Longer and larger studies may be needed to offset concerns that placebo groups in clinical trials of potential therapies for Alzheimer’s disease are not showing enough of a cognitive decline to make meaningful comparisons of effectiveness, two new studies suggest.

The issue was addressed in presentations at the Alzheimer’s Association International Conference on Alzheimer’s Disease in Chicago, US this week. As the association notes, there is an emerging perception that patients with Alzheimer’s in the placebo arms of current studies are showing less deterioration over the time period given.

This may, it has been speculated, be due to inadequate sensitivity of the standard scales used for measuring cognition in the trials; differences in disease severity and co-existing medical conditions between the trial populations now being recruited and participants in previous Alzheimer’s trials; or the relatively common use of approved Alzheimer’s drugs (cholinesterase inhibitors and memantime) among the populations involved in current trials.

In one of the studies presented at ICAD 2008, a team led by Dr Lon Schneider, professor of psychiatry and neurology at the University of Southern California Keck School of Medicine, analysed data from 87 randomised, double-blind, placebo-controlled Alzheimer’s trials carried out between 1991 and 2005.

Encouragingly, they found no overall changes in the degree of cognitive decline across the 15-year time period. However, smaller placebo groups were associated with less likelihood of cognitive deterioration over the course of the trial.

Specifically, sample sizes of fewer than 100 had a 37% chance of not showing any significant change during the trial, while placebo groups with more than 200 patients all displayed significant deterioration at six months. In 12-month studies, 95% of placebo groups with sample sizes of more than 100 demonstrated significant cognitive decline.

Among other determining factors were the location of the trial sites – non-English speaking sites were associated with less cognitive worsening over the course of the trial – and more frequent cognitive assessments, which reduced variance in the amount of decline. The existing use of cholinesterase inhibitors was not associated with any less cognitive deterioration in the trials reviewed.

No need to worry

“Experts have no need to worry that people with Alzheimer’s in clinical trials are less likely to worsen than they have before, or that cholinesterase inhibitors are lessening decline of placebo-treated patients in recent trials,” Schneider said. “Based on our analysis, the most reliable approach to maximize the likelihood for demonstrating efficacy is to have a placebo group size greater than 200 and to use the ADAS-cog [Alzheimer’s Disease Assessment Scale cognitive subscale] at least four times and in English.”

The second presentation was of a meta-analysis led by Professor Roy Jones, director of the Research Institute for the Care of Older People at the Royal United Hospital in Bath, UK.

Building on discussions and analyses from an expert working initiated and funded by Eisai and Pfizer, which manufacture and market the Alzheimer’s disease therapy Aricept (donepezil), Jones and colleagues looked at individual patient data from randomised, double-blind, placebo-controlled studies of Aricept between 1990 and 1999.

Data on 3,403 patients from 13 Alzheimer studies were grouped according to the year the trials started. Group 1 constituted studies initiated in 1990-1994 and Group 2 studies initiated in 1996-1999. Changes from baseline Mini Mental State Examination (another standard measure of cognition) and ADAS-cog scores up to week 24 were compared between Groups 1 and 2 for placebo arms only, then between donepezil and placebo.

Later, slower

The decline on the MMSE from baseline to week 24 was significantly greater among the placebo patients in Group 1 (-1.28 points) than in Group 2 (-0.56 points; p = 0.024). Placebo decline on the ADAS-cog score was also more pronounced in Group 1 than in Group 2, although here the difference was non-significant.

“Our results indicate that patients with Alzheimer’s entering the later clinical trials appear to be experiencing a slower rate of decline in memory and thinking processes,” Jones commented.

“These observations are potentially important for the future design of clinical trials in people with Alzheimer’s. For example, it may be necessary to conduct longer duration research studies – more than 24 weeks – to ensure any effects of treatment can be fully evaluated.”