology Meeting,Health

BROOMFIELD, Colo., May 01, 2007 /PRNewswire/ -- Accera Inc. is presenting topline data today from its Phase IIb study of its lead compound AC-1202 in Alzheimer's disease (AD) at the American Academy of Neurology (AAN) 59th Annual Meeting in Boston. Judged by the AAN to be in the top five percent of the program, the data will also be featured in the Scientific Highlights Plenary Session.

The randomized, double-blinded, placebo-controlled Phase IIb trial evaluated 152 subjects that had previously been diagnosed with mild to moderate AD. Consistent with the findings of Accera's Phase IIa study, subjects who did not have the ApoE4 genotype-a known genetic risk factor that occurs in half of all AD patients-responded particularly well to treatment, as reflected in statistically significant improvement in AD Assessment Scale- Cognitive (ADAS-Cog) scores.

Subjects underwent pharmacogenomic analysis for a variety of genetic markers and were evaluated through a battery of neuropsychometric tests at Day 0, 45, and 90. Compared to the placebo group, the ADAS-Cog scores of the AC- 1202-treated ApoE4(-) population improved 3.5 points in twelve weeks (p=0.01), and statistically significant improvement was seen in just 45 days.

Interestingly, ApoE4(-) subjects who also exhibited a genetic variation that affects glucose regulation showed a 5 point improvement in ADAS-Cog scores compared to placebo, providing further insight into the disease. "The profound effect we see in the population without the ApoE4 risk factor supports the findings of an earlier study linking efficacy to a certain pharmacogenomic profile," said Dr. Lauren Costantini, Accera's vice president of clinical development. "It also provides further evidence of the link between Alzheimer's disease and glucose metabolism."

Regardless of genotype, subjects treated with AC-1202 showed a trend toward improvement (p=.072), suggesting the com
'"/>