After six months of number crunching, Neurochem Inc. confirmed what most analysts already suspected: the North American Phase III trial of Alzhemed (tramiprosate) in Alzheimer's disease did not meet its primary endpoints.

Neurochem completed the trial in February, but results were repeatedly delayed as the Laval, Quebec-based company made adjustments to its statistical models. Last month, analysts Brian Lian of CIBC World Markets Corp. and Jonathan Aschoff of Brean Murray, Carret & Co. LLC both told BioWorld Today that the trial had likely failed.

Top-line data released on Sunday showed the 18-month trial did not meet its primary endpoints in disease modification as measured by magnetic resonance imaging or in clinical efficacy as measured by the Alzheimer's Disease Assessment Scale, cognitive subpart (ADAS-cog) and the Dementia Rating, sum of boxes rating scale (CDR-SB). The trial had randomized 1,052 mild-to-moderate Alzheimer's patients to receive placebo, 100 mg of Alzhemed, or 150 mg of Alzhemed along with stable doses of their conventional treatments.

In a conference call, Neurochem's Senior Vice President of Drug Development Denis Garceau, called the data "inconclusive" but pointed to a number of positive trends. He said the ADAS-cog and CDR-SB numbers favored Alzhemed, though not with statistical significance, and the data suggested a pharmacological effect on spinal fluid levels of amyloid-beta and Tao biomarkers. Additionally, the difference in hippocampal volume, which has been linked to disease improvement, approached statistical significance.

Neurochem also said Alzhemed was generally safe and well tolerated, with the most common adverse events being nausea, falling, diarrhea, dizziness, body weight decrease and vomiting. However, the company did not provide specific data regarding safety and efficacy and said it is not possible to draw definitive conclusions regarding the treatment effect due to a number of complicating factors. Those included high variations between clinical sites, a large number of covariates identified during the modeling process, the inherently variable rate of disease progression in Alzheimer's, and the effects of concomitant medications.

During the call, principal investigator Paul Aisen of the Georgetown University Medical Center noted that this was the first long-term clinical trial of a disease-modifying agent for Alzheimer's and that the "issues faced by this trial will be shared by other trials aiming to demonstrate disease modification."

Existing Alzheimer's drugs treat the symptoms of the disease rather than modifying its course. These drugs include Aricept (donepezil, Pfizer Inc.), Exelon (rivastigmine, Novartis AG), Cognex (tacrine, First Horizon Pharmaceutical Corp.), Reminyl (galantamine, Shire Pharmaceuticals Group plc) and Namenda (memantine, Forest Laboratories Inc.).

Alzhemed is a small molecule amyloid B antagonist designed to interfere with amyloid deposition in the brain, a primary characteristic of Alzheimer's. Other disease-modifying Alzheimer's drugs in development include Flurizan (tarenflurbil, from Myriad Genetics Inc.), a selective amyloid lowering agent in Phase III, and bapineuzumab (Elan Corp. plc and Wyeth), a monoclonal antibody slated to begin Phase III later this year.

The FDA advised Neurochem in a recent meeting that no adjustment of statistical models would allow its current data to be used to support a claim of efficacy. However, the agency was open to Neurochem adjusting its ongoing European Phase III trial to increase the likelihood of success based on knowledge gained in the North American trial.

To that end, Neurochem's Chairman, President and CEO, Francesco Bellini, said the company is establishing a special advisory board to fully analyze the data and provide recommendations regarding next steps. Bellini said the board will attempt to address important questions, such as "why more than 50 percent of patients on control groups have not declined, and why some patients in the control group were declining for many months and suddenly started to improve." The board's report is expected by the end of the year.

Shares of Neurochem (NASDAQ:NRMX) fell $2.40, or 43 percent, to close at a new 52-week low of $3.16 on Monday.

Neurochem's troubles do not stop with Alzhemed: just last month the company received its second approvable letter for Kiacta (eprodisate) in the treatment of AA amyloidosis. Although a complete response filing is anticipated next month, analysts are skeptical that the FDA will approve the drug without an additional Phase III trial. Kiacta is also under review in Europe. (See BioWorld Today, July 19, 2007.)

In fact, Eun Yang of Jefferies & Co. Inc. dropped coverage of Neurochem last week, even before the Alzhemed news, citing a "lack of confidence" in the eventual success of either Alzhemed or Kiacta. Beyond those two products, Neurochem has a Phase II product for hemorrhagic stroke due to cerebral amyloid angiopathy sitting on the back burner due to a lack of resources.

Neurochem reported cash, cash equivalents and marketable securities of $90.8 million at the end of the second quarter, which Bellini said will last through the year.