Overview. A consequence of Alzheimer’s disease is neuronal loss, the result of a cascade of complex intracellular events. Excitotoxic reactions due to excessive glutamate release are thought to contribute to this neuronal cell death. Blocking glutamate receptors (NMDA receptors) is one approach to reducing cell death. In the past, however, drugs targeting NMDA receptors have been associated with toxicity and significant side effects, and many such drugs have failed during clinical development.
Mechanism of Action. NMDA receptor antagonists function as neurotransmitter modulators; in theory, their activity should improve cognitive symptoms of Alzheimer’s disease by reducing the effects of excessive glutamate levels that occur during intracellular stress in Alzheimer’s disease patients. Glutamate binds to the NMDA receptor expressed on the postsynaptic neuronal membrane. The amino acid glycine binds to a separate site on the NMDA receptor and acts as a coagonist to glutamate to activate the receptor and open its coupled calcium ion channel, allowing an influx of calcium ions (Ca++) into the postsynaptic neuron.
Under normal conditions, presynaptic neurons prohibit excessive glutamate release via negative feedback mechanisms, but such mechanisms are compromised under conditions of cellular stress such as in Alzheimer’s disease. In addition, beta-amyloid (Aβ) plaques found in the synaptic clefts of Alzheimer’s disease patients are believed to increase glutamate levels released from the presynaptic neuron upon depolarization and to inhibit its uptake by glial cells ().
Excess glutamate present in the synaptic cleft of Alzheimer’s disease patients causes postsynaptic calcium channels to be continually opened. Prolonged elevation of intracellular calcium levels within neurons triggers a series of concurrent biochemical pathways — the end points of these pathways are severe cellular damage and/or cellular death. Major steps along these pathways include the activation of calcium-activated proteases (e.g., calpain), the loss of membrane phospholipases, and the generation of free radicals. Moreover, continually opened calcium channels cause a continuous low level of activation (depolarization) in the post-synaptic neuron, analogous to a low level of background electrical noise. This background activation masks bona fide glutamate bursts from the presynaptic neuron as a result of signaling to the postsynaptic neuron; these signals are required for learning and memory formation.
Memantine (Merz’s Axura, Lundbeck’s Ebixa, Forest’s Namenda) () was originally launched by Merz under the brand name Akatinol in Germany in 1989 for the treatment of cognitive insufficiencies, including Alzheimer’s disease. Currently, Merz markets the drug for dementia in Germany under the brand name Axura. Since the drug’s launch, Merz has entered licensing agreements with Forest Laboratories, Lundbeck, and Suntory for the marketing of memantine in the United States, Europe, and Japan, respectively. In 2002, the EMEA approved memantine’s use for moderately severe to severe Alzheimer’s disease, and Lundbeck launched the drug under the name Ebixa at the end of that year. Memantine is available in Europe and the United States (but not Japan) for the treatment of moderate to severe Alzheimer’s disease. The drug is in Phase lib trials in Japan for moderate to severe Alzheimer’s disease, and Suntory plans a 2006 launch. In October 2004, Lundbeck filed with the EMEA for the approval of memantine for mild to moderate Alzheimer’s disease in the European Union. In October 2005, European approval was granted for memantine in moderate but not mild Alzheimer’s disease. In September 2004, Forest submitted an sNDA to the FDA for approval of memantine for mild to moderate Alzheimer’s disease in the United States. The sNDA was accepted in November 2004. In July 2005, however, Forest announced that the FDA had issued a non-approvable letter in response to the sNDA for memantine in mild Alzheimer’s disease (Forest Laboratories press release, July 25, 2005). The FDA had informed Forest that a single positive study in patients with mild to moderate Alzheimer’s disease would be adequate to support extending memantine’s labeling for patients with mild Alzheimer’s disease; Forest submitted data from a six-month study of mild to moderate Alzheimer’s disease, in which the efficacy of memantine versus placebo reached statistical significance, but also from additional studies in mild to moderate Alzheimer’s disease in which memantine performed numerically — but not statistically significantly — better than placebo on primary end points. Data from these studies influenced the FDA’s decision to issue the non-approvable letter for Forest’s sNDA in mild Alzheimer’s disease. In July 2005, Forest announced it would meet with the FDA to further discuss the non-approvable letter.
In Japan, Suntory is planning supportive studies and anticipates launching memantine for mild to moderate Alzheimer’s disease in 2008. Memantine is also in Phase III trials for vascular dementia, diabetic neuropathies, and glaucoma and Phase II trials for neuropathic pain.
For its original U.S. regulatory submission for treatment of moderate to severe Alzheimer’s disease, Forest Laboratories submitted an NDA to the FDA in August 2002 but voluntarily withdrew it in September 2002 to include new data from a Phase III Alzheimer’s disease trial that evaluated a combination therapy of donepezil and memantine. Upon resubmission in 2003, the Peripheral and Central Nervous System Drugs Advisory Committee unanimously recommended memantine to the FDA for approval for moderate to severe Alzheimer’s disease despite the drug’s failure to reach statistical significance in one of the three trials under review. The approval underlines the high level of unmet need in this area — there are no other drugs approved for moderate to severe Alzheimer’s disease — and the benign clinical safety profile of memantine.
Memantine is a moderate-affinity, noncompetitive, NMDA receptor antagonist. Because it is voltage dependent, memantine can enter only open NMDA receptors to exert its blocking effects; therefore, it exclusively inhibits overly active receptors, thus lowering the background noise generated in postsynaptic neurons by excess glutamate in the synaptic cleft of Alzheimer’s disease patients (). Thus, an electrical signal necessary to initiate learning and memory formation can be detected in the presence of memantine above the background electrical noise, much like a person can hear a soft voice over television static once the static noise has been turned down.
Memantine effectively functions as a noncholinergic neurotransmitter modulator, an effect that should improve the cognitive symptoms of Alzheimer’s disease. While many NMDA receptor antagonists have shown toxicity in later-phase studies, it is probably the low to moderate affinity for the NMDA receptor that both reduces memantine’s toxicity and contributes to its moderate efficacy. In animal models, the drug demonstrates neuroprotective qualities, presumably by protecting neurons from damage associated with excessive stimulation by glutamate without altering glutamate’s role in normal brain function (). Furthermore, memantine is reported to restore normal tau hyperphosphorylation in vitro (), an effect that may protect against neurofibrillary degeneration.
A Phase III trial demonstrated that memantine is able to enhance cognition and reduce functional decline in patients suffering from severe dementia (). A 28-week, placebo-controlled, double-blind, parallel-group study randomized 252 patients with moderate to severe Alzheimer’s disease (MMSE scores, 3-14) to receive memantine or placebo. Primary end points were the CIBIC-plus global score and the Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory modified for severe dementia (ADCS/ADLsev). The CIBIC-plus scores failed to reach significance for efficacy (mean difference of 0.3; p = 0.06), although Forest argues that a higher dropout rate in the placebo group (34 versus 21 dropouts) may have skewed results in favor of placebo. Re-evaluation of the data by including only the participants who remained in the trial to 28 weeks showed a mean difference of 0.3, p = 0.03. More significantly, this trial showed that patients treated with memantine (10 mg twice daily) were better able to participate in ADLs, as evidenced by a mean difference in ADCS/ADL scores of 2.1 points (p = 0.02) between treated patients and placebo recipients at the six-month end point. Secondary end points of the trial included the Severe Impairment Battery (SIB), which evaluates cognitive performance in advanced Alzheimer’s disease. The mean difference in SIB scores was 6.1 points (p < 0.001) for the memantine -treated group compared with the placebo group. Memantine was well tolerated with minimal side effects; there were no clinically relevant differences between patient groups in terms of side effects. However, more placebo patients than memantine-treated patients (22 versus 13) withdrew from the study.
Data also suggest that memantine is safe and provides improved efficacy as combination therapy with donepezil than donepezil as monotherapy. A 24-week study showed that a combination treatment of memantine and donepezil in moderate to severe Alzheimer’s disease (MMSE scores, 5-14) was more effective than donepezil monotherapy (). The randomized, double-blind, placebo-controlled, parallel group study included 400 patients treated with donepezil for the previous six months (and on stable doses of 5 or 10 mg/day for the three months before that). Patients were randomly assigned to receive either placebo or memantine in addition to donepezil; the memantine-treated group received 10 mg twice daily in addition to donepezil. The primary end points for the trial included the SIB and the ADCS/ADLsev. The study showed that improvements in cognition and daily function were significantly better in the combination-treated patients than in those treated with donepezil alone. Throughout the 24-week study, the SIB scores for patients treated with memantine/donepezil combination therapy were, to a statistically significant degree, better than the SIB scores of patients treated with donepezil monotherapy. Most importantly, the results showed that combination treatment improved cognitive scores (mean change from baseline score, 0.9) versus a decline in cognition in those patients treated with donepezil alone (mean change from baseline score, — 2.5). The ADCS/ADLsev scores for memantine-treated patients were also statistically significant at weeks 4, 8, and 24 of the study and showed less decline from baseline (mean change — 2.0 for combination treatment versus — 3.4 for donepezil monotherapy).
In anticipation of expanding the drug’s labeling, Lundbeck and Forest have completed placebo-controlled Phase III studies in Europe and the United States in patients with mild to moderate Alzheimer’s disease. Preliminary analysis of a six-month, placebo-controlled, double-blind study in Europe that followed 470 patients failed to demonstrate superiority of memantine (as measured by the ADAS-cog and the CIBIC-plus) over placebo at the study’s conclusion (Lundbeck, press release, January 7, 2004). However, statistical significance was reached at several interim time points. A second six-month, double-blind, placebo-controlled Phase III study conducted in the United States was designed to evaluate the efficacy of memantine as monotherapy (). The study, which followed 403 Alzheimer’s disease patients (MMSE scores, 10-22), demonstrated that patients receiving 10 mg of memantine twice daily performed significantly better on the ADAS-cog (p = 0.003) and the CIBIC-plus (p = 0.004) than patients receiving placebo.
According to the drug’s package insert, memantine is generally associated with mild or moderate adverse effects for doses up to 20 mg/day. Side effects include dizziness (7% of patients receiving memantine versus 5% of placebo recipients), headache (6% versus 3%), constipation (5% versus 3%), and hypertension (4% versus 2%). Side effects for patients receiving both memantine and donepezil include confusion (7.9% for treatment versus 2% for placebo), and headache (6.4% versus 2.5%) ().