Memantine is one of only four commonly prescribed drugs for the treatment of memory loss in Alzheimer’s disease. The FDA approved the drug for moderate to severe AD in 2003, but since then many physicians have taken to prescribing it off-label for people at mild stages of the disease as well. Does memantine (trade name Namenda) help in the earliest stages of dementia? A comprehensive meta-analysis just put paid to that idea. In the April 11 Archives of Neurology online, researchers led by Lon Schneider at the University of Southern California, Los Angeles, and colleagues Rupert McShane at the University of Oxford and Julian Higgins at the Medical Research Council, Cambridge, both in the U.K., report that memantine is no better than placebo for people with mild AD, and only marginally effective in moderate cases.

The study is important for two reasons, Schneider told ARF. It counters prior claims that memantine, a glutamate receptor antagonist, works “across the spectrum of the disease, from mild to severe cases.” And it provides physicians and patients with independent information to make informed decisions. Many people with mild AD are taking the drug; for example, 45 percent of volunteers enrolled in the Alzheimer’s Disease Neuroimaging Initiative and 25 percent of those registered with the National Alzheimer’s Coordinating Center at the National Institutes of Health (Schneider et al., 2011) are on it. Many patients ask for memantine, and physicians can feel compelled to prescribe it. “We are not in the business of crushing hope [by not prescribing it off-label or for patients who ask],” said Schneider, “but this study provides a reality check.”

The drug maker would not provide raw data for the researchers to analyze the effects of memantine on people with mild AD, but Schneider and colleagues were able to use analyses published by the sponsors to tease out the necessary data.

Sponsors in the U.S. (Forest Laboratories, Inc.) and Europe (Merz Pharma) used the same three clinical trials of people with mild to moderate AD (Mini-Mental State Exam scores of 10-23) to argue for an expansion of approval to mild AD. In July 2005, the U.S. Food and Drug Administration rejected that application. In November 2005, the European Medicines Agency offered limited expansion of their approval from “moderately severe to severe,” to “moderate to severe” AD, but also declined approval for mild forms. One meta-analysis by the European sponsor had focused on patients only with moderate to severe AD (MMSE less than 20), which allowed Schneider and colleagues to obtain data on mild AD patients (MMSE 20-23) by subtraction. On analysis, they found no significant difference between the drug and placebo groups on a variety of outcomes, including cognitive, activities of daily living, psychiatric, and global measures. People with moderate AD who took memantine did marginally better on cognitive and global measures.

The study speaks to the thorny issue of healthcare management. Who should decide whether someone gets to take a drug of limited benefit? It may depend on who is paying for it. In the U.K., there was a furor in 2005 when the National Institute for Clinical Excellence (NICE), which sets guidance for the National Health Service, shot down cholinesterase inhibitors for AD because they deemed them of little benefit over the long term. NICE recanted a year later, officially recommending cholinesterase inhibitors for patients with moderate disease. Last October, NICE had another change of heart, extending their recommendation to mild AD as well (see related news from the BBC). NICE does not, however, recommend memantine for mild AD. The U.S. does not have a national healthcare system, though taxpayers subsidize the cost of memantine through Medicare prescription drug coverage. Memantine typically costs $170 a month. Medicare can cover up to 95 percent.—Tom Fagan


  1. Schneider et al. present a reanalysis of three studies of memantine in mild to moderate Alzheimer’s disease (AD) in which they were able to isolate the data for the mild subjects. They find no benefits for memantine in milder AD dementia patients. Schneider and I had drawn the same conclusion from the industry-sponsored meta-analysis of the same studies published in 2007 (Knopman, 2007; Schneider, 2007).

    What is particularly disturbing to me is another finding of detective Schneider: that a large fraction of patients with mild AD in the research projects of the Alzheimer’s Disease Neuroimaging Initiative or the Alzheimer’s Disease Centers’ programs are prescribed memantine. The widespread use of memantine in mild AD and even in people with mild cognitive impairment (MCI) points out how desperate patients, families, and treating physicians are to employ a medication for these disorders, even when the data show the drug has no benefits. As Schneider et al. illustrate in their timeline, there were long gaps between the completion of the mild to moderate AD trials with memantine and their eventual publication. And, the reports of the memantine trials, for whatever reason, were published in journals that are far out of the mainstream, including one to which my institution didn’t even subscribe. Physicians bear responsibility for ignoring the lack of demonstrated efficacy of memantine in mild AD dementia patients.

    I am acutely aware of the frustration of families, patients, and physicians when it comes to therapy for mild AD dementia. I have to face patients and families with dementia or MCI and tell them how bleak the therapeutic landscape is. But why waste resources and expose patients to unnecessary medications? Physicians interested in the best care of AD patients should speak out against the against the wasteful use of memantine in mild AD dementia, and I applaud Schneider et al. for calling attention to this embarrassment.


    . Commentary on "Meta-analysis of six-month memantine trials in Alzheimer's disease." Memantine has negligible benefits in mild to moderate Alzheimer's disease. Alzheimers Dement. 2007 Jan;3(1):21-2. PubMed.

    . Commentary on "Meta-analysis of six-month memantine trials in Alzheimer's disease." Wuthering forest plots: distinguishing the forest from the plots. Alzheimers Dement. 2007 Jan;3(1):18-20. PubMed.

    View all comments by David Knopman
  2. Memantine has been approved by the U.S. Food and Drug Administration as a safe and effective treatment for moderate to severe Alzheimer’s disease based upon evidence from randomized, double-blind, placebo-controlled trials (1,2). The recent meta-analysis (3) by Schneider et al. is an attempt to use post hoc subpopulations from previous clinical trials to examine the potential benefits and risks of memantine in patients with mild AD. In a previously published meta-analysis (4), we demonstrated that memantine was associated with significant cognitive and global benefits compared to placebo in patients with mild to severe Alzheimer’s disease. This meta-analysis included six randomized, placebo-controlled trials of memantine: three in mild to moderate patients and three in moderate to severe patients. The moderate ranges of the two groups obviously overlapped, although each study recruited patients independently. It is worth noting that the mild to moderate trials analyzed in our previous meta-analysis are the same ones that were re-analyzed by Schneider et al. In our analysis of the mild to moderate studies, patients treated with memantine outperformed those given placebo on tests of both cognitive and global function. Furthermore, these effects were homogeneous, or consistent across the individual included studies, which heightened our confidence in these findings. Because no memantine trials recruited and randomized only mild AD patients, we did not investigate this subgroup alone.

    Schneider et al. defined mild AD patients as those with MMSE scores ranging from 20 to 23; in comparison, moderate AD patients had MMSE scores from 10 to 19. Clearly, the range of MMSE scores in the mild AD sample was quite limited; this restriction of MMSE range probably resulted in a subsequent restriction of the range on other cognitive (e.g., ADAS-cog) and functional measures, thereby reducing the researchers’ ability to detect differences between changes over time in the drug and placebo groups. There was also a relatively small number of patients in the mild AD category; for example, on the ADAS-cog measure, across the three studies, the average number of patients per treatment group per study was 71; in the moderate category, the average number of patients per treatment group per study was 114. It is likely that the meta-analysis of the mild AD subgroup was relatively underpowered to detect differences between memantine and placebo on the selected outcome measures.

    The meta-analysis by Schneider et al. also combined two trials of memantine versus placebo monotherapy with one trial of background cholinesterase inhibitors (ChEIs) plus memantine versus placebo. Including these two different types of trials in the single category of “memantine treatments” ignores the potential interaction between memantine and cholinesterase inhibitors (5), as well as the possible benefits of cholinesterase inhibitors alone. Equating the comparison of memantine to placebo with that of memantine plus ChEI to placebo plus ChEI assumes that the effects of these two classes of drugs are simply additive and ignores any potential interaction between them. The meta-analysis by Schneider et al. did not examine this issue.

    In conclusion, the recent meta-analysis, while interesting and well executed, leaves open the question of whether or not memantine, either as a monotherapy or combined with cholinesterase inhibitors, is beneficial to patients with mild Alzheimer’s disease. These criticisms of the current meta-analysis should not be interpreted as advocacy for the off-label use of memantine in patients with mild AD. Indeed, our standard practice is not to prescribe memantine to mild patients, but rather to do so only when they show rapid progression on a cholinesterase inhibitor alone, or a decline to the moderate stage of disease. We agree with the authors that studies specifically designed to look at the efficacy of memantine in a range of mild AD patients are necessary in order to fully answer this question. Our response is motivated by the unsupported use of this recently published meta-analysis by some in the lay press to suggest that memantine is ineffective and should not be used by any patients with AD. Clearly, the authors of the current meta-analysis were not questioning either the wisdom of the FDA or their decision to approve memantine as a treatment for patients with moderate to severe Alzheimer’s disease.


    . Memantine in moderate-to-severe Alzheimer's disease. N Engl J Med. 2003 Apr 3;348(14):1333-41. PubMed.

    . Memantine treatment in patients with moderate to severe Alzheimer disease already receiving donepezil: a randomized controlled trial. JAMA. 2004 Jan 21;291(3):317-24. PubMed.

    . Lack of evidence for the efficacy of memantine in mild Alzheimer disease. Arch Neurol. 2011 Aug;68(8):991-8. PubMed.

    . Meta-analysis of six-month memantine trials in Alzheimer's disease. Alzheimers Dement. 2007 Jan;3(1):7-17. PubMed.

    . Long-term course and effectiveness of combination therapy in Alzheimer disease. Alzheimer Dis Assoc Disord. 2008 Jul-Sep;22(3):209-21. PubMed.

    View all comments by Rachelle Doody

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Treatment with cholinesterase inhibitors and memantine of patients in the Alzheimer's Disease Neuroimaging Initiative. Arch Neurol. 2011 Jan;68(1):58-66. PubMed.

External Citations

  1. Forest Laboratories, Inc.
  2. Merz Pharma
  3. related news from the BBC

Further Reading


  1. . Meta-analysis of six-month memantine trials in Alzheimer's disease. Alzheimers Dement. 2007 Jan;3(1):7-17. PubMed.
  2. . Memantine in moderate to severe Alzheimer's disease: a meta-analysis of randomised clinical trials. Dement Geriatr Cogn Disord. 2007;24(1):20-7. Epub 2007 May 10 PubMed.
  3. . Commentary on "Meta-analysis of six-month memantine trials in Alzheimer's disease." Memantine has negligible benefits in mild to moderate Alzheimer's disease. Alzheimers Dement. 2007 Jan;3(1):21-2. PubMed.

Primary Papers

  1. . Lack of evidence for the efficacy of memantine in mild Alzheimer disease. Arch Neurol. 2011 Aug;68(8):991-8. PubMed.