• MEMANTINE: PHARMACOLOGY AND BIOLOGICAL PROPERTIES
• CLINICAL THERAPY
• MEMANTINE AND VASCULAR DEMENTIA
• CONCLUSION
• REFERENCES

Alzheimer’s disease (AD) is the most common cause of dementia in older adults, the incidence doubling every 5 years after the age of 65 [1]. United States census data from the year 2000 revealed that 4.5 million Americans had AD: 7% were age 65-74 years, 53% were 75-84 years, and 40% were 85 years or older [2]. Contrary to public perception, the typical individual with AD is not profoundly cognitively impaired: 48% of cases are considered mild, 31% are moderate, and only 21% are severe [2]. Both the prevalence of AD (determined by its incidence and the duration of the disorder) and the proportion of severe AD cases increase with age (Figure 1). Population trends ensure that the challenges of AD will also increase as older age groups continue to expand—nearly 19 million Americans will be ≥85 years old by the year 2050—with the result that this already substantial societal and public health burden will become enormous.

Figure 1. Prevalence of dementia.

Since the introduction in 1984 by the Work Group of the National Institute of Neurological and Communicative Diseases and Stroke and the Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) of standard clinical diagnostic criteria for AD [3], much has been learned about its clinical and behavioral signs and symptoms. Current clinical diagnostic methods can achieve a correct diagnosis of AD in ≥85% of cases (rates confirmed by autopsy) [4], and AD now should be considered an inclusionary, rather than an exclusionary, diagnosis [5]. However, notwithstanding the publication of the NINCDS-ADRDA clinical diagnostic criteria, the more numerous cases, and the availability of US Food and Drug Administration (FDA)-approved treatments, AD remains woefully underdiagnosed in the community: perhaps ≥50% of individuals with dementia go undetected by their physicians [6,7].
The American Academy of Neurology (AAN) published revised practice parameters in 2001 for the diagnosis of dementia [8]. The clinical approach involves first establishing the presence of dementia, followed by differential diagnosis. Indeed, dementia remains a clinical diagnosis; no test replaces an assessment by an experienced clinician. Recommended criteria for routine diagnosis of dementia are provided in the Diagnostic and Statistical Manual [9]:

The essential feature of dementia is impairment in short- and long-term memory, associated with impairment in abstract thinking, impaired judgment, other disturbances of higher cortical function, or personality change. The disturbance is severe enough to interfere significantly with work or usual social activities or relationships with others.

Once dementia is diagnosed, standard criteria can be used to determine that AD is the likely cause. AD is so prevalent, that the constellation of gradual onset and progression of memory deficits, impaired executive function, and other cognitive loss without altered consciousness occurring late in life can usually be attributed to AD (Figure 2).

Figure 2. Diagram of the most frequent varieties of dementia.

Dementia beginning earlier than about age 60 is uncommon, but early-onset disease is often determined by autosomal dominant point mutations in genes encoding the amyloid precursor protein or the presenilin proteins (see below). So-called susceptibility genes (e.g., the gene encoding apolipoprotein E) can modify disease risk for both early- and late-onset dementia (Figure 3).

Figure 3. Pathology of Alzheimer's disease.

Pathologically, AD is characterized by the accumulation of neurofibrillary tangles and neuritic plaques in specific brain regions. A key biochemical abnormality is the accumulation of amyloid-beta peptide (Aβ) as soluble fibrils or as insoluble aggregates within plaque cores. Although the etiology of AD may involve many processes (e.g., oxidative stress, inflammation), most investigators favor targeting Aβ or hyperphosphorylated tau, the principal components of plaques and tangles, respectively (Figure 4). Aβ is a 40- to 42-amino-acid peptide that is part of a transmembrane protein called amyloid precursor protein (APP). Three secretase enzymes cleave APP in the region of the Aβ segment: β- and γ-secretase cleave at the two ends of the Aβ sequence, while α-secretase cleaves within Aβ. Thus sequential cleavage of APP by β- and γ-secretase releases the intact Aβ peptide, while cleavage by α-secretase precludes release of the intact peptide. β-secretase, also known as BACE1 or memapsin 2, is an aspartic protease. γ-Secretase is a complex of proteins capable of intra membranous proteolytic cleavage of APP. Intact Aβ released by cleavage of APP can exist in a number of conformations; it auto-aggregates into oligomers and ultimately fibrils (the constituents of plaques), assuming an antiparallel β sheet conformation characteristic of an insoluble amyloid tissue deposit.

	Figure 4a. AD brain tissue, with amyloid deposits.
	Figure 4b. The amyloid cascade hypothesis.

Tau is a structural neuronal protein necessary for the stabilization of microtubules. In AD and other neurodegenerative diseases, tau becomes hyperphosphorylated. Hyperphosphorylated tau changes conformation, condensing into paired helical filaments that comprise the tangles. Tangle accumulation more closely reflects antemortem clinical manifestations of the disease than does plaque pathology [10].
Specific mutations of three different genes can cause familial autosomal dominant AD: the gene encoding APP on chromosome 21, and two genes named presenilin 1 and 2 on chromosomes 14 and 1, respectively. The mutations of APP that cause familial AD all result in altered rates of secretase cleavage, increasing β- and γ-secretase activity or reducing α-secretase activity; each of these mutations increases the release of intact Aβ peptide. Presenilins 1 and 2 are components of the γ-secretase complex, and AD-causing presenilin mutations—the most common cause of familial AD—augment γ-secretase activity, again resulting in increased release of intact Aβ. This further suggests that cleavage of APP to release Aβ must be the pivotal step in sporadic AD, with other yet-to-be-elucidated factors contributing to the accumulation of Aβ. Most investigators and pharmaceutical companies have concluded that the most promising target for disease-modifying therapy in AD is Aβ.
There is an intriguing connection between amyloid generation and tau pathology: immunologic reduction of Aβ in a triple-transgenic APP-PS1-tau mouse with both amyloid plaque and tangle-like pathology not only reduces amyloid accumulation, but also early tau pathology [11,12]. This supports the amyloid hypothesis that Aβ reduction may slow tangle formation and neurodegeneration.
Secretase modulation has appropriately received the greatest attention as a potential therapeutic target. Elucidation of the role of the presenilins in the γ-secretase complex [13] lends weight to the validity of this target. However, development of drugs targeting γ-secretase has been slowed by the recognition that the enzyme has other essential substrates (such as Notch) [14], and that its inhibition produces gastrointestinal and perhaps other toxicities [15]. Nonetheless, there is recent evidence that γ-secretase activity can be modulated, with consequent reduction in full-length Aβ generation and without inhibiting essential secretase activity. Indeed, this modulating activity is shared by a number of nonsteroidal anti- inflammatory drugs (at high concentrations) [16]. Drug development in this area continues [17], and at least one γ-secretase inhibitor has reached phase 1 testing [18].
In some ways, β-secretase is more promising as a therapeutic target. β-Secretase is in the cathepsin family of aspartic proteases, along with the HIV protease that was successfully targeted with a specific inhibitor. It thus seems highly plausible that a drug-like β-secretase inhibitor can be developed, and that it will be tolerable and effective in slowing the disease process. Unfortunately, the protein’s large binding site [19] may be an obstacle to drug development, and while several groups have reported progress, no β-secretase inhibitor has reached the stage of clinical testing.
Success in reducing amyloid accumulation has also been achieved via immunization, in 1999 [20]. A vaccine was prepared by mixing aggregated Aβ with an adjuvant. Mice vaccinated at a young age were protected against amyloid deposition. Still more exciting (from the perspective of therapeutic research), mice with established amyloid pathology treated with the vaccine showed significant regression of plaques.
Studies have suggested that the vaccine worked by inducing antibodies against the N-terminal region of Aβ [21]. Some of the antibodies penetrated the central nervous system (CNS) and bound to plaques, rendering the plaques susceptible to phagocytosis by microglial cells. This work led to rapid testing of an amyloid vaccine in humans with AD. Unfortunately, a large phase 2 trial of the vaccine was halted because of the development of meningoencephalitis in 6% of subjects [22]. This serious adverse effect was apparently mediated by T-cell autoimmunity induced by the vaccine. Examination of brain tissue from several subjects who died following participation in the trial showed evidence of amyloid clearance [23,24]. Post-hoc analysis of cognitive data from participants who developed significant anti-amyloid antibody titers suggests that the vaccine had clinical benefits [23,24].
As a result, studies of alternative immunotherapies continue. A number of pharmaceutical companies are developing passive immunization treatments, consisting of infusion of monoclonal anti-amyloid antibodies. This offers important advantages: unlike active vaccination, no T-cell immune response is expected, the therapy can target specific epitopes on the amyloid peptide, and therapy can be discontinued should evidence of adverse effects appear [25].
Active and passive immunotherapies generate antibodies that bind to Aβ, principally aggregated Aβ in plaques. Infusion of anti-amyloid FAB fragments (antibodies lacking FC receptors, so that phagocytosis is not induced) resulted in a reduction of brain amyloid load in transgenic mice; one postulated mechanism to explain this finding is that amyloid binding results in plaque disruption [26]. But there is also evidence that amyloidbinding agents may sequester Aβ in the periphery, with depletion of the brain amyloid pool [27,28].
Fibrillization of Aβ is strongly influenced by heavy metal ions; chelators of such ions can markedly retard polymerization. Clioquinol is an effective metal chelator that markedly reduced amyloid polymerization in vitro, and reduced brain amyloid load in transgenic mice. A small, phase 2-type randomized controlled trial of clioquinol treatment in mild to moderate AD subjects had positive (though preliminary) results: there was evidence of cognitive benefits with treatment, and alterations of plasma Aβ were demonstrated [29]. Interestingly, the cognitive benefits appeared rapidly with treatment, suggesting a symptomatic benefit; this may suggest that some cognitive dysfunction in AD is directly mediated by Aβ, so that alteration of Aβ conformation or distribution may have short-term benefits. Unfortunately, enthusiasm for development of clioquinol is tempered by toxicity concerns: it was linked to a rare form of optic nerve damage called subacute myelo-opticoneuropathy (SMON).
Fibrillization of Aβ also depends on the molecular environment. In the brain, interaction between Aβ and proteoglycans promotes tissue deposition. Tramiprosate (Alzhemed®) was developed as a glycosaminoglycan (GAG) mimetic, interacting with the GAG-binding region of soluble Aβ to reduce fibrillization [30,31]. Of all anti-amyloid strategies, tramiprosate has advanced the furthest, and is now in phase 3 testing. A phase 2 study of tramiprosate in subjects with mild to moderate AD demonstrated tolerability, safety, and brain penetration, and produced a reduction in the level of Aβ42 in the cerebrospinal fluid (CSF) in a dose-dependent manner [32], providing a strong indication that the treatment influences brain amyloid. A large phase 3 trial has failed in the United States and Canada; despite descriptive data showing numerical differences, the study did not demonstrate a statistical difference in favor of tramiprosate, which was a major reason for discontinuation of the European trial. Neurochem is also currently developing a tramiprosate pro-drug that it claims achieves higher levels in the brain; the company intends to continue an open-label extension of the North American trial.
In addition to the chelation and GAG mimetic approaches, other strategies for reducing Aβ polymerization are being pursued. These include peptide analogues of Aβ [33], compounds related to amyloid histological dyes such as Congo Red [34], as well as a number of other types of molecules with demonstrated anti-polymerization effects in vitro [35].
Disease-modifying therapy is the major goal of therapeutic research on AD today (Figure 5). Progress in understanding amnestic mild cognitive impairment (MCI), which is essentially the prodrome of AD, will allow therapeutic trials of anti-amyloid therapy before AD diagnosis. The combination of disease-modifying therapy with currently available cognitive-enhancing treatments may be the optimal treatment, and gives hope that this disease may, to a substantial degree, be brought under control.

Figure 5. Treatment outcomes in Alzheimer's disease.

The rationale for the initial development of drugs for AD was based on the cholinergic hypothesis [36], which states that acetylcholine is the primary neurotransmitter mediating cognitive function, and that depletion of cholinergic neurotransmission is the immediate cause of cognitive impairment in AD. This hypothesis was supported by demonstration in the 1970s that the cholinergic nucleus basalis of Meynert is a site of neuropathology in AD, and that markers of cholinergic activity are depleted in the AD brain [37,38]. It was further demonstrated that cognitive deficits similar to those of AD can be induced in animals and humans by anticholinergic treatments [36].
After a cholinergic impulse, acetylcholine is cleared from the synapse by the action of acetylcholinesterase, splitting the molecule into choline and acetic acid for re-uptake by the presynaptic neuron. Cholinesterase inhibitors augment the activity of acetylcholine by slowing this clearance, thus partially compensating for the disease-induced cholinergic depletion. This class of drug became the first to be approved for the treatment of the primary cognitive symptoms of AD [39] (Figure 6).

Figure 6. Normal cholinergic function.

The first of these drugs, tacrine, showed efficacy in the treatment of AD and was approved by the FDA in 1993. Approval of donepezil in 1997 represented a substantial advance in cholinesterase inhibitor therapy. Donepezil has a very long half-life, allowing once-daily dosing. Furthermore, donepezil is free of major toxicity, and blood-test monitoring is not required. The side effects of donepezil, which include gastrointestinal symptoms (particularly diarrhea), urinary frequency, muscle cramps, and vivid dreams, are generally mild and tolerable.
Rivastigmine was the third cholinesterase inhibitor to gain FDA approval. It has an intermediate half-life, and is administered twice daily. Whereas donepezil is a selective inhibitor of acetylcholinesterase, rivastigmine is a dual inhibitor, inhibiting butyrylcholinesterase as well. Because butyrylcholinesterase levels in the CNS rise with progression of dementia, it is plausible that lack of selectivity may confer added efficacy; this has not been tested in long-term head-to-head comparison studies.
Galantamine is the latest cholinesterase inhibitor to reach the US market. Galantamine differs from both donepezil and rivastigmine in its mechanism of action: in addition to inhibiting acetylcholinesterase, galantamine is an allosteric modulator of the nicotinic cholinergic receptor. Because presynaptic nicotinic receptors regulate acetylcholine release, this second mechanism may contribute to the drug’s efficacy.
Pivotal trials of each of the three cholinesterase inhibitors in common use have shown similar results: modest improvement in cognitive function and global clinical status, and stabilization of function and behavior for patients with mild to moderate AD. These results justify treatment of the majority of patients who tolerate the generally mild side effects. There is some evidence that cholinesterase inhibitors are helpful in patients with severe AD, and many clinicians prescribe them for all but end-stage AD patients. While no placebo-controlled study has evaluated treatment longer than two years, there is no evidence that patients become refractory to the symptomatic benefits of cholinesterase inhibitor therapy [40].

MEMANTINE: PHARMACOLOGY AND BIOLOGICAL PROPERTIES

Memantine, representing a second class of drug for AD treatment, was approved by the FDA. Memantine is a relatively weak inhibitor of the N-methyl-D-aspartate (NMDA) glutamate receptor. Glutamate is a neurotransmitter essential to cognition and behavior, but overstimulation of NMDA receptors by glutamate leads to neurodegeneration [41]; thus, NMDA inhibitors have been sought for neuroprotection in AD and other illnesses. However, strong NMDA antagonists have been associated with toxicity. Memantine has been demonstrated to be useful in the treatment of AD without intolerable side effects. Studies in patients with moderate to severe AD demonstrate that memantine treatment improves cognition and function, whether given alone [42] or in conjunction with donepezil [43]. This effect may be mediated through amelioration of symptoms rather than through neuroprotection; there is no evidence that memantine treatment slows the rate of clinical decline.
It is now generally accepted that glutamate is the major fast excitatory neurotransmitter in the mammalian CNS. It activates four major types of ionotropic receptors, namely α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), kainate, and NMDA and several types of metabotropic receptors. AMPA receptors are probably involved in all forms of fast glutamatergic neurotransmission [44]. There are four known subunits, GluR1 to GluR4, which form functional receptors as tetrameric subunit assemblies. All AMPA receptors are permeable to Na+ and K+, whereas complexes lacking GluR2 subunits are also permeable to Ca2+. NMDA-sensitive ionotropic glutamate receptors are coupled to high-conductance cationic channels permeable to Na+, K+, and Ca2+ [44].
The NMDA channel is blocked in a use- and voltage-dependent manner by Mg2+ and many exogenous agents. NMDA receptors are only activated following depolarization of the postsynaptic membrane that physiologically follows AMPA receptor stimulation, which relieves blockade by Mg2+. This unique feature and the high Ca2+ permeability renders NMDA receptors inherently suitable as mediators of synaptic plasticity (e.g., learning and memory) (Figure 7). Similar to Mg2+, noncompetitive NMDA receptor antagonists such as ketamine, dextromethorphan, memantine, phencyclidine, and (+)-5-methyl-10,11-dihydro-5H-dibenzocyclohepten-5,10-imine maleate (+)MK-801 block the NMDA channel in the open state, although the blocking kinetics and voltage dependence of this effect vary considerably [44].

Figure 7. NMDA receptors: structure and properties.

Glutamate is the principal excitatory neurotransmitter in the brain and is active in about one-third of all synapses in the CNS. Glutamate and NMDA receptors are involved in long-term potentiation (LTP), a fundamental process for memory consolidation whereby brief high-frequency stimulation leads to an increased response after subsequent activation. NMDA receptors are involved in mediating the postsynaptic components of LTP in the hippocampus, for example in the Schaffer collateral projection from CA3 to CA1 [45-47].
Under normal physiological resting conditions, the ion channels of the NMDA receptors are blocked by Mg2+ ions in a voltage-dependent manner. The small amounts of glutamate that are released are not sufficient to displace the Mg2+ ions and remove the blockade of the NMDA receptor channels; there is a so-called ‘low-background noise.’ During normal synaptic activity, larger concentrations of glutamate are released, the postsynaptic membrane is depolarized, the Mg2+ blockade is transiently inactivated, Ca2+ ions enter the cell, and a signal is generated.
Under conditions of impaired metabolism, there is a sustained release of glutamate, and excessive glutamate activity is associated with ‘excitotoxicity’, in which the Mg2+ ions are lost from the NMDA receptors, allowing a continuous influx of Ca2+ ions into the cell, in turn creating a high level of background noise and impairing the recognition of signals from physiological activation of the receptor [48]. High intracellular concentrations of Ca2+ eventually lead to neuronal degeneration and cell death, as is observed in many types of acute and chronic neuro-degenerative disorders of the CNS [49].
Disturbance of glutamate homeostasis probably also plays a pivotal role in the neuropathology triggered by other factors, such as energy deficit and formation of free radicals, that facilitate the neurotoxic potential of endogenous glutamate [44]. Three aspects help to demonstrate this effect:

1. There is evidence in AD of increased levels of glutamate (caused by decreased cellular uptake and/or increased release) or other endogenous glutamate- receptor agonists in the vicinity of neurons. In postmortem AD brains there are reduced levels of the astroglial glutamate carrier EAA2 in the frontal cortex [50]. Production of glutamate and oxygen free radicals by macrophages and of tyrosine kinases in microglial cells, monocytes and monocytic cell lines, is enhanced by Aβ (1-40) [51]. In vitro, constituents of senile plaques stimulate microglia to produce an unknown neurotoxin having agonistic properties at NMDA receptors [52].
2. Amyloid-beta peptide either activates NMDA receptors or enhances their sensitivity (Figure 8): Aβ (1-40) stimulates nitric oxide (NO) production by microglia [53]; NO is known to enhance glutamate release and to inhibit its uptake [54]. In vitro, Aβ enhances the toxicity of glutamate [55,56] and augments NMDA receptor-mediated transmission [57]. In vivo, injection of beta-amyloid intracerebrovascularly produces long-lasting depression of excitatory postsynaptic potentials (EPSPs) in the hippocampus—an expression of ongoing mild excitotoxicity—that is prevented by the NMDA receptor antagonist 3-(2-carboxypiperazin-4-yl) propyl-1-phosphonic acid (CPP) [57].
3. Postmortem and epidemiological studies suggest a strong association between glutamate dysfunction and AD. Several authors observed co-localization of glutamatergic neurons and pathological alterations (neurofibrillary tangles and senile plaques) in postmortem brains of Alzheimer’s patients [58,59]. Head trauma (associated with glutamatergic dysfunction) has been suggested to be an important risk factor for AD by some [60] but not other studies [61,62] (Figure 9).

	Figure 8. Glutamate's role in neuronal death.
	Figure 9. Role of glutamate in AD.

Memantine was first studied in psychogeriatric patients more than 10 years ago [63-68] (Figure 10). It is a moderate- affinity noncompetitive NMDA antagonist with voltage-dependent binding characteristics, and has been used in the treatment of patients with dementia, Parkinson’s disease, and neurogenic bladder dysfunction in spasticity [5,47]. The drug was also found to be stimulating or vigilance-enhancing in comatose patients [6,69,70].

Figure 10. Memantine.

The pharmacodynamic properties of memantine have been investigated in patch-clamp experiments in rat retinal ganglion cells [71,72], rat hippocampal and striatal slices [73], rat cortical neurons [74], and neurons obtained from mouse [75] and rat embryos [76]. Memantine blocks NMDA receptor channels in cultured neurons in a voltage-dependent manner, as measured by the patch-clamp technique [72,75,76]. The effects of memantine were not reversed by exposure to glycine <100 μmol/L [76]. Memantine induces open-channel blockade of NMDA receptors [67] and has been observed to be ‘partially trapped’ in NMDA receptor channels [72,74] (Figure 11). ‘ Partial trapping’ may discriminate between classes of drugs that otherwise share the mechanism of trapping channel block, and may influence the effects of a blocker on synaptic transmission. Rather than becoming trapped in 100% of channels, memantine could be washed from approximately one-sixth of channels during in vitro experiments [74].

Figure 11. Glutamate receptor families.

Memantine protects cultured neurons from excitotoxin-induced cell death [77]. The drug dose-dependently prevented glutamate-induced cell death in rat cerebellar, cortical, mesencephalic and hippocampal neurons [78-80] and Ca2+ ion-induced death in retinal ganglions obtained from rats [81].
Memantine exerts neuroprotective effects in several models of brain injury. The drug attenuated loss of cholinergic neurons in the CNS induced by injection of NMDA into the basal forebrain of rats [82]. Memantine also attenuated neuronal injury in various rat models, including traumatic brain injury [83], ischemic stroke induced by occlusion of cerebral [84,85] or carotid arteries [86,87], a photo-induced thrombotic model of cerebral focal ischemia [88], and quinolinic acid-induced hippocampal damage [89].
At a dosage of 5-50 mg/kg in rats, memantine induced production of brain-derived neurotrophic factor (BDNF), a substance shown to promote survival and differentiation of CNS neurons, and trkB, a tyrosine kinase receptor for BDNF [89]. mRNA for BDNF and trkB was detected in limbic cortex slices by in situ hybridization [89].
Memantine is completely absorbed from the gastrointestinal tract (absolute biovailability = 100%) [90]. Maximum plasma concentrations occur between 3 and 8 hours after oral administration, and the relationship between dose and plasma concentration is linear over the range of 10-40 mg in volunteers. Food does not influence the bioavailability of the drug [90].
There is considerable interindividual variation in steady-state plasma concentrations of memantine, which ranged from 70-150 μg/L after administration of a single dose of 20 mg/day under fasting conditions [91].
The mean plasma protein binding ratio is around 45% and the mean volume of distribution of memantine is a round 9-10 L/kg, suggesting extensive distribution of memantine into tissues. The mean CSF/serum ratio was 0.52 after administration of memantine 5-30 mg/day over a range of months [90].
Approximately 80% of an administered dose of memantine circulates as the unchanged parent drug [90]. Several inactive metabolites have been identified (N-3,5-dimethyl-gludantan and the isomeric mixture of 4- and 6-hydroxy-memantine), although cytochrome P450 (CYP)-catalyzed metabolism has not been detected in vitro [90].
The mean terminal elimination half-life of memantine is 60-100 hours [90]. Memantine is eliminated renally and undergoes both renal tubular secretion and reabsorption via cationic transport proteins. In healthy volunteers with normal renal function, total body clearance (CL_tot) was 170 mL/min/1.73 m² (10.2 L/h/1.73 m²) [90]. In elderly volunteers with normal or reduced renal function [creatinine clearance (CrCl), 50-100 mL/min/1.73 m² (3-6 L/h/1.73 m²)], CL_tot of memantine was correlated with CrCl [90].
Memantine did not inhibit CYP1A2, 2A6, 2C9, 2D6, 2EI or 3A in vitro [90]. In addition, memantine had no in vitro inhibitory effect on acetylcholinesterase inhibitors [91]. Furthermore, in healthy volunteers receiving donepezil 10 mg/ day, a single concomitant 10-mg dose of memantine had no effect on the in vitro cholinesterase activity of donepezil; nor did concomitant administration have any relevant effects on the pharmacokinetic profile of either agent [92].

CLINICAL THERAPY

Memantine is a noncompetitive antagonist with moderate affinity for NMDA receptors, and it demonstrates voltage-dependency and relatively fast on/off receptor kinetics. Agents such as memantine that mimic some of the features of the endogenous antagonist Mg2+ may be an optimal treatment, combining both neuroprotective activity with symptomatologic improvement. The latter can be explained on the basis of a reduction in the noise level and restoration of a sufficient ‘signal-to-noise’ ratio.
Owing to memantine’s anti-ischemic and anti-excitotoxic properties, recent clinical efficacy of the drug has been demonstrated in patients with advanced dementia of vascular origin (Figure 12). Memantine is indicated for the treatment of moderate to severe AD, reflecting a license extension granted by the European Commission in October 2005 to include the moderate AD patient population.

Figure 12. Channel blocking kinetics: memantine (intermediate) and dizolcipine (slow).

In two multicenter placebo-controlled trials, memantine demonstrated efficacy and safety in AD patients. The first study enrolled 252 patients, with equal numbers (n = 126) receiving memantine 20 mg/day or placebo for 28 weeks. Efficacy was assessed using the Clinician’s Interview-Based Impression of Change (NYU CIBIC-plus), the modified Alzheimer’s Disease Cooperative Study Activities of Daily Living (ADCS-ADLsev) inventory, the Severe Impairment Battery (SIB), and Functional Assessment Staging (FAST) [93,94]. In the second study, 166 patients were enrolled and randomized to receive either memantine 10 mg/day or placebo for 12 weeks. The primary efficacy endpoints were measured using the Clinical Global Impression of Change (CGI-C) and the Behavioral Rating Scale for Geriatric Patients (BGP). The modified D-scale (Arnold/ Ferm) was used to assess basic activities of daily living as a secondary endpoint. Both studies demonstrated that patients treated with memantine showed improvements in the three main AD domains: function, cognition, and global response. Memantine was well tolerated in both studies; the incidence of adverse events and serious adverse events was similar in both treatment groups [95]. In conclusion, memantine treatment offered functional improvements and reduced care-dependence for patients with moderately severe to severe AD.
Until recently, treatment for AD had been limited to acetylcholinesterase inhibitors (AChEls). Since numerous neurotransmitters are affected in AD, it is logical to investigate the combination of an AChEI (donepezil) with an NMDA antagonist (memantine). A pharmacokinetic study in 24 healthy volunteers showed no pharmacokinetic or pharmacodynamic interactions and the combination was well tolerated. A 24-week, randomized, double-blind, parallel-arm, placebo-controlled trial was performed in 37 US centers to study the safety and efficacy of memantine in patients with moderate to severe AD treated with donepezil [93]. Inclusion criteria of the study were: a diagnosis of probable AD by NINCDS-ADRDA, Mini- Mental State Examination (MMSE) scores of 5-14, magnetic resonance imaging (MRI) or computed tomography (CT) scan consistent with probable AD, and 6-month daily AChEI (donepezil) therapy (stable dose for the past 3 months). Primary outcome assessments were: cognition (SIB) and function (ADCS-ADLsev).The CIBIC-plus global assessment was also performed. Of 403 patients randomized and treated with memantine 10 mg bid (n = 202) or placebo (n = 201), 85% of memantine-treated patients and 75% of placebo patients completed the trial. At endpoint (last observation carried forward , LOCF), memantine/AChEI patients improved significantly (p <.001) in cognition (SIB) compared with placebo/AChEI patients, and declined significantly less (p = .028) in function (ADCS-ADLsev). A significant difference favoring memantine/AChEI was also seen with the CIBIC-plus (p = .027). Memantine/AChEI was safe and well tolerated [96].
Another well designed study involving patients with moderate- to- severe Alzheimer’s disease who were randomly assigned to receive placebo or 20 mg of memantine daily for 28 weeks was conducted [97]. Of the 252 subjects enrolled, 181 (72%) completed the study and were evaluated at week 28, while 71 patients discontinued treatment prematurely (42 taking placebo and 29 taking memantine). Patients receiving memantine had a better outcome than those receiving placebo, according to the results of the CGI-C (p = .06 with LOCF; p = .03 for observed cases), of the ADCS-ADLsev (p = .02 with LOCF; p = .003 for observed cases), and the SIB (p <.001 with LCOF; p = .002 for observed cases). Memantine was not associated with a significant frequency of adverse events. The conclusions of the authors are well captured by the following sentence: “Antiglutamatergic treatment reduced clinical deterioration in moderate-to-severe Alzheimer’s disease, a phase associated with distress for patients and burden on caregivers, for which other treatments are not available” [97].
A postmarketing surveillance study was performed in Germany to assess the tolerability of memantine in combination with an AChEI (84% donepezil and 16% rivastigmine). One hundred fifty-eight of 200 questionnaires were returned, and the data analyses showed that tolerability was rated as good or very good by 98% of the patient respondents. The results demonstrate that combining memantine with a commonly used AChEI is safe and superior to AChEI alone in moderate to severe AD [98-102].
To investigate the effects of memantine on cognitive function, Schmitt et al [103] conducted a post-hoc exploratory reanalysis of a 24-week randomized, double-blind, placebo-controlled, parallel group clinical trial comparing memantine to placebo in patients with moderate to severe AD receiving treatment with donepezil. Their results suggested an effect of memantine on memory, language, and praxis in patients with moderate to severe AD, and support the efficacy of memantine for the treatment of cognitive defects in AD [103], consistent with a previous clinical trial [99].
The results of a recently published study demonstrate that once-daily dosing of memantine is as safe and well tolerated as a twice-daily dosing in patients with moderate to severe AD [104]. The data showed no significant differences in rates of withdrawal or adverse events between the groups receiving memantine dosed either once or twice daily. In a 12-week, randomized, double-blind study of 78 patients with moderate to severe AD the safety and tolerability of three different dosing schedules of memantine were examined. The data demonstrated equal efficacy for once-daily and twice-daily dosing as assessed by CGI-C. Statistically significant benefits of memantine given twice daily to treat moderate to severe AD have already been shown by multiple studies in pivotal AD domains (function, cognition and global performance) [105]. Safety and tolerability were assessed on the basis of the number of withdrawals, adverse events, and monitoring of vital signs. The number of patients withdrawing from the study was low, and adverse effects were transient, mild or moderate, and typical for the patient population studied. No clinically relevant differences in adverse effects or vital signs were observed between the different dosing schedules [104,105].

MEMANTINE AND VASCULAR DEMENTIA

In the course of a cerebrovascular event, there is a rundown of energy in neurons. A number of microdialysis studies indicate that there is also a consistent increase in extracellular glutamate concentration. In humans, there is also an increase in glutamate content in both the CSF and plasma of patients with progressive, but not stable, stroke.
Other factors may increase neuronal vulnerability to physiological levels of glutamate by, for example, decreasing the resting potential of the membrane or the intracellular buffering of Ca2+ ions; oxidative stress, inflammatory reactions, and breakdown of the blood–brain barrier may also play a role.
Studies performed in the 1950s during carotid artery clamping for carotid endarterectomy using intra-carotid xenon-133 injections reported that hemiparesis occurred when regional cerebral blood flow (CBF) fell to 30-50% of normal, and permanent neurologic deficit occurred if mean CBF fell below 30% of normal. It was also shown that development of permanent neurological sequelae is a time-dependent process; for any given blood flow level, low CBF values are tolerated only for a short period of time whereas higher CBF values require a longer time for infarction to occur. Electroencephalographic (EEG) activity slows down if mean CBF falls below 23 mL/100 g/min, whereas at values below 15 mL/100 g/min the EEG becomes flat.
It is important to understand that the blood-flow thresholds studied in most animal and human experiments refer to ischemic tolerance of the brain cortex. Threshold values for deep white matter or basal ganglia have not been studied rigorously and are simply not known. However, it is believed that gray matter is more susceptible to infarction than white matter, and within the gray matter the basal ganglia have a lower ischemic tolerance than does the cortex.
Normal CBF studies differentiate gray and white matter. The basal ganglia and thalamus are easily identified. Cerebral blood flow values in normal gray matter are in the range of 60-100 mL/100 g/min, whereas the CBF of white matter is lower, at 18-22 mL/100 g/min. Regions of interest can be drawn that encompass vascular territories such as the distribution of the middle cerebral artery (MCA), or smaller more discrete areas can be measured. Occlusions of the MCA typically cause large areas of reduced CBF whereas branch occlusions appear as regional areas of low CBF with normal values elsewhere in the MCA territory.
In patients with acute stroke, CBF is typically reduced in the vascular territory appropriate to the clinical deficit observed. Not infrequently, CBF in the opposite hemisphere is also reduced, making measurement of absolute values rather than relative hemispheric differences very important.
Cortical regions with flow values of 8 mL/100 g/min or less in the first few hours after stroke consistently develop infarction, despite eventual reperfusion and thrombolysis. In addition, patients with an MCA territorial flow of 9 mL/100 g/min or less before thrombolytic therapy have a high incidence of hemorrhage and aggressive edema which, in a significantly increased number of cases, causes clinically definable tentorial herniation and brainstem compression. Conversely, when the mean CBF of the MCA territory is 18 mL/100 g/min, patients are less likely to develop massive infarct, massive edema, or clinical herniation. Patients with M1 occlusion of the MCA had a significantly lower territorial CBF than when the occlusion was not in the M1 segment (12 mL/100 g/min vs 30 mL/100 g/min, p <.01, t test). All patients with a mean CBF <20 mL/100 g/min for the entire MCA territory had angiographically proven M1 occlusions [106].
The notion that in acute stroke, depending on the extent and duration of hypoperfusion, the tissue supplied by the occluded artery is compartmentalized into areas of irreversibly damaged brain tissue and areas of brain tissue that are hypoperfused but viable, led to the concept of ischemic core and ischemic penumbra. The ischemic core represents tissue that is irreversibly damaged. Positron emmisson tomography (PET) studies in humans suggest that beyond a certain time limit (probably no longer than one hour) the ischemic core will include those areas having CBF values of <7 mL/100 g/min to 12 mL/100 g/min.
The ischemic penumbra represents tissue that is functionally impaired but structurally intact, and as such potentially salvageable. It corresponds to a high CBF limit of 17-22 mL/100 g/min and a low CBF limit of 7-12 mL/100 g/min. Salvaging this tissue by restoring its flow to nonischemic levels is the aim of acute stroke therapy. Another compartment, termed “oligemia” by Astrup et al. [106], represents mildly hypoperfused tissue from the normal range down to around 22 mL/100 g/min. It is believed that under normal circumstances this tissue is not at risk of infarction. However, it is conceivable that under certain circumstances such as hypotension, fever, or acidosis, oligemic tissue can be incorporated into the penumbra and subsequently undergo infarction.
Evidence in the literature suggests there is a temporal evolution of the core, which can grow at the expense of the penumbra. One of the proposed mechanisms for growth of the ischemic core is progressive recruitment of penumbral areas into the core caused by ischemic edema. It is known that the ischemic penumbra represents a dynamic phenomenon. If vessel occlusion persists, then the penumbra may shrink due to progressive recruitment into the core. Alternatively it may return to a normal state following vessel recanalization or possibly neuroprotectant interventions [107].
Although the brain accounts for only 2% of the body’s mass it utilizes 25% of the body’s energy stores and receives 15-20% of the total cardiac output. Given these high demands it makes sense that even brief periods of anoxia should have dire consequences. An ischemic injury associated with stroke or, more globally, total circulatory arrest results in energy failure (loss of mitochondrial ATP production), which then leads to malfunctioning membrane pumps (Na+ / K+ pumps). The latter leads to depolarization of the cell with subsequent release of glutamate, and consequent neuronal death.
The toxic effects of glutamate are mediated largely through NMDA receptors. Activation of NMDA-associated channels leads to the passage of Na+ and Cl- into the cell followed by the obligatory movement of water, resulting in cytotoxiedema. At this stage, injury is not necessarily irreversible. If Ca2+ is present early it tends to pass through the channel into the cell where it recruits second messengers that activate kinases and proteases that eventually lead to irreversible injury. This is believed to be the underlying mechanism of delayed glutamate toxicity. Glutamate excitotoxicity can lead to a state of selfamplification. In this case, glutamate release increases intracellular Ca2+, leading to second messenger activation, giving rise to changes that make the cell more permeable to additional Ca2+ entry and further glutamate release. Glutamate release also gives rise to oxygen free-radical production, which in turn results in the further release of glutamate. The degree of glutamate release has been shown to be temperature sensitive. Cooling the brain by only a few degrees can prevent the release of glutamate during global ischemia. Other neurotransmitters that modulate glutamate function include glycine and gamma-aminobutyric acid (GABA). Glycine is needed for the normal function of glutamate on the NMDA receptors, whereas GABA is inhibitory.
Global ischemia also results in alterations to glutamate receptor subtypes. Twenty-four hours after ischemia in the hippocampus, there is a reduction in the R2 subtype of the glutamate AMPA receptor. The R2 subtype has the effect of making the AMPA channels impermeable to Ca2+; thus, absence of this receptor subtype would lead to an increase of intracellular Ca2+, causing further neuronal injury.
Following the return of spontaneous cardiac rhythm the brain is also subject to secondary damage as a result of reperfusion–reoxygenation. At this stage, three sets of pathophysiological derangements occur. Initially, there is the so-called post-stroke or post-arrest cerebral reperfusion failure [108-111]. Over the next one to 12 hours a state of global hypoperfusion follows in spite of maintaining normal blood pressure [111]. The hypoperfusion is due to a functional disturbance of the cerebrovascular system and is characterized by an impairment of CO₂ reactivity and increased vascular tone [112]. It has been shown that the increase in vascular tone is mediated by endothelin type A receptors (ETA) [113]. As a result, global CBF can be reduced to as low as 50% of normal. This hypoperfusion tends to be inhomogeneous and can result in regional reduction of blood flow to lethal levels [114]. During this period, the global cerebral metabolic rates for oxygen and glucose recover to levels equal to or even higher than those present at baseline [111]. The increased metabolism may possibly be due to spreading depression [115].
A second wave of metabolic derangements that occurs after reperfusion includes delayed excitotoxicity of vulnerable neurons [116] as well as calcium loading that results in lipid peroxidation of membranes leading to necrosis [114]. Finally apoptosis, or programmed cell death, may be the final common pathway by which all of the effects of ischemia are expressed [117,118].
The results from animal studies are controversial: often, short intervals (1-3 days) for analysis of infarct size are used. There are data showing that some treatments delay, but ultimately do not prevent, neuronal death; that is, protective effects are seen when analyzed at three days but not 7-28 days after the insult. Infarct volume or cell damage is not always predictive of functional (neurological) outcome. There is significant strain and vendor variability of infarct size and neuroprotective efficacy of NMDA receptor antagonists, adding to the already large methodological diversity [119] (Figure 13).

Figure 13. An increased basal level of glutamate induces ‘noise' at the glutamatergic synapse.

Moreover, there is an open debate on the possible extension of glutamate-dependent insults in ischemia beyond the period of reperfusion. In fact, after 30 min of three-vessel occlusion in rabbits, a transient, moderate single peak in brain glutamate levels was observed shortly after its initiation, but two hours of occlusion produced a second, delayed peak at 2-4 hours following reperfusion. This secondary glutamate peak correlated better with infarct size than the primary one, indicating that beneficial effects of glutamate antagonists could possibly be seen even if the treatment is delayed by 2-4 hours after reperfusion. This finding also indicates that maintenance of treatment is necessary for several or more (20 to 24) hours [119].
In summary, the present understanding of glutamate antagonists in brain ischemia is that AMPA receptor antagonists seem to be more active in global ischemia models, whereas both NMDA and AMPA receptor antagonists show moderate activity in focal ischemia models. In post-ischemia treatment regimes, NMDA receptor antagonists in general show much better efficacy in permanent ischemia models [119].
Excitotoxicity, defined as excessive exposure to the neurotransmitter glutamate or overstimulation of its membrane receptors, has been implicated as one of the key factors contributing to neuronal injury and death in a wide range of both acute and chronic neurologic disorders. Excitotoxic cell death is due, at least in part, to excessive activation of NMDA-type glutamate receptors and hence excessive Ca2+ influx through the receptor’s ion channel [120]. Physiological NMDA receptor activity, however, is also essential for normal neuronal function; potential neuroprotective agents that block virtually all NMDA receptor activity will very likely have unacceptable clinical side effects. For this reason many NMDA receptor antagonists have failed advanced clinical trials for a number of diseases, including stroke and neurodegenerative disorders such as Huntington’s disease.
In contrast, studies by Lipton [121] were the first to show that memantine preferentially blocks excessive NMDA receptor activity without disrupting normal activity [121] (Figure 14). Memantine does this through its action as an open-channel blocker (Figure 15): it enters the receptor-associated ion channel preferentially when the channel is excessively open, and, most importantly, memantine’s off-rate is relatively fast so that it does not substantially accumulate in the channel to interfere with normal synaptic transmission. A series of second-generation memantine derivatives is currently in development and may prove to have even greater neuroprotective properties than does memantine itself [121].

	Figure 14. NMDA channel blockers.
	Figure 15. Glutamate's role in vascular damage.

Based on the hypothesis of glutamate-induced neurotoxicity (excitotoxicity) in cerebral ischemia, the efficacy and tolerability of memantine in the treatment of mild to moderate vascular dementia was studied in a multicenter, 28-week trial carried out in France. Three-hundred twenty-one patients received 10 mg/d memantine or placebo twice a day; 288 patients were valid for intent-to-treat analysis. Patients had to meet the criteria for probable vascular dementia and have an MMSE score between 12 and 20 at inclusion. The two primary end points were the cognitive subscale of the Alzheimer’s Disease Assessment Scale (ADAS-cog) and the global Clinician’s Interview Based Impression of Change (CIBICplus) [122]. After 28 weeks, the mean ADAS-cog scores in the treatment arm were significantly improved relative to the placebo arm. In the intention-to-treat population, the memantine group mean score had gained an average of 0.4 points, whereas the placebo group mean score had declined by 1.6 points, i.e., a difference of 2.0 points (95% CI 0.49-3.60). The response rate for CIBIC-plus, defined as improved or stable, was 60% with memantine compared to 52% with placebo (p = .227, intention to treat). Among the secondary efficacy parameters that were analyzed in the per- protocol subset, MMSE was significantly improved with memantine compared with deterioration with placebo (p = .003). The Gottmes-Brane-Steen Scale intellectual function subscore and the Nurses’ Observation Scale for Geriatric Patients disturbing behavior dimension also showed differences in favor of memantine (p = .04 and p = .07, respectively) [122]. Memantine was well tolerated with a frequency of adverse events comparable to that of the placebo. In patients with mild to moderate vascular dementia, the authors concluded that memantine 20 mg/d improved cognition consistently across different cognitive scales, with at least no deterioration in global functioning and behavior. It was devoid of concerning side effects [104].
Another trial was conducted to investigate the safety and efficacy of memantine in mild to moderate vascular dementia (VaD) [123]. It was a 28-week, double-blind, parallel, randomized controlled trial of memantine 20 mg daily versus placebo conducted in 54 centers in the United Kingdom. Patients with a diagnosis of probable VaD and MMSE total scores between 10 and 22 were eligible for inclusion. Primary measures of efficacy were improvements in ADAS-cog and CGI-C scores. A total of 579 patients were randomized and 548 patients with at least one post-baseline efficacy assessment qualified for the intent-to-treat analysis. At endpoint, memantine was shown to improve cognition relative to placebo in VaD: the change in ADAS-cog from baseline differed by a mean of -1.75 points (95% CI 3.023 to -0.49) and a median of 2 points between the two groups, while CGI-C ratings showed no significant differences between treatment groups. A total of 77% of all memantine-treated patients experienced an adverse event versus 75% of the placebo-treated patients, dizziness being the most frequent adverse event (11% versus 8%, respectively) [123].

CONCLUSION

Memantine protects cultured neurons from excitotoxin-induced cell death. Memantine exerts neuro protective effects in several models of brain injury. The drug attenuated loss of cholinergic neurons in the CNS induced by injection of NMDA into the basal forebrain of rats. Memantine induces production of BDNF, a substance shown to promote survival and differentiation of CNS neurons. Owing to its anti-ischemic and anti-excitotoxic properties, clinical efficacy of memantine has recently been demonstrated in patients with advanced dementia of vascular origins. In several trials in subjects with vascular dementia, memantine has shown a potential benefit and no unbearable side effects.
Memantine had a small, clinically detectable beneficial effect on cognitive function and functional decline measured at six months in patients with moderate to severe AD. In patients with mild to moderate dementia, the small beneficial effect on cognition was not clinically detectable in those with vascular dementia and barely detectable in those with AD. It is well tolerated. Slightly fewer patients with moderate to severe AD taking memantine develop agitation, but there is no evidence either way as to an effect on agitation that is already present.
The Specialized Register of the Cochrane Dementia and Cognitive Improvement Group was searched on February 8, 2006. This register contains references from all major healthcare databases and many ongoing trial databases and is updated regularly. In addition, the search engines Copernic and Google were used to identify unpublished trials through inspection of the web sites of licensing bodies, including the FDA, the European Medicines Agency (EMEA), and the National Institute for Health and Clinical Excellence (NICE) as well as those of pharmaceutical companies (Lundbeck, Merz, Forest, Suntori, and others) and clinical trials registries. Data from double-blind, parallel group, placebo- controlled, randomized trials of memantine in subjects with dementia were pooled where possible. Intention- to- treat (ITT) and observed case (OC) analyses have been reported [124]. The main results can be summarized as follows:

1. Moderate to severe AD. Two out of three six-month studies showed a small beneficial effect of memantine. Pooled data indicate a beneficial effect at six months on cognition (2.97 points on the 100- point SIB, 95% CI 1.68-4.26, p <.00001), activities of daily living (1.27 points on the 54-point ADCS-ADLsev, 95% CI 0.44-2.09, p = .003), and behavior (2.76 points on the 144-point NPI [Neuropsyciatric Inventory], 95% CI 0.88-4.63, p = .004), supported by clinical impression of change (0.28 points on the 7-point CIBIC-plus, 95% CI 0.15-0.41, p <.0001).
2. Mild to moderate AD. Pooled data from three unpublished studies indicate a marginal benefical effect at six months on ITT cognition (0.99 points on the 70-point ADAS-cog, 95% CI 0.21-1.78, p = .01), which was barely detectable clinically (0.13 CIBIC-plus points, 95% CI 0.01-0.25, p = .03), but no effect on behavior, activities of daily living, or OC analysis of cognition.
3. Mild to moderate vascular dementia. Pooled data from two six-month studies indicated a small beneficial effect of memantine on cognition (1.85 ADAS-cog points, 95% CI 0.88-2.83, p = .0002) and behavior (0.84 95% CI 0.06-0.91, p = .03) but this impression of general amelioration of cognition and behavior was not supported by clinical global measures.
4. Patients taking memantine were slightly less likely to develop agitation than patients given placebo [134/1739 (7.7%) vs 175/1873 (9.3%); OR 0.78, 95% CI 0.61-0.99, p = .04]. This effect was slightly larger, but still small, in moderate to severe AD [58/506 (12%) vs 88/499 (18%); OR = 0.6, 95% CI 0.42-0.86, p = .005]. There is no evidence either way as to an effect of memantine on agitation that is already present.

A more recent review [125] considered the evidence for the effectiveness of AChEIs and memantine in achieving clinically relevant improvements, primarily in cognition, global function, behavior and quality of life for patients with dementia. Ninety-six publications representing 59 unique studies were considered. Both AChEIs and memantine had consistent effects in the domains of cognition and global assessment, but the studies were marred in some cases by limitations of duration, inclusion of patients with only mild to moderate AD, poor reporting of adverse effects, lack of clear definitions for statistical significance, and limited direct comparison of different treatments [125]. Treatment of dementia with AChEIs and memantine can result in statistically significant but clinically marginal improvement in measures of cognition and global assessment of dementia [125].
Importantly, however, it does appear that the dementia caused by brain vascular disease may share similar anatomic substrates with AD, supporting the notion of a common substrate to dementia. Different studies suggest that memantine should be studied in a broader well-selected population based on etiopathogenetic differences in degeneration, and that different variables should also be included in future studies, including behavioral scores and assessments of executive function and independence in daily living, all of them particularly involved in VaD.

REFERENCES

1. Kukull WA, Higdon R, Bowen JD, et al. Dementia and Alzheimer disease incidence: A prospective cohort study. Arch Neurol 2002;59:1737-1746. [Medline]

2. Hebert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA. Alzheimer disease in the US population. Arch Neurol 2003;60:1119-1122. [Medline]

3. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s disease. Neurology 1984;34:939-944. [Medline]

4. Berg L, McKeel DW Jr, Miller JP, et al. Clinicopathologic studies in cognitively healthy aging and Alzheimer disease: relation of histologic markers to dementia severity, age, sex, and apolipoprotein E genotype. Arch Neurol 1998;55:326-335. [Medline]

5. Small GW, Rabins PV, Barry PP, et al. Diagnosis and treatment of Alzheimer disease and related disorders: consensus statement of the American Association for Geriatric Psychiatry, the Alzheim er’s Association, and the American Geriatrics Society. JAMA 1997;278:1363-1371. [Medline]

6. Sternberg SA, Wolfson C, Baumgartner M. Undetected dementia in community-dwelling older people: the Canadian study of Health and Aging. J Am Geriatr Soc 2000;48:1430-1434. [Medline]

7. Ganguli M, Rodriguez E, Mulsant B, et al. Detection and management of cognitive impairment in primary care: the Steel Valley Seniors Survey. J Am Geriatr Soc 2004;52:1668-1675. [Medline]

8. Knopman DS, DeKosky ST, Cummings JL, et al. Practice parameter: diagnosis of dementia (an evidence-based review ). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2001;56:1143-1153. [Medline]

9. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4 ed. 1994. Washington, D.C.: American Psychiatric Association.

10. Samuel W, Terry RD, DeTeresa R, et al. Clinical correlates of cortical and nucleus basalis pathology in Alzheimer dementia. Arch Neurol 1994;51:772-778. [Medline]

11. Oddo S, Billings L, Kesslak JP, et al. A! immunotherapy leads to clearance of early, but not late, hyperphosphorylated tau aggregates via the proteasome. Neuron 2004;43:321-332. [Medline]

12. Small GW, Kepe V, Ercoli L, et al. PET of brain amyloid and tau in mild cognitive impairment. N Engl J Med 2006;355:2652-2663. [Medline]

13. Takasugi N, Tomita T, Hayashi I, et al. The role of presenilin cofactors in the gamma-secretase complex. Nature 2003;422:438-441. [Medline] 

14. De Strooper B, Annaert W, Cupers P, et al. A presenilin-1-dependent gamma-secretase-like protease mediates release of Notch intracellular domain. Nature 1999;398:518-522. [Medline]

15. Milano J, McKay J, Dagenais C, et al. Modulation of Notch processing by "-secretase inhibitors causes intestinal goblet cell metaplasia and induction of genes known to specify gut secretory lineage differentiation. Toxicol Sci 2004;82:341-358. [Medline]

16. Weggen S, Eriksen JL, Das P, et al. A subset of NSAIDs lower amyloidogenic Abeta42 independently of cyclooxygenase activity. Nature 2001;414:212-216. [Medline]

17. Barten DM, Guss VL, Corsa JA, et al. Dynamics of β-amyloid reductions in brain, cerebrospinal fluid and plasma of β-amyloid precursor protein transgenic mice treated with a γ-secretase inhibitor. J Pharmacol Exp Ther 2005;312:635-643. [Medline]

18. Siemers E, Dean RA, Satterwhite J, et al. Safety, tolerability, and changes in amyloid beta concentrations after administration of a gamma-secretase inhibitor in volunteers. Clin Neuropharmacol 2005;28:126-132. [Medline]

19. Hong L, Koelsch G, Lin X, et al. Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor. Science 2000;290:150-153. [Medline]

20. Schenk D, Barbour R, Dunn W, et al. Immunization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 1999;400:173-177. [Medline]

21. Bard F, Barbour R, Cannon C, et al. Epitope and isotype specificities of antibodies to beta-amyloid peptide for protection against Alzheimer’s disease-like neuropathology. Proc Natl Acad Sci U S A 2003;100:2023-2028. [Medline]

22. Orgogozo JM, Gilman S, Dartigues JF, et al. Subacute meningoencephalitis in a subset of patients with AD after A beta42 immunization. Neurology 2003;61:46-54. [Medline]

23. Nicoll JA, Wilkinson D, Holmes C, et al. Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: a case report. Nat Med 2003;9:448-452. [Medline]

24. Gilman S, Koller M, Black RS, et al. Clinical effects of Abeta immunization (AN1792) in patients with AD in an interrupted trial. Neurology 2005;64:1553-1562. [Medline]

25. Geylis V, Steinitz M. Immunotherapy of Alzheimer’s disease (AD): from murine models to anti-amyloid beta (Abeta) human monoclonoal antibodies. Autoimmunity Rev 2006;5:33-39. [Medline]

26. Bacskai BJ, Kajdasz ST, McLellan ME, et al. Non-Fc-mediated mechanisms are involved in clearance of amyloid-beta in vivo by immunotherapy. J Neurosci 2002;22:7873-7878. [Medline]

27. DeMattos RB, Bales KR, Cummins DJ, et al. Brain to plasma amyloid-beta efflux: a measure of brain amyloid burden in a mouse model of Alzheimer’s disease. Science 2002;295:2264-2267. [Medline]

28. Lemere CA, Spooner ET, LaFrancois J, et al. Evidence for peripheral clearance of cerebral Abeta protein following chronic, active Abeta immunization in PSAPP mice. Neurobiol Dis 2003;14:10-18. [Medline]

29. Ritchie CW, Bush AI, Mackinnon A, et al. Metal-protein attenuation with iodochlorh y d roxyquin (clioquinol) targeting Abeta amyloid deposition and toxicity in Alzheimer disease: a pilot phase 2 clinical trial. Arch Neurol 2003;60:1685-1691. [Medline]

30. Gervais F, Chalifour R, Garceau D, et al. Glycosaminoglycan mimetics: a therapeutic approach to cerebral amyloid angiopathy. Amyloid 2001;(Suppl 1):28-35. [Medline]

31. Aisen PS, Gauthier S, Vellas B, et al. Alzhmed: a potential treatment for Alzheimer’s disease. Curr Alzh Res 2007;4:473-478. [Medline]

32. Aisen PS. The development of anti-amyloid therapy for Alzheimer’s disease: from secreatase modulators to polymerisation inhibitors. CNS Drugs 2005;19:989-996. [Medline]

33. Findeis MA. Peptide inhibitors of beta amyloid aggregation. Curr Top Med Chem 2002;2:417-423. [Medline]

34. Klunk WE, Debnath ML, Koros AM, et al. Chrysamine-G, a lipophilic analogue of Congo red, inhibits A beta-induced toxicity in PC12 cells. Life Sci 1998;63:1807-1814. [Medline]

35. Findeis MA. Approaches to discovery and characterization of inhibitors of amyloid beta-peptide polymerization. Biochim Biophys Acta 2000;1502:76-84. [Medline]

36. Bartus RT, Dean RL 3rd, Beer B, Lippa AS. The cholinergic hypothesis of geriatric memory dysfunction. Science 1982;217:408-414. [Medline]

37. Whitehouse PJ, Price DL, Clark AW, et al. Alzheimer disease: evidence for selective loss of cholinergicneurons in the nucleus basalis. Ann Neurol 1981;10:122-126. [Medline]

38. Smith CM, Swash M. Possible biochemical basis of memory disorder in Alzheimer disease. Ann Neurol 1978;3:471-473. [Medline]

39. Gauthier S, Emre M, Farlow MR, et al. Strategies for continued successful treatment of Alzheimer’s disease: switching cholinesterase inhibitors. Curr Med Res Opin 2003;19:707-714. [Medline]

40. Voisin T, Reynish E, Portet F, et al. What are the treatm ent options for patients with severe Alzheimer’s disease? CNS Drugs 2004;18:575-583. [Medline]

41. Hynd MR, Scott HL, Dodd PR. Glutamate-mediated excitotoxicity and neurodegeneration in Alzheimer’s disease. Neurochem Int 2004;45:583-595. [Medline]

42. Reisberg B, Doody R, Stoffler A et al. Memantine in moderateto - severe Alzheimer’s disease. N Engl J Med 2003;348:1333-1341. [Medline]

43. Tariot PN, Farlow MR, Grossberg GT et al. Memantine treatment in patients with moderate to severe Alzheimer disease already receiving donepezil: a randomized controlled trial. JAMA 2004;291:317-324. [Medline]

44. Danysz W, Parsons CG, Mobius HJ , Stoffler A, Quack G. Neuroprotective and symptomatological action of memantine relevant for Alzheimer’s disease: a unified glutamatergic hypothesis on the mechanism of action. Neurotoxicity Res 2000;2:85-97. [Medline]

45. Danysz W, Zajaczkowski W, Parsons CG. Modulation of learning processes by ionotropic glutamate receptor ligands. Behav Pharmacal 1995;6:455-474. [Medline]

46. Parsons CG, Danysz W, Quack G. Glutamate in CNS disorders as a target for drug development. An update. Drug News Perspect 1998;11:523-569. [Medline]

47. Fischer P-A, Jacobi P, Schneider E, et al. [Effects of intravenous administration of memantine in Parkinsonian patients]. Arzneimittelforschung 1977;27:1487-1489. [Medline]

48. Rothman SM, Olney JW. Excitotoxicity and the NMDA receptor. Trends Neurosci 1987;10:299-302.

49. Choi, DW. Calcium: still center-stage in hypoxicischemic neuronal death. Trends Neurosci 1995;18:58-60. [Medline]

50. Li S, Mallory M, Alford M, Tanaka S, Masliah, E. Glutamate transporter alterations in Alzheimer disease are possibly associated with abnormal APP expression. J Neuropathal Exp Neurol 1997;56:901-911. [Medline]

51. Klegeris A, McGeer PL. Beta-amyloid protein enhances macrophage production of oxygen free radicals and glutamate. J Neurosci Res 1997;49:229-235. [Medline]

52. Giulian D., Haverkamp LJ, Li L, et al. Senile plaques stimulate m i c roglia to release a neurotoxin found in Alzheimer brain. Neurochem Int 1995;27:119-137. [Medline]

53. Goodwin JL, Demura E, Cunnick JE. Microglial release of nitric oxide by the synergistic action of beta-amyloid and IFN-gamma. Brain Res 1995;692:207-214. [Medline]

54. Brorson JR, Bindokas YP, Iwama I, Marcuccilli CJ, Chisholm JC, Miller RJ. The Ca2+ influx induced by beta-amyloid peptide 25-35 in cultured hippocampal neurons results from network excitation. J Neurobiol 1995;26:325-338. [Medline]

55. Mattson MP, Cheng B, Davis D, Bryant K, Lieberburg L, Rydel RE. Beta-amyloid peptides destabilize calcium homeostasis and render human cortical neurons vulnerable to excitotoxicity. J Neurosci 1992;12:376-389. [Medline]

56. Wu J, Anwyl R, Rowan MJ. Beta-amyloid selectively augments NMDA receptor-mediated synaptic transmission in rat hippocampus. Neuroreport 1995;6:2409-2413. [Medline]

57. Cullen WK, Wu J, Anwyl R, Rowan MJ. Beta-amyloid produces a delayed NMDA receptor-dependent reduction in synaptic transmission in rat hippocampus. Neuroreport 1996;8:87-92. [Medline] 

58. Braak H, Braak E, Bohl J. Staging of Alzheimer related cortical destruction. Eur Neurol 1993;33:403-408. [Medline]

59. Francis PT, Sims NR, Procter AW, Bowen DM. Cortical pyramidal neurone loss may cause glutamatergic hypoactivity and cognitive impairment in Alzheimer’s disease: investigative and therapeutic perspectives. J Neurochem 1993;60:1589-1604. [Medline]

60. Mortimer JA, van Duijn CM, Chandra V, et al. Head trauma as a risk factor for Alzheimer’s disease: a collaborative re-analysis of case control studies. Int J Epidemiol 1991;20(Suppl. 2):S28-S35. [Medline]

61. Mendez MF, Underwood KL, Zander BA, Mastri AR, Sung JH, Frey WH. Risk factors in Alzheimer’s disease: a clinicopathologic study. Neurology 1992;42:770-775. [Medline]

62. Parsons C, Danysz W, Quack G. Glutamate in CNS Disorders as a target for a drug development: an update. Drug News Perspect 1998;11:523-569. [Medline]

63. Ditzler K. Efficacy and tolerability of memantine in patients with dementia syndrome: a double-blind, placebo controlled trial. Arzneimittelforschung 1991;41:773-780. [Medline]

64. Gortelmeyer R, Erbler H. Memantine in the treatment of mild to moderate dementia syndrome: a double-blind placebocontrolled study. Arzneimittelforschung 1992;42:904-913. [Medline]

65. Pantev M, Ritter R, Gortelmeyer R. Clinical and behavioural evaluation in long-term care patients with mild to moderate dementia under memantine treatment. Zeitschrift für Gerontopsychologie und -psychiatrie 1993;6:103-117.

66. Grossmann W, Schutz W. Memantine and neurogenic bladder dysfunction in spastic conditions. Arzneimittelforschung 1982;32:1273-1276. [Medline]

67. Miltner FO. [Valve of symptomatic therapy with memantine in cerebral coma. 1. Correlation with coma stages and EEG spectral patterns]. Arzneimittelforschung 1982;32:1268-1270. [Medline]

68. Winblad B, Mobius HJ, Stoffler A. Glutamate receptors as a target for Alzheimer’s disease: are clinical results supporting the hope? J Neural Transm 2002;62:217-225. [Medline]

69. Doraiswamy PM. Non-cholinergic strategies for treating and preventing Alzheimer’s disease. CNS Drugs 2002;16:811-824. [Medline]

70. Ambrozi L, Danieczyk W. Treatment of impaired cerebral function in psychogeriatric patients with memantine: results of a phase II double-blind study. Pharmacopsychiatry 1988;21:144-146. [Medline]

71. Chen H-S, Pellegrini JW, Aggarwal SK, et al. Open-channel block of N-methyl-D-aspartate (NMDA) responses by memantine: therapeutic advantage against NMDA receptor-mediated neurotoxicity. J Neurosci 1992;12:4427-4436. [Medline]

72. Chen H-S, Lipton SA. Mechanism of memantine block of NMDA-activated channels in rat retinal ganglion cells: uncompetitive antagonism. J Physiol 1997;499:27-46. [Medline]

73. Rohrbacher J, Bijak M, Misgeld D. Suppression by memantine and amantadine of synaptic excitation intrastriatally evoked in rat neostriatal slices. Neurosci Lett 1994;182:95-98. [Medline]

74. Frankiewicz T, Parsons CG. Memantine restores long term potentiation impaired by tonic N-methyl-D-aspartate (NMDA) receptor activation following reduction of Mg2+ in hippocampal slices. Neuropharmacology 1999;38:1253-1259. [Medline]

75. Blanpied TA, Boeckman FA, Aizenman E, et al. Trapping channel block of NMDA-activated responses by amantadine and memantine. J Neurophysiol 1997;77:309-323. [Medline]

76. Bormann J. Memantine is a potent blocker of N-methyl-D-aspartate (NMDA) receptor channels. Eur J Pharmacol 1989;166:591-592. [Medline]

77. Parsons CG, Gruner R, Rozental J, et al. Patch clamp studies on the kinetics and selectivity of N-methyl-D aspartate recepto r antagonism by memantine (1-amino-3,5- dimethyladamantan). Neuropharmacology 1993;32:1337-1350. [Medline]

78. Krieglstein J, Lippert K, Poch G. Apparent independent action of nimodipine and glutamate antagonists to protect culture d neurons against glutamate-induced damage. Neuropharmacology 1996;35:1737-1742. [Medline]

79. Erdo SL, Schafer M. Memantine is highly potent in protecting cortical cultures against excitotoxic cell death evoked by glutamate and N-methyl-D-aspartate. Eur J Pharmacol 1991;198:215-217. [Medline]

80. Weller M, Finiels-Marlier F, Paul SM. NMDA receptor-mediated glutamate toxicity of cultured cerebellar, cortical and mesencephalic neurons: neuroprotective properties of amantadine and memantine. Brain Res 1993;613:143-148. [Medline]

81. Pellegrini JW, Lipton SA. Delayed administration of memantine prevents N-methyl-D-aspartate receptor-mediated neurotoxicity. Ann Neurol 1993;33:403-407. [Medline]

82. Wenk GL, Zajaczkowski W, Danysz W. Neuroprotection of acetylcholinergic basal forebrain neurons by memantine and neurokinin B. Behav Brain Res 1997;83:129-133. [Medline]

83. Rao VL, Dogan A, Todd KG, Bowen KK, Dempsey RJ. Neuroprotection by memantine, a non-competitive NMDA receptor antagonist after traumatic brain injury in rats. Brain Res 2001;911:96-100. [Medline]

84. Gorgülü A, Kins T, Cobanoglu S, et al. Reduction of edema and infarction by memantine and MK-80l after focal cerebral ischaemia and reperfusion in rat. Acta Neurochir (Wien) 2000;142:1287-1292. [Medline]

85. Dogan A, Eras MA, Rao VLR, et al. Protective effects of memantine against ischemia-reperfusion injury in spontaneously hypertensive rats. Acta Neurochir (Wien) 1999;141:1107-1113. [Medline]

86. Seif el Nasr M, Peruche B, Rossberg C, et al. Neuroprotective effect of memantine demonstrated in vivo and in vitro. Eur J Pharmacol 1990;185:19-24. [Medline]

87. Heim C, Sontag KH. Memantine prevents progressive functional neurodegeneration in rats. J Neural Transm 1995;46(Suppl 1):117-130. [Medline]

88. Stieg PE, Sathi S, Warach S, et al. Neuroprotection by the NMDA receptor-associated open-channel blocker memantine in a photothrombotic model of cerebral focal ischemia in neonatal rat. Eur J Pharmacol 1999;375:115-120. [Medline]

89. Keilhoff G, Wolf G. Memantine prevents quinolinic acidinduced hippocampal damage. Eur J Pharmacol 1992;219:451-454. [Medline]

90. H. Lundbeck AlS. Summary of product characteristics. Ebixa (memantine hydrochloride). H. Lundbeck AIS, Valby, Denmark 2002:1-13.

91. Wenk GL, Quack G, Moebius HJ, et al. No interaction of memantine with acetylcholinesterase inhibitors approved for clinical use. Life Sci 2000;66:1079-1083. [Medline]

92. Periclou A, Ventura D, Sherman T, et al. Pharmaco kinetic study of memantine and donepezil in healthy young subjects [poster no. 32]. 26th Annual Meeting of the American Medical Directors Association; 2003.

93. Reisberg B, Doody R, Stamer A, et al. A randomized, placebo-controlled study of memantine, an uncompetitive NMDA antagonist, in patients with moderate to severe Alzheimer’s disease. N Engl J Med 2003;348;1333-1341. 

94. Reisberg B, Mobius HJ, Stomer A, et al. Long-term treatment with the NMDA antagonist memantine: results of a 24-week, open-label extension study in moderately severe to severe Alzheimer’s disease [abstract]. 8th International Conference on Alzheimer’s Disease and Related Disorders, Stockholm 2000.

95. Winblad B, Poritis N. Memantine in severe dementia: results of the 9M-Best Study (benefit and efficacy in severely demented patients during treatment with memantine). Int J Geriatr Psychiatry 1999;14:135-146. [Medline]

96. Hartmann S, Mobius H. Tolerability of memantine in combination with cholinesterase inhibitors in dementia therapy. Int Clin Psychopharmacol 2003;18:81-85. [Medline]

97. Bleich S, Wiltfang J, Kornhuber J, Reisberg B, Doody R, Möbius HJ. Memantine in moderate-to-severe Alzheimer’s disease. N Engl J Med 2003;349:609-610. [Medline]

98. Tariot P, Farlow M, Grossberg G, et al. Memantine/donepezil dual therapy is superior to placebo/donepezil therapy for treatment of moderate to severe Alzheimer’s disease. Poster presented at the American Psychiatric Association Annual Meeting, San Francisco, USA, 2003.

99. Tariot PN, Farlow MR, Grossberg GT, Graham SM, McDonald S, Gergel I. Memantine treatment in patients with moderate to severe Alzheimer disease already receiving donepezil: a randomized controlled trial. JAMA 2004;291:317-324. [Medline]

100. Knopman D. Where we stand with symptomatic treatments for Alzheimer disease. American Academy of Neurology, 2005 San Francisco: 1KP.003;1-34.

101. Pomara N, Peskind ER, Potkin SG, et al. Memantine monotherapy is effective and safe for the treatment of mild to moderate Alzheimer’s disease: a randomized controlled trial. Neurobiol Aging 2004;25:S1-S32.

102. Kumar S. Memantine: pharmacological properties and clinical uses. Neurology India 2004;52:307-309. [Medline]

103. Schmitt F, van Dyck C, Wichems C, Olin J; for the Memantine-MD-02 Study Group. Cognitive response to memantine in moderate to severe Alzheimer disease patients a l ready receiving donepezil: an exploratory reanalysis. Alzh Dis Ass Dis 2006;20:255-262. [Medline]

104. Jones RW, Bayer A, Inglis F, et al. Safety and tolerability of once-daily versus twice-daily memantine: a randomised, double-blind study in moderate to severe Alzheimer’s disease. Int J Geriatr Psychiatr 2007;22:258-262. [Medline]

105. Winblad B, Jones RW, Wirth Y, et al. Meta-analysis of clinical trial data with memantine in moderate to severe Alzheimer’s disease. Poster presented at EFNS, September, 2006.

106. Astrup J, Siesjo BK, Symon L. Thresholds in cerebral ischemia- the ischemic penumbra. Stroke 1981;12:723-725. [Medline]

107. Moretti R (Ed.). Subcortical vascular dementia. NOVA SCIENCE Publishers Inc. ed. New York, 2006.

108. Lind B, Snyder J, Safar P. Total brain ischemia in dogs. Cerebral physiologic and metabolic changes after 15 minutes of circulatory arrest. Resuscitation 1975;4:97-113. [Medline]

109. Nemoto E, Hossmann K-A, Cooper H. Postischemic hypermetabolism in cat brain. Stroke 1981;12:666-676. [Medline]

110. Sterz F, Leonov Y, Safar P, et al. Multifocal cerebral blood flow by Xe-CT and global cerebral metabolism after prolonged cardiac arrest in dogs. Reperfusion with open-chest CPR of cardiopulmonary bypass. Resuscitation 1992;24:27-47. [Medline]

111. Oku K, Kuboyama K, Safar P, et al. Cerebral and systemic arteriovenous oxygen monitoring after cardiac arrest. Inadequate cerebral oxygen delivery. Resuscitation 1994;27:141-152. [Medline]

112. Wiesniewski H, Pluta R, Lossinsky A, et al. Ultrastructural studies of cerebral vasculature spasm after cardiac arrest-related global cerebral ischemia in rats. Acta Neuropathol (Berl) 1995;90:432-440. [Medline]

113. Spatz M, Yasuma Y, Strasser A, et al. Cerebral postischemic hypoperfusion is mediated by ETA receptors. Brain Res 1996;726:242-246. [Medline]

114. Graffagnino C. The pathophysiology of anoxic encephalopathy. American Academy of Neurology (AAN), San Francisco 2004;2PC-001:545-549.

115. Iijima T, Mies G, Hossmann K-A. Repeated negative DC deflections in rat cortex following middle cerebral art e ry occlusion are abolished by MK-801. Effect on volume of ischemic injury. J Cereb Blood Flow Metab 1992;12:727-733. [Medline]

116. Meldrum B. Possible therapeutic applications of antagonists of excitatory amino acid neurotransmitters. Clin Sci 1985;68:113-122. [Medline]

117. Chen J, Graham S, Chan P, et al. bcl-2 is expressed in neurons that survive focal ischemia in the rat. Neuro Report 1995;6:394-398. [Medline]

118. Nitatori T, Sato N, Waguri S, et al. Delayed neuronal death in CA1 pyramidal cell layer of the gerbil hippocampus following transient ischemia is apoptosis. J Neurosci 1995;15:1001-1011. [Medline]

119. Mortimer JA, VanDuijn CM, Chandra V, et al. Head trauma as a risk factor for Alzheimer’s disease: a collaborative re-analysis of case control studies. Int J Epidemiol 1991;20(Suppl. 2):S28-S35. [Medline]

120. Orgogozo JM, Rigaud AS, Stoffler A, Mobius HJ, Forette F. Efficacy and safety of memantine in patients with mild to moderate vascular dementia: a randomized, placebo-controlled trial (MMM 300). Stroke 2002;33:1834-1839. [Medline]

121. Lipton SA, Chen HS. Paradigm shift in neuroprotective drug development: clinically tolerated NMDA receptor inhibition by memantine. Cell Death Differ 2004;11:18-20. [Medline]

122. Wilcock G, Mobius HJ, Stoffler A; MMM 500 group. A double-blind, placebo-controlled multi centre study of memantine in mild to moderate vascular dementia (MMM500). Int Clin Psychopharmacol 2002;17:297-305. [Medline]

123. Jarvis B, Figgitt DP. Memantine. Drugs Aging 2003;20:465-476. [Medline]

124. McShane R, Areosa Sastre A, Minakaran N. Memantine for dementia. Cochrane Database Syst Rev 2006;(2):CD003154. [Medline]

125. Raina P, Santaguida P, Ismaila A, et al. Effectiveness of cholinesterase inhibitors and memantine for treating dementia: evidence review for a clinical practice guideline. Ann Intern Med 2008;148(5):379-397. [Medline]