Conflict of Interest: The study was conducted with an independent research grant from Forrest Laboratories. Dr. Bigal is now a full time employee of Merck Research Laboratory.
Memantine in the Preventive Treatment of Refractory Migraine
Article first published online: 8 APR 2008
© 2008 the Authors. Journal compilation © 2008 American Headache Society
Headache: The Journal of Head and Face Pain
Volume 48, Issue 9, pages 1337–1342, October 2008
How to Cite
Bigal, M., Rapoport, A., Sheftell, F., Tepper, D. and Tepper, S. (2008), Memantine in the Preventive Treatment of Refractory Migraine. Headache: The Journal of Head and Face Pain, 48: 1337–1342. doi: 10.1111/j.1526-4610.2008.01083.x
- Issue published online: 3 OCT 2008
- Article first published online: 8 APR 2008
- Accepted for publication January 22, 2008.
- refractory migraine;
Objectives.— To assess the efficacy and tolerability of memantine (MEM) in the preventive treatment of refractory migraine.
Background.— Glutamate is of importance in migraine pathophysiology and may be related to progression from episodic to chronic mirgraine. Furthermore, individuals with chronic pain often report cognitive problems. MEM has the potential to address both issues, justifying this pilot study.
Methods.— We included subjects with refractory migraine (episodic migraine with 8-14 days of headache per month or transformed migraine, who had previously failed at least 2 trials of adequate preventive therapy). Other preventive drugs were allowed if the patient had been on a stable dose for more than 30 days. MEM dose ranged from 10 mg to 20 mg per day. The treatment phase lasted 3 months. The primary endpoint was number of days with headache at month 3. Cognitive performance was assessed with the trail making tests A and B (TMT-A and B). Statistical analyses were performed on the intent-to-treat (ITT) population, using data subjected to the last observation carried forward algorithm. We also conducted per protocol analyses.
Results.— In the ITT population (n = 28), monthly headache frequency was reduced from 21.8 days at baseline to 16.1 (P < .01) at 3 months. The mean number of days with severe pain was reduced from 7.8 to 3.2 at 3 months (P < .01). The mean disability scores were significantly reduced at 3 months, compared with baseline (36.6 vs 54.9, P < .01). There was a significant reduction in the time to complete TMT-A at termination vs baseline (28.4 vs 23.2, P = .02) and also TMT-B (70.1 vs 50.4, P = .04). Side effects were present in 37.5% of the patients; 5.5% dropped out the study because of poor tolerability. Most adverse events were mild.
Conclusion.— This study offers preliminary evidence for the use of MEM in the prevention of refractory migraine. Double-blind studies are now required.
Many migraineurs who seek care in headache clinics are refractory to treatment.1-3 Refractoriness is defined based on the chronicity, frequency, and severity of the headaches, as well as on subjects experiencing less than expected benefit from standard therapies.4,5 Defining refractory migraine (RM) has been the subject of a great deal of interest.4,6
The pharmacological treatment of RM poses a challenge to the physician. Most preventive agents used in this context have not been examined specifically for the treatment of this syndrome.7-9 Medications for RM are used empirically based on their efficacy in the treatment of episodic migraine,7,8 and patients are often treated with multiple drugs.10 As a consequence, side effects, poor compliance, and disappointing outcomes are common. Cognitive symptoms also emerge frequently in this context, either as a consequence of chronic pain11,12 or of the medications used to treat RM.13,14
Glutamate has been mentioned prominently in theories of migraine pathophysiology.15,16 Glutamate is implicated in cortical spreading depression, trigeminovascular activation, central sensitization, and may be linked to migraine chronification. Recent studies suggest that the cerobrospinal fluid levels of glutamate are reduced in individuals with chronic migraine, and further reduced in individuals overusing medication.17
Memantine (MEM) is the first in a novel class of Alzheimer's disease medications acting on the glutamatergic system. MEM is a moderate-affinity voltage-dependent noncompetitive antagonist at glutamatergic N-methyl-D-aspartate (NMDA) receptors.18 By binding to the NMDA receptor with a higher affinity than magnesium (Mg2+) ions, MEM is able to inhibit the prolonged influx of calcium (Ca2+) ions associated with neuronal excitotoxicity.
Because glutamate is probably important in migraine pathophysiology, and individuals with RM frequently complain of cognitive problems, either as a function of the disease itself or as a side effect of medications, herein we conducted a pilot study to prospectively assess the efficacy and tolerability of MEM in the preventive treatment of RM. We also aimed to screen the influence of MEM on neurocognitive performance.
This was a prospective, open-label study conducted in a tertiary headache center from 2006 to 2007, where we consecutively enrolled subjects who agreed to participate. Participants were selected among patients attending the headache center for scheduled visits. Inclusion and exclusion criteria were:
- 1RM, defined as follow:
- A. At least one of i or ii:
- i. Episodic migraine with headaches happening on 8-14 days per month.
- ii. Transformed migraine according to the criteria proposed by Silberstein and Lipton.19
- B. Previous failure to at least one standard migraine preventive medication, used in adequate doses for at least 3 months.
- 2No previous use of MEM.
- 3Using preventive medication on a stable regimen for at least one month.
- 4No major depression.
The baseline observation period consisted of one month, and information was prospectively collected using headache calendars. After the baseline period, MEM was started. Therapeutic doses ranged from 10 mg to 20 mg. The initial dose was 10 mg. If after one month the participant had no side effects and had empirical perception about incomplete satisfaction, dose was forced to 20 mg. The treatment phase lasted 3 months. Patients were reevaluated in person monthly.
Headache information was collected using daily headache calendars. Disability was assessed using the Migraine Disability Assessment (MIDAS) questionnaire. Screening for depression was conducted with the PRIME-MD questionnaire.
Cognitive performance was assessed with the trail making test (TMT) parts A and B. In TMT-A, after a training test, individuals are timed when connecting numbers from 1 to 25 in crescent order. TMT-A assesses spatial scanning, psychometric speed, and praxis.20 In TMT-B, individuals connect numbers and letters in alternating pattern. In addition to the domains assessed by TMT-A, it also assesses retention.21 TMT is a standardized set that is heavily influenced by attention, concentration, resistance to distraction, and cognitive flexibility (or set-shifting).20,21 It has been used to measure the influence of drugs such as atorvastatin and several anti-epileptic drugs on cognition.22,23
The primary endpoint was number of headache days (headache frequency) after 3 months, compared with the baseline period. Secondary endpoint was the number of days with severe headache. Other endpoints included MIDAS score and assessments at 1 and 2 months. We also contrasted the monthly scores of the TMTs.
Statistical analyses were performed on the intent-to-treat (ITT) population, using data subjected to the last observation carried forward (LOCF) algorithm. We also conducted per protocol analyses (PPA) in those who completed the entire study (completers). To compare several time points, we used one-way analysis of variance after normality test. Two groups were compared with the paired T-test, assuming a 5% 2-tailed significance level. We powered our study to capture a 20% change in the primary endpoint, at the 5% 2-tailed significance level, using the paired T-test. Based on the power assumption, we planned an ITT population of 30 individuals.
This study and all the forms were approved by an Investigational Review Board.
Of 38 subjects enrolled, 28 completed the baseline assessment and were given drug, consisting of our ITT population (75% females, mean age of 43.5 years). A total of 23 (82.1%) subjects completed the entire study; 2 (7.1%) dropped out the study because of side effects; 2 (7.1%) withdrew the consent; and one (3.6%) was lost to follow up (Fig. 1). A total of 8 (28.5%) had episodic migraine, while the remaining had TM. No patients had pure episodic or chronic tension-type headache, although superimposed tension-type attacks were allowed. Headaches fulfilling tension-type criteria happened on an overall mean of 10.3 days per month (SD = 7.1). Onset of headaches happened a mean of 12.3 (SD = 12.9 years) before enrollment in the study. All subjects were using other preventive medications. A total of 12 (42.8%) were using one preventive drug, while 51.8% were using 2 or more. They were on stable doses of medication for a minimum of 3 months (Mean = 4.1, SD = 1.3).
ITT Analyses.— Among the 28 patients, 19 (67.8%) used 10 mg, while 9 (32.1%) used 20 mg. At baseline, patients had a mean of 21.8 headache days per month. A significant reduction in headache frequency was observed in 2 months (15.4, P < .01) and 3 months (16.1, P < .01). The reduction at 1 month was not significant (19.0) (Fig. 2).
The mean number of days with severe pain was reduced from 7.8 at baseline to 5.9 after 1 month (NS), 4.0 at 2 months (P < .01), and 3.2 at 3 months (P < .01) (Fig. 3).
The mean MIDAS scores were significantly reduced at 3 months, compared with baseline (36.6 vs 54.9, P < .01). The headache score at baseline was 44, and was significantly reduced at day 90 (27.9, P = .01).
PPA.— For the PPA, we correlated endpoints at baseline and upon completion (Table 1). Mean headache frequency was reduced from 21.5 at baseline to 14.3 days per month at the end of the study (P < .01). We also found a significant difference in number of days with severe headache per month, comparing baseline and termination (7.5 vs 3.0, P < .01), headache score (44.3 vs 27.1, P < .01), and MIDAS scores (52.0 vs 31.3, P = .01).
|Endpoint||Baseline Mean (SD)||3 months Mean (SD)||P value|
|Headache frequency||21.5 (4.1)||14.3 (4.8)||<.01|
|Severe headaches||7.5 (3.3)||3.0 (2.7)||<.01|
|Headache score||44.3 (19.2)||27.1 (15.0)||<.01|
|MIDAS scores||52 (22.4)||31.3 (17)||=.01|
TMT Assessment.— As mentioned in the methods section, TMT-A measures visual space scanning and psychometric speed. There was a significant reduction in the time to complete TMT-A at termination vs baseline (28.4 vs 23.2, P = .02). TMT-B also assesses memory and praxis. Time to complete TMT-B was also decreased at termination, compared with baseline (70.1 vs 50.4, P = .04) (Fig. 4).
Tolerability.— Side effects were reported by 10 patients (35.7%) and 2 (7.4%) dropped out the study because of poor tolerability. We did not observe serious side effects. The most common side effects were somnolence (7 patients, 25%), asthenia, and anxiety (3 patients each, 10.7%). Increased weight (n = 3, 10.7%), depression (n = 2, 7.1%), and a subjective complain of emotional instability (n = 2, 7.1%) also occurred. The following side effects were reported by one patient each: constipation, vertigo, and imbalance. One patient had a cholecystectomy during the study, considered to be not related to study drug.
There were no significant changes in the blood pressure (systolic, 119.5 mmHg at baseline, 118.8 at termination; diastolic: 75.45 mmHg at baseline, 75.41 at termination), pulse rate (69.7 vs 70.1 per minute), and mean weight (147.9 pounds at baseline vs 148.1 at day 90).
Defining RM has emerged as an important issue in headache research.4 Standardized definition of RM is necessary to identify individuals in need of more aggressive therapies, as well as for conducting clinical trials to justify these therapies.
As previously stated, we decided to study MEM in the preventive treatment of RM based on the glutamatergic dysfunction that may occur in severe headache syndromes, as well as on the cognitive problems that are common in this population.15,16,24
Migraine pain-relay centers, including the trigeminal ganglion, trigeminal nucleus caudalis, and thalamus, contain glutamate-positive neurons, and glutamate activates the trigeminal nucleus caudalis.24,25 Glutamate is also implicated in cortical spreading depression and central sensitization.25 Glutamate receptor-subtype antagonists are effective in preclinical models of migraine.26
It has been suggested that chronic pain can be maintained by a state of sensitization within the central nervous system that is mediated in part by glutamate and aspartate binding to the NMDA receptor. A number of antagonists to the NMDA receptor are antinociceptive in animal models but are associated with significant dose-limiting side effects.24-26 Commercially available NMDA-receptor antagonists include ketamine, dextromethorphan, MEM, and amantadine.
Although glutamate dysfunction has been associated with excitotoxic neuronal cell damage mediated in part by over activation of NMDA receptors, physiological NMDA receptor activity is essential for normal neuronal function.26,27 This means that potential neuroprotective agents that block virtually all NMDA receptor activity will very likely have unacceptable clinical side effects. For this reason, many previous NMDA receptor antagonists have disappointingly failed advanced clinical trials for a number of disorders.27 In contrast, studies suggest that MEM preferentially blocks excessive NMDA receptor activity without disrupting normal activity. MEM does this through a noncompetitive, low-affinity, open-channel blocker; it enters the receptor-associated ion channel preferentially when it is excessively open, and, most importantly, its off-rate is relatively fast so that it does not substantially accumulate in the channel to interfere with subsequent normal synaptic transmission.28
Many patients with chronic pain complain about memory disturbances, and cognitive side effects are common during standard migraine therapies.11 Because MEM has been studied in patients with severe memory problems, using this drug experimentally was further justified.
In this open-label, exploratory trial, we found that MEM may be an effective drug in the prevention of RMs, pending controlled studies. Our findings support a prior pilot study.29 We reached clinical significance in the primary and secondary endpoints, using both ITT and PP analyses. The magnitude of the effect was certainly clinically relevant. Furthermore, patients significantly improved over the study in the domains measured by the TMT. Finally, the drug was well tolerated. Our study is similar to another pilot study that has been conducted independently (Andrew Charles, personal communication30).
Several precautions should be considered when analyzing the implications of our results. First, although this is a prospective study, it is not controlled by placebo. It is a pilot, open-label study. Second, we included RM with and without medication overuse. Those overusing medication had failed detoxification protocols in the past. No patient stopped medication overuse during our study. Because studies show that preventive medication is often ineffective in patients overusing acute medication,10 we may have underestimated the benefits of MEM by including over users. Finally, in the ITT primary endpoint, reduction in headache frequency was greater at 2 than at 3 months. The differences were not significant and this may be a consequence of our small sample size.
The results of this study indicate that MEM is a promising drug for the treatment of RM, although it has neither been submitted to nor approved by the Food and Drug Administration for this use. A necessary future step forward is to conduct a double-blind placebo-controlled study of MEM in the treatment of RM.
- 3Chronic daily headache in a tertiary care population: Correlation between the International Headache Society diagnostic criteria and proposed revisions of criteria for chronic daily headache. Cephalalgia. 2002;22:432-438., , , , .
- 9Levetiracetam as prophylaxis for resistant headaches (abstract). Cephalalgia. 2001;21:373..
- 20Neuropsychological impairment scale (NIS): Initial validation study using trailmaking test (A & B) and WAIS digit symbol (scaled score) in a mixed grouping of psychiatric, neurological, and normal patients. J Clin Psychol. 1983;39:746-748., , .
- 23Long-term cognitive and mood effects of zonisamide monotherapy in epilepsy patients. Epilepsy Behav. 2007; [in press., , , , , .
- 29Memantine for the prophylaxis of migraine: A report of 3 cases and discussion of pharmacology. Headache. 2005;16:167-170., , .