PDF: ESCAPE-NA1
(2020)Objective
In patients with acute ischaemic stroke due to large vessel occlusion undergoing endovascular thrombectomy, does a single IV dose of nerinetide (2.6 mg/kg) improve 90-day functional outcomes compared with placebo.
Study Summary
• In the pre-specified alteplase-stratified subgroup, nerinetide improved mRS 0–2 in the no-alteplase stratum (59.3% vs 49.8%; adjusted RR 1.18, 95% CI 1.01–1.38; p-interaction=0.033).
• Safety was similar: any serious AE 33.1% vs 35.7% (RR 0.92, 0.79–1.09); symptomatic ICH 3.5% vs 4.3% (RR 0.80, 0.44–1.45); 90-day mortality 12.2% vs 14.4%.
Intervention
Single IV dose of nerinetide 2.6 mg/kg (max 270 mg) over 10 min vs saline placebo, given as adjunct to endovascular thrombectomy (with or without IV alteplase per usual care).
Inclusion Criteria
Adults ≥18 y with disabling acute ischaemic stroke (NIHSS >5) due to ICA or M1 occlusion on multiphase CTA, ASPECTS 5–10, moderate-to-good CT-angiographic collaterals, functionally independent at baseline (Barthel >90), within 12 h of last-seen-well, selected for endovascular thrombectomy.
Study Design
Arms: Nerinetide 2.6 mg/kg IV single dose (n=549) vs saline placebo IV (n=556)
Patients per Arm: 549 nerinetide vs 556 placebo (N=1105 randomised; ITT analysis)
Outcome
• Secondary: NIHSS 0–2 58.3% vs 57.6% (RR 1.01, 0.92–1.11); mBI 95–100 62.1% vs 60.3% (RR 1.03, 0.94–1.12); mRS 0–1 40.4% vs 40.6% (RR 0.98, 0.85–1.12); 90-d mortality 12.2% vs 14.4% (RR 0.84, 0.63–1.13).
• Alteplase-stratum interaction (exploratory): no-alteplase stratum — mRS 0–2 59.3% vs 49.8% (adj RR 1.18, 1.01–1.38), mortality 12.8% vs 20.3% (adj RR 0.66, 0.44–0.99); alteplase stratum — mRS 0–2 62.7% vs 65.7% (RR 0.97, 0.87–1.08); p-interaction=0.033.
• eTICI 2b–3 reperfusion 87.2% vs 86.3%; symptomatic ICH 3.5% vs 4.3% (RR 0.80, 0.44–1.45); any SAE 33.1% vs 35.7% (RR 0.92, 0.79–1.09).
Bottom Line
In LVO stroke selected for endovascular thrombectomy, a single IV dose of nerinetide (2.6 mg/kg) did not improve the primary outcome of mRS 0–2 at 90 days (61.4% vs 59.2%; adjusted RR 1.04, 95% CI 0.96–1.14; p=0.35). An exploratory subgroup analysis suggested benefit in patients who did not receive alteplase (p-interaction=0.033), hypothesised to reflect plasmin-mediated cleavage of nerinetide when co-administered with alteplase.
Major Points
- Multicentre, double-blind, placebo-controlled, single-dose phase 3 trial at 48 acute care hospitals in 8 countries (Canada, USA, Germany, Australia, South Korea, Sweden, Ireland, UK).
- Enrolled 1105 patients with LVO ischaemic stroke (ICA or M1, ASPECTS 5–10, moderate/good collaterals) within 12 h and selected for EVT; 549 nerinetide and 556 placebo; randomisation stratified by IV alteplase use and declared first thrombectomy device.
- Primary outcome mRS 0–2 at 90 d: 337/549 (61.4%) nerinetide vs 329/556 (59.2%) placebo; adjusted RR 1.04 (95% CI 0.96–1.14); p=0.35 — trial NEGATIVE for the primary endpoint.
- Secondary efficacy outcomes were also neutral overall: NIHSS 0–2 58.3% vs 57.6% (RR 1.01); mBI 95–100 62.1% vs 60.3% (RR 1.03); mRS 0–1 40.4% vs 40.6% (RR 0.98); 90-d mortality 12.2% vs 14.4% (RR 0.84, 0.63–1.13).
- Prespecified subgroup by alteplase (exploratory after negative primary): in the no-alteplase stratum (n=446), mRS 0–2 was 130/219 (59.3%) nerinetide vs 113/227 (49.8%) placebo (adjusted RR 1.18, 1.01–1.38); mortality 12.8% vs 20.3% (adjusted RR 0.66, 0.44–0.99). In the alteplase stratum (n=659), no benefit (mRS 0–2 62.7% vs 65.7%; RR 0.97, 0.87–1.08). p-interaction 0.033 for mRS 0–2 and 0.040 for infarct volume.
- Pharmacokinetic substudy in 22 patients showed lower peak plasma nerinetide concentrations when co-administered with alteplase, consistent with plasmin-mediated cleavage — a biologically plausible explanation for the effect modification.
- Reperfusion quality was high and similar in both groups (eTICI 2b–3 in 87.2% nerinetide vs 86.3% placebo); median study-drug-to-reperfusion time was ~22 min.
- Safety was similar: any serious adverse event 181/547 (33.1%) nerinetide vs 198/554 (35.7%) placebo, RR 0.92 (0.79–1.09). Symptomatic ICH 19/547 (3.5%) vs 24/554 (4.3%); RR 0.80 (0.44–1.45). Numerically more transient hypotension (1.3% vs 0.2%), pneumonia (4.6% vs 3.1%), and congestive cardiac failure (1.6% vs 0.7%) with nerinetide.
- First large trial of any neuroprotectant in the modern EVT era; overall negative but generated the hypothesis, later tested in ESCAPE-NEXT, that nerinetide might benefit patients treated without alteplase.
Study Design
- Study Type
- Randomised, double-blind, placebo-controlled, parallel-group, single-dose phase 3 multicentre trial
- Randomization
- Yes
- Blinding
- Double-blind; all trial personnel and patients masked. Nerinetide and placebo supplied as visually identical colourless solutions in numbered refrigerated vials; internet-based dynamic minimisation randomisation stratified by IV alteplase use and declared first EVT device.
- Sample Size
- 1105
- Follow-up
- 90 days (primary); imaging at 24 h; clinical follow-up at 30 and 90 days
- Centers
- 48
- Countries
- Canada, USA, Germany, Australia, South Korea, Sweden, Ireland, UK
Primary Outcome
Definition: Favourable functional outcome defined as modified Rankin Scale (mRS) 0-2 at 90 days post-randomisation (in-person, or telephone if in-person visit not possible)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 329/556 (59.2%) | 337/549 (61.4%) | - (0.96-1.14) | 0.35 |
Limitations & Criticisms
- Primary endpoint negative; the 'benefit in the no-alteplase stratum' is an exploratory subgroup finding after a failed primary and should be considered hypothesis-generating.
- The alteplase stratum is confounded by shorter onset-to-treatment times and different clinical selection (contra-indications to alteplase drive who is in the no-alteplase group), so effect modification could reflect population differences, not only a drug-drug interaction.
- Enrollment window (≤12 h) predates the DAWN/DEFUSE-3 based extended-window standard, and inclusion required moderate-to-good CTA collaterals — limiting generalisability to broader LVO populations.
- Randomisation stratified by 'declared' first device rather than actual device used; declared device proved unreliable as devices/practice evolved during the trial.
- Sponsor NoNO (manufacturer) participated in study design and had two members on the Publication Committee; the corresponding author had final publication responsibility but analyses were run jointly by an independent statistical group and academic investigators.
- Pharmacokinetic support for the plasmin-cleavage hypothesis came from only 22 patients.
- Not all patients had MRI-based infarct volumes (57.9% CT, 42.1% MRI).
Citation
Lancet 2020;395(10227):878-887. DOI: 10.1016/S0140-6736(20)30258-0