Acute stroke is a leading cause of mortality and severe disability in developed countries. Acute anticoagulation within three hours of the onset of symptoms can improve outcomes but carries significant risk. The results of some studies indicate that metabolic events in the hypoperfused zone (the ischemic penumbra) are critical to the extent of cell death. This finding indicates that early initiation of neuroprotective therapy could alter neurochemistry and improve outcomes in patients with acute stroke. Because animal models and results from several small pilot studies have shown neuroprotective effects of magnesium, a large international study (the Intravenous Magnesium Efficacy in Stroke Study) examined the ability of intravenous magnesium to improve outcomes in patients with acute stroke.

The investigators enrolled 2,589 patients within 12 hours of onset of stroke symptoms. Patients were previously healthy adults who were conscious and had limb weakness for at least one hour on entry to the study. On the National Institutes of Health stroke scale, these patients had a score of 1 or more. Use of anticoagulation was allowed. Brain imaging had to be completed within seven days of entering the study. Pregnant patients, and those with significant comorbidities, renal impairment, or contraindication to magnesium, were excluded from the study. Eligible patients were assigned randomly after stratification for age, type of stroke, and time since onset of symptoms to receive intravenous magnesium(MgSO₄ solution) or an identical placebo solution. Magnesium was given as a bolus of 4 g (16 mmol) ofMgSO₄ infused over 15 minutes followed by 16.15 mg (65 mmol) infused over 24 hours. Patients were followed for up to 90 days to assess mortality and change in stroke morbidity, using the Barthel and Rankin scores.

Ninety-eight percent of patients received the intended intervention, and only 0.3 percent were lost to follow-up. Patients in both study groups were well matched for important characteristics. The median time from symptom onset to treatment was seven hours. The proportion of patients dead or disabled at 90 days was not significantly different in the 1,188 patients who received magnesium compared with the 1,198 who received placebo (227 deaths compared with 196 deaths, respectively). The odds ratio for death during the study was 1.22. The proportion of patients with disability as measured by Barthel and Rankin scores was not significantly better in patients who received magnesium than in those who received placebo. No treatment effect was identified overall or after adjustment for significant variables. In subgroup analysis, a possible beneficial effect was detected in patients with ischemic lacunar strokes. However, the clinical reliability of this observation is uncertain. Otherwise, no subgroups were identified with positive treatment effect. The two groups also had equivalent serious adverse events within 48 hours of study entry.

The authors conclude that early administration of intravenous magnesium does not reduce mortality or disability in the 90 days following onset of acute stroke. They speculate that the relatively late median treatment (seven hours after onset of symptoms) could have contributed to the failure to demonstrate efficacy, and call for further trials using shorter delays in initiation of treatment.