In American Journal of Cardiology 76:172-173, 1995
ISIS-4 and LIMIT-2
The disparity of published results of magnesium salt infusions in treatment of suspected acute myocardial infarction, AMI(1-5), has evoked controversy.(5-10) Important clinical questions need resolution. Why did the Fourth International Study of Infarct Survival (ISIS-4) of 58,050 suspected AMI patients find no overall benefit from magnesium infusions, and even some increased risk,(5) while the Second Leicester Intravenous Magnesium Intervention Trial (LIMIT-2) in 2,316 suspected AMI patients reported that magnesium infusion resulted in improved left ventricular function, as well as lowered short-term mortality(1) that persisted, in a long-term follow up(2)? Why did a meta-analysis of the results of magnesium infusion given to >1300 AMI patients with suspected AMI in small controlled studies show a dramatic reduction in mortality(3) and also decreased post-AMI arrhythmias that were not seen in either large study? Do the negative findings of the mega-trial [ISIS-4] negate the protective effects of magnesium, that were reported in the earlier trials? The findings in LIMIT-2 led to the recommendation that, because it is a safe, simple, inexpensive intervention, magnesium should be entered into clinical practice without delay.(4) Is it justified to conclude, as did authors of ISIS-4, largely from statistical interpretation of pooled findings, that there are no longer grounds for use of magnesium in AMI?(5-9) The following 4 important dissimilarities among the protocols used in ISIS-4,(5) LIMIT-2,(1-2) and in the meta-analyzed small studies(3) deserve scrutiny: 1. time of initiation of magnesium treatment in relation to thrombolytic therapy; 2. total dosage in the first 24 hours; 3. duration of post-AMI magnesium infusions, and 4. treatment of control groups, that affects comparative mortality rates.
1. In the small studies and in LIMIT-2, magnesium infusion was begun as soon as possible after onset of symptoms, and before or without subsequent fibrinolysis. Since reperfusion injury begins immediately on reestablishing blood flow, and magnesium is protective,(1,2,6-2) Woods et al(1) attributed the 24% reduction in 28-day mortality in LIMIT-2 to its prompt administration, before thrombolysis, the timing of magnesium treatment in relation to thrombolysis being critical.(2) Reduced 28-day left ventricular failure - a strong predictor of subsequent mortality - might explain the 21% reduction in mortality over a mean follow up of 2.7 years, of the LIMIT-2 patients who had received early magnesium treatment. Yusuf and Flather(9) denied that the failure of ISIS-4 to confirm the earlier favorable findings was caused by delay in providing magnesium. They commented that recanalisation of occluded arteries would have been almost simultaneous with the magnesium infusion in a large proportion of ISIS-4 magnesium-treated patients who underwent thrombolysis. However, the report(5) indicates that magnesium was begun within two hours of fibrinolytic therapy (which takes about 90 minutes for 50-60% restoration of arterial patency) in about half of all randomized patients. One fourth of the patients seem to have received magnesium between two to six hours after onset of symptoms and fibrinolysis; the remaining patients may have started treatment up to 24 hours after symptom-onset. Because even slight delay may lessen the beneficial effects of magnesium,(2) fibrinolysis, given before magnesium, might have reduced or eliminated the protection that could have been afforded by prompt magnesium administration.
2. The amounts of magnesium infused over a 24-hour period might account for differences in adverse effects. Galloe and Graudal(7) analyzed the 24-hour magnesium dosages in the small studies and showed the range to be 50-65 mM. They suggest that the optimal amount of magnesium in the first 24 hours could be 55 mM, and that > 75mM may increase mortality. They referred to the double-blind study of 298 patients with AMI studied by Feldstedt et al,(11) who had reported that 24-hour magnesium infusion of 80mM significantly increased bradyarrhythmias and led to fatal pump failure. They observed that this could have been avoided if lower infusion doses of magnesium had been given. In LIMIT-2, the 24-hour dosage was 73 mM; 24-hour post-infusion serum levels averaged 1.55 mM/L. In ISIS-4, 80 mM were given over 24 hours; serum magnesium levels were not reported. Hypermagnesemia, by inducing generalized vasodilatation, can cause profound hypotension, which could increase risk of shock and of extension of myocardial necrosis. Hypermagnesemia was not a problem in LIMIT-2;(1) it occurred significantly (p <0.001) more often (sometimes severe enough to terminate magnesium treatment) than in controls in magnesium treated patients in ISIS-4.(5) Hypermagnesemia also causes blocks of conduction and bradycardia. Magnesium treated patients in ISIS-4, when compared to controls, had slightly increased second or third degree heart block, small but significant increased heart failure (p <0.001), and increased deaths due to cardiogenic shock (p <0.001) that emerged during or just after the infusion. Thus, the amount of magnesium over 24 hours may be a major factor in morbidity and mortality.
3. Another difference between the small and large studies is in duration of post-AMI magnesium treatment. Magnesium infusions were continued after the first 24 hours, but at lower doses for 12 hours to two days longer or more, by groups reporting the small trials,(12-14), a point that was raised in an earlier editorial.(15) Because the anti-tachydysrhythmic effect of magnesium(16) was seen only in the small studies, a longer duration of post-AMI magnesium infusion might be required to prevent tachydysrhythmias.
4. The treatment and mortality rates of control groups affect rates of improvement in the test group. The higher control mortality rate in LIMIT-2, than that in ISIS-4, might reflect the fact that fewer LIMIT-2 controls underwent thrombolysis or received anti-platelet therapy than did ISIS-4 controls: 35% versus 70%, and 66% versus 94%, respectively. The >50% reduction in mortality in magnesium-treated patients versus controls in the small studies reflects high mortality rates in controls, possibly because anti-platelet and fibrinolytic therapy was not customary in the 1980s, when the small studies were performed.
In the ISIS-4 discussion,(5) the benefit that was conventionally significant (p=0.04)(8,9) in LIMIT-2 was reevaluated and deemed uncertain, being statistically compatible with both halving of mortality and there being no benefit. The possibility that negative findings were not published, and that this might have biased statistical interpretation of the meta-analysis, has been suggested.(10) However, an important negative clinical study, that dealt with increased mortality of post-AMI magnesium infusions providing 80mM in the first 24 hours was published in 1991.(11)
Some of the activities of magnesium, that should protect against extension of AMI-induced myocardial damage, were considered by Rasmussen(13,17) and by Woods(12) in their rationales for use of magnesium in AMI. Antiarrhythmic and calcium-blocking effects, reduction of potassium loss, antiplatelet aggregation and vasodilating effects (mediated in part by platelet-derived and endothelium-derived factors), and protection against free radical-induced injury were cited as activities of magnesium that should protect against ischemia-induced cardiac dysfunction and tissue damage. The administration of magnesium promptly after development of signs and symptoms of AMI, and before other therapies, should maximize protective effects of magnesium. Adverse effects, that increase morbidity and mortality, might be minimized by excluding patients with sinoatrial atrioventricular blocks, and by not exceeding the 24-hour optimal dosage of magnesium infusions (range 50-65 mM): however this requires further study, as does the optimum duration of magnesium therapy.
1. Woods KL, Fletcher S, Roffe C, Haider Y. Intravenous magnesium sulphate in suspected acute myocardial infarction: results of the second Leicester Intravenous Magnesium Intervention Trial (LIMIT-2). Lancet 1992;339:1553-1558.
2. Woods KL, Fletcher S: Long-term outcome after intravenous magnesium sulphate in suspected acute myocardial infarction: the second Leicester Intravenous Magnesium Intervention Trial (LIMIT-2). Lancet 1994;243:816- 819.
3. Teo KK, Yusuf S, Collins R, Held PH, Peto R: Effects of intravenous magnesium in suspected acute myocardial infarction: overview of randomized trials. Brit Med J 1991;303:1499-1503.
4. Teo KK, Yusuf S: Role of magnesium in reducing mortality in acute myocardial infarction. A review of the evidence. Drugs 1993;46:347-359.
5. ISIS-4: A randomized factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet 1995;345:669-682.
6. Casscells W: Magnesium and myocardial infarction (commentary). Lancet 1994;343:807-809.
7. Galloe A, Graudal N (letter to editor): Magnesium and myocardial infarction. Lancet 1994;343:1286-1287.
8. Unsigned: Teething problems for two innovations. Science 1994;265:1540.
9. Yusuf S, Flather M: Magnesium in acute myocardial infarction. ISIS-4 provides no grounds for its use. (Editorial) Brit Med J 1995;310:750-752.
10. Egger M, Smith GD: Misleading meta-analysis. Lessons from "as effective, safe, simple" intervention that wasn't. (Editorial) Brit Med J 1995; 310:752-754.
11. Feldstedt M, Boesgaard S, Bouchelouche P, Svenningsen A, Brooks L, Lech Y, Aldershvile J, Skagen K, Godtfredsen J: Magnesium substitution in acute ischaemic heart syndromes. Eur Hear J 1991;12:1215-1218.
12. Morton BC, Nair RC, Smith FM, McKibbon TG, Poznanski WJ: Magnesium therapy in acute myocardial infarction - a double-blind study. Magnesium 1984; 3:346-352.
13. Rasmussen HS, Suenson M, McNair P, Norregard P, Balsey S: Magnesium infusion reduces the incidence of arrhythmias in acute myocardial infarction. A double-blind placebo-controlled study. Clin Cardiol 1987; 10:351-356.
14. Ceremuzynski L, Jurgiel R, Kulakowski P, Gebalska J: Threatening arrhythmias in acute myocardial infarction are prevented by intravenous magnesium sulfate. Am Heart J 1989;118:1333-1334.
15. Seelig MS. Editorial: Magnesium in acute myocardial infarction. (International Study of Infarct Survival 4). Am J Cardiol 1991;68:1221- 1222.
16. Iseri LT: Magnesium and Dysrhythmias. Magnesium Bul 1986;8:223-229.
17. Rasmussen HS: Justification for intravenous magnesium therapy in acute myocardial infarction. Magnesium Res 1988;1:59-73.
18. Woods KL: Possible pharmacological actions of magnesium in acute myocardial infarction. Br J Clin Pharmacol 1991;32:3-10.
Address for reprints: Ronald J. Elin M.D., Ph.D., Chief, Clinical Pathology Department, National Institutes of Health, Building 10, Room 2C-306, 10 Center Drive-MSC 1508 Bethesda, MD 20892-1508
This page was first uploaded to The Magnesium Web Site on October 17, 1995