Idebenone treatment fails to slow cognitive decline in Alzheimer’s disease
Thal LJ, Grundman M, Berg J, Ernstrom K, Margolin R, Pfeiffer E, et al.;
Commented by , 28 Jan 2004
Aim of the study
To assess the effect of idebenone on the rate of decline in Alzheimer’s disease (AD) over one year in a placebo-controlled multicentre randomized clinical trial.
Method
Patients with probable AD ages 50 and above, severity range 12 to 25 on the Mini Mental State Examination (MMSE), were given placebo, 120, 240 or 360mg tid of idebenone. Cholinesterase inhibitors (AChEI) were not allowed. The primary outcome measures were cognitive (AD Rating Scale – cognitive subcomponent; ADAS-cog) and global (Clinical Global Impression of Change; CGIC); secondary outcomes included an activities of daily living scale, a behavior scale and the MMSE.
Results
536 subjects were enrolled and assigned to the four treatment groups, with no differences in characteristics at baseline between groups except for age. Drop out rates over one year were 26% for placebo, 30-32% on active drug. For the primary outcomes there was no statistical difference between the treatment groups in the pre-specified four-group design.
When combining all three active treatment groups to placebo, there was a difference in ADAS-cog in favor of idebenone in both the intent-to-treat (ITT) and the completers analysis. There was no difference in the secondary outcomes in either analysis.
Discussion
There were good reasons to believe that idebenone, an antioxidant structurally related to coenzyme Q and inhibiting lipid peroxidation through free radical scavenging activity, could delay progression of AD.
Other classes of compounds with epidemiological and/or biological plausibility for efficacy in delaying progression of AD failed in prospective randomized trials, including NSAIDS, COX-2 selective inhibitors and estrogens.
In this study the decline on ADAS-cog used for calculation of sample size was 7.3 and the actually observed decline on placebo was 6.3. In new studies AChEI must be allowed as standard treatment in mild to moderate AD, and the new drug under testing are added to stable doses of one of the AChEI.
This combination may modify the slope of ADAS-cog decline, making it harder to detect a placebo vs new drug difference at one year. One solution may be to extend studies to 18 months. Unfortunately this would increase the drop out rate, which was 26-32% in the current study (pre-study projection was 15%). This would also require a change in the primary analysis which would change from ITT to completers.
Another approach to detect efficacy would be to analyze slopes of decline using outcomes with demonstrated linearity over time, such as the ADAS-cog, some ADL scales such as the Disability Assessment in Dementia and global scales such as the Clinical Dementia Rating sum of boxes.
This would not be possible with a behavioral scale. These clinical efficacy measures would be supported by surrogate outcomes such as rate of whole brain atrophy on serial MRI. Other imaging-based measure of disease progression have been proposed and reviewed by Matthews, Siemers and Mozledy, Am J Geriatr Psychiatry 2003; 11: 146-159.
The Harmonization Work Group for Anti-dementia Drug Guidelines (HWGADG) will be putting together this year a consensus document to facilitate clinical research aiming at modifying disease progression. Interested persons can contact us at serge.gauthier@mcgill.ca.
(Editors note: See the trial report summary here.)