Therapeutic Potential of Magnesium in the Treatment of Acute Stroke Keith W. Muir, md, mrcp
Magnesium is a key cation in multiple biological processes, including membrane excitability, protein synthesis, and cellular bioenergetics. Parenterally administered magnesium crosses the blood-brain barrier, raises brain concentrations to supraphysiological levels, and is neuroprotective in preclinical models of cerebral and spinal cord ischemia, excitotoxic injury, and head trauma. Neuronal and vascular effects of therapeutic magnesium may be pertinent, including inhibition of presynaptic release of excitatory neurotransmitters, presynaptic potentiation of adenosine, blockade of the N-methyl d-aspartate receptor, block of voltage-gated calcium channels, relaxation of vascular smooth muscle with vasodilatation of large and small vessel vascular beds causing increased cerebral blood flow, antagonism of endothelin-1 and other vasoconstrictors, enhanced postischemic recovery of tissue adenosine triphosphate, buffering of intracellular calcium ions (especially antagonism of mitochondrial calcium entry), and inhibition of deleterious ion shifts in white matter. Wide clinical experience in obstetrics and myocardial infarction confirms safety and tolerability as a therapeutic agent. Clinical trials in stroke are ongoing following encouraging results of pilot studies. Key Words: Magnesium— Stroke—Cerebrovascular disease—Clinical trials—Pharmacology.
Multiple metabolic and neurochemical events influence the viability of ischemic neurones in experimental models of stroke, and intervention in many pathways is capable of tipping the balance in favor of neuronal survival. Although many agents that act by single, welldefined mechanisms have potent neuroprotective effects in animal models, these have not yet translated into benefits in clinical practice. Magnesium (Mg2⫹) is neuroprotective in preclinical models of cerebral ischemia both in vitro and in vivo. There are multiple potential sites at which Mg2⫹ may act, encompassing both neuronal and
From the Department of Neurology, Institute of Neurological Sciences, Southern General Hospital, Glasgow, Scotland. Received April 19, 2000; accepted September 30, 2000. K.W.M. is coprincipal investigator in the Intravenous Magnesium Efficacy in Stroke (IMAGES) trial, which is funded by the UK Medical Research Council. Address reprint requests to Keith W. Muir, MD, MRCP, Department of Neurology, Southern General Hospital, Glasgow G51 4TF, Scotland. Copyright © 2000 by National Stroke Association 1052-3057/00/0906-0008$3.00/0 doi:10.1053/jscd.2000.20669
vascular mechanisms of relevance to acute stroke. Unlike most synthetic neuroprotective compounds, parenteral magnesium has no major adverse effects in doses that achieve serum levels in the range of preclinical neuroprotective concentrations. Extensive clinical experience with magnesium sulfate (MgSO4) in obstetrics confirms central nervous system (CNS) penetrance, tolerability, safety, and clinical effects of supraphysiological serum concentrations of Mg2⫹. Clinical studies in myocardial infarction (MI) confirm the safety and tolerability of Mg2⫹ in therapeutic doses in populations with acute cardiovascular disease. This paper reviews the pharmacological properties of magnesium of potential relevance to its neuroprotective properties and summarizes the preclinical and early clinical evidence of its potential in acute stroke treatment.
Distribution and Biological Functions of Mg2ⴙ in the CNS Magnesium ions (Mg2⫹) are involved in many essential biochemical processes and act especially as cofactors in protein and energy synthesis.1 In the CNS, Mg2⫹ is
Journal of Stroke and Cerebrovascular Diseases, Vol. 9, No. 6 (November-December), 2000: pp 257-267
257
KEITH W. MUIR
258
predominantly intracellular, and approximately 80% bound to adenosine triphosphate (ATP). Neuronal free Mg2⫹ concentration is proportional to that of ATP and is influenced by intracellular pH, since hydrogen ions compete with Mg2⫹ for binding to phosphates. Magnesium may contribute to regulation of ATP concentrations. Active transport of Mg2⫹ across the blood-brain barrier2 maintains a concentration gradient between cerebrospinal fluid (CSF) and serum, such that [Mg2⫹] in CSF is approximately 1.1 mmol/L, and in serum is around 0.8 mmol/L, except in disease states that compromise the blood-brain barrier. In dietary magnesium deficiency, CSF [Mg2⫹] is maintained preferentially. Significant reductions in extracellular fluid and total brain Mg2⫹ have been found in animal models of traumatic brain injury and stroke.3,4 Phosphorus magnetic resonance spectroscopy in stroke patients indicates increased intracellular free [Mg2⫹] in the acute phase of stroke (up to day 5), correlating with the period of intracellular acidosis, and probably consequent to displacement of ATP-bound Mg2⫹ together with competition with hydrogen ions.5 Excitotoxicity Synaptic accumulation of neurotoxic glutamate concentrations and pathological stimulation of postsynaptic N-methyl d-aspartate (NMDA) and ␣-amino-3-hydroxy5-methylisoxazole-4-propionic acid (AMPA) receptors are thought to be key events in the evolution of ischemic neuronal cell death. The activated NMDA receptor ion channel conducts both sodium (Na⫹) and Ca2⫹ and NMDA stimulation is coupled to sustained intracellular free Ca2⫹ increases that trigger deleterious intracellular events. Pharmacological inhibition of glutamate release, or of NMDA or AMPA ion channel opening, is consistently neuroprotective in preclinical focal ischemia models, irrespective of mechanism. High concentrations of Mg2⫹ (around 10 mmol/L) block synaptic activity nonspecifically, thereby preventing death and functional impairment of neurones in primary culture or isolated brain slices,6-8 but are unattainable therapeutically. Physiological extracellular Mg2⫹ concentrations (250 mol/L to 1 mmol/L) inhibit glutamate release.9 Adenosine-mediated inhibition of glutamate release via A1 receptors is potentiated by Mg2⫹ and may be magnesium-dependent.10 Physiological Mg2⫹ concentrations specifically antagonize NMDA receptor activation via a voltage-dependent ion channel block that must be overcome by membrane depolarization.11,12 At resting membrane potentials, current flow through the NMDA receptor is completely abolished by physiological Mg2⫹ concentrations. The behavior of the NMDA ionophore in patch-clamp indicates
2⫹
the likelihood of two distinct Mg binding sites within the ion channel,13 with the more external Mg2⫹ site overlapping the binding site for synthetic uncompetitive NMDA antagonists (ion channel blockers).14 In heteromeric receptors expressed in Xenopus oocytes, the susceptibility of NMDA receptor proteins to Mg2⫹ block is dependent on the presence of asparagine residues within the M2 (channel-lining) domain of the fundamental NR1 subunit, but especially of potentiating NR2 subunits.15-17 Additional amino acid groups further modulate Mg2⫹ sensitivity of different NR2 subtypes.18 There is some evidence that different modulatory sites on the NMDA receptor complex influence different intracellular signalling pathways: in cortical neuronal culture, extracellular Mg2⫹ (in micromolar concentrations) inhibited NMDAinduced Ca2⫹ increases but not cyclic guanosine monophosphate (cGMP).19 Receptor sensitivity to Mg2⫹ is also influenced by prior ischemic conditioning and neuronal maturity. Increasing extracellular [Mg2⫹] to supraphysiological concentrations noncompetitively antagonizes NMDA conductance.20 The toxicity of NMDA applied directly to neuronal culture is attenuated by physiological levels of extracellular Mg2⫹,19,21 whereas reduction of extracellular [Mg2⫹] enhances toxicity.22 Loss of NMDA receptor sensitivity to Mg2⫹ block may be a major factor in the delayed ischemic death of CA1 pyramidal cells.23 In addition to the inhibitory effect of extracellular Mg2⫹ on NMDA channel opening, intracellular Mg2⫹ also reduces NMDA ionic conductance at physiological concentrations.13,24 NMDA receptor activation causes a 20-fold increase in intracellular [Mg2⫹].25 The elevation of intracellular free Mg2⫹ is Ca2⫹-dependent and has two components, one consequent to shifts of intracellular Mg2⫹ (possibly due to release from Mg-ATP) and the other to entry of extracellular Mg2⫹. Changes in extracellular [Mg2⫹] may thus reduce NMDA channel opening by raising intracellular as well as extracellular [Mg2⫹]. Intracellular [Mg2⫹] increases are sufficiently large to block L-type voltage-gated calcium channels (VGCCs), as well as several voltage-gated potassium and Na⫹ channels.25 Systemic Mg2⫹ administration alters NMDA receptor affinity. In rats receiving MgSO4 270 mg/kg every 4 hours by intraperitoneal (IP) injection, radiolabeled CGP 39653 (an NMDA receptor glutamate recognition site ligand) binding was reduced by 50% compared with controls after 24 hours.26 The mechanism is unclear, but Mg2⫹ potentiates glycine binding to the NMDA receptor at depolarizing membrane potentials and inhibits binding at resting potentials,27 possibly through interactions with the polyamine modulatory site that are subtypespecific.
MAGNESIUM FOR STROKE
Parenteral MgSO4 inhibits epileptic electroencephalographic activity induced by cortical penicillin in dogs, cats, and primates in a dose-dependent fashion.28 In rats, MgSO4 at 1 mmol/kg IP raised CSF [Mg2⫹] by 26% (from 0.99 to 1.25 mmol/L) and raised the threshold for seizures induced via stimulation of implanted hippocampal electrodes.29 The mechanism for changes in seizure threshold may involve modulation of NMDA-mediated events, since intravenous (IV) MgSO4 at 90 mg/kg significantly delayed and shortened seizures induced by direct intracerebroventricular injection of NMDA in rats30 (and also prevented mortality, which was 50% in controls). In a study of oral organic Mg2⫹ salts in mice, NMDA and strychnine-induced seizures were delayed by doses equivalent to 300 mg/kg Mg2⫹ (doubling serum [Mg2⫹]), results qualitatively similar to those obtained with the uncompetitive NMDA antagonist dizocilpine (MK 801).31 Non-Excitotoxic Neuronal Effects of Magnesium Increased free intracellular Ca2⫹ may result from extracellular calcium entry via VGCCs or receptor-operated Ca2⫹ channels (NMDA and AMPA receptors possessing the GluR2 protein subunit) and intracellular release from endoplasmic reticulum. Magnesium competes with Ca2⫹ at VGCCs on both intracellular and cell surface membranes and is an antagonist at N-, P-, and (at high concentrations) L-type VGCCs.32 It may thereby impede Ca2⫹-dependent presynaptic glutamate release and VGCC-mediated Ca2⫹ entry to ischemic neurones. Excessive Ca2⫹ entry to mitochondria appears to be a crucial event in late ischemic cell death, and Mg2⫹ enhances mitochondrial buffering of intracellular free Ca2⫹33 and is required for uptake of Ca2⫹ by endoplasmic reticulum (via membrane Mg2⫹-Ca2⫹ ATPase) in ischemic neurones.34 Recovery from white matter anoxic injury in an isolated rat optic nerve model is enhanced significantly by 10 mmol/L MgCl2 35 but not by other polyvalent cations or dihydropyridine calcium antagonists. Magnesium ions probably act by blocking voltage-sensitive Na⫹ channels and thereby preventing reversal of the Na⫹-Ca2⫹ exchanger. CNS Pharmacokinetics of Parenteral Magnesium Parenteral administration of magnesium raises serum, CSF, and brain [Mg2⫹] significantly, even in the presence of an intact blood-brain barrier; however, CSF [Mg2⫹] increases are less than those in serum. IV infusion of magnesium chloride (MgCl2) or MgSO4 to dogs increases plasma [Mg2⫹] to approximately 3 times baseline and increases CSF concentration by 21%,2 similar to increases in human subjects receiving Mg infusions. In animals,
259 2⫹
peripherally administered Mg appears within the CSF within minutes and peaks 2 to 3 hours after a bolus dose.2 Slower equilibration of CSF with brain Mg2⫹ suggests that further active transport processes govern distribution within the CNS.29 The half-life in CSF of intracisternally administered Mg2⫹ is around 70 minutes.2 Serum [Mg2⫹] correlates with CSF, cortical, and hippocampal [Mg2⫹]. Cardiovascular Effects of Magnesium Magnesium infusion causes vasodilatation while increasing cardiac output, with increased cardiac index despite transient lowering of blood pressure.36,37 Pressor effects of angiotensin II and norepinephrine are antagonized by MgSO4.38 Magnesium has direct effects on large, medium, and small vessels of the cerebral circulation. Carotid vasoconstriction caused by endothelin 1, neuropeptide Y, and angiotensin II in rats is reversed by IV MgSO4 or MgCl2 39,40 at plasma [Mg2⫹] of 2.4 to 2.7 mmol/L. Endothelin-1–mediated vasoconstriction is attenuated by Mg2⫹41 and is abolished at plasma concentrations of 3.8 to 4.6 mmol/L.39 Magnesium infusions cause cerebral arteriolar vasodilatation.42 Vasodilatation of mediumsized vessels may occur as a direct Ca2⫹ antagonist effect on vascular smooth muscle43 or may be mediated by prostacyclin.44 Physiological magnesium concentrations have significant anticonstrictor effects against prostaglandin F2␣ and serotonin in isolated postmortem human middle cerebral artery (MCA).45 Pial vessel vasoconstriction in response to excitatory amino acids is antagonized by IV or intra-arterial (IA) MgCl2 in rats.46 Basilar artery vasoconstriction was reversed by IV MgSO4 in a rat model of subarachnoid hemorrhage at mean plasma [Mg2⫹] of 4.32 mmol/L.47 Antiplatelet Activity Magnesium inhibits calcium-mediated ADP- and collagen-triggered platelet aggregation in vitro.48,49
Preclinical Studies in Cerebral Ischemia Focal Cerebral Ischemia Models Three full studies of magnesium in focal ischemia models have been published in full50-52 and one in abstract.53 These are summarized in Table 1. Magnesium was administered before onset of ischemia51,52 or immediately after MCA occlusion (MCAO),50,53 with repeat dosing at 1 hour50,53 or reperfusion.52 Additional unpublished studies by Wester and by Shuaib have found significant neuroprotection with systemically administered MgSO4 in mechanical MCAO and thromboembolic MCAO, respectively. Significant protection was seen
KEITH W. MUIR
MgCl2 Sprague-Dawley rats MCAO (90 min)-r Schmid-Elsaesser et al52
MgSO4 MgCl2 Sprague-Dawley rats Harland-Olac mice MCAO (90 or 120 min)-r pMCAO Marinov et al51 Roffe et al53
Significant infarct volume reductions are in bold. Abbreviations: MCAO, middle cerebral artery occlusion; pMCAO, permanent middle cerebral artery occlusion; MCAO-r, middle cerebral artery occlusion with reperfusion (occlusion time in parentheses); IP, intraperitoneal; IA, intra-arterial; IV, intravenous.
ⴚ37% Pre & 90 min post IV
ⴚ60% (90 mg/kg, 90 min ischemia) ⫹30% Pre Pre IA IP
Pre & 1 h post IP
1 mmol/kg (repeated) 30 or 90 mg/kg 1 mmol/kg (repeated) 1 mmol/kg (repeated) MgCl2 Fischer-344 rats pMCAO Izumi et al50
Route Dose Salt Species Model Study
Table 1. Magnesium in focal cerebral ischemia models
Timing
ⴚ26%
Infarct Volume
260
with administration delayed for 2 hours after onset of ischemia. Wester also reported significant elevation of CSF and brain [Mg2⫹], especially within the penumbra.54 Additive neuroprotective effects of MgCl2 and the freeradical scavenging aminosteroid compound tirilazad, with or without hypothermia, have been reported.52,55 Clear evidence of dose-dependent neuroprotection was found by Marinov et al.51 The overall treatment effect in this study represents a conservative estimate, since rigorous statistical analysis excluded animals without established cortical infarcts: 7 of 8 animals so excluded were in magnesium-treated groups. Different magnesium salts have had differing effects on glycemic control in reported studies. Magnesium chloride has been reported consistently to cause hyperglycemia in focal ischemia models50,52,53 and in other models of ischemia.56 No such trend has been found for MgSO4.51 Hyperglycemia is recognized as exacerbating ischemic neuronal damage, and adjunctive insulin administration to maintain normoglycemia has resulted in enhanced neuroprotection in some studies, eg, Izumi et al50 found 26% infarct volume reduction with MgCl2 and 44% reduction with MgCl2 plus insulin. After permanent MCAO in rats, Chi et al57 reported abolition of significant reduction in cerebral blood flow (CBF) by IV MgSO4 (total dose of 1.95 mmol/kg over 30 minutes, serum [Mg2⫹] increased from 0.86 to 3.21 mmol/L), measuring CBF using the [14C]-iodoantipyrine method. However, another group reported no difference in cortical CBF during ischemia by laser Doppler flowmetry between MgCl2-treated and control rats subjected to transient MCAO.52
Other In Vivo Cerebral Ischemia and Excitotoxicity Models Global Cerebral Ischemia Models In rats subjected to 30 minutes of forebrain ischemia, intrastriatal infusion of 2.5 mmol/L MgCl2 inhibited striatal glutamate and aspartate release, as did dizocilpine (but not the AMPA antagonist NBQX).58 In a rat global cerebral ischemia model (20 minutes of forebrain ischemia followed by reperfusion), direct injection of 50 mmol/L MgCl2 to the hippocampus reduced death of CA1 neurones when administered 24 hours after onset of ischemia.59 Functional neurological outcome at day 7 was also improved by 2 mmol/kg MgSO4 in dogs subjected to 18 minutes of global ischemia followed by reperfusion.60 However, in another study, pretreatment with 5 mmol/kg MgCl2 (peak plasma [Mg2⫹], 5.22 to 6.29 mmol/L) did not prevent loss of hippocampal CA1 neurones assessed 7 days after 10 minutes of forebrain ischemia in Sprague-Dawley rats, despite correcting significant hyperglycemia with insulin.56
MAGNESIUM FOR STROKE
Other Models In a pure excitotoxic brain injury model (stereotactic intrastriatal injection of NMDA to perinatal day 7 rats), there was dose-dependent reduction of infarcted brain tissue by MgSO4 up to 4 mmol/kg IP 15 minutes postNMDA injection.61 A cocktail of 0.3g/kg MgSO4, mannitol, and L-methionine IP after hypoxic and ischemic insults to perinatal day-8 rats reduced overall cerebral hemispheric damage.62 One hour postinjury, MgSO4 (600 mg/kg administered subcutaneously) improved evoked potentials and attenuated lipid peroxide products in rats.63 Intrathecal 1 mol/L MgSO4 improved functional recovery 1 week after spinal cord injury in rats without significant histological effects.64 Preischemic IV 5 mmol/kg MgCl2 increased the duration of recoverable ischemia in a rabbit spinal cord ischemia model but caused profound respiratory depression.65 In several models of traumatic brain injury, magnesium has been neuroprotective or facilitated recovery. McIntosh et al66 found improved neurological scores up to 4 weeks after focal fluid-percussion injury using 15minute duration IV infusions of 12.5 mol or 125 mol MgCl2 commenced 30 minutes after insult, improved performance of rats in memory testing paradigms with 125 mol MgCl2,67 and reduced cerebral edema after 15-minute postinjury MgCl2 at 300 mol/kg IV.68 Preinjury regimens of 1 mmol/kg MgCl2 IP daily enhanced neurological recovery, even when the final doses were administered 24 hours before electrolytic lesions to the sensorimotor cortex, and also reduced the volume of associated subcortical striatal degeneration 43 days postictus.69 One hour postinjury treatment with MgCl2 increased brain [Mg2⫹], reduced edema formation, and improved neurological recovery after focal contusional injury in rats.70 Enhanced motor recovery has been observed in a diffuse axonal injury model in rats receiving 100 mol/kg of either MgCl2 or MgSO4 30 minutes after injury.71 Infusion of IV MgCl2 at 4 mol/min (plasma [Mg2⫹] 3 times baseline) prevented intracerebral hemorrhage and death consequent to direct alcohol injection into rat brain.72 Several magnesium regimens prevent the vasospasm and microvascular hemorrhage induced by phencyclidine73 and cocaine74 in rats, and both IV and IA MgCl2 or Mg aspartate hydrochloride exhibit dose-dependent inhibition of phencyclidine-mediated pial arteriolar contraction at serum [Mg2⫹] between 0.99 and 2.72 mmol/L.73
Clinical Studies Clinical use of magnesium has been almost exclusively confined to the sulfate salt rather than the chloride. Mag-
261
nesium sulfate has been used widely in the treatment of eclamptic seizures, prevention of preterm labor, acute MI, and cardiac dysrhythmias. Extensive data from randomized, controlled trials are available for MI and preeclampsia/eclampsia. Several smaller studies in acute stroke have been reported, and a large multicenter trial in stroke is ongoing. Magnesium is excreted renally, and dose adjustment is required in renal impairment. Toxicity is manifest as depression of deep tendon reflexes at serum levels of over 5 mmol/L, with respiratory depression, EKG changes (T wave flattening, PR interval increase, and diminished R wave amplitude), and neuromuscular weakness at concentrations of approximately 7 mmol/L.75 Calcium gluconate can be administered IV in emergencies to counteract these effects, but rapid renal clearance ensures that symptoms of hypermagnesemia are transient. Muscular strength returns rapidly and is normal 6 hours after terminating IV infusion.75 Cases of iatrogenic overdose can usually be managed with supportive care alone. Volunteers In patients undergoing craniotomy, a 15-minute IV infusion of MgSO4 at 60 mg/kg (peak serum [Mg2⫹], 1.24 mmol/L) raised CSF [Mg2⫹] from 0.95 to 1.13 mmol/L. CSF concentration peaked at 4 hours.76 In patients with hypertension or stable ischemic heart disease, IV MgSO4 infusions that produce serum concentrations of 1.5 to 2.0 mmol/L increase cardiac index,77 reduce systemic vascular resistance,77 increase arterial blood flow,78 cause flushing,36 inhibit aldosterone release,79 and potentiate insulin-mediated oxidative glucose metabolism.80 Blood pressure and heart rate effects are variable. Bleeding time was prolonged postinfusion in eclamptic patients receiving MgSO481 and also in healthy volunteers infused with 8 mmol MgSO4 over 15 minutes followed by 3 mmol/h (achieving mean serum [Mg2⫹], 1.50 mmol/L).82 Antiaggregant effects were independent of aspirin use in vitro.48,49 Acute Myocardial Infarction An overview of seven small trials of MgSO4 in MI, involving 1,301 patients, reported reduced mortality and incidence of ventricular dysrhythmias.83 Two large trials followed: the Second Leicester Intervention in Myocardial Infarction Trial (LIMIT-2),84 which included 2,316 patients and suggested a myocardial cytoprotective effect, and subsequently the fourth International Study of Infarct Survival (ISIS-4),85 which randomized 58,050 patients but showed no effect of treatment on any outcome. Both LIMIT-2 and ISIS-4 used similar dose regimens: IV
KEITH W. MUIR
262
boluses of 8 mmol being followed by IV infusions of 65 or 72 mmol, respectively over 24 hours. Both trials confirmed a low incidence of magnesium-related adverse effects. A small excess of flushing was noted in both trials, affecting 0.3% of patients in ISIS-4. In ISIS-4, there were small but statistically significant increases in hypotension (1.1%), cardiac failure (1.2%), and cardiogenic shock (0.5%). However, these may have been subject to pharmacological confounding—patients received concomitant captopril, nitrates, or streptokinase— or reporting bias, since the trial used an open rather than blinded control group. The LIMIT-2 study, in contrast, reported a significant reduction in the incidence of cardiac failure (3.7% absolute risk reduction) and nonsignificant reduction of cardiogenic shock (1.3%). Although ISIS-4 was much larger than LIMIT-2, concerns regarding differences in administration of magnesium and concomitant treatments have led to persistent debate. Pre-Eclampsia/Eclampsia Several large randomized, controlled trials involving over 2,000 women and a meta-analysis86 confirm the superiority of MgSO4 over placebo and conventional anticonvulsants (diazepam and phenytoin) for prevention of seizures in pre-eclamptic87 and eclamptic88,89 patients. Typical dose regimens involved initial IM MgSO4 at 10 g (with or without 4 g IV) and maintenance IM doses (4 to 5 g every 4 hours) or IV infusion (1 g/h for 24 hours). For heptahydrated MgSO4, 1 g approximates 4 mmol. IM regimens produce a high initial serum peak,90 with subsequent stabilization of serum levels at around 2 to 3 times physiological. Pregnancy-related changes in volume of distribution and renal excretion are further affected in pre-eclampsia and eclampsia, where edema and oliguria are features. These factors caution against direct extrapolation of these regimens to nonpregnant patients. In pre-eclamptic women, IV MgSO4 increases CSF concentration significantly (for example, from 1.05 to 1.25 mmol/L).91 The mechanism of action is uncertain. Transcranial and color-flow Doppler studies show changes consistent with vasodilatation of the maternal cerebral circulation distal to the MCA92 in pregnancy-induced hypertension and pre-eclampsia. Magnesium also reduces angiotensin-converting enzyme concentrations,93 restores urinary cGMP to that of normotensive pregnant women,94 and reduces endothelin-1 expression in vascular endothelium.95 Magnesium’s efficacy in retarding preterm labor is debatable, with several trials finding negative results. Perinatal Hypoxic Injury Magnesium use in mothers treated for pre-eclampsia or preterm labor has been associated in several studies
with reduced incidence of cerebral palsy in very low birth-weight babies.96,97 Others have reported neutral effects98,99 and the absence of an effect on neonatal white matter lesions on transcranial ultrasound.100 Reduced perinatal mortality with MgSO4 in preterm labor has been reported in a case-control study,101 but a recent randomized, controlled study was terminated after 3 deaths in 46 pregnancies treated with MgSO4.102 Nonsignificant reductions in perinatal deaths compared with phenytoin have been reported on post hoc analyses of pre-eclampsia/eclampsia trials.87,88 A study of MgSO4 in perinatal hypoxia was stopped early due to an excess of poor outcomes in neonates receiving magnesium. The trial remains unpublished to date. Subsequent analysis of the study has disclosed misunderstandings of the magnesium dose due to transAtlantic differences in pharmacy calculation of MgSO4 dose based on weight (differences arising between anhydrous and heptahydrated MgSO4). A proportion of neonates with adverse outcomes appears to have received an excessive dose of magnesium. Stroke Five small trials of magnesium in stroke have been reported, with three reported as abstracts only. A large multicenter trial is ongoing (the Intravenous Magnesium Efficacy in Stroke [IMAGES] trial). Wester et al reported open-label use of IV infusions of MgSO4 in 12 patients without cardiovascular or other tolerability problems and went on to conduct a doubleblind trial in 26 stroke patients103 receiving 4 mmol MgSO4 as a bolus followed by a 5 mmol/h infusion over 8 hours and then oral magnesium supplements for 5 days. Full details of these studies remain unpublished. Muir and Lees have reported two small randomized, double-blind, placebo-controlled studies of MgSO4.104,105 The first trial used a loading infusion of 8 mmol, followed by 65 mmol over 24 hours in 60 patients within 12 hours of stroke. No significant treatment-related adverse effects were reported. A mean serum [Mg2⫹] of 1.42 mmol/L was achieved, but maximal concentration was only obtained at the end of the 24 hours of infusion. A subsequent dose-optimization study105 used higher loading infusions of 8, 12, and 16 mmol, each followed by 65 mmol over 24 hours in 25 patients within 24 hours of stroke. No adverse cardiovacular effects or notable tolerability problems were encountered. The highest loading dose group elevated serum [Mg2⫹] to 1.84 mmol/L. Results of the IMAGES pilot trial have been reported in abstract, including 51 patients treated within 12 hours of stroke. The trial intends to recruit 2,700 patients, one third within 6 hours of stroke, and uses the 16 mmol loading plus 65 mmol over 24 hours regimen. No tolerability issues have been reported to date.
MAGNESIUM FOR STROKE
A single-center Greek study has been reported in abstract in which 233 subjects received IV magnesium aspartate hydrochloride equivalent to 7.5 mmol of Mg2⫹ with improved outcome at days 30 and 90 compared with 277 controls. However, details of the study design and randomization procedures were not stated.106
Discussion There are multiple potentially beneficial pharmacological effects of supraphysiological magnesium concentrations in stroke. Peripherally administered magnesium crosses the intact blood-brain barrier and is therefore available in the acute phase of human stroke. Selective local increases in ischemic tissue are also reported. Systemically administered magnesium can block specific excitotoxic mechanisms of brain injury, and enhanced NMDA receptor block may be attainable with well-tolerated brain concentrations. Magnesium has also been found to reduce presynaptic glutamate release, block N-, P-, and L-type calcium channels; prevent intracellular free Ca2⫹ increase; enhance mitochondrial Ca2⫹ buffering; enhance recovery of neuronal ATP after ischemia; and antagonize the vasoconstrictor effects of several mediators, including endothelin-1. Magnesium reduces histological infarct volume in standard animal models of focal cerebral ischemia when administered before or up to 2 hours after onset of ischemia and also reduces neuronal loss in several other models of cerebral ischemia. Magnesium sulfate is neuroprotective in animal focal ischemia models and in several models of global ischemia and traumatic brain injury. Magnesium chloride is less consistently neuroprotective in animal models, probably by inducing hyperglycemia, but has been neuroprotective when normoglycemia is maintained. The mechanism of this apparently anion-specific effect is unknown, although physicochemical differences that may influence membrane transport of Mg2⫹ are recognized. Magnesium is excreted renally, and significant toxicity is unlikely in patients with normal renal function. The specific effects of hypermagnesemia are well-documented and can be reversed pharmacologically. Safety and tolerability of magnesium sulfate have been established in clinical trials in pre-eclampsia/eclampsia and acute MI. Trials in stroke have found no tolerability problems or evidence of potentially detrimental hypotension, bradycardia, or hyperglycemia. Preliminary data suggest the potential for improved clinical outcome in stroke patients, and the ongoing multicenter IMAGES trial is being conducted to establish efficacy. Several criticisms may be made of the existing clinical and preclinical data. Animal data for magnesium are less complete than is desirable, because, unlike pharmaceuti-
263
cal compounds, it has never been subject to a comprehensive development program. There is only one study showing significant neuroprotection with delayed administration, most treating before or concurrently with ischemia. However, the consistency of reports of neuroprotection across a broad range of models, laboratories, and investigators is a strength. The magnitude of infarct reduction and maintenance of efficacy with 2 hours of delay is comparable to most NMDA antagonist drugs in focal cerebral ischemia models, and magnesium sulfate shares neither the toxicity nor the thermoregulatory problems associated with this drug class. However, in the absence of proof of clinical utility, animal models of neuroprotection continue to have a credibility gap. Trials in MI failed to establish therapeutic benefit for this indication, and although debate continues regarding the possibility that the lack of effect of magnesium in ISIS-4 may be attributable to delayed administration after reperfusion, subgroup analysis suggests that significant benefit is unlikely even in patients treated before thrombolysis. The negative MI trial results should probably not detract greatly from the potential for stroke treatment, since there are multiple additional CNS targets for Mg2⫹, and the weight of preclinical evidence is considerably greater than was the case for MI. Nevertheless, they emphasize the need for caution in interpretation of small trials and the necessity of confirmation by larger pragmatic studies. The failures of several moderate to large neuroprotective trials raise concerns about the viability of both the neuroprotection hypothesis and of current trial designs. Neuroprotective agents have failed for a variety of reasons. Trials of glutamate antagonists have been terminated prematurely due to concerns about toxicity or where interim analysis has suggested low probability of demonstrating significant benefit. Most of these drugs have had significant dose-limiting toxic effects (CNS or cardiac) that have prevented administration of doses likely to produce serum levels in the neuroprotective range derived from animal data being attained in humans (aptiganel HCl, lubeluzole, selfotel, and eliprodil are all examples). Calcium antagonists have similarly been ineffective in trials, but most studies (mainly of nimodipine) have used delayed oral dosing. Significant hypotensive effects of IV nimodipine were dose-limiting and detrimental. Magnesium has multiple relevant modes of action besides calcium-channel and glutamate antagonism and, unlike most agents in clinical trials to date, may achieve these pharmacological goals without major side-effects. Experience in pre-eclampsia suggests that magnesium, unlike most neuroprotective agents in trials to date, may exert significant CNS effects at very well tolerated serum concentrations. Nevertheless, in the absence of evidence of the relevance of neuroprotective
264
mechanisms in human stroke, it remains possible that pharmacological targets simply cannot be extrapolated from animal models. Whether the neuroprotection hypothesis has yet been tested adequately is open to question.107 Clinical trial design has to date been predicated on the anticipation of large treatment effects, with most trials having been powered to detect effect sizes of approximately 10%. If full recruitment is achieved, IMAGES will have a sample size 50% greater than any individual trial to date and is powered to detect a 6% difference in outcome. However, it is compromised by the extended time window of 12 hours, which may be too long to expect benefit. As the time window lengthens, the proportion of patients with viable tissue decreases (possibly exponentially), and proven therapeutic benefit in stroke at present is limited to treatment within 3 hours. Nevertheless, the IMAGES sample size is sufficient to include in excess of 1,000 patients treated within 6 hours, which represents the standard window for neuroprotective trials. The IMAGES protocol is necessarily a compromise, since academic studies cannot be funded to a level that permits the intensive, short-window studies run by pharmaceutical industry sponsors: extension of the trial to centers not heavily involved in commercial trials has the advantage of improving the generalizability of results, but at the price of reduced data collection and an extended time window. Since even modest benefit in acute stroke would have a very substantial impact on stroke care due to the broad applicability of a safe, readily available, inexpensive, and familiar treatment, testing the efficacy of magnesium in a paradigm closely similar to that of existing neuroprotective trials is worthwhile.
References 1. Ebel H, Gunther T. Magnesium metabolism: A review. J Clin Chem Clin Biochem 1980;18:257-270. 2. Oppelt WW, MacIntyre I, Rall DP. Magnesium exchange between blood and cerebrospinal fluid. Am J Physiol 1963;205:959-962. 3. Vink R, McIntosh TK, Demediuk P, et al. Decrease in total and free magnesium concentration following traumatic brain injury in rats. Biochem Biophys Res Commun 1987;149:594-599. 4. Hossmann KA, Grosse Ophoff B, Csiba L, et al. Regional pH and electrolyte homeostasis of cat brain after prolonged ischemia. Neurochem Pathol 1988;9:127-137. 5. Helpern JA, Vande Linde AM, Welch KM, et al. Acute elevation and recovery of intracellular [Mg2⫹] following human focal cerebral ischemia. Neurology 1993;43: 1577-1581. 6. Rothman SM. Synaptic activity mediates death of hypoxic neurons. Science 1983;220:536-537. 7. Clark GD, Rothman SM. Blockade of excitatory amino acid receptors protects anoxic hippocampal slices. Neuroscience 1987;21:665-671.
KEITH W. MUIR 8. Kass IS, Cottrell JE, Chambers G. Magnesium and cobalt, not nimodipine, protect neurons against anoxic damage in the rat hippocampal slice. Anesthesiology 1988;69:710-715. 9. Smith DAS, Connick JH, Stone TW. Effect of changing extracellular levels of magnesium on spontaneous activity and glutamate release in the mouse neocortical slice. Br J Pharmacol 1989;97:475-482. 10. Bartrup JT, Stone TW. Presynaptic actions of adenosine are magnesium-dependent. Neuropharmacology 1988; 27:761-763. 11. Evans RH, Francis AA, Watkins JC. Selective antagonism by Mg2⫹ of amino acid-induced depolarization of spinal neurones. Experientia 1977;33:489-491. 12. Nowak L, Bregestovski P, Ascher P, et al. Magnesium gates glutamate-activated channels in mouse central neurones. Nature 1984;307:462-465. 13. Johnson JW, Ascher P. Voltage-dependent block by intracellular Mg2⫹ of N-methyl-D-aspartate-activated channels. Biophys J 1990;57:1085-1090. 14. Yamakura T, Mori H, Masaki H, et al. Different sensitivities of NMDA receptor channel subtypes to noncompetitive antagonists. NeuroReport 1993;4:687-690. 15. Burnashev N, Schoepfer R, Monyer H, et al. Control by asparagine residues of calcium permeability and magnesium blockade in the NMDA receptor. Science 1992; 257:1415-1419. 16. Mori H, Masaki H, Yamakura T, et al. Identification by mutagenesis of a Mg2⫹-block site of the NMDA receptor channel. Nature 1992;358:683-675. 17. Kawajiri S, Dingledine R. Multiple structural determinants of voltage-dependent magnesium block in recombinant NMDA receptors. Neuropharmacology 1993;32: 1203-1211. 18. Kuner T, Schoepfer R. Multiple structural elements determine subunit specificity of Mg2⫹ block in NMDA receptor channels. J Neurosci 1996;16:3549-3558. 19. Frandsen A, Schousboe A. Effect of magnesium on NMDA mediated toxicity and increases in [Ca2⫹]i and cGMP in cultured neocortical neurons: Evidence for distinct regulation of different responses. Neurochem Int 1994;25:301-308. 20. Harrison NL, Simmonds MA. Quantitative studies on some antagonists of N-methyl D-aspartate in slices of rat cerebral cortex. Br J Pharmacol 1985;84:381-391. 21. Ascher P, Nowak L. The role of divalent cations in the N-methyl-D-aspartate responses of mouse central neurones in culture. J Physiol (Lond) 1988;399:247-266. 22. Rose K, Christine CW, Choi DW. Magnesium removal induces paroxysmal neuronal firing and NMDA receptor-mediated neuronal degeneration in cortical cultures. Neurosci Lett 1990;115:313-317. 23. Hori N, Carpenter DO. Functional and morphological changes induced by transient in vivo ischemia. Exp Neurol 1994;129:279-289. 24. Li-Smerin Y, Johnson JW. Kinetics of the block by intracellular Mg2⫹ of the NMDA-activated channel in cultured rat neurons. J Physiol (Lond) 1996;491:121-135. 25. Brocard JB, Rajdev S, Reynolds IJ. Glutamate-induced increases in intracellular free Mg2⫹ in cultured cortical neurons. Neuron 1993;11:751-757. 26. Hallak M, Berman RF, Irtenkauf SM, et al. Magnesium sulfate treatment decreases N-methyl-D-aspartate receptor binding in the rat brain: An autoradiographic study. J Soc Gynecol Invest 1994;1:25-30.
MAGNESIUM FOR STROKE 27. Wang LY, MacDonald JF. Modulation by magnesium of the affinity of NMDA receptors for glycine in murine hippocampal neurones. J Physiol (Lond) 1995;486:83-95. 28. Borges LF, Gucer G. Effect of magnesium on epileptic foci. Epilepsia 1978;19:81-91. 29. Hallak M, Berman RF, Irtenkauf SM, et al. Peripheral magnesium sulfate enters the brain and increases the threshold for hippocampal seizures in rats. Am J Obstet Gynecol 1992;167:1605-1610. 30. Mason BA, Standley CA, Irtenkauf SM, et al. Magnesium is more efficacious than phenytoin in reducing N-methyl-D-aspartate seizures in rats. Am J Obstet Gynecol 1994;171:999-1002. 31. Decollogne S, Tomas A, Lecerf C, et al. NMDA receptor complex blockade by oral administration of magnesium: Comparison with MK-801. Pharmacol Biochem Behav 1997;58:261-268. 32. Iseri LT, French JH. Magnesium: Nature’s physiologic calcium blocker. Am Heart J 1984;108:188-193. 33. Favaron M, Bernardi P. Tissue-specific modulation of the mitochondrial calcium uniporter by magnesium ions. FEBS Lett 1985;183:260-264. 34. Parsons JT, Churn SB, DeLorenzo RJ. Ischemia-induced inhibition of calcium uptake into rat brain microsomes mediated by Mg2⫹/Ca2⫹ ATPase. J Neurochem 1997; 68:1124-1134. 35. Stys PK, Ransom BR, Waxman SG. Effects of polyvalent cations and dihydropyridine calcium channel blockers on recovery of CNS white matter from anoxia. Neurosci Lett 1990;115:293-299. 36. Mroczek WJ, Lee WR, Davidov ME. Effect of magnesium sulfate on cardiovascular hemodynamics. Angiology 1977;28:720-724. 37. Reinhart RA. Clinical correlates of the molecular and cellular actions of magnesium on the cardiovascular system. Am Heart J 1991;121:1513-1521. 38. Lee MI, Todd HM, Bowe A. The effects of magnesium sulfate infusion on blood pressure and vascular responsiveness during pregnancy. Am J Obstet Gynecol 1984; 149:705-708. 39. Kemp PA, Gardiner SM, Bennett T, et al. Magnesium sulphate reverses the carotid vasoconstriction caused by endothelin-I, angiotensin II and neuropeptide-Y, but not that caused by N(G)-nitro-L-arginine methyl ester, in conscious rats. Clin Sci 1993;85:175-181. 40. Kemp PA, Gardiner SM, March JE, et al: Assessment of the effects of endothelin-1 and magnesium sulphate on regional blood flows in conscious rats, by the coloured microsphere reference technique. Br J Pharmacol 1999; 126:621-626. 41. Torregrosa G, Perales AJ, Salom JB, et al. Different effects of Mg2⫹ on endothelin-1- and 5-hydroxytryptamine-elicited responses in goat cerebrovascular bed. J Cardiovasc Pharmacol 1994;23:1004-1010. 42. Altura BT, Altura BM. Interactions of Mg and K on cerebral vessels-aspects in view of stroke. Review of present status and new findings. Magnesium 1984;3: 195-211. 43. Altura BT, Altura BM. The role of magnesium in etiology of strokes and cerebrovasospasm. Magnesium 1982;1:277-291. 44. Nadler JL, Goodson S, Rude RK. Evidence that prostacyclin mediates the vascular action of magnesium in humans. Hypertension 1987;9:379-383.
265 45. Alborch E, Salom JB, Perales AJ, et al. Comparison of the anticonstrictor action of dihydropyridines (nimodipine and nicardipine) and Mg2⫹ in isolated human cerebral arteries. Eur J Pharmacol 1992;229:83-89. 46. Huang QF, Gebrewold A, Zhang A, et al. Role of excitatory amino acids in regulation of rat pial microvasculature. Am J Physiol 1994;266:158-163. 47. Ram Z, Sadeh M, Shacked I, et al. Magnesium sulfate reverses experimental delayed cerebral vasospasm after subarachnoid hemorrhage in rats. Stroke 1991;22:922927. 48. Ravn HB, Vissinger H, Kristensen SD, et al. Magnesium inhibits platelet activity-an in vitro study. Thromb Haemost 1996;76:88-93. 49. Ravn HB, Kristensen SD, Vissinger H, et al. Magnesium inhibits human platelets. Blood Coagul Fibrinol 1996;7: 241-244. 50. Izumi Y, Roussel S, Pinard E, et al. Reduction of infarct volume by magnesium after middle cerebral artery occlusion in rats. J Cereb Blood Flow Metab 1991;11: 1025-1030. 51. Marinov MB, Harbaugh KS, Hoopes PJ, et al. Neuroprotective effects of preischemia intraarterial magnesium sulfate in reversible focal cerebral ischemia. J Neurosurg 1996;85:117-124. 52. Schmid-Elsaesser R, Zausinger S, Hungerhuber E, et al. Neuroprotective effects of combination therapy with tirilazad and magnesium in rats subjected to reversible focal cerebral ischemia. Neurosurgery 1999;44:163-171. 53. Roffe C, Thomas L, Fotheringham A, et al. The effect of magnesium on infarct size and oedema after middle cerebral artery occlusion. Cerebrovasc Dis 1996;6:42 (abstr, suppl 2). 54. Sjostrom LG, Wester P. Systemically induced hypermagnesemia increases the bioavailability of magnesium in rat brain. Soc Neurosci 1994;20(Part 1):181 (abstr). 55. Schmid-Elsaesser R, Hungerhuber E, Zausinger S, et al. Combination drug therapy and mild hypothermia: A promising treatment strategy for reversible, focal cerebral ischemia. Stroke 1999;30:1891-1899. 56. Blair JL, Warner DS, Todd MM. Effects of elevated plasma magnesium versus calcium on cerebral ischemic injury in rats. Stroke 1989;20:507-512. 57. Chi OZ, Pollak P, Weiss HR. Effects of magnesium sulfate and nifedipine on regional cerebral blood flow during middle cerebral artery ligation in the rat. Arch Int Pharmacodyn Ther 1990;304:196-205. 58. Ghribi O, Callebert J, Verrecchia C, et al. Blockers of NMDA-operated channels decrease glutamate and aspartate extracellular accumulation in striatum during forebrain ischaemia in rats. Fund Clin Pharmacol 1995; 9:141-146. 59. Tsuda T, Kogure K, Nishioka K, et al. Mg2⫹ administered up to twenty-four hours following reperfusion prevents ischemic damage of the CA1 neurons in the rat hippocampus. Neuroscience 1991;44:335-341. 60. Okawa M. [Effects of magnesium sulfate on brain damage by complete global brain ischemia]. Masui—Jpn J Anesthesiol 1992;41:341-355. 61. McDonald JW, Silverstein FS, Johnston MV. Magnesium reduces N-methyl-D-aspartate (NMDA)-mediated brain injury in perinatal rats. Neurosci Lett 1990;109: 234-238. 62. Thordstein M, Bagenholm R, Thiringer K, et al. Scavengers of free oxygen radicals in combination with mag-
KEITH W. MUIR
266
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
nesium ameliorate perinatal hypoxic-ischemic brain damage in the rat. Pediatr Res 1993;34:23-26. Suzer T, Coskun E, Islekel H, et al. Neuroprotective effect of magnesium on lipid peroxidation and axonal function after experimental spinal cord injury. Spinal Cord 1999;37:480-484. Follis F, Miller K, Scremin OU, et al. NMDA receptor blockade and spinal cord ischemia due to aortic crossclamping in the rat model. Can J Neurol Sci 1994;21: 227-232. Vacanti FX, Ames A III. Mild hypothermia and Mg⫹⫹ protect against irreversible damage during CNS ischemia. Stroke 1984;15:695-698. McIntosh TK, Vink R, Yamakami I, et al. Magnesium protects against neurological deficit after brain injury. Brain Res 1989;482:252-260. Smith DH, Okiyama K, Gennarelli TA, et al. Magnesium and ketamine attenuate cognitive dysfunction following experimental brain injury. Neurosci Lett 1993; 157:211-214. Okiyama K, Smith DH, Gennarelli TA, et al. The sodium channel blocker and glutamate release inhibitor BW1003C87 and magnesium attenuate regional cerebral edema following experimental brain injury in the rat. J Neurochem 1995;64:802-809. Hoane MR, Irish SL, Marks BB, et al. Preoperative regimens of magnesium facilitate recovery of function and prevent subcortical atrophy following lesions of the rat sensorimotor cortex. Brain Res Bull 1998;45:45-51. Feldman Z, Gurevitch B, Artru AA, et al. Effect of magnesium given 1 hour after head trauma on brain edema and neurological outcome. J Neurosurg 1996;85: 131-137. Heath DL, Vink R. Neuroprotective effects of MgSO4 and MgCl2 in closed head injury: A comparative phosphorus NMR study. J Neurotrauma 1998;15:183-189. Altura BM, Gebrewold A, Altura BT, et al. Role of brain [Mg2⫹]i in alcohol-induced hemorrhagic stroke in a rat model: A 31P-NMR in vivo study. Alcohol 1995;12:131136. Huang QF, Gebrewold A, Altura BT, et al. Magnesium ions prevent phencyclidine-induced cerebrovasospasms and rupture of cerebral microvessels: Direct in-vivo microcirculatory studies on the rat brain. Neurosci Lett 1990;113:115-119. Altura BM, Zhang A, Cheng TP, et al. Cocaine induces rapid loss of intracellular free Mg2⫹ in cerebral vascular smooth muscle cells. Eur J Pharmacol 1993;246:299-301. Somjen G, Hilmy M, Stephen CR. Failure to anesthetize human subjects by intravenous administration of magnesium sulfate. J Pharmacol Exp Ther 1966;154:652-659. Fuchs-Buder T, Tramer MR, Tassonyi E. Cerebrospinal fluid passage of intravenous magnesium sulfate in neurosurgical patients. J Neurosurg Anesthesiol 1997;9:324328. Rasmussen HS, Larsen OG, Meier K, et al. Hemodynamic effects of intravenously administered magnesium on patients with ischemic heart disease. Clin Cardiol 1988;11:824-828. Fujita T, Ito Y, Ando K, et al. Attenuated vasodilator responses to Mg2⫹ in young patients with borderline hypertension. Circulation 1990;82:384-393. Ichihara A, Suzuki H, Saruta T. Effects of magnesium on the renin-angiotensin-aldosterone system in human subjects. J Lab Clin Med 1993;122:432-440.
80. Paolisso G, Gambardella A, Balbi V, et al. Effects of magnesium and nifedipine infusions on insulin action, substrate oxidation, and blood pressure in aged hypertensive patients. Am J Hypertens 1993;6:920-926. 81. Kynczl-Leisure M, Cibils LA. Increased bleeding time after magnesium sulfate infusion. Am J Obstet Gynecol 1996;175:1293-1294. 82. Ravn HB, Vissinger H, Kristensen SD, et al. Magnesium inhibits platelet activity—An infusion study in healthy volunteers. Thromb Haemost 1996;75:939-944. 83. Teo KK, Yusuf S, Collins R, et al. Effects of intravenous magnesium in suspected acute myocardial infarction: An overview of the randomised trials. BMJ 1991;303: 1499-1503. 84. 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; 343:816-819. 85. ISIS-4 Collaborative Group. ISIS-4: A randomised factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet 1995;345:669-685. 86. Chien PF, Khan KS, Arnott N. Magnesium sulphate in the treatment of eclampsia and pre-eclampsia: An overview of the evidence from randomised trials. Br J Obstet Gynaecol 1996;103:1085-1091. 87. Lucas MJ, Leveno KJ, Cunningham FG. A comparison of magnesium sulfate with phenytoin for the prevention of eclampsia. N Engl J Med 1995;333:201-205. 88. The Eclampsia Trial Collaborative Group. Which anticonvulsant for women with eclampsia? Evidence from the Collaborative Eclampsia Trial. Lancet 1995;345: 1455-1463. 89. Coetzee EJ, Dommisse J, Anthony J. A randomised controlled trial of intravenous magnesium sulphate versus placebo in the management of women with severe pre-eclampsia. Br J Obstet Gynaecol 1998;105: 300-303. 90. Sibai BM, Graham JM, McCubbin JH. A comparison of intravenous and intramuscular magnesium sulfate regimens in preeclampsia. Am J Obstet Gynecol 1984;150: 728-733. 91. Thurnau GR, Kemp DB, Jarvis A. Cerebrospinal fluid levels of magnesium in patients with preeclampsia after treatment with intravenous magnesium sulfate: A preliminary report. Am J Obstet Gynecol 1987;157:14351438. 92. Belfort MA, Moise KJ Jr. Effect of magnesium sulfate on maternal brain blood flow in preeclampsia: A randomized, placebo-controlled study. Am J Obstet Gynecol 1992;167:661-666. 93. Fuentes A, Goldkrand JW. Angiotensin-converting enzyme activity in hypertensive subjects after magnesium sulfate therapy. Am J Obstet Gynecol 1987;156:13751379. 94. Barton JR, Sibai BM, Ahokas RA, et al. Magnesium sulfate therapy in preeclampsia is associated with increased urinary cyclic guanosine monophosphate excretion. Am J Obstet Gynecol 1992;167:931-934. 95. Mastrogiannis DS, Kalter CS, O’Brien WF, et al. Effect of magnesium sulfate on plasma endothelin-1 levels in normal and preeclamptic pregnancies. Am J Obstet Gynecol 1992;167:1554-1559.
MAGNESIUM FOR STROKE 96. Nelson KB, Grether JK. Can magnesium sulfate reduce the risk of cerebral palsy in very low birthweight infants? Pediatrics 1995;95:263-269. 97. Schendel DE, Berg CJ, Yeargin-Allsopp M, et al. Prenatal magnesium sulfate exposure and the risk for cerebral palsy or mental retardation among very low-birthweight children aged 3 to 5 years. JAMA 1996;276:18051810. 98. Cox SM, Sherman ML, Leveno KJ. Randomized investigation of magnesium sulfate for prevention of preterm birth. Am J Obstet Gynecol 1990;163:767-772. 99. Paneth N, Jetton J, Pinto-Martin J, et al. Magnesium sulfate in labor and risk of neonatal brain lesions and cerebral palsy in low birth weight infants. Pediatrics 1997;99:e1. 100. Leviton A, Paneth N, Susser M, et al. Maternal receipt of magnesium sulfate does not seem to reduce the risk of neonatal white matter damage. Pediatrics 1997;99:e2. 101. Grether JK, Hoogstrate J, Selvin S, et al. Magnesium sulfate tocolysis and risk of neonatal death. Am J Obstet Gynecol 1998;178:1-6.
267 102. Mittendorf R, Pryde P, Khoshnood B, et al. If tocolytic magnesium sulfate is associated with excess total pediatric mortality, what is its impact? Obstet Gynecol 1998; 92:308-311. 103. Wester PO, Asplund K, Eriksson S, et al. Infusion of magnesium in patients with acute brain infarction. Acta Neurol Scand 1984;70:143. 104. Muir KW, Lees KR. A randomised, double-blind, placebo-controlled pilot trial of intravenous magnesium sulfate in acute stroke. Stroke 1995;26:183-188. 105. Muir KW, Lees KR. Dose optimization of intravenous magnesium sulfate after acute stroke. Stroke 1998;29: 918-923. 106. Galeas T, Ziogas G, Valotasiou B, et al. The role of magnesium (Mg)—a natural calcium (Ca) antagonist in the Ca channels of the platelets—in the treatment of acute ischaemic stroke. Consensus Conference on Stroke 1998;15. Edinburgh, UK, Royal College of Physicians of Edinburgh. 107. Gorelick PB. Neuroprotection in acute ischaemic stroke: A tale of for whom the bell tolls? Lancet 2000;355:1925-1926.