Interest in the therapeutic use of magnesium in patients with acute myocardial infarction comes from non-controlled studies carried out in Australia, Asia and Europe concerning a supposed decrease in arrhythmic events and an increase in survival in patients treated with magnesium.57,58 Epidemiological studies have also reported that the incidence of sudden death is greater in geographical areas characterized by soft water, which are relatively poor in calcium and magne-sium.59 An epidemiological perspective study has been carried out into the relation between plasma levels of magnesium and the incidence of coronary artery disease,60 and another study investigated the relationship between dietary magnesium and coronary artery disease.61 In the latter study, 13 922 middle-aged subjects were studied for a follow-up period of from 4 to 7 years. The risk of the onset of coronary artery disease was significantly higher in subjects with low plasma levels of magnesium and there was a significant inverse correlation between the two parameters. Furthermore, patients with plurimetabolic syndrome, that is, a syndrome characterized by obesity, dyslipi-demia, reduced peripheral glucose tolerance in the presence or absence of arterial hypertension, and patients who are characterized by increased risk of coronary artery disease show a decrease in ionized intralymphocyte magnesium and the level of magnesium is inversely related to peripheral insulin resistance.62 65
A weak inverse correlation was also found between dietary intake of magnesium and the incidence of coronary artery disease. It is well known that a key element in the physiopathology of acute myocardial infarction is platelet activation. Under normal conditions, platelets circulate in a 'resting' condition without interacting with each other or with other cells such as leucocytes or endothelial cells. After contact with agonists such as thrombin, collagen and ADP, the platelets become active and exhibit fibrinogen binding sites on the IIb-IIIa glycoprotein. At this point plasma fibrinogen may bind to the platelets thus allowing them to form microaggregates. This is the first, still reversible, step in platelet aggregation. After this, the platelets release into circulation a multitude of molecules which were stored in the granules.
Some studies have shown that magnesium is able to inhibit platelet activation both by inhibiting certain factors which stimulate the platelets such as throm-boxanes A2 and by inhibiting the production of platelet-inhibiting factors such as prostacyclin.66,67
Many years ago it was reported that magnesium was able to prolong coagulation time if added in small quantities to human plasma.68 Adams and Mitchel found that magnesium inhibited thrombus formation and increased the minimum concentration of ADP necessary to start the formation of a thrombus.69 In an experimental in vitro model of platelet-dependent thrombosis, Ravn and collaborators recently described how the in vitro administration of magnesium was able to inhibit the formation of arterial thrombosis.70 Hypomagnesaemia (low magnesium plasmatic levels) seems to selectively inhibit the release of nitric oxide by the endothelium.71 Since nitric oxide is a powerful inhibitor of platelet aggregation and adhesion, hypomagnesaemia could be accompanied by a predisposition to thrombosis.
A global analysis by Horner72 of magnesium therapy in acute myocardial infarction found that this treatment is safe and useful. However, two doubleblind controlled mega-trials have cast doubt on this point of view. The first is the Leicester intravenous magnesium interventional trial (LIMIT-2) which recruited 2316 patients.73 Magnesium was infused either at the moment of thrombolitic therapy or before spontaneous riperfusion (when an acute myocardial infarction is caused by thrombosis on an atherosclerotic plaque in a coronary artery leading to complete vessel occlusion) a spontaneous riperfusion may occur. Total mortality measured at day 28 was 7.8% in the group randomized to magnesium and 10.3% in the group randomized to placebo with a relative reduction of 24%. Moreover, in the group randomized to magnesium there was a 25% decrease in episodes of left cardiac failure. In the group treated with magnesium, no improvement was recorded in terms of episodes of hypotension or of arrhythmic events and, in fact, in this group there was no reduction in the use of anti-arrhythmic drugs or cardiac pacing. The effect of the reduction of mortality obtained with magnesium was long lasting, as shown by a reduction in total mortality of 16% with a mean follow-up of 2.7 years. Side-effects of the treatment included an increase in bradiarrhythmia (i.e. a series of arrhythmia characterized by either decrease in frequency of ventricular contractions or an increased duration of atrio-ventricular conduction, i.e. a slower propagation of excitation along the specialized tissue of conduction) and skin flushing on administration.
The second mega-trial to deal with the question of treatment with infusions of magnesium in patients with acute myocardial infarction was called the Fourth International Study of Infarction Survival (ISIS-4).74 In addition to the infusion of magnesium, patients were allocated randomly to thrombolitic, nitroderivative or ACE (angiotensin-converting enzyme)-inhibitor therapy. The patients were recruited with a mean interval of 8 hours (maximum 24 hours) from the onset of symptoms and treatment with magnesium was interrupted during riperfusion. The death rate measured after 5 weeks in patients randomized to treatment with magnesium did not show any improvement compared to placebo, and there was even a trend towards an increase in the death rate, with an increase in the incidence of cardiogenic shock and heart failure in the group treated with magnesium (7.6% vs 7.2%), although there was a significant reduction in the incidence of ventricular fibrillation. Since this study, the use of magnesium in the treatment of myocardial infarction has been significantly scaled down.
In any case, the two studies described above, which show clearly contrasting results, are characterized by differences in their design. In LIMIT-2, the dose of magnesium was a bolus of 8 mmol followed by 65 mmol infused over 24 hours, while in ISIS-4 the dose was of 8 mmol in bolus followed by 72 mmol over 24 hours. The timing of the infusion of magnesium was different since in LIMIT-2 it was earlier than in ISIS-4. Some authors hypothesized that the delay in starting the infusion of magnesium may have limited the ability of the magnesium to inhibit the damage from riperfusion, which is one of its main beneficial actions. During the first minutes of riperfusion, the calcium accumulates in the myocardium, particularly in the mitochondria, and there is depletion of high-energy phosphates and a contractile dysfunction. If magnesium is to carry out in full its protective effect, the concentration of magnesium in the plasma must be increased before riperfusion. The timing of the infusion of magnesium is therefore critical in relation to both spontaneous riperfusion and after thrombolysis. The ideal therapeutic window for the infusion of magnesium is probably limited to the first 1-2 min after riperfusion and it has been shown that the protective effect of magnesium is reduced when it is administered after an hour from the beginning of riperfusion.75 Some authors have shown in an experimental model that magnesium is able to reduce the influx of calcium in myocardial cells only when it is administered within 15 min from the onset of riperfu-sion. The delicate question of the exact timing of the infusion of magnesium has been reconsidered in the MAGIC (magnesium in coronaries) trial.
Unfortunately, the recent results of the MAGIC study are not encouraging.76 MAGIC is a vast-scale trial designed to show the potential beneficial effect of the infusion of magnesium. The experiment involves the enrolment of 10 400 patients with a high risk of acute myocardial infarction and elevation of the ST-segment, and their random assignment to early treatment with infusion of magnesium or placebo. The primary end-point was the mortality at 30 days.76 At 30 days, 15.3% patients in the magnesium group and 15.2% in the placebo group had died. No benefit or harm due to magnesium was observed. Extensive univariate analysis did not identify any subgroup that benefited from magnesium nor were there beneficial effects from magnesium in a multivariate model that adjusted for factors shown to affect mortality risk.
From all these data taken together, in current coronary care practice, there is no indication for the routine administration of intravenous magnesium in patients with acute myocardial infarction.
Was this article helpful?