Memantine in the treatment of mild-to-moderate Alzheimer’s disease
Memantine is the first and only medication that has been approved by European, US and Canadian regulatory agencies for the treatment of moderate-to-severe Alzheimer’s disease (AD). It is an NMDA receptor antagonist that works to prevent excitotoxicity and cell death, which are mediated by the excessive influx of calcium during a sustained release of glutamate. Preclinical studies of memantine reveal that it has the potential to improve memory and learning processes after impairment has occurred, as well as to prevent further neuronal damage. Although memantine has been considered for the treatment of earlier AD, it has not yet been approved for this. Randomized controlled trials of memantine in the treatment of mild-to-moderate AD have demon- strated small treatment effects in measures of cognition, global assessment and behavior favoring the use of memantine. However, the differences between treatment groups were not consistently significant. Two ongoing long-term trials are further investigating the efficacy of memantine in the treatment of mild-to-moderate AD.
Keywords: Alzheimer’s disease, dementia, glutamate, memantine, NMDA
1. Introduction
Results from the 2000 census reveal that 4.5 million US residents are diagnosed with Alzheimer’s disease (AD) [1]. An analysis completed in 1998 found that the number of new diagnoses of AD made each year was between 200,000 and 600,000; it was also estimated that this incidence could increase to 1.14 million new cases in 2047 [2]. The prevalence and incidence of AD have been shown to generally increase with age, as does the severity of disease symptoms [1,3,4]. Contemporary research has shifted from focusing mostly on apparent signs and symptoms to incorporating data gained from neuroimaging techniques and other analyses of biomarkers when diag- nosing AD [5-7]. As this shift permeates from the research to the clinical environ- ment, physicians will probably be able to make more accurate diagnoses of AD much earlier in the stages of the disease. Brain imaging and other biomarkers are likely to identify pathologic signs of AD much earlier than the time that cognitive deficits are apparent, especially for individuals who have greater so-called brain reserve, which enables them to hide or compensate for any underlying problems [6]. Consequently, it might be expected that the future prevalence of AD, particularly mild AD, may increase due to more sensitive diagnostic measures.
The pathology of AD is chiefly characterized by the proteins -amyloid (A) and hyperphosphorylated tau, which lead to the excessive presence of neuritic plaques and neurofibrillary tangles. Neuronal loss and reduced synaptic density also typically occur in AD, resulting in deficits in the cholinergic, noradrenergic and serotonergic transmitter systems [8]. Strong correlations have been found between Braak stage, a method of measuring the severity of neurofibrillary pathology at autopsy, and cognitive status before time of death [9]. Promis- ing advances have been made in the use of neuroimaging tech- niques, including structural and functional magnetic resonance imaging (MRI), water and flurodeoxyglucose positron emission tomography (PET) and single photon emission computed tomography (SPECT), to detect abnormal changes in atrophy, activity level and metabolism in various regions of the brain that may be suggestive of AD or other neurologic disorders [7]. As the information gained from these techniques accumulates, it will probably be easier to better understand the pathologic features that develop over the course of the disease.
Glutamate, a critical excitatory neurotransmitter in the cerebral cortex and hippocampus, has also been identified as a component involved in the etiology of AD. Glutamatergic activation at the NMDA receptor is known to be an essential step in learning and memory, and is involved in a mechanism of synaptic plasticity called long-term potentiation [10,11]. The NMDA receptors are present at high concentrations through- out the brain, especially in the neocortex and hippocampus, and are necessary for normal function.
However, pathologic activation of these receptors through excessive stimulation by glutamate results in an overload of intracellular calcium and impaired neuronal homeostasis. Prolongation of the calcium overload can cause gradual damage to the neurons, a process that has been termed ‘excitotoxicity’. If the excessive intraneuronal calcium levels continue to be present, cellular apoptosis, a programmed mechanism of cell death, may also occur [12]. In addition to the damage that glutamate excitotoxicity may cause to undamaged neurons, excessive glutamate activity and activa- tion of the NMDA receptor have been thought to enhance the production of pathologic forms of A and tau [13-15]. Furthermore, several preclinical studies have shown that the presence of A adversely affects calcium homeostasis in the presence of glutamate, consequently exacerbating the neuro- toxic effects of prolonged glutamatergic stimulation [16,17]. As a result, each separate pathologic condition seems to com- pound the other, worsening the factors that contribute to AD progression. Modification of glutamate activity at the NMDA receptor site may help to reduce the severity of these pathologic processes and consequently delay the disease progression.
2. Overview of the market
At present, the only drugs that are available for the treatment of AD are the cholinesterase inhibitors (ChEIs), three of which are approved for the treatment of mild-to-moderate AD, one (donepezil) is approved for mild-to-severe AD, and memantine, which has FDA approval for the treatment of moderate-to-severe AD. ChEIs offer modest therapeutic benefits and delay the onset of severe deficiencies in cognition and function, potentially playing a more effective role when treatment is initiated in the earliest stages of the disease [18]. ChEIs that have been on the market include tacrine, donepezil, rivastigmine and galantamine. Tacrine is rarely used, due to the associated high risk of hepatotoxicity and gastrointestinal side effects [19,20]. Overall the ChEIs are thought to be similar in efficacy, but slightly different in side-effect profile. Thus, ChEI selection in clinical practice tends to be based on preference and the tolerance of the patient to each treatment option. Although the ChEIs provide some therapeutic benefit for the patient, this benefit is moderate, and the ultimate progression of the disease cannot be halted with this treatment [21-24]. Successful development of an agent that has the ability to cure, halt or slow progression of AD is lacking.
Even with early treatment of AD with a ChEI, the progres- sion of the disease will eventually outweigh the benefit provided by medical therapy. In addition to attempting to slow the progression of AD, physicians will often resort to alternative treatment options to address the non-cognitive symptoms that often occur in the later stages of the disease, when non-pharmacologic interventions are not sufficiently effective. Antipsychotic medications, most commonly the atypical antipsychotics, are frequently used for the treatment of behavioral disturbances, but do not have FDA or European Medicines Agency (EMEA) indications for this use. It should also be noted that concerns have been raised regarding cerebrovascular adverse events, and even excess mortality, associated with these agents. Antidepressants should be considered to treat depressive symptoms; however, treatment guidelines for depression associated with AD based on empirical data are yet to be established [22,23,25-27].
There are several avenues under investigation as potential means by which the progression of AD can be altered. A predominant area of research involves the modulation of the neuropathologic features of the disease, with the amyloid pro- tein serving as a specific target. Preclinical data and one clini- cal trial offer evidence to support the theory that vaccination, with the intent of producing A antibodies, would slow the rate of disease progression [28-30]. The same study was prema- turely terminated due to a high incidence rate of meningoen- cephalitis [31]. Alternative possibilities being researched for anti-amyloid therapy that may be safer than the active vaccination include passive immunization [32,33], -secretase inhibitors [34], -secretase inhibitors [35], statins [36], insu- lin-degrading enzymes [37] and other selective amyloid-lower- ing agents. Prevention of the neurotoxicity associated with A is also a possible strategy that may develop into a future treat- ment option. Preventative agents, including antioxidants [38] and anti-inflammatory agents [39], might be effective in individuals with mild (or no) cognitive impairment, and need to be evaluated more thoroughly. Furthermore, compounds that may normalize tau phosphorylation, such as lithium [40] and valproate [41], might reduce hyperphosphorylated tau-induced neurotoxicity by inhibiting the enzyme tau kinase. The use of combination therapy has been considered, but additional research is needed to better assess the benefit of this tactic in the treatment of AD [22,42].
3. Memantine
3.1 Introduction to the compound
The glutamate neurotransmitter system offers another treat- ment option for AD. Ongoing research efforts identify the glutamatergic system as a promising candidate for both enhancement of cognition and interruption of certain patho- physiologic processes associated with AD. Moderate stimula- tion at the NMDA receptors (postsynaptic receptors that are stimulated by glutamate) may improve memory and cogni- tion. However, excessive stimulation at this site will poten- tially result in neuronal injury and toxicity. At rest, under normal physiologic conditions, the receptor channel is occupied by Mg2+, which blocks the entry of Ca2+ into the neuron. When glutamate concentrations are elevated in the synapse (which occurs during physiologic learning and memory processes), Mg2+ exits the NMDA receptor, allowing normal Ca2+ influx. The blockade by Mg2+ is highly voltage dependent, and Mg2+ leaves the channel relatively easily. When there is a continuous release of low concentrations of glutamate, such as under pathologic conditions associated with AD, depolarization is prolonged, and Mg2+ is displaced from the NMDA receptor channel for an abnormally long period of time.
Consequently, Ca2+ continually flows into the neuron, causing an overload of intraneuronal calcium. This overload impairs neuronal homeostasis, which may lead to excitotoxicity (neuronal damage) and even cell death. The intraneuronal Ca2+ excess may additionally cause an increase in slow after hyperpolarization (sAHP). This results in diminished frequency of neuronal firing and, due to an ineffective signal-to-noise ratio, the transmission of physio- logic signals is inhibited. Thus, possibly due to the same processes, cognitive deficits and neuronal loss occur in neurodegenerative dementia [11,43-45].
With the intention of preventing glutamate excitotoxicity, efforts have been made to block the NMDA receptor with a compound that has a higher affinity to the receptor than Mg2+. First attempts with such a compound resulted in agents that completely block glutamatergic activity at the NMDA receptor. As a result, in addition to preventing the pathologic activity of glutamate, physiologic activity (such as learning activity and the formation of memories) was also inhibited. These high-affinity compounds, including dizocilpine and phencyclidine, impair learning in animal models, and can produce psychotomimetic effects. Glutamate antagonists with a moderate affinity for the NMDA receptor have led to more successful outcomes [10,11,43,46,47].
Memantine (1-amino-3, 5-dimethyladamantane hydro- chloride) is one such antagonist. It is non-competitive and has a moderate affinity for the NMDA receptor. Memantine exerts voltage-dependent effects with rapid blocking and unblocking kinetics. It preferentially blocks the NMDA receptor when glutamate levels are pathologically high, but has minimal effect in normal physiologic environments. Thus, normal glutamatergic neurotransmission is permitted under physiologic circumstances, and excitotoxicity is inhibited in pathologic conditions when glutamate chronically stimulates the receptor [44,48,49].
Memantine is available in 5- and 10-mg tablets in the US, and in 10-mg divisible tablets elsewhere. Recommended dos- ing begins at 5 mg/day, and is titrated to a final target dose of 20 mg/day. Doses should be increased in 5-mg increments, with at least 1 week elapsing between dose increases. Titration steps should include 10 mg/day (5 mg b.i.d.), 15 mg/day (5 and 10 mg, as separate doses) and finally, 20 mg/day (10 mg b.i.d.). Memantine is also available in liquid form for patients who have difficulty swallowing [101].
3.2 Chemistry
Memantine hydrochloride, C12H21N•HCl, has a molecular weight of 215.76 Da. It is found as fine white or off-white powder and is soluble in water. The chemical structure of memantine is shown in Figure 1 [101].
3.3 Pharmacodynamics
Memantine has an affinity for the NMDA receptor at the same channel as Mg2+, but has a higher affinity than Mg2+ and is less voltage dependent. Therefore, it does not exit the channel as easily with minor depolarization. Accordingly, it is more effec- tive at blocking the sustained influx of Ca2+ that accompanies the pathologic activation of the NMDA receptors. Memantine does leave the channel during physiologic glutamate stimula- tion, allowing the normal influx of Ca2+. Furthermore, the reduction in synaptic noise that occurs when memantine is present improves the recognition and conduction of physiologic signals [11,48,49].
Memantine reduces acute excitotoxic damage in vitro and in vivo following the administration of glutamate agonists. In models of global and focal ischemia, it has conferred protection against excitotoxicity and improved cognitive performance in rats [44,48,49]. Improvement in cognitive performance following memantine treatment has also been demonstrated in transgenic mice models of AD [50]. Therapeutic doses of memantine have also shown protective effects against pathologic conditions that are specifically relevant to AD, including A toxicity, inflammation, inhibition of mitochondrial function and reduced blood flow to the brain [11].
In addition to its action at the NMDA receptor site, memantine also blocks 5-HT3 receptors at physiologic concentrations. Antagonism at 5-HT3 has been postulated to facilitate long-term potentiation, to have antipsychotic and antinausea effects and to decrease gastric hypermotility [48,49].
3.4 Pharmacokinetics and metabolism
After oral administration, memantine is rapidly absorbed, with a Tmax of 3 – 7 h. It has a high oral bioavailability that is unaffected by food. Memantine shows linear pharmaco- kinetics over the therapeutic dose range, and its terminal elimination half-life is 60 – 80 h. It is extensively distributed in tissue, and readily crosses the blood–brain barrier. Memantine is 45% protein bound [101].
Figure 1. Chemical structure of memantine.
Memantine is partially metabolized by the liver, but the CYP450 enzyme system is not significantly involved in this metabolism. Data have indicated that there are no expected pharmacokinetic interactions with drugs metabolized by the CYP450 enzymes. When ingested, 48% of memantine remains unchanged, and is excreted in urine. Much of the rest of the drug is broken down to form three polar metabolites that are minimally active NMDA receptor antagonists. Memantine clearance is reduced with increasing degrees of renal impairment, and dosage adjustments are recommended for patients with moderate or severe renal impairment. Medications or conditions that raise urine pH may decrease the urinary elimination of memantine, resulting in increased plasma levels [101].
3.5 Clinical efficacy
Preclinical data have provided evidence supporting the hypothesis that memantine should prevent further neuronal damage, and should also improve learning and long-term potentiation when those processes are impaired. Experimental models that mimic the deficiencies associated with AD have paved the way for a number of clinical trials of memantine in the treatment of AD [11].
3.5.1 Memantine in moderate-to-severe Alzheimer’s disease
Three randomized, placebo-controlled studies have been conducted to evaluate the use of memantine in the treatment of moderate-to-severe AD. One of these, MEM-MD-01, which assessed the safety and efficacy of memantine mono- therapy in 350 participants with Mini-Mental State Examina- tion (MMSE) scores ranging between 5 and 14 (mean 10.1) over 24 weeks, failed to show statistically significant differences between the memantine and placebo groups on all measures of cognition, function and behavior [102]. Results from the other two studies were more promising, offering statistically significant results favoring the use of memantine. Study MRZ9605 included 252 patients with MMSE scores between 3 and 14 (mean 7.9) who were not concurrently receiving a ChEI. At the end of the 28-week trial period, significant differences (p < 0.05) were demonstrated on observed cases (OC) analyses of measures of global assessment, function and cognition, as well as last observation carried forward (LOCF) analyses of daily function and cognition [51]. A 24-week, open-label extension following the placebo-controlled phase further supported the efficacy of memantine, as participants who were previously in the placebo group experienced significant benefits in all measures relative to their rate of decline in the placebo-controlled phase (p < 0.05) [52]. The third randomized, placebo-controlled study of memantine in the treatment of moderate-to-severe AD, MEM-MD-02, was conducted to assess the efficacy of memantine when used in combination with the ChEI donepezil. In total, 404 participants with MMSE scores of 5 – 14 (mean 10.2 for placebo, 9.9 for memantine) were enrolled in this 24-week trial. Statistically significant (p < 0.05) differences in favor of memantine were found in all analyses (LOCF and OC) of all instruments used, including measures of cognition, daily function, global assessment and behavior. Highly statistically significant (p < 0.001) differ- ences were found on the Severe Impairment Battery – an assessment scale for cognitive function. Notably, the effect sizes (shown as the difference between means divided by the mean standard deviation) calculated for all assessments were quite large (effect sizes 2.78), suggesting that the benefit provided by this therapy is clinically significant as well [53]. A fourth study that included patients diagnosed with moder- ate-to-severe AD, the M-BEST (Benefit and Efficacy in Severely Demented Patients during Treatment with Meman- tine) study, investigated the effects of memantine treatment in patients with dementia. Of all the patients in the study, 49% had AD, and 51% had dementia of the vascular type. A total of 167 participants with MMSE scores of < 10 (mean 6.3) were enrolled in this study, which was 12 weeks in duration. Statistically significant results favoring the use of memantine were demonstrated on both primary outcome measures used, showing efficacy in areas of behavior and global outcome. No method was used to monitor any changes in cognition [54].
3.5.2 Memantine in vascular dementia
Memantine has also been studied in populations with vascular dementia, which, similar to early AD, tends to have a slower rate of decline. Results from two randomized, controlled studies found significant differences favoring memantine in measures of cognition, but no significant differences in measures of global assessment [55,56]. Safety analyses from these studies further supported the safety and tolerance of memantine.
3.5.3 Memantine in mild-to-moderate Alzheimer’s disease
Three randomized, controlled trials have been completed to assess the safety and efficacy of memantine specifically in participants with mild-to-moderate AD (MMSE scores ranging from 10 to 23). Two of these looked at the use of memantine monotherapy for the treatment of AD [57,58]. The third investigated the efficacy of memantine when used in combination with one of the ChEIs (Porsteinsson AP, Grossberg GT, Mintzer J, Olin JT: Memantine treatment in patients with mild-to-moderate Alzheimer’s disease already receiving a cholinesterase inhibitor: a randomized controlled trial. Manuscript in preparation). The results from these stud- ies are mixed and provide only minimal statistically significant evidence supporting the use of memantine in this population. Table 1 illustrates the highlights from each of these trials.
3.5.3.1 Monotherapy
The MEM-MD-10 study was a 24-week study that enrolled participants with MMSE scores ranging from 10 to 22 [57]. In total, 403 out-patient participants were randomized, of which 394 participants completed at least one postbaseline efficacy assessment. The mean MMSE scores at baseline were similar between the memantine (17.4) and placebo (17.2) groups (p = 0.66). Use of ChEIs or investigational medications within 30 days of screening and any previous treatment with memantine were not permitted. Dosing began at 5 mg/day and increased by 5 mg/day each week, reaching a final dose of 20 mg/day (10 mg b.i.d.). Primary outcomes were the Alzhe- imer’s Disease Assessment Scale – Cognitive Subscale (ADAS-cog), which measures cognitive function, and the Cli- nician’s Interview-Based Impression of Change Plus Caregiver Input (CIBIC-plus), which provides a global assessment. Sec- ondary outcomes included the Neuropsychiatric Inventory (NPI), which measures behavioral symptoms, and the Alzhe- imer’s Disease Cooperative Study Activities of Daily Living Inventory (ADCS-ADL23), which assesses daily functioning. The ADAS-cog, CIBIC-plus and NPI assessments all showed statistically significant results supporting memantine at end point according to LOCF analyses. Only CIBIC-plus demon- strated statistical significance according to the OC analysis. ADCS-ADL23 did not show a statistically significant differ- ence between groups at week 24. There were no statistically significant differences in adverse events between the treatment groups. However, there was a trend for a greater rate of dis- continuation due to an adverse event in the memantine group than in the placebo group (9.5 versus 5.0%; p = 0.09). The only adverse event that was present in 5% of the memantine group, with an incidence of at least twice that for the placebo group, was somnolence (7 versus 1%, respectively; p = 0.02). There were no clinically meaningful differences between the groups’ changes in lab test values, vital sign measurements or electrocardiogram results. However, there was a potentially sig- nificant difference in postbaseline blood urea nitrogen levels (p = 0.23). Overall, the results from the study suggested that the use of memantine in the treatment of mild-to-moderate AD is both effective and safe [57].
Study 99679 was a 24-week study conducted to evaluate the use of memantine in the treatment of mild-to-moderate AD [58,104]. In this study, 470 participants were randomized in a 2:1 memantine-to-placebo ratio, and 409 participants fully completed the study. The MMSE range was 11 – 23, and base- line mean MMSE was 18.7. Participants who were previously treated with ChEIs were required to have discontinued ChEI treatment for at least 30 days prior to enrolling. Dosing of memantine began at 5 mg/day, and was up-titrated weekly by 5-mg increments, reaching the final dose of 20 mg/day (10 mg b.i.d.). The primary outcomes used for efficacy analysis were the change in ADAS-cog score from baseline to week 24, and the week 24 CGIC-plus total score. The memantine group scored slightly better on both measures compared with placebo, although neither difference was statistically significant. It was noted that during the course of the study, the placebo group did not decline as anticipated and instead improved on both meas- ures at certain data points. Results from the safety analysis have not been reported. However, there was a higher discontinuation rate due to adverse events in the memantine group (9%) than in the placebo group (4%) [58,104].
3.5.3.2 Combination therapy
The MEM-MD-12 study was a 24-week trial that required concurrent treatment with one of the ChEIs. Of the 433 randomized participants, 427 completed at least one postbaseline assessment, and 385 continued until the end of the study. The MMSE score range was 10 – 22, and the mean scores at baseline were 16.7 for the memantine group and 17.0 for the placebo group. Participants were required to have been receiving treatment with a ChEI for at least 6 months, with the last 3 months being at a stable dose. The initial dose of the study drug was 5 mg/day, which was increased each week by 5 mg/day, reaching a final dose of 20 mg/day given once at night. This was the first randomized, controlled trial of memantine in which the treatment was given in a single daily dose. The primary measures used were the change in ADAS-cog and CIBIC-plus scores from baseline to week 24. Neither measure elicited statistically significant differences using the initially planned analyses; however, both measures displayed numerical advantages in favor of the use of meman- tine. Furthermore, when considering differences due to the type of ChEI and the duration of treatment, additional analyses revealed statistically significant findings in support of memantine. For example, significant differences between memantine treatment groups were demonstrated in ADAS-cog and CIBIC-plus at certain points other than at week 24 for participants who were treated with donepezil, but not for participants treated with the other ChEIs. Another post hoc analysis examined differences when the treatment groups were stratified by duration of prior ChEI treatment. A statistically significant difference (LOCF: p = 0.032) in ADAS-cog was found within the group treated for a shorter duration (< 76 weeks), whereas there was no significant differ- ence found in the group treated for a longer duration ( 76 weeks). There were no significant differences in adverse events, and no clinically significant differences between the groups’ laboratory findings, vital sign measurements and electrocardiogram results. Overall, memantine given as a single dose of 20 mg/day in addition to ChEI treatment was found to be safe and well tolerated (Porsteinsson AP, Grossberg GT, Mintzer J, Olin JT: Memantine treatment in patients with mild-to-moderate Alzheimer’s disease already receiving a cholinesterase inhibitor: a randomized controlled trial. Manuscript in preparation).
3.5.3.3 Meta-analyses
One meta-analysis of memantine in the treatment of mild-to-moderate AD combined the results from the two monotherapy trials to form a pooled population. The authors used the intent-to-treat population of the MEM-MD-10 study and the full analysis set (FAS) of the Study 99679 population to form the overall pooled FAS group, which had 506 participants in the memantine group and 349 in the placebo group. The mean MMSE at baseline for the pooled analysis was 18.1. Statistically significant differences favoring memantine were found on LOCF and OC analyses of ADAS-cog and CIBIC-plus at the end point, as well as at weeks 12 and 18, using both parametric and non-parametric methods of analysis. The memantine group performed better numerically on the NPI at weeks 12 and 24. However, the difference between groups was statistically significant only at week 12 and not at the end point. The authors also performed a responder analysis, defining the single-criterion response as an improvement in the ADAS-cog by 4 points or a CIBIC-plus score of 4, and defining dual criterion as the accomplishment of both of the sin- gle criteria. From week 12 until end point, the response rate in the memantine group was statistically significantly greater than that in the placebo group for each of the defined single and dual criteria, which lends support towards a clinical significance of the benefits associated with memantine treatment [104].
Another meta-analysis of studies targeting mild-to-moder- ate AD combined data from the three studies outlined above: MEM-MD-10, 99679 and MEM-MD-12. This analysis shows small, but statistically significant, advantages of memantine over placebo on the CIBIC-plus (0.13 points; p = 0.03) and the ADAS-cog scores (0.99 point; p = 0.01) in the intent-to-treat population. No statistically significant effects were found on measures of behavior or function. There were also no significant differences between the dropout rates, percentage of patients experiencing at least one adverse event or number of patients experiencing agitation in the two groups [58].
A third analysis incorporated data from six studies of memantine in the treatment of AD, including studies MEM-MD-01, MRZ-9605, MEM-MD-02, MEM-MD-10,
99679 and MEM-MD-12. Statistically significant and clini- cally meaningful results favoring memantine were found on measures of cognition (p < 0.001), global assessment (p < 0.001) and function (p < 0.01) using LOCF and OC methods of analysis. The effect sizes attributed to memantine were said to be comparable to those found with ChEIs for each of these measures. Memantine was also associated with the slowing of clinical decline, as 11% of memantine-treated participants and 21% of placebo-treated participants experi- enced marked clinical worsening in all measures of cognition, global assessment and function (p < 0.0001) [104].
3.6 Safety and tolerability
.
Data from preclinical trials, pharmacokinetic studies, pivotal trials and postmarketing studies suggest that the combined use of memantine and ChEI is safe and well tolerated [59-61]. No significant pharmacokinetic interaction between memantine and donepezil was found in one in vivo study, further support- ing the safety of combination therapy with these two compounds [62]. The completion rates for the three mild-to-moderate AD studies were all relatively high for partic- ipants in the memantine treatment groups (85.0 – 89.4%), lending additional support towards the tolerability of memantine [57,58] (Porsteinsson AP, Grossberg GT, Mintzer J, Olin JT: Memantine treatment in patients with mild-to-mod- erate Alzheimer’s disease already receiving a cholinesterase inhibitor: a randomized controlled trial. Manuscript in preparation). The most frequently encountered adverse events from the MEM-MD-10 and MEM-MD-12 studies are shown in Table 2. Specific adverse events experienced by participants in Study 99679 were not indicated, with the exception of agita- tion, which had a greater incidence rate in the placebo group than in the memantine group. Table 2 includes any adverse events experienced by 5% of the memantine group when the incidence of the adverse event for the memantine group was 1% above that for the placebo group.
Of the listed adverse events, confusion was the only symptom meeting these criteria in both the MEM-MD-10 and MEM-MD-12 studies. Somno- lence (MEM-MD-10) was the only symptom that was present in the memantine group at a frequency that was at least twice as great as the frequency experienced by the placebo group. According to a cumulative report of adverse events experienced in clinical studies of memantine in mild-to-severe dementia, headache, somnolence, constipation and dizziness were the most commonly encountered adverse reactions, occurring in < 1/10, but > 1/100 individuals. Fatigue, confusion, hallucina- tions, vomiting, gait abnormalities and seizures were also associ- ated with memantine use, but they were ranked as uncommon (> 1/1000 and < 1/100) or very rare (< 1/10,000) in occurrence. It was noted that the overall incidence rate of adverse events was similar between treatment groups, and that the severity of the adverse events was usually mild-to-moderate [57,58,104] (Porsteinsson AP, Grossberg GT, Mintzer J, Olin JT: Meman- tine treatment in patients with mild-to-moderate Alzheimer’s disease already receiving a cholinesterase inhibitor: a randomized controlled trial. Manuscript in preparation).
Questions have been raised regarding the safety of meman- tine in individuals with renal impairment, as clearance depends on renal function and the metabolism of memantine is minimal. To address this question, one study analysed the predicted steady-state plasma concentration of memantine in groups of participants with renal function that was either normal (creatinine clearance of > 80 ml/min) or impaired; mild, moderate or severe renal impairment (creatinine clear- ances of 50 – 80, 30 – 49 and 5 – 29 ml/min, respectively). Each group contained eight participants. The predicted plasma concentrations in the normal and mild-impairment groups were comparable, but concentrations were elevated in participants with moderate (60%) or severe (115%) impairment. Based on these results, it was suggested that > 50 countries worldwide [108]. Applications have been submitted to extend the indication of memantine to include the treatment of mild AD, but neither the FDA nor EMEA have granted approval for this indication [109-111]. Memantine has not been approved for the treatment of other forms of dementia.
4. Conclusion
Preclinical data have demonstrated support for the hypoth- esis that memantine elicits therapeutic effects by improving cognition and learning in models of impairment. It also has been shown to provide neuroprotection against several forms of neuronal insult that potentially lead to irreversible damage. The evidence shows that memantine may indeed be effective at both compensating for present deficiencies and delaying further damage. Clinical studies have sup- ported the tolerability of memantine in participants with AD, and have further endorsed the evidence from preclini- cal research by showing treatment effects in favor of patients with mild or moderate renal impairment could tolerate the typical dosing regimen of 10 mg b.i.d., but that patients with severe renal impairment should modify meman- tine dosing to 5 mg b.i.d [63]. FDA regulations have endorsed this suggestion, as FDA-monitored labeling indicates that patients with severe renal impairment (creatinine clearance of 5 – 20 ml/min) should limit the target dose to 10 mg/day (5 mg b.i.d.) [101]. Memantine labeling monitored by the EMEA suggests that patients with moderate renal impairment (creatinine clearance of 40 – 60 ml/min/1.73 m2) should reduce dosing to 10 mg/day, and that patients with severe renal impairment (creatinine clearance of < 9 ml/min/1.73 m2) should refrain from using memantine treatment [105].
Studies MEM-MD-10 and 99679 have progressed into ongoing open-label extension trials: MEM-MD-11 and Study 99819. Study MEM-MD-11 is expected to continue for a total of 132 weeks, and Study 99819 is expected to continue for 3 years [104]. It is hoped that the results from these two long-term trials will reveal valuable information regarding the efficacy and safety of memantine in the treatment of mild-to-moderate AD over a longer duration of treatment.
3.7 Regulatory affairs
Memantine was authorized in Europe by the EMEA for the treatment of moderately severe-to-severe AD in 2002. In 2005, the indication was extended in Europe for the memantine, especially in study populations with greater levels of impairment. The effect of memantine found in studies of mild-to-moderate AD was not as great as that seen in studies of moderate-to-severe AD, and sometimes the difference found did not reach significance. Stronger support for the efficacy of memantine in the treatment of mild-to-moderate AD is needed before memantine is approved for use with this particular patient population. It is possible that in this population, any treatment effect may not be detectable over a shorter duration of time, and that future studies should incorporate longer placebo-controlled phases. Further information regarding long-term safety and tolerability will be available at the completion of two long-term, open-label studies of memantine in the treatment of mild-to-moderate AD.
5. Expert opinion
Memantine is an acceptable and approved option for the treatment of moderate-to-severe AD. It has shown modest benefit when used in this patient population and, overall, is safe and well tolerated. Based on preclinical data, memantine has the potential to alter the progression of AD. Its protective effects against A toxicity and other sources of neuronal insult implies the capacity to actually prevent or delay future decline, rather than simply being therapeutic in nature while compensating for present deficits. This characteristic makes memantine an attractive possibility for use earlier on in the course of a neurodegenerative disease, such as AD. The challenge is to establish that these properties translate into clinically meaningful effects.
Efforts have been made to gain approval for the use of memantine in the treatment of mild AD; however, applica- tions thus far have been denied. Moderate treatment effects have been demonstrated in measures of cognition, global assessment and behavior in this patient group. However, the significance of these effects has not been consistent across studies. The duration of randomized, controlled trials of memantine in patients with mild-to-moderate AD has been similar to the duration used in earlier trials of memantine in patients with moderate-to-severe AD. Although it is typical for patients with AD to decline in areas of cognition, behavior and function, regardless of the stage of the disease they are in, it is not atypical for the rate of decline to be steeper later in the course of the disease. Perhaps a longer study period is necessary to detect a treatment effect when assessing the potential for neuroprotection and course-modifying impact in the earlier stages of dementia. Similarly, the assessments that have been used in AD clinical trials might not have the sensitivity that is required to accurately depict subtle changes that occur earlier in the course of the disease. Thus, a longer duration of treatment and more sensitive measures of assessment may improve the detection of clinical efficacy, if any, provided by memantine in patients with early AD.
Although memantine is not approved for the treatment of mild AD by regulatory agencies, it seems to be safe and well tolerated in this population. As treatment options are limited,some clinicians have relied on preclinical data (that suggests that memantine may have broad, short- and long-term effects on A toxicity and neuronal insult), as well as on the data from clinical trials to support the use of memantine in early AD and vascular dementia. At this point, we do not know for certain whether early treatment in patients with mild AD or mild cognitive impairment may significantly delay the onset of moderate or severe deficits. It is possible that any beneficial outcome of the protective effects may not be clinically or statistically detectable until months or even years after the ini- tiation of treatment. Further evidence of the efficacy of memantine in mild AD and other forms of dementia is needed before memantine might gain approval for the treatment of these indications.
Conflict of interest
AP Porsteinsson has received grants/research support from Abbott, AstraZeneca, Bristol Myer-Squibb, Eisai, Elan, Eli Lilly, Forest, Janssen, Merck, Mitsubishi, Myriad Neuro- sciences, Neurochem, Novartis, Ono Pharma, Pfizer and Sanofi. AP Porsteinsson is a Consultant for, or on the advis- ery board of, Abbott, AstraZeneca, Bristol Myer-Squibb, Eisai, Forest, Janssen, Novartis, Organon and Pfizer. AP Porsteinsson is on the speakers’ bureau for Abbott, Astra- Zeneca, Bristol Myer-Squibb, Eisai, Janssen, Novartis, Organon and Pfizer.