Another Alzheimer’s Drug Fails in Large-Scale Trials

Flurizan's maker says it will discontinue work on the drug

by Jim Schnabel

June 30, 2008

Flurizan, a drug that reduced the production of apparently harmful amyloid in laboratory and animal experiments, has failed to show a significant benefit in a large-scale, “Phase III” trial in about 1,700 people with Alzheimer’s disease. Its maker, Myriad Pharmaceuticals of Salt Lake City, announced the results today and indicated that it would discontinue development of the drug.

The results were among the most eagerly awaited of all clinical trial results for Alzheimer’s drugs this year, since no truly effective treatments are yet available for the disease. Flurizan’s failure also will have come as a surprise to some, since the drug, according to Myriad, had shown signs of being moderately effective in some patients in a smaller-scale Phase II clinical trial.

That smaller trial, following approximately 200 people with early-stage Alzheimer’s over 12 months, initially had indicated no significant effect for Flurizan for its primary, predefined endpoint, the slowing of cognitive decline. But in what has become a common practice in these high-stakes drug-development efforts, Myriad performed a “post-hoc analysis” and declared that a subset of a few dozen patients with early-stage, mild Alzheimer’s, who had taken Flurizan for an extended period of 24 months, showed significant improvement. Largely on that basis the Phase III study went forward.

Why did it fail?

Flurizan (generic name tarenflurbil), is an “enantiomer,” or mirror-image molecule, of the non-steroidal anti-inflammatory drug flurbiprofen. It was selected for drug development because, like flurbiprofen and ibuprofen, it appears to slow the production of amyloid-beta-42 proteins in laboratory and animal tests; but because of its mirror-image structure it lacks most of the other effects of those anti-inflammatory drugs, including most of their adverse side effects.

There are at least several possible explanations for why Flurizan reduced amyloid production in preclinical tests but still failed to benefit humans with early-stage Alzheimer’s. First, at the doses used it may simply have failed to affect amyloid production in patients’ brains.

“One cannot draw any conclusions about the role of amyloid in Alzheimer’s unless there are direct data to show whether amyloid burden increased, decreased or remained unchanged during the trial,” says Sam Gandy, an Alzheimer’s researcher and clinician at Mount Sinai School of Medicine. In practice, data on amyloid pathology are impossible to obtain until a patient has died; if Myriad has obtained such data from treated and control-group patients who died during the study, the company has not released them.

Another possible explanation for the failure is that the patients in the trial, though considered to have early-stage Alzheimer’s, were actually treated relatively late in their disease courses. Alzheimer’s-related processes in the brain, including amyloid deposition, are now believed to start years, perhaps even decades, before the first outward signs of disease are noticed, and by the time those signs appear multiple mechanisms of neuronal destruction may be at work, so that stopping one of them—for example amyloid—may not be enough.

“Initiation of therapy even in humans with mild cognitive impairment may still be too late for anti-amyloid intervention to be effective,” says Gandy. The practice of testing drugs in mice prior to the accumulation of amyloid, he adds, “may be giving us false positives in terms of which drugs to take to the clinic.”

The Flurizan failure may also be relevant to a recent shift in the “amyloid hypothesis” of Alzheimer’s. In the past several years, evidence has mounted that amyloid-beta-42, long considered the culprit in the disease, affects memory-related functions only when it has formed multi-protein conglomerations called “oligomers.” In the light of this concept, it is possible that Flurizan affects amyloid-beta-42 production in the brain and reduces the formation of insoluble amyloid deposits but has little or no effect on amyloid oligomer levels. As Gandy points out, “We have as yet no useful biomarker for brain oligomer levels.”

Of course, it is also possible that in humans, neither form of amyloid is particularly relevant to the ongoing disease process after diagnosis. As neuropathologist Mark Smith of Case Western Reserve University, a long-time critic of the amyloid hypothesis, notes, “We still don’t know which is the toxic species” in Alzheimer’s.