CHARM Trial: Intravenous Glibenclamide for Cerebral Edema After Large Hemispheric Stroke
In patients with large hemispheric infarction (ASPECTS 1-5 or DWI core 80-300 mL), does IV glibenclamide 8.6 mg over 72 hours improve 90-day functional outcomes compared with placebo?
Sheth et al. (Lancet Neurol 2024) · 535 patients
Population
Included
- Age 18-80 (primary efficacy analysis age 18-70)
- Large hemispheric infarction: ASPECTS 1-5 or DWI core 80-300 mL
- Study drug initiated within 10 hours of stroke onset
- Patient or surrogate able to provide consent
- Aged 18 to 85 years
- Clinical diagnosis of acute ischaemic stroke in at least the middle cerebral artery territory
- Within 10 h of stroke onset (in patients with wake-up stroke, stroke onset time was defined as the midpoint between sleep onset and waking)
- US National Institutes of Health Stroke Scale (NIHSS) score of 10 or higher
- Absence of significant premorbid disability as assessed by the site investigator
- A core lesion of 80 to 300 mL, as assessed by CT perfusion (RAPID, IschemaView; or Golden, CO, USA) or diffusion-weighted imaging
- If CT perfusion or MRI was not available to determine core lesion volume, non-contrast CT was used with an Alberta Stroke Program Early CT Score (ASPECTS) of 1 to 5 involving at least two cortical regions
- For patients who underwent thrombectomy, a post-thrombectomy MRI was required for ascertainment of the qualifying infarct volume
Excluded
- Core volume below 80 mL or above 300 mL
- Severe hepatic or renal impairment
- Known sulfonylurea hypersensitivity
- Planned early hemicraniectomy precluding 72-hour drug administration
- Pregnancy
- Core lesion volume below 80 mL or above 300 mL on qualifying imaging
- NIHSS score below 10
- Significant premorbid disability
- Stroke onset more than 10 h before planned study-drug initiation
- Known hypersensitivity to glibenclamide or sulfonylurea agents
- Severe hepatic or renal impairment precluding study-drug administration
- Pregnancy
- Any condition precluding the planned 72-h study-drug infusion
Source: Sheth KN et al., Lancet Neurol 2024;23(12):1205–1213; NCT02864953· Retrieved 2026-06-09
Stopped early: findings are inconclusive
CHARM was halted before planned enrollment (535 of approximately 750 patients) due to COVID-19 operational disruptions. The trial was neither futile nor stopped for safety -- it was underpowered. Results should be interpreted with this context: the confidence interval (0.80-1.71) excludes neither meaningful benefit nor meaningful harm.
Primary Outcome — mRS Ordinal Shift at 90 Days (age 18-70)
535 patients; IV glibenclamide vs placebo within 10 hours of onset
mRS Ordinal Shift at 90 Days
cOR 1.17
95% CI 0.8 to 1.71 · P = 0.42
Mortality: numerically higher with glibenclamide · Hypoglycemia: 6% vs 2% · Stopped early for COVID-19
Study Arms
- Agent
- Glibenclamide (intravenous formulation, BIIB093)
- Dose
- Total dose 8.6 mg
- Route
- IV
- Frequency
- Bolus followed by a continuous infusion
- Duration
- 72 h
- Co-interventions
- Standard-of-care reperfusion interventions (rtPA and endovascular thrombectomy) permitted in accordance with local stroke guidelines; additional measures for large stroke and swelling (for example, mannitol) as clinically indicated
Study drug started within 10 h of stroke onset. The trial was stopped early by the sponsor for strategic and operational reasons (slow enrolment because of COVID-19), before unblinding or knowledge of trial results, at approximately 71% of planned enrolment, leaving it underpowered. The result was neutral: the primary mRS ordinal shift at 90 days was not improved (common OR 1.17, 95% CI 0.80 to 1.71, p=0.42) in the modified intention-to-treat population aged 18 to 70 years.
- Agent
- Matching placebo infusion
- Route
- IV
- Frequency
- Bolus followed by a continuous infusion
- Duration
- 72 h
- Co-interventions
- Standard-of-care reperfusion interventions and supportive measures identical to the active arm; drug vials, preparation bags, and tubing were identical for both groups
Hypoglycaemia occurred in 6% (15 of 259) of glibenclamide patients versus 2% (four of 259) of placebo patients. 90-day mortality was numerically higher with glibenclamide (32% vs 29%; hazard ratio 1.20, 95% CI 0.85 to 1.70, p=0.30).
Trial Design
Type
- Phase 3 double-blind placebo-controlled randomized trial
- Large hemispheric infarction defined by ASPECTS 1-5 or core 80-300 mL
- Study drug started within 10 hours of stroke onset
- Primary efficacy analysis focused on patients aged 18-70 years
Timeline
143 stroke centers in 21 countries; stopped early for operational reasons
N
535
Enrollment
535 patients at 143 stroke centers across 21 countries. Phase 3 double-blind placebo-controlled trial. IV glibenclamide 8.6 mg over 72 hours started within 10 hours of onset. Stopped early for COVID-19 operational disruptions. Published Lancet Neurol 2024.
ClinicalTrials.gov
NCT02864953Bedside Pearl
CHARM stopped early for COVID and was underpowered; results are inconclusive. Glibenclamide did not significantly improve disability and caused 3-fold higher hypoglycemia. Do not use IV glibenclamide for malignant edema outside a clinical trial. Hemicraniectomy (DESTINY II, HAMLET, DECIMAL) remains the only intervention with survival evidence in eligible malignant MCA infarction patients.
See also