Magnesium Deficiency in Patients With Recent Myocardial Infarction and Provoked Coronary Artery Spasm

Transcription

1 Jpn Circ J 2001; 65: Magnesium Deficiency in Patients With Recent Myocardial Infarction and Provoked Coronary Artery Spasm Shozo Sueda, MD; Hiroshi Fukuda, MD; Kouki Watanabe, MD; Jun Suzuki, MD; Hideyuki Saeki, MD; Takashi Ohtani, MD; Tadao Uraoka, MD* This study sought to clarify the relationship between magnesium (Mg) deficiency and coronary artery spasm provoked by pharmacologic agents in patients with a recent acute myocardial infarction (AMI). Twenty-three consecutive patients suffering from AMI were investigated with a Mg retention test (Mg: 0.1 mmol/kg for 4 h) in both the acute phase (within 1 week (3±2 days) of onset) and the subacute phase (3 4 weeks (24±6 days) of the onset). Early coronary arteriography was performed in all patients. Coronary stenosis in the infarct-related artery was less than 90% in all patients in the subacute phase. The spasm provocation test was performed in the subacute phase and coronary spasm was defined as transient subtotal or total occlusion in association with angina or electrocardiographic ST-segment deviation. Coronary artery spasm was provoked in only 13 of the 23 patients. Compared with the control subjects (12 patients without coronary artery disease or coronary spasm), the 24-h Mg retention was significantly higher in patients with AMI (acute phase: 78±27%, subacute phase: 66±32%, vs control: 48±12%, p<0.05). In the subacute phase, the 24-h Mg retention decreased in patients without coronary spasm (43±26%), but a high level of Mg retention was still observed in patients with coronary spasm (84±25%). There was no difference in the serum concentrations of Mg, calcium and phosphorus between the 2 groups on both phases. In conclusion, both Mg deficiency and provoked coronary artery spasm were noted in more than half of the Japanese patients with a recent AMI, suggesting a close association between Mg deficiency and AMI. (Jpn Circ J 2001; 65: ) Key Words: Acute myocardial infarction; Coronary spasm; Magnesium deficiency Magnesium (Mg) deficiency is one of the most important triggers of coronary artery spasm in patients with vasospastic angina. 1 3 Some researchers have reported that patients with variant angina have a Mg deficiency and intravenous administration of Mg suppresses the anginal attacks. 4 6 Magnesium lowers systemic vascular resistance, dilates the coronary arteries, improves myocardial metabolism and stabilizes cell membranes. 7 Variant angina has been associated with the incidence of myocardial infarction (MI), 8 and in patients who have had an acute MI (AMI), Mg administration improves serious fatal arrhythmias and may thus decrease the mortality rate. 9 Coronary artery spasm is implicated in the genesis of acute coronary syndrome, but no data exist concerning the relationship between Mg deficiency and coronary artery spasm in patients with a recent AMI. In order to clarify the possible role of Mg deficiency and coronary artery spasm in the genesis of AMI, a provocative test for coronary artery spasm was performed during the subacute phase of AMI, and the 24-h Mg retention after a Mg loading test was determined during both the acute and subacute phase in patients with AMI. (Received January 22, 2001; revised manuscript received April 11, 2001; accepted April 17, 2001) Department of Cardiology, Saiseikai Saijo Hospital and *Department of Cardiology, Kita Medical Association Hospital, Ehime, Japan Mailing address: Shozo Sueda, MD, Department of Cardiology, Saiseikai Saijo Hospital, Tsuitachi, Saijo City, Ehime , Japan Methods Study Patients Twenty-three consecutive patients (19 men, 4 women; mean age, 64±7 years) with a recent AMI comprised the study group. All patients suffered from AMI and Killip I, and those whose AMI had been complicated with heart failure were excluded. The value of mean creatine kinase was 2,957±2,014 IU/L. Early coronary arteriography was performed during the acute phase in all 23 patients. Conventional balloon angioplasty was performed in 6 patients, intracoronary thrombolysis in 8 and the remaining 9 patients had conservative therapy. The infarct-related artery was the right coronary artery in 7 patients, the left circumflex artery in 3 and the left anterior descending artery in 13 patients. In the subacute phase, 18 patients had significant organic stenosis (>75% <90% luminal narrowing): 14 with 1-vessel and 4 with 2-vessel disease. The remaining 5 patients had no significant organic stenosis. In all patients, coronary stenosis in the infarct-related artery was less than 90%. Three patients had a history of old MI and 8 patients had post infarction angina attacks. As previously reported, subjects were excluded and the spasm provocation test was not performed if patients had left main narrowing (>50%), 3-vessel disease, 2-vessel disease with total occlusion, heart failure (New York Heart Association functional class III or IV), or renal failure (creatinine >2.0 mg/dl), or if isosorbide dinitrate was initially used to relieve spasm in the coronary artery tested. During this study, neither death nor other complications occurred in the AMI patients. Twelve patients (9 men, 3 women; mean age, 58±5 years)

2 644 SUEDA S et al. Table 1 Patient Characteristics AMI (n=23) Control (n=12) Male gender 19 (83) 9 (75) Age (years) 64±7 58±5 Hypertension 14 (61)* 3 (25) Smoking 18 (78)** 4 (33) Hyperlipidemia 9 (39) 4 (33) Diabetes mellites 5 (22) 3 (25) T cholesterol (mg/dl) 186±28 176±22 Triglyceride (mg/dl) 125±43 109±35 HDL-cholesterol (mg/dl) 37±9 42±8 Organic stenosis ( 75%) 18 (78)* 0 (0 VD: 1 VD: 2 VD) 5 : 14 : 4 Pharmacologic spasm provocation test ACh : ER : both 14 (61) : 7 (30) : 2 (9) 8 (66) : 2 (17) : 2 (17) Spasm provoked vessel 16 0 Right coronary artery 4 0 Left circumflex artery 3 0 Left anterior descending artery 9 0 Old MI 3 (13) 0 Post infarction angina 8 (35) 0 Medication Ca-channel blocker 23 (100)** 5 (42) ISDN 23 (100)** 4 (33) ACE inhibitor 2 (9) 0 -blocker 0 0 Aspirin 23 (100)** 2 (17) Statin 3 (13) 1 (8) AMI, acute myocardial infarction; HDL, high-density lipoprotein; VD, vessel disease; ACh, acetylcholine; ER, ergonovine; ISDN, isosorbide dinitrate; ACE, angiotensin converting enzyme. *p<0.05, **p<0.01 vs Control. Values in parenthesis are %. without coronary artery disease or provoked coronary vasospasm were matched for age and sex with the AMI patient group as control subjects. All 12 patients were studied because of chest pain or discomfort at rest and/or during exercise, and all 12 had less than 75% luminal narrowing and were diagnosed with atypical chest pain. The clinical data of the patients with AMI and control subjects are listed in Table1. For at least 1 month before the study, none of the patients had taken diuretics, digitalis, -blockers or other agents, such as gentamicin or carbenicillin, that would affect Mg metabolism. Written informed consent was obtained from all patients before the study, and the protocol was in agreement with the guidelines of the institutional ethical committee. Magnesium Loading Test All the study participants were allowed to eat their regular diet during hospital stay. In the group with AMI, Mg loading tests were performed during both the acute (within 1 week (3±2 days) of the onset) and subacute phase (3 4 weeks (24±6 days) of the onset). All medications, including sublingual nitroglycerin, were continued during the Mg loading test. In the control group, the Mg loading tests were performed at least 1 week after admission. No conditions, including heart failure, intestinal malabsorption, loss of body fluids, liver cirrhosis, renal disease, or alcoholism (chronic excessive alcohol intake >81g/day), that potentially lead to Mg depletion were found in any of the study subjects. The Mg loading test was based on the method described by Ryzen et al. 10 Twenty-four-hour urine samples were collected to determine Mg excretion and creatinine clearance, after which subjects were given 0.1 mmol/kg bodyweight of Mg sulfate in 50ml of 5% glucose intravenously over 4 h (between h and h). Blood samples were obtained for determination of serum Mg, calcium, and phosphorus before and immediately after completion of the Mg sulfate infusion. Twenty-four-hour urine samples, beginning with the infusion, were then collected to determine urine Mg. The serum and urine Mg concentrations were determined with an atomic absorption spectrophotometer, and the 24-h Mg retention (%) was calculated according to the formula: 24-h Mg retention (%) = (1 (postinfusion urine Mg preinfusion urine Mg)/total infusion of Mg) 100 Spasm Provocation Test Repeat coronary arteriography was performed from 10.00h to 12.00h using the Sones technique while patients were in the fasting state under no medication for 24 h, other than nitroglycerine during subacute phase. After baseline coronary arteriograms were obtained, an intracoronary injection of acetylcholine or ergonovine was performed: acetylcholine was injected in incremental doses of 20, 50, and 80 g into the right coronary artery and 20, 50, and 100 g into the left coronary artery over 20s, and ergonovine was administered in total doses of 40 g into the right coronary artery and 64 g into the left coronary artery over 4 min. A USCI bipolar electrode catheter was inserted into the right ventricular apex through the femoral or antecubital vein and was connected to a temporary pacemaker, setting the rate at 45beats/min during acetylcholine testing. After the spasm provocation test was completed, intracoronary isosorbide dinitrate (5.0 mg) was administered, and coronary arteriography was again performed in multiple projections. The coronary arteriograms were analyzed separately by 2 independent observers and their findings were classified according to the reporting system of the American Heart

3 Magnesium Deficiency and Coronary Spasm 645 Table 2 Serum Electrolyte Levels in Patients With AMI and Control Subjects AMI Acute phase Subacute phase Control Mg (before) (mg/dl) 2.1± ± ±0.2 Mg (after) 2.8± ± ±0.2 Mg (next day) 2.2± ± ±0.1 Ca (before) (mg/dl) 8.5± ± ±0.4 Ca (after) 8.4± ± ±0.2 Ca (next day) 8.4± ± ±0.4 P (before) (mg/dl) 3.2± ± ±0.6 P (next day) 3.2± ± ±0.6 Cr (mg/dl) 0.9± ± ± h CCr (ml/min) 86.4± ± ±23.1 AMI, acute myocardial infarction; Mg, magnesium; Ca, calcium; P, phosphorus; Cr, creatinine; CCr, creatinine clearance. *p<0.05 vs control. Association. 16 The percentage of luminal diameter narrowing of coronary arteries was measured by an automatic edge-contour detection computer analysis system (CARDIO 500, Kontron Instrument, Tokyo, Japan). The size of the coronary catheter was used to calibrate the image in millimeters and the measurement was performed in the same projection at each stage. Positive coronary spasm was defined as more than 90% transient luminal narrowing demonstrated angiographically in association with either anginal attacks or ischemic ST-segment deviation, or both. All data are shown as mean ± SEM. The statistical significance of differences between 2 groups was tested by Student s unpaired t test. One-way analysis of variance followed by Fisher s least significant difference test were performed to determine differences among 3 groups. A p value <0.05 was considered statistically significant. Results Clinical Characteristics, Serum Electrolyte, and 24-h Mg Retention Rate Between Patients and Controls The prevalence of hypertension and a history of smoking in the AMI patients was significantly higher than in the control subjects, but the prevalence of hyperlipidemia and diabetes mellites did not differ between the 2 groups (Table 1). The percentage of those receiving calcium-channel antagonists, isosorbide dinitrate and aspirin in the AMI Table 3 Comparison of 24-h Mg Retention Rates Between Patients With AMI and Control Subjects Variables 24-h Mg retention rate (%) Control (n=12) 48±12 AMI (n=23) Acute phase 78±27* Subacute phase 66±32* AMI, acute myocardial infarction; Mg, magnesium. *p<0.05 vs control. patients with positive spasm was significantly higher than in those without provoked spasm. Serum calcium, phosphorus, and creatinine concentrations were within normal ranges for all the study subjects. As shown in Table 2, there were no significant differences in baseline serum Mg concentration between the patient group with AMI and the control group (2.1±0.2 (acute), 2.2±0.2 (subacute) vs 2.1± 0.1 mg/dl, p=ns), but patients with AMI retained significantly more Mg than did control subjects (78±27 (acute), 66±32 (subacute) vs 48±12%, p<0.05) (Table3). Comparisons Between AMI Patients With and Without Provoked Spasm Patients with AMI were further divided into 2 groups: 13 patients with provoked coronary artery spasm and 10 patients without spasm during the subacute phase. As shown Table 4 Clinical Characteristics of AMI Patients With and Without Spasm AMI spasm (+) AMI spasm ( ) (n=13) (n=10) Male gender 11 (85) 8 (80) Age (years) 66±7 62±8 Hypertension 7 (54) 7 (70) Smoking 10 (77) 8 (80) Hyperlipidemia 4 (31) 5 (50) Diabetes mellites 2 (15) 3 (30) Organic stenosis ( 75%) 11 (85) 7 (70) (0 VD: 1 VD: 2 VD) 2 : 8 : 3 3 : 6 : 1 PTCA : ICT : C 2 (15) : 6 (46) : 5 (38) 4 (40) : 2 (20) : 4 (40) Old MI 2 (15) 1 (10) Post infarction angina 7 (54) 1 (10) Peak creatine kinase (IU/L) 3,054±2,186 2,830±1,874 Left ventricular ejection fraction (%) 54±12 52±6 Pharmacologic spasm provocation test ACh : ER : both 8 (62) : 4 (31) : 1 (8) 6 (60) : 3 (30) : 1 (10) AMI, acute myocardial infarction; VD, vessel disease; PTCA, percutaneous transluminal coronary angioplasty; ICT, intracoronary thrombolysis; C, conservative; ACh, acetylcholine; ER, ergonovine. Values in parenthesis are %.

4 646 SUEDA S et al. in Table 4, there was no significant difference in the peak value of creatine kinase (3,054±2,186 vs 2,830±1,874IU/L, p=ns) or the left ventricular ejection fraction (54±12% vs 52±6%, p=ns) between the 2 groups. Nor was there a significant difference between the 2 groups regarding the other items. Patients with coronary vasospasm retained more Mg than those without spasm (84±25% and 43±26%, p<0.05) during the subacute phase, whereas in the acute phase no significant difference existed between the 2 groups (82±25% and 73±29%, NS; Fig1). However, there was no significant difference in the Mg retention of patients without coronary vasospasm in the subacute phase and that of control subjects. No difference in Mg levels existed between the 2 groups in both phases (Table 5). Moreover, there was no significant difference in other electrolyte levels. Coronary spastic angina and old MI were each observed in 2 AMI patients with positive spasm, but only one of the AMI patients without a provoked spasm had a history of old MI. Of the 13 patients with a positive response, 12 (92%) had coronary artery spasm in the infarct-related artery. Spasm was provoked in 16 vessels: 12 in the infarct-related artery and 4 in non-infarct-related vessels. Post infarction angina was determined in 7 (54%) of the 13 AMI patients with a positive spasm reaction, whereas only 1 patient (10%) had an angina attack after AMI in the 10 patients without a provoked spasm (Table 4). No significant difference in the 24-h Mg retention rate existed between patients with (acute: 83±25%, subacute: 77±31%) or without (75± 28, 60±32%) post infarction angina attacks. Comparison Among Acute Therapy Table6 shows that there was no difference in the 24-h Mg retention rates of the 3 groups, and, moreover, the electrolyte levels measured during both phases did not differ among the groups. Comparison Between AMI Patients With and Without Organic Stenosis in the Subacute Phase There was no difference in the 24-h Mg retention rate and the serum electrolyte levels between patients with and without organic stenosis in both phases (Table6). Fig 1. Twenty-four-hour Mg retention (%) in AMI patients with and without spasm. There is no difference in the 24-h Mg retention rate between the acute and subacute phases in AMI patients with a positive spasm response (n=13), whereas in AMI patients without a provoked spasm (n=10) the 24-h Mg retention rate during the subacute phase is significantly decreased compared with during the acute phase. Discussion The Mg loading test revealed that high levels of Mg retention existed in patients with AMI in both the acute and subacute phase, and furthermore, the 24-h retention of Mg was much greater in those with a provoked coronary vasospasm in the subacute phase. There was no significant difference in the 24-h Mg retention test between AMI patients without a provoked coronary vasospasm in the Table 5 Serum Electrolyte Levels in AMI Patients With and Without Provoked Coronary Artery Spasm AMI spasm (+)* AMI spasm ( )* (n=13) (n=10) Acute phase Subacute phase Acute phase Subacute phase Mg (before) 2.2± ± ± ±0.2 Mg (after) 2.9± ± ± ±0.3 Mg (next day) 2.2± ± ± ±0.3 AMI, acute myocardial infarction; Mg, magnesium. *Duration (days) for AMI (+) was 21±6 and 20±4 for AMI ( ). Table 6 Comparison of 24-h Mg Retention Rate Among Acute Therapies and Between Patients With and Without Organic Stenosis No. of patient Positive spasm 24-h Mg retention rate (%) Acute phase Subacute phase PTCA 6 2 (33%) 72±33 55±47 Intracoronary thrombolysis 8 6 (75%) 89±22 71±33 Conservative treatment 9 5 (56%) 72±26 68±27 Organic stenosis in subacute phase Without 5 2 (40%) 66±32 50±42 With (61%) 80±25 70±29 PTCA, percutaneous transluminal coronary angioplasty.

5 Magnesium Deficiency and Coronary Spasm subacute phase and control subjects. These findings suggest that Mg deficiency may play a role in the occurrence of AMI by means of coronary artery spasm. Mg Deficiency and Coronary Vasospasm Goto et al reported Mg deficiency in patients with variant angina 1 and recently, Satake et al reported that the 24-h Mg retention rate and intracellular concentrations of Mg in mononuclear cells and erythrocytes correlated well with the frequency of anginal attacks in patients with variant angina. A previous study revealed that Mg deficiency was one of the cause of coronary artery spasm, and clinical studies have also demonstrated that intravenous administration of Mg to patients with variant angina suppresses hyperventilation-induced anginal attacks, 4 exercise-induced angina, 5 and, in part, accelerated exercise following mild hyperventilation-induced angina 6 with ST-segment elevation. Mg deficiency has been detected in patients with AMI, as well as in those with coronary spastic angina. However, it is unknown whether the mechanism of the deficiency in patients with AMI is similar to that in patients with coronary spastic angina. Most likely, the pathogenesis of coronary spasm is multifactorial. AMI and Coronary Vasospasm Coronary vasospasm is one of the most important triggers of acute coronary syndrome and according to previous reports, 18,19 the frequency of provoked coronary artery spasm in patients with AMI is significantly higher than in controls. Thus, patients with AMI had higher coronary vasoactivity. Moreover, the incidence of coronary artery spasm provoked with intracoronary acetylcholine is 2 3 times higher in Japanese patients with a recent MI than in Caucasian patients. In the present study, coronary spasm was observed in 57% (13/23) of patients with a recent AMI, which is a similar frequency of coronary spasm to that reported in other studies (69 80%) AMI and Mg Deficiency It has been reported that patients with AMI have a transient hypomagnesemia and an intravenous Mg-loading test revealed a Mg deficiency compared with control subjects Elevated levels of catecholamines and the subsequent changes in the renal handling of Mg could lead to Mg deficiency in patients with AMI, as well as in those with chronic heart failure associated with high levels of norepinephrine. 21,23 Mg deficiency was found in the present patients with AMI in both the acute and subacute phases. Elevated catecholamines may be one of the causes of Mg deficiency in the acute phase, but for the subacute phase, the mechanism of Mg deficiency is controversial. Although a causal relation between Mg deficiency, coronary artery spasm and AMI remains to be elucidated, our study clearly demonstrated that coronary spasm was provoked in association with Mg deficiency in patients with a recent MI. Clinical Implications Mg deficiency may play a role in the induction of coronary vasospasm during the acute phase in some patients with AMI. Therefore, it is beneficial to administer Mg to reduce coronary vasoactivity in the acute phase. In previous large studies of Caucasian patients, there was no clinical evidence for the routine use of Mg for patients with AMI, but intravenous administration of Mg may have a different result in Japanese patients with AMI because of 647 their greater incidence of coronary artery spasm. 26 Some conclusions can be drawn from the results of the present study. First, -blocker should be administered prudently in recent AMI patients with spasms provoked during the acute and subacute phase, because -blockers may aggravate coronary arterial spasm. Second, not only aspirin but also per os or intravenous Mg may be helpful in some oriental patients experiencing an AMI. The prognosis did not differ between the 2 groups during the hospital stay. Moreover, the prevalence of post infarction angina, a history of old MI, or a history of coronary spastic angina did not differ in the AMI patients with and without a provoked spasm. In AMI patients with a positive spasm response during the acute and subacute phases, not only calcium-channel blockers but also intravenous or per os Mg may be effective in reducing coronary vasoactivity. Study Limitations This study has 4 limitations. First, we did not measure the concentration of serum cathecolamines in the acute and subacute phases. We excluded patients with heart failure or those requiring catecholamine administration as part of their therapy. All patients had an AMI classified into Killip I. Second, we ignored the dietary intake of Mg and the administration of calcium antagonists. Alcohol intake and smoking were discontinued during the hospital stay and did not occur during the Mg loading test. Third, the study included a small number of patients, and fourth, we did not measure the intracellular Mg concentration. Only 1% of the total body Mg exists in the extracellular space. In conclusion, there may be a close association between Mg deficiency and provoked coronary artery spasm in half of the Japanese patients with recent AMI during the subacute phase. Further studies are needed to clarify the correlation between Mg deficiency and coronary spasm. Acknowledgments We thank Yuji Shigematsu, MD, Associate Professor Mareomi Hamada, MD, and Professor Kunio Hiwada, MD, for their helpful comments. References 1. Goto K, Yasue H, Okumura K, Matuyama K, Kugiyama K, Miyagi H, et al: Magnesium deficiency detected by intravenous loading test in variant angina pectoris. Am J Cardiol 1990; 65: Miwa K, Igawa A, Miyagi Y, Fujita M: Importance of magnesium deficiency in alcohol-induced variant angina. Am J Cardiol 1994; 73: Igawa A, Miwa K, Miyagi Y, Fujita M, Inoue H: Comparison of frequency of magnesium deficiency in patients with vasospastic angina and fixed coronary artery disease. Am J Cardiol 1995; 75: Miyagi H, Yasue H, Okumura K, Ogawa H, Goto K, Oshima S: Effect of magnesium on anginal attack induced by hyperventilation in patients with variant angina. Circulation 1989; 79: Kugiyama K, Yasue H, Okumura K, Goto K, Minoda K, Miyagi H, et al: Suppression of exercise-induced angina by magnesium sulfate in patients with variant angina. J Am Coll Cardiol 1988; 12: Sueda S, Saeki H, Otani T, Mineoi K, Kondo T, Yano K, et al: Limited efficacy of magnesium for the treatment of variant angina. J Cardiol 1999; 34: Turlapaty PDMV, Altura BM: Magnesium deficiency produces spasms of coronary arteries: Relationship to etiology of sudden death ischemic heart disease. Science 1980; 208: Madias JE: The syndrome of variant angina culminating in acute myocardial infarction. Circulation 1979; 59: Horner SM: Efficacy of intravenous magnesium in acute myocardial infarction in reducing arrhythmias and mortality: Meta-analysis of magnesium in acute myocardial infarction. Circulation 1992; 86:

6 648 SUEDA S et al. 10. Ryzen E, Elbaum N, Singer FR, Rude RK: Parenteral magnesium tolerance testing in the evaluation of magnesium deficiency. Magnesium 1985; 4: Sueda S, Saeki H, Otani T, Ochi N, Kukita H, Kawada H, et al: Investigation of the most effective provocation test for patients with coronary spastic angina: Usefulness of accelerated exercise following hyperventilation. Jpn Circ J 1999; 63: Sueda S, Saeki H, Otani T, Mineoi K, Kondou T, Yano K, et al: Major complications during spasm provocation tests with an intracoronary injection of acetylcholine. Am J Cardiol 2000; 85: Sueda S, Suzuki J, Watanabe K, Mineoi K, Kondou T, Yano K, et al: Clinical characteristics of female patients with coronary spastic angina: Comparison with male patients. Jpn Circ J 2000; 64: Sueda S, Ochi T, Yano K, Mineoi K, Kondou T, Ochi N, et al: New combined spasm provocation test in patients with rest angina: Intracoronary injection of acetylcholine after intracoronary administration of ergonovine. Jpn Circ J 2000; 64: Sueda S, Fukuda H, Watanabe K, Ochi N, Kawada H, Hayashi Y, et al: Usefulness of accelerated exercise following mild hyperventilation for the induction of coronary artery spasm: Comparison with an acetylcholine test. Chest 2001; 119: AHA Committee report: A reporting system on patients evaluated for coronary artery disease. Circulation 1975; 51: Satake K, Lee JD, Shimizu H, Ueda T, Nakamura T: Relation between severity of magnesium deficiency and frequency of anginal attcks in men with variant angina. J Am Coll Cardiol 1996; 28: Beltrame JF, Sasayama S, Maseri A: Racial heterogeneity in coronary artery vasomotor reactivity: Differences between Japanese and Caucasian patients. J Am Coll Cardiol 1999; 33: Okumura K, Yasue H, Matsuyama K, Ogawa H, Morikami Y, Obata K, et al: Effect of acetylcholine on the highly stenotic coronary artery: Difference between the constrictor response of the infarct-related coronary artery and that of the noninfarct-related artery. J Am Coll Cardiol 1992; 19: Pristipino C, Beltrame JF, Finocchiaro ML, Hattori R, Fujita M, Mongiardo R, et al: Major racial differences in coronary constrictor response between Japanese and Caucasians with recent myocardial infarction. Circulation 2000; 101: Rasmussen HS, Aurap P, Hojberg S, Jensen EK, McNair P: Magnesium and acute myocardial infarction: Transient hypomagenesemia not induced by renal magnesium loss in patients with acute myocardial infarction. Arch Intern Med 1986; 146: Rasmussen HS, McNair P, Goransson L, Balslov S, Larsen OG, Anrup P: Magnesium deficiency in patients with ischemic heart disease with and without acute myocardial infarction uncovered by an intravenous loading test. Arch Intern Med 1988; 148: Madias JE, Sheth K, Choudry MA, Berger DO, Madias NE: Admission serum magnesium level does not predict the hospital outcome of patients with myocardial infarction. Arch Intern Med 1996; 156: ISIS-4 Collaborative Group: ISIS-4: A randomized factorial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in patients with suspected acute myocardial infarction. Lancet 1995; 345: 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; 343: Sueda S, Ochi N, Kawada H, Matsuda S, Hayashi Y, Tsuruoka T, et al: Frequency of provoked coronary vasospasm in patients undergoing coronary arteriography with spasm provocation test of acetylcholine. Am J Cardiol 1999; 83: