Introduction

Excitotoxicity is known to contribute to neuropathology in acute neurological insults such as traumatic brain injury (TBI) and stroke, as well as chronic and insidious neurodegenerative disorders such as dementia of the Alzheimer’s type (AD). Although experimental models have provided compelling evidence that N -methyl- d -aspartate (NMDA) antagonism confers neuroprotection from excitotoxic insults, initial clinical trials of NMDA antagonists in TBI and stroke proved disappointing, at least in part owing to unacceptable psychotomimetic side effects. More recently, the NMDA antagonist memantine has demonstrated therapeutic efficacy in dementia, with a relatively benign side effect profile. It is currently Food and Drug Administration (FDA) approved for use in moderate to severe AD and has been investigated in clinical trials for several other neurological and psychiatric disorders.

Two conceptually distinct uses for memantine in TBI can be proposed for consideration and empirical study. The first involves use as a neuroprotective agent in the acute phase of TBI. This potential application is underpinned by substantial support from preclinical research, which has shown that memantine confers neuroprotection from traumatic, hypoxic-ischemic, and chemical lesions, ultimately preserving memory function, in experimental models ( ). The second potential use of memantine involves longer-term administration in postacute or chronic TBI for cognitive improvement. This indication would be akin to its use for symptomatic improvement during the gradual neurodegeneration seen in AD, purportedly by improving the signal-to-noise ratio at the NMDA receptor ( ). Currently, this potential application in TBI is more speculative, with limited theoretical or empirical support in the literature specific to TBI. Nonetheless, the actual efficacy of this approach in TBI remains an empirical question.

The primary purpose of this chapter is to review theoretical rationales and empirical findings suggesting potential utility of memantine in TBI. Fundamental principles surrounding excitotoxicity, theoretical rationales for NMDA antagonist use, and empirical studies with early NMDA antagonists will be briefly reviewed to provide background conceptual information and shed light on the historical context. NMDA receptor structure, function, and pharmacologic manipulation have been investigated intensively over the past several decades; as such, the depth and breadth of the literature cannot truly be conveyed in the course of a cursory review. Once the conceptual context has been established, pharmacological properties, proposed mechanisms of action, and empirical studies relating to memantine in particular will then be summarized. Findings relevant to ischemia or stroke will sometimes be included, given that excitotoxicity is an important contributor to secondary neurological injury in both ischemia and TBI, that the pathophysiology overlaps between these conditions (particularly with respect to excitotoxic cascades), and this author’s belief that ischemia exacerbates excitotoxicity in TBI. Finally, conclusions will be drawn and future directions for research addressed.

NMDA Receptors: Crucial for Memory, But Potentially Excitotoxic

Glutamate, an excitatory amino acid, is the endogenous ligand for metabotropic and ionotropic glutamate receptors in the brain. Activation of metabotropic glutamate receptors causes mobilization of calcium from intracellular stores, while activation of ionotropic glutamate receptors causes influx of extracellular sodium, potassium, and calcium into the cell ( ). Abnormally elevated levels of extracellular glutamate cause calcium influx and intracellular calcium overload, ultimately leading to cell death ( ). The neurotoxicity of excessive excitatory amino acids such as glutamate was first demonstrated by Olney, who coined the term “excitotoxicity” for this phenomenon ( ). Since that time, the existence and deleterious effects of excitotoxicity have been extensively demonstrated both in vivo and in vitro, yielding compelling evidence that NMDA-mediated excitotoxicity contributes to neuropathology and hippocampal cell death in models of ischemic stroke, dementia, and TBI ( ).

There are three types of ionotropic glutamate receptors: N -methyl- d -aspartate (NMDA), 2-amino-3-(3-hydroxy-5-methylisox-azol-4-yl) propionate (AMPA), and kainate. Although various sources and mechanisms can contribute to excess intracellular calcium, calcium influx via the NMDA receptor in particular appears to be a potent driver of downstream signaling pathways leading to cell death ( ). The NMDA receptor is a heteromeric, transmembrane, protein complex consisting of four subunits, typically two NR1 subunits and two NR2 subunits (NR2A-D), which form the ion channel ( ). At the physiologic resting membrane potential, the ion channel is blocked by Mg ²⁺ . Opening of the ion channel requires depolarization to release the block by Mg ²⁺ and concurrent binding of two agonists: glycine to the NR1 subunit and glutamate to the NR2 subunit. Opening of the ion core permits the entry of monovalent cations such as Na ⁺ and divalent cations such as Ca ²⁺ ( ). Neurotoxic cascades arising from calcium overload due to trauma or ischemia include uncoupling of mitochondrial electron transfer from ATP synthesis and activation of enzymes such as calpains, protein kinases, nitric oxide synthase, and endonucleases, ultimately causing production of reactive oxygen species, cytoskeletal disruption, mitochondrial dysfunction, DNA fragmentation, cell swelling, malfunction of the endoplasmic reticulum, and induction of apoptosis ( ).

Numerous pharmacologic agents with NMDA antagonistic properties are able to block the action of NMDA receptors, including the high-affinity uncompetitive antagonist dizocilpine (MK-801), the competitive antagonist Selfotel acting at the glutamate binding site, and the high-affinity uncompetitive ion channel blocker Aptiganel ( ). Putative targets for mitigation of NMDA-mediated excitotoxicity are shown in Fig. 16.1 . These include drugs that act directly upon binding sites on the NMDA receptor subunits or channel, as well as upstream or downstream modulators of NMDA function ( ). It is important to note that early NMDA antagonists entered into TBI and stroke clinical trials were potent competitive agonists acting directly at the glutamate site (ie, Selfotel) and high-affinity channel blockers (ie, Aptiganel), which yielded unacceptable psychotomimetic side effects ( ). In contrast, memantine is an open-channel blocker, which yields qualitatively distinct pharmacological properties and a more benign side effect profile.

Potential sites for drug action within the NMDA receptor protein complex. Extracellular sites include the glycine (Gly) binding site on NR1 subunits, the glutamate (Glu) binding site on NR2 subunits, and the binding sites within the channel pore which overlap with the site for magnesium binding (Mg ²⁺ ). d -serine is an endogenous coagonist at the glycine-binding site. The sites of action of previously studied or currently used drugs are indicated. NR2 subunits also contain sites of action for polyamines, zinc, and protons. The site of action of NR2B-selective antagonists, such as traxoprodil, overlaps with a zinc-binding site on NR2B subunits. Intracellular targets include signaling molecules such as kinases, phosphatases, other enzymes, and scaffold proteins. These molecules are upstream modulators of NMDA receptor function or downstream effectors of NMDA receptor activity.

From Kalia, L. V., Kalia, S. K., & Salter, M. W. (2008). NMDA receptors in clinical neurology: excitatory times ahead. Lancet Neurology , 7 (8), 742–755. http://dx.doi.org.easyaccess1.lib.cuhk.edu.hk/10.1016/S1474-4422(08)70165-0 .

Although pathological overactivation of NMDA receptors induces excitotoxicity, their function is crucial for a variety of normal physiological processes, including synaptic plasticity and learning. In particular, NMDA receptors appear to permit coincidence detection and induce long-term potentiation (LTP), an activity-dependent strengthening of synaptic connections. Hippocampal LTP has been the most widely studied model of LTP and is viewed as the electrophysiological basis of episodic memory ( ). Hence, although excessive and protracted NMDA activation is excitotoxic, transient, and physiologically appropriate activation of NMDA receptors is required for adaptive neuroplasticity and learning.

NMDA Antagonists Vis-à-Vis Excitotoxicity in TBI and Ischemia

Excitotoxicity is widely recognized as an important contributor to secondary injury in TBI, as well as acute stroke and hypoxic-ischemic injury. Indeed, the overlap between pathophysiological mechanisms in TBI and ischemia has been explicitly delineated, with the implication that similar neuroprotective strategies may be beneficial in both disorders ( ). Following acute TBI in humans, glutamate levels are elevated in human CSF and microdialysate, which persists for at least 7 days after injury and correlates with outcome. The increases in glutamate are most dramatic for patients with secondary ischemic events caused by low cerebral perfusion pressure (CPP), high intracranial pressure (ICP), and/or hypoxemic events ( ). These secondary ischemic insults are known to powerfully increase morbidity and mortality secondary to TBI. Multiple factors in TBI can contribute to focal or global ischemia in TBI, including systemic hypotension secondary to hemorrhage leading to reduced CPP, increased ICP secondary to mass lesions or vasogenic edema, focal compression underlying mass lesions such as extra-axial hematomas, and herniation syndromes (eg, infarcts in the posterior cerebral artery distribution secondary to uncal herniation).

Although the initial sharp rise in glutamate levels occurs soon after mechanical injury to the brain, it is important to recognize that secondary insults, including ischemia and intracranial hypertension, can and do occur in the days and weeks following the initial injury, often in the ICU ( ). This potentially provides an opportunity for therapeutic intervention. Brain atrophy (including hippocampal atrophy) and Wallerian degeneration also appear to progress over the first several months after TBI ( ), further reinforcing the possibility of intervention during progressive secondary degeneration. Because excitotoxicity is maximal in TBI patients with secondary ischemic events, it could be hypothesized that NMDA antagonists would be most helpful for patients with superimposed ischemia.

An extensive body of literature has confirmed the potent neuroprotective effects of NMDA antagonists in various experimental models of neuronal injury. For example, NMDA antagonists can reduce hippocampal cell death and cortical damage, while improving recovery and preserving memory, in animal models of TBI ( ). Given promising results for NMDA antagonists in preclinical models of neurologic insults, clinical trials of NMDA antagonists were initiated in stroke and TBI ( ). Some of these drugs were competitive NMDA antagonists acting at the glutamate site, while others were high-affinity uncompetitive blockers. Unacceptable psychotomimetic side effects for these drugs, such as hallucinations and coma, were identified during clinical trials for stroke. Clinical trials in TBI were, therefore, abandoned before meaningful results were obtained ( ). In retrospect, pharmacologic agents that potently and completely blocked the function of NMDA receptors were ill-advised, given their crucial role in normal function and cognition. From the therapeutic perspective, chronic, persistent, or extreme overactivation of NMDA receptors that leads to excitotoxcity must be prevented, while transient and physiologically appropriate responses of NMDA receptors must be preserved to permit normal learning and memory processes.

Memantine: An NMDA Antagonist Without the Undesirable Side Effects

Memantine is an adamantane derivative similar to amantadine, a drug with antiviral and antiparkinsonian properties ( ). Memantine has now been characterized as an uncompetitive, moderate affinity, open-channel NMDA antagonist, which demonstrates strong voltage dependency and rapid blocking and unblocking kinetics ( ). It is safe and generally well tolerated, without the problematic side effect profile of early NMDA antagonists. It has been approved by the FDA for use in moderate-to-severe AD, in which it attenuates decline in memory, language, global outcome, behavior, and activities of daily living ( ).

Several of memantine’s pharmacodynamics properties have been proposed to account for its benign side effect profile relative to other NMDA antagonists, such as MK-801. First, the fact that it acts as an open-channel blocker, rather than a direct antagonist at the glutamate or glycine sites, suggests that it will only block NMDA receptor function when the channel is opened by agonist binding (ie, in the presence of glutamate). In theory, this suggests that memantine will limit excitotoxicity secondary to excess glutamate, while leaving intact the function of unaffected tissue in the absence of glutamate because the drug is unable to access the binding site within the ion channel ( ). Furthermore, the uncompetitive nature of memantine suggests that the proportion of the current blocked by the drug should increase in response to higher levels of glutamate, such as those seen in acute ischemia, while the block should be minimal when agonist levels are low ( ). Additionally, memantine demonstrates rapid blocking and unblocking kinetics at low concentrations. This should permit maintenance of normal physiological NMDA receptor function, such as induction of LTP, when appropriate. By comparison, MK-801 maintains the block and does not permit normal physiologic function ( ).

The mechanism of action of memantine is not fully understood. One early hypothesis was formulated to account for the symptomatic improvement observed with memantine (eg, improved memory), rather than prevention of ongoing chronic neurodegeneration or neuroprotection from an acute insult. This hypothesis posited that hyperglutamatergic states increased synaptic “noise,” thereby limiting the capacity to detect true physiological signal, effectively decreasing the signal-to-noise ratio. According to this line of reasoning, memantine improved the signal-to-noise ratio by decreasing background glutamatergic noise, thereby ameliorating NMDA receptor function ( ).

A more recent theory suggests that activation of extrasynaptic NMDA receptors initiates calcium influx that triggers proapoptotic mechanisms, ultimately leading to cell death. In contrast, activation of synaptic NMDA receptors is implicated in induction of LTP required for plasticity; it triggers prosurvival pathways, including generation of neurotrophins and inhibition of apoptotic processes ( ). Memantine is believed to preferentially block overactivation of the extrasynaptic NMDA receptors, which elicit excitotoxic and proapoptotic mechanisms, while exerting lesser impact on the synaptic NMDA receptors responsible for LTP and generation of neurotrophins ( ). In this context, it is also important to note that ischemia appears to preferentially increase extrasynaptic NMDA receptor currents, thereby promoting prodeath signaling ( ).

More specifically, activation of synaptic NMDA receptors has been proposed to induce a persistent acquired neuroprotective action via several mechanisms. Calcium influx from synaptic NMDA receptors is augmented by internal stores of calcium and ultimately alters gene transcription via nuclear calcium signaling. This mechanism involves cyclic AMP response element binding protein (CREB) and brain-derived neurotrophic factor (BDNF), which have been implicated in neuroprotection against excitotoxicity and apoptosis, synaptic plasticity, neurogenesis, and memory. Synaptic NMDA receptor activation also suppresses apoptosis by transcription of Puma, which limits the activity of caspases such as caspase-9. It further suppresses expression of transcription factors such as forkhead box O (FOXO), which promote cell death in response to excitotoxic and other insults. It also boosts antioxidant defenses. In contrast, prodeath signaling pathways appear to be initiated by activation of extrasynaptic NMDA receptors, which involves CREB shut-off, inactivation of the ERK1/2 pathway, activation of the FOXO pathway, and calpain activation ( ).

The reasons for memantine’s preferential blocking of extrasynaptic NMDA receptors remain unclear, as do the reasons for differential engagement of prodeath and prosurvival signaling. It is possible that synaptic and extrasynaptic receptors are linked to different signaling complexes or that different subunit compositions (eg, GluN2A vs GluN2B) are associated with different molecular cascades. Additionally, it is possible that synaptic receptors are typically activated by transient synaptic release of glutamate, while extrasynaptic receptors are typically activated by chronic elevations of glutamate outside the synaptic cleft ( ).

Memantine: Formulations, Dosages, Contraindications, and Side Effects

Memantine hydrochloride has been marketed as Namenda by Forest Laboratories in the United States and as Ebixa by Lundbeck in Europe, although generic formulations of the tablets are now available. Memantine is currently available in three formulations: oral solution, tablet, and extended-release capsule. The recommended maintenance dose of memantine in AD is 10 mg twice daily for oral solution or tablet, or 28 mg once daily for the extended-release capsule. The manufacturer recommends gradual titration, beginning at one-quarter of the full dose, with incremental increases of the same dose at weekly intervals, until the target dose is reached. At these doses, the most common side effects include dizziness, headache, confusion, and constipation.

According to Forest’s package insert for Namenda in the United States, peak concentrations of memantine are reached within 3–7 h of oral ingestion for tablets and its terminal elimination half-life is about 60–80 h. The drug undergoes partial hepatic metabolism, without significant involvement of the hepatic microsomal CYP450 enzyme system. Renal clearance involves active tubular secretion moderated by pH-dependent tubular reabsorption. The clearance of memantine was reduced by about 80% under conditions of alkaline urine (pH > 8). As such, dosage reduction should be considered in conditions that alkalinize the urine (eg, urinary tract infections, use of carbonic anhydrase inhibitors, or administration of sodium bicarbonate) since these could lead to increased drug accumulation in the plasma. According to manufacturer recommendations, dosage modification is not required for patients with mild to moderate renal failure, but the dosage should be halved for patients with several renal failure, as defined by creatinine clearance of 5–29 mL/min. A dosage adjustment is not recommended for patients with mild to moderate hepatic failure, but the drug should be used with caution in patients with severe renal failure.

Caution would also be indicated in treating pregnant patients, nursing mothers, children, and patients with concurrent seizure disorder since safety and efficacy have not been established in well-validated studies involving these populations. Concurrent use of other medications with NMDA antagonistic properties (eg, amantadine, ketamine, dextromethorphan) should probably be avoided or used with caution. Co-administration of drugs that rely on the same renal cationic system for elimination of memantine could affect plasma levels of both drugs.