What now for Alzheimer's disease?

Editor, – The review of the AD2000 trial (Aust Prescr 2005;28:134-5) fails to note that this trial has been heavily criticised. It permitted enrolment of people with cerebrovascular disease, enrolled less than 20% of its recruitment target and carried on with too few patients for too short a time to tell whether the drugs delayed institutionalisation. There was a complicated double randomisation method with an extra four-week washout period every 12 months. Of 566 people entering the study only 111 completed two years of the trial and only 20 completed the third year of a planned five-year study. Many prominent researchers in the UK chose not to be involved because of the questionable ethics of offering treatment only as part of a randomised controlled trial. The researchers skirted this ethical dilemma by asking doctors to recruit only patients about whom they were 'substantially uncertain that the individual would gain a worthwhile clinical benefit from donepezil'! About the only conclusion that can be drawn from this study is that donepezil produces a measurable but small improvement in a crude cognitive measure which is sustained in individuals receiving treatment compared to those receiving placebo over at least one and possibly two years.

The review contains a footnote saying that the results of a recent study were 'very similar to those of the AD2000 trial'. This is misleading. The recent trial assessed the usefulness of donepezil and vitamin E for a completely different indication (mild cognitive impairment, not Alzheimer's disease) and returned negative, not weakly positive, results on measures of cognition.¹

Cholinesterase inhibitors have modest efficacy for some patients with Alzheimer's disease, but it is not possible to tell in advance who will respond. It is therefore appropriate to offer people with mild to moderate Alzheimer's disease a trial of treatment, monitor their response and then decide about continuation. The requirement for at least a 2-point improvement in the mini-mental state examination goes some way towards ensuring that the patients who receive continuing treatment will be those who have shown some response.

David Ames
Professor of Psychiatry of Old Age
University of Melbourne
Melbourne

Henry Brodaty
Professor, Academic Department for Old Age Psychiatry
University of New South Wales
Sydney

Professor Ames and Professor Brodaty have both received research support, honoraria and financial assistance to attend conferences from companies marketing cholinesterase inhibitors.

Professor J. Attia and Professor P. Schofield, authors of the editorial, comment:

In our editorial, we clearly acknowledged the drawbacks of the trial, including the low recruitment and the complex design. Despite the contention that the trial was too short, it was the only trial up to that point to have looked at outcomes beyond one year. Despite criticism of the inclusion criteria, even Ames and Brodaty acknowledge the difficulty of prospectively identifying responders. We would contend that the study has some strengths, including the focus on clinical end points, caregiver burden, and economic analyses. It tempers the enthusiasm generated by the results from short-term, largely drug company-sponsored studies and this cautionary note has been sounded by others.²

The recent study examined the effect of donepezil and vitamin E on conversion rates from incipient to diagnosed Alzheimer's disease, and thus was concerned with the same disease entity as was AD2000, albeit at a milder stage. The similarity that we see between the two three-year trials is that they both indicated a small beneficial effect in primary outcomes at 6-12 months, which was not sustained in the long term.

However, 'evidence alone is never sufficient to make a clinical decision'.³ The translation of evidence into practice is subject to an evaluation of the risks and benefits, costs, patient values and circumstances. We agree with Ames and Brodaty that an N of 1 trial is the highest level of evidence to apply!