Recommendations
Class I
Class IIa
1. Correction of documented magnesium (and/or potassium)
deficits, especially in patients receiving diuretics before
onset of infarction.
2. Episodes of torsades de pointes-type VT associated with
a prolonged QT interval should be treated with 1 to 2 g
magnesium administered as a bolus over 5 minutes.
Class IIb
1. Magnesium bolus and infusion in high-risk patients such
as the elderly and/or those for whom reperfusion therapy is
not suitable.
Comment: The available data suggest that mortality
reduction may be seen in high-risk patients, provided
magnesium therapy is administered soon after onset of
symptoms (preferably less than 6 hours). The optimum dose has
not been established, but a bolus of 2 g over 5 to 15 minutes
followed by an infusion of 18 g over 24 hours has been used
with success.
Background
Supplemental administration of magnesium for reducing
morbidity and mortality in patients with acute MI is a
reasonable avenue to pursue because of abundant data relating
magnesium to cardiovascular disease.548 It is the second most
abundant intracellular cation and is involved in more than 300
enzymatic processes. Evidence exists that magnesium produces
systemic and coronary vasodilatation, possesses antiplatelet
activity, suppresses automaticity in partially depolarized
cells, and protects myocytes against calcium overload under
conditions of ischemia by inhibiting calcium influx especially
at the time of reperfusion.548-552
Meta-analyses of the seven randomized trials published
between 1984 and 1991 suggest a significant mortality benefit
of magnesium (odds ratio 0.44, CI 0.27 to 0.71).553,554 The
Second Leicester Intravenous Magnesium Intervention Trial
(LIMIT-2) trial555 subsequently reported a 24% reduction in
mortality with magnesium treatment (P<.04). The
magnesium-treated patients in LIMIT-2 had a 25% lower incidence
of CHF in the CCU and a 21% lower rate of ischemic heart
disease-related mortality over 4 years, consistent with the
hypothesis that magnesium exerts its beneficial effects, at
least in part, via a myocardial protective action.555,556
The results of one large trial were negative. The ISIS-4
investigators enrolled 58 050 patients, 29 011 allocated to
magnesium, and 29 039 to control. There were 2216 deaths
(7.64%) by 35 days in the magnesium group and 2103 deaths
(7.24%) in the control group (odds ratio 1.06; CI 0.99 to
1.13), suggesting no mortality benefit of magnesium
administration and even the possibility of slight harm.421 When
ISIS-4 is added to the preceding randomized trials,
meta-analysis indicates no beneficial effect of magnesium.
Possible sources of heterogeneity that could explain these
differences include:
1. The relatively late administration of magnesium in
ISIS-4.557
2. The control group mortality in ISIS-4 was only 7.2%.
Regression analyses of the available data predict a null effect
of magnesium when the control mortality is about 7% and
increasing benefit of magnesium for higher control mortality
rates.558
Shechter and colleagues559 recently reported a randomized
trial of 194 patients with acute MI unsuitable for
thrombolysis. There was a significant reduction in mortality in
the magnesium group (4.2% versus 17.3%, P<.01), largely due
to a lower incidence of cardiogenic shock and CHF.
An NHLBI-sponsored trial (Magnesium in Coronary Disease
[MAGIC]) is planned to further evaluate the role of magnesium
in acute MI, especially with early administration before
thrombolysis in higher-risk patients.557
© 1999 by the American College of Cardiology and
American Heart Association, Inc.
Healthy Water
Association--USA