Memantine: updates from the past decade and implications for future novel therapeutic applications

Abstract

This review provides an update on advances during the past decade in our knowledge of the pharmacology and clinical use of memantine. It covers aspects related to preclinical research, such as new findings about the mechanism of action, and the efficacy of memantine in animal models of various diseases beyond dementia. In particular, new emerging indications since the publication of our latest review almost a decade ago were investigated. Where possible, preclinical findings were reinforced by data obtained in clinical studies in patients with Alzheimer’s disease, as well as in other indications reported by researchers who initiated off-label studies. This comprehensive narrative review demonstrates that memantine continues to show robust efficacy in its primary indication of Alzheimer's disease. However, while numerous potential new indications have emerged from preclinical studies, only a limited number show sufficient evidence to warrant further clinical investigation.

Introduction

This narrative review provides an overview of advances in our understanding of the pharmacology and clinical use of memantine hydrochloride (3,5-dimethyladamantan-1-amine) (Fig. 1) since ca. 2014, focusing on three key aspects:

1. New insights into memantine’s mechanism of action (MoA).

2. Update on the standard indication, Alzheimer’s disease (AD), based on preclinical and clinical findings.

3. New potential indications emerging from preclinical or clinical studies.

Fig. 1.

Chemical structure of memantine (3,5-dimethyladamantan-1-amine)

The time period for publications included in this review (after 2014) was chosen to coincide with the time since our last reviews on memantine were published (Parsons et al. 2013; Danysz and Parsons 2012; Rammes et al. 2008; Danysz and Parsons 2003; Parsons et al. 1999b). This update is not intended to repeat data already discussed in these previous reviews; however, salient information is reiterated in certain instances. In this review, preclinical studies or group of studies were cited for a given indication and the therapeutic relevance was judged based on the concentration/doses of memantine used. This is particularly relevant for potential new indications based on preclinical results. Studies which support previous data are emphasized, in addition to those which are completely novel. Studies supporting the use of memantine in AD are discussed first, followed by other potential indications.

Overview of memantine history and current marketing/indications status

Memantine is an uncompetitive N-methyl-D-aspartate (NMDA)-receptor antagonist that has been marketed in Germany for dementia since 1989 (Witt et al. 2004), and subsequently in several other countries as Akatinol Memantine^® for use in a broad range of indications such as cerebral disorders. In 2002, memantine was approved for the treatment of moderately severe to severe AD in the European Union under the trade names Axura® and Ebixa® by Merz Pharmaceuticals and H Lundbeck A/S, respectively. In November 2005, the label indication was extended to include moderate-to-severe AD, and memantine has been available in the USA as Namenda® (since 2003) and in Japan as Memary® (since 2011) for this indication. As of 2025, memantine had been approved in more than 88 countries and is available as immediate-release (tablets and oral solution) and extended-release (ER) formulations (Plosker 2015).

Currently, memantine is sold worldwide under nearly 100 trade names, and since the patent expired in 2015, it is now also available as a generic drug. The clinical studies that led to the approval of memantine were conducted using the originator formulation; however, currently there are several available generics of memantine, and their quality and level of regulatory control might vary.

Mechanism of action of memantine: focus on new findings

Knowledge of which targets are affected by therapeutically relevant plasma concentrations is crucial when clarifying the MoA of any given substance and when narrowing down potential new indications. Memantine, at therapeutic doses in humans (20–30 mg/day), produces a maximum plasma level of 1 µM and slightly lower concentrations in the brain cerebrospinal fluid (CSF) (Parsons et al. 2007; Danysz et al. 2000; Kornhuber and Quack 1995). This concentration is relevant for targets reached from the extracellular space e.g., receptors or transporters on cellular membranes. Based on animal studies, a brain extracellular fluid concentration of about 0.8–1.0 µM can be anticipated (Hesselink et al. 1999; More et al. 2008) and any protein showing memantine affinity up to this range should be considered as a potential target. In fact, there are only two clearly known targets fulfilling this criterion; the most likely is the NMDA receptor channel, but also 5-hydroxytryptamine type 3 (5-HT3) receptors (Rammes et al. 2001). However, due to lysosomal trapping, memantine is significantly concentrated intracellularly (Parsons et al. 2007; Danysz et al. 2000; Honegger et al. 1993; Keshavarz et al. 2020; Trapp et al. 2008), hence intracellular targets may be affected by memantine at a higher concentration range.

In vitro, in human cells in culture, up to 200-fold intracellular accumulation of memantine has been observed (Honegger et al. 1993). In rats, Misztal et al. showed ca. 40 µM brain levels at 1.2 µM plasma steady-state levels indicating a 36-fold accumulation (Misztal et al. 1996). However, it should be noted that drug binding in the rat brain is higher than in humans, hence a rat brain may accumulate lipophilic drugs at a higher concentration than a human brain (Gustafsson et al. 2019). In a single post-mortem human case study (Wesemann et al. 1980), only a 4-fold accumulation of memantine was found. However, in this study, total blood (not plasma) was compared with homogenates, potentially leading to an underestimation of the brain accumulation factor. Overall, no definitive statements can be made on the degree of memantine accumulation in the human brain, but this most likely lies in the range of 10- to 30-fold. Higher concentrations of memantine are therefore reached intracellularly versus extracellularly, and diverse targets may be affected that would not be affected if outside the cell (Fig. 2).

Fig. 2.

Major targets of memantine based on in vitro activity and concentrations achieved at therapeutic doses extracellularly or intracellularly (graphic presentation inspired by (Kornhuber et al. 1994)

In rats, treatment with memantine 6.27 mg/rat/day (ca. 23 mg/kg/day) as continuous infusion resulted in plasma levels of 1.03 µM and brain extracellular levels of 0.99 µM, as measured with brain microdialysis (More et al. 2008). At this plasma concentration, 33% NMDA receptor occupancy was observed in vivo (More et al. 2008). In ex vivo experiments, comparable data have been obtained (Tober 2007a, b; More et al. 2008; Lord et al. 2013). Hence, at therapeutic doses, only partial block of NMDA receptors can be expected which likely contributes to the favorable tolerability profile of memantine (Danysz et al. 1997; Parsons et al. 2007).

NMDA receptors

Memantine is a voltage-dependent, uncompetitive NMDA receptor antagonist which exhibits rapid blocking/unblocking kinetics (Parsons et al. 2007; Kornhuber et al. 1989). Numerous publications have confirmed these biophysical properties in different in vitro systems (Losi et al. 2006; Zhao et al. 2006; Wrighton et al. 2008; Puddifoot et al. 2009; Parsons et al. 2008; Parsons and Gilling 2014; Otton et al. 2011; Otton et al. 2009; McClymont et al. 2012; Kotermanski et al. 2009; Kotermanski and Johnson 2009; Gilling et al. 2007; Gilling et al. 2009; Emnett et al. 2013; Emnett et al. 2014; Giorgetti et al. 2010; Valverde et al. 2014; Heusler et al. 2015; Tari et al. 2023). The onset and offset kinetics of NMDA receptor blockade by memantine are clearly dependent on agonist concentration (Gilling et al. 2009; Gilling et al. 2007). Fast agonist concentration-dependent kinetic properties, in addition to the clear voltage-dependence of memantine, have been proposed to be important for the therapeutic tolerability of this compound in the treatment of AD (Gilling et al. 2009; Gilling et al. 2007; Parsons and Gilling 2007).

Memantine has been claimed to selectively target extra-synaptic “death” receptors which are mainly composed of NR2B subunits and coupled to different signaling pathways than the physiologically more relevant sub-synaptic receptors (Leveille et al. 2008; Zhao et al. 2006; Okamoto et al. 2009; Xia et al. 2010). Interestingly, in substantia nigra neurons in rat brain slices, memantine’s preference to inhibit extrasynaptic NMDA receptors was not shared by the high affinity NMDA blocker, (+)MK-801 (Wu and Johnson 2015). This feature of memantine may be of particular relevance for pathological processes in AD (Leveille et al. 2008; Bordji et al. 2010; Rammes et al. 2011; Mota et al. 2012). The moderate selectivity for extrasynaptic receptors probably has little to do with the subtype selectivity of memantine for NR2B receptors (Bresink et al. 1996; Parsons et al. 1999a), but rather is determined by moderate and prolonged membrane depolarization at these receptor loci and the strong voltage-dependency of memantine (Parsons et al. 2007; Danysz and Parsons 2012). In turn, memantine selectivity is between sustained versus transient activation of all receptors, whether extra- or sub-synaptic (Wroge et al. 2012) or both (Zhou et al. 2013). Memantine antagonism of NMDA receptors in hippocampal slices has also been reported to be stronger in principal excitatory neurons (e.g., pyramidal) than in inhibitory interneurons, an effect attributed to specific differences in the molecular make-up of the NMDA receptors in these two neuron types (Martina et al. 2013). This conclusion, however, contrasts somewhat with the claim that tonically active NMDA receptors contribute more prominently to intrinsic excitation of gamma-aminobutyric acid (GABA)ergic interneurons than pyramidal cells, and that readily reversible NMDA receptor open-channel blockers like memantine preferentially inhibit such tonically active receptors (Riebe et al. 2015).

The 20% partial untrapping previously reported for memantine at rodent receptors (Blanpied et al. 1997) was confirmed at the human receptors GluN1/GluN2A, whereas ketamine was confirmed to show little or no partial untrapping (Gilling et al. 2009). Johnson’s group extended their mechanistic studies on partial trapping, as reviewed previously, see Johnson et al. 2014 (Johnson et al. 2014), showing that a superficial site on NMDA receptors contributes to this phenomenon (Kotermanski et al. 2009).

The dominant negative NR3 subunits of the NMDA receptor are known to reduce the Ca²⁺ permeability and Mg²⁺ sensitivity of NR1/NR2 subunit-containing receptors. Open-channel blockade by memantine was somewhat less effective at NR1/NR2/NR3B receptors than at NR1/NR2/NR3A receptors (2-fold weaker at −75 mV) (McClymont et al. 2012). However, memantine did not block glycine-activated currents in HEK 293 cells expressing only NR1 and NR3 subunits in the absence of GluN2 subunits (Smothers and Woodward 2007). It has been suggested that hyperactivity of NR3A is involved in AD pathogenesis (Zhong et al. 2022). In this context, it is interesting that in adult/aging GluN3A (NR3A) knockout mice, memantine (10 mg/kg/day in drinking water for 3 months) starting at the preclinical stage prevented or decreased age-dependent changes such as elevated intracellular Ca²⁺, neuroinflammation, impaired synaptic integrity/plasticity, cognitive dysfunction, and neuronal loss (Zhong et al. 2022).

Higher concentrations of memantine (20–50 µM) have also been reported to directly interact with phosphoinositide lipids present in the inner monolayer of erythrocyte membranes, inducing the formation of stomatocytes (Zambrano et al. 2016) (Zambrano et al. 2018). It is possible that such physiochemical properties might underlie effects reported for amyloid precursor protein (APP) processing (Butler et al. 2011; Kanazirska et al. 2012) due to intracellular accumulation – see below.

Additionally, it is clear that Mg²⁺, memantine and (+)MK-801 all interact with the same, or similar, NMDA receptor channel binding site, and that attaining a new steady-state equilibrium between different blockers takes time (McKay et al. 2013). For example, both memantine and Mg²⁺ accelerate the unblock of (+)MK-801 in the presence of NMDA. This suggests a model whereby, upon dissociation from its binding site in the pore – although (+)MK-801 is normally able to re-bind—this process is antagonized by other uncompetitive NMDA receptor antagonists (McKay et al. 2013). Even when all NMDA receptors were preblocked by (+)MK-801, incubation of neurons with 100 µM NMDA in the presence of Mg²⁺ for 2.5 h eventually triggered relief of block by (+)MK-801 (McKay et al. 2013). Song and colleagues showed that memantine binds within the ion channel, promoting closure of the ion channel gate, and locates between the M3 helix crossing and the M2 pore loops to physically block ion permeation (Song et al. 2018). An interesting, if somewhat controversial paper, has claimed that spermidine and Ca²⁺, but not Na⁺, can permeate NMDA receptors in the presence of Mg²⁺ (100 µM) and that this effect is blocked by memantine (10µM) (Hirose et al. 2015). The effects of physiological concentrations of Mg²⁺ (1.5 mM) were apparently not addressed in this study.

The efficacy of memantine combined with its favorable tolerability has been discussed extensively in previous reviews (Danysz et al. 1997; Parsons et al. 2007). Memantine blocks the NMDA receptor channel transiently and this block is relieved during temporally and/or spatially convergent/co-operative activation of glutamatergic synapses – for example, during learning and memory processes. This occurs due to memantine’s uncompetitive NMDA receptor antagonist activity, with strong voltage dependency and rapid unblocking kinetics (Danysz et al. 1997; Parsons et al. 2007).

Dysfunction of post-synaptic neurons may lead to weaker NMDA channel blockade by Mg²⁺, caused by partial depolarization due to, e.g., an energy deficit. This may also trigger prolonged Ca²⁺ influx and neuronal loss (Danysz and Parsons 2003; Rogawski and Wenk 2003). Memantine block is more potent and less voltage dependent than that of Mg²⁺ and, in turn, may serve as an improved substitute for Mg²⁺ due to the monovalent charge (Parsons et al. 1993). Memantine is therefore more effective than Mg²⁺ in blocking tonic pathological activation of NMDA receptors at moderately depolarized membrane potentials. However, following transient but strong synaptic activation, memantine, like Mg²⁺, can unblock the NMDA channel. Thus, memantine suppresses tonic pathological synaptic “noise”, allowing the relevant physiological synaptic signal to be better detected e.g., during the learning process (Parsons et al. 1999b; Danysz and Parsons 2012; Danysz and Parsons 2003).

Wilcox and colleagues proposed that memantine, in addition to a hydrophilic path through the open channel gate, may also enter NMDA receptor channels through a hydrophobic path from plasma membrane to channel through a gated fenestration (Wilcox et al. 2022). Glasgow and colleagues investigated the role of the second binding site for memantine within NMDA receptor channels (Glasgow et al. 2018). Their data support the association of memantine with this second site, but this binding does not lead to channel inhibition. Once bound, memantine can dissociate and move to the extracellular solution, or bind fast to the deep site resulting in channel block (Glasgow et al. 2018).

Possible targets beyond NMDA receptors

5-HT3 serotonergic receptors

5-HT3 receptors (ligand-gated monovalent cation channels) regulate neurotransmitter release, particularly dopamine in mesolimbic pathways. Blockade of 5-HT3 receptors has been suggested/shown to improve cognitive deficits and depression (Rammes et al. 2008; Thompson and Lummis 2007). Memantine antagonizes 5-HT3 receptors with a potency similar to its effects at NMDA receptors. This effect is not voltage- or use-dependent, and there is no evidence for open channel blockade (Rammes et al. 2008). It is possible that memantine’s effects on 5-HT3 receptors contribute to its therapeutic efficacy and tolerability in AD by increasing cognitive enhancement (Wilde and Markham 1996).

AMPA receptors

Previous publications failed to reveal effects of memantine on AMPA receptors (Parsons et al. 1993); however, these studies were performed using calcium-impermeable AMPA receptors lacking auxiliary subunits (Jackson and Nicoll 2011). New findings from electrophysiological experiments carried out by Carrillo and colleagues showed that currents mediated by AMPA receptors permeable to Ca²⁺ expressed in HEK-293 cells, with the presence of auxiliary subunits, were inhibited by memantine at a half-maximal inhibitory concentration (IC₅₀) of 10–15 µM (Carrillo et al. 2024). The auxiliary subunits γ2 or γ8 were used to stabilize AMPA receptors in the open state and for receptor desensitization. Moreover, the effect of memantine seemed to be voltage dependent and IC₅₀ decreased to ca. 4 µM at −100mV. Single channel recordings support the idea of open channel block. This inhibition of AMPA receptors was not seen with other NMDA channel blockers such as ketamine or MK-801 (Carrillo et al. 2024). In whole-cell voltage-clamp recordings in cortical neuronal cultures, memantine inhibited excitatory synaptic currents. Experiments using cryo-electron microscopy of activated calcium-permeable-AMPA receptors showed that memantine interacts directly with the AMPA receptor Q/R site (at the glutamine residue at site 607). In genetically labeled spinal VGluT2 neurons of mice subjected to spared nerve injury, memantine inhibited calcium-permeable-AMPA receptors starting at 10 µM, however, full inhibition was not seen at a 5-fold higher concentration (Carrillo et al. 2024). It is not clear whether the potency of memantine shown in these experiments is sufficient to play a role in the therapeutic MoA; nevertheless, a supportive effect cannot be excluded.

GDNF/BDNF

Given the aforementioned intracellular accumulation of memantine, targets such as glial cell line-derived neurotrophic factor (GDNF) become relevant even if the effects seen in vitro are at concentrations above 1 µM. Exposure of C6 glial cells to memantine for 24 h (but not 8 h) induced an over 3-fold increase in GDNF release at ca. 10 µM concentration with EC₅₀ of 6.4 µM (Caumont et al. 2006). This was accompanied by an increase in respective messenger ribonucleic acid (mRNA) levels, as shown in a rat GDNF promoter luciferase reporter assay (Caumont et al. 2006). Given the intracellular accumulation of memantine, such concentrations may be achieved following therapeutic doses.

In rats treated with memantine (intraperitoneal [IP] 5–50 mg/kg) 4 h before sacrifice, an increase in brain-derived neurotrophic factor (BDNF) mRNA was found in hippocampal areas, the thalamus, hypothalamus, and dopaminergic regions such as the substantia nigra and ventral tegmental area (Marvanova et al. 2001). At the lowest dose of memantine, an increase in BDNF mRNA was seen in parietal, cingulate, and retrosplenial cortices. At 10 mg/kg, an increase was also seen in the hippocampal region. When BDNF was analyzed at the protein level by enzyme-linked immunosorbent assay (ELISA), only the highest memantine dose, 50 mg/kg, which is well above the therapeutic range, produced an increase. The same dose also increased trkB (BDNF receptor) levels. It is to be explored whether lower doses of memantine given repetitively to produce steady-state exposure for days or weeks result in increased BDNF protein expression at therapeutic plasma levels.

Memantine given for 5 days at a dose of 20 mg/kg (above the therapeutic range at time to maximum plasma concentration) significantly attenuated the decrease in GDNF induced by 3-nitropropionic acid (Ranju et al. 2015). It would be interesting to determine whether a lower dose treatment leading to constant body concentrations would produce the same effect. Due to a fast metabolism in mice, such an effect cannot be achieved by one- or two-times daily treatment, but requires constant infusion (e.g., using Alzet osmotic minipumps). In rats, memantine given repetitively at 20 mg/kg/day partially reversed the decrease in BDNF induced by valproate in a model of autism (Zohny et al. 2023). In summary, preclinical in vitro data suggest that memantine at therapeutic concentrations may increase GDNF levels; however, in vivo preclinical experiments to date have failed to confirm this at therapeutically relevant dose ranges. Similarly, the effect on BDNF is to be confirmed.

Sigma-1 receptors

Sigma-1 receptors are also located intracellularly, e.g., on membranes of the endoplasmic reticulum, and serve to control Ca²⁺ signaling (Monnet 2005). Activation of these receptors modulates NMDA-stimulated dopamine release (Gonzalezalvear and Werling 1995), increases dopamine levels in vivo in the striatum (Gudelsky 1995), and decreases dopamine uptake (Thompson et al. 2001). Sigma-1 receptors have also been implicated in neuroprotection (Francardo 2014) (Mori et al. 2012; Rousseaux and Greene 2015).

Kornhuber and colleagues were the first to show that memantine has relevant affinity at sigma-1 receptors (Kornhuber et al. 1993). In a binding study, they reported that memantine inhibited ³H-pentazocine binding in the human cortex with an inhibition constant of 19.98 µM. Considering memantine intracellular accumulation, such concentrations can be considered therapeutically relevant. In a different receptor binding study using [³H]-(+)SKF-10,047, Peeters and colleagues reported that memantine binds to the sigma-1 receptor in rat forebrain homogenates with an inhibition constant of 2.6µM (Peeters et al. 2004). In neuroblastoma NG108-15 cells, at 10 µM concentration, memantine potentiated the mobilization of intracellular Ca²⁺ produced by bradykinin, similar to the effect of the sigma-1 receptor agonist PRE-084 (Peeters et al. 2004). The authors therefore concluded that memantine behaves as a sigma-1 receptor agonist.

In an MTT assay (a colorimetric test used to determine cell cytotoxicity) in SH-SY5Y cells, Keshavarz et al. showed that memantine at a concentration of 2.5 µM decreased the neurotoxic effects of amyloid beta peptide (Aβ) (Keshavarz et al. 2020). The sigma-1 antagonist, BD1063, decreased the protective effect of memantine, indicating involvement of these receptors. A similar attanuation of memantine’s effects was observed after addition of chloroquine, indicating the role of lysosomal accumulation of memantine. The concentration used in these experiments (2.5 µM) is slightly above the maximal therapeutic range, hence it not known whether lower concentrations, such as 1 µM, would produce a similar effect.

DTG, a sigma-1/sigma-2 receptor agonist, enhanced the effect of memantine (2.5 mg/kg) in the forced swimming test in a rat model of depression; this effect was attenuated by a sigma-1 receptor antagonist. These data indicate that the sigma-1 receptor subtype may be involved in the behavioral response induced by memantine in Porsolt’s swim test in rats (Keshavarz et al. 2020; Maurice 2015; Skuza and Rogoz 2003).

Other targets

Memantine likely has anti-inflammatory actions that may also be important for its beneficial effects in AD. Patch-clamp recordings of inwardly rectifying K⁺ currents (IK(IR)) in RAW 264.7 macrophages and in BV2 microglial cells showed that memantine directly reduces their amplitude in a concentration-dependent manner (IC₅₀ 12 µM), by decreasing mean open time, increasing mean closed time, and slowing the inactivation time rate evoked with membrane hyperpolarization (Tsai et al. 2013). Additionally, memantine (1–30 µM) has been reported to suppress murine lymphocyte function through cross-inhibition of Kv1.3 and KCa3.1 voltage-gated potassium channels (Kahlfuss et al. 2014; Simma et al. 2014). Memantine also inhibits potassium Kv1.3 channel currents in human T cells under various conditions (stimulated, unstimulated) with IC₅₀ values between 14 and 40 µM (Lowinus et al. 2016). However, in rodent microglia cells in culture, memantine did not affect intracellular nitric oxide (NO) and Ca²⁺ mobilization or phagocytic activity (Murakawa-Hirachi et al. 2021).

In an animal AD model, APP23 mice, memantine increased hippocampal long-term potentiation (LTP) and CaMKII activity, which was antagonized by a specific adenosine triphosphate (ATP)-sensitive potassium channel opener (Moriguchi et al. 2018). Memantine also inhibited Kir6.1 and Kir6.2 ATP-sensitive potassium channels (at 10 nM concentration) and decreased intracellular Ca²⁺ in neuro2A cells overexpressing Kir6.1 or Kir6.2 channels. The authors suggested that the Kir6.2 channel is a novel potential target contributing to the efficacy of memantine in AD (Moriguchi et al. 2018).

Memantine was confirmed to block responses of α7, α4β2 and α3β2 receptors with IC₅₀ values of 6, 12 and 9 µM, respectively (Lee et al. 2012b). These affinities are clearly above the therapeutic range and can therefore be regarded as irrelevant. Such effects are also in clear contrast to evidence suggesting that the neuroprotective effect of memantine is partially dependent on enhancement of cholinergic transmission through nicotinic receptors (α9α10) (Ferrer-Acosta et al. 2022). Memantine also blocked receptors with an IC₅₀ of 1.2 µM and this effect was not mediated via interactions with the orthosteric agonist binding site (Plazas et al. 2007). However, this effect is probably only of potential relevance for diseases such as tinnitus due to limited localization of these channels in inner ear hair cells.

Using in vitro methods such as molecular docking, dynamics simulation and in vitro binding studies, interaction of memantine with human transferrin was suggested; however, the reported inhibition constant was 10 µM, rendering this target less therapeutically relevant (Shamsi et al. 2021). In APPswe/PS1 transgenic (APP/PS1) mice, memantine enhanced hippocampal neurogenesis and increased the number of radial glial-like cells (Sun et al. 2015). Additionally, memantine decreased amyloidogenesis and apoptosis in cultured radial glia cells induced by Aβ and inhibited activation of microglial cells in APP/PS1 mice. The authors suggested a role for reelin in this process (Sun et al. 2015).

In electrophysiological ionophoresis experiments in the rat hippocampus, Naggy et al. showed that low doses of memantine (0.1 mg/kg) increased spontaneous and acetylcholine-stimulated firing activity (Nagy et al. 2024). The combination of NMDA antagonism and enhancement of alpha7 (α7)-nicotinic acetylcholine receptor (nAChR) activity has been suggested to be beneficial in the treatment of AD (Nagy et al. 2024).

Functional effects

Amyloid beta peptide

Toxicity

The effects of memantine on Aβ toxicity have been reviewed previously (Danysz and Parsons 2012; Klein et al. 2007). Memantine (20 µM) reversed and reduced spontaneous activity induced by 24 h exposure to Aβ_1–42 (5µM), as recorded using microelectrode arrays from cultured murine neurons (Charkhkar et al. 2015). Memantine also reduced Aβ-induced astrocytic glutamate release, extrasynaptic NMDA receptor activation, and synaptic loss (Talantova et al. 2013). In human neuroblastoma SH-SY5Y cells, the toxic effects of Aβ fragment (25-35 at 50 µM) were only modestly attenuated by memantine (20 µM) alone, but an enhanced effect was reported in combination with folic acid (Chen et al. 2013). Additionally, a clear neuroprotective effect of memantine infusion (20 mg/kg/day) against Aβ toxicity was reported in septal neurons in vivo (Colom et al. 2013).

Synaptic plasticity

In ex vivo hippocampal slices from transgenic mice (CRND8 mice, Swedish-Indiana APP mutation) memantine (1 µM) selectively restored LTP in the dentate gyrus at 6 months, but not at 2 months (Tozzi et al. 2015). Moreover, memantine exerted anti-autophagic and anti-apoptotic functions, via mTOR-dependent and independent autophagic signaling pathways in SHSY5Y cells overexpressing the Swedish mutant of Aβ-precursor protein (APP695swe) (Song et al. 2015).

Aβ production

AD has some analogy to type 2 diabetes mellitus, e.g., insulin receptor signaling is associated with Aβ pathology (Sebastiao et al. 2014). In cultured astrocytes, memantine (1–10 µM) reversed streptozotocin (STZ)-induced loss of neurotrophic factors (BDNF, GDNF) and improved insulin receptor dysfunction (Rajasekar et al. 2016). Memantine also attenuated STZ-induced APP, β-secretase cleavage, and Aβ expression (Rajasekar et al. 2016). These changes were accompanied by attenuation of the effects of STZ such as astrocyte activation, and reduction of kappaB translocation, glial fibrillary acidic protein, COX-2, tumor necrosis factor-α (TNF-α) levels, and oxidative stress (Rajasekar et al. 2016).

Tau

Investigations of effects on tau showed that memantine (20 µM) prevented toxicity and tau hyper-phosphorylation in primary cortical neurons exposed to the mitochondrial toxin NaN₃ (0.1 mM) in combination with the oxidative stressor H₂O₂ (30 µM) (Selvatici et al. 2013). In vitro, memantine (100 nM) restored NR2B expression patterns and function produced by human AD-tau (Marshall et al. 2022). In primary cortical neurons, memantine (20 µM) attenuated the consequences of oxidative stress produced by H₂O₂ such as tau hyperphosphorylation and neurosecretory dysfunction (Selvatici et al. 2013).

Other effects in in vitro models relevant for AD

Memantine (10 µM) reversed nerve growth factor IB (Nur77) upregulation and sub-cellular translocation of Nur77/Cyt c/HSP60, and restored nuclear receptor related1 (Nurr1) down-regulation in 6-OHDA-lesioned PC12 cells (Wei et al. 2015). The authors concluded that memantine effectively suppresses Nur77-mediated neurodegeneration and promotes survival signaling through post-translational modification of Nurr1. In cultured astrocytes, memantine (1–10 μM) improved STZ-induced neurotrophic factor (BDNF, GDNF) decline. Moreover, memantine attenuated STZ-APP-induced β-site APP-cleaving enzyme-1 and Aβ_1–42 expression and restored insulin-degrading enzyme expression (Rajasekar et al. 2016).

Summary of the MoA

The NMDA receptor remains the major target of memantine; however, some supportive effects could result from action at other targets such as 5-HT3 receptors and possibly sigma-1 receptors.

New relevant publications supporting the use of memantine in AD/dementia

Many animal studies show clearly that therapeutically relevant doses of memantine can attenuate learning deficits produced by various insults. These finding are in agreement with clinical data on symptomatological efficacy; however, to date clinical support for neuroprotective effects of memantine has not been obtained (Parsons et al. 2013; Danysz and Parsons 2012; Rammes et al. 2008; Danysz and Parsons 2003; Parsons et al. 1999b; Ferris 2003; Grossberg et al. 2009; Bullock 2006; Cosman et al. 2007; Matsunaga et al. 2015; McKeage 2009; Robinson and Keating 2006; Wilkinson 2011).

Preclinical

In vitro

The neuroprotective properties of memantine (10 µM) were investigated in Aβ-induced neurotoxicity in hippocampal neuron cultures (Ovey and Naziroglu 2021). Ca²⁺ concentration, apoptosis, caspase 3 activation, caspase 9 activation, mitochondrial membrane depolarization and intracellular reactive oxygen species production were all decreased by memantine, while the Aβ-induced decrease in cell viability level was restored by memantine (Ovey and Naziroglu 2021). In human endothelial cell cultures, memantine (10, 20 µM) improved microvascular dysfunction induced by the pro-inflammatory cytokine, TNF-α (Wang et al. 2017a). Specifically, memantine decreased the attachment of monocyte THP-1 cells to human brain microvascular endothelial cells and inhibited TNF-α-induced disruption in a blood-brain barrier (BBB) model. Additionally, memantine interfered with monocyte transmigration across the BBB model. TNF-α induced expression of cell adhesion molecules (ICAM-1, VCAM-1, E-selectin) was also prevented by pre-treatment with memantine. These data indicate anti-inflammatory effects of memantine at the level of endothelial brain vasculature which may contribute to its neuroprotective activity (Wang et al. 2017a). Unfortunately, the very high concentrations used do not allow direct extrapolation to therapeutic use. In ex vivo cortical slices, the effects of memantine on excitatory/inhibitory balance in the prefrontal cortex (PFC) were investigated (Povysheva and Johnson 2016). It was reported that 10 µM memantine shifted this balance away from inhibition to excitation, an effect which appeared to be NMDA-receptor mediated (Povysheva and Johnson 2016).

Memantine (2.5 µM) attenuated the neurotoxic effects of Aβ in SH-SY5Y cells (Keshavarz et al. 2020). In rat astrocytoma cells (C6) STZ was used to induce astroglial inflammation. Memantine (5 µM) and ibuprofen (200 μM) prevented STZ-induced change in expression of CaMKIIα, cyclic-AMP response element-binding protein (CREB), calpain, and caspase 3 in C6 astrocytoma cells (Mishra et al. 2021a). Memantine also prevented changes in NMDA receptor subunit expression. Both memantine and ibuprofen attenuated increases in glial fibrillary acidic protein, TNF-α, and iNOS, and produced a decrease in IL-10 (Mishra et al. 2021a).

In HeLa cells stably expressing green fluorescent protein-microtubule-associated protein light chain 3, memantine increased the levels of LC3-II and upregulated autophagic flux (Hirano et al. 2019). The pharmacological effects of memantine on autophagy were independent of mTORC1 activity and NMDA receptor activation. A VPS34 inhibitor suppressed the effect of memantine, indicating that memantine may act through VPS34 complex activity. Specifically, memantine decreased aggregates of elongated huntingtin (Hirano et al. 2019). Memantine also enhanced elimination of degraded mitochondrial in neurons derived from stem cells of PARK2 or PARK6 patients, who exhibited defective PINK1/parkin-mediated mitophagy (Hirano et al. 2019). Unfortunately, in this study, memantine was used at a concentration of 10 µM or more.

All data presented in this section support the use of memantine in AD due to beneficial effects on Aβ action; however, the high concentrations used are a major obstacle in permitting direct therapeutic implications. In one study which used a concentration relevant for therapeutic situations, memantine (1 µM) rescued the deficit produced by acute exposure to soluble Aβ42 aggregates in dentate gyrus LTP in rat hippocampal slices (Bellingacci et al. 2024).

In vivo

Cognition, neurogenesis

Memantine treatment (20 mg/kg per day IP) in mice increased the density of newborn granule cells and the content of chondroitin sulfate proteoglycan in the hippocampus, accompanied by increased expression of the enzyme responsible for chondroitin sulfate proteoglycan synthesis (Maeda et al. 2022). Memantine also increased neurotrophin-related molecules and improved learning performance (Maeda et al. 2022). Similarly, single treatment with memantine (25 mg/kg) enhanced social recognition memory and increased neurogenesis in the hippocampus 7 days after treatment (Jaimes et al. 2020)

Cahill and colleagues tested memantine on neurogenesis in male and female rats (Cahill et al. 2018). Memantine transiently increased neurogenesis for 1 month, while 2 months of running followed by memantine increased neurogenesis in males. In another study from this group, also in rats, running + memantine (35 mg/kg, four injections once a week) increased neurogenesis by 17%, and blocked experience-dependent c-Fos expression (Cahill et al. 2019).

Learning in animals without cognitive deficits

Performance in a taste-avoidance task in chicks (pecking a bead coated with an aversive substance) was improved by memantine with a bell shaped curve i.e., a significant positive effect was observed at 1 mmol/L while lower and higher doses were ineffective (Barber and Kimbrough 2015). In the same task, memantine at a dose of 0.1 mM injected shortly after weekly aversive training enhanced retention for up to 24 h; it was also effective if given before the reminder trial (Samartgis et al. 2012). The study authors commented that the results suggest that memantine may facilitate memory consolidation and reconsolidation.

In another study, in vivo hippocampal LTP was enhanced by memantine at 5 and 10 mg/kg doses and this effect was not blocked by scopolamine (5 mg/kg) (Ma et al. 2015). Memantine at the same doses also enhanced gamma oscillations and acoustic startle and, at a higher dose (10 mg/kg), increased pre-pulse inhibition (Ma et al. 2015). Co-administration of inactive doses of memantine with galantamine (1 mg/kg each) facilitated rat’s set-shifting performance and reversed delay-induced deficits in object recognition (Nikiforuk et al. 2016). Methyllcaconitine antagonized this effect, suggesting that the cognitive enhancement is a7-nAChR dependent. Similar enhancement was seen in an object recognition test after 1 mg/kg of memantine in combination with galantamine 0.3 mg/kg (Nikiforuk et al. 2016).

Memantine at 5 mg/kg (but not 1 mg/kg) enhanced fear-conditioned context learning and step-down passive avoidance acquisition, as measured by retention in mice (Vignisse et al. 2014). Single administration of memantine at 50 mg/kg enhanced hippocampus-dependent spatial water maze learning and social recognition 3–6 weeks later, but not at 3 days or 4 months later (Ishikawa et al. 2014). It should be noted, however, that this is a very high dose not relevant for clinical use. In an allothetic place avoidance alternation task, in which rats learned to avoid entering a place where shocks were presented on a rotating arena, memantine at 5 mg/kg enhanced working memory; however, at a higher dose (20 mg/kg) an impairment was observed (Wesierska et al. 2013).

Several studies reported no effect of memantine in animals without deficits and a few using high doses demonstrated impairment; however, these mostly occurred at the doses producing unspecific effects on task performance. No specific effect on 5-choice serial reaction time accuracy was observed after memantine doses up to 3 mg/kg, i.e., the dose which started to produce response omissions and premature response, indicating an unspecific effect on performance and increase in impulsivity (Benn and Robinson 2014). In the Morris water maze task, memantine at a high acute dose (20 mg/kg), but not at a low dose (5 mg/kg), impaired working memory in an interval-independent manner in rats (Duda et al. 2016). In rats, memantine 5 mg/kg 30 min before a training session in two versions of a place avoidance task improved avoidance in the room+arena task but not in the arena+ task (Wesierska et al. 2019). Hence, it seems that memantine improved performance in a task requiring the segregation of spatial stimuli into coherent subsets.

Deficit reversal

The majority of the studies described below showed positive effects of memantine in various models in aged animals, expressed as reversal of functional deficits measured by electrophysiological or behavioral methods. First, deficits produced by pharmacological treatments are discussed, followed by lesion-induced deficits, age, and finally transgenic animals.

Intracellular single unit recordings were performed in rats after nucleus basalis magnocellularis lesion (Zamani et al. 2022). The mean frequency of CA1 pyramidal neuron firing was decreased, but this effect was attenuated by memantine 10 and 20 mg/kg (Zamani et al. 2022). Memantine (0.1 mg/kg) and PHA-543613 (0.3 mg/kg) reversed scopolamine-induced short-term memory deficits in rats (Bruszt et al. 2021). Co-administration of subeffective doses of both agents produced stronger effects than either treatment alone (Bruszt et al. 2021). In the spatial working memory spontaneous alternation paradigm in T-maze in rats, scopolamine-induced amnesia was attenuated by administration of a very low dose of memantine (0.03 and 0.1 mg/kg) and also at 0.003 mg/kg in combination with a subeffective dose of the α7-nAChR agonist, PHA-543613 (Bali et al. 2019). A higher dose of memantine (0.3 mg/kg) failed to show efficacy alone but increased the action of PHA-543613. The authors suggested that memantine and PHA-543613 may exert their cognitive effects through α7-nAChRs, which supports the benefits of combination therapy (Bali et al. 2019). Interpretation of these data is difficult due to the use of such low doses of memantine which do not assure action at any known target according to current knowledge.

In rats, intracerebroventricular (ICV) injection of okadaic acid impaired memory function in dual-solution plus-maze tasks, and decreased expression of the α7 subunit of nAChR and expression of the NR2B subunit of NMDA glutamate receptors in the hippocampus (Dashniani et al. 2020). Administration of memantine (5 mg/kg/d for 13 days) prevented this hippocampal-dependent spatial memory impairment and improved expression of the a7 and NR2B subunits in the hippocampus (Dashniani et al. 2020). A learning deficit in autoshaping associative learning task produced by intraventricular STZ was also attenuated by memantine at 5 mg/kg (Rai et al. 2013b, a). In the Morris water maze task, memantine 20 mg/kg for 2 weeks in rats, attenuated the deficit produced by scopolamine and attenuated the increase in acetylcholinesterase (AChE) activity, glutamate levels, and change in other transmitters produced by scopolamine in the hippocampus, cortex and striatum (Al-Hazmi et al. 2015). Spontaneous alteration deficit in T-maze in mice produced by methylcaconitine was reversed by memantine (1 or 5 mg/kg) treatment (Andriambeloson et al. 2014). Memantine was much less effective against deficits produced by scopolamine in the same task, producing maximum 20% effect. A combination of sub-active doses of memantine (0.56 mg/kg) and galantamine (0.1 mg/kg) in mice attenuated deficits produced by 1 mg/kg of scopolamine in spontaneous alteration and the object recognition tasks (Busquet et al. 2012). Memantine improved spatial working memory deficits in the Y-maze in rats produced by repeated immobilization stress, decreased serum level of stress markers, and modified hippocampal synaptic plasticity markers (Amin et al. 2015). In the hippocampus of rats treated with doxorubicin (to produce learning impairment) memantine 2.5 mg/kg for 7 days improved discrimination index (DI) in novel location recognition task compared with doxorubicin alone (Elbeltagy et al. 2021). Moreover, memantine increased the number of Ki67-positive cells, indicating enhanced cell proliferation (Elbeltagy et al. 2021). Following amphetamine withdrawal, treatment with memantine (5 mg/kg) partially attenuated a learning deficit in novel object recognition task (Marszalek-Grabska et al. 2016).

In mice after olfactory bulbectomy (OBX), which is a model of early AD, memantine 1–3 mg/ kg by mouth for 14 days starting at 10 days after surgery rescued cognitive deficits, possibly via ATP-sensitive potassium channel inhibition, as assessed in Y-maze, novel object recognition, and passive avoidance tasks (Moriguchi et al. 2021). LTP in the hippocampal CA1 region was also improved, and immunoblotting assay showed improvement in CaMKII and CaMKIV (Moriguchi et al. 2021). In the same model, combined administration of memantine and donepezil improved depressive-like behaviors, social interaction and cognitive deficits more potently than either treatment alone (Yabuki et al. 2017). Using brain microdialysis, the authors observed reduced nicotine-induced serotonin (5-HT) in the hippocampus which was also restored by combined treatment (Yabuki et al. 2017). Moreover, decreased autophosphorylation of CaMKII was restored by this treatment (Yabuki et al. 2017).

In senescence-accelerated prone-8 mice, memantine given orally at 20 mg/kg daily for 8 weeks improved learning in a Morris water maze task (Dong et al. 2012). It also reduced neurofibrillary tangles and APP in the hippocampus. A similar effect was seen after environmental enrichment and both treatments combined produced even stronger effects. Unfortunately, lower doses have not been tested. Memantine moderately improved cognition in a delayed matching to sample task in aged rhesus macaques with longer delays, and produced a benefit in combination with galantamine in paired associated learning task (Schneider et al. 2013).

Injection of different forms of Aβ has also been used to investigate the cognitive effects of memantine in animals. Memantine attenuated contextual fear conditioning deficit after ICV injection of Aβ in mice (Ceccom et al. 2012). In another study, memantine administered for 7 days (p.o. 10 mg/kg) rescued oxidative stress and memory impairment in a novel object recognition task produced by high molecular weight Aβ oligomers (injected ICV), but not that produced by low molecular weight oligomers (Figueiredo et al. 2013). In a novel object recognition task in rats, the deficit induced by soluble Aβ was attenuated by memantine (5 mg/kg IP) given shortly after the first object presentation, but not when administered before (Tucci et al. 2014). Memantine treatment (2 mg/kg) for 7 days prevented passive avoidance learning impairment produced by ICV administration of Aβ in rats (Hemmati et al. 2013). The effect of memantine (5 mg/kg, 17 days) on spatial memory and neuroinflammation was studied in rats after an intra-hippocampal injection of Aβ_1-42 oligomers (Budni et al. 2017). In the radial arm-maze task, performance was improved and reductions in the levels of interleukin (IL) in the frontal cortex were reversed similarly to TNF-α in the hippocampus (Budni et al. 2017).

Most relevant for the primary AD indication of memantine are studies in different types of transgenic animals modelling this disease. In PS2Tg2576 transgenic mice (5–6 months old), memantine (30 mg/kg/day in drinking water) administered for 3–4 weeks improved performance in spatial test (Morris water maze) and the water-finding test for latent inhibition (Matsumura et al. 2021). APP/PS1 transgenic mice treated with memantine (20 mg/kg/day) showed improvement in acquisition in the Morris water maze, enhanced entorhinal cortex to hippocampal CA1 synaptic transmission, and enhanced dendritic spine regeneration of this projection (Li et al. 2021). In the double transgenic APP/PS1 (5xFAD) mouse model, memantine administered for 32 days in drinking water with melatonin decreased Aβ aggregates and reactive microgliosis, and improved episodic memory in the object recognition test more potently than either drug alone (Jurgenson et al. 2019). Working memory deficit in diacylglycerol kinase β knockout (involved in spine formation) mice in Y-maze was attenuated by memantine at 2.5 mg/kg (Kakefuda et al. 2016). In APP/PS1 transgenic mice, memantine (5 mg/kg for 4 weeks p.o.) improved water Morris maze learning and decreased APP and Aβ expression in the brain, increased NGF/TrkA and inhibited p75 signaling (Liu et al. 2014). It also improved working memory deficits (2.5 and/or 5 mg/kg) in 4-month-old mice, while donepezil was ineffective; however, both compounds were effective in 6-month-old animals in the Y-maze and Morris water maze (Nagakura et al. 2013). In APP23 mice in a complex dry-land maze, both memantine (10 mg/kg) and donepezil (1 mg/kg) improved learning following combined treatment (Neumeister and Riepe 2012).In tgCRND8 mice developing amyloid plaques, memantine 5 mg/kg attenuated false object recognition and normalized enhanced NMDA receptor dependent LTD (Romberg et al. 2012).

Neuroprotection

Most of the studies addressing the neuroprotective effects of memantine have been performed in animals subjected to injection of various toxins or different forms of Aβ, or were conducted in transgenic animals.

Stress Memantine (20 mg/kg followed by 2 mg/kg/day for 14 days) prevented postoperative cognitive impairments of spatial learning in the radial arm maze in rats (Chi et al. 2013). In Octodon degus, treatment with memantine (10 mg/kg IP) before sleep deprivation prevented deficits in the Barnes maze, radial arm maze learning, and novel object recognition tasks (Tarragon et al. 2014). Memantine (5, 10 mg/kg) in mice subjected to chronic restraint stress for 16 or 28 consecutive days (1 h daily) decreased the elevation of phosphorylated tau, attenuated upregulation of GSK3β and CDK5 expression, and reversed downregulation of protein phosphatase 2A (PP2A) activity (Liu et al. 2019). Memantine also upregulated the NMDA receptor subunits GluN2A and GluN2B in the frontal cortex. In contrast, memantine enhanced tau phosphorylation in the cortical regions following stress. At the behavioral level, memantine produced anxiolytic effects in the elevated plus maze test (Liu et al. 2019).

Age A model of progressive dementia was induced in Wistar 15-month-old rats which were hippocampally loaded with pore former polymeric 1,3-alkylpyridinium salts (poly-APS) and recombinant human tau under pharmacological inhibition of phosphatase activity by okadaic acid (Mietelska-Porowska et al. 2019). This treatment resulted in an increase of tau and its hyperphosphorylation, changes in tau compartmentalization, breakdown of the cytoskeleton, and cognitive impairment. In this rat model, memantine (20 mg/kg for 33 days) reversed deficits in spatial learning and produced normalization of tau alterations (Mietelska-Porowska et al. 2019). In senescence-accelerated mice, memantine (10 mg/kg) reduced anxiety and improved active avoidance performance (Zeng et al. 2020). Memantine at the same dose, also attenuated age-related reduction in CD4+CD28+ cells and reduced pro-inflammatory factors in plasma, namely TNF-a, IL-1 beta (IL-1β), and G-CSF (Zeng et al. 2020). Yamada and colleagues studied the role of the perineuronal net, an aggregate of the extracellular matrix, in neuroprotection against oxidative stress and in turn considered contributors to decline in brain function with age (Yamada et al. 2017). The results suggest involvement of the perineuronal net in brain aging and indicate that memantine (20 mg/kg for 21 days) counteracts the age-related increase in expression of Cat-315 (an epitope which recognizes human natural killer-1) through modification of subcellular localization (Yamada et al. 2017).

In a rat model of late onset Alzheimer's disease memantine (20 mg daily p.o.) was administered from 4 to 10 months (Souchet et al. 2022). Memantine promoted a non-amyloidogenic cleavage of APP and decreased soluble Abeta42 and also improved LTP and learning. There was no effect on tau (Souchet et al. 2022).

Toxin injections Memantine (2.5 mg/kg IP) injected for 3 weeks at the same time as intrathecal methotrexate (given for the first 2 weeks) prevented development of long-lasting spatial memory deficits, as seen in an object recognition task in rats (Cole et al. 2013). Prevention of motor performance decline produced by 3-nitropropionic acid was also shown with memantine given for 5 days at the dose of 20 mg/kg (above the therapeutic range) to mice (Ranju et al. 2015). Associated effects included decreased striatal glutamate content, improved SDH activity in the striatum and cortex, and decreased apoptosis (Ranju et al. 2015). Mice that received an ICV injection of ouabain, an inhibitor of Na⁺/K⁺-ATPase, showed hyperactivity via overactivation of glutamatergic neurons, which was decreased by chronic memantine treatment (Kurauchi et al. 2019). Additionally, the decrease in Na⁺/K⁺-ATPase activity was also improved in the hippocampus and accompanied by neuroprotection (Kurauchi et al. 2019). In mice treated with a mitochondrial toxin, 3-nitropropionic acid, memantine (20 mg/kg) administered for 4-5 days prior, restored motor functions, increased brain succinate dehydrogenase activity, and decreased glutamate content (Ranju et al. 2015). Moreover, striatal neurodegeneration and neuronal apoptosis were also decreased.

When Wistar rats were treated with memantine for 3 days prior to intrahippocampal infusion of okadaic acid (a PP2A and PP1 inhibitor used for inducing tau hyperphosphorylation and cognitive decline), memantine’s neuroprotective effect was associated with astrocytic activation, decreased glutamate uptake in the hippocampus, and increased release of S100B protein in the CSF (Torrez et al. 2019). The authors highlighted astrocytes as possible targets for memantine (Torrez et al. 2019). Additionally, while ICV injection of the PP2A inhibitor, okadaic acid leads to increased phosphorylation of tau, memantine 10 mg/kg/day for 13 days starting one day after okadaic acid injection has been shown to attenuate changes in tau, tau-C terminal, tau-N terminal, CaMKII, calpain, PP2A, GSK3β, TNF-α, IL-β, total nitrite, iNOS, nNOS I, malondialdehyde, reduced glutathione, reactive oxygen species, mitochondrial calcium levels, apoptosis, neuronal death (histology), and learning impairments in the water Morris maze (Kamat et al. 2013; Kamat et al. 2012a; Kamat et al. 2012b; Kamat et al. 2010; Kamat et al. 2011).

Diabetes Memantine (10 mg/kg) administered p.o. in rats for 13 days starting shortly before STZ treatment prevented spatial memory deficits in the Morris water maze task and normalized TNF-α, ILβl, and reactive oxygen species levels, glial call activation, apoptotic cell death, and neurotoxicity (Rai et al. 2013b, a). Obesity associated with type 2 diabetes and aging have emerged as risk factors for the development of AD and similarities exist in the pathophysiology of AD and type 2 diabetes which are Aβ related (Sebastiao et al. 2014). In a rat model of type 2 diabetes and AD, expression of APP and phosphorylated tau protein was decreased by memantine (10 mg/kg) alone as well as in combination with vildagliptin (Khalaf et al. 2019). In an ICV STZ-induced AD rat model, after memantine 30 mg/kg/day (for 28 days) Bace1-as gene expression was increased in brain tissue and blood, suggesting utility as a biomarker during progression of AD, while the level of the Bace1 protein was significantly increased in the brain tissue and was low in plasma, suggesting utility as a biomarker for prognosis (Azadfar et al. 2020).

In rats following STZ memantine (5, 10 mg/kg) reversed memory deficit in Morris water maze from day 2. Insulin increased the potency of memantine in preventing memory deficit. (Bahramian et al. 2016).

Özsoy and co-workers showed in rats after STZ memantine at 5 mg/kg improved passive avoidance learning and hippocampal morphology (neuronal density) (Ozsoy et al. 2023).

Aβ injection in Wistar rat AD model based on ICV injection of Aβ_1-42, memantine (20 mg/kg) administered for 21 days in combination with baicalein (a natural product) reduced escape latency in the Morris water maze (also transfer latency) and passive avoidance deficit, and decreased neurodegeneration in the hippocampus and cortex (Jadhav and Kulkarni 2023a). The combination treatment decreased oxidative stress and Aβ plaque formation, and increased the expression of BDNF in the brain. Combined with baicalein, memantine additionally normalized malondialdehyde, superoxide dismutase, catalase, and reduced glutathione (Jadhav and Kulkarni 2023a). Similar effects have been reported by the same group using a rat model based on aluminum chloride-induced neurotoxicity (Jadhav and Kulkarni 2023b). Memantine (bolus 20 mg/kg followed by 20 mg/kg/day infusion for 8 days) blocked damage to the septal cholinergic neurons produced by a direct injection of Aβ_1–40 in rats 12 h after initiation of memantine treatment (Colom et al. 2013). Additionally, memantine infused continuously for 2 weeks (9.4–14.4 mg/kg/day) prevented neurodegeneration produced by ICV injection of Aβ_1-40 in rats as evidenced by histology, decrease of DNA fragmentation, and decrease in Bcl-2 staining and caspase 8 immunostaining (Miguel-Hidalgo et al. 2012).

TG mice In transgenic APPswe/PS1dE9 mice, memantine given in drinking water at 30 mg/kg restored brain insulin signaling and protein kinase B and CREB expression, and reduced amyloidogenesis (Ettcheto et al. 2018). Ito and colleagues showed that in 9-month-old Tg2576 mice, memantine 10 or 20 mg/kg/day in drinking water for 1 month significantly decreased CHAPS-soluble and CHAPS-insoluble Aβ in the brain (Ito et al. 2017). A similar effect was seen with memantine 10 mg/kg/day in aged F344 rats. In rat primary cortical cultures, memantine also reduced Aβ and sAPPβ levels in conditioned media without affecting the enzymatic activities of secretases (Ito et al. 2017). The authors concluded that memantine reduces Aβ production and plaque deposition through the regulation of intracellular trafficking of APP (Ito et al. 2017).

Quantitative proteomic analysis was used to investigate protein profiles in the hippocampus and the cerebral cortex of a transgenic mouse model (3×Tg-AD) (Zhou et al. 2019). Memantine (5 mg/kg, twice daily [BID]) for 4 months improved learning and memory retention. Memantine altered expression of 233 proteins out of over 3000 that were changed in transgenic mice. Specifically, memantine seems to modulate pathways associated with the cytoskeleton and ErbB signaling in the hippocampus, and those associated with axon guidance, ribosome, cytoskeleton, calcium and MAPK signaling in the cerebral cortex (Zhou et al. 2019).

The effect of 12 weeks of memantine (1 mg/mL, gavage assumed 50 mg/kg) was enhanced by the probiotic Lactobacillus plantarum in APP/PS1 mice, decreasing the concentration of Αβ_1-42 and Αβ_1-40, protecting hippocampal neurons and plasticity, inhibiting trimethylamine-N-oxide synthesis, and alleviating neuroinflammation (Wang et al. 2020). In the Tg4-42 mouse model of AD, memantine (20 mg/kg per os for 4 months) decreased hippocampal CA1 neuron loss, restored hippocampal neurogenesis and improved learning in the Morris water maze test and a novel object recognition task (Stazi and Wirths 2021). Aβ deposition in the hippocampus, hippocampal CA1 and dentate gyrus histology, Morris water maze test was also improved by memantine (5 mg/kg for 4 weeks) in APP/PS1 mice (Qiao et al. 2021).

Memantine treatment (10 mg/kg) for 30 days to younger 5XFAD mice reversed memory impairments in contextual fear conditioning and spontaneous alternation Y-maze (Devi and Ohno 2016). Soluble Abeta oligomer and total Abeta42 were not affected. In older (12-15 month) 5XFAD mice with well-established deficits memantine had no effect.

Section conclusions

Most of the studies discussed above show that memantine, at therapeutically relevant doses, can attenuate learning deficits produced by various insults and decrease neurodegeneration. Several studies support NMDA receptors as the primary target for these effects. In general, the preclinical studies provide support for the use of memantine in AD and its suggested neuroprotective potential.

Other effects

In an electroencephalogram study, a low dose of memantine (2.5 mg/kg) enhanced pedonculopontine-elicited theta oscillations in the hippocampus in mice, while a higher dose (10 mg/kg) decreased it (Guadagna et al. 2012). This may be related to cognition-enhancing activity and corresponds to doses shown in rodents to enhance and disrupt learning respectively. Interestingly, the high-affinity NMDA antagonist (+)MK-801 only demonstrated disruptive effects.

Clinical

Effects on cognition and functional changes

In an open-label, single-arm, multicenter study in 451 patients with early, middle, or late-stage AD, 12-weeks’ treatment with memantine (up-titrated from 5 to 20 mg/day) increased the Relevant Outcome Scale for AD scores across all severity stages, with the exception of early stage (Holthoff et al. 2012). Although changes from baseline on the Clinical Global Impressions of Change (CGI-C) demonstrated minimal improvement or effect for most of the patients, memantine was well tolerated (Holthoff et al. 2012).

In post-hoc analyses of nine randomized, double-blind, placebo-controlled trials in a total of 2506 patients with moderate-to-severe AD, estimated odds ratio and corresponding confidence intervals were based upon a random-effect model for three individual domains: cognition, activities of daily living [ADL], and global impression of change (Hellweg et al. 2012). Compared with placebo, memantine treatment resulted in a significant delay in clinical worsening in these domains for which fewer patients faced clinical worsening (p < 0.05). The study duration was 16–28 weeks (depending on the center). It could not be ascertained whether the observed effect was due to a delay in disease progression since necessary additional measures were missing, e.g., analysis after wash out or cross over.

Memantine treatment (10 mg BID) was assessed for effects on functional communication in patients with moderate AD in a 12-week, randomized, double-blind, placebo-controlled trial using the Functional Linguistic Communication Inventory as the primary measure (Saxton et al. 2012). Additionally, combined subscales (Social Communication and Communication of Basic Needs) from the American Speech-Language-Hearing Association Functional Assessment of Communication Skills for Adults (a secondary measure) were administered to caregivers. Patients treated with memantine (n = 133) demonstrated a trend for improvement on the Functional Linguistic Communication Inventory and a significant improvement on the secondary measure.

Memantine was studied for its effects on agitation in 149 patients with moderate-to-severe AD and significant agitation in a randomized, double-blind, placebo-controlled trial (Fox et al. 2012). The primary outcome was mixed model autoregressive analysis of the Cohen-Mansfield Agitation Inventory at 6 weeks and secondary outcomes were: effects at 12 weeks on the Cohen-Mansfield Agitation Inventory; 6- and 12-week effects on the instruments: Neuropsychiatric Inventory (NPI); CGI-C; Mini-Mental State Examination (MMSE), and Severe Impairment Battery (SIB). Relative to placebo, the authors found no significant differences in the primary outcome (agitation) at 6 or 12 weeks. However, in the NPI measure of neuropsychiatric symptoms, a benefit with memantine was observed. Additionally, memantine showed a positive effect on cognition (Fox et al. 2012).

A 16-week study in 146 probable AD patients evaluated AD Assessment Scale-Cognitive Subscale (ADAS-cog) scores relative to baseline and tested whether the butyrylcholinesterase genotype affected rivastigmine (transdermal patch) alone or with memantine (20 mg/day) (Han et al. 2015). Interestingly, BCHE-K carriers showed a lower responder rate on the ADAS-cog than non-carriers and this was evident in patients with apolipoprotein E ε 4. APOE genotypes were determined in a subset of 415 subjects who participated in a randomized controlled trial of vitamin E and memantine in 613 veterans with mild-to-moderate AD during c.a. 3 years’ study (Belitskaya-Levy et al. 2018). There was no difference in the rate of functional decline between APOE ε4 allele carriers and non-carriers, although a significant interaction was observed between treatment and the APOE genotype on disease progression: ε4 carriers declined faster than non-carriers in the vitamin E plus memantine treatment group.

A study in 22 AD patients treated with memantine (20 mg/day) evaluated CSF drug concentrations (Valis et al. 2019). The IC₅₀ value for NMDA antagonism was achieved in only three cases, while 25% of the IC₅₀ value was reached in five cases; hence, NMDA antagonism as the MoA of memantine was questioned by the authors. However, it cannot be excluded that the therapeutic effect requires less than 50% of the inhibition and/or additional, above discussed mechanisms could play a role. This is supported by the fact that memantine positively affected the levels of some oxidative stress parameters, especially non-protein thiols and 3-nitrotyrosine, in this study.

Wilkinson and colleagues conducted a randomized, double-blind, placebo-controlled, fixed-dose study in patients with probable AD to evaluate brain atrophy with magnetic resonance imaging (MRI) during 52 weeks of treatment with memantine (20 mg) or placebo (Wilkinson et al. 2012). Secondary assessments included cognitive and behavioral scales. Of 518 patients screened, 278 were randomized and 217 completed the study. The primary efficacy analysis, total brain atrophy, showed no differences between treatment with memantine and placebo. However, there was a statistically significant correlation between change in total brain atrophy and in cognitive and behavioral scale scores. Patients not treated with AChEIs had a significantly lower total brain atrophy rate than patients treated with AChEIs (Wilkinson et al. 2012).

In a 24-week, double-blind, placebo-controlled, randomized trial (11 patients with moderate to severe probable AD per arm), the potential disease-modifying effects of memantine 10 mg BID were investigated using the clinical ratings, fluorodeoxyglucose positron emission tomography measurements and CSF assays (Wang et al. 2013). Compared with placebo, memantine was associated with less cognitive decline on the SIB and significantly less declines in the cerebral metabolic rate for glucose in regions preferentially affected in AD; there were no significant effects on CSF Aβ₁₋₄₂, CSF Aβ₁₋₄₀, and total or phosphorylated tau levels or ratios. The investigators suggested an association between the clinical benefits of memantine and its effects on FDG-PET measurements in AD-affected brain regions.

In an open-label exploratory study, a correlation between proton magnetic resonance spectroscopy and cognition was investigated in 11 patients with mild-to-moderate AD for the first 24 weeks after AChEI treatment and for the next 24 weeks after adding memantine (Gordon et al. 2012). Following the addition of memantine, there was an increase in myo-inositol and a decrease in N-acetylaspartate/myo-inositol ratio, but there were no significant changes in other metabolites or ratios, or neurocognitive measures.

Combination with AChEIs or AChE enhancers

The response to rivastigmine transdermal patch alone or in combination with memantine was studied in relation to apolipoprotein E genotype in 206 patients with AD (Han et al. 2012). The study showed a better response to memantine in patients with the APOE ε4 allele.

A multicenter study of donepezil (10 or 23 mg) with memantine or donepezil alone assessed the benefit of the combination therapy (Doody et al. 2012). For the SIB, there was a statistically significant treatment difference that favored donepezil 23 mg/day over donepezil 10 mg/day, regardless of concomitant memantine use. In another study in 295 community-dwelling moderate or severe AD patients treated with donepezil for at least 3 months, the following treatment options were offered: continue donepezil, discontinue donepezil, discontinue donepezil and start memantine, or continue donepezil and start memantine for 52 weeks (Howard et al. 2012). The coprimary outcomes were scores on the standardized MMSE and on the Bristol Activities of Daily Living Scale. Results revealed that the efficacy of donepezil and memantine did not differ significantly in the presence or absence of the other. There were no significant benefits of the combination of donepezil and memantine over donepezil alone (Howard et al. 2012).

Memantine (20 mg) was given to 201 outpatients with moderate-to-severe AD with some patients (n = 93) additionally treated with AChEIs. Analysis (MMSE, ADL, Instrumental ADL, NPI) was performed at 6 months and 1, 2, and 3 years (Sinforiani et al. 2012). In each group, ca. 20% of patients showed no deterioration at 6 months and 1 year (slightly less at 2 years) and this result was similar in the combination and memantine monotherapy groups. Grossberg and colleagues studied the safety, tolerability, and efficacy of an ER formulation of memantine (28 mg/day) in a randomized, double-blind, placebo-controlled study in 677 patients with moderate-to-severe AD taking AChEIs for up to 24 weeks (Grossberg et al. 2013). Memantine treatment was superior on the SIB, NPI, and verbal fluency test (p = 0.004) but failed to reach significance on the Alzheimer’s Disease Cooperative Study-Activities of Daily Living Inventory (ADCS–ADL₁₉; p = 0.177). Authors concluded that this higher dose of ER memantine was efficacious and well tolerated (Grossberg et al. 2013). In a retrospective study, the combination of memantine and AChEIs for up to 6 months was assessed in 240 AD patients (Gareri et al. 2014). The results revealed a significant increase in the MMSE total score after treatment with the combination of donepezil plus memantine compared with rivastigmine plus memantine, with no effect on ADL or the Cumulative Illness Rating Scale (which is used to evaluate comorbidities and their impact). However, there were no groups treated with any treatment alone, which precludes any elaboration of the implications of these study findings (Gareri et al. 2014).

In a 52-week, multicenter, randomized, double-blind, placebo-controlled trial in moderate-to-severe AD patients receiving donepezil, clinical decline after donepezil cessation was compared to continuation of donepezil or an addition or substitution of memantine (Tariot 2013). The study was terminated and in turn only 295 patients were analyzed out of a planned 800. Patients after donepezil withdrawal showed marked decline in both cognition and daily function in comparison with the other three treatment groups. Memantine treatment was associated with fewer behavioral symptoms (Tariot 2013). The effects of donepezil alone (n = 19) and in combination with memantine 5->20 mg (n = 18) were assessed in patients with moderate-to-severe AD at weeks 0, 4, 12, and 24 using change from baseline on the CGI-Improvement (CGI-I) scale, MMSE, Clock Drawing Test, NPI, Japanese version of the Zarit Burden Interview (J-ZBI), and near-infrared spectroscopy (Araki et al. 2014). A significant benefit of the combination therapy was found in the score variation of the CGI-I scale, MMSE, Clock Drawing Test, NPI, and J-ZBI.

Memantine (up to 20 mg) was given to 110 AD patients in addition to placebo, donepezil, rivastigmine, galantamine, or huperzine A for 24 weeks (Shao 2015). At 24 weeks, patients treated with memantine plus huperzine A showed better MMSE and ADL scores than those treated with memantine plus placebo. In a prospective study, the efficacy of memantine was analyzed in galantamine-treated AD patients (n = 2045) using the MMSE, Disability Assessment for Dementia scores and mortality endpoints (Hager et al. 2016). At 24 months, the decline from baseline in MMSE and Disability Assessment for Dementia scores were lower in galantamine patients as compared to placebo recipients.

A multicenter, 24-week, prospective, randomized, open-label study in 147 patients with AD assessed the effect of memantine as add on to rivastigmine patch therapy using MMSE scores (Yoon et al. 2016). Agitation symptoms were analyzed using the Korean Version of the Cohen Mansfield Agitation Inventory. The rivastigmine patch monotherapy group showed significantly decreased non-aggressive agitated behaviors. The combination treatment showed significantly increased Korean Version of the Cohen Mansfield Agitation Inventory total scores and non-aggressive agitated behaviors. Neither monotherapy nor combination therapy affected the appearance of new agitation symptoms.

Oxidative DNA damage, DNA susceptibility to oxidation, and the ratio of reduced to oxidized glutathione measured in blood were compared in 67 patients with AD treated with an AChEI as monotherapy and in combination with memantine 20 mg/day (Akkaya et al. 2017). DNA strand breaks and H₂O₂-induced DNA damage were found to be lower in the combination therapy recipients compared with AChEI therapy alone. No effect was seen on oxidized purines and the ratio of reduced to oxidized glutathione.

Memantine ER was evaluated in a 24-week, randomized trial in patients with moderate-to-severe AD receiving an AChEI (Grossberg et al. 2018). Post-hoc analyses were carried out using Clinician’s Interview-based Impression of Change Plus Caregiver Input, baseline to endpoint changes on the SIB and ADL. A significantly higher number of patients achieved an early response that was maintained on SIB, NPI, and the Clinician’s Interview-based Impression of Change Plus Caregiver Input in the combination group who maintained a clinically notable response on ADL/NPI, SIB/ADL/NPI.

The efficacy of add-on memantine (20 mg/day final dose) to donepezil for 6 months was tested in a retrospective study in 483 patients with a concomitant diagnosis of moderate-to-severe AD and mild-to-moderate chronic obstructive pulmonary disease (Cao et al. 2020). As assessed by standardized MMSE scores, the combination was more effective than donepezil alone in this patient population. A prospective, randomized, 6-month clinical trial in 85 patients with moderate AD compared the effects of memantine and donepezil on behavioral and psychological symptoms of dementia (Bago Rozankovic et al. 2021). The NPI total score and all subdomains were improved from baseline to month 6 in both treatment groups, with the exception of euphoria and apathy, for which improvements were seen with donepezil only.

Youn and colleagues tested combination treatment with memantine and AChEIs in moderate-stage AD in 148 patients in a double-blind, randomized, placebo-controlled study for 12 weeks (Youn et al. 2021). The combination treatment was superior to AChEIs alone in the NPI-disinhibition score. A multicenter, randomized, open-label, prospective study evaluated the effect of memantine (titrated to 20 mg/day) on speech function in moderate-to-severe AD patients (164 completed) on donepezil (Kim et al. 2023). The efficacy of the combination therapy was not better than donepezil at 24 weeks as measured by K-WAB scores used to evaluate language function.

According to a study by Havreng-Théry and colleagues, the combination of memantine with AChEIs results in reduced mortality in nursing home residents, as shown by a longitudinal analysis of over 25,000 patients with dementia (Havreng-Thery et al. 2024). Residents who had received monotherapy (memantine or an AChEI), combination therapy, or no anti-dementia therapy were included in the study analysis. The adjusted risk ratio for mortality showed that each anti-dementia treatment was effective and the best effect was observed after combination treatment. In an unadjusted model, memantine was not shown to be effective.

Gareri and colleagues analyzed the effect of addition of citicoline (an acetylcholine precursor) to memantine (10 to 20 mg/day) plus AChEIs in combined datasets of two studies (CITIMEM and CITIDEMAGE) involving 295 patients to increase analysis power (Gareri et al. 2024). The primary outcome measure was cognitive function assessed by MMSE. Secondary measures involved daily life functions, mood, and behavioral symptoms assessed by ADL and Instrumental ADL, Geriatric Depression Scale, and NPI Scale. A positive effect of citicoline was noted on several measures (MMSE, Geriatric Depression Score, ADL).

Combination with vitamins

In a review by Annweiler and Beauchet it was proposed that the combination of memantine and vitamin E may provide greater neuroprotective effects than either treatment alone by simultaneously addressing several different pathophysiological processes leading to neurodegeneration (Annweiler et al. 2012). The authors reported the start of recruitment for a double-blind, randomized, placebo-controlled, parallel-group, intent-to-treat, clinical trial addressing this aspect (AD-IDEA trial; ClinicalTrials.gov identifier: NCT01409694). Thereafter, in a pilot study in 43 white outpatients (mean age 84.7 ± 6.3 years; 65.1% women) with a new diagnosis of AD and naïve to anti-dementia drugs or vitamin D supplements, it was shown that patients who took memantine plus vitamin D for 6 months had a statistically and clinically relevant gain in cognition using MMSE scores compared with each individual treatment alone (Annweiler et al. 2012). A retrospective pilot analysis was performed on 48 patients with AD treated with memantine for 6 months in relation to the absence or presence of vitamin D deficiency (Lemire et al. 2018). After 6 months, memantine was associated with improved cognitive performance, as assessed by MMSE scores, only in patients with vitamin D deficiency.

In a double-blind, placebo-controlled, parallel-group, randomized clinical trial in 613 subjects (ca. 150 per arm) with mild to moderate AD, treatment with vitamin E, memantine and their combination was tested for 4 years (Dysken et al. 2014). The ADCS-ADL score was used as the primary outcome measure; secondary outcomes included cognitive, neuropsychiatric, functional, and caregiver measures. A significant negative treatment interaction between alpha tocopherol and memantine was seen for the primary outcome. Compared with placebo, alpha tocopherol produced a significant delay in decline, as assessed by the ADCS-ADL. In secondary outcomes, there was no significant effect except a comparison of alpha tocopherol and memantine Caregiver Activity Survey.

Behavioral disturbances

Memantine produces beneficial behavioral changes and is effective in treating and preventing the behavioral symptoms of moderate-to-severe AD (Gauthier et al. 2008). In a retrospective analysis of more than 21,000 AD patients, Lachaine and coworkers compared the need for antipsychotic use before and after treatment with memantine and acetylcholinesterase inhibitors (AChEIs) (Lachaine et al. 2013). They found that the use of memantine, but not AChEIs, was associated with a significant decrease in the use of antipsychotics. In another study in 38 patients with AD, behavioral and psychological symptoms of dementia and cognitive function were analyzed in relation to concomitant use of antipsychotics (Suzuki et al. 2013). Memantine use allowed significant reductions in the dosages of concomitantly used psychotropic drugs (e.g., risperidone, diazepam) (Suzuki et al. 2013).

Ballard and colleagues compared memantine (20 mg) and antipsychotics for the treatment of neuropsychiatric symptoms in a double-blind, randomized, placebo-controlled withdrawal study in 200 patients with AD (Ballard et al. 2014). At 24 weeks, there was a trend in favor of memantine on the Bristol Activities of Daily Living Scale, a measure of function, and in favor of antipsychotic use on the Cohen-Mansfield Agitation Inventory, a measure of agitation. Individuals treated with antipsychotics were less likely to show relapse of neuropsychiatric symptoms, while memantine use was associated with a trend for improvement on the MMSE at 24 weeks (Ballard et al. 2014).

In 12 patients with AD treated with memantine (20 mg/day for 4 weeks) the NPI was performed for assessment of behavioral and psychological symptoms of dementia, the MMSE for cognitive function, and polysomnography for evaluation of sleep architecture (Ishikawa et al. 2016). Treatment with memantine decreased the NPI score (e.g., anxiety and irritability/lability), prolonged sleep time, increased sleep efficiency and time spent in stage II, and decreased nocturnal awakening.

Results of systematic reviews and metal analyses

Results of a meta-analysis by Kishi et al. (Kishi et al. 2017b) of randomized controlled trials of memantine (as monotherapy or as add-on treatment) versus control in patients with AD suggested the benefits of memantine in the treatment of most of the behavioral disturbances assessed. The individual behavioral disturbances were delusion, hallucination, agitation/aggression, dysphoria, anxiety/phobia, euphoria, apathy, disinhibition, irritability/lability, aberrant motor activity/activity disturbances, nighttime disturbance/diurnal rhythm disturbances, and eating disturbances.

A network meta-analysis (Dou et al. 2018) and a systematic review with individual patient data network meta-analysis (Veroniki et al. 2022) of randomized controlled trials of cognitive enhancers approved for the management of AD (memantine and the AChEIs, galantamine, donepezil, rivastigmine, alone or in combination) compared and ranked the efficacy and tolerability of these drugs; for moderate-to-severe AD, memantine showed the best profile of acceptability, i.e., best balance between efficacy and safety (Dou et al. 2018). Additionally, a subsequent meta-analysis of 54 randomized, placebo-controlled trials found that the combination of memantine and donepezil was superior for effects on cognitive and neuropsychiatric symptoms, global assessment, and daily activities, but was less acceptable than memantine or placebo monotherapy (Guo et al. 2020).

Recent clinical trials of the safety and efficacy of memantine in patients with AD

Clinical trials from the clinical trails.gov database (http://clinicaltrials.gov/study/NCT02553928, NCT02580305, NCT01626391, NCT02080364; accessed on 28 December 2024, Table 1) were identified as investigating the safety and efficacy of memantine dosing and comparing memantine with other established and new drugs.

Table 1.

Summary of recent clinical studies at clinical trails.gov database which included the safety and efficacy of memantine in patients with AD

AD	Study type	Study duration	Treatment	N	Redouts/scales	Results	References
Probable AD, aged ≥50 yrs	r, db, Phase IV	12–16 weeks	Memantine 20 mg once daily and placebo once daily vs. memantine 10 mg BID or placebo BID	62	Difference in ADCS-CGIC, incidence of adverse events	The efficacy and safety of memantine did not differ between dosing as a single dose of 20 mg/day or as 10 mg BID	http://clinicaltrials.gov/study/NCT02553928
Moderate AD	mc, r, pc, db, Phase IIa	26 wks	SUVN-502 50 mg (n = 190) or 100 mg (n = 185) vs. placebo (n = 189) as adjunct treatment with memantine and donepezil	564	Change from baseline in ADAS-Cog at wk 26	Neither low- nor high-dose SUNV502 was superior to placebo as adjunct treatment with memantine and donepezil, according to a mixed model analysis.	http://clinicaltrials.gov/study/NCT02580305
Mild-to-moderate AD	r, pc, db, Phase II	4 wks	TRx0237 125 mg BID (n = 5) in addition to AChEIs (donepezil, galantamine, or rivastigmine) and/or memantine vs. placebo (n = 4)	9	Adverse events	TRx0237 showed a favorable safety and tolerability profile	http://clinicaltrials.gov/study/NCT01626391
Mild AD	mc, pc, db, RCT, Phase III	72 wks	Azeliragon (TTP488) in patients receiving AChEIs and/or memantine	880	ADAS-Cog, CDR-sb	TTP488 was not superior to placebo on a background of an AChEI and/or memantine.	https://clinicaltrials.gov/study/NCT02080364

AD Alzheimer’s disease, ADAS-cog AD Assessment scale-cognitive subscale, ADCS-CGIC AD Cooperative study-clinical global impression of change, BID Twice daily, CDR-sb Clinical dementia rating scale sum of boxes, db Double-blind, mc Multicenter, pc Placebo-controlled, r Randomize

Memantine dosage

A randomized, double-blind study in 62 patients aged ≥ 50 years with probable AD(according to the National Institute of Neurological and Communicative Disorders andStroke and Alzheimer's Disease and Related Disorders Association [NINCDS-ADRDA]criteria) investigated whether cognition, behavior, social competence, and daily livingskills were affected differently if a 20 mg/day memantine dose was administered once daily or as 10 mg BID (http://clinicaltrials.gov/study/NCT02553928, accessed on 28December 2024). The structured interview, the Alzheimer's Disease Cooperative Study—Clinical Global Impression of Change (ADCS-CGIC) was used to evaluate these criteria;differences in the frequency of adverse events was also assessed. At 12 weeks, nodifferences between the dosing groups in the ADCS-CGIC scale were detected. The meanMMSE was also similar between the single and twice-daily dosing groups, 11.4 ± 4.2 and11.4 ± 4.5, respectively, and there was no between-group difference in the incidence ofadverse events (50% each). It was concluded that the dose distribution throughout the dayis not critical in determining the efficacy and safety of memantine; memantine can beadministered once or twice a day at a total dose of 20 mg/day

Comparison of memantine with other new and established compounds

In another multicenter, placebo-controlled, double-blind, phase 2A study (Nirogi et al.2022 )(NCT02580305); Table 1 ) conducted over 26 weeks, the efficacy and safety ofSUVN502 (5‐HT6 receptor antagonist, masupirdine) as adjunct treatment in 564 patientswith moderate AD concomitantly treated with donepezil and memantine was comparedwith placebo. The primary outcome was cognitive performance, measured on the 11-itemcognitive subscale of the ADAS-Cog (ADAS-Cog 11). Compared with the placebo group(n = 189), there were no significant differences in change from baseline in ADAS-Cog 11scores at week 26 in the low-dose SUNV502 50 mg (n = 190) or the high-dose SUNV502100 mg (n = 185) groups. SUNV502 is therefore not considered better than standardtreatment with memantine or donepezil. In a small double-blind, placebo-controlled,randomized 4-week study (http://clinicaltrials.gov/study/NCT01626391, accessed on 28December 2024), the safety and efficacy of an inhibitor of tau aggregation, TRx0237 30 or 60 mg/day (n = 5), as adjunctive therapy with AChEIs and/or memantine, was compared with placebo (n = 5) in patients with mild-to-moderate AD. Adverse events were observedin 80% of the TRx0237 group and 100% of the control group; no serious adverse events were observed in either group. Authors described TRx0237 as safe and tolerable despite the low number of patients in the study. Another study investigated the efficacy and safety of azeliragon (TTP488) (http://clinicaltrials.gov/study/NCT02080364, accessed on 28 December 2024) in 880 patients with mild AD over 18 months in a double-blind,randomized, Phase III study; azeliragon was compared to placebo treatment that included intake of either an AChEI and/or memantine. At study end, scores for the primary outcome parameters, ADAS-Cog and Clinical Dementia Rating Scale Sum of Boxes, did not differsignificantly between groups; azeliragon treatment was not superior to placebo.In summary, clinical trials investigating memantine between 2014 and 2024 have shown,firstly, that the distribution of doses throughout the day is not critical to efficacy and safety, and memantine can be administered once or twice daily at a total dose of 20mg/day. Furthermore, it was demonstrated that several new therapeutic agents such asSUNV502 and azeliragon did not appear to be superior to memantine treatment.

New publications supporting the use of memantine beyond AD

Preclinical and clinical

For details of the design of the clinical studies and a summary of results, see Table 2. Memantine was generally administered at a starting dosage of 5 mg/day, and titrated up by 5 mg weekly to 20 mg/day. A scoring system was used to evaluate the strength of evidence for different indications on a scale of 0–5. Scoring was performed separately for preclinical and clinical evidence and multiple criteria were considered: (1) the number and quality of available studies, (2) the use of therapeutically relevant doses, (3) the validity of employed models and tests, (4) the study design, particularly for clinical studies, (5) the effect size of observed outcomes, and (6) the statistical power based on patient numbers. In this scoring system, “0” indicates no or very weak evidence, while “5” represents very strong evidence supported by multiple high-quality studies showing consistent positive results. In our scoring system, a value of 1 represents minimal evidence and 5 represents maximal evidence. We applied a factor of 5 to clinical scores to reflect the higher evidentiary value of clinical studies compared to preclinical data, in line with established hierarchies of evidence in research. This weighting is intended to ensure that robust clinical findings have a proportionally greater influence on the overall assessment, acknowledging that clinical data are generally more directly relevant to patient care than preclinical results The overall rating for an indication is derived from the weighted combination of preclinical and clinical scores, with clinical evidence weighted 5-fold higher (Table 3). The scoring system is based on the authors’ expert judgment and is intended as an orientation tool rather than a validated ranking of evidence. While similar approaches are sometimes used in narrative reviews to synthesize heterogeneous data, this particular method has not been formally validated and should be interpreted accordingly. Importantly, none of this evidence should be used as a recommendation to support the use of memantine beyond licensed indications.

Table 2.

Summary of clinical studies of memantine since 2014 by indication: not related to Alzheimer’s disease

Subpopulation/aspect	Study type	Study duration	Treatment (n)a	N	Redouts/scales	Results	Reference
Attention deficit hyperactivity disorder (ADHD)
Children	db, RCT	6 wks	Memantine 5 mg BID to 10 mg BID (n = 22) Methylphenidate (n = 18)	40	Attention deficit hyperactivity rating scale, CGI-S scale	At 6 wks, better response with methylphenidate than with memantine; time-dependent reduction of symptoms with memantine; however, the significance could not be judged due to the lack of a placebo control	Mohammadi et al. (2015)
Adults	Pilot, open-label prospective study	12 wks	Memantine titrated to 10 mg BID (n = 34)	34	Adult ADHD investigator symptom report (AISRS), CGI, neuropsychological battery of executive function, and the CANTAB cognitive battery	Memantine improved AISRS (total symptoms, inattentive symptoms, hyperactive symptoms), cognitive performance (attention, working memory, and some executive domains). Simple reaction time declined	Surman et al. (2013)
Adults	Pilot, db pc, RCT	12 wks	Memantine 5 to 10 mg/day (n = 12) Placebo (n = 14)	26	Executive functions (BRIEF scale) and Cambridge neuropsychological test automated battery	No significant effect was observed	Biederman et al. (2014).
Alcohol use disorder
Adults with alcohol use disorder recently detoxified	db, pc, escalating-dose, pilot study	4 wks	Memantine 10 to 30 mg/day (n = 18) Placebo (n = 18)	18	Alcohol consumption/compliance; other outcomes included effects on cognition, alcohol craving and affective measures	The memantine and placebo groups reported equivalent adverse events and clinical severity; no effects of memantine were observed in secondary outcome measures; memantine was not effective in this trial	Lewis et al. (2020)
Adults with or without a family history of alcoholism	db, pc, RCT	Single dose	Memantine 40 mg (n = 15) Placebo (n = 15)		Go/No-Go behavior and functional MRI	In functional MRI, right cingulate activation differed between subjects with and without a family history of alcoholism during No-Go correct rejects; memantine decreased effects in the former group	Jamadar et al. (2012)
Bipolar II disorder with comorbid alcohol dependence	Single-arm, pilot study	12 wks	Add-on memantine 5 mg/day to ongoing valproic acid (500-1000 mg) (n = 45)	45	HAM-D, YMRS, alcohol use, plasma cytokine levels, BDNF levels	Compared with baseline, memantine decreased HAM-D and YMRS scores, reduced alcohol use and plasma cytokine levels (IL-8), and increased plasma BDNF levels	Lee et al. (2018)
Heavy drinkers and positive family history of alcoholism	r, db, ph II, crossover study; 6–8 days of treatment 1, 6-day washout, 6–8 days of treatment 2	4 wks	Memantine up-titrated to 20 mg/day + naltrexone, washout, naltrexone (n = 26) Naltrexone, washout, naltrexone + memantine up-titrated to 20 mg/day (n = 30)	56	Number of drinks consumed, alcohol craving, alcohol-induced stimulation/sedation	Memantine enhanced the efficacy of naltrexone and reduced alcohol use and craving; benefits continued after memantine termination	Krishnan-Sarin et al. (2020)
Autism spectrum disorder
Adults	Open-label	12 wks	Memantine 5 to 20 mg/day	18	CGI-I, Cambridge neuropsychological test automated battery	Memantine produced a reduction in CGI-I subscale, decreased ADHD, improved anxiety, nonverbal communication, and self-rated executive function and neuropsychological assessments	Joshi et al. (2016)
Children	db, pc, RCT with open-label follow-up	12 wks/48 wks follow-up	Memantine 3 to 15 mg/day	121	Caregiver/parent ratings on the social responsiveness scale (SRS). Core autism treatment scale-improvement (CATS-I). core and associated autism symptom treatment scale (CAASTS)	No improvements over baseline were shown for any measure and one communication measure (secondary outcome) revealed significant worsening after 12 wks	Aman et al. (2016)
Children	db, r, pc	10 wks	Memantine titrated to 20 mg/day as add on to risperidone (3 mg/day)	20/20	Irritability subscale of aberrant behavior checklist–community	Memantine enhanced reduction of ABC-C subscale scores for irritability, stereotypic behavior and hyperactivity	Ghaleiha et al. (2013a)
Children (aged ≤ 14 yrs)	R, single-blind, pc	3 mos	Memantine 5 mg/day as adjunct to applied behavior analysis	30	Aberrant behavior checklist–community (ABC-C).	In the memantine group, there were greater reductions in subscale scores for irritability, stereotypic behavior and hyperactivity than in the placebo group	Karahmadi et al. (2018)
Children	r	8 wks	Memantine or risperidone	30	Autism behavior checklist (ABC), childhood autism rating scale (CARS), CGI-I and CGI-S	Both treatments were effective; the memantine benefit was weaker	Nikvarz et al. (2016)
Children	Variable design mc study	12–50 weeks	Memantine ER 3 to 15 mg/day	765	Social Responsiveness Scale	No effect was observed	Hardan et al. (2019)
Children with autism spectrum disorder	db, pc, RCT	24 wks	Memantine 3 mg/day to 6 to 12 mg/day	23	Apraxia and expressive language; secondary outcomes, memory and adaptive behavior measures	Memantine improved verbal recognition memory (p=0.03), as measured by the Narrative Memory-Recognition	Soorya et al. (2021)
Bipolar disorder
Elderly patients (aged >60 years) with bipolar disorder hospitalized for acute phase of mania	db, pc, RCT	8 wks	Memantine 5 to 20 mg/day (n = 35) or placebo (n = 35) as add on to sodium valproate	70	Mania severity score, cognition, and quality of life	Mania severity score was reduced significantly more in the memantine than the placebo group (p=0.038)	Omranifard et al. (2018)
Manic phase of bipolar disorder type I with obsessive compulsive symptoms	db, pc, RCT	16 wks	Memantine 5 to 20 mg/day (n = 19) or placebo (n = 19) as add on to routine medication (lithium + olanzapine + clonazepam)	38	Y-BOC behavior scale and adverse events	More memantine than placebo recipients showed >34% improvement in obsessive compulsive symptoms (78.9% vs 36.8%; p < 0.01) with no serious adverse events	Sahraian et al. (2017)
Compulsive buying disorder
Compulsive buying	Open-label	10 wks	Memantine 10 to 30 mg/day (n = 9)	9	Y-BOCS–shopping version (Y-BOCS-SV) scores	Y-BOCS-SV scores were decreased by ca. 50% at study end; memantine decreased impulsive buying and improved cognitive tasks of impulsivity	Grant et al. (2012)
Coughing
Healthy subjects or subjects with acute viral upper respiratory tract infection	db, r, pc, crossover study		Memantine 20 mg or placebo 6 h prior to capsaicin cough challenge	14/14	Cough reflex sensitivity	After capsaicin cough challenge, in healthy subjects, memantine inhibited cough reflex sensitivity (p = 0.034), in the viral infection group, only a trend was seen	Dicpinigaitis et al. (2015)
COVID-19
Dementia patients	Retrospective study	Various	Various	279	Mortality	Analysis revealed no effect on mortality in COVID patients	Park and Kwon (2021)
Dementia with Lewy bodies and Parkinson’s disease with dementia
Dementia with Lewy bodies and Parkinson’s disease dementia	db, pc, RCT	24 wk rct, 36-mo open- label treatment	Memantine 20 mg/day	75	CGI-C, survival	At 36-mo follow-up, compared with placebo, survival was increased in the memantine-treated group (p = 0.045); memantine responders had higher rates of survival compared with the non-responders (p = 0.010)	Stubendorff et al. (2014)
Depression and bipolar II disorder
Bipolar II disorder	db, pc, RCT	8 wks	Memantine 5 to 20 mg/day over 4 wks, 20 mg/day for 4 wks (n = 15) or placebo (n = 15) as adjunct to lamotrigine 100 mg or more	29	HAM-D, and other behavioral measures, weekly for up to 8 wks.	At 8 wks, there was no significant effect of memantine vs placebo in HAM-D (trend for improvement), at wk 4 during the titration phase there was a significant reduction with memantine in HAM-D scores compared with baseline (p = 0.007)	Anand et al. (2012)
Bipolar II disorder	db, pc, RCT	12 wks	Memantine 5 mg/day (n = 106) or placebo (n = 108) as add on to valproate	214	HAM-D and YMRS (clinical symptoms), plasma cytokine levels – TNF-α, IL-6, IL-8, and IL-1 levels.	Memantine decreased TNF-α and other cytokine levels; clinical symptoms were not improved	Lee et al. (2014a)
Bipolar II disorder	db, pc, RCT	12 wks	Memantine 5 mg/day (n = 115) or placebo (n = 117) as add on to valproate	157	HAM-D and YMRS. Tested BDNF genotype variant (66Met vs. Val66) on response to memantine	In the Met variant, a stronger effect of memantine was observed	Lee et al. (2014b)
Bipolar II disorder, moderate mood symptoms	db, pc, RCT	12 wks	Memantine 5 mg/day as adjunct to valproate (n = 66) Dextromethorphan as adjunct to valproate (n = 69) Placebo as adjunct to valproate (n = 69)	201	HAM-D and YMRS, blood TNF-α, CRP, BDNF, metabolic profile analysis (cholesterol, triglycerides, glycosylated hemoglobin [HbA1C], fasting serum glucose, BMI).	Memantine + dextromethorphan decreased depressive symptoms and increased plasma BDNF levels compared with placebo; neither treatment alone was effective	Lee et al. (2020)
Bipolar II disorder	db, pc, RCT	12 wks	Memantine 5 mg/day (n = 155) or placebo (n = 170) as add on to valproate	325	YMRS, HAM-D, cognition and inflammation (cytokine plasma levels: TNF-α, CRP, IL-8, and BDNF).	In a subpopulation of middle-aged to elderly patients, memantine improved cognitive performance attention and decreased plasma TNF-α levels.	Lu et al. (2021)
Depression, elderly patients aged >60 years	db, pc, RCT	8 wks	Memantine 20 mg/day (n = 28) or placebo (n = 29) add- on to citalopram	57	Geriatric depression scale, HAM-D, World Health Organization quality of life (WHOQOL-BREF)	No significant antidepressant effect or improvement in QOL was shown with memantine as add-on to citalopram	Omranifard et al. (2014)
Depression Elderly patients	db, pc	12 wks	Memantine 10 then 20 mg/day	35	HAM-D, apathy evaluation scale, functional recovery score, helplessness scale.	There was no effect of memantine on any outcome	Lenze et al. (2012)
Elderly patients with memory complaints	db, pc	6 mo	Memantine 5 to 20 mg/day	62	Depression HAM-D scores and serum cytokine levels	There was no effect of memantine on severity HAM-D score and on cytokine factor scores. Memantine was suggested to improve executive functions.	Van Dyk et al. (2020)
Depression Elderly subjects	db, pc, RCT	3 mo	Memantine 5 to 20 mg/day (n = 23) or placebo (n = 18) as adjunct to escitalopram	41	Structural neuroplastic changes: gray matter volume and cortical thickness (assessed by high-resolution T1-weighted MRI imaging), HAM-D	Higher remission rate (HAM-D ≤ 6) in memantine than in placebo recipients, 62% vs 43%; no effect in HAM-D scoring; with memantine relative to placebo, gray matter volume and thickness were increased more in the left middle and inferior temporal lobe, right medial, and lateral orbito-frontal cortex	Krause-Sorio et al. (2020a)
Follow-up study, Depression	pc, pilot study	6 mo	Memantine 5 to 20 mg/day as adjunct to escitalopram (n = 22) Placebo as adjunct to escitalopram (n = 16)	38	Gray matter volume and cortical thickness (assessed using high-resolution T1-weighted MRI imaging); HAM-D	No differences in remission or HAM-D scoring; increase in fractional anisotropy corresponded to greater improvements in depressive symptoms with memantine +escitalopram-treated patients, but not placebo+escitalopram	Krause-Sorio et al. (2020b)
Elderly patients, major depression	db, pc, RCT	6 mo	Memantine 10 to 20 mg/day as add on to citalopram	62	HAM-D, Apathy Evaluation Scale, Connor-Davidson Resilience Scale, Cumulative Illness Rating Scale for Geriatrics, California Verbal Learning, Memory Functioning Questionnaire scale	No effect of memantine on HAM-D but stronger improvement in delayed recall and executive function	Lavretsky et al. (2020)
Major depression	db, pc, RCT	8 wks	Memantine titrated from 5 to 20 mg/day	15+16	Montgomery-Asberg Depression Rating Score; secondary measures: other depression and anxiety rating scales, suicidal and delusional ideation.	No effect of memantine was observed in any applied measures	Smith et al. (2013)
Depression—electroconvulsive therapy-induced cognitive deficit
Depression ECT-induced cognitive impairment	db, pc	9 days	Memantine 5 mg/day (n = 20) Placebo (n = 20)	40	MMSE	Memantine significantly improved MMSE scores, suggesting that memantine may reduce cognitive impairment following ECT	Abbasinazari et al. (2015)
Diabetic neuropathy/diabetes
Type 2 diabetes mellitus with confirmed diabetic neuropathy	Open label r, pc	8 wks	Memantine 5 mg/day to 10 mg BID (n = 72) or placebo (n = 71) as adjunct to gabapentin (300 mg)	143	Douleur Neuropathique 4 questionnaire	At 8 wks, memantine decreased the questionnaire score and the number of patients with diabetic neuropathy	Jafarzadeh et al. (2023)
Down syndrome
Down syndrome	db, pc, RCT	16 wks	Memantine 20 mg/day (n = 81) Placebo (n = 79)	160	Cognitive and adaptive effects. California verbal learning test	There was no significant effect relative to placebo, possibly due to low memantine plasma levels	Costa et al. (2022)
Down syndrome	db, pc, RCT	52 wks	Memantine 5 mg/day followed by 10 mg/day (n = 88) Placebo (n = 85)	173	Change in cognition and function—measured by the DAMES scores and the adaptive behavior scale (ABS) parts I and II.	Both groups declined in cognition and function; no benefit with memantine was observed	Hanney et al. (2012)
Dyskinesia after L-DOPA in Parkinson’s disease
Dyskinesia after L-DOPA in Parkinson’s disease	db, pc, RCT crossover study	3 wks	Memantine 20 mg/day	15	Primary outcome: change in dyskinesia ratings (video rating), as assessed by two independent and blinded investigators; secondary outcome: diary recording of changes in time spent in ‘on-with-dyskinesia’	Data from self-administered diaries indicated significant reduction in the percentage of time per day spent in ‘on-with-dyskinesia’ (p = 0.005); the primary outcome failed to show significance	Wictorin and Widner (2015)
ECT cognitive impairment
ECT cognitive impairment	db, RCT	With ECT treatment for 12 days	Memantine 5 mg, melatonin 3 mg	18+22	MMSE for cognition	Memantine treatment was better than melatonin at the last ECT treatment	Sarraf et al. (2020)
Epilepsy
Children with epileptic encephalopathy	db, r, pc, crossover study	6 wks, 2-wk washout, cross over 6wks	Memantine 5 to 15 mg/day	27	Seizure frequency, electroencephalography, caregiver clinical impression, serum inflammatory markers, neuropsychological evaluation	Memantine improved seizures and electroencephalography; neuropsychological evaluation suggested improvements in symptoms of ADHD and autism	Schiller et al. (2022)
Focal-onset epilepsy	db, pc, RCT and 13 wks’ open-label treatment	13 wks	Memantine titrated to 10 mg/day (n = 8) Placebo (n = 9)	17	Selective reminding test (SRT) continuous long-term retrieval (CLTR) score and 7/24 spatial recall test	No effect was observed in primary measures and there was no effect on depression, sleepiness, side effects, or seizure frequency	Leeman-Markowski et al. (2018)
Epileptic patients with mild to moderate cognitive impairment	db, pc, RCT, phase IIIB pilot study	16 wks	Memantine 5 to 10 mg/day (n = 29) Placebo (n = 30)	59	MMSE score, the Weshler's Memory Scale total score, self-rated quality of life and memory	Memantine significantly improved total MMSE score and the Weshler's memory scale; self-rated quality of life and memory were also improved	Marimuthu et al. (2016)
Epilepsy	r, parallel group	16 wks	Memantine 5 to 10 mg/day (n = 35) Donepezil 10 mg (n = 35)	70	MoCA test	The mean MoCA score was significantly improved (p < 0.001) from baseline in the memantine, but not in the donepizil-treated group	Oustad et al. (2020)
Fibromyalgia
Fibromyalgia	db, pc, RCT	6 mo	Memantine 5 to 20 mg/day (n = 13) Placebo (n = 12)	25	Patient-reported outcomes (by questionnaires) on pain, anxiety, depression, quality of life, and cognitive impairment, and single‐voxel magnetic resonance spectroscopy changes of the brain	Compared with placebo, memantine significantly increased glutamate (p = 0.010), glutamate/creatine ratio (p = 0.013), combined glutamate + glutamine (p = 0.016) and total N-acetyl-aspartate levels (p = 0.034 in the posterior cingulate cortex; and increased creatine and choline in the right posterior insula. Interpreted as positive effect	Fayed et al. (2014)
Fibromyalgia	db, pc, RCT	6 mo	Memantine 20 mg/day Placebo	63	Pain threshold (sphygmomanometer), pain visual analogue scale, secondary measures: the Cognition Mini-Exam (MEC), The Fibromyalgia Impact Questionnaire (FIQ), The Hospital Anxiety Depression Scale, The EuroQol-5D questionnaire, CGI, The UKU side effect rating scale	Memantine decreased ratings on a pain visual analogue scale and pain measured with a sphygmomanometer; secondary outcomes, except for anxiety, were also improved, including depression	Olivan-Blazquez et al. (2014)
Fragile X
Fragile X-associated tremor/ataxia syndrome	db, pc, RCT	1 yr	Memantine 5 to 10 mg BID (n = 43) Placebo (n = 45)	94	Behavioral dyscontrol scale score; computer-assisted tremor system measure of intention tremor severity	No benefit with memantine relative to placebo was observed with respect to the behavioral dyscontrol scale score or intention tremor severity	Seritan et al. (2014)
Fragile X-associated tremor/ataxia syndrome	db, pc, RCT	1 yr	Memantine 5 to 10 mg BID (n = 21)	41 (21)	California verbal learning test, behavioral dyscontrol scale score, controlled oral word association test; electroencephalography and event-related potential recordings	Cued-recall memory and N400 repetition effect amplitude were improved with memantine treatment	Yang et al. (2014)
Fragile X-associated tremor/ataxia syndrome	db, pc, RCT	1 yr	Memantine 5 to 20 mg/day (n = 24) Placebo (n = 24)	48	Attention and working memory, electroencephalogram	Memantine, but not placebo, improved attention/working memory, enhanced event related potential P2 amplitude; P2 amplitude was positively correlated with improvement of behavioral measure of attention/working memory	Yang et al. (2016)
Frontotemporal lobar degeneration
Frontotemporal lobar degeneration	db, pc, RCT	26 wks	Memantine 20 mg/day (n = 39) Placebo (n = 42)	81	NPI-Q or CGI-C	Memantine failed to affect either measure	Boxer et al. (2013)
Mild and moderate-to-severe frontotemporal lobar degeneration	open-label, self-controlled	6 mo	Memantine 20 mg/day	42	NPI-Q scores	Memantine had no effect on NPI-Q scores, with the exception of a positive effect on agitation subdomain; patients with moderate-to-severe status responded better to memantine (showed cognitive and functional benefits)	Li et al. (2016)
Huntington’s disease
Cognitive decline in patients with Huntington’s disease	Retrospective, treatment vs. no treatment	5 yrs	Memantine up to 30 mg/day	192/96	Symbol digit modalities test, stroop color naming test, stroop word reading test, and stroop interference test	No effect of memantine	Ogilvie and Schultz (2023)
Kleptomania
Individuals with kleptomania	Open-label	8 wks	Memantine 10 to 30 mg/day	12	Y-BOCS modified for kleptomania; secondary outcomes: kleptomania symptom assessment Scale (K-SAS), CGI-I and CGI-S scales, sheehan disability scale (SDS), hamilton anxiety rating Scale (HAM-A), HAM-D, perceived stress scale, quality of life inventory (QoLI), CANTAB, stop-signal reaction time (SSRT).	Kleptomania disease severity scores were decreased across all study measures, 91.7% of the 11 participants met the responder criteria (35% improvement on the primary effectiveness measure plus CGI improved/very much improved); improvements were also observed in mood, anxiety, disability scores and stop-signal response inhibition	Grant et al. (2013)
3,4-Methylenedioxymethamphetamine memory impairment
MDMA-induced memory impairment	db, r, pc intra-subject study	Single dose	Memantine 20 mg	15	Memory function assessments: visual verbal learning task, a prospective memory task, the sternberg memory task and the abstract visual pattern learning task	Pretreatment with memantine did not affect any MDMA-induced deficits	de Sousa Fernandes Perna et al. (2014)
Migraine
Migraine	db, pc, RCT	24 wks	Memantine 10 mg/day (n = 30) Placebo (n = 30)	60	Migraine frequency/mo, 50% responder rate, rescue medication use	At week 24, compared with placebo memantine improved migraine frequency/mo (p = 0.003) , rescue medication use (p = 0.0001) and the 50% responder rate (85.7% vs 51.7%; p = 0.005)	Shanmugam et al. (2019)
Patients without aura	db, pc, RCT	12 wks	Memantine 10 mg/day	60	Primary outcome: change from baseline in the attack frequency using migraine diary; secondary measures were assessed using several clinical, functional, and psychological tools	Memantine reduced attack frequency and the number of days absent from work, severity, and disability	Noruzzadeh et al. (2016)
Mild cognitive impairment
Mild cognitive impairment	db, pc	2–52 weeks	Memantine 10 mg BID as add on to galantamine (16 mg)	232	ADAS-cog	Memantine in combination with galantamine resulted in cognitive improvement in a subgroup with presumed AD etiology, the improvement was greater than that seen with galantamine alone	Peters et al. (2012)
Multiple sclerosis
Fatigue in patients with relapsing-remitting MS	db, pc, RCT	3 mo	Memantine 20 mg/day (n = 30) Placebo (n = 32)	64	Change in severity of fatigue assessed by the modified fatigue impact scale	Memantine failed to show efficacy in the management of MS-related fatigue	Falsafi et al. (2020)
Cognitive disorders in patients with relapsing-remitting MS	db, r, p-c multicenter study	52 wks	Memantine 20 mg/day (n = 50) Placebo (n = 43)	93	Primary endpoint: paced auditory serial addition test (PASAT), secondary: neuropsychological tests and the relapse rate	No therapeutic effect of memantine was noted on selected readouts; incidence of adverse events was increased	Peyro Saint Paul et al. (2016)
Neuropathic pain
Complex regional pain syndrome/neuropathic pain	Retrospective, open- label	2 yrs or more	Memantine titrated from 5 to 10 mg/day to maximum 40 to 60 mg/day	56	Quantitative visual analog scale, number of exaggerated burning attacks per month, Percentage of decrease of unpleasing feelings related to soft touch of the skin of the area affected by complex regional pain syndrome, time spent in the shower	13 patients showed complete remission from pain and the disappearance of allodynia for ≥ 9 months post-treatment; 18 patients showed partial improvement of visual analog scale and allodynia; eight patients showed no response In subjective improvement in short-term memory, nine patients showed much improvement, 14 some improvement, three showed no changes; in subjective feelings, 17 patients reported having a better quality of life, 3 felt no change	Ahmad-Sabry and Shareghi (2015)
Refractory neuropathic pain after ketamine infusion	r, p-c multicenter study	12 wks	Memantine 20 mg/day (n = 20), dextromethorphan (90 mg/day) (n = 20) or placebo (n = 20) as follow on after ketamine infusion	60	Primary outcome: pain intensity at one month; secondary endpoints: pain, sleep, anxiety, depression, cognitive function, and quality of life evaluations	Pain was lower after 12 wks of treatment with no between-group difference (placebo, dextromethorphan, or memantine); positive effects were observed in cognitive-affective domains and quality of life with memantine and dextromethorphan (p < 0.05)	Martin et al. (2019)
Neuropathic pain in women post mastectomy	db, pc, RCT	Follow-up 3 mo	Memantine 5 to 20 mg/day (n = 20) or Placebo (n = 20) from 2 wks prior to surgery and 2 wks after	40	Primary endpoint: pain intensity measured on a numerical rating scale (0–10)	Compared with placebo, memantine markedly improved post-mastectomy pain intensity, reduced the need for rescue analgesia and improved emotional state	Morel et al. (2016)
Nicotine abuse
Nicotine abuse	db, pc, RCT	1 year	Memantine 10 mg/day was combined with or without reactivation of associative cue-smoking memories, or reactivation with placebo on smoking cessation day	59	Primary outcome: time to relapse	Memantine failed to affect any measure of smoking outcome	Das et al. (2015)
Nystagmus
Acquired pendular nystagmus	Cross-over trial		Memantine 10 mg 4×day Gabapentin (300 mg 4×day)	16	Complete ophthalmic examination, oscillopsia measurements, vision-specific quality of life questionnaire, eye movement recordings/analysis	Memantine and gabapentin improved acquired pendular nystagmus, ocular motor parameters and distance oscillopsia; median near oscillopsia was not affected	Nerrant et al. (2020)
Oculopalatal tremor in Guillain-Mollaret triangle lesion	Open-label	6 Month	Memantine 20 mg	6	Video-oculography, subjective visual vertical assessed through the “bucket test”, vision-specific quality of life questionnaire, palatal tremor and facial movements, MRI	Memantine produced only a modest and transient improvement in nystagmus	Martins et al. (2022)
Obsessive-compulsive disorder
OCD	db, pc, RCT	12 wks	Memantine 10 mg BID (n = 35) or placebo (n = 35) as adjunct to sertraline	70	Y-BOCS as a measure of OCD symptoms, WCST as a measure of executive functions	The Y-BOCS total score and obsession and compulsion subscale scores were significantly improved, but there was no between-treatment difference; compared with placebo, the memantine-treated group had a greater response in the number of completed categories subscale of the WCST(p < 0.001)	Askari et al. (2022)
OCD	db, pc, RCT	8 wks	Add-on memantine 5 to 10 mg/day (n = 33), gabapentin (n = 33) or placebo (n = 33) to standard therapy (fluoxetine)	99	Y-BOCS	Neither memantine nor gabapentin treatment provided additional significant effects than that of the selective serotonin reuptake inhibitor, fluoxetine, on Y-BOC scores	Farnia et al. (2018)
OCD After termination of ketamine IV treatment	Open- label	6–12 wks	Memantine 5 to 10 mg BID (n = 12)	15	OCD symptoms-Y-BOCS, depression–HAM-D and anxiety–HAM-A	Overall, there was no effect	Rodriguez et al. (2016)
OCD	db, pc, RCT	12 wks	Memantine 5 to 10 mg/day (n = 20) or placebo (n = 20) as add-on to SSRI or clomipramine	40	Y-BOCS, CGI	Memantine had no effect on Y-BOCS scores and CGI. There was time related improvement and significant interaction between time and treatment.	Haghighi et al. (2013)
Treatment-resistant OCD patients	Open-label	12 wks	Memantine 5 to 10 mg/day as an augmenting agent	12	Y-BOCS	In eight patients, improvements were observed, ≥ 25% reduction in Y-BOCS	Bakhla et al. (2013)
Moderate-to-severe OCD	db, pc, RCT	8 wks	Memantine 10 to 20 mg/day (n = 21) or placebo (n = 21) as add-on to fluvoxamine	42	Y-BOCS	Y-BOCS showed significant effect of time and treatment interaction in total scale and obsession, by study end, all patients in the memantine group and six (32%) in the placebo group had partial or complete response	Ghaleiha et al. (2013b)
SSRI-refractory OCD patients	db, pc, RCT	12 weeks	Memantine 20 mg/day (n = 16) or placebo (n = 16) as add on to SSRI	32	Y-BOCS	Memantine significantly decreased mean Y-BOCS total score at the end of wks 8 and 12 (by 40.9%), with 73.3% of patients achieving treatment response; no effect was seen with placebo	Modarresi et al. (2018)
Opioid dependence/abuse
Opioid dependence/abuse	db, pc, RCT	12 wks	Memantine 40 mg/day as add-on to naltrexone ER	82	Primary outcome: retention in the 12-wk trial; secondary measures included: weekly rates of opiate use, asymptomatic on CGI-S scale and craving; ratings of opiate withdrawal symptoms and depression (HAM-D 21)	Memantine was not effective when used in combination with naltrexone ER as a relapse prevention strategy for opioid dependence	Bisaga et al. (2014)
Cognitive performance in opioid-dependent patients undergoing methadone-maintenance therapy	db, pc, RCT	12 wks	Memantine 5 mg/day (n = 42) or placebo (n = 38) as add-on to methadone	80	The Wisconsin card sorting test, the Conners’ continuous performance test, opiate treatment index	Memantine decreased amphetamine use and improved cognitive performance; other measures were not changed	Chang et al. (2015)
Opioid dependence/abuse	db, pc, RCT	13 wks	Memantine 15 mg/day (n = 24) or 30 mg/day (n = 27) or placebo (n = 29) as add-on to buprenorphine/naloxone	80	Primary outcomes: change in weekly mean proportion of opioid use, and cumulative abstinence rates after rapid buprenorphine discontinuation on week 9; secondary measures included: retention in treatment, opioid withdrawal symptoms, and cognitive control measured using the anti-saccade task	Memantine at 30 mg decreased likelihood for relapse after buprenorphine discontinuation and reduced opioid use during the last 2 weeks	Gonzalez et al. (2015)
Opioid dependence/abuse	db, pc, RCT	12 wks	Memantine 5 mg/day (n = 53) or placebo (n = 75) as add-on to methadone maintenance therapy	128	Primary outcomes: methadone dose required, retention rates, concomitant opioid use; secondary outcome: plasma TNF-α, CRP, IL-6, IL-8, and TGF-β1 levels and plasma BDNF levels	Add-on memantine permitted methadone dose lowering and decreased plasma TNF-α and increased TGF-β1 levels in opioid-dependent patients undergoing methadone maintenance therapy	Lee et al. (2015)
Optic neuritis
First episode of acute optic neuritis	db, pc, RCT	6 wks (+ 2-mo follow-up)	Memantine 10 mg/day (n = 20) Placebo (n = 18)	38	Retinal nerve fiber layer thickness, visual evoked potential, and visual acuity	Memantine reduced retinal nerve fiber layer thinning in three quadrants; visual acuity was not affected	Motamedi et al. (2022).
Parkinson's disease
Parkinson's disease	db, pc, RCT	16 wks	Memantine 20 mg/day (n = 11) Placebo (n = 14)	25	Primary measures, patient-reported outcomes: Goal Attainment Scaling (GAS), the Parkinson’s disease questionnaire-8 and the Zarit burden inventory	Memantine significantly improved GAS scoring and caregiver burden.	Leroi et al. (2014)
Parkinson's disease	db, pc, RCT	3 mo	Memantine 20 mg/day (n = 13) Placebo (n = 12)	25	UPDRS score, its axial sub score, axial hypertonia, the axial and overall dyskinesia rating scale and axial strength	Memantine lowered axial motor symptoms and dyskinesia, but did not affect other measures including gait	Moreau et al. (2013)
Parkinson's disease dementia and dementia with Lewy bodies	db, pc, RCT	24 wks	Memantine 20 mg/day (n = 21) Placebo (n = 30)	51	Attention (simple and choice reaction time) and word recognition (immediate and delayed) tests from the CDR system	Memantine improved choice reaction time, immediate and delayed word recognition	Wesnes et al. (2015)
Post-traumatic stress disorder
Civilian PTSD – all were women	Open-label	12 wks	Memantine 5 to 20 mg/day (n = 10) add-on to current medication	10	PTSD diagnosis and severity assessed with the Posttraumatic Diagnostic Scale (PDS)	Mean PDS scores were significantly reduced from baseline (p < 0.002)	Hori et al. (2021)
War veterans with PTSD and cognitive impairment	Open-label proof-of-concept study	16 wks	Memantine 5 to 20 mg/day	26	Several instruments for assessment of cognition; RBANS—measure of attention, language, visuospatial skills, and immediate and delayed memories; PTSD and disability scales	Memantine was associated with significant improvements (p < 0.05) in memory, cognition, core symptoms of PTSD, and depression	Ramaswamy et al. (2015)
Radiation therapy
Prevention of cognitive decline in adult patients with brain metastases receiving WBRT	db, pc, RCT	24 wks	WBRT + memantine 20 mg/day (n = 235) or placebo (n = 238) 3 days after radiotherapy initiation for 24 weeks	508	Serial standardized tests of cognitive function: Hopkins Verbal Learning Test-Revised, processing speed (Trail Making Test Part A), executive function (Trail Making Test Part B), verbal fluency (Controlled Oral Word Association), MMSE	Memantine produced a trend for positive effect but only 149 patients were available for analysis, memantine extended time to cognitive decline and improved executive functions	Brown et al. (2013)
Prevention of cognitive decline in adult patients with brain metastases receiving WBRT	r, Phase III study	Mean 7.9 mo	Hippocampal avoidance + WBRT + memantine up-titrated from 5 to 10 mg BID (ER 7 to 28 mg/day) (n = 261) WBRT plus memantine (n = 257)	518	Primary outcome, time to cognitive function failure; secondary outcome, overall survival, intracranial progression-free survival, toxicity, and patient-reported symptom burden	Cognitive deficits were lower after hippocampal avoidance + WBRT + memantine vs. WBRT + memantine, and there was less deterioration in executive function at 4 months and learning and memory at 4 and 6 months, respectively, and at 6 months there was less fatigue, less memory difficulties, and less difficulty with speaking and using imputed data, less interference of neurologic symptoms in daily activities and fewer cognitive symptoms	Brown et al. (2020)
Adult patients with brain metastases receiving WBRT	db, pc, RCT	24 wks	Memantine 20 mg/day (n = 246) or placebo (n = 246) within 3 days of initiating radiotherapy	508	Health-related quality of life and cognitive function	All domains of cognitive function declined over time, and quality of life	Laack et al. (2019)
Schizophrenia
Chronic psychotic disorders	db, r, pc cross-over study		Memantine 10 or 20 mg single dose (n = 41)	82	MATRICS Consensus Cognitive Battery (MCCB)	Memantine (20 mg) reduced MCCB performance in healthy subjects and in psychotic patients, improved prepulse-inhibition; single nucleotide polymorphism rs1337697 (in NMDA, GRIN3A gene) predicted greater positive response	Bhakta et al. (2016).
In patients with schizophrenia	pc, RCT	12 wks	Memantine 5 to 20 mg/day (n = 32) or placebo (n = 32) as add on to atypical antipsychotics	64	Global assessment of functioning and quality of life scale	Memantine increased global assessment of functioning and quality of life scale scores	Omranifard et al. (2015)
Negative symptoms in patients with stable schizophrenia	db, pc, RCT	8 wks	Memantine 20 mg/day (n = 20) or placebo (n = 20) as add-on to risperidone (6 mg/day)	40	Positive and negative syndrome scale, HAM-D extrapyramidal symptom rating scale	Memantine improved negative subscale, total score, and general psychopathology subscale scores significantly more than placebo (p < 0.002); there was no effect on any other measure	Rezaei et al. (2013)
Acute and chronic schizophrenia	Proof-of-concept, db, pc, RCT	6 or 24 wks	Memantine 10 mg BID or placebo as add on to risperidone (2-8 mg) for 6 wks in pts with acute episodes and mostly positive symptoms (n = 11) and for 24 wks in pts with chronic disease and negative symptoms (n = 13)	23	Psychopathological changes were assessed with the PANSS	Memantine in acute schizophrenia patients improved attention intensity (p = 0.043), problem-solving (p = 0.043), verbal learning (p = 0.050), and flexibility (p = 0.049); patients with chronic schizophrenia showed significantly higher immediate memory (p = 0.033) and a significantly greater reduction of the PANSS sum score	Schaefer et al. (2020)
Schizophrenia	db, pc, RCT	6 wks	Memantine 10 to 20 mg/day (n = 30) or placebo (n = 30) as add on to olanzapine 15 to 20 mg/day	60	PANSS	Memantine improved both positive and negative functions compared with olanzapine alone (p < 0.001), especially in female patients	Fakhri et al. (2016)
Stabilzed chronic schizophrenia	db, pc, RCT	8 wks	Memantine 20 mg/day (n = 20) or placebo (n = 20) as add-on to standard antipsychotic regimen	40	PANSS and brief assessment of cognition scale	Memantine improved verbal memory, learning, and verbal letter fluency, with no effects on psychotic symptoms	Hassanpour et al. (2019)
Chronic schizophrenia	db, pc, RCT	12 wks	Memantine 5 to 20 mg/day (n = 13) or placebo (n = 13) as add-on to conventional antipsychotic therapy	26	Korean version of the MMSE, PANSS, HAM-D, CGI-S and CGI-I scales	Compared with placebo, memantine was not associated with significantly improved cognitive test scores; a trend for improvement in the scores on the PANSS negative subscale was seen with memantine	Lee et al. (2012a)
Male hospitalized patients with schizophrenia	db, pc, RCT	12 wks	Memantine 5 to 20 mg/day (n = 23) or placebo (n = 23) as add on to risperidone 4 to 6 mg/day	46	PANSS, MMSE	Negative symptoms and cognitive symptoms, but not positive symptoms, improved significantly with memantine; no between-group difference in general psychopathologic symptoms	Mazinani et al. (2017)
Clozapine-treated refractory schizophrenia	Open label, 1-yr extension	52 wks	Add-on to clozapine	31	Memory and executive function using the CANTAB, the PANSS, and the CGI-S scale	Significant improvements were seen in PANSS negative and positive symptoms and PANSS total symptoms; no significant effect was seen on CGI-S scores	Veerman et al. (2016)
Stroke, hemorrhage, post-stroke recovery
Recovery - aphasia	db, pc, RCT	20 wks	Memantine 20 mg/day or placebo alone (0–16 wks), followed by combination with constraint-induced aphasia therapy wks 16–18, then memantine or placebo alone (wks 18–20)	28	Event-related potentials (P100 and N400), root mean square during a silent reading task	Memantine decreased event-related potentials/root mean square values (wk 16) which was enhanced with constraint-induced aphasia therapy (wks 16–18); changes were maintained afterwards (wks 18–20); changes corresponded to improvements in language performance	Barbancho et al. (2015)
Mild-to- moderate cerebral thromboembolic event	r, open-label	Acute	Memantine 20 mg TID for 5 days as add on to conventional treatment (n = 24) Control (conventional treatment) (n = 29)	53	Changes in National Institute of Health Stroke Scale (NIHSS) scores at day 1 and day 5	Changes from baseline in NIHSS scores were significantly different between the memantine and control group (p < 0.0001), suggesting improved neurological function	Kafi et al. (2014)
Traumatic brain injury or brain injury
Prevention of brain injury following open-heart surgery	pc, RCT	3 days	Memantine 20 mg/day 72 hr before surgery	34	Serum concentration of S100-B, measured by ELISA	Memantine decreased S100-B levels at the termination of CPB and 6 and 12 h later	Ziabakhsh Tabary et al. (2022)
Moderate TBI	db, RCT	7 days	Memantine 30 mg BID enteral (plus standard management) for 7 days (n = 22) Control (standard management) (n = 19)	41	Serum levels of neuron specific enolase (indicator of neuronal damage), Glasgow Coma Scale (symptoms)	Memantine lowered levels of enolase on day 3 and 7 and markedly improved Glasgow Scale Scores on day 3	Mokhtari et al. (2018)
Trichotillomania
Trichotillomania or skin picking	db, pc, RCT	8 wks	Memantine 10 to 20 mg/day (n = 55) Placebo (n = 45)	100	Pulling and picking severity assessed using the NIMH trichotillomania symptom severity scale, sheehan disability scale, CGI-S scale	Memantine reduced hair pulling and skin-picking symptoms (based on the number needed to treat = 1.9)	Grant et al. (2023)
Vascular dementia
Moderate to moderately-severe VaD, aged 50–85 years	r, no pc	24 wks	Memantine 20 mg/day (n = 65) Memantine 10 mg BID (n = 65)	130	ADAS-cog (primary endpoint), MMSE, CGI-C, CGI-S, HAM-D	Statistically significant benefit compared with baseline in both groups for primary endpoint (p < 0.01) and other scores (except HAM-D), no significant between-group differences; similar safety	Gavrilova et al. (2024)
Data for hospital patients diagnosed with VaD	Retrospective	6 wks	Memantine 10 mg BID as add on to scalp electroacupuncture (n = 55) Scalp electroacupuncture (n = 50) Memantine 10 mg BID (control) (n = 45)	150	Cognitive function, ADL and quality of life assessed by MoCA, Barthel index and dementia quality of life questionnaire; blood superoxide dismutase, lipid peroxide and NO	Post-treatment/follow-up scalp electroacupuncture with memantine showed a significantly greater therapeutic effect (improved questionnaire scores) than either treatment alone (p < 0.01 and p < 0.05)	Yue et al. (2020)

a Memantine was generally up-titrated from the lower dose to the upper dose over a period of weeks.

AD Alzheimer’s disease, ADAS-cog AD Assessment scale-cognitive subscale, ADHD Attention deficit hyperactivity disorder, BDNF Brain-derived neurotrophic factor, BID Twice daily, BRIEF Behavior rating inventory of executive functions CANTAB Cambridge neuropsychological test automated battery, CGI Clinical global impressions, CGI-C CGI of change, CGI-I CGI-improvement, CGI-S CGI-severity, CRP C-reactive protein, db Double-blind, ECT, Electroconvulsive therapy; ELISA Enzyme-linked immunosorbent assay, HAM-D Hamilton depression rating scale, IL Interleukin, mc Multicenter, MMSE Mini-mental status examination, mo Months, MoCA Montreal cognitive assessment, NPI Neuropsychiatric inventory, OCD Obsessive compulsive disorder, PANSS Positive and negative syndrome scale, pc Placebo-controlled, PTSD Post-traumatic stress disorder; r Randomized, RBANS Repeatable battery for the assessment of neuropsychological status, TBI Traumatic brain injury, TNF-α Tumor necrosis factor-α, WBRT Whole-brain radiotherapy, WCST Wisconsin card sorting test, wks Weeks, Y-BOCS Yale-Brown obsessive-compulsive scale, YMRS Young mania rating scale

Table 3.

Scoring of preclinical and clinical evidence for utility of memantine in different indications

Indication	Scoring preclinical	Scoring clinical	Total scoring (clin × 5)
ADHD		0	0
Alcohol use disorder	2	2	12
Aminoglycoside ototoxicity	1		1
Autism spectrum disorder	0	2	10
Bipolar disorder		2	10
Cardiovascular	1		1
Cerebellar ataxia	0		0
Chemotherapy	1		1
Chronic obstructive pulmonary disease	2		2
Compulsive buying disorder		1	5
Coughing		1	5
COVID-19	0	0	0
Dementia with Lewy bodies and Parkinson’s disease dementia		2	10
Depression and bipolar II disorders	3	3	18
Depression—electroconvulsive therapy-induced cognitive deficit		1	5
Diabetic neuropathy/diabetes	2	1	7
Down syndrome		0	0
Dyskinesia after L-DOPA in Parkinson’s disease		3	15
ECT cognitive impairment		3	15
Epilepsy	1	1	6
Fibromyalgia		2	10
Fragile X		2	10
Frontotemporal lobar degeneration		1	5
Gastric ulcers	1		1
Glaucoma	1		1
HIV	1		1
Huntington’s disease		0	0
Hydrocephalus	1		1
Inborn errors of metabolism	1		1
Kleptomania		1	5
Liver injury	2		2
Macular degeneration	1		1
Methamphetamine toxicity	2		2
Methylenedioxymethamphetamine memory impairment		0	0
Migraine		2	10
Mild cognitive impairment		1	5
Multiple sclerosis	1	0	1
Neuromyelitis optica spectrum disorders	1		1
Neuropathic pain	1	1	6
Nicotine abuse	0	0	0
Nystagmus		0	0
Obsessive-compulsive disorder	1	2	11
Opioid dependence/abuse		1	5
Optic neuritis		1	5
Osteoarthritis	1		1
Parkinson’s disease	1	1	6
Post-operative cognitive dysfunction	1		1
Post-traumatic stress disorder	0	1	5
Radiation therapy	1	1	6
Schizophrenia	1	3	16
Sepsis	2		2
Skin transplantation	0		0
Sporadic cerebral amyloid angiopathy	2		2
Stress-related disorder	0		0
Stroke, hemorrhage, post-stroke recovery	2	2	12
Tardive dyskinesia	0		0
Tinnitus	2		2
Traumatic brain injury	2	1	7
Trichotillomania		1	5
Trypanosoma	1		1
Tumor	1		1
Vascular dementia		1	5
Viral infections	1		1

Clinical scoring was increased by factor 5 in relation to preclinical data

Attention deficit hyperactivity disorder

Clinical Three clinical studies of memantine in patients with attention deficit hyperactivity disorder (ADHD) (Table 2) generally showed minor effects (Mohammadi et al. 2015; Surman et al. 2013; Biederman et al. 2014), including in children (Mohammadi et al. 2015). In adults with ADHD, in a small, prospective, open-label study, memantine was associated with improvements in symptoms and neuropsychological performance (Surman et al. 2013), and in a pilot, placebo-controlled trial, executive function deficits showed a trend for improvements with memantine on the Behavior Rating Inventory of Executive Functions scale (Biederman et al. 2014). Based on these studies, the evidence for the efficacy of memantine in ADHD is poor, which is reflected by the score below.

Rating: 0 (no effect observed)

Alcohol use disorder

Preclinical In a chronic alcohol exposure model in rats (28 consecutive days of alcohol [20%] and 6 h withdrawal) memantine (4 mg/kg) inhibited alcohol-induced phosphorylation of the NR1-CaMKII-ERK pathway in the prefrontal cortex and nucleus accumbens (Yuanyuan et al. 2018). Moreover, anxiety was reduced, as assessed by open field and elevated plus maze tests (Yuanyuan et al. 2018). In C57BL/ 6 mice provided 10% or 20% of ethanol as drinking fluid for 60, 90, or 180 days and treated with memantine (10 mg/kg per os) for the same duration, ethanol-induced spatial memory impairment (Morris water maze and radial maze), caspase-3 activation, and apoptosis in the hippocampus were decreased (Wang et al. 2018b). In another study, alcohol relapse-like behavior was investigated in 50 Wistar rats trained to associate several discrete stimuli with ethanol delivery; half were tested for memory reactivation followed by memantine 20 mg/kg administered twice (Vengeliene et al. 2015). Memantine reduced response on an ethanol-associated lever in a cue-induced ethanol-seeking test, which was followed by repeated extinction sessions and a reacquisition test. Overall, memantine disrupted the drug-cue association, which consequently interfered with relapse-like behavior (Vengeliene et al. 2015).

The effect of memantine (10 mg/kg) over 7 days for a binge-eating disorder was studied in rats through analysis of a licking microstructure using a 10% sucrose solution (Galistu and D'Aquila 2020). Treatment before testing reduced ingestion but an increase in burst number resulted in return of ingestion levels by the end of treatment. Post-session treatment induced a decrease in activation of licking behavior, followed by a slow recovery after treatment termination. These results suggest a reduced hedonic response. The inhibition in post-session administration mode may be due to the development of conditioned taste aversion (Galistu and D'Aquila 2020).

Rating: 2 (only 4 papers, two used impractically high doses)

Clinical A family history of alcoholism is associated with higher alcoholism rates and impulsiveness than in those without such a family history, potentially due to altered NMDA receptor function (Jamadar et al. 2012). Memantine (single-dose, 40 mg) was tested on impulsivity, Go/No-Go behavior and fMRI in 30 subjects with and without a family history of alcoholism (Jamadar et al. 2012). On fMRI, right cingulate activation differed between subjects with and without a family history of alcoholism during No-Go correct rejects; memantine decreased effects in the first group. In the case of No-Go correct rejects, compared with placebo, memantine decreased left cingulate and caudate activity in subjects with, but not in those without, a family history of alcoholism (Jamadar et al. 2012). Memantine (up-titrated to 20 mg/day) enhanced the efficacy of naltrexone in the suppression of alcohol drinking and craving in a randomized, cross-over, Phase II study in 56 heavy drinkers with a family history of alcoholism (Krishnan-Sarin et al. 2020); stimulation or sedation were not affected. Interestingly, the effect of memantine extended beyond treatment discontinuation. Lee and colleagues assessed memantine (5 mg/day) as add-on to valproic acid in patients with bipolar II disorder with comorbid alcohol dependence in a 12-week, one-arm trial (Lee et al. 2018). Compared with baseline, add-on memantine improved clinical symptoms, attenuated alcohol use and plasma cytokine levels, and increased BDNF levels. By contrast, Lewis et al. showed that memantine (10 to 30 mg/day) in 18 patients with alcohol use disorder who had recently detoxified, was no different than placebo in a 4-week, double-blind study (Lewis et al. 2020).

Rating: 2 (relevant doses, but mixed results)

Aminoglycoside ototoxicity

Preclinical The neuroprotective effect of memantine (20 mg/kg) was tested on spiral ganglion cell apoptosis induced by gentamicin injection in guinea pig cochleae (Kim et al. 2016). Memantine injected 12 and 1 h prior to gentamicin decreased FasL expression (a transmembrane protein belonging to the TNF family) 1-week later and activation of caspases 3, 8, and 9, indicating reduced apoptosis (Kim et al. 2016). The effect of memantine (10 mg/kg/day, 2 weeks) was also tested on amikacin cochlear toxicity in rats (Pavlidis et al. 2024). Histological analysis showed that memantine treatment parallel to amikacin, and less efficiently starting after amikacin, suppressed cochlea damage. However, distortion product otoacoustic emissions assessment failed to reveal a convincing effect supporting the histology results.

Rating: 1 (mixed results from only 2 papers)

Autism spectrum disorder

Preclinical Identifying novel therapeutic targets for autism spectrum disorder has gained attention due to a suggested role of aberrant glutamatergic transmission in the pathogenesis of associated cellular and behavioral deficits. In a valproic acid-induced rat model of autism, memantine administered repetitively at 20 mg/kg in combination with aripiprazole partially reversed reductions in BDNF expression (Zohny et al. 2023). Shank2, an excitatory postsynaptic scaffolding protein implicated in autism spectrum disorders, was studied in Shank2-KO mice, which have decreased NMDA receptor function and autistic-like behaviors. Repetitive memantine (20 mg/kg BID for 2 weeks) during postnatal days 7–21 prevented temporal NMDA receptor hypofunction and autistic-like behaviors at later stages. Additionally, memantine upregulated chromatin-related genes but suppressed mitochondria, extracellular matrix, and actin-related gene expression (Yoo et al. 2021). In prenatal valproic acid-induced autism in rats, memantine (10, 20 mg/kg per os on days 21 to 50) attenuated impairments in social interaction, spontaneous alteration, exploratory activity, intestinal motility, serotonin levels and prefrontal cortex mitochondrial complex activity (Kumar and Sharma 2016). Additionally, memantine attenuated increases in locomotion, anxiety, brain oxidative, inflammation, calcium, and BBB permeability (Kumar and Sharma 2016).

Rating: 0 (studies used very high dosing)

Clinical In a prospective, 12-week, open-label study in 18 adults with autism, Joshi et al. investigated the efficacy of memantine (15 to 20 mg/day) for social deficits (Joshi et al. 2016). A reduction in informant-rated and clinician-rated measures of autism severity were shown with memantine. Additionally, memantine improved ADHD and anxiety symptoms, and non-verbal communication, self-rated executive function, and neuropsychological assessments on the Cambridge Neuropsychological Test Automated Battery (Joshi et al. 2016).

The safety and efficacy of memantine 3 to 15 mg/day was evaluated in 121 children in a 12-week, randomized, placebo-controlled trial; by study end, there were no improvements relative to baseline in any of the outcome measures (Aman et al. 2016). Ghaleiha and colleagues studied a higher dose of memantine (titrated to 20 mg/day) as add-on to risperidone in children with autism in a 10-week placebo-controlled trial (Ghaleiha et al. 2013a). In the primary outcome measure (irritability subscale of Aberrant Behavior Checklist–Community) memantine resulted in greater reductions in subscale scores for irritability, stereotypic behavior and hyperactivity. Karahmadi and colleagues showed in a randomized, single-blind study that memantine (5 mg/day) as adjunct therapy for 3 months in 60 children with autism resulted in improvements, as analyzed by the autism spectrum disorder score (Karahmadi et al. 2018). Memantine and risperidone were compared in 30 autistic children in an 8-week randomized trial (Nikvarz et al. 2016). Both treatments resulted in similar, significant reductions in the scores of four subscales of the Autism Behavior Checklist, as well as the 10-item and the total score of the Childhood Autism Rating Scale. More risperidone than memantine recipients showed "very much improvement" in the CGI-I scale. Hardan et al. assessed the efficacy and safety of memantine in children with autism spectrum disorder in a multicenter study with variable design. In general, there was no significant effect of treatment (Hardan et al. 2019). A pilot, 24-week, randomized, controlled study assessed the neurocognitive effects of memantine in 23 children with autism spectrum disorder (Soorya et al. 2021). Memantine was well tolerated; however, dropout rates were high. Although, improvements in apraxia and expressive language were not seen, memantine was associated with improvements in verbal recognition memory (p=0.03) (Soorya et al. 2021).

Rating : 2 (relevant doses used, mixed results)

Bipolar disorder

Clinical The efficacy of memantine was compared with placebo as adjunct treatment to sodium valproate in the acute treatment of 70 elderly patients (aged >60 years) with bipolar disorder hospitalized for mania (Omranifard et al. 2018). By 4 and 8 weeks of treatment, the mania severity score was reduced significantly in both groups (p < 0.001), although the reduction was greater with memantine than placebo (p = 0.038). Cognition and quality-of-life scores were improved from baseline but did not differ significantly between groups (Omranifard et al. 2018). A preliminary placebo-controlled study showed the favorable tolerability and efficacy of memantine 5 to 20 mg/day over 16 weeks as an adjuvant treatment in reducing obsessive-compulsive symptoms in 38 patients in the manic phase of bipolar disorder type I (Sahraian et al. 2017).

Rating: 2 (encouraging results)

Cardiovascular

Preclinical In an atherosclerotic model, human umbilical vascular endothelial cells were stimulated with low-density lipoprotein (LDL) (Hao et al. 2021). LDL stimulation reduced the viability of vascular endothelial cells and resulted in excessive inflammation, oxidative stress, and apoptosis, all of which were attenuated by memantine (5, 10 and 50 mM). Moreover, memantine activated the BDNF/TrkB pathway, which was blocked by K252a (Hao et al. 2021). Of note, concentrations used were 5000 times higher than expected plasma levels after therapeutic treatment. In an ischemia-reperfusion injury model of the peripheral vasculature (human umbilical cord endothelial cells), memantine (5 or 10 µM) attenuated the expression of IL-6 and IL-8 and prevented mitochondrial dysfunction and oxidative stress (Lv et al. 2020). Mitochondrial membrane potential was normalized, and production of reactive oxygen species was reduced. Memantine also increased cell viability and prevented reduction in microtubule formation induced by oxygen glucose deprivation. The PI3K/protein kinase B pathway may be crucial for the effect on angiogenesis (Lv et al. 2020). It has also been suggested that memantine prevents and terminates atrial fibrillation in a variety of models by inhibiting glutamate channels in atrial cardiomyocytes (Xie et al. 2022). However, a significant effect was only obtained with concentrations above the therapeutic range (10 µM or more). The cardioprotective effects of memantine were investigated in the isolated perfused rat heart, a model of myocardial infarction (Jannesar et al. 2020). Memantine (in vivo 5 or 20 mg/kg) increased heart rate, left ventricular systolic pressure and left ventricular maximal rate of pressure, and decreased cardiac arrhythmia, malondialdehyde levels, and infarct size (Jannesar et al. 2020). In an in vivo model induced by isoproterenol, malondialdehyde levels and myeloperoxidase activity were decreased by memantine (5, 10 and 20 mg/kg) as well as expression of cardiac TNFα (at 20 mg/kg). Moreover, memantine (10 mg/kg) decreased cardiac fibrosis and hypertrophy (Jannesar et al. 2020).

Rating: 1 (only one of 4 papers showed an effect using memantine dosing within the therapeutic range)

Cerebellar ataxia

Preclinical Reactive transformation of cerebellar Bergmann glia was produced using an optogenetic approach in vivo in mice (Shuvaev et al. 2021). Memantine (90 mg/kg in drinking water) for 4 days prevented thickening of the processes and attenuated length reduction produced by light. Similarly, a decrease of glutamate transporter EAAT1 in Bergmann glia was attenuated. Importantly, memantine reduced the loss of Purkinje cells (Shuvaev et al. 2021).

Rating: 0 (very high dose used, 1 paper only)

Chemotherapy

Preclinical In a paclitaxel model in B6 mice (Sung et al. 2021), pretreatment with memantine (50 mg/kg) improved neurogenesis (increased BDNF) and restored spatial learning deficit but failed to inhibit expression of peripheral and central TNF-α and IL-1β, and depressive-like symptoms (Sung et al. 2021). Co-treatment with memantine 10 mg/kg for 6 days with paclitaxel improved spatial learning deficit, restored hippocampal neurogenesis (increased BDNF), and modulated inflammation (Sung et al. 2021). Memantine (5, 10 mg/kg per os) was given for 30 days in BALB/c mice as pretreatment before cisplatin (Salih and Al-Baggou 2020). At 10 mg/kg, improvement was observed with memantine in the open field activity, negative geotaxis, hole-board test, swimming test, and body weight, while at 5 mg/kg only partial improvement was observed. Memantine treatment normalized nAChRs score (Salih and Al-Baggou 2020).

Rating: 1 (2 papers only, very high doses)

Chronic obstructive pulmonary disease

Preclinical The effects of memantine on lung inflammation were studied in RAW264.7 cells and in a chronic obstructive pulmonary disease mouse model developed using cigarette smoke combined with lipopolysaccharide (LPS) as an insult (Cheng et al. 2019). Memantine attenuated increase in cytokine (TNF-a, IL-6, and IFN-g) levels and elevated glutamate release, NR-1 and xCT protein levels, Ca²⁺ influx, and activation of the ERK1/2 pathway, both in in vitro and in vivo (5 mg/kg) (Cheng et al. 2019).

Rating: 2 (one paper only with a practical dose and multiple parameters showed a positive effect)

Compulsive buying disorder

Clinical In nine patients with compulsive buying behavior, memantine (10 mg/day titrated to 30 mg/day) was assessed in a 10-week open-label study and analyzed using the Yale-Brown Obsessive Compulsive Scale–Shopping Version (Grant et al. 2012). It was concluded that memantine reduced impulsive buying and improved cognitive tasks of impulsivity.

Rating: 1 (only 1 study with a low number of patients and open-label design)

Coughing

Clinical Memantine 20 mg was tested in capsaicin-evoked coughing in healthy subjects or subjects with acute viral upper respiratory tract infection in a randomized, double-blind, crossover study (Dicpinigaitis et al. 2015). In 14 healthy subjects, memantine significantly inhibited cough sensitivity following capsaicin cough challenge (p=0.034); in 14 subjects with viral infection, only a trend for a positive effect was seen.

Rating: 1 (no effect in disease condition)

COVID-19

Preclinical Although memantine is not believed to have value as a direct treatment for COVID-19 infections, some of the papers tackling this topic have been included in this review for completeness. In VeroE6 cells, memantine inhibited COVID-19 infections with an EC₅₀ of 80 µM. Similar results were observed in human hepatoma Huh7.5 and lung carcinoma A549-hACE2 cells (Zhou et al. 2021). Interestingly, it was proposed that memantine may decrease angiotensin-converting enzyme 2 receptor expression in respiratory epithelium and hence prevent invasion of pulmonary epithelial cells by α7-nAChR antagonism (Hasanagic and Serdarevic 2020).

Rating: 0 (very high concentration used, very poor evidence status)

Clinical In a subanalysis of data for 279 patients with dementia from a retrospective review of memantine use and COVID-19-associated mortality from a national South Korean database, no statistically significant relationship between COVID‐19‐associated mortality and memantine treatment was shown after adjustment for clinically relevant potential confounders (age, sex, and/or comorbidities) (Park and Kwon 2021).

Rating: 0 (no effect)

Dementia with Lewy bodies and Parkinson’s disease with dementia

Clinical The effect of memantine 20 mg/day on survival in 75 patients with dementia with Lewy bodies and Parkinson's disease with dementia was investigated in a prospective 24-week, randomized, placebo-controlled trial (Stubendorff et al. 2014). Treatment response was assessed by CGI-C at 24 weeks from baseline and patients were classified as responders (CGI-C 1–3) or non-responders (CGI-C 4–7). After open-label treatment, survival was recorded at 36 months. Compared with placebo, memantine-treated patients had a longer length of survival (p = 0.045) and the memantine responders had higher rates of survival compared with the non-responders (p = 0.010); similar results were not seen with placebo. The investigators concluded that early treatment with memantine and a positive clinical response predicted longer survival, suggesting a possible disease-modifying effect of memantine in patients with dementia with Lewy bodies and Parkinson's disease with dementia (Stubendorff et al. 2014).

Rating: 2 (one study but strong design and encouraging results)

Depression and bipolar II disorders

Preclinical The α7-nAChRs have been suggested to be involved in neuroinflammation and microglial activation associated with depression (Alzarea et al. 2022). Memantine (1, 3 mg/kg IP) prevented cognitive deficits and depressive-like behaviors induced by LPS in mice (Alzarea et al. 2022). Memantine’s potential effects on mitochondrial function and depressive-like symptoms were studied in the chronic unpredictable stress (28 days) model of depression (Mishra et al. 2021b). Depressive-like symptoms in the forced swim test and a decline in the spatial learning in the Morris water maze test were prevented with memantine (10 mg/kg/day) (Mishra et al. 2021b). Memantine also attenuated chronic unpredictable stress-induced increase in plasma cortisol, synaptosomal Ca²⁺ ion levels, increase in oxidative stress, decrease in mitochondrial electron transport chain enzymes activity, and mitochondrial membrane potential. Additionally, memantine restored expression of cell survival genes, CREB, and BDNF (Mishra et al. 2021b).

In an olfactory bulbectomized mouse model of depression, memantine (1–3 mg/kg; per os for 14 days) enhanced neurogenesis in the hippocampus (dentate gyrus BrdU-positive neurons) and improved depressive-like behaviors (tail-suspension and forced-swim tests) (Moriguchi et al. 2020). Further analysis showed that memantine increased phosphorylation of CaMKIV and protein kinase B. Interestingly, the improvements in depressive-like behavior and increase in BrdU-positive neurons were not observed in Kir6.1 heterozygous (+/−) mice. Furthermore, the increase in CaMKIV (Thr-196) and protein kinase B (Ser-473) phosphorylation and BDNF protein expression levels were not observed in Kir6.1 +/− mice, suggesting the role of these channels in the MoA of memantine (Moriguchi et al. 2020).

In another olfactory bulbectomized mouse model of depression, depressive-like behaviors were observed 42 days after surgery (Takahashi et al. 2018). At the molecular level, a decrease of monoamines, reduced cell proliferation, and lower levels of tyrosine hydroxylase, pKA, p-DARPP-32, p-ERK1/2, p-CREB, BDNF, NeuN, and Bcl-2 were seen. There was also an increase in the number of activated microglia and astrocytes, and levels of Iba1, glial fibrillary acidic protein, p-IkB-a, p-NF-kB p65, TNF-a, IL-6, Bax, and cleaved caspase-3 in the hippocampus. These effects were reversed by memantine (10, 20 mg/kg) with the exception of Bax. These changes suggest that the antidepressant-like action of memantine may be associated with increased hippocampal cell proliferation and neuroprotection (Takahashi et al. 2018). In a repeated unpredictable stress (10 days) rat model of depression, memantine (20 mg/kg) reversed depression-like behavior and learning impairment, and increased BDNF and TrkB mRNA in the prefrontal cortex and hippocampus (Amidfar et al. 2018). The authors suggested that upregulated BDNF and TrkB may be important for the antidepressant-like activity of memantine in this model (Amidfar et al. 2018). In a study designed to compare ketamine and memantine, despite that only ketamine produces rapid antidepressive responses in humans, ketamine but not memantine (10 mg/kg) produced rapid responses in the forced swim test in rats (Zhang et al. 2017). Memantine and ketamine similarly increased GluA1 S845 phosphorylation and CA1 synaptic transmission; however, only ketamine increased the expression of GluA1. Furthermore, in hippocampal slices, only ketamine enhanced mTOR phosphorylation similar to the time course of increase in GluA1. These data suggest that regulation of mTOR phosphorylation differentiates the antidepressant responses of memantine and ketamine (Zhang et al. 2017).

Rating: 3 (several papers showed positive effects at low doses)

Clinical In a proof-of-concept, randomized, double-blind trial in patients in the depressive phase of bipolar II disorder, memantine (5 to 20 mg/day) as add on to lamotrigine showed no significant effect on the 17-item Hamilton Depression Rating Scale (HAM-D) at 8-weeks compared with placebo; however, compared with baseline, there was a trend for improvement during the first 4 weeks (p = 0.007) (Anand et al. 2012). Lee and colleagues tested the 12-week effect of low-dose memantine (5 mg/day) as adjunct to valproate on blood cytokines and clinical symptoms in a double-blind, randomized study in 214 patients with bipolar depression (Lee et al. 2014a). Memantine treatment decreased TNF-a levels; however, clinical symptoms were not improved. The same group also tested the role of a genotype variant in bipolar disorder II on therapeutic response to low-dose memantine as adjunct to valproate in a randomized, double-blind, 12-week study (Lee et al. 2014b). A stronger effect of memantine on the HAM-D scores was shown in the BDNF 66Met genotype as compared to BDNF Val66 (Lee et al. 2014b). In a further 12-week, randomized, controlled study in patients with bipolar spectrum disorder, Lee and colleagues investigated whether a combination of agents with anti-inflammatory and neurotrophic effects was more efficacious than a mood stabilizer alone in improving clinical symptoms, plasma BDNF, cytokine levels, and metabolic profiles (Lee et al. 2020). The combination of low-dose memantine plus dextromethorphan as adjunct to valproate decreased depressive symptoms and increased plasma BDNF levels as compared to placebo. Neither memantine nor dextromethorphan alone were effective (Lee et al. 2020). In 325 patients with bipolar II disorder, memantine (5 mg/day) was evaluated as add-on in two 12-week, randomized, double-blind, placebo-controlled studies that assessed inflammation and cognition (Lu et al. 2021). Overall, memantine did not affect cognitive functions in all patients but decreased plasma TNF-α levels. In a subset of middle-aged to elderly patients, in addition to decreased plasma TNF-α levels with memantine, there was a significant time and group interaction effect on omission T-scores, hit reaction time T-scores, and hit reaction time standard error T-scores on continuous performance tests (Lu et al. 2021).

The severity of depression and quality of life were evaluated following add-on memantine 20 mg/day during citalopram therapy in 57 elderly patients with depression in an 8-week, placebo-controlled trial (Omranifard et al. 2014). Change from baseline and between-group differences in mean scores of the Geriatric Depression Scale, HAM-D and WHO-QOL-BREF scales were not significant with memantine use (p >0.05), indicating a lack of antidepressant effect of add-on memantine in elderly patients with depression receiving citalopram.

The effects of 3-months‘ of memantine (5 mg/day up-titrated to 20 mg/day) as add-on to escitalopram on depression scores, gray matter volume and cortical thickness were evaluated in a double-blind, randomized, placebo-controlled study in 41 elderly patients with depression (Krause-Sorio et al. 2020a). In memantine-treated patients, remission rates were higher than that with placebo (62% vs 43%); however, there were no difference in HAM-D scoring. Structural neuroplastic changes were greater after memantine in the left middle and inferior temporal lobe, right medial, and lateral orbito-frontal cortex. In a follow-up study, investigators found no differences in remission or HAM-D scoring after 6 months‘ treatment in 38 patients (Krause-Sorio et al. 2020b). In general, increase in fractional anisotropy corresponded to greater improvements in depressive symptoms in memantine- (+escitalopram) treated patients, but not placebo+citalopram-treated patients.

In 95 older patients with major depression (62 of whom completed the study), memantine (10 to 20 mg/day) was tested as add-on to citalopram in a randomized, double-blind, placebo-controlled 6-month study (Lavretsky et al. 2020). There was no effect of memantine for change in depression, as assessed by the HAM-D, but stronger improvements in delayed recall and executive function were observed. In a double-blind, placebo-controlled, 12-week study in 35 elderly patients following a disabling medical event, memantine (10 mg/day followed by 20 mg/day) showed no effect on depression and apathy (Lenze et al. 2012).

A recent systemic review and meta-analysis of the efficacy and safety of memantine for depressive symptoms in patients with major mental disorders revealed that memantine decreases depressive symptoms; however, the effect size was modest and memantine was well-tolerated (Hsu et al. 2022).

Rating: 3 (relevant therapeutic range, double-blind design, antidepressant effect in several studies, supportive meta-analysis)

Depression: electroconvulsive therapy-induced cognitive deficit

Clinical In 40 patients with a major depressive disorder, memantine (5 mg/day) or placebo was given during four sessions of electroconvulsive therapy (ECT) to assess potential use in protection against the adverse cognitive effects of ECT (Abbasinazari et al. 2015). MMSE was used to evaluate memory dysfunction/cognition. Analysis revealed that, compared with placebo, MMSE total scores (p = 0.02) and item 3 MMSE (related to recent memory, p < 0.001) were significantly improved with memantine.

Rating: 1 (one study only but encouraging results)

Diabetic neuropathy/diabetes

Preclinical In a diabetic retinopathy model induced by alloxan in mice, oral memantine (5, 10 mg/kg) for 28 days attenuated vacuolization and restored normal retinal cell layers (ElSayed et al. 2023). At the molecular level, memantine reduced TXNIP, NLRP3, IL-1β and MDA concentrations, all supportive for neuroprotection (ElSayed et al. 2023). The anti-allodynic effects of memantine were studied in a alloxan-induced mouse model of diabetic neuropathy; after 5 weeks, memantine (10 mg/kg) attenuated pain, inhibited NMDAR1 activation, and suppressed glutamate and pro-inflammatory cytokine release in the spinal cord (Alomar et al. 2021). An examination of whether antenatal blockade of NMDA receptors by memantine could reduce the risk of diabetes induced by a high-fat diet was conducted in pregnant rats exposed to hypoxic conditions for 8 h/day (Huang et al. 2017). Memantine injections (5 mg/kg IP) were started before hypoxia exposure from embryonic day 14 (Huang et al. 2017). The offspring were than fed a high-fat diet from 4 to 12 weeks, which produced a high rate of diabetes in adulthood; however, memantine decreased this rate. This effect of memantine was accompanied by rescue of glucose tolerance, increased insulin release, improvement of insulin sensitivity, and increase of expression of genes related to β-cell function in the pancreas (Huang et al. 2017).

Rating: 2 (relevant doses and 3 papers available)

Clinical In a randomized clinical trial in 143 diabetic patients with diabetic neuropathy, compared with control, memantine (5 mg BID followed by 10 mg BID from the week 2) as add-on to gabapentin for 8 weeks significantly lowered the Douleur Neuropathique 4 questionnaire mean score and the number of patients with diabetic neuropathy, indicating significant potential in the treatment of this condition (Jafarzadeh et al. 2023).

Rating: 1 (one study only but encouraging results)

Down syndrome

Clinical In a randomized, double-blind, placebo-controlled trial in 185 patients with Down syndrome, memantine (20 mg/day) was tested for cognitive and adaptive effects (Costa et al. 2022). For the primary outcome measure, no significant effect compared with placebo was seen (p=0·61). Notably, the concentration of memantine in plasma was lower than typically seen in patients with AD. In general, memantine was well tolerated. Memantine (5 mg/day up-titrated to 10 mg/day) was further tested on cognition and function in 88 patients with Down syndrome in a 52-week, randomized, double-blind, placebo-controlled trial; however, there was no effect of memantine on any of the scores applied (Hanney et al. 2012).

Rating: 0 (no effect observed)

Dyskinesia after L-DOPA in Parkinson’s disease

Clinical The partial glutamate antagonist, amantadine, is currently used in clinical practice to reduce dyskinesia developing in levodopa-treated patients with Parkinson’s disease. The results of a 3-week, randomized, double-blind, placebo-controlled, crossover trial suggest that memantine 20 mg/day may have antidyskinetic effects (Wictorin and Widner 2015). Although the primary outcome measure, change in observed dyskinesia ratings, did not reach significance, seven of 15 patients showed reduced (32%) L-dopa-induced dyskinesias, in three patients there was a 33% increase, and in five patients there was no change. Data from self-administered diaries (a secondary outcome measure) showed a significant reduction (35%) in the percentage of time per day spent with ‘on-with dyskinesias’, from 25% with placebo to 16% with memantine (p = 0.005) (Wictorin and Widner 2015).

Rating: 3 (one paper, only the change in the measure of time with dyskinesia was significant)

Electroconvulsive therapy-induced cognitive impairment

Clinical

The efficacy and safety of memantine 5 mg/day were compared with those of melatonin for the relief of cognitive disorders in patients with major depressive disorder undergoing electroconvulsive therapy (ECT) (Sarraf et al. 2020). Patients (n = 40) undergoing ECT for the treatment of major depressive disorder randomly received melatonin or memantine (5 mg/day). Compared with baseline, the memantine group scored significantly higher at the end of the ECT sessions on the MMSE or item 3 of the MMSE, p = 0.04 and p = 0.03, respectively, suggesting the efficacy of memantine in alleviating cognitive disorders induced by ECT.

Rating: 3 (encouraging results, but one paper only and no placebo group)

Epilepsy

Preclinical Young 15-day-old rats were treated with memantine (20 mg/kg IP) at different times relative to pilocarpine induction of status epilepticus (Zenki et al. 2018). Memantine increased latency of status epilepticus only in animals treated 3 h before or together with pilocarpine. Memantine also significantly reduced neurodegeneration in CA1 hippocampal region when administered 3 h prior to, together, and 15 min after pilocarpine. Post-treatment, memantine also attenuated neurodegeneration in the rat amygdala and thalamus (Zenki et al. 2018). In the lithium–pilocarpine model of epilepsy in rats, memantine (10 mg/kg) attenuated cognitive impairment, seen as prevention of impairments in exploratory behavior and spatial memory, and prevented a decline of extinction of orienting behavior (Kalemenev et al. 2016).

Rating: 1 (very high doses used)

Clinical Among 27 children with epileptic encephalopathy who received memantine for 6 weeks, followed by a 2-week washout period then crossover to placebo for 6 weeks, nine (33%) were classified as responders to memantine compared with two (7%) with placebo (p < 0.02) (Schiller et al. 2022). A responder was defined by two or more of the following measures: >50% seizure frequency reduction, electroencephalography improvement, caregiver clinical impression of improvement or clear neuropsychological testing improvement. This study indicates the potential of memantine to improve seizure control and cognitive function (Schiller et al. 2022). Memantine, titrated to 10 mg/day, was assessed for effects on memory in 17 patients with focal-onset epilepsy in a 13-week, randomized, double-blind, placebo-controlled trial that was followed by 13 weeks of open-label treatment (Leeman-Markowski et al. 2018). No significant between-group differences were observed in primary measures such as selective reminding test, continuous long-term retrieval score, and 7/24 Spatial Recall Test. Additionally, there was no effect on depression, sleepiness, side effects, or seizure frequency (Leeman-Markowski et al. 2018). Marimuthu and colleagues investigated the effect of memantine (5 to 10 mg/day) on cognition in 59 epileptic patients on antiepileptics in a 16-week, double-blind, placebo-controlled study (Marimuthu et al. 2016). Memantine improved the mean total MMSE score and the Weshler's Memory Scale score; self-rated quality of life and memory were also improved. In 70 patients with temporal lobe epilepsy, cognitive function, as assessed by the Montreal Cognitive Assessment test, was significantly improved from baseline with 16-weeks of memantine treatment 5 to 10 mg/day (p < 0.001), but not donepezil, in a parallel-group study (Oustad et al. 2020).

Rating: 1 (only effect on cognition was observed in one study)

Fibromyalgia

Clinical Elevated levels of glutamate have been found in several brain regions in patients with fibromyalgia, suggesting the utility of glutamate-blocking drugs as treatment (Olivan-Blazquez et al. 2014). In a double-blind, randomized, placebo-controlled study with a 6-month follow-up, Fayed et al. studied magnetic resonance spectroscopy changes in 25 fibromyalgia patients (Fayed et al. 2014). Memantine (5 to 20 mg/day) significantly increased the glutamate, glutamate/creatine ratio, combined glutamate + glutamine and total N-acetyl-aspartate levels in the posterior cingulate cortex, and increased creatine and choline in the right posterior insula. Interestingly, a correlation was seen between choline and the Fibromyalgia Impact Questionnaire. These changes were interpreted as indicating a positive effect of memantine in fibromyalgia (Fayed et al. 2014). Olivan-Blázquez and colleagues compared memantine (20 mg/day) with placebo in 63 fibromyalgia patients in a 6-month, double-blind, randomized trial (Olivan-Blazquez et al. 2014). Memantine decreased ratings on a pain visual analogue scale (at 6 months) and pain measured with a sphygmomanometer. Several secondary outcomes e.g., depression were also improved. The absolute risk reduction with memantine was 16.1% (95% CI, 2.0% to 32.6%), and the number needed to treat, 6.2 (95% CI, 3 to 47).

Rating: 2 (only two studies but encouraging)

Fragile X

Clinical In the first randomized, placebo-controlled trial of memantine 5 mg/day to 10 mg BID in patients with fragile X-associated tremor/ataxia syndrome (n = 85), over 1 year, no significant effects on clinical and neuropsychological outcome measures, including intention tremor severity, executive function, verbal learning/memory, and working memory were shown (Seritan et al. 2014). A substudy of the trial used the N400 repetition effect, a glutamate-related electrophysiological marker of semantic priming, to assess the effects of chronic (1-year) memantine (5 mg/day titrated to 10 mg BID) treatment on verbal memory (Yang et al. 2014). Cued memory retrieval, potentially via regulation of the FMR1 mRNA associated glutamatergic signaling abnormalities, was shown with memantine, but with no benefits on executive dysfunction (Yang et al. 2014). The effects of memantine on attention and working memory were also examined in patients with fragile X-associated tremor/ataxia syndrome (Yang et al. 2016). In the memantine, but not the placebo group, attention/working memory performance were significantly improved after 1 year. The event-related potential P2 amplitude elicited by non-targets was significantly enhanced, indicating improvement in attentional processes. The increase in P2 amplitude was positively correlated with improvement on the behavioral measure of attention/working memory (Yang et al. 2016).

Rating: 2 (mixed results)

Frontotemporal lobar degeneration

Clinical Memantine (20 mg/day for 26 weeks) was investigated in 81 patients with frontotemporal lobar degeneration in a multicenter, randomized, double-blind, placebo-controlled trial (Boxer et al. 2013). Memantine did not affect symptoms measured on either the NPI Questionnaire (NPI-Q) or the CGI-C. In a 6‑month, open‑label, self‑controlled clinical trial, memantine (20 mg/day) was tested on neuropsychiatric symptoms in 42 patients with behavioral variant frontotemporal dementia (Li et al. 2016). Memantine treatment had no effect on NPI-Q scores, with the exception of a positive effect on the agitation subdomain. In general, as shown in neuropsychiatric scores, patients with moderate-to-severe status responded more favorably to memantine (cognitive and functional benefits) than patients with mild symptoms (Li et al. 2016).

Rating: 1 (mixed results)

Gastric ulcers

Preclinical Following indomethacin-induced peptic ulcer in rats, memantine monotherapy (10 mg/kg) or in combination with glibenclamide or a NO synthase inhibitor, reduced ulcer index and MDA, increased SOD and total nitrites, and reduced expression of both TNF-α and NF-κB (Rofaeil and Gaber 2019). The gastroprotective effect of memantine was attenuated by beclamide. The authors indicated that the MoA of memantine may involve potassium channels, antioxidative stress and anti-inflammatory actions (Rofaeil and Gaber 2019). In a mouse model of ulcerative colitis induced by trinitrobenzene sulfonic acid, memantine (12.5, 25 and 50 mg/kg IP) given 24 h before insult and daily for 4 days attenuated body weight loss, colon weight, plasma levels of IL-1β, IL-6, and colon TNF-α and myeloperoxidase (MPO), and macroscopic and microscopic signs of colitis (Motaghi et al. 2016).

Rating: 1 (high doses used)

Glaucoma

Preclinical In a mouse optic nerve crush model, 2-µL intravitreal injections of memantine (1 mg/mL) were administered in experimental eyes (Maciulaitiene et al. 2017). Analysis carried out one week later showed attenuation of retinal astrocyte count decrease compared with control eyes, suggesting a glioprotective effect (Maciulaitiene et al. 2017). Memantine (10 mg/kg for 6 weeks) administered in a rat model of glaucoma induced by 3 weeks’ of sodium hyaluronate injection into the anterior chamber was associated with attenuating ultrastructural changes in the retina and prevention of cell death (Celiker et al. 2016).

Rating: 1 (high dose)

HIV

Preclinical In a model of combined toxicity of cocaine in HIV-1 Tg rats, memantine (10 mg/kg) attenuated cocaine effects on neuronal and hemodynamic responses and attenuated enhancement of neuronal Ca²⁺ influx, hypoxemia, and ischemia (Du et al. 2022).

Rating: 1 (high dose, one paper only)

Huntington’s disease

Clinical Although memantine was speculated to provide a clinical benefit in Huntington’s disease, Ogilvie et al. showed no effect of memantine (up to 30 mg/day) on cognitive decline over 5 years in Huntington’s patients in the enroll-HD study (Ogilvie and Schultz 2023).

Rating: 0 (no effect)

Hydrocephalus

Preclinical In an investigation in juvenile rats with kaolin-induced hydrocephalus, memantine 10 or 30 mg/kg/day inhibited ventricular enlargement at both doses. At the behavioral level, memantine improved some parameters but did not reduce brain tissue (Di Curzio et al. 2018).

Rating: 1 (high dose, one paper only)

Inborn errors of metabolism

Preclinical Inborn errors of metabolism which lead to hyperactivity and intellectual disability were studied in Xpnpep1 mutant mice (knockout of aminopeptidase P1) (Bae et al. 2022). Memantine (2 μM) applied ex vivo in hippocampal slices during electrophysiological recordings normalized exaggerated long-term potentiation but had no effect on theta burst stimulation induced long-term potentiation of exaggerated NMDAR activity. In vivo, acute memantine treatment (10 mg/kg IP) reversed hyperactivity of knockout mice but failed to improve learning. Chronic memantine treatment (10 mg/kg, IP, BID) decreased hippocampal neurodegeneration, hyperactivity, and improved learning (Bae et al. 2022).

Rating: 1 (One paper only, high dose)

Kleptomania

Clinical In 12 kleptomania subjects in an 8-week, open-label study, as assessed using validated measures and select neurocognitive tests, memantine (10 to 30 mg/day) decreased all kleptomania severity scores and 11 subjects met the responder criteria, accompanied by improvements in mood, anxiety, and disability scores (Grant et al. 2013).

Rating: 1 (one study only but encouraging data)

Liver injury

Preclinical Acute lung injury in septic mice was attenuated by memantine (1, 5, 10 mg/kg), likely involving normalization of Ca²⁺ influx, inhibition of macrophage Nlrp3 inflammasome activation and pyroptosis (Ding et al. 2021).

Rating: 2 (one paper only but low doses and several measures affected)

Macular degeneration

Preclinical An in vitro model of macular degeneration was used based on 2-ethylpyridine-induced oxidative stress and mitochondrial dysfunction in human RPE cells (Bardak et al. 2018). Memantine, similarly to melatonin, decreased apoptosis, caspase-3, 9 activities and pro-caspase and poly(ADP-ribose) polymerase expression; there was some indication of a stronger effect with the combination of memantine and melatonin (Bardak et al. 2018).

Rating: 1 (only in vitro study)

Methamphetamine toxicity

Preclinical In a model of repeated exposure to methamphetamine neurotoxicity, memantine (5 or 10 mg/kg per os) given in combination with methamphetamine improved cognitive function, decreased Bcl-2 expression and increased caspase-3 expression in the prefrontal cortex (Long et al. 2017).

Rating: 2 (one paper only but low doses were active)

3,4-Methylenedioxymethamphetamine memory impairment

Clinical In 15 subjects affected by 3,4-methylenedioxymethamphetamine (MDMA) impairment, the efficacy of memantine (20 mg/day) was investigated in a double-blind, placebo-controlled, intra-subject study. Memory function was assessed using a battery of tasks including the Visual Verbal Learning Task. Pre-treatment with memantine did not affect any deficits (memory impairment and mood) induced by MDMA (de Sousa Fernandes Perna et al. 2014)

Rating: 0 ( (negative results)

Migraine

Clinical In a randomized, placebo-controlled, double-blind study in 60 patients with 3–12 migraine headaches in the previous 6 months, memantine (10 mg/day) was evaluated over 24 weeks (Shanmugam et al. 2019). By 24 weeks, compared with placebo, memantine significantly improved migraine frequency/month, rescue medication use, and the 50% responder rate.

Noruzzadeh and colleagues showed in a 12-week, randomized, double-blind, placebo-controlled trial in 60 migraine patients without aura that memantine (10 mg/day) reduced the monthly attack frequency and afforded a greater reduction in the number of work absenteeism days, severity, and disability. The authors therefore suggested that memantine might be an option for prophylaxis in patients with migraine without aura (Noruzzadeh et al. 2016).

Rating: 2 (two studies with encouraging results)

Mild cognitive impairment

Clinical Mild cognitive impairment is considered a transitional phase between normal cognitive aging and dementia (Peters et al. 2012). In a double-blind, placebo-controlled study, 232 subjects with mild cognitive impairment, presenting with an amnestic syndrome not necessarily a result of presumed prodromal AD, randomly received galantamine plus memantine, galantamine alone, or placebo for 2–52 weeks; the variation in treatment duration was a result of the trial being halted (Peters et al. 2012). Cognitive changes did not differ between treatment groups; however, for the subgroup of patients with amnestic MCI with presumed AD etiology, a significant improvement in ADAS-cog scores compared with placebo was shown; the cognitive improvement was greater with combination treatment than with galantamine alone. Cognitive decline was seen following galantamine discontinuation, either as part of the combination regimen or as monotherapy, while discontinuation of memantine did not result in changes in cognitive status.

Rating: 1 (one paper only, well designed but an effect was seen only in a subpopulation with AD)

Multiple sclerosis

Preclinical In a rat model of experimental autoimmune encephalomyelitis resembling mechanisms of multiple sclerosis (MS), memantine 60 mg/kg/day improved locomotor activity, decreased paralysis of tail and hind limbs, and normalized oxidative stress parameters such as MDA, -SH groups, and SODs (Dabrowska-Bouta et al. 2021). Memantine (60 mg/kg), administered to young and aged rats from day 7 after immunization improved symptoms, reactivation, and apoptosis of CD4+ T cells in aged EAE rats (Bufan et al. 2024). Additionally, memantine increased mRNA expression in brain tissue regulated by Nrf2 and Nrf2-enzymes.

Rating: 1 (two papers only, very high dose)

Clinical In the absence of an approved pharmacotherapy for MS-related fatigue, Falsafi and colleagues studied memantine (20 mg/day) in a 3-month, pilot, randomized, double-blind, placebo-controlled trial in 64 patients with MS-related fatigue (Falsafi et al. 2020). The analysis showed that memantine failed to modify MS-related fatigue. Similarly, long-term (52-week) administration of memantine (20 mg/day) for the symptomatic treatment of cognitive disorders in patients with relapsing-remitting MS was shown not to be effective in a placebo-controlled study (Peyro Saint Paul et al. 2016).

Rating: 0 (no effect)

Neuromyelitis optica spectrum disorders

Preclinical In a neuromyelitis optica spectrum disorders model in mice, which is a model of autoimmune inflammatory demyelinating diseases, memantine (60 mg/kg per os) ameliorated the motor impairments induced by AQP4-IgG, reduced AQP4 and astrocyte loss, attenuated demyelination, and decreased axonal loss in the spinal cord (Yick et al. 2020). Additionally, decreased apoptosis and neuroinflammation were observed, accompanied by a decrease in microglia activation and neutrophil infiltration, reduction in of proinflammatory cytokines (IL-1β, IL-6, TNF-α) and increase in BDNF and GDNF (Yick et al. 2020).

Rating: 1 (one paper only and high dose)

Neuropathic pain

Preclinical In a rat model of trigeminal pain based on electrical stimulation of dura mater adjacent to the middle meningeal artery neurons, responses to stimulation were recorded extracellularly during microiontophoretic application of memantine. Memantine applied to the trigeminocervical complex inhibited mechanically and electrically stimulated responses. In contrast, memantine (10 mg/kg IV) was not effective (Hoffmann et al. 2019). In a spared nerve injury pain model, low-dose memantine (10 nmol, intrathecal) prevented the induction of dynamic allodynia but not punctate allodynia; at a higher dose (30 nM), memantine prevented the induction of both (Chen et al. 2019). Memantine (10 nmol) inhibited injury-induced overactivation of microglia in spinal dorsal horn. In a complete Freund’s adjuvant model, memantine was not effective (Chen et al. 2019). In a sciatic nerve crush injury model in rats, memantine (5, 10 mg/kg IP) for 7 postoperative days had no effect on the regenerative process rate and functional recovery in the sciatic functional index (Ghayour et al. 2016).

Rating: 1 (mixed results at relevant doses)

Clinical The efficacy of memantine (5 to 10 mg/day up-titrated to 40 to 60 mg/day for at least 2 months, as tolerated) on complex regional pain syndrome was evaluated in a retrospective study involving 56 patients (Ahmad-Sabry and Shareghi 2015). Almost a quarter of patients (n = 13) showed complete remission from pain and the disappearance of allodynia for at least 9 months after memantine treatment; 18 patients showed partial improvement of the visual analogue scale and allodynia; and eight patients showed no response. In available subjective feeling results, 17 patients reported having a better quality of life while three felt no change (Ahmad-Sabry and Shareghi 2015). In 60 peripheral neuropathic pain patients, memantine (20 mg/day) and dextromethorphan were tested as follow on after ketamine infusion for pain relief in a placebo-controlled, randomized trial for 12 weeks (Martin et al. 2019). After 12 weeks, pain was lower than at study inclusion but did not differ between the treatment groups (placebo, dextromethorphan, memantine). Positive effects on cognitive-affective domains and quality of life were observed with memantine and dextromethorphan. Memantine (5 to 20 mg/day) was administered to patients for the prevention of post-surgical pain following mastectomy (Morel et al. 2016) in a randomized, blinded, placebo-controlled trial in 40 women. The primary endpoint was pain intensity measured on a numerical rating scale 3 months post-mastectomy (range 0–10). Memantine treatment resulted in an improvement of post-mastectomy pain intensity, lowered the need for rescue analgesia, and improved emotional wellbeing. Also, some effect on pain related to cancer chemotherapy was seen (Morel et al. 2016).

Rating: 1 (mixed results)

Nicotine abuse

Preclinical In the self-administration procedure in rats, acute treatment with memantine increased nicotinic self-administration while, with chronic memantine (10 mg/kg) treatment, it was inhibited starting from the week 2 of treatment. Some positive effect was also retained after termination of memantine administration (Levin et al. 2019). The effect of acute and chronic combination of the D1-antagonist SCH-23390 and memantine (10 mg/kg) on nicotine self-administration in rats was investigated (Natarajan et al. 2024). SCH-23390 attenuated nicotine self-administration and memantine given as acute administration increased it, while chronic administration tended to decrease it. In combination, memantine attenuated the effect of chronic SCH-23390.

Rating: 0 (two papers, both at high dose, and one showed negative effect)

Clinical In a study by Das et al., 59 smokers were randomized to one of the following treatments: memantine with reactivation of associative cue-smoking memories, memantine without reactivation of associative cue-smoking memories, or reactivation with placebo, on smoking cessation day in a double-blind manner up to 1 year measuring time to relapse (Das et al. 2015). Memantine failed to affect any smoking outcome measures.

Rating: 0 (no effect)

Nystagmus

Clinical In a cross-over trial, 16 patients with acquired pendular nystagmus were treated with gabapentin or memantine (10 mg, four times per day) (Nerrant et al. 2020). Acquired pendular nystagmus, ocular motor parameters, and distance oscillopsia improved after both treatments; median near oscillopsia was not affected. In a study evaluating the efficacy of memantine (20 mg/day for 6 months) as treatment for nystagmus in six patients with oculopalatal tremor in Guillain-Mollaret triangle lesion, memantine treatment was associated with only a modest and transient improvement in nystagmus. (Martins et al. 2022).

Rating: 0 (no or transient effect)

Obsessive-compulsive disorder

Preclinical In a model of obsessive-compulsive disorder (OCD) based on chronic quinpirole sensitization, neither memantine (1, 5 mg/kg) nor riluzole (1, 5 mg/kg) attenuated the deficit induced by quinpirole, and memantine even increased it (Janikova et al. 2019). In an OCD rodent model, Macedo and colleagues tested the role of gender on the effect of memantine on compulsive behavior (Macedo et al. 2024). Memantine (3, 5, and 10 mg/kg) reduced marble-burying behavior in both male and female mice without affecting activity. Additionally, in the nest-building test, a similar effect was observed; however, the highest memantine dose reduced nest-building behavior. Pretreatment with L-arginine (a NO precursor) decreased the anti-compulsive effect of memantine in male mice only in the marble burying test.

Rating: 1 (mixed results, 2 studies only)

Clinical In a 12-week, placebo-controlled, randomized clinical trial, memantine (10 mg BID) was evaluated as adjunct to sertraline in 70 patients with OCD, diagnosed using the Diagnostic and Statistical Manual of Mental Disorders (DSM– 5) criteria and a Yale-Brown Obsessive Compulsive Scale (Y-BOCS) score of >21 (Askari et al. 2022). Memantine showed no effect on OCD symptom severity, but there was some improvement in executive function, as assessed by the Wisconsin Card Sorting Test (WCST) (Askari et al. 2022). Similarly, Farnia and colleagues showed no additional efficacy of adjunctive gabapentin or memantine to that of standard therapy in 99 patients with OCD, as assessed by Y-BOCS scores (Farnia et al. 2018). An open-label pilot trial in adults with OCD who had recently completed participation in a trial of IV ketamine, investigated whether response to one NMDA receptor antagonist (IV ketamine) might predict response to a second (memantine) (Rodriguez et al. 2016). Clinical effects of treatment with memantine after ketamine were not seen in this 6- to 12-week trial (Rodriguez et al. 2016).

Haghighi and colleagues tested the addition of memantine to serotonin reuptake inhibitors or clomipramine in a 12-week, double-blind, randomized, placebo-controlled study in 40 patients with refractory OCD. There was no effect of memantine in Y-BOCS scores; full response was more likely in the treatment group; and CGI symptom severity (CGI-S) did not differ between groups. In general, there were time-related improvements as well as a significant interaction between time and treatment (Haghighi et al. 2013).

Nevertheless, memantine has been shown to improve OCD symptoms in several studies. In an open-label, 12-week trial of memantine (5 to 10 mg/day) in 12 patients with treatment-resistant OCD (poor responders) Bakhla et al. observed clear improvement in eight patients (minimum 25% reduction in Y-BOCS); no adverse events were reported (Bakhla et al. 2013). Memantine (10 to 20 mg/day) added to fluvoxamine was studied in 42 patients with moderate-to-severe OCD in an 8-week, randomized, double-blind, placebo-controlled study (Ghaleiha et al. 2013b). Analysis using Y-BOCS revealed significant effect for time and treatment interaction in total scale score and obsession subscale scores. By the end of the study, all patients in the memantine group and six patients in the fluoxetine alone group had partial or complete responses, indicating a positive therapeutic effect of memantine in these conditions (Ghaleiha et al. 2013b). In a 12-week, double-blind, placebo-controlled study, memantine was assessed at a dose of 20 mg/day in OCD patients refractory to serotonin reuptake inhibitors (Modarresi et al. 2018). OCD symptoms analyzed using Y-BOCS were reduced in the memantine group but not in the placebo group. Notably, a recent meta-analysis indicated that memantine 20 mg/day is effective as augmentation to first-line anti-OCD treatments in patients with moderate-to-severe OCD (Modarresi et al. 2019).

Rating: 2 (mixed results but mostly positive, positive meta-analysis)

Opioid dependence/abuse

Clinical There is some indication of an interaction between opioid and dopamine systems, hence opioid use may negatively affect cognitive functioning (Chang et al. 2015). Bisaga and colleagues evaluated memantine (40 mg/day) as add-on to naltrexone-based (ER) treatment of opioid dependence in 82 patients in a 12-week, double-blind study (Bisaga et al. 2014). Rates of trial completion were lower with memantine use and in general, memantine was not effective when used in combination with naltrexone ER as a relapse prevention strategy for opioid dependence (Bisaga et al. 2014). Chang et al. (Chang et al. 2015) tested low-dose add-on memantine (5 mg/day) to methadone on cognitive performance in 81 opioid-dependent patients for up to 12 weeks. By study end, cognitive improvement in memantine-treated patients was greater than that seen with placebo control. Memantine (15 or 30 mg/day) was tested in a 13-week, double-blind, placebo-controlled trial as add-on to buprenorphine/naloxone in 80 young adults dependent on opioids (Gonzalez et al. 2015). Treatment with memantine at 30 mg/day decreased the likelihood of relapse after buprenorphine discontinuation and reduced opioid use during the last 2 weeks. Hence, memantine at higher doses may provide benefit in opioid-dependent young adults (Gonzalez et al. 2015). In opioid-dependent patients undergoing methadone maintenance therapy, memantine (5 mg/day) treatment allowed for a significantly lower methadone dose in a 12-week, double-blind, randomized controlled study (Lee et al. 2015). Memantine also decreased plasma TNF-α and increased TGF-β1.

Rating: 1 (mixed results)

Optic neuritis

In a double-blind, randomized, placebo-controlled study in 38 patients with acute optic neuritis, memantine 10 mg/day was administered for 6 weeks with a 3-month follow-up (Motamedi et al. 2022). Memantine reduced retinal nerve fiber layer thinning in three quadrants; visual acuity was not affected.

Rating: 1 (only one study but encouraging data)

Osteoarthritis

Preclinical In osteoarthritis, inflammation and cartilage damage seem to be induced by advanced glycation end products. Memantine (5 or 10 µM) ameliorated advanced glycation end product-induced degradation of collagen II and aggrecan in human SW1353 chondrocytes (Zhao et al. 2017). Memantine also decreased upregulation of the transcriptional factor, interferon response factor-1 (IRF-1) which is induced by advanced glycation end products, and activation of the JAK2/STAT1 pathway (Zhao et al. 2017).

Rating: 1 (high concentrations, one paper in vitro only)

Parkinson’s disease

Preclinical The neuroprotective effects of memantine (10 µM) against alpha-synuclein-induced cell death were seen on nigral dopaminergic neurons, in addition to decreased levels of internalized cytosolic alpha-synuclein; the involvement of clathrin, EEA1, and NR2A was suggested (Lee et al. 2021). In mice, neuroprotective effects of memantine on propagation of extracellular α-synuclein in α-synuclein-inoculated animals was observed with memantine at 20 mg/kg (for 30 days) (Lee et al. 2021). Continuous intrastriatal administration of memantine (with gradual increase to a final dose of 4.48 ng/µL at a flow rate of 1.0 µL/h from day 9), attenuated levodopa-induced dyskinesias and Parkinsonian signs in mice with unilateral SNc lesions (Ogawa et al. 2019).

Rating: 1 (high dose or difficult to judge)

Clinical Dementia is a common development in people with Parkinson’s disease. Patient-reported outcomes were used to assess the effect of memantine (20 mg/day) over 16 weeks of treatment in a double-blind, randomized, controlled study in 25 patients with Parkinson’s disease with dementia (Leroi et al. 2014). Significantly more of the memantine group than the placebo group (64% vs 7%; p = 0.007) achieved improved Goal Attainment Scaling outcomes. Although caregiver burden (assessed with the Zarit Burden Inventory) was significantly lower in those treated with memantine compared with placebo, no significant between-group difference in health-related quality of life (the Parkinson’s Disease Questionnaire-8) was shown (Leroi et al. 2014). Memantine (20 mg/day) for 3 months was assessed in a randomized, double-blind, placebo-controlled study for effects on axial symptoms in 25 patients with Parkinson’s disease (Moreau et al. 2013). Memantine lowered axial motor symptoms and dyskinesia but failed to affect gait as evidenced by UPDRS scores, its axial subscore, axial hypertonia, the axial and overall Dyskinesia Rating Scale and axial strength (Moreau et al. 2013). Attentional dysfunction is a core clinical feature, along with disrupted episodic memory, in dementia with Lewy bodies and Parkinson's disease dementia. A 24-week, randomized, double-blind, placebo-controlled study evaluated the cognitive effects of memantine in dementia with Lewy bodies and Parkinson's disease dementia using automated tests of attention and episodic memory (Wesnes et al. 2015). Among 51 patients (21 DLB and 30 PDD), memantine treatment was associated with statistically significant medium to large effect sized improvements in choice reaction time, and immediate and delayed word recognition.

Rating: 1 (mixed data)

Post-operative cognitive dysfunction

Preclinical In a mouse model of postoperative cognitive dysfunction, 4 weeks’ treatment with memantine (30 mg/kg/day, per os) prior to surgery attenuated impairment of learning in the Morris water maze test, reversed depression in the tail suspension test, and corrected impaired novelty preference in the three-chamber test; increased anxiety was not improved (Almahozi et al. 2019). In aged C57BL/6 J male mice (18 months) treated with sevoflurane anesthesia, sevoflurane + memantine (20 mg/kg) and sevoflurane plus necrostatin-1, memantine and necrostatin-1 decreased neuronal damage, apoptosis and expression of necroptosis-related proteins (Liu et al. 2024). Memantine also decreased Ca²⁺ and expression of NMDAR2A/NMDAR2B subunits of NMDA receptors, while cognitive function was improved.

Rating: 1 (high doses used)

Post-traumatic stress disorder

Preclinical In a mouse post-traumatic stress disorder (PTSD) model (PTSD stress for 10 days), memantine (50 mg/kg) once a week for 4 weeks enhanced forgetting of traumatic memories and improved PTSD (anxiety)-like behavior (Ishikawa et al. 2019).

Rating: 0 (very-high dose)

Clinical In a small, 12-week, open-label, efficacy and safety trial in civilian women with PTSD, addition of memantine (5 to 20 mg/day) to each patient's current medication resulted in significantly reduced PTSD symptoms and treatment was well tolerated (Hori et al. 2021). Similarly, in 26 war veterans with PTSD and cognitive impairment, memantine (5 to 20 mg/day) significantly improved cognitive deficits and reduced PTSD symptoms as shown by PTSD scale scores and Post-traumatic Diagnostic Scale in a 16-week open-label study (Ramaswamy et al. 2015).

Rating: 1 (two studies, encouraging, but open-label)

Radiation therapy

Preclinical In hippocampal cultures expressing enhanced green fluorescent protein, memantine (50 µM) prevented radiation-induced changes in synaptic structure (Duman et al. 2018). In vivo, in irradiated brains at a postnatal stage, memantine (5 mg/kg) had no effect on baseline spine density in dentate gyrus but attenuated spine shortening in the CA1.

Rating: 1 (one paper, high concentration in vitro, in vivo at low dose partial effect only)

Clinical Brown et al. tested memantine (20 mg/day, maximum 3 days after initiating radiotherapy for 24 weeks) in a randomized, placebo-controlled trial for the prevention of cognitive dysfunction in 508 patients receiving whole-brain radiotherapy (Brown et al. 2013). At 24 weeks, memantine tended to decrease decline in delayed recall but the effect failed to reach significance as only 149 patients were available for analysis. However, memantine delayed the time to cognitive decline. The positive effect of memantine on executive function was observed at 8 and 16 weeks, and for processing speed and delayed recognition at 24 weeks (Brown et al. 2013). Subsequently, Brown and colleagues assessed the effects of memantine on the negative adverse effects of radiotherapy (cognitive toxicity) in 518 patients with brain metastases during mean follow-up of 7.9 months (Brown et al. 2020). Compared with whole-brain radiotherapy + memantine, use of hippocampal avoidance during whole-brain radiotherapy + memantine better preserved cognitive function and patient-reported symptoms, with no difference in toxicity, intracranial progression-free survival or overall survival. Due to the study design, conclusions on the effects of memantine could not be made. In another study in 508 patients with brain metastases, memantine 20 mg/day for 24 weeks after whole-brain radiotherapy was assessed for effects on health-related quality of life and cognitive function in a randomized, placebo-controlled study (Laack et al. 2019). On average, although cognitive decline was evident, a decline in health-related quality of life was not reached or was not discernible.

Rating: 1 (mixed results)

Schizophrenia

Preclinical In the early maternal deprivation rat model of schizophrenia, neonatal administration of memantine (10 mg/kg) prevented behavioral changes, such as deficits in active linking (social cognition) and anxiety in plus maze (Uribe et al. 2019). In the same model, memantine (10 mg/kg) administrated to neonatal rats reduced deficits in social interaction in adulthood (Uribe et al. 2016) In vivo, using brain microdialysis, it was found that cystine/glutamate antiporter system xc activity was enhanced by memantine (3 or 10 mM), but not by MK-801, and thalamocortical glutamatergic transmission was inhibited (Okada et al. 2019). The authors suggested that the effects of memantine on thalamocortical glutamatergic transmission are predominantly caused by activation of system cysteine/glutamate antiporter system xc and not action at NMDA receptors (Okada et al. 2019).

Rating: 1 (applied high dose or difficult to judge)

Clinical In 41 patients with chronic psychotic disorder and 41 healthy subjects, the effect of acute (single-dose) treatment with memantine (10 or 20 mg) was analyzed, based on MATRICS Consensus Cognitive Battery (MCCB) performance in a double-blind, randomized cross-over study (Bhakta et al. 2016). Memantine (20 mg) reduced MCCB performance in healthy subjects while in psychotic patients, positive effects of memantine were seen on prepulse-inhibition and single nucleotide polymorphism rs1337697 (in NMDA, GRIN3A gene), predicting greater positive procognitive effects. Memantine (5 to 20 mg/day) as add-on to ongoing atypical antipsychotic therapy was tested on global function and quality of life in 64 in patients with schizophrenia in a randomized, double-blind, placebo-controlled study (Omranifard et al. 2015). Memantine increased Global Assessment of Functioning and quality of life scores significantly (p < 0.001 for both) more than placebo and was well tolerated.

In an 8-week study, 40 patients with schizophrenia who were stabilized on risperidone were randomized to memantine (20 mg) or placebo as add-on to risperidone (Rezaei et al. 2013). At study end, the memantine group showed significantly greater improvements on the negative subscale scores than the placebo group (p < 0.001), the total PANSS score (p < 0.001), and the general psychopathology subscale score (p = 0.002). No significant between-group differences were observed for reduction of positive symptoms, changes in the HMDRS, Extrapyramidal Symptom Rating Scale scores, and frequency of adverse events (Rezaei et al. 2013). In a proof-of-concept study which assessed the effects on cognition and negative symptoms of a 6- or 24-week course of memantine (10 mg BID) as add-on treatment to risperidone in 23 patients with acute or chronic schizophrenia, neuroprotective effects and improvements in cognitive function were shown (Schaefer et al. 2020). Compared with placebo, patients with acute schizophrenia who received add-on memantine showed a significantly higher performance in attention intensity, problem-solving, verbal learning, and flexibility (p ≤ 0.05). Patients with chronic schizophrenia who received add-on memantine showed significantly higher immediate memory (p < 0.05) and a significantly greater reduction of the PANSS sum score (Schaefer et al. 2020).

In an 8-week study, 40 patients with schizophrenia who were stabilized on risperidone were randomized to memantine (20 mg) or placebo as add-on to risperidone (Rezaei et al. 2013). At study end, the memantine group showed significantly greater improvements on the negative subscale scores than the placebo group (p < 0.001), the total PANSS score (p < 0.001), and the general psychopathology subscale score (p = 0.002). No significant between-group differences were observed for reduction of positive symptoms, changes in the HMDRS, Extrapyramidal Symptom Rating Scale scores, and frequency of adverse events (Rezaei et al. 2013). In a proof-of-concept study which assessed the effects on cognition and negative symptoms of a 6- or 24-week course of memantine (10 mg BID) as add-on treatment to risperidone in 23 patients with acute or chronic schizophrenia, neuroprotective effects and improvements in cognitive function were shown (Schaefer et al. 2020). Compared with placebo, patients with acute schizophrenia who received add-on memantine showed a significantly higher performance in attention intensity, problem-solving, verbal learning, and flexibility (p ≤ 0.05). Patients with chronic schizophrenia who received add-on memantine showed significantly higher immediate memory (p < 0.05) and a significantly greater reduction of the PANSS sum score (Schaefer et al. 2020).

In a double-blind, placebo-controlled study, memantine (10 to 20 mg/day) as add-on to olanzapine in patients with schizophrenia improved positive and negative PANSS scores after 6 weeks compared with olanzapine alone (p < 0.001); this effect was more prominent in female patients (Fakhri et al. 2016). Hassanpour and colleagues assessed the safety and efficacy (on psychotic symptoms and cognitive function) of memantine (20 mg/day) as add-on to standard antipsychotic therapy in 40 patients with chronic schizophrenia in an 8-week, double-blind, randomized placebo-controlled study (Hassanpour et al. 2019). Memantine treatment improved verbal memory, learning, verbal letter fluency, and working memory, as shown by PANSS and Brief Assessment of Cognition Scale scores. There were no improvements in psychotic symptoms (Hassanpour et al. 2019). In a 12-week, placebo-controlled study in 26 patients with chronic schizophrenia, the effects of memantine as add-on to conventional treatment on cognitive impairment were investigated (Lee et al. 2012a). In this study, memantine did not improve neurocognition or affect psychopathology, but a trend for improvement was seen in the negative subscale of the PANSS. In 46 male patients with schizophrenia, memantine (20 mg/day) was assessed as add on to risperidone in a 12-week randomized, double-blind, placebo-controlled study (Mazinani et al. 2017). Positive, and general psychopathologic symptoms were not affected by memantine, however negative symptoms and cognitive function were improved. Further to the significant beneficial effects of memantine as adjunct to clozapine on memory and negative symptoms in 52 patients with refractory schizophrenia in a 12-week randomized, double-blind, placebo-controlled cross-over study, the long-term effects and tolerability of memantine as add-on to clozapine were investigated in an open-label 1-year extension study (Veerman et al. 2016). In 31 patients with clozapine-refractory schizophrenia, the positive effects of adjunctive memantine on memory were sustained and further improvement of negative, positive and overall symptoms were demonstrated (Veerman et al. 2016).

A systematic review and meta-analysis performed by Kishi and coworkers concluded that memantine may be effective (as add-on treatment) in patients with schizophrenia for treating psychopathological symptoms, particularly negative symptoms (Kishi et al. 2017a).

Rating: 3 (mixed results but encouraging data on negative symptoms, positive meta-analysis)

Sepsis

Preclinical The effects of memantine (5, 10, 20 mg/kg three days before and 3 days after sepsis) were studied on the long-term consequences of sepsis on the brain in mice (Bardaghi et al. 2023). Memantine normalized MDA, mRNA NF-κB, TNF-α, and IL-1. In contrast, the total thiol content and superoxide dismutase activity increased. Memantine at higher doses also improved behaviors such as cognition in the novel object recognition task, depressive-like behavior in the tail suspension test, and anxiety in the elevated plus maze task (Bardaghi et al. 2023).

Rating: 2 (one paper only but encouraging data)

Skin transplantation

Preclinical Following skin flap transplantation in a rat model of wound repair after surgery, memantine (10, 20 mg/kg) increased skin survival, VEGF expression, microvascular density, angiogenesis, blood perfusion, and superoxide dismutase (Fan et al. 2021). Memantine also decreased levels of neutrophil density and levels of IL-1 β, IL-6, TNF-alpha, TLR4, NF-kappaB, MAPK and malondialdehyde (Fan et al. 2021).

Rating: 0 (high doses used)

Sporadic cerebral amyloid angiopathy

Preclinical Sporadic cerebral amyloid angiopathy is associated with cerebrovascular Aβ deposits and dementia in the elderly (Inoue et al. 2019). In APP23 transgenic mice, memantine (30 mg/kg/d in drinking water) for 6–18 months reduced cerebrovascular Aβ, hemosiderin and Aβ₄₀, and increased levels of hippocampal and vascular insulin-degrading enzyme (Inoue et al. 2019)

Rating: 2 (Rational dose, encouraging data).

Stress-related disorder

Preclinical Impaired social behavior produced by social defeat stress during the juvenile period was studied in adult mice (Yoshida et al. 2022). Memantine (10 mg/kg) was effective when given before a social interaction test, but not when given before social defeat stress. Moreover, memantine inhibited increase in phosphorylated GluN2A and ERK1/2 (Yoshida et al. 2022).

Rating: 0 (one paper, high dose)

Stroke, hemorrhage, post-stroke recovery

Preclinical In an in vitro ischemia–reperfusion injury model in primary human brain microvascular endothelial cells, memantine (5, 10 µM) reduced expression of TNF-α and IL-1b, both analyzed at mRNA and protein levels (Liu et al. 2018). Memantine reduced increased endothelial monolayer permeability and expression of MMP-2, and increased the expression of the transcriptional factor Krueppel-like factor 2 through activating extracellular signal regulated kinase (ERK5). The also data showed improvements in brain vascular barrier integrity (Liu et al. 2018). In a study in an in vitro model of ischemia, rat organotypic hippocampal slices were exposed to a 30-minute period of oxygen-glucose deprivation and memantine (10–30 µM) decreased CA1 hippocampal damage (Landucci et al. 2018). The effect was stronger after combination with hypothermia. In an in vivo model in 7-day-old rat pups based on permanent common carotid artery occlusion combined with hypoxia, memantine (20 mg/kg IP) given directly after the insult and 2h after hypoxia reduced the infarct area. Again, the combination of memantine with hypothermia produced a stronger effect (Landucci et al. 2018).

The effects of memantine on secondary thalamic damage were studied in a permanent distal middle cerebral artery occlusion (MCAO) stroke model in young and old mice. Memantine given 4 and 24 h after stroke (100 and 50 mg/kg, respectively) reduced the infarct area and gliosis (microglia and astroglia) at day 14 in the thalamus (Kim et al. 2021). In the same model in rats, memantine (20 mg/kg) given 24 h after the insult once daily for 7 days decreased sensory decline and reduced p-tau-231, glycogen synthase kinase3βpY216 and protein phosphatase 2ApY307. These changes were accompanied by reduction of neuronal loss, inhibition of reactive astrogliosis in the thalamus, and a decrease in apoptosis (Liang et al. 2020). Sertoli cells implants were combined with memantine (1 and 10 mg/kg at 24 h interval) in a rat MCAO model. Combination attenuated neurological defects, infarct volume, BBB deficiency, and edema (SafialHosseini et al. 2020). Similar effects were observed with each treatment alone, cell therapy, or memantine. In mice exposed to MCAO, memantine treatment (4, 20 mg/kg/day for 28 days starting 72 h post-stroke) at the higher dose improved motor coordination and spatial memory, and reduced striatal atrophy (Wang et al. 2017b). Treatment also reduced astrogliosis and increased capillary formation. Moreover, memantine increased levels of BDNF, GDNF, and VEGF in the striatum and cortex. Interestingly, subunit GluN2B expression of extra-synaptic NMDA receptors was reduced. These data indicate the utility of memantine for stroke recovery (Wang et al. 2017b). In an MCAO model of ischemic stroke in rats, memantine (directly after insult at 20 mg/kg followed by 1 mg/kg during the following days) prevented neurological deficits, brain infarct, and reduced neuronal death (Chen et al. 2017). Activation of the calpain-caspase-3 pathway and cell apoptosis were also reduced (Chen et al. 2017). In a mouse model of permanent MCAO ischemia, memantine decreased brain infarct, neuronal injury, and cortical neuronal apoptosis (Chen et al. 2016). Decreases in damaged astrocytes and over activated microglia were additionally observed. Memantine (20 mg/kg) also decreased MMP-9 secretion, degradation of collagen IV and PSD-95 cleavage. These results indicate the role of the neurovasculature in the action of memantine (Chen et al. 2016).

A model of hypoxia was produced keeping rats at Tibet Plateau at 4300 meters above the sea which, after 4 weeks, resulted in increased NMDA receptor expression but decreased AMPA receptor expression (Ji et al. 2021). Apoptosis in the hippocampus and frontal cortex was observed, as well as an increase in oxidative stress, and later, behavioral disturbances; memantine (5 mg/kg) normalized all of these alterations (Ji et al. 2021). In a hyperoxia-induced brain injury model in immature rats, memantine (20 mg/kg IP single-dose followed by 1 mg/kg) reduced apoptosis as evaluated by TUNEL and caspase-3 analysis (Polat et al. 2020). Memantine (25 µmol/kg, IP) given 30 minutes before global ischemia (bilateral common carotid artery occlusion) increased regional cerebral blood flow and decreased the level of 2,3-dihydroxybenzoic acid (Tanaka et al. 2018). There was no effect on NO2, NO3 or survival (Tanaka et al. 2018). In an intracerebral hemorrhage model produced by intra-striatal collagenase injection in rats, memantine (20 mg/kg/day) administered 30 minutes after the insult and once daily thereafter, improved neurologic deficits, normalized brain water content, and reduced MMP-9, NLRP3, and IL-1β levels and neuronal death (Chen et al. 2021).

Rating: 2 (only a few papers used therapeutically relevant does)

Clinical In a randomized, open-label study in patients with mild-to-moderate cerebral thromboembolic event, the addition of memantine (20 mg TID for 5 days) to conventional treatment was assessed (n = 29 in control arm; n = 24 in memantine arm). The National Institute of Health Stroke Scale (NIHSS) analysis revealed significant improvement with memantine versus control (p < 0.0001), indicating improved neurological function (Kafi et al. 2014).

Considering post-stroke recovery, Barbancho and colleagues studied memantine in 28 chronic post-stroke aphasia patients for changes in event-related potentials (P100 and N400) and root mean square (RMS) during a silent reading task in a randomized, double-blind, placebo-controlled trial (Barbancho et al. 2015). Memantine was given for 16 weeks, followed by combination with constraint-induced aphasia therapy (weeks 16–18) and memantine alone (weeks 18–20). Memantine decreased event-related potentials /root mean square values at week 16. In the second phase (weeks 16–18) memantine amplified improvements in aphasia severity and event-related potentials /root mean square values. These changes remained visible until week 20. Changes corresponded to improvement in language performance (Barbancho et al. 2015).

Rating: 2 (two studies only but encouraging)

Tardive dyskinesia

Preclinical Memantine (0.5, 5, 25 mg/kg) failed to affect tremulous jaw movements in rats after chronic treatment interpreted as anti-tremor activity; however, it was effective when combined with Δ⁹ -tetrahydrocannabinol (THC) (Ionov et al. 2020).

Rating: 0 (negative data)

Tinnitus

Preclinical In a salicylate-induced tinnitus model in rats, memantine (5 mg/kg for 7 days) improved gap-prepulse inhibition of the acoustic startle reflex, indicating tinnitus suppression (Jang et al. 2019). Additionally, expression of NR2B was decreased. In an in vitro study in SH-SY5Y cells and in an animal model, memantine attenuated the increase in expression of inflammatory cytokine genes (TNFα and immediate-early gene ARC) (Jang et al. 2019). Administration of piribedil (a dopamine agonist) and/or memantine was tested in noise-induced loss of inner hair cell ribbons in rats (Altschuler et al. 2016). Combined treatment with piribedil and memantine (3 mg/kg) reduced loss of ribbons and corresponding changes in ABR sensitivity and dynamic responsiveness; memantine alone was not effective (Altschuler et al. 2016).

Rating: 2 (relevant doses)

Traumatic brain injury or brain injury

Preclinical In brain slices from mice hippocampus, memantine (10–300 µM) did not affect the induction of spreading depolarization but impaired spreading depolarization propagation and recovery (Reinhart et al. 2021). These results may be relevant for secondary insult after acute brain injury. After repetitive mild traumatic brain injury (TBI) in mice, memantine (10 mg/kg) decreased oligodendrocyte loss and myelin basic protein expression shortly after injury. Memantine also attenuated axon damage, as evidenced by neurofilament light chain (NF-l) expression (Ma et al. 2019). In the same model, memantine at 10 mg/kg 1 hr after the last insult reduced tau phosphorylation shortly after injury and at 1 month, decreased glia activation and reduced LTP deficit (Mei et al. 2017). Memantine also normalized NMDA receptor expression. At the behavioral level, no effect of memantine was observed (Mei et al. 2017). Repetitive mild TBI was induced by two insults (24 h apart) in vitro in organotypic hippocampal slice cultures (Effgen and Morrison Iii 2016). Memantine 1.5 µM 1 h after each stretch insult decreased the effect of insults for all outcomes measured, including cell death, impairment of LTP and astrogliosis (Effgen and Morrison Iii 2016). The authors concluded that memantine warrants further preclinical and clinical investigations for its potential in preventing cognitive deficits and neuropathology from multiple mild TBIs.

Delayed (1h) treatment with memantine in a controlled cortical impact brain injury model prevented neuronal degeneration 1 day after injury at 10 mg/kg (but not at 2.5 or 5 mg/kg) (Abrahamson et al. 2019). Seven days after injury, memantine decreased loss of synaptophysin immunoreactivity in the hippocampus. At 21 or 90 days, no effect on histopathology was seen. Similarly, there was no effect on vestibulomotor function and spatial memory acquisition in the water Morris maze (Abrahamson et al. 2019). In a fluid percussion TBI in rat model, memantine was given initially at 20 and then at 1 mg/kg IP every 12 h for a total of six doses (Wang et al. 2018a). At 72 h, memantine attenuated motor and proprioception deficits and infarction volume, and improved apoptosis, astrogliosis, and microgliosis (Wang et al. 2018a). In the same model, memantine (5 mg/kg, IV) given after trauma induction increased neuronal survival in the cortex, but failed to restore learning deficit; 17-estradiol (E2) enhanced the effect of memantine at lower doses (0.05 and 0.5 mg/kg) (Day et al. 2017). In a mild TBI model based on weight drop with rotational acceleration (5x in 5 days), acute and delayed treatment with memantine were assessed (Boucher et al. 2024). Animals randomly received memantine directly, 3 or 6 months post-injury for a treatment duration of 1 month (Boucher et al. 2024). Memantine (20 mg/kg) given directly after insult or 6 months later produced better rescue of neuroinflammatory microglia reaction than vehicle control or treatment at 3 months.

Memantine was evaluated for prevention of spinal cord injury in a rabbit paraplegic model induced by clamping both the infrarenal aorta and inferior vena cava (Panthee et al. 2020). Preoperative oral memantine (60 mg) for 7 days, followed by gradual decrease in dose, improved the mean Tarlov score. This effect correlated with plasma levels of memantine and the authors concluded that memantine was protective against spinal cord injury at serum levels ≥ 4.5 ng/mL in the rabbit model used, and may be a potential adjunct for spinal protection during thoracicthoracoabdominal aortic surgeries (Panthee et al. 2020).

Rating: 2 (in some papers high doses were used)

Clinical In a randomized controlled trial in 41 TBI patients, the effects of memantine on serum levels of neuron-specific enolase, as an indicator of brain damage, and on the Glasgow Coma Scale, as a measure of symptom expression, were assessed (Mokhtari et al. 2018). On days 3 and 7, levels of enolase were lower in memantine-treated patients than in the control group, and Glasgow Coma Scale scores were better in the memantine group on day 3. Interestingly, there was negative correlation between biomarker levels and symptom expression. Brain injury is one of the complications following open-heart surgery, with glutamate considered to play a role. In a study in 34 consecutive patients undergoing on-pump coronary artery bypass graft surgery, compared with control, memantine (10 mg BID) at least 72 h before surgery significantly attenuated post-operative serum S100-B concentrations. The authors commented that this may reduce cerebral damage during surgery (Ziabakhsh Tabary et al. 2022).

Rating: 1 (one study, mild effect)

Trichotillomania

Clinical Memantine (10 to 20 mg/day) was tested in an 8-week, double-blind, placebo-controlled trial in 100 patients (86% women) with the compulsive behavior disorders, trichotillomania or skin picking (Grant et al. 2023). Compared with placebo, significant improvements were seen in scores on the NIMH Trichotillomania Symptom Severity Scale, modified to include skin picking, and on secondary outcome measures. In addition, at the study end, over 60% of the memantine-treated group were much or very much improved versus 8.3% of the placebo group, with no between-group safety concerns (Grant et al. 2023).

Rating: 1 (one study but encouraging)

Trypanosoma

Preclinical The effects of memantine in models of Chagas disease, caused by Trypanosoma cruzi, were studied both in vitro (infection of RAW 264.7 macrophages) and in vivo (in BALBc mice) (Santos Souza et al. 2019). In vitro, memantine (starting at 10 µM) inhibited NO and Ca²⁺ entry in Escherichia coli-derived LPS-activated cells. In mice, memantine (10 mg/kg) decreased parasitemia, cardiac parasitic load and inflammatory infiltration, and increased survival (Santos Souza et al. 2019).

Rating: 1 (one paper, high dose)

Tumor

Preclinical Several types of cancer cells have been reported to express NMDA receptors. Memantine (and metformin) were tested on glucose and glutamine metabolism in androgen-dependent LNCaP cells and on prostate cancer cell proliferation in vitro (Albayrak et al. 2018). Memantine showed cytotoxic effects at 250 µM, which the authors suggested involved the Bax-dependent apoptotic pathway (Albayrak et al. 2018). Similar effects of memantine (250 µM) were observed on proliferation of 4T1 cells (breast cancer model) as demonstrated by the MTT assay (Albayrak and Korkmaz 2020). Memantine also modified Bcl-2, Bax, Casp3, and Casp9 protein expression (Albayrak and Korkmaz 2020). Memantine (250 µM) was studied for effects on cell cycle progression and cell death in A549 cells; decreased cell viability by induction of G0/G1 cell cycle arrest was shown. Additionally, memantine decreased cancer cell metabolism (HIF1A, B-catenin and PKM) (Albayrak and Demirtas Korkmaz 2021). In T98G and U87-MG human glioblastoma cells, memantine induced cytotoxicity with an IC₅₀ of 250 µM (Albayrak et al. 2021). In mice with 4T1 tumors, memantine at 10 (but not 5) mg/kg reduced tumor volume and decreased genome-wide DNA methylation levels (Albayrak et al. 2022). The induction of autophagic cell death by memantine was studied in glioma cell lines (T-98 G and U-251 MG) (Yoon et al. 2017). Memantine produced an antiproliferative effect in T-98 G cells (expressing NMDA receptors) and increased proteins involved in autophagy (Yoon et al. 2017). Memantine studied in the breast cancer cell line MCF-7 starting at 20 µM concentrations reduced cell survival, decreased tau and stathmin expression (known regulators of cell motility), and inhibited cell migration; some synergy with paclitaxel was also observed (Seifabadi et al. 2017).

Rating: 1 (high doses/concentrations)

Vascular dementia

Clinical In the absence of an effective treatment for vascular dementia, which is characterized by cognitive impairment and cerebrovascular pathologies, several studies have assessed the utility of memantine. In middle-aged and older patients with confirmed moderate to moderately-severe vascular dementia, memantine 20 mg once daily or 10 mg BID were compared for efficacy and safety over 24 weeks using a battery of dementia of assessment scales (ADAS-cog, MMSE, CGI-C, CGI-S and HAM-D). Both treatment regimens showed similar efficacy, which was statistically significant compared with baseline, and safety; however, there was no placebo comparator in this study (Gavrilova et al. 2024). In a retrospective study, memantine (20 mg/day) was used as add-on to scalp electroacupuncture and patients were assessed for cognitive function, daily living and quality of life (Montreal Cognitive Assessment [MoCA]), Barthel index, dementia quality of life questionnaire, blood superoxide dismutase, lipid peroxide and NO. Memantine, in combination with electroacupuncture, showed a significantly greater therapeutic effect than either treatment alone (p < 0.01 and p < 0.05). (Yue et al. 2020).

Rating: 1 (design of the studies not optimal, mixed results)

Viral infections

Preclinical In a Zika virus mouse model of temporary flaccid paralysis in IFNa/β-receptor KO mice, memantine (60 mg/kg/day) on days 4–9 after infection reduced the incidence of paralysis; however, in mice cell cultures, memantine did not affect viral titers (Siddharthan et al. 2020). Memantine (25 µM) in neuronal cultures infected with the rabies virus was neuroprotective (Sun et al. 2019). In vivo, memantine (20 mg/kg) showed only a marginal effect on survival in mice. However, memantine prolonged survival in mice infected with the Japanese encephalitis virus and reduced the amount of virus in the mice brain (Sun et al. 2019).

Rating: 1 (high doses/concentration of memantine used)

Ongoing/planned studies

A search conducted on the clinicaltrials.gov database in March 2025 revealed 29 studies with memantine registered since 2014 that have a status of active/not recruiting (n = 2), not yet recruiting (n = 7), or recruiting (n = 20) (Table 4) (clinicaltrials.gov). Notably, five studies are in AD and two in cognitive impairment or neurocognitive dysfunction. Other indications under investigation in ongoing clinical studies of memantine are: cancer, including forms of breast, liver, lung and brain cancer (n = 14); cardiovascular or cerebrovascular disease including stroke (n = 2); autoimmune/inflammatory diseases (n = 2); and Fragile X syndrome (n = 1), trichotillomania (n = 1), schizophrenia (n = 1), and psychosis (n = 1). These ongoing studies highlight the sustained interest in memantine and its potential clinical applications, while the current research reflects and extends the progress made in the last decade.

Table 4.

Clinical trials registered at clinicaltrials.gov (since ca. 2014) including the use of memantine in Alzheimer’s disease or other indications

Clinical trial identifier	Conditions/indications	Interventions	Start date	Trial status
NCT06862960	Alzheimer’s disease	Ozanimod Conventional medication	1/05/2025	Not yet recruiting
NCT06789757	Hepatocellular carcinoma	Memantine Bevacizumab Atezolizumab	25/02/2025	Recruiting
NCT06727773	Breast cancer (stage 1–3)/locally advanced Cognitive impairment/decline/change	Memantine Placebo Exercise	2025-03	Not yet recruiting
NCT06679387	Breast cancer early stage (stage 1–3)	Memantine Placebo	30/10/2024	Recruiting
NCT06594172	Radiation disease/cognitive impairment/drug effect	Memantine Pioglitazone Hippocampal avoidance WBRT	10/09/2024	Not yet recruiting
NCT06501638	Atrial premature beats/contractions/systoles	Memantine Placebo	7/07/2024	Not yet recruiting
NCT06275035	Neurocognitive dysfunction	Memantine	22/02/2024	Recruiting
NCT06007846	Hepatocellular carcinoma/cirrhosis	Namenda	31/07/2023	Recruiting
NCT05834231	Cancer	Memantine Placebo	2023-08	Not yet recruiting
NCT05807165	Brain oligometastases	Hypofractionated stereotactic radiotherapy Hippocampus-sparing WBRT plus memantine	1/08/2023	Not yet recruiting
NCT05801380	Alzheimer’s disease	AChEI AChEI plus NMDA receptor antagonist	14/02/2022	Recruiting
NCT05796752	Skin-picking/trichotillomania (Hair-pulling disorder)	Memantine ComB Behavioral Therapy	2/08/2023	Recruiting
NCT05664464	Glioblastoma	Gabapentin Sulfasalazine Memantine Temozolomide Radiotherapy	1/01/2023	Recruiting
NCT05647473	Orthostatic hypotension/Alzheimer’s disease	Routine treatment Add-on low- or high-dose Astragalus	20/02/2024	Recruiting
NCT05564169	Alzheimer’s disease	Placebo Masitinib Standard of care	2024-01	Not yet recruiting
NCT05418049	Fragile X syndrome	Baclofen Memantine Roflumilast Placebo	8/09/2022	Recruiting
NCT05063851	Alzheimer’s disease	Memantine Placebo	11/10/2021	Recruiting
NCT05045950	Brain metastases	Memantine - BID or ER WBRT - PRDR technique	17/11/2021	Recruiting
NCT05013892	Brain metastases	Normal tissue sparing WBRT Memantine	8/02/2022	Recruiting
NCT04939597	Central nervous system carcinoma	Biospecimen collection Cognitive assessment MRI Memantine Placebo	10/05/2022	Recruiting
NCT04857983	Schizophrenia/schizoaffective disorder	Memantine Placebo	6/07/2021	Recruiting
NCT04804644	Metastatic or recurrent lung small cell carcinoma/metastatic malignant brain neoplasm/stage IV lung cancer	Biospecimen Collection MRI Memantine Neurocognitive Assessment Stereotactic radiosurgery WBRT	8/06/2021	Recruiting
NCT04789915	Psychosis/negative symptoms with primary psychotic disorder	Memantine Placebo	26/05/2021	Recruiting
NCT04302870	Motor neuron disease, amyotrophic lateral sclerosis	Memantine oral solution Trazodone oral solution Placebo oral solution Amantadine oral solution	27/02/2020	Recruiting
NCT03703856	Alzheimer’s disease	Memantine Placebo	31/01/2019	Recruiting
NCT03550391	Brain metastases	Memantine Hippocampal-avoidant WBRT Stereotactic Radiosurgery	22/11/2018	Recruiting
NCT03527472	Lupus erythematosus, Systemic	Memantine Placebo	23/08/2018	Recruiting
NCT02144584	Ischemic stroke/upper extremity weakness	Memantine XR Placebo	2014-01	Active not recruiting
NCT01430351	Glioblastoma/gliosarcoma/supratentorial glioblastoma	Mefloquine Memantine Metformin Temozolomide	14/09/2011	Active not recruiting

AChEI Acetylcholinesterase inhibitor, MRI Magnetic resonance imaging, WBRT Whole-brain radiotherapy

Conclusions

• In general, new evidence on the MoA of memantine supports the previous dominant role of NMDA receptor antagonism. In this context, the supportive evidence for preference of memantine for extrasynaptic NMDA receptors is noteworthy. Increase of GDNF is a plausible supportive mechanism given that this effect is induced at a therapeutic dose/concentration range of memantine.

• Both preclinical and clinical studies support the use of memantine for AD and dementia in general. An interesting aspect is the accumulating evidence for neuroprotective (disease-modifying) potential; however, a final proof-of-concept clinical study is still lacking.

• Analysis of evidence from combined preclinical and clinical data (since 2014) of new potential indications for memantine revealed promising preliminary findings for depression and bipolar II disorders; schizophrenia (negative symptoms); dyskinesia after L-DOPA in patients with Parkinson’s disease; ECT-induced cognitive impairment; alcohol use disorder; stroke, hemorrhage, post-stroke recovery; and OCD, followed by autism spectrum disorder; bipolar disorder; Dementia with Lewy bodies and Parkinson’s disease with dementia; fibromyalgia; fragile X syndrome; and migraine.

Abbreviations

5-HT3

5-hydroxytryptamine type 3

AChE

Acetylcholinesterase

ADL

Activities of daily living

α7-nAChR

Alpha7-nicotinic acetylcholine receptor

Aβ

Amyloid beta peptide

AChEIs

Acetylcholinesterase inhibitors

AD

Alzheimer’s disease

ADAS-cog

AD assessment scale-cognitive subscale

ADCS-ADL

Alzheimer’s disease cooperative study-activities of daily living inventory

ADL

Activities of daily living

APP

Amyloid precursor protein

ADHD

Attention deficit hyperactivity disorder

BBB

Blood-brain barrier

BDNF

Brain-derived neurotrophic factor

BID

Twice daily

CGI-C

Clinical Global Impressions of Change

CGI-I

Clinical global impressions-improvement

CGI-S

Clinical global impressions-severity

CREB

cyclic-AMP response element-binding protein

CSF

Cerebrospinal fluid

ELISA

Enzyme-linked immunosorbent assay

ER

Extended release

GABA

Gamma-aminobutyric acid

GDNF

Glial cell line-derived neurotrophic factor

HAM-D

 Hamilton depression rating scale

IL

Interleukin

IC₅₀

Half-maximal inhibitory concentration

ICV

Intracerebroventricular

IP

Intraperitoneal

LTP

Long-term potentiation

MoA

Mechanism of action

mRNA

Messenger ribonucleic acid

MMSE

Mini-mental state examination

NMDA

N-methyl-D-aspartate

NO

Nitric oxide

NPI

Neuropsychiatric inventory

OCD

Obsessive-compulsive disorder

Per os

By mouth

PP2A

Protein phosphatase 2A

PTSD

Post-traumatic stress disorder

SIB

Severe impairment battery

STZ

Streptozotocin

TNF-α

Tumor necrosis factor-α

Declarations

Competing interest

WD, AS and AG are employees of Merz Therapeutics. JH, JW and JK have no competing interest in relation to this manuscript.