CLINICAL STUDIES

INTRAVENOUS MAGNESIUM VERSUS NIMODIPINE IN THE

TREATMENT OF PATIENTS WITH ANEURYSMAL
SUBARACHNOID HEMORRHAGE: A RANDOMIZED STUDY
Robert Schmid-Elsaesser, M.D.
Department of Neurosurgery,
Ludwig-Maximilians-Universität,
Munich, Germany
Matthias Kunz, M.D.
Department of Neurosurgery,
Ludwig-Maximilians-Universität,
Munich, Germany
Stefan Zausinger, M.D.
Department of Neurosurgery,
Ludwig-Maximilians-Universität,
Munich, Germany
Stefan Prueckner, M.D.
Institute for Anesthesiology,
Ludwig-Maximilians-Universität,
Munich, Germany
Josef Briegel, M.D.
Institute for Anesthesiology,
Ludwig-Maximilians-Universität,
Munich, Germany
Hans-Jakob Steiger, M.D.
Department of Neurosurgery,
Heinrich-Heine-Universität,
Düsseldorf, Germany
Reprint requests:
Robert Schmid-Elsaesser, M.D.,
Department of Neurosurgery,
Ludwig-Maximilians-Universität,
Klinikum Grosshadern,
Marchioninistraße 15,
81377 Munich, Germany.
Email: Schmid-Elsaesser@med.uni-
muenchen.de
Received, July 5, 2005.
Accepted, February 9, 2006.
OBJECTIVE: The prophylactic use of nimodipine in patients with aneurysmal sub-
arachnoid hemorrhage reduces the risk of ischemic brain damage. However, its
efficacy seems to be rather moderate. The question arises whether other types of
calcium antagonists offer better protection. Magnesium, nature’s physiological cal-
cium antagonist, is neuroprotective in animal models, promotes dilatation of cerebral
arteries, and has an established safety profile. The aim of the current pilot study is to
evaluate the efficacy of magnesium versus nimodipine to prevent delayed ischemic
deficits after aneurysmal subarachnoid hemorrhage.
METHODS: One hundred and thirteen patients with aneurysmal subarachnoid hem-
orrhage were enrolled in the study and were randomized to receive either magnesium
sulfate (loading 10 mg/kg followed by 30 mg/kg daily) or nimodipine (48 mg/d)
intravenously until at least postoperative Day 7. Primary outcome parameters were
incidence of clinical vasospasm and infarction. Secondary outcome measures were the
incidence of transcranial Doppler/angiographic vasospasm, the neuronal markers
(neuron-specific enolase, S-100), and the patients’ Glasgow Outcome Scale scores at
discharge and after 1 year.
RESULTS: One hundred and four patients met the study requirements. In the magne-
sium group (n ⫽ 53), eight patients (15%) experienced clinical vasospasm and 20
(38%) experienced transcranial Doppler/angiographic vasospasm compared with 14
(27%) and 17 (33%) patients in the nimodipine group (n ⫽ 51). If clinical vasospasm
occurred, 75% of the magnesium-treated versus 50% of the nimodipine-treated pa-
tients experienced cerebral infarction resulting in fatal outcome in 37 and 14%,
respectively. Overall, the rate of infarction attributable to vasospasm was virtually the
same (19 versus 22%). There was no difference in outcome between groups.
CONCLUSION: The efficacy of magnesium in preventing delayed ischemic neurolog-
ical deficits in patients with aneurysmal subarachnoid hemorrhage seems to be com-
parable with that of nimodipine. The difference in their pharmacological properties
makes studies on the combined administration of magnesium and nimodipine seem
promising.
KEY WORDS: Intracranial aneurysm, Magnesium sulfate, Nimodipine, Subarachnoid hemorrhage,
Vasospasm
Neurosurgery 58:1054-1065, 2006 DOI: 10.1227/01.NEU.0000215868.40441.D9 www.neurosurgery-online.com

D
espite recent advances in the manage-
ment of aneurysmal subarachnoid
hemorrhage (SAH), delayed ischemic
neurological deficits (DIND) remain a major
cause of morbidity and mortality. The patho-
genesis of DIND has not yet been elucidated,
but it is at least partially related to cerebral
vasospasm. Treatment of cerebral vasospasm
aims to prevent or reverse arterial narrowing
and to improve cerebral perfusion, and, ulti-
mately, to prevent ischemic neurological def-
icits (86).
No treatment for vasospasm is entirely safe
or without side effects. Currently, the most
common therapeutic approaches to vaso-
spasm are the use of calcium antagonists and

1054 | VOLUME 58 | NUMBER 6 | JUNE 2006 www.neurosurgery-online.com

 MAGNESIUM VERSUS NIMODIPINE IN SUBARACHNOID HEMORRHAGE

variations of normovolemia to hypervolemia, hypertension, Munich, Germany, with the approval of the local ethics commit-
and hemodilution (triple-H therapy). Balloon or chemical an- tee. Patients were enrolled from September 2000 through Sep-
gioplasty is usually reserved for selected refractory patients tember 2002. Informed consent was obtained from either the
(21). Early surgery for aneurysm clipping or endovascular patient or a relative in all cases.
treatment facilitates the management of vasospasm using
triple-H therapy. The efficacy of triple-H therapy, however, in Study Protocol
preventing the onset of cerebral vasospasm and improving
Inclusion in the study required the patient to have an an-
ischemic deficits or survival remains uncertain, and conflict-
eurysmal SAH. The diagnosis of SAH was made by the pres-
ing results have been reported (85). Prophylactic use of cal-
ence of blood in the basal cisterns on computed tomographic
cium antagonists, especially nimodipine, in patients with rup-
(CT) scans or by lumbar puncture, and the aneurysm was
tured intracranial aneurysms may reduce the risk of ischemic
confirmed by conventional or CT angiography. Patients were
damage (10). A recent meta-analysis showed that, overall,
entered in the study within 96 hours of ictus. In patients with
calcium antagonists provide an absolute risk reduction for
presumed sentinel hemorrhages, the time point of the ictus
poor outcome of 5%. The corresponding number of patients
leading to hospitalization was taken. Exclusion criteria were
needed to treat to prevent a single poor outcome event is 20.
pregnancy, renal failure (serum creatinine ⬎150 ␮mol/L),
Despite statistical significance, the efficacy of nimodipine in
congestive heart failure (New York Heart Association Class 3
improving outcome after SAH seems rather moderate, and
or 4), known neuromuscular disease such as myasthenia gra-
improved treatment strategies are needed (24, 76).
vis or other known adverse clinical effects of hypermag-
Among pharmacological agents, the role of magnesium is
nesemia, pre-existing medication with magnesium or nimo-
still under experimental (46, 70, 79, 80, 91, 94, 95) and clinical
dipine, or imminent death.
(9, 14, 16, 18, 57, 89, 92) investigation. Magnesium is a natu-
On admission, the clinical condition was assessed by the
rally occurring calcium antagonist and potent vasodilator (37).
World Federation of Neurological Surgeons (WFNS) scale (22)
It exhibits a range of vascular and neuronal actions that may
and the amount of blood on CT scans was documented ac-
ameliorate ischemic central nervous system insults (53, 54).
cording to the grading system of Fisher (26). Eligible patients
Experimentally, magnesium can reverse delayed cerebral va-
were randomly assigned to one of two treatment arms with
sospasm and reduces the extent of acute ischemic cerebral
the use of sealed opaque envelops to receive either magne-
lesions after experimental SAH in rats (73, 90). Furthermore,
sium sulfate or nimodipine. Magnesium sulfate was adminis-
magnesium ions (Mg2⫹) participate in voltage-sensitive block-
tered as a bolus infusion of 10 mg/kg (0.4 mmol/kg) MgSO4
ade of ion channels, resulting in noncompetitive antagonism
during a period of 30 minutes under blood pressure monitor-
of N-methyl-D-aspartate receptors (37, 59), compete with ex-
ing followed by a continuous infusion of 30 mg/kg daily (1.2
tracellular calcium ions to reduce calcium entry into cells (5),
mmol/kg daily). This body weight-adjusted dosage regimen
and inhibit the release of intracellular calcium ions (36) and
was adopted from the Intravenous Magnesium Efficacy in
excitatory amino acids (28, 78). Clinically, magnesium has a
Stroke trial for acute stroke (55, 56). Nimodipine was admin-
wide therapeutic index and is inexpensive, and its established
istered as a continuous infusion of 1.0 mg/h for 6 hours under
usefulness in obstetrical and cardiovascular practice confirms
blood pressure monitoring, then the dosage was increased to
its safety and tolerability (47). Its use in neurological patients
2.0 mg/h, provided that there was no inadvertent systemic
is also known to be safe (14, 55–57, 92).
hypotension (2, 60). Magnesium or nimodipine was adminis-
These properties make magnesium an interesting candidate
tered intravenously until postoperative Day 7. When there
to evaluate its influence on cerebral vasospasm and to assess
were no clinical symptoms and transcranial Doppler (TCD)
its effectiveness in preventing DIND and improving outcome
revealed no evidence of vasospasm, intravenous treatment
in patients with aneurysmal SAH. To our knowledge, magne-
was overlapped and followed by oral administration of nimo-
sium has not been evaluated as a single agent in this setting
dipine (60 mg every 4 h) or magnesium (12 mmol every 12 h),
yet. A comparison with a placebo would be considered un-
which then was phased out within 1 week. When TCD re-
ethical in light of the proven efficacy of nimodipine. We were
vealed evidence of vasospasm, intravenous treatment was
aware that any single center trial would probably be under-
continued until flow velocities dropped to less than 120 cm/s
powered to detect a significant difference, and we decided to
or Day 21 after SAH.
perform a pilot trial to identify possible trends. We present the
results of a randomized patient-blinded study of magnesium
sulfate (MgSO4) versus nimodipine therapy in patients with Monitoring Parameters
aneurysmal SAH who were admitted to our hospital during a On admission, serum baseline levels of magnesium, nerve
2-year period. tissue protein S-100, and neuron-specific enolase (NSE) were
obtained and then checked daily as long as the patient had a
PATIENTS AND METHODS central venous catheter, then with every blood sample until
discharge. If an intraventricular catheter or lumbar drainage
This study was performed in the Department of Neurosur- had to be inserted for other reasons, these parameters also
gery, Klinikum Grosshadern, Ludwig-Maximilians-University, were determined in the cerebrospinal fluid (CSF). Patients

NEUROSURGERY VOLUME 58 | NUMBER 6 | JUNE 2006 | 1055

SCHMID-ELSAESSER ET AL.
received daily monitoring for vasospasm with the aid of TCD
ultrasonography recordings and hourly neurological exami-
nations with specific checks for development of DIND. Mean
flow velocities of the intracranial arteries were measured
through the temporal bone window. The Lindegaard Index
was calculated by dividing the blood velocity in the middle
cerebral artery (MCA) by the blood velocity in the ipsilateral
distal extracranial internal carotid artery (42, 43).
Study Independent Management Protocol
All patients were treated according to a standardized man-
agement protocol for aneurysmal SAH, including early oblit-
eration of the aneurysm by surgical clipping or endovascular
techniques within 24 hours after admission. The patients were
monitored in a neurosurgical intensive care setting. Routine
monitoring included compilation of records on hemodynamic
parameters, 24-hour input/output measurements, intracranial
pressure monitoring (Glasgow Coma Scale score ⬍ 9 and/or
hydrocephalus), and serum electrolytes. Vasospasm was ag-
gressively treated with hemodynamic optimization, and bal-
loon or chemical angioplasty, or both, was performed in re-
fractory cases. An angiography and balloon or chemical
angioplasty, or both, was performed in patients with DIND
not reversible within 1 hour under hemodynamic optimiza-
tion. In patients who could not be assessed neurologically,
angiography was performed when TCD indicated vasospasm
or a control CT scan showed signs of cerebral ischemia. In
these patients, angioplasty was performed when perfusion CT
scan or perfusion-weighted magnetic resonance imaging scan
showed a mismatch with decreased regional cerebral blood
flow by 30% or more. Findings from cerebral angiograms, CT
scans, or other imaging studies were documented. At least one
postoperative control CT or magnetic resonance imaging scan
was performed before discharge in every patient.
Study End Points
To analyze the results of our trial, we focused on the inci-
dence of clinical vasospasm and infarction attributable to va-
sospasm as primary outcome measures.
Clinical vasospasm was defined as a DIND that could not be
accounted for by other causes such as hydrocephalus, hema-
toma, seizure, or electrolyte abnormality combined with ultra-
sonographic or angiographic evidence of vasospasm. A mean
velocity of more than 140 cm/s or Lindegaard Index of more
than 4 or an increase in the flow velocity by more than 40 cm/s
within 24 hours was regarded as pathological. Angiographic
evidence of vasospasm was defined as any moderate-to-severe
(i.e., ⬎30%) narrowing of the diameter of the arterial vessel
lumen on high-resolution images using a digital subtraction
method. The diagnosis of clinical vasospasm was made only
in patients who could be assessed neurologically, because in
comatose patients with or without analgosedation, the criteria
of secondary deterioration or new focal neurological deficit
cannot be met.
1056 | VOLUME 58 | NUMBER 6 | JUNE 2006
Infarction attributable to vasospasm was defined as the
occurrence of a new spontaneous hypodense lesion as re-
vealed by a CT scan (or infarction according to diffusion- or
perfusion-weighted magnetic resonance imaging scan) that
was not associated with the initial hemorrhage or was caused
by operation or endovascular procedure.
Secondary outcome parameters were the incidence of TCD/
angiographic vasospasm according to the criteria mentioned
above, the neuronal markers (NSE, S-100), and the patients’
Glasgow Outcome Scale (GOS) scores as a measure of the
effectiveness for prevention of long-term sequelae of vaso-
spasm from aneurysmal SAH. The GOS score (5, good recov-
ery; 4, moderate disability but independent; 3, severe disabil-
ity; 2, persistent vegetative state; or 1, death) was recorded at
discharge and at the 1-year follow-up examination.
Statistical Analysis
Statistical analyses were performed with the use of SigmaS-
tat statistical software (Version 2.0; SPSS Science, Inc., Chi-
cago, IL). Parametric data were analyzed using the t test, and
nonparametric data (e.g., Fisher, WFNS, GOS scores) were
analyzed using the Mann-Whitney rank-sum test. Incidence
rates (such as vasospasm, infarction, mortality) were analyzed
using the ␹2 test. Differences were considered significant at the
P ⬍ 0.05 level. Correlation between neuronal markers and
GOS score was calculated using the Spearman rank-order
correlation.
RESULTS
One hundred and sixty-one patients with aneurysmal SAH
were admitted to our hospital between September 2000 and
September 2002. Of these, 113 patients were eligible and were
enrolled. Five patients in the magnesium group and four
patients in the nimodipine group were withdrawn from the
study because study requirements were not met. In contrast to
the primarily available information, it was discovered that
nimodipine had already been administered by the transferring
hospital in four patients. In three patients, it was discovered
that the ictus was outside the 96-hour enrollment window.
Another patient who had been randomized to the nimodipine
group received a bolus infusion of MgS04 because of cardiac
arrhythmia. Finally, one patient was removed from the study
at his own request.
The clinical data of the remaining 104 patients (51 patients
in the nimodipine group and 53 patients in the magnesium
group) are summarized in Table 1. Mean values for groups are
given as mean ⫾ standard deviation or medians with inter-
quartile range (difference between the 25th and 75th percen-
tiles), respectively.
There were no significant differences between the two
groups regarding age, sex, Fisher grade, WFNS grade, or
treatment method. In 78 patients, the aneurysm was obliter-
ated with neurosurgical clipping, and in 22 patients, the an-
eurysm was obliterated with endovascular coiling. Four pa-
www.neurosurgery-online.com
 MAGNESIUM VERSUS NIMODIPINE IN SUBARACHNOID HEMORRHAGE

tients with persisting WFNS Grade 5 were managed rebled before definite treatment (one scheduled for surgery,
conservatively. the other for endovascular coiling) and six patients rebled
As expected, magnesium serum levels were comparable at during an endovascular procedure. Rebleeding occurred dur-
baseline and were significantly higher in the magnesium ing diagnostic angiography in one patient and during endo-
group during treatment. There was no difference regarding vascular coiling in five patients.
magnesium CSF levels (Table 2). Other than flushing sensation, The mean maximum levels of the neuronal markers (S-100
no side effects of magnesium infusion, such as bradycardia, and NSE) in serum and CSF were not significantly different
hypotension, or bradypnea, were observed. Six patients (12%) between the two groups. There was a trend toward higher
treated with nimodipine experienced inadvertent systemic hy- levels in WFNS Grades 4 and 5 compared with WFNS Grades
potension, and the dosage had to be reduced. No patients in 1 and 3 that reached statistical significance for S-100 (serum) in
the nimodipine group and eight patients in the magnesium the nimodipine and magnesium group and for NSE (serum
group experienced an in-hospital rebleeding. Two patients and CSF) in the magnesium group (Table 3). However, there
was no correlation between
any marker and patient out-
TABLE 1. Baseline characteristics of the study populationa come.
Nimodipine Magnesium The data concerning blood
Characteristic P value flow velocities as measured by
group (n ⴝ 51) group (n ⴝ 53)
TCD and the incidence of va-
Age (yr)b 54 ⫾ 18 52 ⫾ 14 0.490
sospasm are presented in Table
Male-to-female ratio 20:31 23:30 0.815
4. There was no statistically
Fisher grade 0.787
significant difference between
Median 3 3
c the two groups. The mean av-
Interquartile range 1 1
erage MCA velocity was 105
1, no. (%) 1 (2.0) 1 (1.9)
⫾ 39 cm/s in the nimodipine
2, no. (%) 16 (31.4) 17 (32.1)
group and 104 ⫾ 45 cm/s in
3, no. (%) 32 (62.8) 30 (56.6)
the magnesium group. The
4, no. (%) 2 (3.9) 5 (9.4)
mean maximum MCA veloc-
WFNS grade 0.682
ity was 141 ⫾ 39 cm/s in the
Median 2 2
c nimodipine group and 145 ⫾
Interquartile range 4 3
55 cm/s in the magnesium
1, no. (%) 14 (27.5) 15 (28.3)
group. Fourteen patients
2, no. (%) 15 (29.4) 16 (30.2)
(27%) in the nimodipine group
3, no. (%) 0 (0.0) 2 (3.8)
and eight patients (15%) in the
4, no. (%) 8 (15.7) 9 (17.0)
magnesium group experi-
5, no. (%) 14 (27.5) 11 (20.8)
enced clinical vasospasm. Sev-
Treatment 0.55
enteen patients (33%) in the ni-
Clip, no. (% ) location of aneurysms 38 (74.5) AC, 37; PC, 1 40 (75.5) AC, 35; PC, 5
modipine group and 20
Coil, no. (%) location of aneurysms 10 (19.6) AC, 6; PC, 4 12 (22.6) AC, 7; PC, 5
patients (38%) in the magne-
Conservative, no. (%) location of 3 (5.9) AC, 3; PC, 0 1 (1.9) AC, 0; PC, 1
sium group experienced
aneurysms
TCD/angiographic vaso-
a
WFNS, World Federation of Neurological Surgeons; AC, anterior circulation; PC, posterior circulation. spasm. This includes five pa-
b
Values are expressed as the mean ⫾ standard deviation.
c
tients of each group in whom
The interquartile range is the difference between the 25th and 75th percentiles.
the diagnosis of clinical vaso-
spasm was not made because
the patients remained continu-
ously at Glasgow Coma Scale
TABLE 2. Magnesium levels at baseline and in the course of treatmenta score of 3 with or without se-
Nimodipine group Magnesium group dation and, therefore, the cri-
P value
magnesium level (mmol/L) magnesium level (mmol/L) teria of secondary deteriora-
Baseline (serum) 0.81 ⫾ 0.01 0.78 ⫾ 0.09 0.243 tion or new focal neurological
Baseline (CSF) 1.03 ⫾ 0.16 1.02 ⫾ 0.27 0.912 deficit was not met. Four of
Treatment (serum) 0.83 ⫾ 0.15 1.46 ⫾ 0.30 ⱕ 0.001 the five patients in each group
Treatment (CSF) 1.14 ⫾ 0.21 1.19 ⫾ 0.20 0.256 experienced cerebral infarc-
tion as demonstrated by CT or
a
CSF, cerebrospinal fluid. Values are expressed as mean ⫾ standard deviation.
magnetic resonance imaging
scans. In patients with clinical

NEUROSURGERY VOLUME 58 | NUMBER 6 | JUNE 2006 | 1057

SCHMID-ELSAESSER ET AL.

vasospasm, seven (50%) of 14

TABLE 3. Neuronal markers (S-100 and neuron-specific enolase)a in the nimodipine group and
Nimodipine Magnesium six (75%) of eight in the mag-
P value nesium group experienced
group (␮g/L) group (␮g/L)
cerebral infarction. Three pa-
All WFNS grades
tients in each group experi-
S-100 (serum) 1.0 ⫾ 2.1 1.3 ⫾ 3.4 0.534
enced cerebral infarction with-
S-100 (CSF) 25.6 ⫾ 28.7 53.8 ⫾ 63.1 0.051
out evidence of vasospasm
NSE (serum) 17.9 ⫾ 26.2 18.6 ⫾ 13.5 0.862
because of other causes such
NSE (CSF) 40.3 ⫾ 55.2 62.3 ⫾ 59.1 0.226
as permanent vessel occlusion
WFNS Grades 1–3
during surgery (n ⫽ 1) or the
S-100 (serum) 0.39 ⫾ 0.5b 0.55 ⫾ 1.0b 0.480
endovascular procedure (n ⫽
S-100 (CSF) 24.3 ⫾ 45.1 21.34 ⫾ 22.7 0.878
2), after evacuation of an intra-
NSE (serum) 12.67 ⫾ 5.3 15.15 ⫾ 5.3b 0.082
cerebral hematoma (n ⫽ 2),
NSE (CSF) 15.02 ⫾ 5.4 28.73 ⫾ 19.3b 0.155
and cardiogenic sources of
WFNS Grades 4 –5
embolism (n ⫽ 1). Overall, the
S-100 (serum) 1.68 ⫾ 2.9b 2.56 ⫾ 5.2b 0.519
incidence of cerebral infarction
S-100 (CSF) 26.18 ⫾ 0.88 55.69 ⫾ 48.5 0.103
was comparable (27 versus
NSE (serum) 24.48 ⫾ 38.3 24.39 ⫾ 19.9b 0.993
25%). Eleven patients in the ni-
NSE (CSF) 48.78 ⫾ 61.8 68.84 ⫾ 44.2b 0.353
modipine group (22%) and 10
a
WFNS, World Federation of Neurological Surgeons; CSF, cerebrospinal fluid; NSE, neuron-specific enolase. Values are patients in the magnesium
expressed as mean ⫾ standard deviation of the maximum value of each patient in the respective group. group (19%) experienced a ce-
b
P ⬍ 0.05 for WFNS 1–3 versus WFNS 4 – 5.
rebral infarction attributable to
vasospasm (Table 5).
The outcomes at discharge
and at the 1-year follow-up ex-
TABLE 4. Blood flow velocities and incidence of vasospasma amination are presented in Ta-
Nimodipine Magnesium ble 6. Overall, there was no dif-
P value
group (n ⴝ 51) group (n ⴝ53) ference between groups. At
TCD
discharge, in the nimodipine
VMCA average (cm/s)b 105 ⫾ 39 104 ⫾ 45 0.687
group, 28 patients (55%) had a
VMCA maximum (cm/s)b 141 ⫾ 39 145 ⫾ 55 0.610
good outcome (GOS 4–5) and
Incidence of vasospasm
23 patients (45%) a poor or fa-
TCD/angiographic vasospasm, no. (%) 17 (33) 20 (38) 0.792
tal outcome (GOS 1–3) versus
Clinical vasospasm, no. (%) 14 (27) 8 (15) 0.193
29 patients (55%) and 24 pa-
tients (45%) in the magnesium
a
TCD, transcranial Doppler; VMCA, blood flow velocity in the middle cerebral artery. group. In patients with clinical
b
Values are expressed as mean ⫾ standard deviation.
vasospasm, seven (50%) of 14
patients in the nimodipine
group and five (62.5%) of eight
patients in the magnesium
TABLE 5. Incidence of cerebral infarctiona group had a poor or fatal out-
come (GOS 1–3). In patients
Nimodipine Magnesium
P value with TCD/angiographic vaso-
group, no. (%) group, no. (%)
spasm, 11 (65%) of 17 patients
Patients with TCD/angiographic vasospasm 4 out of 51 (8)b 4 out of 53 (8)b 0.755 in the nimodipine group and
out of 17 (24)c out of 20 (20)c 0.888 10 (50%) of 20 patients in the
Patients with clinical vasospasm 7 out of 51 (14)b 6 out of 53 (11)b 0.941 magnesium group had a poor
out of 14 (50)d out of 8 (75)d 0.486 or fatal outcome.
Other identified causes 3 out of 51 (6)b 3 out of 53 (6)b 0.710 After 1 year, 10 patients
Total 14 out of 51 (27) b 13 out of 53 (25)b 0.908 (seven patients in the nimo-
a
TCD, transcranial Doppler. dipine group and three pa-
b
Percentage of all patients in the respective group. tients in the magnesium
c
Percentage of patients with TCD/angiographic vasospasm in the respective group. group) were lost to follow-up.
d
Percentage of patients with clinical vasospasm in the respective group.
All patients lost to follow-up
had a poor outcome (GOS 2 or

1058 | VOLUME 58 | NUMBER 6 | JUNE 2006 www.neurosurgery-online.com

 MAGNESIUM VERSUS NIMODIPINE IN SUBARACHNOID HEMORRHAGE

TABLE 6. Outcome at discharge and after 1 yeara

At discharge After 1 year

Nimodipine group, Magnesium group, Nimodipine group, Magnesium group,

P value P value
n ⴝ 51 (%) n ⴝ 53 (%) n ⴝ 44 (%) n ⴝ 50 (%)
GOS scores 0.807 0.373
Median 4 4 5 5
Interquartile rangeb 2 2.25 2 3
5 19 (37.3) 24 (45.3) 29 (65.9) 27 (54.0)
4 9 (17.7) 5 (9.4) 2 (4.5) 7 (14.0)
3 11 (21.6) 11 (20.8) 5 (11.4) 3 (6.0)
2 5 (9.8) 3 (5.7) 0 (0) 1 (2.0)
1 7 (13.7) 10 (18.9) 8 (18.2) 12 (24.0)
Poor outcome after
TCD/angiographic
vasospasm
GOS 2 and 3 8 out of 51 (16)c 5 out of 53 (9)c 0.505 4 out of 44 (9)c 1 out of 50 (2)c 0.285
out of 17 (47)d out of 20 (25)d 0.291 out of 15 (27)d out of 18 (5)d 0.231
GOS 1 3 out of 51 (6)c 5 out of 53 (9)c 0.755 3 out of 44 (7)c 5 out of 50 (10)c 0.856
out of 17 (18)d out of 20 (25)d 0.888 out of 15 (20)d out of 18 (28)d 0.911
Total 11 out of 51 (22)c 10 out of 53 (19)c 0.921 7 out of 44 (16)c 6 out of 50 (12)c 0.804
out of 17 (65)d out of 20 (50)d 0.571 out of 15 (47)d out of 18 (33)d 0.672
Poor outcome after
clinical vasospasm
GOS 2 and 3 5 out of 51 (10)c 2 out of 53 (4)c 0.403 1 out of 44 (2)c 1 out of 50 (2)c 0.532
out of 14 (36)e out of 8 (25)e 0.966 out of 11 (9)e out of 7 (14)e 0.669
GOS 1 2 out of 51 (4)c 3 out of 53 (6)c 0.965 2 out of 44 (5)c 3 out of 50 (6)c 0.883
out of 14 (14)e out of 8 (37)e 0.471 out of 11 (18)e out of 7 (43)e 0.549
Total 7 out of 51 (14)c 5 out of 53 (10)c 0.706 3 out of 44 (7)c 4 out of 50 (8)c 0.860
out of 14 (50)d out of 8 (62.5)e 0.903 out of 11 (27)e out of 7 (57)e 0.440
a
GOS, Glasgow outcome scale; TCD, transcranial Doppler.
b
The interquartile range is the difference between the 25th and 75th percentiles.
c
Percentage of all patients in the respective group.
d
Percentage of patients with TCD/angiographic vasospasm in the respective group.
e
Percentage of patients with clinical vasospasm in the respective group.

3) at discharge. Therefore, the outcome of nimodipine-treated
patients improved when compared with magnesium-treated pa-
tients.
DISCUSSION
The clinical use of intravenous MgSO4 in the prevention of
vasospasm after aneurysmal SAH was compared with intra-
venous nimodipine in this randomized trial. No statistically
significant difference was observed in the proposed end
points, which included incidence of clinical vasospasm and
cerebral infarction as primary outcome measures and inci-
dence of TCD/angiographic vasospasm, neuronal markers,
and GOS scores as secondary outcome measures.
Tolerability and Adverse Effects
Aside from a temporary flushing sensation, the use of
MgSO4 was well tolerated, as reported in many other studies
(14, 47, 55–57, 87, 92). Thus, we can confirm its safety and
tolerability in SAH patients. In contrast, nimodipine use is
known to have several adverse effects. In our study, six pa-
tients (12%) treated with nimodipine experienced systemic
hypotension, a well-known side effect that is more pro-
nounced after intravenous rather than oral administration (69,
72, 98). Other serious side effects of nimodipine described in
the literature, including pulmonary ventilation or perfusion
mismatch, hypoxia, and reduction of brain tissue partial pres-
sure of oxygen, were not observed (19, 27, 44, 83).
Another potentially adverse effect is that the use of calcium
antagonists may be associated with an increased frequency of
rebleeding (7, 8, 29, 41). In our study, no patients in the
nimodipine group and eight patients in the magnesium group
experienced an in-hospital rebleeding. There was no postop-
erative hemorrhage. Two patients admitted on the day of
hemorrhage rebled on the same day after admission to the
hospital, but before definite treatment of the aneurysm. This is

NEUROSURGERY VOLUME 58 | NUMBER 6 | JUNE 2006 | 1059

SCHMID-ELSAESSER ET AL.
within, or even well below, the range reported in the literature
(23, 35, 52, 77, 82). The other six patients rebled during an
endovascular procedure, one patient during diagnostic an-
giography and five patients during endovascular coiling. The
difference between magnesium- and nimodipine-treated pa-
tients concerning rehemorrhage is statistically significant.
However, it is difficult to explain. Reports on platelet-
inhibiting effects exist for both magnesium and nimodipine,
and these effects are most likely too negligible to actively
cause a rehemorrhage (25, 38, 75, 81). One may argue that the
hypotensive effect of intravenous nimodipine possibly pro-
tected the patients. However, most rebleeding occurred under
anesthesia and there was no difference in blood pressure. In a
large randomized multicenter trial of magnesium sulfate ver-
sus nimodipine for the prevention of eclampsia, there was no
difference concerning coagulopathy between groups; how-
ever, magnesium-treated patients were more likely to require
additional medication for blood pressure control and there
was also a higher incidence of postpartum hemorrhage in the
magnesium group (11). The data of the latest Cochrane Data-
base analysis of calcium antagonists for aneurysmal SAH sug-
gest that the use of calcium antagonists in patients with SAH
is not associated with an increased frequency of rebleeding
and that there is, in contrast, a tendency toward a reduction in
the rate of rebleeding (76). This reduction was independent
from the clinical condition on admission of patients and from
the type of calcium antagonists used. However, an antiplatelet
effect may reduce the risk of delayed cerebral ischemia (DCI)
in patients with SAH (20). Nevertheless, close surveillance of
the proportion of patients with rebleeding or postoperative
hemorrhage seems justified in future studies on magnesium
therapy after SAH (41, 87).
Study End Points
The incidence of clinical vasospasm was less in the magne-
sium group than in the nimodipine group (15 versus 27%).
However, if clinical vasospasm occurred in magnesium-
treated patients, it was more likely to cause cerebral infarction
(75 versus 50%) and result in fatal outcome (37 versus 14%)
than in nimodipine-treated patients. This would favor the
theory that nimodipine acts via direct neuroprotective mech-
anisms rather than by preventing vasospasm (50, 74), whereas
magnesium, apart from its multiple potentially neuroprotec-
tive properties, may be effective on account of its cerebrovas-
cular dilatory activity (37, 90). Magnesium is beneficial in the
treatment of eclampsia, a disease with a pathophysiology
comparable with that of DCI after SAH, by reducing cerebral
vasoconstriction and ischemia (12). Interestingly, in a large
multicenter trial, magnesium sulfate was more effective than
nimodipine for prophylaxis against seizures in women with
pre-eclampsia (11). Because magnesium did not affect TCD
flow velocities in our study or other studies (12, 16, 92), it is
conceivable that it acts by dilation of smaller arteries or of
collateral or anastomotic blood flow pathways, or both (46). In
our pilot study, the difference in incidence of clinical vaso-
1060 | VOLUME 58 | NUMBER 6 | JUNE 2006
spasm was not statistically significant. Concerning this end
point, a total sample size of 332 patients would be required to
reach statistical significance with the usual ␣ ⫽ 0.05 and a
desired power of 0.8. Using the Yate’s correction factor, the
required sample size increases to 466 patients. However, the
rate of infarction attributable to vasospasm was virtually the
same (19 versus 22%).
Furthermore, this study might have underestimated the
incidence of TCD/angiographic vasospasm because control
angiography was not performed routinely. Only very low or
very high TCD velocities in the MCA (⬍120 or ⬎200 cm/s)
negatively (94%) or positively (87%) predict angiographic and
clinical vasospasm in a reliable manner (93). Okada et al. (62)
compared TCD velocity with angiographic features and cere-
bral circulation time. Overall, TCD had a sensitivity of 84%
and a specificity of 89% for detecting vasospasm of the MCA
compared with angiography. Although TCD sonography may
warn of spasm development, TCD sonography itself is not
entirely accurate. The technique is quite operator dependent,
and the results may be misleading because of false-negative
findings attributed to vasospasm of peripheral sites (34). How-
ever, in our study, the latter should have been indicated by the
Lindegaard Index (1, 42, 43).
The neuronal markers NSE and S-100 are potentially capa-
ble of providing information about the extent of neurological
injury (17, 51, 66, 97). Some studies suggest that these markers,
especially S-100, may be used as an early measure of the
degree of brain damage after SAH and that they may have
prognostic value for SAH patients (33, 45, 64, 84, 96). In
accordance with these studies, we found higher levels in
WFNS Grade 4 and 5 patients compared with WFNS Grade
1–3 patients. However, as also reported by Kuroiwa et al. (39),
there was no correlation between any marker and patient
outcome in our study.
At discharge, 45% of the magnesium-treated patients and
37% of nimodipine-treated patients had an excellent outcome
(GOS 5), whereas 45% in both groups had a poor outcome
(GOS 1–3). Considering that 40% of the patients had a poor
clinical grade (WFNS 4–5) in our study, the results are in the
range of, or compare favorably with, other studies with 10 to
25% of poor-grade SAH patients (63, 68, 89).
Other Clinical Studies on Nimodipine and Route of
Administration
Recently, Rinkel et al. (76) presented a meta-analysis of
eight nimodipine trials versus placebo that included one trial
with intravenous administration (49), two trials with intrave-
nous followed by oral administration (32, 61), and five trials
with oral administration (3, 48, 58, 65, 67, 68). The results of
these trials with regard to DIND (17–66%) and infarction
(33–71%) in the control groups (Table 7) were clearly inferior to
our results with magnesium (15 and 19%, respectively). This
supports the assumption that magnesium is effective in pre-
venting the sequelae of vasospasm after aneurysmal SAH.
www.neurosurgery-online.com
 MAGNESIUM VERSUS NIMODIPINE IN SUBARACHNOID HEMORRHAGE

TABLE 7. Earlier studies on nimodipine versus placebo according to Rinkel et al. (76)a
Study (ref. no.) Event Treatment n/N (%) Control n/N (%) Relative risk (95% CI)
Intravenously only
Messeter et al., 1987 (49) DIND 1/13 (8%) 3/7 (43%) 0.18 (0.02, 1.42)
Infarction — — —
Intravenously followed by orally
Han et al., 1993 (32) DIND 29/142 (20%) 51/180 (28%) 0.72 (0.48, 1.07)
Infarction — — —
Ohman et al., 1991 (61) DIND 14/104 (13%) 31/109 (28%) 0.47 (0.27, 0.84)
Infarction 34/88 (39%) 49/92 (53%) 0.73 (0.52, 1.00)
Orally only
Allen et al., 1983 (3) DIND 5/56 (9%) 10/60 (17%) 0.54 (0.20, 1.47)
Infarction — — —
Neil-Dwyer et al., 1987 (58) DIND 3/25 (12%) 5/25 (20%) 0.60 (0.16, 2.25)
Mee et al., 1988 (48) Infarction — — —
Petruk et al., 1988 (65) DIND 33/72 (46%) 54/82 (66%) 0.70 (0.52, 0.94)
Infarction 21/33 (64%) 32/45 (71%) 0.89 (0.65, 1.23)
Philippon et al, 1986 (67) DIND 7/31 (23%) 17/39 (44%) 0.52 (0.25, 1.09)
Infarction — — —
Pickard et al., 1989 (68) DIND — — —
Infarction 61/278 (22%) 92/276 (33%) 0.66 (0.50, 0.87)
a
CI, confidence interval; DIND, delayed ischemic neurological deficits.

Only limited information is available concerning the opti-
mal route of administration. Although the oral form of nimo-
dipine is, by far, the most commonly used form in North
America, intravenous administration is still widely spread in
Europe and other countries. In their review, Rinkel et al. (76)
emphasize that the available data provide robust evidence
only for oral nimodipine administration and that calcium
antagonists reduce the risk of poor outcome (relative risk, 0.7)
and secondary ischemia after aneurysmal SAH, but this evi-
dence depends mainly on one large trial. If the nimodipine
studies were analyzed without the largest trial with oral ni-
modipine (68), the benefit in terms of reduction in poor out-
come is no longer statistically significant. Furthermore, poor-
grade patients (WFNS 4 and 5) who have the highest risk of
experiencing vasospasm and who have to receive oral nimo-
dipine through a nasogastric tube are not sufficiently repre-
sented in these studies. The efficacy of ground nimodipine
tablets administered through a nasogastric tube is limited.
According to the manufacturer, the bioavailability of the ac-
tive substance is substantially decreased by grinding because
short-term dissolved nimodipine recrystallizes (71). Others
who favor intravenous administration argue that oral versus
intravenous nimodipine has never been formally tested and
directly compared. In a review of several thousand reported
cases, the overall incidence of DIND was somewhat lower
when intravenous rather than oral nimodipine was used (10,
21). Intracarotid injection of nimodipine during angiography
for treatment of cerebral vasospasm has been reported earlier
with conflicting results (15, 30). In a more recent series, dila-
tation of infused vessels occurred in 13 of 30 procedures (13).
The only available data concerning intraarterial magnesium
administration in humans indicate that infusion improves
endothelium-dependent vasodilation in the forearm without
affecting systemic hemodynamics (31).
Other Clinical Studies on Magnesium in Aneurysmal
SAH
There are three recent studies of magnesium in patients
with aneurysmal SAH (14, 89, 92). In all these studies, mag-
nesium sulfate was given in addition to nimodipine, in con-
trast to our study, in which magnesium and nimodipine were
directly compared. The first study established administration
guidelines for the use of MgSO4 in aneurysmal SAH (14). Ten
patients with Fisher Grade 3 aneurysmal SAH were given a
bolus as well as a constant infusion of intravenous MgSO4 up
to 10 days after ictus. The goal was to raise the serum level to
2.0 to 2.5 mmol/L or twice the baseline serum level. Five
patients exhibited evidence of vasospasm on TCD and three
patients exhibited clinical evidence of vasospasm. In the sec-
ond trial of 40 patients with SAH (92), the authors compared
nimodipine and MgSO4 (delivered as a 6-g intravenous bolus
followed by a 2-g/h infusion) for 10 days with nimodipine
and placebo. Clinical vasospasm occurred in six out of 20
patients receiving MgSO4 and in five of 16 patients receiving
placebo. The authors reported a trend in which a higher
percentage of patients obtained GOS scores of 4 or 5 in the
group treated with MgSO4, but the trend did not reach a

NEUROSURGERY VOLUME 58 | NUMBER 6 | JUNE 2006 | 1061

SCHMID-ELSAESSER ET AL.
statistically significant level. More importantly, in this last
trial, patients in the placebo group received routine infusions
of MgSO4 if serum Mg was less than 0.8 mmol/L. Thus, these
two studies were not suitable to compare the efficacy of mag-
nesium and nimodipine or to detect a beneficial effect of
supplementary magnesium administration in addition to ni-
modipine. In the third study, the primary outcome measure
was DCI, defined as the occurrence of a new hypodense lesion
on CT scan compatible with clinical features of DCI (89). Two
hundred eighty-three patients were randomized to receive a
continuous intravenous dose of MgSO4 (64 mmol/L per day
until 14 d after occlusion of the aneurysm) or placebo. Admin-
istration of magnesium in addition to nimodipine reduced the
risk of DCI by 34% with a number of 14 patients needed to
treat to avoid one delayed cerebral ischemic event. After 3
months, the risk reduction for poor outcome was 23%. At that
time, 18 patients in the treatment group and six in the placebo
group had an excellent outcome. This study suggests that
magnesium reduces DCI and subsequent poor outcome, but
the results are not yet definitive.
Besides possibly additive neuroprotective mechanisms, there
are several arguments that favor a combined administration of
nimodipine and magnesium. In one comparison between 13
patients with SAH who were treated with MgSO4 in addition to
nimodipine to maintain a plasma level of Mg at 1 to 1.5 mmol/L
and 10 historical controls, the authors found a significant reduc-
tion in angiographically diagnosed vasospasm in response to
MgSO4 treatment (18). Unfortunately, the magnesium serum
levels of the historical controls are not known. Barile et al. (9)
reported that intravenous magnesium sulfate administration re-
versed otherwise refractory vasospasm after SAH in a patient
receiving intravenous nimodipine. Furthermore, low serum
magnesium predicts neurological events in patients with ad-
vanced atherosclerosis (6), and there is a relationship between a
low magnesium concentration in CSF during the first 48 hours
after onset of ischemic stroke and the intensity of the neurological
deficit (40). Hypomagnesemia may increase neuronal damage
after SAH through vasoconstriction, calcium influx, and lack of
energy stores (4, 88). van der Bergh et al. (88) demonstrated that
hypomagnesemia frequently is present after SAH, and, occurring
between Days 2 and 12 after SAH, it is a strong predictor for DCI.
In conclusion, this pilot study suggests that the efficacy of
magnesium in preventing DIND in patients with aneurysmal
SAH is comparable with that of nimodipine. Large, multi-
center trials are needed to define the role of magnesium in
SAH better, and especially studies on the combined adminis-
tration of nimodipine and magnesium seem to be promising.
In the meantime, magnesium levels should be routinely mon-
itored and, at least, kept within the high normal range in SAH
patients.
REFERENCES
1. Aaslid R: Transcranial Doppler assessment of cerebral vasospasm. Eur J
Ultrasound 16:3–10, 2002.
1062 | VOLUME 58 | NUMBER 6 | JUNE 2006
2. Ahmed N, Nasman P, Wahlgren NG: Effect of intravenous nimodipine on
blood pressure and outcome after acute stroke. Stroke 31:1250–1255, 2000.
3. Allen GS, Ahn HS, Preziosi TJ, Battye R, Boone SC, Boone SC, Chou SN,
Kelly DL, Weir BK, Crabbe RA, Lavik PJ, Rosenbloom SB, Dorsey FC,
Ingram CR, Mellits DE, Bertsch LA, Boisvert DP, Hundley MB, Johnson RK,
Strom JA, Transou CR: Cerebral arterial spasm—A controlled trial of
nimodipine in patients with subarachnoid hemorrhage. N Engl J Med
308:619–624, 1983.
4. Altura BT, Altura BM: Withdrawal of magnesium causes vasospasm while
elevated magnesium produces relaxation of tone in cerebral arteries.
Neurosci Lett 20:323–327, 1980.
5. 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
3:195–211, 1984.
6. Amighi J, Sabeti S, Schlager O, Mlekusch W, Exner M, Lalouschek W,
Ahmadi R, Minar E, Schillinger M: Low serum magnesium predicts neuro-
logical events in patients with advanced atherosclerosis. Stroke 35:22–27,
2004.
7. Arbit E, Kaplan R, Gillich A, Brem S, Hier M: Can calcium antagonists affect
the hemostatic mechanism and promote rebleeding? An experimental study.
Neurosurgery 20:732–734, 1987.
8. Assaley J, Baron JM, Cibils LA: Effects of magnesium sulfate infusion upon
clotting parameters in patients with pre-eclampsia. J Perinat Med 26:115–
119, 1998.
9. Barile M, van De Wyngaert F, Mbia JJ, Jativa M, Grandin C, Rooijakkers H,
Hantson P: Intravenous magnesium sulfate administration in a patient with
refractory vasospasm following subarachnoid hemorrhage. Intensive Care
Med 29:1182–1185, 2003.
10. Barker FG, Ogilvy CS: Efficacy of prophylactic nimodipine for delayed
ischemic deficit after subarachnoid hemorrhage: A metaanalysis.
J Neurosurg 84:405–414, 1996.
11. Belfort MA, Anthony J, Saade GR, Allen JC: A comparison of magnesium
sulfate and nimodipine for the prevention of eclampsia. N Engl J Med
348:304–311, 2003.
12. 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 167:661–666, 1992.
13. Biondi A, Ricciardi GK, Puybasset L, Abdennour L, Longo M, Chiras J, van
Effenterre R: Intra-arterial nimodipine for the treatment of symptomatic
cerebral vasospasm after aneurysmal subarachnoid hemorrhage: Prelimi-
nary results. AJNR Am J Neuroradiol 25:1067–1076, 2004.
14. Boet R, Mee E: Magnesium sulfate in the management of patients with
Fisher Grade 3 subarachnoid hemorrhage: A pilot study. Neurosurgery
47:602–606, 2000.
15. Boker DK, Solymosi L, Wassmann H: Immediate postangiographic intraar-
terial treatment of cerebral vasospasm after subarachnoid hemorrhage with
nimodipine. Report on 3 cases. Neurochirurgia (Stuttg) 28 [Suppl 1]:118–
120, 1985.
16. Brewer RP, Parra A, Lynch J, Chilukuri V, Borel CO: Cerebral blood flow
velocity response to magnesium sulfate in patients after subarachnoid hem-
orrhage. J Neurosurg Anesthesiol 13:202–206, 2001.
17. Butterworth RJ, Wassif WS, Sherwood RA, Gerges A, Poyser KH,
Garthwaite J, Peters TJ, Bath PM: Serum neuron-specific enolase,
carnosinase, and their ratio in acute stroke. An enzymatic test for predicting
outcome? Stroke 27:2064–2068, 1996.
18. Chia RY, Hughes RS, Morgan MK: Magnesium: A useful adjunct in the
prevention of cerebral vasospasm following aneurysmal subarachnoid
haemorrhage. J Clin Neurosci 9:279–281, 2002.
19. Devlin JW, Coplin WM, Murry KR, Rengachary SS, Wilson RF: Nimodipine-
induced acute hypoxemia: Case report. Neurosurgery 47:1243–1246, 2000.
20. Dorhout Mees SM, Rinkel GJ, Hop JW, Algra A, van Gijn J: Antiplatelet
therapy in aneurysmal subarachnoid hemorrhage. A systematic review.
Stroke 34:2285–2289, 2003.
21. Dorsch NW: Therapeutic approaches to vasospasm in subarachnoid hem-
orrhage. Curr Opin Crit Care 8:128–133, 2002.
www.neurosurgery-online.com
 MAGNESIUM
22. Drake CG, Hunt WE, Kassell N, Pertuiset B, Sano K, Teasdale G, De Villiers
JC: Report of World Federation of Neurological Surgeons Committee on a
universal subarachnoid hemorrhage grading scale. J Neurosurg 68:985–986,
1988.
23. Edner G, Ronne-Engstrom E: Can early admission reduce aneurysmal
rebleeds? A prospective study on aneurysmal incidence, aneurysmal
rebleeds, admission and treatment delays in a defined region. Br
J Neurosurg 5:601–608, 1991.
24. Feigin VL, Rinkel GJ, Algra A, Vermeulen M, van Gijn J: Calcium antago-
nists for aneurysmal subarachnoid haemorrhage. Cochrane Database Syst
Rev 2:CD000277, 2000.
25. Feinberg WM, Bruck DC: Effect of oral nimodipine on platelet function.
Stroke 24:10–13, 1993.
26. Fisher CM, Kistler JP, Davis JM: Relation of cerebral vasospasm to subarach-
noid hemorrhage visualized by computerized tomographic scanning.
Neurosurgery 6:1–9, 1980.
27. Gerloni R, Copetti R: Easily reversible hypoxemia and hypotension induced
by nimodipine. Eur J Emerg Med 11:295–297, 2004.
28. Ghribi O, Callebert J, Verrecchia C, Plotkine M, Boulu RG: Blockers of
NMDA-operated channels decrease glutamate and aspartate extracellular
accumulation in striatum during forebrain ischaemia in rats. Fundam Clin
Pharmacol 9:141–146, 1995.
29. Gore JM, Sloan M, Price TR, Randall AM, Bovill E, Collen D, Forman S,
Knatterud GL, Sopko G, Terrin ML: Intracerebral hemorrhage, cerebral
infarction, and subdural hematoma after acute myocardial infarction and
thrombolytic therapy in the Thrombolysis in Myocardial Infarction Study.
Thrombolysis in Myocardial Infarction, Phase II, pilot and clinical trial.
Circulation 83:448–459, 1991.
30. Grotenhuis JA, Bettag W, Fiebach BJ, Dabir K: Intracarotid slow bolus
injection of nimodipine during angiography for treatment of cerebral vaso-
spasm after SAH. A preliminary report. J Neurosurg 61:231–240, 1984.
31. Haenni A, Johansson K, Lind L, Lithell H: Magnesium infusion improves
endothelium-dependent vasodilation in the human forearm.
Am J Hypertens 15:10–15, 2002.
32. Han DH, Lee SH, Lee SH: Effect of nimodipine treatment on outcome in
surgical cases of aneurysmal SAH. J Neurosurg 78:346A, 1993.
33. Hardemark HG, Almqvist O, Johansson T, Pahlman S, Persson L: S-100
protein in cerebrospinal fluid after aneurysmal subarachnoid haemorrhage:
Relation to functional outcome, late CT and SPECT changes, and signs of
higher cortical dysfunction. Acta Neurochir (Wien) 99:135–144, 1989.
34. Harrod CG, Bendok BR, Batjer HH: Prediction of cerebral vasospasm in
patients presenting with aneurysmal subarachnoid hemorrhage: A review.
Neurosurgery 56:633–654, 2005.
35. Hillman J, Fridriksson S, Nilsson O, Yu Z, Säveland H, Jakobsson KE:
Immediate administration of tranexamic acid and reduced incidence of early
rebleeding after aneurysmal subarachnoid hemorrhage: A prospective ran-
domized study. J Neurosurg 97:771–778, 2002.
36. Huang QF, Gebrewold A, Altura BT, Altura BM: Mg2⫹ protects against
PCP-induced cerebrovasospasms and vascular damage in rat brain. Magnes
Trace Elem 9:44–46, 1990.
37. Iseri LT, French JH: Magnesium: Nature’s physiologic calcium blocker. Am
Heart J 108:188–193, 1984.
38. Juvela S, Kaste M, Hillbom M: Effect of nimodipine on platelet function in
patients with subarachnoid hemorrhage. Stroke 21:1283–1288, 1990.
39. Kuroiwa T, Tanabe H, Arai M, Ohta T: Measurement of serum neuron-
specific enolase levels after subarachnoid hemorrhage and intracerebral
hemorrhage [in Japanese]. No Shinkei Geka 22:531–535, 1994.
40. Lampl Y, Geva D, Gilad R, Eshel Y, Ronen L, Sarova-Pinhas I: Cerebrospinal
fluid magnesium level as a prognostic factor in ischaemic stroke. J Neurol
245:584–588, 1998.
41. Legault C, Furberg CD, Wagenknecht LE, Rogers AT, Stump DA, Coker L,
Troost BT, Hammon JW: Nimodipine neuroprotection in cardiac valve
replacement: Report of an early terminated trial. Stroke 27:593–598, 1996.
42. Lindegaard KF, Nornes H, Bakke SJ, Sorteberg W, Nakstad P: Cerebral
vasospasm diagnosis by means of angiography and blood velocity measure-
ments. Acta Neurochir (Wien) 100:12–24, 1989.
43. Lindegaard KF, Sorteberg W, Nornes H: Transcranial Doppler in neurosur-
gery. Adv Tech Stand Neurosurg 20:39–80, 1993.
NEUROSURGERY
VERSUS NIMODIPINE IN SUBARACHNOID HEMORRHAGE
44. Lovell AT, Smith M: Pulmonary vasoconstriction following intravenous
nimodipine. Anaesthesia 47:409–410, 1992.
45. Mabe H, Suzuki S, Mase M, Umemura A, Nagai H: Serum neuron-specific
enolase levels after subarachnoid hemorrhage. Surg Neurol 36:170–174,
1991.
46. Macdonald RL, Curry DJ, Aihara Y, Zhang ZD, Jahromi BS, Yassari R:
Magnesium and experimental vasospasm. J Neurosurg 100:106–110, 2004.
47. McLean RM: Magnesium and its therapeutic uses: A review. Am J Med
96:63–76, 1994.
48. Mee E, Dorrance D, Lowe D, Neil-Dwyer G: Controlled study of nimodipine
in aneurysm patients treated early after subarachnoid hemorrhage.
Neurosurgery 22:484–491, 1988.
49. Messeter K, Brandt L, Ljunggren B, Svendgaard NA, Algotsson L, Romner
B, Ryding E: Prediction and prevention of delayed ischemic dysfunction
after aneurysmal subarachnoid hemorrhage and early operation.
Neurosurgery 20:548–553, 1987.
50. Meyer FB: Calcium antagonists and vasospasm. Neurosurg Clin N Am
1:367–376, 1990.
51. Meynaar IA, Straaten HM, van der WJ, Verlooy P, Slaats EH, Bosman RJ,
van der Spoel JI, Zandstra DF: Serum neuron-specific enolase predicts
outcome in post-anoxic coma: A prospective cohort study. Intensive Care
Med 29:189–195, 2003.
52. Molyneux A, Kerr R, Stratton I, Sandercock P, Clarke M, Shrimpton J,
Holman R: International Subarachnoid Aneurysm Trial (ISAT) of neurosur-
gical clipping versus endovascular coiling in 2143 patients with ruptured
intracranial aneurysms: A randomised trial. Lancet 360:1267–1274, 2002.
53. Muir KW: New experimental and clinical data on the efficacy of pharma-
cological magnesium infusions in cerebral infarcts. Magnes Res 11:43–56,
1998.
54. Muir KW: Magnesium for neuroprotection in ischaemic stroke: Rationale for
use and evidence of effectiveness. CNS Drugs 15:921–930, 2001.
55. Muir KW, Lees KR: A randomized, double-blind, placebo-controlled pilot
trial of intravenous magnesium sulfate in acute stroke. Stroke 26:1183–1188,
1995.
56. Muir KW, Lees KR: Dose optimization of intravenous magnesium sulfate
after acute stroke. Stroke 29:918–923, 1998.
57. Muir KW, Lees KR, Ford I, Davis S: Magnesium for acute stroke (Intrave-
nous Magnesium Efficacy in Stroke trial): Randomised controlled trial.
Lancet 363:439–445, 2004.
58. Neil-Dwyer G, Mee E, Dorrance D, Lowe D: Early intervention with
nimodipine in subarachnoid haemorrhage. Eur Heart J 8 [Suppl K]:41–47,
1987.
59. Nowak L, Bregestovski P, Ascher P, Herbet A, Prochiantz A: Magnesium
gates glutamate-activated channels in mouse central neurones. Nature 307:
462–465, 1984.
60. Ohman J, Heiskanen O: Effect of nimodipine on the outcome of patients
after aneurysmal subarachnoid hemorrhage and surgery. J Neurosurg 69:
683–686, 1988.
61. Ohman J, Servo A, Heiskanen O: Long-term effects of nimodipine on cere-
bral infarcts and outcome after aneurysmal subarachnoid hemorrhage and
surgery. J Neurosurg 74:8–13, 1991.
62. Okada Y, Shima T, Nishida M, Yamane K, Hatayama T, Yamanaka C,
Yoshida A: Comparison of transcranial Doppler investigation of aneurysmal
vasospasm with digital subtraction angiographic and clinical findings.
Neurosurgery 45:443–449, 1999.
63. Osawa M, Hongo K, Tanaka Y, Nakamura Y, Kitazawa K, Kobayashi S:
Results of direct surgery for aneurysmal subarachnoid haemorrhage: Out-
come of 2055 patients who underwent direct aneurysm surgery and profile
of ruptured intracranial aneurysms. Acta Neurochir (Wien) 143:655–663,
2001.
64. Persson L, Hardemark H, Edner G, Ronne E, Mendel-Hartvig I, Pahlman S:
S-100 protein in cerebrospinal fluid of patients with subarachnoid haemor-
rhage: A potential marker of brain damage. Acta Neurochir (Wien) 93:116–
122, 1988.
65. Petruk KC, West M, Mohr G, Weir BK, Benoit BG, Gentili F, Disney LB,
Khan MI, Grace M, Holness RO: Nimodipine treatment in poor-grade an-
eurysm patients. Results of a multicenter double-blind placebo-controlled
trial. J Neurosurg 68:505–517, 1988.
VOLUME 58 | NUMBER 6 | JUNE 2006 | 1063
SCHMID-ELSAESSER ET AL.
66. Petzold A, Keir G, Lim D, Smith M, Thompson EJ: Cerebrospinal fluid (CSF)
and serum S100B: Release and wash-out pattern. Brain Res Bull 61:281–285,
2003.
67. Philippon J, Grob R, Dagreou F, Guggiari M, Rivierez M, Viars P: Prevention
of vasospasm in subarachnoid haemorrhage. A controlled study with
nimodipine. Acta Neurochir (Wien) 82:110–114, 1986.
68. Pickard JD, Murray GD, Illingworth R, Shaw MD, Teasdale GM, Foy PM,
Humphrey PR, Lang DA, Nelson R, Richards P: Effect of oral nimodipine on
cerebral infarction and outcome after subarachnoid haemorrhage: British
aneurysm nimodipine trial. BMJ 298:636–642, 1989.
69. Porchet F, Chiolero R, de Tribolet N: Hypotensive effect of nimodipine
during treatment for aneurysmal subarachnoid haemorrhage. Acta
Neurochir (Wien) 137:62–69, 1995.
70. Pyne GJ, Cadoux-Hudson TA, Clark JF: Magnesium protection against in
vitro cerebral vasospasm after subarachnoid haemorrhage. Br J Neurosurg
15:409–415, 2001.
71. Raabe A, Beck J, Berkefeld J, Deinsberger W, Meixensberger J, Schmiedek P,
Seifert V, Steinmetz H, Unterberg A, Vajkoczy P, Werner C: Recommenda-
tions for the management of patients with aneurysmal subarachnoid hem-
orrhage [in German]. Zentralbl Neurochir 66:79–91, 2005.
72. Radhakrishnan D, Menon DK: Haemodynamic effects of intravenous
nimodipine following aneurysmal subarachnoid haemorrhage: Implications
for monitoring. Anaesthesia 52:489–491, 1997.
73. Ram Z, Sadeh M, Shacked I, Sahar A, Hadani M: Magnesium sulfate
reverses experimental delayed cerebral vasospasm after subarachnoid hem-
orrhage in rats. Stroke 22:922–927, 1991.
74. Rasmussen G, Bergholdt B, Dalh B, Sunde N, Cold G, Voldby B: Effect of
nimodipine on cerebral blood flow and cerebrovascular reactivity after
subarachnoid haemorrhage. Acta Neurol Scand 99:182–186, 1999.
75. Ravn HB, Vissinger H, Kristensen SD, Wennmalm A, Thygesen K, Husted
SE: Magnesium inhibits platelet activity—An infusion study in healthy
volunteers. Thromb Haemost 75:939–944, 1996.
76. Rinkel G, Feigin V, Algra A, Bergh W, Vermeulen M, Gijn J: Calcium
antagonists for aneurysmal subarachnoid haemorrhage. Cochrane Database
Syst Rev CD000277, 2005.
77. Rosenorn J, Eskesen V, Schmidt K, Ronde F: The risk of rebleeding from
ruptured intracranial aneurysms. J Neurosurg 67:329–332, 1987.
78. Rothman S: Synaptic release of excitatory amino acid neurotransmitter
mediates anoxic neuronal death. J Neurosci 4:1884–1891, 1984.
79. Schmid-Elsaesser R, Hungerhuber E, Zausinger S, Baethmann A, Reulen HJ:
Combination drug therapy and mild hypothermia: A promising treatment
strategy for reversible focal cerebral ischemia. Stroke 30:1891–1899, 1999.
80. Schmid-Elsaesser R, Zausinger S, Hungerhuber E, Baethmann A, Reulen HJ:
Neuroprotective effects of combination therapy with tirilazad and magne-
sium in rats subjected to reversible focal cerebral ischemia. Neurosurgery
44:163–172, 1999.
81. Serebruany VL, Herzog WR, Schlossberg ML, Gurbel PA: Bolus magnesium
infusion in humans is associated with predominantly unfavourable changes
in platelet aggregation and certain haemostatic factors. Pharmacol Res
36:17–22, 1997.
82. Steiger HJ, Fritschi J, Seiler RW: Current pattern of in-hospital aneurysmal
rebleeds. Analysis of a series treated with individually timed surgery and
intravenous nimodipine. Acta Neurochir (Wien) 127:21–26, 1994.
83. Stiefel MF, Heuer GG, Abrahams JM, Bloom S, Smith MJ, Maloney-Wilensky
E, Grady MS, LeRoux PD: The effect of nimodipine on cerebral oxygenation
in patients with poor-grade subarachnoid hemorrhage. J Neurosurg 101:
594–599, 2004.
84. Takayasu M, Shibuya M, Kanamori M, Suzuki Y, Ogura K, Kageyama N,
Umekawa H, Hidaka H: S-100 protein and calmodulin levels in cerebrospi-
nal fluid after subarachnoid hemorrhage. J Neurosurg 63:417–420, 1985.
85. Treggiari MM, Walder B, Suter PM, Romand JA: Systematic review of the
prevention of delayed ischemic neurological deficits with hypertension,
hypervolemia, and hemodilution therapy following subarachnoid hemor-
rhage. J Neurosurg 98:978–984, 2003.
86. Treggiari-Venzi MM, Suter PM, Romand JA: Review of medical prevention
of vasospasm after aneurysmal subarachnoid hemorrhage: A problem of
neurointensive care. Neurosurgery 48:249–261, 2001.
1064 | VOLUME 58 | NUMBER 6 | JUNE 2006
87. van den Bergh WM, Albrecht KW, Berkelbach van der Sprenkel JW, Rinkel
GJ: Magnesium therapy after aneurysmal subarachnoid haemorrhage a
dose-finding study for long term treatment. Acta Neurochir (Wien) 145:
195–199, 2003.
88. van den Bergh WM, Algra A, van der Sprenkel JW, Tulleken CA, Rinkel GJ:
Hypomagnesemia after aneurysmal subarachnoid hemorrhage. Neurosurgery
52:276–281, 2003.
89. van den Bergh WM, Algra A, van Kooten F, Dirven CM, van Gijn J,
Vermeulen M, Rinkel GJ: Magnesium sulfate in aneurysmal subarachnoid
hemorrhage: A randomized controlled trial. Stroke 36:1011–1015, 2005.
90. van den Bergh WM, Dijkhuizen RM, Rinkel GJ: Potentials of magnesium
treatment in subarachnoid haemorrhage. Magnes Res 17:301–313, 2004.
91. van den Bergh WM, Zuur JK, Kamerling NA, van Asseldonk JT, Rinkel GJ,
Tulleken CA, Nicolay K: Role of magnesium in the reduction of ischemic
depolarization and lesion volume after experimental subarachnoid hemor-
rhage. J Neurosurg 97:416–422, 2002.
92. Veyna RS, Seyfried D, Burke DG, Zimmerman C, Mlynarek M, Nichols V,
Marrocco A, Thomas AJ, Mitsias PD, Malik GM: Magnesium sulfate therapy
after aneurysmal subarachnoid hemorrhage. J Neurosurg 96:510–514, 2002.
93. Vora YY, Suarez-Almazor M, Steinke DE, Martin ML, Findlay JM: Role of
transcranial Doppler monitoring in the diagnosis of cerebral vasospasm
after subarachnoid hemorrhage. Neurosurgery 44:1237–1247, 1999.
94. Westermaier T, Hungerhuber E, Zausinger S, Baethmann A, Schmid-
Elsaesser R: Neuroprotective efficacy of intraarterial and intravenous mag-
nesium sulfate in a rat model of transient focal cerebral ischemia. Acta
Neurochir (Wien) 145:393–399, 2003.
95. Westermaier T, Zausinger S, Baethmann A, Schmid-Elsaesser R: Dose find-
ing study of intravenous magnesium sulphate in transient focal cerebral
ischemia in rats. Acta Neurochir (Wien) 147:525–532, 2005.
96. Wiesmann M, Missler U, Hagenstrom H, Gottmann D: S-100 protein plasma
levels after aneurysmal subarachnoid haemorrhage. Acta Neurochir (Wien)
139:1155–1160, 1997.
97. Wunderlich MT, Wallesch CW, Goertler M: Release of neurobiochemical
markers of brain damage is related to the neurovascular status on admission
and the site of arterial occlusion in acute ischemic stroke. J Neurol Sci
227:49–53, 2004.
98. Zygmunt SC, Delgado-Zygmunt TJ: The haemodynamic effect of transcra-
nial Doppler-guided high-dose nimodipine treatment in established vaso-
spasm after subarachnoid haemorrhage. Acta Neurochir (Wien) 135:179–
185, 1995.
COMMENTS
T he authors are to be commended for the tremendous work asso-
ciated with a nonsponsored, single-center trial. However, accruing
patients in a single-center trial is often unpredictable, and the weak-
ness of this design is usually either a small sample size or an unreal-
istic timeframe to include the sample size required for answering the
study question. This is also true for this study.
Magnesium is interesting as a substance for the prophylaxis and
treatment of delayed ischemic neurological deficit (DIND) for a num-
ber of reasons. Although a direct vasodilation seems less likely with
the serum concentration used in this study, magnesium is a calcium
antagonist and neuroprotective agent (i.e., our naturally occurring
“nimodipine”). The authors logically challenged nimodipine with
magnesium, and they did what many of us thought one should do.
The problem, however, is statistics. If we had the paradoxical
situation that magnesium increases the absolute risk of poor outcome
after one year by 5% (e.g., from 35 to 40%), and nimodipine leads to an
absolute risk reduction of poor outcome by 5% (e.g., from 35 to 30%),
the trial had to be designed with about 400 patients to show statistical
significance (power of 80 at 5% significance) and to conclude that
nimodipine is superior to magnesium.
The most likely hypothesis is that for prophylaxis and treatment of
DIND, magnesium is as efficient as nimodipine, but I don’t know how
www.neurosurgery-online.com
 MAGNESIUM
this can be shown by analyzing clinical outcome. This is also true for
upcoming trials of other drugs designed to prevent or treat vaso-
spasm. Therefore, it would be more helpful to use an endpoint that is
influenced only by cerebral vasospasm. It is highly likely that cerebral
vasospasm negatively affects long-term outcome only by causing a
cerebral infarction. Using this endpoint, the sample size required for
finding a statistical difference will probably be lower. Therefore, the
19% infarction rate with magnesium (nimodipine, 22%) is a sound
finding that supports the authors’ hypothesis.
Andreas Raabe
Frankfurt, Germany
T he authors present the results of a single-center, randomized study
comparing intravenous magnesium sulfate versus nimodipine in
the treatment of patients with aneurysmal subarachnoid hemorrhage.
The primary outcome measure of clinical vasospasm was reached by
8 of 53 (15%) patients in the magnesium group versus 14 of 51 (27%)
patients in the nimodipine group. The difference was not statistically
significant. Cerebral infarction attributable to vasospasm was the
other primary outcome measure, and it did not reach statistical sig-
nificance (19% for magnesium vs. 22% for nimodipine). Secondary
outcome measures, including incidence of transcranial Doppler/
angiographic vasospasm, neuronal markers (S-100, neuron-specific
enolase), and patients’ Glasgow Outcome Scale scores at discharge
and at 1 year also did not achieve statistical significance. The authors
concluded that magnesium was comparable to nimodipine in the
prevention of delayed ischemic neurological deficits.
This study adds to the current literature on the role of magnesium
in the prevention of vasospasm. The authors have concluded that
magnesium is no different than nimodipine in the prevention of
clinical vasospasm and the resultant cerebral infarction. As the au-
thors readily admit, their study is statistically underpowered to show
a difference between magnesium and nimodipine. Magnesium could
be either superior or inferior to nimodipine. Therefore, one must be
very careful in interpreting the results of this study.
The major concern of this study remains its design. The fact that
two active treatments are involved make the results a little harder to
interpret. As the authors hint at in their closing statement, the com-
bined use of magnesium and nimodipine may be a more worthwhile
study to perform. A controlled, randomized trial with both groups
receiving nimodipine, and with only one of them receiving magne-
sium, would have been a more robust study design.
Nevertheless, this study and others that have been previously published
lend support to magnesium as a potential beneficial treatment for cerebral
vasospasm, and the time has come for a larger multicenter trial.
Michael Chow
J. Max Findlay
Edmonton, Canada
NEUROSURGERY
VERSUS NIMODIPINE IN SUBARACHNOID HEMORRHAGE
S o far, calcium antagonists in peroral (more in the United States and
at our department), and the intravenous form are the only proven
agents with some effect on vasospasm. However, it still remains a
major cause of morbidity and mortality in SAH patients. Intravenous
magnesium presented in the study seems to have a comparable effect
on vasospasm as nimodipine. However, with fewer adverse effects
and with lower costs, magnesium could be more involved, either
alone or in combination with calcium antagonists, in the fight against
vasospasm. However, this has to be confirmed in larger trials, and we
still need to be looking for novel and more potent agents in the future.
Mika Niemelä
Tarja Randell
Reza Dashti
Juha Hernesniemi
Helsinki, Finland
S chmid-Elsaesser et al. have performed a randomized trial investi-
gating intravenous magnesium versus nimodipine in the treatment
of patients with SAH. After enrolling 113 patients in their study, the
authors found a trend toward a lesser incidence of symptomatic
vasospasm in magnesium treated patients and a lesser incidence of
fatal outcome from symptomatic vasospasm in nimodipine treated
patients. These results support both the predominantly vasodilatory
action of magnesium and neuroprotective effect of nimodipine. Al-
though no significant difference in outcome was observed between
groups, this trial further suggests the importance of magnesium in
preserving cerebrovascular homeostasis following SAH. To this end,
magnesium levels should probably be monitored in this patient pop-
ulation and maintained within the high limits of normal.
It is important to note that prior studies have failed to demonstrate
clinical efficacy for magnesium supplementation (1). It may be, how-
ever, that magnesium’s calcium antagonist effects are redundant and
of little additional benefit for patients already receiving nimodipine.
Thus, in order to accurately assess the benefit of magnesium, it is
necessary to randomize patients to receive either magnesium or ni-
modipine, as was done in this trial. Unfortunately, withholding ni-
modipine from certain individuals may breach ethical boundaries, as
this agent has been shown to improve clinical outcomes and is cur-
rently considered standard of care. Regardless, the continued study of
intravenous magnesium administration in SAH patients is clearly
needed, with more definitive conclusions requiring a phase III trial.
Ricardo J. Komotar
E. Sander Connolly, Jr.
New York, New York
1. Veyna RS, Seyfried D, Burke DG, Zimmerman C, Mlynarek M, Nichols V,
Marrocco A, Thomas AJ, Mitsias PD, Malik GM: Magnesium sulfate therapy
after aneurysmal subarachnoid hemorrhage. J Neurosurg 96:510–514, 2002.
VOLUME 58 | NUMBER 6 | JUNE 2006 | 1065