Tramiprosate Falls Short in Phase III Alzheimer's Trial : Unusually large placebo effect could be a recurring problem in studies that allow concomitant medications.


Tramiprosate, the first antiamyloid drug to enter a phase III trial, was not significantly better than placebo in improving cognitive function in patients with Alzheimer's disease, according to officials of Neurochem Inc., manufacturer of the investigational agent and sponsor of the North American trial.

The negative results are a blow to Alzheimer's disease (AD) researchers and patient advocacy groups, said Dr. Richard J. Caselli, chair of neurology at the Mayo Clinic, Scottsdale, Ariz. “Tramiprosate was the first antiamyloid drug to reach this point, and as such, was widely watched by the Alzheimer's disease community,” Dr. Caselli said in an interview. “Its failure to achieve its therapeutic outcomes is therefore very disappointing.”

The question now is whether this failure poses a serious challenge to the amyloid hypothesis of AD pathogenesis, he said. “Possibly not, as findings from a newly released study suggest that in addition to its antiamyloid effects, tramiprosate may have a competing effect favoring tau aggregation. It remains too early to heavily discount the amyloid hypothesis, and other trials in progress will be watched expectantly.”

The North American Phase III study included 1,052 patients with mild to moderate AD, recruited from 67 sites in Canada and the United States. Patients were randomized to placebo or 100 mg or 150 mg twice daily of tramiprosate. They continued all their concomitant AD drugs during the 18-month study period.

Although there were numerical differences in favor of tramiprosate, those differences failed to reach statistical significance in any of the three primary end points: the Alzheimer's Disease Assessment Scale (ADAS-Cog), the Dementia Rating-Sum of Boxes rating scale (CDR-SB), or magnetic resonance imaging. The MRI analysis showed a trend toward larger hippocampal volume in the active group, although the investigators have not assessed that finding's possible relationship to cognitive performance, Dr. Francesco Bellini, Neurochem's president and chief executive officer, said during the teleconference during which the data were released.

Dr. Bellini hesitated to describe the study as negative, pointing out that the statistical analysis was complicated by improvement in more than 30% of the control patients. “This complicated the analysis beyond expectation, so that our results are inconclusive,” Dr. Bellini said.

Dr. Paul S. Aisen, principal investigator of the North American trial, noted the unexpected improvement of so many control patients is probably a result of the effect of concomitant medications, and will be a recurring problem in all long-term studies of disease-modifying agents for AD.

“This problem will not be unique to this program,” said Dr. Aisen, professor of neurology at Georgetown University Medical Center, Washington. “It will be faced by anyone who tries to conduct these trials. A number of the approved medications have significant effects on the primary outcomes, and during 18-month trials, these effects will be unavoidable.”

Dr. Caselli noted drug companies need to factor this into their trial designs in light of the reality that patients will not be giving up their approved medications while taking a study drug. “We should not infer that study limitations imply the study 'was not really negative, even though it appeared to be,'” he said. “Hopefully the European trial [of tramiprosate] will show something different, but after all is said and done, if both fail and we still encounter the 'Can't tell because they're on treatment' argument, then the study design needs to be changed to accommodate this. But the study was negative as performed.”

Dr. Marwan Sabbagh, director of clinical research at the Sun Health Research Institute in Sun City, Ariz., noted the real difficulty with the tramiprosate trial design centers on the lack of objective clinical end points.

“That is the peril of using cognitive outcomes of primary measures,” he said in an interview. “Many in the industry would like to see other measures used, but there are none validated or universally agreed upon.”

Future studies will need to give equal weight to specific biomarker outcomes, he said, including PET scans, MRI volumetry, and cerebrospinal fluid. “This negative study will make a lot of companies consider their development strategies more carefully.”

Neurochem's Dr. Denis Garceau said the company delayed the release of its findings, which were to have been presented publicly in June. In the meantime, Neurochem reworked the statistical analysis and sought advice from the Food and Drug Administration, said Dr. Garceau, who is senior vice president of drug development. “While recognizing the challenges of a trial of this magnitude, the FDA advised that neither the proposed adjusted models nor any further adjustments could be used for this trial to support a positive effect of tramiprosate,” Dr. Garceau said.

Next Article: