← Back
NeuroTrials.ai
Neurology Clinical Trial Database

IMASH

Intravenous Magnesium Sulphate for Aneurysmal Subarachnoid Hemorrhage (IMASH): A Randomized, Double-Blinded, Placebo-Controlled, Multicenter Phase III Trial

Share Share Copied! Compare

Year of Publication: 2010

Authors: George Kwok Chu Wong, Wai S. Poon, Matthew T.V. Chan, ..., for the IMASH Investigators

Journal: Stroke

Citation: Stroke. 2010;41:921-926

PDF: https://www.ahajournals.org/doi/epub/10....EAHA.109.571125

Clinical Question

Does intravenous magnesium sulfate infusion for 10-14 days, in addition to oral nimodipine, improve neurological outcome (Extended Glasgow Outcome Scale) at 6 months in patients with acute aneurysmal subarachnoid hemorrhage?

Bottom Line

This Phase III multicenter RCT found no clinical benefit of IV magnesium sulfate infusion over placebo in patients with acute aneurysmal SAH. Despite achieving significantly higher serum magnesium levels, there was no difference in the primary outcome (favorable GOSE 5-8 at 6 months: 64% MgSO4 vs 63% saline, OR 1.0). Secondary outcomes including clinical vasospasm, mRS, Barthel Index, and SF-36 also showed no significant differences. No subgroup showed benefit from magnesium sulfate. The results do not support routine use of IV magnesium sulfate in aSAH patients.

Major Points

  • Multicenter (10 centers), randomized, double-blind, placebo-controlled Phase III trial
  • 327 patients recruited between June 2002 and December 2008 (387 total minus 60 pilot study patients)
  • Treatment initiated within 48 hours of SAH onset (mean 31.7±15.5 hours)
  • Primary outcome (GOSE 5-8 at 6 months): no difference (MgSO4 64% vs saline 63%, OR 1.0, 95% CI 0.7-1.6)
  • Clinical vasospasm: no significant difference (MgSO4 25% vs saline 18%, OR 1.5, 95% CI 0.9-2.5)
  • Good outcome mRS 0-2: no difference (MgSO4 57% vs saline 58%, OR 1.0)
  • Excellent outcome mRS 0-1: no difference (MgSO4 46% vs saline 45%, OR 1.0)
  • Barthel Index ≥85: no difference (MgSO4 57% vs saline 61%, OR 0.9)
  • SF-36 physical and mental scores: no significant differences
  • Average serum magnesium significantly higher in MgSO4 group (1.67±0.27 vs 0.91±0.16 mmol/L, p<0.001)
  • Mean TCD middle cerebral artery velocities: no difference (82±31 vs 87±61 cm/s, p=0.487)
  • 91% completed at least 10 days of study drug infusion
  • Inpatient mortality similar: MgSO4 10% vs saline 12% (OR 0.8, 95% CI 0.4-1.6)
  • Hypotension: MgSO4 15% vs saline 13% (p=0.590)
  • No difference in adverse events between groups
  • No subgroup showed benefit from magnesium sulfate in prespecified analyses
  • Study drug stopped in only 3 patients (1%) due to adverse events
  • Follow-up: 96% at 6 months (11 patients lost to follow-up total)

Design

Study Type: Multicenter, randomized, double-blind, placebo-controlled Phase III trial

Randomization: 1

Blinding: Double-blind. Computer-generated randomization list from central office. Hong Kong and Australia sites: internet randomization. China and Southeast Asia sites: sealed envelopes in order. Group assignments and infusion handling by dedicated site research nurse. Patients, assessors, and healthcare staff blinded to group identity. Treatment allocation concealed from outcome assessors

Enrollment Period: June 1, 2002 to December 31, 2008

Follow-up Duration: 6 months

Centers: 10

Countries: Hong Kong, China, Malaysia, Australia

Sample Size: 327

Analysis: Intention-to-treat. Last observation carried forward for missing 6-month data (4% missing). Sample size calculation based on Veyna's pilot study (65% MgSO4 vs 50% placebo favorable outcome): 348 patients needed for 2-tailed test, 5% significance, 80% power. Chi-square test, Mann-Whitney U test, proportional odds analysis as appropriate. P<0.05 considered significant. CONSORT reporting. OR >1.0 indicates advantage of magnesium (except for vasospasm where OR <1.0 indicates advantage). Post hoc power analysis: 0.85 for 15% improvement, 0.49 for 10% improvement

Inclusion Criteria

  • Age ≥18 years
  • Radiological diagnosis of subarachnoid hemorrhage
  • Intracranial aneurysm considered cause of SAH
  • Randomization within 48 hours after SAH onset
  • Woman of nonchildbearing potential (postmenopausal or physiologically incapable of pregnancy) OR woman of childbearing potential with negative urine pregnancy test immediately before randomization

Exclusion Criteria

  • Death within 48 hours after admission anticipated
  • Major hepatic, pulmonary, or cardiac disease
  • Recent myocardial infarction (within 6 months of ictus)
  • Significant renal impairment (plasma creatinine ≥200 μmol/L)
  • Clinical indication or contraindication to magnesium infusion
  • Pre-existing neurological disability from stroke, dementia, or other neurological diseases
  • Concurrent participation in another clinical trial

Baseline Characteristics

Overall:

  • Number: 327
  • Mean age, years (SD): 57.0 (12.5)
  • Median age, years (range): 56 (19-90)
  • Female: 208 (64%)
  • Male: 119 (36%)

MgSO4 Group:

  • Number: 169
  • Mean age, years (SD): 57.0 (12.5)
  • Median age, years (range): 55 (19-90)
  • Female: 108 (64%)
  • Male: 61 (36%)
  • WFNS 1: 50 (30%)
  • WFNS 2: 40 (24%)
  • WFNS 3: 17 (10%)
  • WFNS 4: 42 (25%)
  • WFNS 5: 20 (12%)
  • Fisher grade 3: 141 (83%)
  • Fisher grade 4: 19 (11%)
  • Aneurysm location - ICA: 59 (35%)
  • Aneurysm location - ACA: 53 (31%)
  • Aneurysm location - MCA: 31 (20%)
  • Aneurysm location - posterior circulation: 16 (10%)
  • Treatment - coiling: 85 (50%)
  • Treatment - clipping: 72 (43%)
  • Treatment - no treatment: 11 (7%)

Saline Group:

  • Number: 158
  • Mean age, years (SD): 57.0 (12.5)
  • Median age, years (range): 57 (31-89)
  • Female: 100 (63%)
  • Male: 58 (37%)
  • WFNS 1: 45 (29%)
  • WFNS 2: 43 (27%)
  • WFNS 3: 14 (9%)
  • WFNS 4: 39 (25%)
  • WFNS 5: 17 (11%)
  • Fisher grade 3: 121 (77%)
  • Fisher grade 4: 20 (13%)
  • Aneurysm location - ICA: 57 (36%)
  • Aneurysm location - ACA: 43 (27%)
  • Aneurysm location - MCA: 33 (21%)
  • Aneurysm location - posterior circulation: 15 (10%)
  • Treatment - coiling: 72 (46%)
  • Treatment - clipping: 69 (44%)
  • Treatment - no treatment: 17 (11%)

Arms

FieldMagnesium SulfateControl
InterventionIV magnesium sulfate: 20 mmol MgSO4 bolus over 30 minutes, followed by continuous infusion of 80 mmol MgSO4 per day for up to 14 days after hemorrhage. Plasma magnesium measured daily. Infusion adjusted to raise plasma magnesium to ~2× baseline and ≤2.5 mmol/L. All patients received oral nimodipine 60 mg every 4 hours for 14 days if blood pressure stable. Treatment started mean 31.7±15.5 hours from ictus. Average serum Mg achieved: 1.67±0.27 mmol/L. 91% completed ≥10 days infusion. Study drug stopped in 1% for adverse events. Assessments: daily labs, WFNS grade, Fisher scale. Outcomes assessed at 3 and 6 months by blinded assessor: GOSE, mRS, Barthel Index, SF-36. Clinical vasospasm defined as new focal deficits or GCS decrease ≥2 points for ≥6 hours or new cerebral infarct not related to treatment complications. TCD monitoring when feasibleIV normal saline: equivalent volume to magnesium sulfate infusion for up to 14 days after hemorrhage. Infusion rates occasionally adjusted to maintain blinding. All patients received oral nimodipine 60 mg every 4 hours for 14 days if blood pressure stable. Treatment started mean 31.7±15.5 hours from ictus. Average serum Mg: 0.91±0.16 mmol/L. 91% completed ≥10 days infusion. Assessments: daily labs, WFNS grade, Fisher scale. Outcomes assessed at 3 and 6 months by blinded assessor: GOSE, mRS, Barthel Index, SF-36. Clinical vasospasm defined as new focal deficits or GCS decrease ≥2 points for ≥6 hours or new cerebral infarct not related to treatment complications. TCD monitoring when feasible
Duration10-14 days treatment, 6 months follow-up10-14 days treatment, 6 months follow-up

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Favorable outcome defined as Extended Glasgow Outcome Scale (GOSE) score 5-8 at 6 monthsPrimary100/158 (63%)108/169 (64%)0.61%Not significant
Clinical vasospasm during initial 2 weeksSecondary29/158 (18%)42/169 (25%)OR 1.595% CI 0.9-2.5, NS
Good outcome (mRS 0-2) at 6 monthsSecondary91/158 (58%)97/169 (57%)OR 1.095% CI 0.6-1.5, NS
Excellent outcome (mRS 0-1) at 6 monthsSecondary71/158 (45%)77/169 (46%)OR 1.095% CI 0.7-1.6, NS
Barthel Index ≥85 at 6 monthsSecondary96/158 (61%)97/169 (57%)OR 0.995% CI 0.6-1.4, NS
SF-36 physical score at 6 months (mean±SD) - 189 patients completedSecondary65.5±25.3 (90 patients)67.3±26.1 (99 patients)Mean difference 3.895% CI -5.6 to 9.2, NS
SF-36 mental score at 6 months (mean±SD)Secondary64.5±24.165.4±22.0Mean difference 3.495% CI -5.7 to 7.6, NS
Inpatient mortalitySecondary19/158 (12%)17/169 (10%)OR 0.895% CI 0.4-1.6, NS
Average serum magnesium concentration (mmol/L, mean±SD)Secondary0.91±0.161.67±0.27p<0.001
Mean TCD middle cerebral artery velocity (cm/s, mean±SD) - 210 patientsSecondary87±6182±31p=0.487
Study drug stopped for adverse eventsAdverse2 patients (1 refractory hypernatremia, 1 early discharge)2 patients (1 severe limb weakness, 1 severe hypocalcemia)
Study drug infusion <10 daysAdverseOverall: 32 patients (9%): 2 early discharge, 15 death (7 MgSO4, 8 saline), 11 refusal (7 MgSO4, 4 saline), 3 stopped for adverse events
Hypotension (persistent SBP <90 mmHg requiring inotropes)Adverse21/158 (13%)26/169 (15%)p=0.590
No difference inAdverseNote: Cardiac failure, acute renal failure, pneumonia, sepsis, pulmonary embolism, myocardial infarction, gastrointestinal bleeding
No mortality related to study drugAdverseNote: No drug-related deaths reported

Subgroup Analysis

Prespecified subgroups analyzed for primary outcome (GOSE 5-8 at 6 months). None showed benefit from magnesium sulfate. Subgroups: Age (<65 vs ≥65), WFNS grade (1-2 vs 3-5), pre-existing hypertension (absence vs presence), intracerebral hematoma (absence vs presence), intraventricular hemorrhage (absence vs presence), aneurysm location (anterior vs posterior circulation), aneurysm size (<12mm vs ≥12mm), aneurysm treatment (clipping vs coiling). All p-values for interaction were non-significant (range 0.114-0.924)

Criticisms

  • No clinical benefit demonstrated despite achieving target magnesium levels
  • Low CSF penetration of peripherally infused magnesium (only 11-21% increase) may explain lack of efficacy
  • Brain free intracellular magnesium increased only 13% - may be insufficient
  • No vasodilatory effect demonstrated (TCD velocities similar between groups)
  • Treatment window (within 48 hours) may be too late for neuroprotection from early brain injury
  • Sample size may have been insufficient based on actual control group outcome (63% favorable vs 50% expected)
  • Post hoc power analysis showed only 0.49 power for 10% improvement
  • No assessment of angiographic vasospasm (CT angiography not validated at time of protocol design)
  • Cognitive outcomes not comprehensively assessed (SF-36 completed in only 58% of survivors)
  • Academically funded study without industry support - lack of on-site monitoring
  • Last observation carried forward method for missing data (4%) could introduce bias
  • Magnesium sulfate known to lower blood pressure - potential for adverse effects not fully explored
  • Daily blood pressure data not incorporated in trial design
  • Higher target magnesium (2.5 mmol/L) limited by toxicity concerns
  • Direct intracisternal infusion not evaluated (may achieve better CSF levels)
  • Primary outcome (GOSE) may not be sensitive enough to detect subtle improvements
  • Composite cognitive function assessment would have been better outcome measure
  • Recruitment barriers: late presentation, refusal, unavailable consent
  • Only 327 of 348 planned patients enrolled (94% of target)

Funding

Research Grants Council of Hong Kong (CUHK Ref. No. CUHK4183/02 M). Academically funded, investigator-initiated trial. No industry support. Generic medications used

Based on: IMASH (Stroke, 2010)

Authors: George Kwok Chu Wong, Wai S. Poon, Matthew T.V. Chan, ..., for the IMASH Investigators

Citation: Stroke. 2010;41:921-926

Content summarized and formatted by NeuroTrials.ai.