Abstract
The objective of this randomized, placebo-controlled, double-blind study was to determine the effectiveness of intravenous magnesium sulphate and intravenous metoclopramide in the treatment of acute migraine attacks in the Emergency Department when compared with placebo. Adult patients who presented to the Emergency Department with a headache that met International Headache Society (IHS) criteria for acute migraine were infused with either 10 mg of intravenous metoclopramide, 2 g of intravenous magnesium sulphate or normal saline over 10 min. At 0, 15, and 30 min, patients were asked to rate their pain on a standard visual analogue scale. At 30 min, patients were asked in a standard manner about the need for rescue medication. Adverse affects were also recorded. Patients were followed up by telephone within 24 h for any recurrence after discharge. The primary endpoint of the study was the difference in pain relief between the groups at 30 min. Of the 120 patients who met IHS criteria, seven were excluded from the study due to insufficient data. The number of patients, gender, age and initial visual analogue scale (VAS) scores were comparable between groups. Each group experienced more than a 25-mm improvement in VAS score at 30 min. However, there was no significant difference detected in the mean changes in VAS scores for pain. The rescue medication requirement was higher in the placebo group. The recurrence rate in 24 h was similar between the groups. Although patients receiving placebo required rescue medication more than the others, metoclopramide and magnesium have an analgesic effect similar to placebo in migraine attacks.
Introduction
Headache is a common complaint among emergency department (ED) patients, representing up to 4% of all ED visits (1). While most patients suffer from primary headache syndromes, a physician has to exclude secondary and serious causes of headache that effect 3.8% of patients presenting with headache (2). Migraine is the second most common primary headache syndrome and is about three times more common in women than in men (3).
Many different treatment strategies have been used for acute migraine. Ergotamine and dihydroergotamine are effective and frequently used for pain relief in acute migraine but cause nausea and vomiting in a large number of patients (4). Sumatriptan is another specific and effective treatment but is associated with high cost, frequent but minor side-effects and high recurrence rates (5). Newer triptans are more effective, better tolerated and have lower recurrence rates, yet still are an expensive treatment choice and appear ineffective in 20–30% of patients (6, 7).
Despite side-effects such as drowsiness and rarely seen dystonic reactions, metoclopramide, a dopamine receptor antagonist, is a well-known antiemetic and also safe in pregnancy (8). It is widely used for promoting gastric motility and relieving nausea and vomiting. Like other dopamine receptor antagonists (9), metoclopramide has been studied in acute migraine attacks. It is also used in Europe as a combination with some analgesics for the treatment of acute migraine attacks (10).
Magnesium is an intracellular cation and has been found to have an important role in ischaemic heart disease (11) and arrhythmia (12) pathogenesis. It is commonly used for the treatment of eclampsia. Magnesium also has a role in neurochemical transmission, in the function of serotonin and regulation of vascular tone, which are thought to play important roles in migraine pathophysiology. Recently, magnesium sulphate (MgSO4) has been studied as a sole treatment agent or as an adjunctive medication in acute migraine attacks (13, 14). Only a few well-designed studies have investigated its efficacy in acute migraine attacks.
We designed a prospective, randomized, placebo-controlled study to determine the efficacy of intravenous metoclopramide and magnesium sulphate in the treatment of acute migraine headache in an ED setting. Our primary endpoint was the difference in pain scores between the groups.
Methods
The study was conducted in the ED of a university hospital that has 50 000 patients annually. A non-consecutive sample of patients who presented to the ED with headache, meeting International Headache Society criteria (15) for migraine, were included the study. The study protocol was reviewed and approved by institutional Ethics Committee and informed consent was obtained from all patients. Exclusion criteria were: patients < 18 years old, a known adverse reaction to metoclopramide or MgSO4, consumption of the study drugs within 48 h, a temperature of ≥ 38°C, altered mental status, meningeal signs, and renal or cardiac disease.
A computer-generated program was used for randomization. The study solutions, of identical appearance, were prepared and administered by a nurse who was not part of the study. Patients received 100 ml normal saline solutions with either 2 g of MgSO4, 10 mg metoclopramide, or none. Patients and treating physicians were blind to the study solution. The identification code was kept confidential until the code was broken at the end of the study.
Studies with MgSO4 in the treatment of headache have used various doses ranging from 0.5 to 2 g (14, 16–18). The safety and tolerability of MgSO4 have been well shown in studies on toxaemia of pregnancy and asthma at a dose range of 2–39 g. We preferred to use a dose of 2 g because it is the maximum dose shown to have an effect and is still in the lower limits of the tolerability and safety range.
Patients were asked to rate their pain on an unmarked 100-mm visual analogue scale (VAS) with endpoints of ‘none’ and ‘greatest’ pain. An intravenous (i.v.) line was placed by a nurse. The study solutions were infused over 10 min and afterwards the patients were asked again to rate their pain on the 100-mm VAS at 15 and 30 min after the beginning of the infusion. We preferred to select an outcome at 30 min because with the i.v. route the maximal absorption and distribution occurs within minutes and most patients who respond to treatment do so during this interval (19–21). At 30 min patients were asked by the treating physician for additional analgesic requirement with a standard question written on the study form. Patients needing rescue medication were treated for residual pain with 0.75 mg/kg meperidine. A repeat dose of meperidine was administered 30 min later if needed. Adverse events occurring within the ED were also recorded. Patients were discharged only after they stated they were pain free or did not need any more analgesic medication either after receiving rescue medication or not. Patients were followed up by telephone for any recurrence at 24 h after discharge.
The primary endpoint of the study was the difference in pain relief between the groups in VAS scores at 30 min. Secondary outcome measures were the difference in adverse reactions, need for rescue medications, and recurrence rate at 24 h between the groups.
A sample size of 35 patients in each group allowed the detection of a difference of 20 mm in VAS between groups with a power of 80% (α= 0.05; two-sided; SD = 25). Two-way repeated-measures analysis of variance and Pearson's χ2 tests were used for statistical analysis. Patients with migraine with aura (MA) were analysed as a subgroup. The Kruskall–Wallis test was used to analyse the percentage changes in VAS scores between baseline and 15 min (0–15) and baseline and 30 min (0–30) instead of repeated-measure analysis because of insufficient patient numbers in the MA group.
Results
Of the 120 patients enrolled in the study, seven were excluded due to incomplete records and because 24-h follow-up could not be performed. Patient demographics, baseline VAS scores and migraine types in the treatment groups are shown in Table 1. There was no difference in age, gender or baseline VAS scores between the groups (P > 0.05).
Patient numbers, demographic features, baseline visual analogue scale (VAS) scores and number of patients with migraine with aura in the treatment groups
SD, Standard deviation; MA, migraine with aura.
A significant reduction in pain intensity at 15 and 30 min compared with baseline was seen in all three groups (P < 0.000). This reduction in pain intensity was not different between the groups at the 15 and 30 min (P = 0.619) (Fig. 1).
Visual analogue scale (VAS) scores at 15 and 30 min of treatment. Changes were significant at 30 min in all groups (P < 0.000), but the difference between groups was not significant at either 15 or 30 min. –––, Metoclopramide; – − – –, magnesium; - - - -, placebo.
In patients with MA, the change in VAS score at 15 min was found to be significantly different between the groups (P = 0.04). In order to detect the group that caused the difference, the Mann–Whitney U-test was applied. Magnesium was found to decrease pain intensity significantly when compared with both metoclopramide (P = 0.03) and placebo (P = 0.04). The change in VAS score at 30 min was not different between all groups (P > 0.05).
The percentage of patients who needed rescue medication at the end of 30 min was 38% for metoclopramide, 44% for magnesium and 65% for placebo (Fig. 2). The need for rescue medication was significantly higher in the placebo group compared with the other two groups (P = 0.04).
Need for rescue medication and pain recurrence at 24 h in the treatment groups. ▪, Metoclopramide; □, magnesium; , placebo.
Side-effects were seen in four (4%) of the 113 patients who completed the study. In the metoclopramide group one (3%) patient developed a dystonic reaction shortly after infusion and flushing was seen in three (8%) patients who received magnesium. None of the side-effects warranted discontinuation of the drug infusion. Diphenhydramine was used for the relief of dystonic reaction. None of the side-effects which occurred after magnesium infusion required treatment.
Patients were followed up by telephone and asked whether the headache recurred 24 h after discharge (Fig. 2). The migraine attack recurred in 43, 52 and 52% of the patients in the metoclopramide, magnesium and placebo groups, respectively, and the difference was not significant (P > 0.05).
Discussion
Although migraine pathophysiology is still not fully understood, it is known that the basic mechanism is neurovascular (22). Both neuronal (cerebral cortex, brainstem) and vascular components (the trigeminovascular system) play important roles in migraine pathophysiology (23). Cortical hyperexcitability, characterized by low excitability thresholds and exaggerated response to stimuli, has been proposed as a crucial step in the initiation of an attack (22–24). It has been shown that cortical hyperexcitability is associated with low intracellular magnesium levels (25), increased levels of neurotoxic amino acids and hereditary changes in calcium channels (24).
Metoclopramide, tried in the treatment of migraine attacks, is a dopaminergic agent. In a recent review, Silberstein et al. categorized i.v. metoclopramide as group 2 (moderate statistical and clinical benefit) (26). In another double-blind, prospective, placebo-controlled study, metoclopramide was found statistically and clinically more effective 1 h after initiation of treatment when compared with placebo (20). However, Coppola et al. found that the efficacy of metoclopramide was equal to placebo at 30 min in a study comparing metoclopramide, prochlorperazine and placebo (27).
The results of our study show that i.v. 10 mg metoclopramide is similar to placebo in the treatment of migraine attacks. This finding does not concur with the aforementioned two studies. This may well be the result of different endpoint timings in these studies. Tek et al. used 60 min as an endpoint while we evaluated patients at 30 min.
Magnesium has been investigated as a treatment option for migraine attacks or prophylaxis. Mauskop et al. reported intravenous MgSO4 as having an important role in decreasing pain in patients with low serum ionized magnesium levels in different types of headache (16, 17). In a randomized, placebo-controlled, double-blind study, Ginder et al. reported MgSO4 to be moderately effective and the response to MgSO4 was unrelated to serum levels (18). In a more recent study, Bigal et al. investigated the differences in efficacy of MgSO4 between migraine without aura (MO) and MA. They concluded that MgSO4 could be used for the treatment of all symptoms in MA or as an adjuvant therapy for MO (13).
The results of trials investigating MgSO4 in migraine are discrepant, perhaps because of methodological differences in the way these trials were conducted. Most of them were not randomized, placebo controlled or blind, and lacked power. Furthermore, most trials were not large enough to be conclusive (16–18, 28). Methodological differences such as selection of outcome time, headache type or study design make comparison between trials difficult.
In a previous study, placebo was found effective up to 45% in relieving headache (29). Similarly in our study, placebo was effective (decrease in VAS scores > 50% from 0 to 30 min) in 43% of patients. In a review of placebo response, Thompson states that the therapeutic effect has three components: the effect of the drug, the effect due to the natural course of the disease, and the placebo effect (30). The placebo effect is influenced by many factors such as the reputation of the treating physician, anxiety, culture and expectation and preconditioning (30). In chronic and repetitive diseases such as migraine, the association of past treatments with relief of symptoms may cause expectation of treatment leading to a conditioned response. It should always be kept in mind that the expectation that the drug will cause pain relief increases the placebo effect. Delivery of a drug by the i.v. route in an ED setting might have increased the expectation of treatment in our study, causing the high placebo response we found.
In our study, the need for rescue medication was greater in the placebo group compared with the treatment groups. The mean VAS score at 30 min was also higher in the placebo group compared with the metoclopramide and magnesium groups, but this difference was not statistically significant. A possible explanation for this finding is that the demand for rescue medication may not only result from pain but may also be caused by accompanying symptoms such as nausea, vomiting, photophobia and phonophobia which persist despite a decrease in pain. We did not evaluate accompanying symptoms in this study, so we can draw no conclusions about the effects of our study medications on symptoms other than pain.
The recurrence rate of pain within 24 h in all three patient groups was about 50% in our study. Similarly high recurrence rates have been found in studies with either magnesium (16) or combinations of other drugs such as sumatriptan, aspirin or metoclopramide (31), and lower for ergotamine (32).
Some limitations of our study warrant mention. We did not evaluate accompanying symptoms like nausea, vomiting, photophobia and phonophobia, so we could not evaluate the efficacy of our study medications on symptoms other than pain. We also did not measure blood ionized magnesium levels in our patients. Therefore we could not determine whether patients with low ionized magnesium had gained extra benefit from infusions of magnesium. Another important point is that the number of patients with MA in our study was small, making it difficult to draw firm conclusions from our results with patients with MA.
The results of this randomized, placebo-controlled, double-blind study indicate that i.v. magnesium and metoclopramide are no more effective than placebo in the treatment of acute migraine attacks. Although this study was sufficiently powered (80% power for a 20-mm difference in VAS scores) to detect differences in treatment efficacy, more powered studies are needed to detect smaller differences in efficacy, and further studies are needed to show whether MA responds differently to treatment from MO.