Journal Logo


Hemostatic Changes After Early Versus Late Intracoronary Magnesium During Acute Myocardial Infarction in Swine

Serebruany, Victor; Herzog, William*; Schlossberg, Matthew*; Edenbaum, Lisa*; Gurbel, Paul

Author Information
Journal of Cardiovascular Pharmacology: December 1996 - Volume 28 - Issue 6 - p 817-823
  • Free


Controversy surrounds the use of magnesium (Mg) salts during acute myocardial infarction (AMI). This debate arises because of conflicting results from recent clinical trials (1,2). Several different hypotheses have been advanced to explain the cardioprotective properties of Mg, including the influence of the timing of Mg administration. An attempt to relate the Mg ion with platelet activity and coagulation is not new. Competitive antagonism between cytosolic Mg and calcium channel activity occurs in vascular smooth muscle (3) and platelets (4). Mg provides anticoagulant activity in vitro against prothrombin, factor V, VII, and factor IXa (5). However, the effect of intracoronary Mg supplementation on the hemostatic profile during AMI is not known. In animals, high levels of extracellular Mg in vitro (6), as well as intravenous Mg supplementation ex vivo (7), are associated with a progressive dose-dependent inhibition of platelet aggregability. How Mg affects the plasma levels of the natural antithrombotics (AT-III, protein C, and protein S) is still under investigation. Mg deficiency is associated with an increased urinary excretion of the TXB metabolite (8). Mg supplementation results in a dramatic increase in the urinary prostacyclin breakdown product (9). This association between antagonistic eicosanoids and Mg supplementation during AMI is not fully understood. Although there is agreement on the importance of fibronectin in coronary artery disease, the exact significance of plasma fibronectin levels is a matter of considerable controversy. Plasma fibronectin was significantly decreased after 48 h in patients with AMI complicated by ventricular arrhythmia, left ventricular failure, or death (10). It was shown that supplemental MgSO4 can ameliorate myocardial stunning (11) and reduce infarct size in a swine model (12,13). However, the timing of treatment is very important, and a delay of even an hour results in no effect (13).

Our study was designed and implemented in the same animals to test the hypothesis that the timing of intracoronary Mg supplementation may influence critical hemostatic parameters differently during AMI (13). Such an approach could be useful in understanding the advantages of early Mg delivery in the numerous clinical settings of ischemia-reperfusion in which hemostatic abnormalities are established and well described. A swine model was chosen for several reasons. First, porcine hemostatic parameters, especially fibronectin, AT-III, protein C, and ET-1 are identical or very similar to those in humans and are therefore relevant to thrombosis research (14,15). Second, Yorkshire swine are large enough to obtain multiple blood samples, to produce a model of regional ischemia similar to the human condition, and to allow local intracoronary drug delivery. Finally, the coronary anatomy of the swine mimics the human coronary circulation closely, especially with regard to a relative absence of preexisting collateral flow (16).


This study was approved by the Institutional Animal Care and Use Committee of the University of Maryland. All procedures conformed to the guidelines established by the U.S. Department of Health and Human Services, published by the U.S. National Institute of Health (NIH publication No. 85-23, revised 1985).


Twenty-six purebred Yorkshire female swine (34-42 kg), aged 10-12 weeks, received normal swine food (Purina, Richmond, IN, U.S.A.) and were housed at our institution for a minimum of 1 week before use. The animals were randomized into three groups. The first group (Mg-early group, average weight of 35.9 ± 2.0 kg) received a 12-min intracoronary infusion of MgSO4 (250 mg) with the onset of reperfusion. The second group (Mg-late group; average weight 38.2 ± 2.1 kg) received the same dose of intracoronary MgSO4 beginning 1 h into the reperfusion period. The third group (average weight 37.4 ± 1.7 kg) received intracoronary normal saline at the beginning of reperfusion. They served as the controls. A detailed protocol for the myocardial infarct size experiments can be found elsewhere (13). In brief, all three groups were subjected to 50 min of occlusion of the left anterior descending artery (LAD) followed by 3 h of reperfusion. Hemodynamics (mean arterial pressure, heart rate, and left ventricular end-diastolic pressure) were continuously monitored through the entire protocol. Blood was collected six times from the femoral vein during the experiment: at the baseline; at 25 min, and at 50 min of occlusion; and after 60, 120, and 180 min of reperfusion. To avoid possible observer bias, blood samples were coded and blinded before any measurements. Hemostatic parameters were determined by an individual unaware of the treatment protocol.


Blood was collected into plastic tubes with 7.5 mM EDTA and centrifuged immediately at 2,000 g for 10 min at 4°C. Plasma was stored at -80°C before analysis. Endothelin-1 (ET-1) was extracted from plasma by using Amersham Amper 500 columns (Amersham International, Little Chalfont, England). First the column was equilibrated by washing with 2 ml methanol followed by 2 ml water. Then 1 ml of plasma was acidified with 0.25 ml 2 M HCl, centrifuged at 10,000 g for 5 min, and loaded onto the column. The column was then washed twice with water, 0.1% trifluoroacetic acid, and 80% high performance liquid chromatography (HPLC) grade methanol. Eluent was collected into a polypropylene tube and dried under nitrogen. Finally, the pellet was reconstituted in 250 μl assay buffer, which was taken for analysis. The ET-1 plasma level was determined by using an immunoenzymetric “sandwich” enzyme-linked immunosorbent assay (ELISA) system (Amersham International, Little Chalfont, England).


Fibronectin was measured in EDTA-treated platelet-poor plasma (PPP; 5,000 g for 5 min) with kinetic turbidimetry of the antigen-antibody reaction according to the principle of the fixed time method (Boehringer Mannheim, Mannheim, Germany).


Because eicosanoids have a very short half-life under physiologic conditions, their metabolites were analyzed. TXB2, the stable breakdown product of thromboxane A2 (TXA2), and 6-keto-PGF1a, the stable degradation product of prostacyclin, were measured in the remaining PPP, which was kept at -4°C. In vitro prostaglandin biosynthesis was inhibited with 7.5 mM EDTA and 4 μg/ml indomethacin. Plasma samples were extracted with ethanol and then stored at -80°C before final prostaglandin determination by using TiterZyme enzyme immunoassays (PerCeptive Diagnostics, Inc., Cambridge, MA, U.S.A.).


AT-III was determined in citrated PPP using a quantitative chromogenic assay Accucolor (Sigma Chemical, St. Louis, MO, U.S.A.).

Protein S and protein C

Proteins S and C were measured in citrated plasma by using commercial Asserachrom (Diagnostica Stago, Asnieressur-Seine, France) enzyme immunoassays.

Statistical analysis

All comparisons were done by using repeated-measures analysis of variance. The values are expressed as mean ± SEM; p < 0.05 was considered significant. The post hoc comparison by using the Bonferroni t test was performed to identify specific differences between Mg-supplemented and saline-treated animals.


Exclusion of animals/arrhythmias

Of the 26 swine under investigation, eight animals were excluded because of ventricular fibrillation during late occlusion or early reperfusion (three of nine controls, one of seven Mg-early group, and four of 10 Mg-late group). The remaining analyzed animals were six controls, six Mg-early, and six Mg-late swine.

Hemostatic parameters

The plasma levels of ET-1, fibronectin, eicosanoids, and natural antithrombotics (AT-III, total protein S, protein C) in saline- and both Mg-treated groups of animals are summarized in Table 1. At baseline, there were no differences between the three experimental groups. However, myocardial ischemia followed by reperfusion was associated with the significant disturbances in the hemostatic parameters measured in the systemic circulation. Such changes occurred in each experimental group.

Control group. We observed a decline in ET-1 plasma levels at the end of occlusion. This was followed by a dramatic increase that remained constant during the entire reperfusion. Plasma fibronectin increased significantly from baseline during coronary artery occlusion and remained elevated during reperfusion. At first, TX levels decreased significantly during occlusion and then increased above baseline value throughout reperfusion. Plasma 6-keto-PGF1a decreased during the occlusion phase followed by a significant increase during reperfusion. AT-III plasma levels also significantly decreased during occlusion. No further changes were observed during reperfusion. Total protein S declined during occlusion followed by a significant increase at the reperfusion phase. Meanwhile, plasma protein C increased significantly at the end of occlusion and reached a peak with the beginning of reperfusion. A decline back to the baseline level was noted at the end of the experimental protocol.TABLE 2.

Mg-early group. Progressive significant decreases in plasma levels of ET-1, TX, prostacyclin, AT-III, and protein S occurred during LAD occlusion in all animals. In contrast, fibronectin and protein C levels significantly increased during occlusion. The reperfusion phase was characterized by a significant increase of ET-1, TX, prostacyclin, and total protein S; plasma AT-III remained unchanged, whereas plasma fibronectin and protein C decreased.

Mg-late group. When compared with that of controls, plasma TX concentration after the second hour of reperfusion was reduced, whereas the AT-III level was slightly increased. At all other times during ischemia-reperfusion, there were no differences in the hemostatic parameters between the Mg-late group and controls.

Controls versus Mg-early. There were major differences in the hemostatic parameters between those animals that received early-Mg and controls. Plasma ET-1 was reduced at the beginning of the reperfusion period in the Mg-early group (Fig. 1). We observed also a marked decrease in the plasma fibronectin concentration in the Mg-early group throughout reperfusion phase (Fig. 2). The Mg-early group demonstrated a reduced plasma TXB2 after 1 h of reperfusion (Fig. 3). Despite a similar trend between groups, plasma protein C increased and remained significantly higher during the end of occlusion and reperfusion in the Mg-early group (Fig. 7). There were no significant differences in the AT-III profile during ischemia-reperfusion between both groups (Fig. 5). However, intracoronary administration of MgSO4 was associated with a transitory increase of plasma AT-III. The patterns for plasma prostacyclin (Fig. 4) and total protein S level (Fig. 6) during ischemia-reperfusion showed no differences between experimental groups.


One attractive strategy to achieve further clinical gains in the treatment of AMI is to use adjunctive therapies that may either limit reperfusion injury or prevent repeated episodes of ischemia by maintaining arterial patency. Concurrently, researchers in the field of reperfusion injury have demonstrated a beneficial effect of Mg supplementation in experimental models of ischemia-reperfusion. However, the results of clinical trials evaluating the therapies to decrease reperfusion injury and enhance AMI survival rate have been controversial. The LIMIT-2 trial (2,316 patients) showed a significant (25%) mortality benefit when 1 g of Mg was infused longer and started early (1), whereas the ISIS-4 trial (54,824 patients) did not confirm these data (2).

The data from this laboratory recently showed that intracoronary low-dose Mg delivered immediately at the time of reperfusion can significantly diminish infarct size in swine (13). Our study suggests that intracoronary Mg in the same animals is associated with favorable changes in the hemostatic profile but is dependent on the timing of Mg delivery. The mechanism of such action, however, remains unknown. The proposed mechanisms of the cardioprotective properties of Mg include antiplatelet effects (7), which may help to avoid arterial reocclusion, and vasodilatory effects (11), which may decrease afterload and prevent spasm. The protective action of Mg may also be mediated through antioxidants (17), leading to the free radical hypothesis. Mg has also been described as a “natural” Ca-channel blocker, preventing calcium flux (18). Assuming Ca2+ ion plays a major role in platelet aggregation and the coagulation cascade, we hypothesize that Mg may prevent such activation. Some evidence suggests that Mg may exert its cardioprotective effect by altering the prostaglandin system.

It has been demonstrated that Mg amplifies the synthesis of prostacyclin from umbilical vein endothelial cells (19). In support of this, Mg was also found to increase 6-keto-PGF1a release in hypertensive rats, while lowering blood pressure (20). In normal human subjects, Mg infusion produced a significant increase in the excretion of a prostacyclin metabolite (8).

In our study, there were no differences in plasma 6-keto-PGF1a levels in both Mg-treated groups as compared with controls. These findings contradict the prostacyclin theory of Mg action. On the other hand, Mg administration was associated with a significant reduction in the plasma TXB2 level during reperfusion in both Mg-early and Mg-late groups and was dependent on the timing of Mg infusion. Unchanged 6-keto-PGF1a plasma concentrations together with the decreased TXB2 levels during low-dose intracoronary Mg therapy could partly compensate proaggregatory and hypercoagulemic changes in the hemostatic profile during AMI.

The interest in the determination of plasma fibronectin levels during Mg supplementation is stimulated by recent reports linking Mg and fibronectin. Several nuclear proteins have been shown to interact in a Mg-dependent fashion with a conditionally processed pre-messenger RNA (mRNA) derived from the fibronectin gene (21). Mg alone has no effect on fibronectin binding to endothelial cells but instead acts as an antagonist, suppressing the calcium-stimulated binding (22). We observed a significant decline in the plasma fibronectin level in the early-Mg group. A similar trend in fibronectin concentrations has been observed after uncomplicated AMI in humans (10).

The rationale for plasma AT-III determination during Mg supplementation was based on a recent observation that Mg enhances AT-III activity against thrombin (23). Moreover, acquired AT-III deficiency has been documented in myocardial ischemia-reperfusion, which includes unstable angina and AMI (24,25). We found enhanced AT-III levels shortly after Mg infusion during reperfusion in both early- and late-Mg groups. It seems that changes in AT-III plasma level during AMI are not associated with the cardioprotective properties of Mg in the reduction of infarct size. We measured protein C and protein S because these antithrombotics play an important role in the prevention of arterial reocclusion after thrombolysis in experimental AMI (26). We found that intracoronary Mg supplementation is associated with significant increase of protein C plasma levels in the Mg-early group, whereas the concentration of total protein S in plasma was not found to be different between groups. Similar trends of increased protein C and unchanged total protein S levels have been observed in patients after uncomplicated AMI (24).

Decreases in the total protein S level during coronary artery occlusion may be related to its increased binding to activated platelets (27). The protein C elevation after early-Mg infusion may represent an advantage in preventing further thrombosis and limiting infarct size.

Because vasoconstriction is an important factor in AMI progression, we evaluated the possible association between Mg supplementation and ET-1. Increased plasma concentrations of ET-I have been shown positively to correlate with MI size (28,29). In rats, the vasoconstrictor action of ET-1 was completely abolished by high-dose magnesium sulfate (30). Furthermore, with Mg infusion, ET-1 levels were lower compared with preinfusion values, but incubation with Mg does not alter the ET-1 release from human endothelial cells (31). Results of this study demonstrate that early Mg infusion is associated with a significant decrease of plasma ET-1 levels at 1 h of reperfusion. Prevention of the ET-1 increase during early reperfusion by low-dose intracoronary Mg can be partly responsible for the benefits of Mg supplementation in experimental AMI. The possibility of the delayed effects of Mg delivered later beyond 3 h of reperfusion could not be excluded.

Study limitations

The overall purpose of this project was an attempt to link the changes in the hemostatic profile during AMI with the timing of low-dose intracoronary Mg supplementation. The primary limitation of this study is that we did not measure local intracellular Ca and Mg concentrations in the myocytes or coronary artery endothelial cells. Thus there is no direct evidence that the observed changes in hemostasis were the result of the cellular effects of Mg. However, most of the favorable changes occurred in the early-Mg group compared with both the control and late-Mg group. Because treatment was limited to the reperfusion phase, we were able to avoid systemic hemodynamic effects from the local delivery of low-dose MgSO4. Other limitations may include the choice of hemostatic parameters studied. We limited ourselves to those that have demonstrated experimental or clinical association or both with Mg.

Finally, we do not have any data on the effects of Mg supplementation before or during an occlusion phase of AMI on the same hemostatic parameters. It will be critical in future studies to reproduce the observed changes in the clinical setting of ischemia-reperfusion. Further investigation of the effects of Mg on the hemostatic profile is warranted.

In summary, our data suggest that early-Mg supplementation is associated with a significant favorable change in certain hemostatic factors. The timing of treatment is very important, and a delay of even 1 h may result in selective alterations of hemostatic factors. Although we must be cautious in drawing clinical conclusions from animal models, it may be worthwhile to speculate on the clinical significance of these findings. Supplemental administration of Mg, when given early during reperfusion, may have beneficial effects on morbidity and mortality rates in patients with AMI because of improved hemostatic parameters and elimination of possible thrombotic or thromboembolic complications or both.

Acknowledgment: This study was supported by a faculty seed fund of the University of Maryland School of Medicine, grant 02390510, and a research grant from Medtronic, Inc. We thank Helen Scott for excellent technical assistance. We thank Dr. Warren K. Laskey for his constructive comments about the manuscript.

FIG. 1.
FIG. 1.:
Effect of Mg supplementation on endothelin-1 level in swine during AMI. *Statistical difference between groups at the same time point. Squares, controls; circles, Mg-early animals; triangles, Mg-late swine.
FIG. 2.
FIG. 2.:
Effect of Mg supplementation on fibronectin level in swine during acute myocardial infarction (AMI). Squares, controls; circles, Mg-early animals; triangles, Mg-late swine. *Statistical difference between groups at the same time point.
FIG. 3.
FIG. 3.:
Effect of Mg supplementation on thromboxane B2 level in swine during acute myocardial infarction (AMI). Squares, controls; circles, Mg-early animals; triangles, Mg-late swine. *Statistical difference between groups at the same time point.
FIG. 4.
FIG. 4.:
Effect of Mg supplementation on prostacyclin level in swine during acute myocardial infarction (AMI). Squares, controls; circles, Mg-early animals; triangles, Mg-late swine.
FIG. 5.
FIG. 5.:
Effect of Mg supplementation on antithrombin-III level in swine during acute myocardial infarction (AMI). Squares, controls; circles, Mg-early animals; triangles, Mg-late swine.
FIG. 6.
FIG. 6.:
Effect of Mg supplementation on total protein S level in swine during acute myocardial infarction (AMI). Squares, controls; circles, Mg-early animals; triangles, Mg-late swine.
FIG. 7.
FIG. 7.:
Effect of Mg supplementation on protein C level in swine during acute myocardial infarction (AMI). Squares, controls; circles, Mg-early animals; triangles, Mg-late swine. *Statistical difference between groups at the same time point.


1. Woods KL, Fletcher S. Long-term outcome after intravenous magnesium sulphate in suspected acute myocardial infarction: the second Liecester Intravenous Magnesium Intervention Trial (LIMIT-2). Lancet 1994;343:816-9.
2. ISIS collaborative groups. 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;345:669-85.
3. Durlach J. Magnesium-deficiency thrombosis [Letter]. Lancet 1967;1:13821.
4. White RE, Hartzell HC. Effects of intracellular free magnesium on calcium current in isolated cardiac myocytes. Science 1988;239:778-80.
5. Hwang D, Yen C, Nadler J. Effect of extracellular magnesium on platelet activation and intracellular calcium mobilization. Am J Hypertens 1992;5:700-6.
6. Herrman RG, Lacefield WB, Crwe VG. Effect of ionic calcium and magnesium on human platelet aggregation. Proc Soc Exp Biol Med 1970;135:100.
7. Herzog WR, Atar D, Gurbel PA, Vogel RA, Schlossberg ML, Serebruany VL. Effect of magnesium sulphate infusion on ex vivo platelet aggregation in swine. Magnes Res 1993;6:349-53.
8. Nadler JL, Buchanan T, Natarajan R, Antonipillai I, Bergman R, Rude RK. Magnesium deficiency produces insulin resistance and increased thromboxane synthesis. Hypertension 1993;21:1024-9.
9. Nadler JL, Goodson S, Rude RK. Evidence that prostacyclin mediates the vascular action of magnesium in humans. Hypertension 1987;9:379-83.
10. Dyck RF, Orchard RC, Senger DL. The response of plasma fibronectin to acute myocardial infarction in humans. Clin Invest Med 1990;13:107-10.
11. Atar D, Serebruany V, Poulton J, Godard J, Schneider A, Herzog WR. Effects of magnesium supplementation in a porcine model of myocardial ischemia and reperfusion. J Cardiovasc Pharmacol 1994;24:603-11.
12. Herzog WR, Atar D. Magnesium and myocardial infarction. Lancet 1994;343:1285-6.
13. Herzog WR, Schlossberg ML, MacMurdy KS, et al. Timing of magnesium therapy affects experimental infarct size. Circulation 1995;92:2622-6.
14. Softeland E, Framstad T, Thorsen T, Holmsen H. Porcine platelets in vitro and in vivo studies: relevance to human thrombosis research. Eur J Haematol 1992;49:161-73.
15. Dodds WJ. The pig model for biomedical research. Fed Proc 1982;41:247-56.
16. Fedor JL, McIntosh DM, Rembert JC, Greenfield JC. Coronary and transmural myocardial blood flow responses in awake domestic pigs. Am J Physiol 1978;235:H435-44.
17. Freedman A, Atrakchi A, Cassidy M, Weglicki W. Magnesium deficiency-induced cardiomyopathy: protection by vitamin E. Biochem Biophys Res Commun 1990;170:1102-6.
18. Kimura T, Yasue H, Sakaino N, Rokutanda M, Jougasaki M, Araki H. Effects of magnesium on the tone of isolated human coronary arteries. Circulation 1989;79:1118-24.
19. Watson KN, Moldow CF, Ogburn PL, Jacob HS. Magnesium sulfate: rationale for its use in preeclampsia. Proc Natl Acad Sci USA 1986;83:1075-8.
20. Laurant P, Moussard C, Albert D, Henry JC, Berthelot A. In vivo and in vitro magnesium effects on aortic prostacyclin generation in DOCA-salt hypertensive rats. Prostaglandins Leukot Essent Fatty Acids 1992;47:183-6.
21. Norton PA, Hynes RO. Characterization of HeLa nuclear factors which interact with a conditionally processed rat fibronectin pre-mRNA. Biochem Biophys Res Commun 1993;195:215-21.
22. Wang L, Asem EK, McLaughlin GL. Calcium enhances Acanthamoeba polyphaga binding to extracellular matrix proteins. In vest Ophthalmol Vis Sci 1994;35:2421-6.
23. Sherill GB, Meade JB, Kalayanamit T, Monroe DM, Church FC. Calcium enhances factor Xa activity independent of gamma-carboxyglutamic acid residues. Thromb Res 1988;52:53-60.
24. Vaziri ND, Kennedy SC, Kennedy D, Gonzales E. Coagulation, fibrinolytic, and inhibitory proteins in acute myocardial infarction and angina pectoris. Am J Med 1992;93:651-7.
25. Willerson JT, Frazier OH. Reducing mortality in patients with extensive myocardial infarction. N Engl J Med 1991;325:2266-8.
26. Sakamoto T, Ogawa H, Yasue H, et al. Prevention of arterial reocclusion after thrombolysis with activated protein C. Circulation 1994;90:427-32.
27. Harris KW, Esmon CT. Protein S is required for bovine platelets to support activated protein C binding and activity. J Biol Chem 1985;260:2007-10.
28. Watanabe T, Suzuki N, Shimamoto N, Fujino M, Imada A. Contribution of endogenous endothelin to the extension of myocardial infarct size in rats. Circ Res 1991;69:370-7.
29. Velasco CE, Jackson EK, Morrow JA, Vitola JV, Inagmi T, Forman MB. Intravenous adenosine suppresses cardiac release of endothelin after myocardial ischemia and reperfusion. Cardiovasc Res 1993;27:121-8.
30. Kemp PA, Gardiner SM, Bennett T, Rubin PC. Magnesium sulphate reverses the carotid vasoconstriction caused by endothelin-I, angiotensin II and neuropeptide-Y, but not that caused by NG-nitroarginine methyl ester, in conscious rats. Clin Sci 1993;85:175-81.
31. Mastrogiannis DS, Kalter CS, O'Brien WF, Carlan SJ, Reece EA. Effect of magnesium sulfate on plasma endothelin-1 levels in normal and preeclamptic pregnancies. Am J Obstet Gynecol 1992;167:1554-9.

Myocardial infarction; Magnesium; Hemostasis; Animal model

© Lippincott-Raven Publishers