ESCAPE
(2015)Objective
To evaluate whether rapid endovascular treatment in addition to standard care improves functional outcomes in patients with acute ischemic stroke who have a small infarct core, proximal anterior circulation occlusion, and moderate-to-good collateral circulation on CT/CTA imaging.
Study Summary
• Primary outcome favored intervention with common OR 2.6 (95% CI 1.7-3.8, P<0.001) for shift toward better mRS scores
• 90-day mortality was significantly reduced with intervention (10.4% vs. 19.0%, P=0.04)
• Median time from study CT to first reperfusion was 84 minutes; successful reperfusion (TICI 2b/3) achieved in 72.4%
• Symptomatic ICH rates were similar between groups (3.6% vs. 2.7%, P=0.75)
Intervention
Rapid endovascular treatment using available thrombectomy devices (retrievable stents recommended, 86.1% used; 77% Solitaire stent) plus standard care including IV alteplase when eligible, versus standard care alone.
Inclusion Criteria
Adults with acute ischemic stroke; proximal intracranial anterior circulation occlusion (ICA, M1, or M2 MCA) on CTA; small infarct core (ASPECTS 6-10) on non-contrast CT; moderate-to-good collateral circulation (>=50% MCA pial filling on CTA); pre-stroke functional independence (Barthel Index >=90); enrollment within 12 hours of symptom onset.
Study Design
Arms: 1) Intervention arm: endovascular treatment plus guideline-based standard care (including IV alteplase if eligible). 2) Control arm: guideline-based standard care alone (including IV alteplase if eligible).
Patients per Arm: Intervention: 165 patients; Control: 150 patients
Outcome
• mRS 0-2 at 90 days: 53.0% vs. 29.3% (rate ratio 1.8, 95% CI 1.4-2.4)
• 90-day mortality: 10.4% vs. 19.0% (rate ratio 0.5, 95% CI 0.3-1.0, P=0.04)
• TICI 2b/3 reperfusion: 72.4% in intervention group
• Barthel Index 95-100 at 90 days: 57.7% vs. 33.6%
• Symptomatic ICH: 3.6% vs. 2.7% (P=0.75)
Bottom Line
Rapid endovascular treatment significantly improved functional independence and reduced mortality in carefully selected stroke patients with proximal vessel occlusions.
Major Points
- Stopped early for efficacy after enrolling 316 of planned 500 patients at 22 centers across 5 countries.
- Key innovation: imaging selection using ASPECTS ≥6 (small core) AND moderate-to-good collaterals (≥50% MCA pial filling on multiphase CTA) — first trial to require collateral assessment.
- Eligible vessels: proximal anterior circulation — intracranial ICA, M1 MCA, or M2 MCA on CTA. ICA+M1 occlusions: ~28% ICA with M1, ~68% M1 or M2 alone.
- Functional independence (mRS 0–2) at 90 days: 53.0% vs 29.3% (absolute difference 23.7%, NNT ≈ 4, P<0.001).
- Primary outcome (ordinal mRS shift): common OR 2.6 (95% CI 1.7–3.8, P<0.001). Median mRS: 2 (intervention) vs 4 (control).
- Mortality significantly reduced: 10.4% vs 19.0% (rate ratio 0.5, P=0.04) — ESCAPE was the only early-window thrombectomy trial to show a mortality benefit.
- Fastest workflow of the 2015 thrombectomy trials: median CT-to-reperfusion 84 minutes (target ≤90 min). Emphasized parallel workflow, rapid transfer protocols.
- Successful reperfusion (mTICI 2b–3): 72.4%. sICH: 3.6% vs 2.7% (P=0.75) — no increase.
- IV alteplase used in 72.7% of intervention and 78.7% of control patients.
- No upper age limit — included patients ≥80 years, unlike many other trials.
Study Design
- Study Type
- Multicenter, prospective, randomized, open-label, controlled trial with blinded outcome evaluation (PROBE design)
- Randomization
- Yes
- Blinding
- Outcome assessors and imaging interpreters were blinded to treatment assignment
- Sample Size
- 316
- Follow-up
- 90 days
- Centers
- 22
- Countries
- Canada, United States, South Korea, Ireland, United Kingdom
Primary Outcome
Definition: Score on modified Rankin scale at 90 days (0=no symptoms to 6=death)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| Median score 4 | Median score 2 | - (1.7 to 3.8) | <0.001 |
Limitations & Criticisms
- Stopped early after 316 of planned 500 patients — while ethically necessary given overwhelming benefit, early stopping may overestimate treatment effect (the 'stopping bias' of interim analyses).
- No screening logs maintained — cannot estimate what proportion of stroke patients were eligible, making it impossible to determine the real-world applicability of ESCAPE's strict imaging criteria.
- Limited to selected endovascular centers with efficient workflows — the median 84-minute CT-to-reperfusion time is exceptionally fast and may not be achievable at most centers, limiting generalizability.
- Imaging protocol violations in 8.3% of participants — some patients were randomized without complete adherence to the ASPECTS + collateral criteria.
- Multiphase CTA for collateral assessment was a novel, non-validated technique at the time — while subsequently adopted widely, the learning curve and inter-reader variability were not fully characterized.
- PROBE design (open-label treatment, blinded outcome assessment) — treating physicians knew allocation, potentially introducing placebo/nocebo effects in post-procedural care decisions.
- Small number treated beyond 6 hours — while the 12-hour window was allowed, most patients were treated within 6 hours. The extended window benefit was later definitively established by DAWN and DEFUSE 3 using different imaging selection.
- Control group received IV alteplase in 78.7% — higher than some other thrombectomy trials. This may have attenuated the treatment difference by providing some recanalization in controls.
- Predominantly stent retriever era — most procedures used Solitaire FR. Modern aspiration techniques, combined approaches, and newer devices may achieve different outcomes.
Citation
N Engl J Med 2015;372:1019-30