Ignite Creation Date:
2024-05-05 @ 10:00 AM
Last Modification Date:
2024-10-26 @ 9:02 AM
Study NCT ID:
NCT00000610
Status:
COMPLETED
Last Update Posted:
2016-04-14
First Post:
1999-10-27
Brief Title:
Magnesium in Coronaries MAGIC
Sponsor:
National Heart Lung and Blood Institute NHLBI
Organization:
National Heart Lung and Blood Institute NHLBI
Study Overview
Official Title:
None
Status:
COMPLETED
Status Verified Date:
2005-03
Last Known Status:
None
Delayed Posting:
No
If Stopped, Why?:
Not Stopped
Has Expanded Access:
False
If Expanded Access, NCT#:
N/A
Has Expanded Access, NCT# Status:
N/A
Acronym:
None
Brief Summary:
To determine whether early intravenous magnesium treatment of patients with suspected acute myocardial infarction reduces mortality
Detailed Description:
BACKGROUND
The management of patients with acute myocardial infarction MI has improved dramatically over the last three decades Advances in the general coronary care unit environment treatment with beta blockers and aggressive attempts at reperfusion have all contributed to a reduction in mortality from acute MI Large randomized trials have demonstrated that aggressive reperfusion strategies in conjunction with aspirin can reduce mortality in patients with suspected acute MI to an average of 65 to 75 percent However the mortality rate remains high in two particular subgroups of patients those who do not receive thrombolysis 115 to 13 percent or those over 65 years who do receive thrombolytics 135 to 24 percent
Recently attention has turned to additional adjunctive pharmacologic treatment with agents such as magnesium nitrates and angiotensin converting enzyme inhibitors to determine their potential for reducing mortality further Of these magnesium appears to be particularly promising It is safe even in the hands of physicians who have no prior experience with it and it is easily administered and readily available in any hospital in the United States Further if it has the expected benefit in the high risk groups described it would become an unusually cost-effective intervention costing less than 2500 per year of life saved
Supplemental administration of magnesium very early after the onset of acute myocardial infarction is supported by abundant data indicating potential cardioprotective effects of magnesium Magnesium is considered to be natures physiologic calcium blocker This is because it protects myocytes against calcium overload by inhibiting calcium influx which is particularly important at the time of reperfusion In experimental models of ischemia and reperfusion agents inhibiting calcium influx improved post-ischemic recovery of mechanical function when given prior to or at the time of reperfusion On the other hand little improvement in mechanical function was observed if such agents were given 15-20 minutes after the onset of reperfusion Reduced serum magnesium may also be responsible for a maladaptive increase in coronary tone and an increased response to vasoconstrictors
In the setting of acute myocardial infarction when increased serum magnesium might be beneficial there is actually a decline in free magnesium This comes about because of a sharp rise in free fatty acids brought about by catecholamine induced lipolysis at the onset of chest pain that results in a completing of magnesium in the form of insoluble soaps Thus although total body magnesium does not decrease magnesium available in a free form that is capable of exerting a cardioprotective effect declines Hence there is a strong theoretical rationale for supplemental magnesium administration in this setting
Since 1984 at least 10 randomized control trials RCTs of intravenous magnesium for acute MI have been reported Several statistical models exist for pooling the data from multiple RCTs in a meta-analysis and estimating the treatment effect of magnesium It is important to review the essential features of these models in order to place the RCT findings in proper perspective The fixed effects model assumes that the RCTs are sampled from a homogenous group of trials Under the homogeneity assumption each RCT provides an estimate of the true treatment effect and differences between the estimates from the various RCTs are due only to experimental error within-trial variability The random effects model assumes the RCTs are heterogeneous and that differences between their estimates of the treatment effect are due both to experimental error within-trial variability and real differences among the trials such as trial design and characteristics of the patients enrolled between-trial variability The random effects model is generally favored since heterogeneity that cannot be explained by experimental error often exists among the RCTs and this model takes such heterogeneity into account in estimation and hypothesis testing Meta-analyses of the seven RCTs published between 1984-199l provided an estimated odds ratio OR for mortality of magnesium treated patients of 044 027-071 using the fixed effects model and 045 023-086 using the random effects model The Leicester Intravenous Magnesium Intervention Trial LIMIT-2 published in 1992 reported a 24 percent reduction in mortality with magnesium treatment P004 confirming the benefit of magnesium in reducing mortality in MI and inspiring many clinicians to advocate magnesium treatment programs in their coronary care units The magnesium treated patients in LIMIT-2 experienced a 25 percent lower incidence of congestive heart failure in the coronary care unit suggesting that magnesium exerts its beneficial effects at least in part via a direct protective action on the myocardium Given the potent predictive power of left ventricular function on survival following MI one would anticipate that magnesium-treated patients in LIMIT-2 would have a lower long-term mortality This hypothesis appears to have been confirmed by the recent long-term follow up report from LIMIT-2 showing a 21percent lower rate of ischemic heart disease related mortality in the magnesium group over a median follow up of 27 years The LIMIT-2 investigators have recently examined the mortality rates over a five year follow up and continue to document the same long term benefit of magnesium administered in the acute phase of infarction The absence of any loss of the mortality benefit of magnesium over the long term is consistent with a significant myocardial protective effect achieved during the critical period of myocardial reperfusion
The results of ISIS-4 seemed to contradict the results of the above studies A total of 58043 patients were enrolled in ISIS-4 29009 allocated to magnesium and 29034 to control There were 2196 deaths 76 percent by 35 days in the magnesium group and 2079 deaths 72 percent in the control group OR 106099-113 suggesting no mortality benefit of magnesium administration and even the possibility of slight harm The findings of ISIS-4 have triggered considerable controversy over the reasons why it produced a null effect for magnesium in reducing mortality in suspected acute MI
When ISIS-4 is added to the preceding eight RCTs the fixed effects model driven heavily by the large sample size of ISIS-4 indicates no beneficial effect of magnesium OR 102 096-109 while the random effects model that takes into account the heterogeneity among these trials suggests that magnesium may reduce mortality OR 069047-102 Thus the random effects model suggests that one must search for possible sources of differences between ISIS-4 and the preceding trials Two important differences that appear to be acting in concert to bias ISIS 4 towards a null effect of magnesium include
1 A low control group mortality rate The control group mortality in ISIS-4 was only 72 percent This was probably the result of the combination of extensive use of thrombolysis 70 percent of patients and antiplatelet agents 94 percent of patients combined with the enrollment of intrinsically low-risk patients only 28 percent were over 70 years of age 17 percent had a history of a prior MI 14 percent had clinical congestive heart failure CHF and 2 percent had systolic blood pressure SBP 100 mm Hg Incremental mortality reducing effects of magnesium are difficult to detect against a low background control mortality rate The inability of ISIS 4 to detect any overall benefit of magnesium also apply to specific subgroups such as the 17325 patients who did not receive thrombolytics ISIS 4 had less than 60 percent power to detect even a 10 percent reduction observed in the 93 percent control mortality in this subgroup A detailed analysis relating the mortality rate in the control group and the treatment effect of magnesium observed in the clinical trials published before ISIS-4 shows clearly that the benefit of magnesium therapy increases as the control group mortality increases Using this relationship it was predicted that trials with a control group mortality rate of about 7 percent would show no benefit of magnesium therapy precisely the result observed in ISIS-4 Of note the LIMIT-2 trial observed a control group mortality of 103 percent that was reduced to 78 percent with magnesium The ISIS-4 control group mortality was thus below that of the magnesium treated group in LIMIT-2
This analysis is consistent with the results of the latest RCT recently reported by Shechter and colleagues They randomized 194 patients with acute MI considered unsuitable for thrombolysis to control N98 or intravenous magnesium N96 In addition to the standard contraindications to lytic therapy reasons for exclusion from thrombolysis included either presentation after six hours andor age greater than 70 years
Shechter et al reported 17 deaths 173 percent in the placebo group and 4 deaths 42 percent in the magnesium group corresponding to an OR of 021 007-064 Consistent with the hypothesis that magnesium helps reduce mortality by a direct myocardial protective effect are the data on the causes of death in this latest study In the placebo group 11 patients died from cardiogenic shock 2 from electromechanical dissociation 2 from myocardial rupture and 1 from cardiac arrest In contrast in the magnesium group 1 patient died from cardiogenic shock 1 from myocardial rupture and 2 from electromechanical dissociation Particularly noteworthy are the findings in the subset of 77 patients over the age of 70 a group expected to have a high short-term mortality from acute MI Indeed 10 of the 44 elderly patients treated with placebo died 23 percent while only 3 of the 33 elderly patients treated with magnesium died 9 percentp009 Also in this especially high risk subgroup the incidence of congestive heart failure was reduced from 25 percent in the placebo patients to 18 percent in the magnesium patients
2 Magnesium was administered late in ISIS-4 The ISIS-4 protocol required that acute phase treatments for MI including lytic therapy were administered prior to randomization and initiation of study drug therapy ie magnesium By design therefore magnesium could not be administered in the early lytic phase eg first hour Although the time from onset of symptoms to randomization was recorded in ISIS-4 time from randomization to actual administration of magnesium was NOT recorded The median time to randomization from the onset of chest pain for all patients was 8 hours in the subset of patients who did not receive thrombolytic therapy 30 percent of trial patients the median time to randomization from onset of chest pain was QUITE prolonged at 12 hours No further details of the distribution of time to randomization have been reported by the ISIS-4 investigators In an effort to answer these concerns they conducted a retrospective survey of a 1000 randomly selected patients This revealed that among those receiving thrombolytics only about 50 percent received magnesium within the two hours following the start of thrombolytic therapy LIMIT-2 randomized patients a median of 3 hours from the onset of chest pain and in Schechters study 19 of nonthrombolytic treated patients the average time from chest pain to initiation of treatment was about seven hours in both the treatment and placebo groups a full five hours earlier than in the non thrombolized group in ISIS-4 Attempts at subgroup analyses in ISIS-4 also suffer from a critical lack of precise information on the actual time of administration of magnesium Thus although no apparent benefit of magnesium was seen in the 23000 patients randomized within six hours of the onset of chest pain or among the 17000 who did not receive thrombolytic therapy including 9000 randomized within 12 hours since most of these patients received magnesium several hours after randomization one cannot be confident that reperfusion pharmacologically induced or spontaneous took place in the presence of a raised serum magnesium level in any subgroup Experimental attempts aimed at ameliorating cellular calcium overload have shown that calcium antagonists must be administered before reperfusion or during a critical window of only a few minutes following reperfusion in order to minimize contractile dysfunction Calcium flux inhibitors such as magnesium administered too late after reperfusion appear to be ineffective
The same observations pertain to the subset of patients alluded to by the ISIS-4 Investigators who were randomized within six hours of symptoms onset and had a high multivariate prognosis score In the absence of details on the timing of administration of magnesium with particular reference to the elapsed time from onset of thrombolytic therapy even in high risk subgroups the findings of the ISIS-4 study remain compatible with the hypothesis that early administration of magnesium particularly before reperfusion occurs is associated with a reduction in mortality from acute myocardial infarction
The implications of these observations are that to prevent calcium overload of reperfused myocytes a loading dose of magnesium should be administered before thrombolytic therapy and during the period when spontaneous reperfusion is most likely to occur in patients not receiving thrombolytics The design of ISIS-4 did not permit these conditions to be met Further prior studies suggest that the low risk profile that characterized the ISIS-4 patients would have been likely to preclude obtaining much additional benefit The results of the small trial reported by Schechter et al strongly support the view that high risk MI patients benefit from early treatment with magnesium infusion However definitive proof requires the larger sample size proposed in MAGIC
DESIGN NARRATIVE
The study was a randomised double-blind trial in 6213 patients with acute ST-elevation myocardial infarction STEMI who were assigned a 2 g intravenous bolus of magnesium sulphate administered over 15 minutes followed by a 17 g infusion of magnesium sulphate over 24 hours n3113 or matching placebo n3100 The primary endpoint was 30-day all-cause mortality At randomisation patients were stratified by their eligibility for reperfusion therapy The first stratum included patients who were aged 65 years or older and eligible for reperfusion therapy and the second stratum included patients of any age who were not eligible for reperfusion therapy Analysis was by intention-to-treat At 30 days 475 153 patients in the magnesium group and 472 152 in the placebo group had died No benefit or harm of magnesium was observed in eight prespecified subgroup analyses of patients and in 15 additional exploratory subgroup analyses After adjustment for factors shown to effect mortality risk in a multivariate regression model no benefit of magnesium was observed
The management of patients with acute myocardial infarction MI has improved dramatically over the last three decades Advances in the general coronary care unit environment treatment with beta blockers and aggressive attempts at reperfusion have all contributed to a reduction in mortality from acute MI Large randomized trials have demonstrated that aggressive reperfusion strategies in conjunction with aspirin can reduce mortality in patients with suspected acute MI to an average of 65 to 75 percent However the mortality rate remains high in two particular subgroups of patients those who do not receive thrombolysis 115 to 13 percent or those over 65 years who do receive thrombolytics 135 to 24 percent
Recently attention has turned to additional adjunctive pharmacologic treatment with agents such as magnesium nitrates and angiotensin converting enzyme inhibitors to determine their potential for reducing mortality further Of these magnesium appears to be particularly promising It is safe even in the hands of physicians who have no prior experience with it and it is easily administered and readily available in any hospital in the United States Further if it has the expected benefit in the high risk groups described it would become an unusually cost-effective intervention costing less than 2500 per year of life saved
Supplemental administration of magnesium very early after the onset of acute myocardial infarction is supported by abundant data indicating potential cardioprotective effects of magnesium Magnesium is considered to be natures physiologic calcium blocker This is because it protects myocytes against calcium overload by inhibiting calcium influx which is particularly important at the time of reperfusion In experimental models of ischemia and reperfusion agents inhibiting calcium influx improved post-ischemic recovery of mechanical function when given prior to or at the time of reperfusion On the other hand little improvement in mechanical function was observed if such agents were given 15-20 minutes after the onset of reperfusion Reduced serum magnesium may also be responsible for a maladaptive increase in coronary tone and an increased response to vasoconstrictors
In the setting of acute myocardial infarction when increased serum magnesium might be beneficial there is actually a decline in free magnesium This comes about because of a sharp rise in free fatty acids brought about by catecholamine induced lipolysis at the onset of chest pain that results in a completing of magnesium in the form of insoluble soaps Thus although total body magnesium does not decrease magnesium available in a free form that is capable of exerting a cardioprotective effect declines Hence there is a strong theoretical rationale for supplemental magnesium administration in this setting
Since 1984 at least 10 randomized control trials RCTs of intravenous magnesium for acute MI have been reported Several statistical models exist for pooling the data from multiple RCTs in a meta-analysis and estimating the treatment effect of magnesium It is important to review the essential features of these models in order to place the RCT findings in proper perspective The fixed effects model assumes that the RCTs are sampled from a homogenous group of trials Under the homogeneity assumption each RCT provides an estimate of the true treatment effect and differences between the estimates from the various RCTs are due only to experimental error within-trial variability The random effects model assumes the RCTs are heterogeneous and that differences between their estimates of the treatment effect are due both to experimental error within-trial variability and real differences among the trials such as trial design and characteristics of the patients enrolled between-trial variability The random effects model is generally favored since heterogeneity that cannot be explained by experimental error often exists among the RCTs and this model takes such heterogeneity into account in estimation and hypothesis testing Meta-analyses of the seven RCTs published between 1984-199l provided an estimated odds ratio OR for mortality of magnesium treated patients of 044 027-071 using the fixed effects model and 045 023-086 using the random effects model The Leicester Intravenous Magnesium Intervention Trial LIMIT-2 published in 1992 reported a 24 percent reduction in mortality with magnesium treatment P004 confirming the benefit of magnesium in reducing mortality in MI and inspiring many clinicians to advocate magnesium treatment programs in their coronary care units The magnesium treated patients in LIMIT-2 experienced a 25 percent lower incidence of congestive heart failure in the coronary care unit suggesting that magnesium exerts its beneficial effects at least in part via a direct protective action on the myocardium Given the potent predictive power of left ventricular function on survival following MI one would anticipate that magnesium-treated patients in LIMIT-2 would have a lower long-term mortality This hypothesis appears to have been confirmed by the recent long-term follow up report from LIMIT-2 showing a 21percent lower rate of ischemic heart disease related mortality in the magnesium group over a median follow up of 27 years The LIMIT-2 investigators have recently examined the mortality rates over a five year follow up and continue to document the same long term benefit of magnesium administered in the acute phase of infarction The absence of any loss of the mortality benefit of magnesium over the long term is consistent with a significant myocardial protective effect achieved during the critical period of myocardial reperfusion
The results of ISIS-4 seemed to contradict the results of the above studies A total of 58043 patients were enrolled in ISIS-4 29009 allocated to magnesium and 29034 to control There were 2196 deaths 76 percent by 35 days in the magnesium group and 2079 deaths 72 percent in the control group OR 106099-113 suggesting no mortality benefit of magnesium administration and even the possibility of slight harm The findings of ISIS-4 have triggered considerable controversy over the reasons why it produced a null effect for magnesium in reducing mortality in suspected acute MI
When ISIS-4 is added to the preceding eight RCTs the fixed effects model driven heavily by the large sample size of ISIS-4 indicates no beneficial effect of magnesium OR 102 096-109 while the random effects model that takes into account the heterogeneity among these trials suggests that magnesium may reduce mortality OR 069047-102 Thus the random effects model suggests that one must search for possible sources of differences between ISIS-4 and the preceding trials Two important differences that appear to be acting in concert to bias ISIS 4 towards a null effect of magnesium include
1 A low control group mortality rate The control group mortality in ISIS-4 was only 72 percent This was probably the result of the combination of extensive use of thrombolysis 70 percent of patients and antiplatelet agents 94 percent of patients combined with the enrollment of intrinsically low-risk patients only 28 percent were over 70 years of age 17 percent had a history of a prior MI 14 percent had clinical congestive heart failure CHF and 2 percent had systolic blood pressure SBP 100 mm Hg Incremental mortality reducing effects of magnesium are difficult to detect against a low background control mortality rate The inability of ISIS 4 to detect any overall benefit of magnesium also apply to specific subgroups such as the 17325 patients who did not receive thrombolytics ISIS 4 had less than 60 percent power to detect even a 10 percent reduction observed in the 93 percent control mortality in this subgroup A detailed analysis relating the mortality rate in the control group and the treatment effect of magnesium observed in the clinical trials published before ISIS-4 shows clearly that the benefit of magnesium therapy increases as the control group mortality increases Using this relationship it was predicted that trials with a control group mortality rate of about 7 percent would show no benefit of magnesium therapy precisely the result observed in ISIS-4 Of note the LIMIT-2 trial observed a control group mortality of 103 percent that was reduced to 78 percent with magnesium The ISIS-4 control group mortality was thus below that of the magnesium treated group in LIMIT-2
This analysis is consistent with the results of the latest RCT recently reported by Shechter and colleagues They randomized 194 patients with acute MI considered unsuitable for thrombolysis to control N98 or intravenous magnesium N96 In addition to the standard contraindications to lytic therapy reasons for exclusion from thrombolysis included either presentation after six hours andor age greater than 70 years
Shechter et al reported 17 deaths 173 percent in the placebo group and 4 deaths 42 percent in the magnesium group corresponding to an OR of 021 007-064 Consistent with the hypothesis that magnesium helps reduce mortality by a direct myocardial protective effect are the data on the causes of death in this latest study In the placebo group 11 patients died from cardiogenic shock 2 from electromechanical dissociation 2 from myocardial rupture and 1 from cardiac arrest In contrast in the magnesium group 1 patient died from cardiogenic shock 1 from myocardial rupture and 2 from electromechanical dissociation Particularly noteworthy are the findings in the subset of 77 patients over the age of 70 a group expected to have a high short-term mortality from acute MI Indeed 10 of the 44 elderly patients treated with placebo died 23 percent while only 3 of the 33 elderly patients treated with magnesium died 9 percentp009 Also in this especially high risk subgroup the incidence of congestive heart failure was reduced from 25 percent in the placebo patients to 18 percent in the magnesium patients
2 Magnesium was administered late in ISIS-4 The ISIS-4 protocol required that acute phase treatments for MI including lytic therapy were administered prior to randomization and initiation of study drug therapy ie magnesium By design therefore magnesium could not be administered in the early lytic phase eg first hour Although the time from onset of symptoms to randomization was recorded in ISIS-4 time from randomization to actual administration of magnesium was NOT recorded The median time to randomization from the onset of chest pain for all patients was 8 hours in the subset of patients who did not receive thrombolytic therapy 30 percent of trial patients the median time to randomization from onset of chest pain was QUITE prolonged at 12 hours No further details of the distribution of time to randomization have been reported by the ISIS-4 investigators In an effort to answer these concerns they conducted a retrospective survey of a 1000 randomly selected patients This revealed that among those receiving thrombolytics only about 50 percent received magnesium within the two hours following the start of thrombolytic therapy LIMIT-2 randomized patients a median of 3 hours from the onset of chest pain and in Schechters study 19 of nonthrombolytic treated patients the average time from chest pain to initiation of treatment was about seven hours in both the treatment and placebo groups a full five hours earlier than in the non thrombolized group in ISIS-4 Attempts at subgroup analyses in ISIS-4 also suffer from a critical lack of precise information on the actual time of administration of magnesium Thus although no apparent benefit of magnesium was seen in the 23000 patients randomized within six hours of the onset of chest pain or among the 17000 who did not receive thrombolytic therapy including 9000 randomized within 12 hours since most of these patients received magnesium several hours after randomization one cannot be confident that reperfusion pharmacologically induced or spontaneous took place in the presence of a raised serum magnesium level in any subgroup Experimental attempts aimed at ameliorating cellular calcium overload have shown that calcium antagonists must be administered before reperfusion or during a critical window of only a few minutes following reperfusion in order to minimize contractile dysfunction Calcium flux inhibitors such as magnesium administered too late after reperfusion appear to be ineffective
The same observations pertain to the subset of patients alluded to by the ISIS-4 Investigators who were randomized within six hours of symptoms onset and had a high multivariate prognosis score In the absence of details on the timing of administration of magnesium with particular reference to the elapsed time from onset of thrombolytic therapy even in high risk subgroups the findings of the ISIS-4 study remain compatible with the hypothesis that early administration of magnesium particularly before reperfusion occurs is associated with a reduction in mortality from acute myocardial infarction
The implications of these observations are that to prevent calcium overload of reperfused myocytes a loading dose of magnesium should be administered before thrombolytic therapy and during the period when spontaneous reperfusion is most likely to occur in patients not receiving thrombolytics The design of ISIS-4 did not permit these conditions to be met Further prior studies suggest that the low risk profile that characterized the ISIS-4 patients would have been likely to preclude obtaining much additional benefit The results of the small trial reported by Schechter et al strongly support the view that high risk MI patients benefit from early treatment with magnesium infusion However definitive proof requires the larger sample size proposed in MAGIC
DESIGN NARRATIVE
The study was a randomised double-blind trial in 6213 patients with acute ST-elevation myocardial infarction STEMI who were assigned a 2 g intravenous bolus of magnesium sulphate administered over 15 minutes followed by a 17 g infusion of magnesium sulphate over 24 hours n3113 or matching placebo n3100 The primary endpoint was 30-day all-cause mortality At randomisation patients were stratified by their eligibility for reperfusion therapy The first stratum included patients who were aged 65 years or older and eligible for reperfusion therapy and the second stratum included patients of any age who were not eligible for reperfusion therapy Analysis was by intention-to-treat At 30 days 475 153 patients in the magnesium group and 472 152 in the placebo group had died No benefit or harm of magnesium was observed in eight prespecified subgroup analyses of patients and in 15 additional exploratory subgroup analyses After adjustment for factors shown to effect mortality risk in a multivariate regression model no benefit of magnesium was observed
Study Oversight
Has Oversight DMC:
Is a FDA Regulated Drug?:
Is a FDA Regulated Device?:
Is an Unapproved Device?:
Is a PPSD?:
Is a US Export?:
Is an FDA AA801 Violation?: