Illustration: a mesh of fibrin strands trapping red blood cells and platelets in a small clot inside a tiny vessel.
← EMA Cited
EMA Cited · Neurology · Reperfusion

Acute ischemic stroke

Updated Oct 16, 2026Adults · ED management23 sources · 16 checked in PubMed3 figuresAHA/ASA 2026

Bottom line

  • Image fast. Get a non-contrast CT or MRI, ideally within 25 minutes of arrival. In suspected large-vessel occlusion, don't hold CT angiography for a creatinine result.1,2
  • Thrombolyse disabling deficits within 4.5 hours. Use tenecteplase 0.25 mg/kg (max 25 mg) or alteplase 0.9 mg/kg (max 90 mg). The guideline endorses either.1,2,3,4
  • Minor and non-disabling? Use dual antiplatelets, not a thrombolytic.1,5 In ARAMIS, dual antiplatelet therapy was noninferior to alteplase (excellent outcome 93.8% vs 91.4%).6
  • Thrombectomy for large-vessel occlusion up to 24 hours in selected patients, now including large cores and basilar occlusion. If a patient is eligible for both, give the thrombolytic, then go straight to thrombectomy without delay.7,2,5
  • After reperfusion, don't push systolic BP below 140. Aggressive lowering brings no benefit after thrombolysis and causes harm after thrombectomy. Don't aim for tight glucose control either.2

Recognition and triage

  • Use a validated prehospital stroke scale and pre-notify the receiving hospital (both Class 1).1,2
  • Mobile stroke units (ambulances with a CT scanner) are recommended where available for thrombolysis-eligible patients (Class 1, level A).1,2 In BEST-MSU, median onset-to-treatment was 72 vs 108 minutes, and 55.0% vs 44.4% of eligible patients reached a modified Rankin score of 0–1.8
  • Where the nearest hospital can't do thrombectomy, the guideline focuses on shortening door-in-door-out times at the transferring hospital.7
Bypass or nearest hospital?

The AHA release says EMS should take suspected large-vessel occlusion to the nearest thrombectomy-capable center.7 But where local stroke centers give thrombolysis well and transfer efficiently, the guideline finds no 3-month benefit from bypassing them for a more distant thrombectomy center (Class 3: no benefit).2 In RACECAT (n = 1,401, rural Catalonia), direct transport did not change 90-day disability. Patients taken directly got thrombolysis less often (47.5% vs 60.4%) and thrombectomy more often (48.8% vs 39.4%).9 Follow your regional protocol.

First 25 minutes

  1. Establish the time last known well. It sets every treatment window.2,5
  2. Judge whether the deficit is disabling, not just the NIHSS number. That decides thrombolysis.5
  3. Non-contrast CT or MRI, ideally within 25 minutes of arrival (Class 1).2,7
  4. CT angiography (with or without perfusion) if large-vessel occlusion is possible. Don't wait for creatinine (Class 1).2 Without perfusion imaging, use ASPECTS on CT to select for thrombectomy.7
  5. Thrombolyse without waiting for extra imaging that won't change the decision.2,5
  6. Give oxygen only if hypoxic, and treat fever.2
  7. Without recent anticoagulant or heparin use, don't wait for coagulation results to start the thrombolytic.3,4
FIG 1Treatment windows. Time from last known well, drawn for this page from the 2026 guideline as reported in its published summaries. Lighter bars need imaging selection: perfusion or MRI mismatch for late thrombolysis, and clinical-imaging criteria for late thrombectomy.7,2,5
Axial non-contrast CT of the brain with a label pointing to a hyperdense right middle cerebral artery
FIG 2Dense MCA sign. The right middle cerebral artery looks brighter than the left on non-contrast CT, a sign of thrombus and an early change of MCA infarction.Cerevisae · Wikimedia Commons · CC BY-SA 4.0
Axial non-contrast CT of the brain showing a large hypodense infarct in the right middle cerebral artery territory with mass effect on the right lateral ventricle
FIG 3Established MCA infarct. A large, sharply bordered low-density area fills the right MCA territory, compressing the right lateral ventricle.Lucien Monfils · Wikimedia Commons · CC BY-SA 4.0

Thrombolysis

DrugAdult doseNotes
Tenecteplase0.25 mg/kg IV bolus over 5 s, max 25 mg. By weight: < 60 kg 15 mg · 60–69 kg 17.5 mg · 70–79 kg 20 mg · 80–89 kg 22.5 mg · ≥ 90 kg 25 mgSingle bolus. Half the STEMI dose: don't mix up the tables.2,3
Alteplase0.9 mg/kg IV, max 90 mg: 10% as a bolus over 1 min, the rest over 60 minDoses above 0.9 mg/kg may raise intracranial hemorrhage.4

Unreviewed doses. Verify against the current label and your institution's protocol before giving anything.

  • Window: within 4.5 hours of onset or last known well (Class 1).2 The alteplase label still says within 3 hours. The 3–4.5 hour window rests on ECASS III.4,10
  • Late or wake-up stroke: thrombolysis may be considered at 4.5–9 hours or with unknown onset when perfusion imaging or a DWI-FLAIR mismatch shows salvageable tissue.5 In WAKE-UP (n = 503), MRI-selected alteplase improved favorable outcome (53.3% vs 41.8%).11
  • Monitor and control BP during and after the thrombolytic.3,4 In the NINDS trials, BP was kept at 185/110 mm Hg or lower for 24 hours.4
  • Contraindications on the tenecteplase label (stroke): active intracranial hemorrhage, active internal bleeding, intracranial or spinal surgery or trauma within 2 months, known bleeding diathesis, severe uncontrolled hypertension, and intracranial neoplasm, AVM or aneurysm.3 The 2026 guideline also revised its approach to contraindications.1

What the trials show

  • NINDS (0–3 h, alteplase): at least 30% more patients had minimal or no disability at 3 months. Symptomatic hemorrhage was 6.4% vs 0.6%, and mortality 17% vs 21% (not significant).12
  • ECASS III (3–4.5 h): favorable outcome 52.4% vs 45.2%. Symptomatic hemorrhage 2.4% vs 0.2%.10
  • AcT (tenecteplase vs alteplase, n = 1,600): modified Rankin 0–1 in 36.9% vs 34.8%, meeting noninferiority. Symptomatic hemorrhage was 3.4% vs 3.2%.13 Door-to-needle times were 36 vs 37 minutes.14
Sources disagree

Guidelines treat thrombolysis as standard.1,2 Some emergency physicians remain skeptical. The SGEM review of AcT argues that the trial assumed alteplase works, that the NINDS and ECASS III results are contested, and that open-label design and telephone outcome scoring could favor the newer drug.15 REBEL EM also flags the open-label design, and notes that the evidence for thrombolytic efficacy itself remains debated.14

Minor, non-disabling stroke

  • Thrombolysis is not recommended. Dual antiplatelet therapy is preferred.1,5
  • PRISMS (NIHSS 0–5, not clearly disabling): alteplase did not beat aspirin (favorable outcome 78.2% vs 81.5%), and caused symptomatic hemorrhage in 3.2% vs 0%. The trial stopped early.16
  • ARAMIS (n = 760): dual antiplatelet therapy was noninferior to alteplase (93.8% vs 91.4% excellent outcome). Symptomatic hemorrhage was 0.3% vs 0.9%.6
  • POINT (minor stroke or high-risk TIA): clopidogrel plus aspirin lowered major ischemic events at 90 days (5.0% vs 6.5%), but raised major hemorrhage (0.9% vs 0.4%). Most events came in the first week.17

Thrombectomy

  • Anterior proximal large-vessel occlusion: thrombectomy within 6 hours, and in selected patients at 6–24 hours on imaging and clinical criteria (Class 1, level A).2
  • Large cores: now eligible in selected patients, including those with low ASPECTS.7,5
  • Basilar occlusion: strong recommendation for thrombectomy up to 24 hours.5
  • Older or mildly disabled patients: the 80-year age limit in the early window is gone, and thrombectomy is reasonable in selected patients with a prestroke modified Rankin of 2.5
  • Medium or distal vessel occlusions: not routinely recommended outside trials.7

What the trials show

  • HERMES (5 trials, n = 1,287): thrombectomy reduced 90-day disability (common OR 2.49). The number needed to treat for one patient to improve by at least one Rankin level was 2.6. Benefit held at age 80 and older.18
  • DAWN (6–24 h, clinical-core mismatch): functional independence 49% vs 13%.19
  • SELECT2 (large core, ASPECTS 3–5 or core ≥ 50 mL): functional independence 20% vs 7%, with more vascular complications.20
  • ATTENTION (basilar, within 12 h): good outcome 46% vs 23%, and mortality 37% vs 55%.21

BP, glucose and early care

Issue2026 guidelineClass
SBP < 140 after thrombolysisNo benefit from intensive lowering3: No benefit
SBP < 140 after successful thrombectomyHarmful3: Harm
Intensive glucose lowering (80–130 mg/dL)No benefit3: No benefit
FeverKeep normothermic; treat hyperthermia1
OxygenOnly if hypoxic1
SwallowingBedside dysphagia screen before anything by mouth1
DVT preventionIntermittent pneumatic compression if immobile1
AntiplateletEarly antiplatelet therapy1
Where to admitOrganized stroke unit1

Classes as reported in a published summary of the guideline.2 Intensive glucose control means 80–130 mg/dL (4.4–7.2 mmol/L).5,22

  • SHINE (n = 1,151): intensive IV insulin to 80–130 mg/dL did not improve 90-day outcome (20.5% vs 21.6% favorable), and caused severe hypoglycemia in 2.6%.22
  • BP after thrombectomy: across 4 trials (n = 1,571), intensive systolic targets lowered functional independence (RR 0.81).23

Pitfalls

  • Deciding on thrombolysis by NIHSS number alone. Ask whether the deficit is disabling.5
  • Thrombolysing a minor, non-disabling deficit.1,16
  • Using the STEMI tenecteplase table (up to 50 mg) for stroke (max 25 mg).3
  • Waiting for creatinine before CT angiography.2
  • Holding thrombectomy to see whether the thrombolytic works.7
  • Pushing systolic BP below 140 after reperfusion.2,23

How strong is the evidence?

Thrombectomy for anterior large-vessel occlusion has strong, consistent randomized evidence.18,2 Thrombolysis within 4.5 hours is a Class 1 recommendation, but its trials remain debated in emergency medicine.2,15 Large-core and basilar trials are newer, and SELECT2 stopped early for efficacy.20,21 The guideline's full text was not accessible for this page, so recommendations come from the AHA release and two published summaries.7,2,5

On-shift teaching

Questions to ask a learner at the bedside. Tap a question to see the answer. Every answer uses facts cited above.

Q1What is the stroke dose of tenecteplase, and how does it differ from STEMI?
0.25 mg/kg, max 25 mg, as a single bolus.2,3 The STEMI table goes up to 50 mg, so don't confuse the two.3
Q2A patient has an NIHSS of 2 from mild hand clumsiness that doesn't affect function. Thrombolyse?
No. For minor non-disabling stroke the guideline recommends dual antiplatelet therapy instead.1,5 In ARAMIS, dual antiplatelets were noninferior to alteplase (93.8% vs 91.4% excellent outcome).6
Q3Why not wait for a creatinine before the CT angiogram?
The guideline says vascular imaging for suspected large-vessel occlusion should not be delayed for creatinine (Class 1).2 Thrombectomy is highly effective: in HERMES, the number needed to treat for one patient to improve by at least one Rankin level was 2.6.18
Q4A patient woke up with a stroke. Is thrombolysis off the table?
Not necessarily. With a DWI-FLAIR or perfusion mismatch, thrombolysis may be considered.5 In WAKE-UP, MRI-selected alteplase raised favorable outcome from 41.8% to 53.3%.11
Q5After a successful thrombectomy, should you target systolic BP under 140?
No. The guideline rates intensive lowering after successful thrombectomy as harmful.2 Pooled trials showed less functional independence with intensive targets (RR 0.81).23
Q6Why not run an insulin drip to keep glucose 80–130 mg/dL?
In SHINE, intensive control did not improve outcomes and caused severe hypoglycemia in 2.6%.22 The guideline rates intensive lowering as no benefit.2

References

Evidence base

What this page is built on

23 sources, cited 95 times. Dots are sources, filled by how far each was read; stronger color means a larger share of the citations. Tap a tier to see its sources.

Read in full · 7Abstract read · 15Record only · 1Link checked · 0

Checked by the writer, not a physician. The dots show how each source was read, not that anyone has confirmed the claims.

How this page was made
  • Written from scratch. No source text was copied; facts are restated and cited.
  • The 2026 guideline's full text is blocked to automated access. Its recommendations, and their class and level where stated, come from the AHA's official news release and two published summaries (one in PMC with class and level for each point). Each is cited with the guideline. Where a summary gave no class, none is stated.
  • Trial results were checked against their PubMed abstracts. Tenecteplase and alteplase doses and contraindications were confirmed on the FDA labels. No antiplatelet doses are given, because none could be confirmed from the guideline text.
  • Adults only. The 2026 guideline adds the first pediatric recommendations, which differ; see the guideline.
  • Figure 1 was drawn for this page from the guideline's time windows. Figures 2 and 3 are openly licensed images from Wikimedia Commons, credited under each one, and keep their own licenses.
  • Not for clinical use.
Not for clinical use

AI-generated draft. Not reviewed by a physician. This page is for education only and may contain errors, including in doses. Do not use it to diagnose, treat or make decisions about any patient. Rely on your own clinical judgment, your institution's protocols and current drug labeling. Reading this page does not create a physician-patient relationship. Full terms