← Back
NeuroTrials.ai
Neurology Clinical Trial Database

MASH-2

Magnesium for aneurysmal subarachnoid haemorrhage (MASH-2): a randomised placebo-controlled trial

Share Share Copied! Compare

Year of Publication: 2012

Authors: Sanne M Dorhout Mees, Ale Algra, W Peter Vandertop, ..., on behalf of the MASH-2 study group

Journal: The Lancet

Citation: Lancet 2012; 380: 44-49

PDF: https://www.thelancet.com/action/showPdf...%2812%2960724-7

Clinical Question

Does intravenous magnesium sulphate improve clinical outcome after aneurysmal subarachnoid haemorrhage by reducing the occurrence or improving the outcome of delayed cerebral ischaemia?

Bottom Line

Intravenous magnesium sulphate 64 mmol/day does not improve clinical outcome after aneurysmal subarachnoid haemorrhage. Routine administration of magnesium cannot be recommended based on this adequately powered phase 3 trial and updated meta-analysis.

Major Points

  • Phase 3 randomized, placebo-controlled trial of 1204 patients across 8 centers in Europe and South America
  • Magnesium sulphate 64 mmol/day given continuously for up to 20 days showed no benefit
  • Poor outcome occurred in 26.2% of magnesium group vs 25.3% of placebo (RR 1.03, 95% CI 0.85-1.25)
  • No difference in excellent outcome or distribution of mRS scores between groups
  • Study was adequately powered (1200 patients planned, 1204 enrolled) unlike previous trials
  • Subgroup analyses showed no benefit in any patient subgroup
  • Updated meta-analysis of 7 trials (2047 patients) confirmed no benefit (RR 0.96, 95% CI 0.84-1.10)
  • Over 99% follow-up rate (only 3 patients lost to follow-up)
  • Treatment started median 33-41 hours after symptom onset
  • Previous MASH-1 pilot study had suggested benefit (RR 0.77) but was underpowered
  • Four treatment discontinuations in magnesium group: 1 hypocalcaemia, 2 hypermagnesaemia, 1 suspected symptomatic hypermagnesaemia

Design

Study Type: Phase 3 randomized, placebo-controlled, double-blind trial

Randomization: 1

Blinding: Double-blind. Patients, treating physicians, investigators assessing outcomes, and data analysts were masked to allocation. Computer-generated randomization in blocks of four, stratified by center. Identical sequentially numbered boxes produced by pharmacy

Enrollment Period: April 2004 to September 2011

Follow-up Duration: 3 months

Centers: 8

Countries: The Netherlands, Scotland (UK), Chile

Sample Size: 1204

Analysis: Intention-to-treat analysis. Risk ratios with 95% CIs for primary outcome. Mann-Whitney U test for comparison of mRS score distributions. Planned sensitivity analyses assigning lost-to-follow-up patients to good/poor outcomes. Planned subgroup analyses by age, sex, clinical condition, aneurysm treatment method, and center practice of treating hypomagnesaemia. Two interim analyses performed by data monitoring committee (after 350 and 750 patients). Updated meta-analysis using Cochrane Review Manager version 5.1. SPSS version 15.0 for other analyses

Inclusion Criteria

  • Age 18 years or older
  • Aneurysmal pattern of subarachnoid haemorrhage on brain imaging
  • Admitted to hospital within 4 days of haemorrhage
  • Extravasated blood in basal cisterns on brain CT, or xanthochromia of CSF if CT negative
  • For patients with normal CT and xanthochromia of CSF, proof of aneurysm required
  • Perimesencephalic pattern on CT only included if posterior circulation aneurysm present
  • Written and oral informed consent obtained

Exclusion Criteria

  • Age younger than 18 years
  • Renal failure (serum creatinine >150 µmol/L)
  • Bodyweight less than 50 kg
  • Imminent death

Baseline Characteristics

CharacteristicControlActive
Number randomized597606
Mean age (SD), years57 (12)57 (13)
Female416 (70%)422 (70%)
WFNS grade >4130 (22%)144 (24%)
Aneurysm treatment - Coiling324 (54%)344 (57%)
Aneurysm treatment - Clipping195 (33%)197 (33%)
Aneurysm treatment - None76 (13%)65 (11%)
Aneurysm treatment - Unknown2 (<1%)0 (0%)
Median time to study medication from onset (IQR), hours41 (23-59)33 (21-60)

Arms

FieldMagnesium sulphateControl
Intervention64 mmol magnesium sulphate per day reconstituted in 0.9% saline, administered continuously via intravenous infusion. Treatment started as soon as possible after consent and continued for 20 days after haemorrhage onset, or until hospital discharge or death if sooner. Renal function checked at least once every 2 days. Magnesium concentration monitoring not mandatory. All patients received standard care including oral nimodipine 360 mg/day, bed rest until aneurysm occlusion, early aneurysm occlusion, and normovolaemia. Treatment of hypomagnesaemia with IV magnesium was discouraged but permittedIdentical appearing placebo (saline) administered continuously via intravenous infusion for 20 days after haemorrhage onset, or until hospital discharge or death if sooner. Renal function checked at least once every 2 days. All patients received standard care including oral nimodipine 360 mg/day, bed rest until aneurysm occlusion, early aneurysm occlusion, and normovolaemia. Treatment of hypomagnesaemia with IV magnesium was discouraged but permitted
DurationUp to 20 days treatment, 3 months follow-upUp to 20 days treatment, 3 months follow-up

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Poor outcome defined as modified Rankin Scale score of 4-5 or death at 3 months after subarachnoid haemorrhagePrimary151/596 (25.3%)158/604 (26.2%)Not significant
No symptoms (mRS 0) at 3 monthsSecondary46/596 (7.7%)46/604 (7.6%)RR 0.9995% CI 0.67-1.46
Distribution of all mRS scoresSecondarymRS 0: 46, mRS 1: 210, mRS 2: 142, mRS 3: 47, mRS 4: 44, mRS 5: 22, Death: 85mRS 0: 46, mRS 1: 222, mRS 2: 124, mRS 3: 54, mRS 4: 42, mRS 5: 25, Death: 91Mann-Whitney U testp=0.95
Poor outcome in womenSecondary116/415 (28.0%)122/421 (29.0%)RR 1.0495% CI 0.84-1.29
Poor outcome in menSecondary35/181 (19.3%)36/183 (19.7%)RR 1.0295% CI 0.67-1.54
Poor outcome in patients ≤55 yearsSecondary44/277 (15.9%)47/282 (16.7%)RR 1.0595% CI 0.72-1.53
Poor outcome in patients >55 yearsSecondary107/319 (33.5%)111/322 (34.5%)RR 1.0395% CI 0.83-1.28
Poor outcome in WFNS grade <4Secondary82/465 (17.6%)79/460 (17.2%)RR 0.9795% CI 0.74-1.29
Poor outcome in WFNS grade 4-5Secondary68/130 (52.3%)79/144 (54.9%)RR 1.0595% CI 0.84-1.31
Poor outcome in patients with no aneurysm treatmentSecondary35/76 (46.1%)33/65 (50.8%)RR 1.1095% CI 0.78-1.55
Poor outcome in patients with coilingSecondary66/323 (20.4%)78/342 (22.8%)RR 1.1295% CI 0.84-1.49
Poor outcome in patients with clippingSecondary49/195 (25.1%)47/197 (23.9%)RR 0.9595% CI 0.67-1.34
Treatment stopped for asymptomatic hypocalcaemiaAdverse01
Treatment stopped for asymptomatic hypermagnesaemiaAdverse02
Treatment stopped for suspected symptomatic hypermagnesaemiaAdverse01 (also had renal insufficiency - protocol violation)
Lost to follow-upAdverse12

Subgroup Analysis

Extensive prespecified subgroup analyses showed consistent lack of benefit across all subgroups: women vs men, age ≤55 vs >55 years, good (WFNS <4) vs poor (WFNS 4-5) clinical condition at admission, different aneurysm treatment methods (coiling, clipping, none), and centers that treat hypomagnesaemia with IV magnesium vs those that do not. All subgroup risk ratios ranged from 0.89 to 1.12 with confidence intervals crossing 1.0

Criticisms

  • Study collected only key baseline and outcome data without detailed quality of life assessments
  • Did not include delayed cerebral ischaemia as secondary outcome (only surrogate measure)
  • Did not perform close on-site monitoring or collect complete adherence data
  • Magnesium concentration not routinely checked, though unmasking by checking levels was possible
  • Modified Rankin Scale may miss cognitive symptoms common after SAH
  • Outcome assessed at 3 months; possible later benefits might have been missed
  • Median time to treatment was 33-41 hours, which may be too long to prevent acute ischemic injury cascade
  • Intravenous magnesium may not cross blood-brain barrier well (only 11-21% increase in CSF with doubling serum levels)
  • Factorial design with nimodipine given to all patients may have obscured magnesium effects
  • One randomization code lost (excluded from analysis)
  • Protocol violation in one patient (treatment not stopped for renal insufficiency with creatinine >150 µmol/L)
  • Some centers treated hypomagnesaemia with IV magnesium, potentially confounding results
  • Did not assess mechanism of action or biochemical markers of neuroprotection

Funding

Netherlands Heart Foundation (grant 2005BO16) and UK Medical Research Council (clinician scientist fellowship G108/613). Van Leersumfonds financially supported part of the production of study medication. Study supported by research nurses at Edinburgh Wellcome Trust Clinical Research Facility and UK Stroke Research Network

Based on: MASH-2 (The Lancet, 2012)

Authors: Sanne M Dorhout Mees, Ale Algra, W Peter Vandertop, ..., on behalf of the MASH-2 study group

Citation: Lancet 2012; 380: 44-49

Content summarized and formatted by NeuroTrials.ai.