Acetylcholinesterase inhibitor drugs are unable to slow progression from mild cognitive impairment (MCI) to dementia, according to a paper in the November 27 PLoS Medicine. Reviewing available data, published and unpublished, from six different clinical trials, Roberto Raschetti and colleagues at the National Institute of Health, Rome, Italy, including Emiliano Albanese, Nicola Vanacore, and Marina Maggini, conclude that MCI patients taking the drugs succumb to dementia at about the same rate as those taking placebo. The results may rekindle the debate on whether MCI should be treated, and if so, how.

Acetylcholinesterase inhibitors (donepezil, galantamine, and rivastigmine) represent one of only two classes of drugs approved for treatment of Alzheimer disease, the other being the glutamate receptor antagonist memantine. The former give a boost to a flagging cholinergic system by preventing degradation of released acetylcholine, and as such they temporarily improve cognition in patients with mild to moderate AD. However, they are generally thought not to prevent the inexorable neurodegeneration and associated neurologic decline. Nonetheless, whether they are able to delay conversion from MCI to AD has been considered. While at least one study suggests that donepezil may slow somewhat the rate of conversion to AD (see ARF related news story), the data overall have not been very promising (see, for example, the review by Jelic et al., 2005). This new review supports the idea that this class of drugs has little effect on conversion from MCI to AD, or even on MCI symptoms.

Raschetti and colleagues set out to review all the available data on the use of acetylcholinesterase inhibitors (AChEIs) in MCI. They note that most of the trials testing the efficacy of donepezil, galantamine, and rivastigmine in MCI remain unpublished, even years after they have been completed. But by searching through clinical trial databases, they found usable data for three unpublished trials, two for galantamine and one for rivastigmine. They reviewed these data together with two published trials for donepezil and one for galantamine (see, respectively, Salloway et al., 2004; Petersen et al., 2005; Koontz et al., 2005). The authors took care to ensure that the trials were similar with respect to design, objective, and definition of MCI, which can vary greatly. They found that conversion to AD rates, available for the three unpublished trials and also from the NIA/ADCS Memory Impairment Study (Petersen et al., 2005), are slightly lower in those taking the inhibitors, but write that even so, “no significant differences emerged in the probability of conversion between treated groups and the placebo groups.” It should be noted, that Raschettti and colleagues focused on the 3-year data from the Memory Impairment Study, while Ron Petersen and colleagues had originally reported that there is some delay in conversion over shorter time frames (see ARF related news story). On that point, Raschetti and colleagues write that if donepezil can delay conversion over 12 months but has no effect over 3 years, then logically it must be assumed that it also accelerates conversion over years 2 and 3.

The authors also examined the effects of the drugs on secondary endpoints, including cognition, global, activity, and neuropsychiatric scales, but conclude that the drugs have no significant effect. The only difference between drug and placebo groups that cannot be attributed to chance alone seems to be whole brain atrophy. In one of the galantamine trials (Gal-INT-11), patients on the cholinesterase inhibitor had less brain volume loss, though it is not clear at this point if that has any functional significance.

This review questions not only the value of treating MCI with cholinesterase inhibitors, but the value of the MCI diagnosis itself. There have been attempts to standardize this diagnosis (see, for example, Matthews et al., 2007 and related comments), but different clinics still use different tests and criteria to diagnose MCI. Raschetti and colleagues point out that if MCI criteria used in the Gal-INT-11, InDDEX rivastigmine, and MIS trials are applied to the same cohort of 150 people, they predict a positive diagnosis in 51, 21, and 17 percent of the cohort, respectively (see Visser et al., 2005). Such a disparity suggests that perhaps it is not that cholinesterase inhibitors do not help, but that some of the patients being tested do not have MCI. “Thus, further clinical trials are needed to clarify whether cholinesterase inhibitors can delay the progression of MCI to dementia, but these additional trials should not be done until the diagnosis of MCI has been standardized,” suggest the authors.—Tom Fagan


  1. A systematic review collects all relevant studies on a subject, integrates findings, and allows readers to do their own assessments based on all available evidence. Inferences from systematic reviews stand in sharp contrast to picking and choosing individual studies from which to make decisions such as treatment.

    Mild cognitive impairment (MCI) has been controversial both in terms of clinical validity and treatment. Many physicians choose to treat patients they diagnose with MCI with cholinesterase inhibitors.

    Raschetti and colleagues did a systematic review of randomized, placebo-controlled MCI trials of the marketed cholinesterase inhibitors. They identified eight such trials, including two in which data were unavailable. The authors neatly summarized characteristics and outcomes of the trials and made a number of observations:

    1. Although the trials used the Petersen or Mayo Clinic criteria for MCI, each trial operationalized the criteria differently.

    2. The outcomes and scales used, especially cognitive scales, varied widely beyond the omnipresent the ADAS-cog. The reviewers arranged these in a matrix by MCI trial and showed a forest plot of the outcomes of the trials.

    3. Their most remarkable observation is the distinct lack of evidence for cognitive, ADL, behavior, or global effects on any of the measures over the 2- to 4-year courses of the trials. The one exception was a significant effect on the 13-item ADAS-cog in favor of donepezil in the one short-term, 6-month long trial. (It is notable that the ADCS donepezil and vitamin E trial showed a similar effect at 6 months that was lost after 18 months).

    The reviewers discuss the thorny issue of making inferences about treatment when the condition being treated, MCI in this case, is so heterogeneous. They cite important work by Visser and colleagues, who applied the protocol-specified MCI criteria used by Janssen, Novartis, and the ADCS to their clinic patients and found differing prevalences of 51 percent, 21 percent, and 17 percent, depending on the criteria.

    Finally, they voiced the ethical specter that the uncertainty of currently applied MCI diagnostic criteria prompts questions about the scientific validity and ethical value of the trials. They quote from Jason Karlawish’s paper that the “logic of clinical purpose” of clinical trials is they be “logically grounded in, and ethically justified by, the way they reflect and contribute in clinical practice.” The authors opine that MCI may be an example of a “risk factor conceptualized as a clinical condition,” and caution that “when there is controversy surrounding the definition of a condition or disease, even inconclusive results from RCTs can be used to suggest treatment.” They estimate that in Italy, 27 percent of patients diagnosed with MCI are prescribed cholinesterase inhibitors, and needless to say, they don’t think this is a good thing.


    . Alzheimer's disease--clinical trials and the logic of clinical purpose. N Engl J Med. 2006 Oct 12;355(15):1604-6. PubMed.

    . Vitamin E and donepezil for the treatment of mild cognitive impairment. N Engl J Med. 2005 Jun 9;352(23):2379-88. Epub 2005 Apr 13 PubMed.

    . Cholinesterase inhibitors in mild cognitive impairment: a systematic review of randomised trials. PLoS Med. 2007 Nov 27;4(11):e338. PubMed.

    . Do MCI criteria in drug trials accurately identify subjects with predementia Alzheimer's disease?. J Neurol Neurosurg Psychiatry. 2005 Oct;76(10):1348-54. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Early Intervention Trial Bears Little Fruit, but Sows Hope

Paper Citations

  1. . Clinical trials in mild cognitive impairment: lessons for the future. J Neurol Neurosurg Psychiatry. 2006 Apr;77(4):429-38. PubMed.
  2. . Efficacy of donepezil in mild cognitive impairment: a randomized placebo-controlled trial. Neurology. 2004 Aug 24;63(4):651-7. PubMed.
  3. . Vitamin E and donepezil for the treatment of mild cognitive impairment. N Engl J Med. 2005 Jun 9;352(23):2379-88. Epub 2005 Apr 13 PubMed.
  4. . Effects of galantamine on working memory and global functioning in patients with mild cognitive impairment: a double-blind placebo-controlled study. Am J Alzheimers Dis Other Demen. 2005 Sep-Oct;20(5):295-302. PubMed.
  5. . Operationalization of mild cognitive impairment: a graphical approach. PLoS Med. 2007 Oct;4(10):1615-9. PubMed.
  6. . Do MCI criteria in drug trials accurately identify subjects with predementia Alzheimer's disease?. J Neurol Neurosurg Psychiatry. 2005 Oct;76(10):1348-54. PubMed.

Further Reading

Primary Papers

  1. . Cholinesterase inhibitors in mild cognitive impairment: a systematic review of randomised trials. PLoS Med. 2007 Nov 27;4(11):e338. PubMed.