Pharmacotherapy for Mild Cognitive Impairment: Are We There Yet?

MCI and AD
When Mrs Keyes and her children investigate MCI in the library and on the Internet, their concern may have been heightened by learning that amnestic MCI is considered a possible precursor to AD. This link is based on the compelling epidemiological evidence that persons with amnestic MCI convert to AD at an accelerated rate compared with the general elderly population.⁴ Furthermore, the connection between amnestic MCI and AD is bolstered by findings in patients with amnestic MCI of increased hippocampal atrophy,⁵ specific abnormal positron emission tomography scan findings,⁶ and AD-like neuropathological changes on autopsy.⁷ A recently published study indicated that even the minimal deficit that is termed "very mild cognitive impairment" is associated with an increased rate of conversion to AD, and this finding is sure to fan the flames of public and clinical anxiety about the prognostic implications of "senior moments."⁸

On the basis of this accumulating data, MCI has increasingly seized public and professional interest. We are still at the beginning of sorting out the implications of an MCI diagnosis, however, and there is no standardized approach to treatment. Vascular risk factors and deleterious effects of medications and substances should be controlled. Some evidence supports the value of a healthful diet and appropriate levels of physical exercise⁹ and of cognitive rehabilitation.¹⁰ Pharmacotherapy for patients with MCI, although common in some settings, is undertaken on the basis of limited data that offer minimal encouragement. To answer the Keyes' question about pharmacotherapy, let's review what we can learn from available studies.

Treatment studies
MCI has been treated "by analogy" based on interventions designed for patients with AD or other dementias. Medications that have some efficacy in treating AD's cognitive impairments, in particular, have been explored. The most impressive study of a cholinesterase inhibitor in patients with MCI is that of Petersen and colleagues,¹¹ who compared the effects of vitamin E(Drug information on vitamin e) (2000 IU/d) or donepezil(Drug information on donepezil) (Aricept; target dosage of 10 mg/d) with placebo in patients with amnestic MCI over a 3-year double-blind trial.

The primary outcome measure, conversion to AD, was diminished during the initial 12 months of this study. Converter rates are not specifically reported, but a visual inspection of the study's Figure 1 suggests that the number needed to treat associated with prevention of conversion at 12 months must be greater than 10, which means that at least 10 patients would need donepezil treatment for one to achieve a benefit unexplainable by placebo effect. By the end of 3 years, conversion rates were similar for donepezil, vitamin E, and placebo except for a reduced rate of conversion among the donepezil-treated group of patients with MCI who carried 1 or 2 apolipoprotein E (ApoE) e4 alleles. The result of this secondary analysis has been attributed to the greater statistical power conferred by the higher rate of conversion to AD among ApoE e4 patients and has not been generally suggested to justify routine testing for ApoE genotype or routine prescribing of donepezil for ApoE e4-positive persons with MCI. Furthermore, 36% of patients who are treated with donepezil discontinued the medication because of adverse effects that included GI symptoms, muscle cramps, insomnia, and abnormal dreams.

Enthusiasm for treating MCI with cholinesterase inhibitors has received no stronger support from several other published trials. An earlier placebo-controlled donepezil trial in MCI participants found no significant effect on primary outcome measures in an intention-to-treat analysis of results from a 24-week trial, although some secondary outcomes were significant.¹² A 2-year, randomized, placebo-controlled trial failed to show an effect of rivastigmine(Drug information on rivastigmine) (Exelon) on conversion from MCI to dementia.¹³ Galantamine (Razadyne) in 2 trials was associated with some improvement on secondary measures of cognition but not with any effect on the conversion rate to dementia at 24 months.¹⁴ An unexpectedly high galantamine(Drug information on galantamine)-associated death rate (or perhaps an unexpectedly low placebo-associated death rate) resulted in concern about the balance of galantamine's risks and benefits for patients with MCI.