Research suggests that the hippocampus goes into overdrive in people who may be on the path to Alzheimer's disease (AD). Is the excess neural activity helping these folks to remember, or making them forget? A study in today's Neuron by Michela Gallagher, Johns Hopkins University, Baltimore, Maryland, suggests the latter. When treated for two weeks with a low dose of an activity-dampening anticonvulsant drug, patients with amnestic mild cognitive impairment (aMCI) showed reduced hippocampal activity and performed better on a memory task. Though preliminary, the results suggest that suppressing hippocampal hyperactivity could be a therapeutic approach for early AD.

"Many of us have long suspected that hippocampal hyperactivation is a harmful response in patients with mild cognitive impairment," said Brad Dickerson, Massachusetts General Hospital, Charlestown. "This is the first time anyone has really tested that hypothesis and provided compelling data that this is a target for treatment."

In humans, greater hippocampal activation was observed in people with aMCI (see Dickerson et al., 2005), asymptomatic ApoE4 carriers (see ARF related news story), and people carrying familial AD mutations (see Quiroz et al., 2010). Some researchers believed that, hampered by AD pathology, the hippocampus was recruiting additional, compensatory resources (see Dickerson et al., 2004, and Hämäläinen et al., 2007). More recent research suggested that the hyperactivation could result from a loss of hippocampus inhibitory function (see Andrews-Zwilling et al., 2010) and may be closely tied to neurodegeneration (see ARF related news story). Anti-epileptic drugs, including levetiracetam, dialed back hippocampal hyperactivity in aged, male Long-Evans rats—which also show a boost in hippocampus action and memory impairments (see Wilson et al., 2006). The drugs improved the rats’ spatial memory performance (see Koh et al., 2010). Would a similar reduction in humans help as well?

First author Arnold Bakker and colleagues tested this idea in 17 people diagnosed with aMCI and 17 age-matched healthy controls over two, two-week treatment phases. All controls received placebo for both phases, while patients received levetiracetam in one phase and placebo in the other. The two phases straddled a four-week washout period. Levetiracetam is FDA approved for epilepsy patients. It quiets firing in neurons with both normally and abnormally elevated firing rates, especially in neurons that fire in bursts.

At the end of each treatment phase, the researchers performed high-resolution functional magnetic resonance imaging (fMRI) scans while participants completed a memory task that calls on the dentate gyrus (DG)/CA3 region of the hippocampus. The task was to view a series of images of everyday objects and indicate whether they were "old" (had been shown before), "new" (had not), or whether they were "similar" to one that had already been viewed. The researchers paid special attention to the "similar" items, because mistaking one for an "old" item meant that the memory of a previously viewed object interfered with a new representation of a look-alike. Participants also underwent a series of neuropsychological tests after each treatment phase.

During the "similar" series, aMCI patients on placebo treatment had more activation in the DG/CA3 region than did their age-matched counterparts. They also made more errors, judging more similar items as "old." The researchers hypothesized that if the hippocampal hyperactivity was helpful, then decreasing its function should worsen memory in aMCI patients. Instead, two weeks on a low dose of levetiracetam (less than half of that given to epilepsy patients) improved task performance. Treated patients made fewer memory errors judging similar items. Patients on the drug also activated their hippocampus less during the memory task, suggesting that the hyperactivation hindered performance in the first place. "It was unclear whether hyperactivity was a beneficial compensatory response or just a detrimental overactivation of neuronal networks," said Jorge Palop, University of California, San Francisco (see full comment below). "The data clearly suggest ... a primary and detrimental effect of network hyperactivity in MCI patients."

While performance on the other neuropsychological tests did not differ between placebo and drug patients, it is likely that assessments after a longer period of drug treatment would be needed to observe whether there were any drug-induced improvements in global memory, Gallagher said. Nevertheless, "these data provide evidence that perhaps treating hyperactivity could be a benefit to people with aMCI," she said.

This finding could be but one small piece of a larger puzzle, said Gunnar Gouras, Lund University, Sweden. While some areas of the brain could be hyperactive and need calming down, others could be hypoactive and require a boost, he pointed out. "If we only consider hyperactivity, we're oversimplifying." A fuller understanding of the affected circuits is probably needed in order to affect global memory, he said (see full comment below).

For instance, Gallagher and colleagues previously showed that the entorhinal cortex, which is hit earlier and harder in AD, was hypoactive in patients with MCI performing a pattern separation task (see Yassa et al., 2010). "If you're really going to try to make a strong impact on Alzheimer's disease, one would imagine that drugs that benefit the entorhinal cortex should be developed as well," said Scott Small, Columbia University in New York.

As for immediate implications for treatment, all researchers interviewed for this article agreed that formal clinical trials of levetiracetam's long-term safety profile and effectiveness will be needed in aMCI patients before doctors should consider prescribing it to them. "I think it would be premature for any physician to be prescribing it off-label," said Jessica Langbaum, Banner Alzheimer's Institute, Phoenix, Arizona, who is helping design and execute upcoming presymptomatic Alzheimer's treatment trials. This is just one short-term study with a small sample size that has not yet been replicated, Langbaum pointed out. To find out whether there is a true clinical benefit, the drug needs to be tested on more patients for a longer period of time, she said. The Food and Drug Administration declined comment on its procedures for clinical trials in a case like this.

A separate, one-year study of levetiracetam in AD patients who had seizures suggested that the drug does improve attention level and oral fluency, and is safe for use in controlling seizures in AD patients (see Cumbo and Ligori, 2010). However, Gallagher added that those benefits could have resulted from reduced seizures rather than global cognitive benefits. What's more, the patients had AD rather than MCI, so by that point their hippocampal activity was likely reduced relative to controls. Therefore, studies on the drug's effectiveness at the MCI/prodromal stage of AD are needed, she said.

Gallagher plans on conducting clinical trials in the near future. She is currently testing different doses in the same experimental paradigm. "We want to be very careful to understand what exposure to this drug might do, particularly in this vulnerable population," Gallagher said. "There is a possibility that overactivity itself is bad for the brain, and controlling it could have some long-term benefit, but we don't know that."

In related findings, Palop and colleagues recently reported that genetically reducing abnormal cortical firing by way of boosting inhibitory neurons improved memory in a familial mouse model of AD (see ARF related news story). Both sets of data point to the potential future utility of reducing neural activity in the treatment of very early AD.—Gwyneth Dickey Zakaib


  1. This is a very interesting report showing that treatment with the antiepileptic levetiracetam improves both the task-specific hyperactivity in the DG/CA3 hippocampal circuit seen by fMRI in patients with amnestic MCI and also the cognitive task that reflects the function of this circuit. It is exciting to follow how better functional imaging can provide ever more detailed insights into the brains of living people with MCI, and now even connect therapy to improvement in a selected brain circuit. At the same time, this particular treatment provided no overall improvement on standard neuropsychological testing, including other memory tests. Mucke, Palop, and others have underscored the importance of better understanding of the neural circuitry dysfunction, in particular aberrant hyperactivity, in Alzheimer's disease.

    In Parkinson's disease, research into circuitry dysfunction has led to therapies (such as lesioning of hyperactive subthalamic nucleus). It wasn't as clear why the authors wrote that some have viewed hippocampal hyperactivity in aMCI as beneficial. Dysfunctional brain circuits typically lead to imbalances, including hyper- and hypoactive circuits. The authors previously noted hypoactivity in the ERC-to-DG/CA3 circuit in aMCI (Yassa et al., 2010). One might not expect that this hypoactive ERC circuit would also benefit from low-dose levetiracetam. The relative contributions of elevated versus reduced (and also resting versus task-related) activity to the disease process remain unclear. Current evidence supports that both hyperactive (Mohajeri et al., 2002) as well as hypoactive (Tampellini et al., 2010) circuitry can negatively impact the disease process.


    . Ultrahigh-resolution microstructural diffusion tensor imaging reveals perforant path degradation in aged humans in vivo. Proc Natl Acad Sci U S A. 2010 Jul 13;107(28):12687-91. PubMed.

    . Passive immunization against beta-amyloid peptide protects central nervous system (CNS) neurons from increased vulnerability associated with an Alzheimer's disease-causing mutation. J Biol Chem. 2002 Sep 6;277(36):33012-7. PubMed.

    . Effects of synaptic modulation on beta-amyloid, synaptophysin, and memory performance in Alzheimer's disease transgenic mice. J Neurosci. 2010 Oct 27;30(43):14299-304. PubMed.

  2. Cortical and hippocampal hyperactivity of neuronal networks are early events in AD pathogenesis and are associated with early amyloid deposition in non-demented humans with and without mild cognitive impairment (MCI) (Sperling et al., 2009; Putcha et al., 2011). However, it was unclear whether such hyperactivity was a beneficial compensatory response or just a detrimental overactivation of neuronal networks. Michela Gallagher’s data clearly suggest that attenuation of hippocampal hyperactivity by levetiracetam improves memory performance in amnestic MCI patients, indicating a primary and detrimental effect of network hyperactivity in MCI patients. This is a quite remarkable and important finding that may lead to novel therapies in a short period of time.


    . Hippocampal hyperactivation associated with cortical thinning in Alzheimer's disease signature regions in non-demented elderly adults. J Neurosci. 2011 Nov 30;31(48):17680-8. PubMed.

    . Amyloid deposition is associated with impaired default network function in older persons without dementia. Neuron. 2009 Jul 30;63(2):178-88. PubMed.

  3. In my opinion the manuscript is very thought provoking and describes a problem of great interest. However, the literature on this matter is not homogeneous.

    It would be interesting to do a follow-up study of these patients. This issue needs to be more thoroughly investigated, although the cognitive results obtained with levetiracetam (LEV) are indeed very encouraging. The implications of this study are adequately discussed, and the results do support these conclusions, but a univocal interpretation of such positive results is not easy. The section on patients and methods is fairly good, except that there is no comment made about local authority approval or patient consent. There is also no discussion of adverse events. What are the main dosages of LEV utilized in the study? The limitations, namely the small number of patients, and the possibility that they are not adequately matched in a small study like this, are not fully discussed.

    It has been suggested that LEV may influence the brain's metabolism in areas devoted to attention and language, similar to what has been suggested with piracetam, the other pyrrolidone derivative found to be associated with improved learning, memory, and attention. In fact, there is some evidence that piracetam enhances glucose utilization and cellular metabolism in the brain. LEV was initially studied in animal models of cognitive impairment with the primary objective of finding a drug more effective than piracetam. However, LEV was initially considered to be more effective in preventing seizures.

    Our experience of LEV is clinical. We structured a clinical trial to evaluate efficacy, tolerability, and cognitive effects of levetiracetam (LEV) in patients with seizures and Alzheimer’s disease (AD). This was a prospective, randomized, three-arm parallel group, case-control study in 95 patients: LEV (n = 38), phenobarbital (PB) (n = 28), and lamotrigine (LTG) (n = 29). A four-week dose adjustment was followed by a 12-month evaluation period. The three groups were compared to a control group (n = 68) to evaluate cognitive effects. We examined drug effects cross-sectionally at baseline, six, and 12 months. LEV was well tolerated and maintained its efficacy during all long-term management of AD. LEV improved cognitive performance, specifically attention level and oral fluency items. LEV had a benign and neuropsychological side effect profile, making it a cognitively safe drug to use for controlling established seizures in patients with dementia. The mean daily dose of LEV monotherapy was 956 mg (range: 500 to 2,000 mg/day).

    View all comments by Eduardo Cumbo

Make a Comment

To make a comment you must login or register.


News Citations

  1. ApoE(ε)4 Brains Have to Work Harder
  2. Research Brief: Hippocampal Hyperactivity Tied to Early MCI Atrophy
  3. Needs Salt: Reduced Sodium Channel Linked to Seizures in AD Model

Paper Citations

  1. . Increased hippocampal activation in mild cognitive impairment compared to normal aging and AD. Neurology. 2005 Aug 9;65(3):404-11. PubMed.
  2. . Hippocampal hyperactivation in presymptomatic familial Alzheimer's disease. Ann Neurol. 2010 Dec;68(6):865-75. PubMed.
  3. . Medial temporal lobe function and structure in mild cognitive impairment. Ann Neurol. 2004 Jul;56(1):27-35. PubMed.
  4. . Increased fMRI responses during encoding in mild cognitive impairment. Neurobiol Aging. 2007 Dec;28(12):1889-903. Epub 2006 Sep 25 PubMed.
  5. . Apolipoprotein E4 causes age- and Tau-dependent impairment of GABAergic interneurons, leading to learning and memory deficits in mice. J Neurosci. 2010 Oct 13;30(41):13707-17. PubMed.
  6. . Neurocognitive aging: prior memories hinder new hippocampal encoding. Trends Neurosci. 2006 Dec;29(12):662-70. PubMed.
  7. . Treatment strategies targeting excess hippocampal activity benefit aged rats with cognitive impairment. Neuropsychopharmacology. 2010 Mar;35(4):1016-25. PubMed.
  8. . High-resolution structural and functional MRI of hippocampal CA3 and dentate gyrus in patients with amnestic Mild Cognitive Impairment. Neuroimage. 2010 Jul 1;51(3):1242-52. PubMed.
  9. . Levetiracetam, lamotrigine, and phenobarbital in patients with epileptic seizures and Alzheimer's disease. Epilepsy Behav. 2010 Apr;17(4):461-6. PubMed.

Further Reading


  1. . Longitudinal fMRI in elderly reveals loss of hippocampal activation with clinical decline. Neurology. 2010 Jun 15;74(24):1969-76. PubMed.

Primary Papers

  1. . Reduction of hippocampal hyperactivity improves cognition in amnestic mild cognitive impairment. Neuron. 2012 May 10;74(3):467-74. PubMed.