ASTER
Contact Aspiration vs Stent Retriever for Successful Revascularization
Clinical Question
Does first-line contact aspiration result in an increased successful revascularization rate compared with first-line use of a stent retriever in patients with acute ischemic stroke undergoing thrombectomy?
Study Overview
Objective
To compare efficacy and adverse events of first-line contact aspiration vs stent retriever technique for successful revascularization in patients with acute ischemic stroke and large vessel occlusion
Study Summary
- Contact aspiration was not superior to stent retriever for achieving successful revascularization (85.4% vs 83.1%, OR 1.20, P=0.53)
- No significant differences in functional independence at 90 days (45.3% vs 50.0%)
- Similar safety profiles with no significant differences in symptomatic ICH or mortality
Intervention
First-line contact aspiration (ADAPT technique) vs first-line stent retriever thrombectomy
Patients per Arm
192 contact aspiration, 189 stent retriever
Bottom Line
First-line contact aspiration did not result in increased successful revascularization compared with first-line use of a stent retriever. No significant differences were found in clinical outcomes or adverse events.
Major Points
- First multicenter RCT comparing contact aspiration (ADAPT technique) vs stent retriever as first-line thrombectomy technique — 381 patients at 8 French comprehensive stroke centers.
- Primary endpoint (successful revascularization, mTICI 2b/3) was similar: 85.4% aspiration vs 83.1% stent retriever (OR 1.20, 95% CI 0.68–2.10, P=0.53; risk difference 2.4%, 95% CI −5.4% to 9.7%) — no superiority of aspiration.
- No significant difference in functional independence at 90 days (45.3% aspiration vs 50.0% stent retriever, P=0.38); the trial was not designed to claim equivalence or noninferiority of clinical outcomes.
- Aspiration showed a trend toward faster procedure time (38 vs 45 min puncture-to-revascularization, P=0.10) with numerically more rescue therapy use (32.8% vs 23.8%, OR 1.57, P=0.05) — reported as no significant difference between groups.
- Similar safety profiles: sICH 5.3% vs 6.5%, 90-day mortality 19.3% vs 19.2% — no significant differences in any adverse event category.
- Both techniques achieved high revascularization rates at experienced centers, suggesting operators may reasonably choose based on anatomy, clot characteristics, and personal expertise — with the important caveat that this superiority trial cannot formally support equivalence.
- Rescue therapy was permitted after ≥3 attempts and its use was not significantly different between groups overall or among patients with first-line failure (OR 1.21, P=.72) — the trial was not designed to establish noninferiority.
- Prespecified subgroup analyses (IV tPA, site of occlusion, clot burden score) showed no significant heterogeneity of treatment effect on the primary outcome.
- All centers had high expertise in both techniques — results may not apply to centers where operators are more experienced with one approach over the other.
- Powered for superiority (15% absolute difference) — the trial cannot exclude smaller but clinically meaningful differences between techniques, nor prove noninferiority.
Design
Study Type: Randomized, multicenter, open-label, blinded end-point clinical trial
Randomization: 1
Blinding: Open-label, blinded endpoint assessment (PROBE)
Enrollment Period: October 2015–October 2016
Follow-up Duration: 90 days
Centers: 8
Countries: France
Sample Size: 381
Analysis: Intention-to-treat; mixed logistic regression model adjusted for randomization stratification variables (IV thrombolysis as fixed effect, center as random effect) for the primary outcome
Inclusion Criteria
- Adults with suspected acute ischemic stroke secondary to occlusion of the anterior circulation
- Imaging evidence of occlusion of the intracranial internal carotid artery or the M1 or M2 branches of the middle cerebral artery
- Within 6 hours of symptom onset
- IV thrombolysis permitted (not required)
Exclusion Criteria
- Cerebral infarction of the posterior circulation
- Proven occlusion of the cervical carotid artery
- Prestroke modified Rankin Scale score >3
- Enrollment in another interventional trial
Arms
| Field | Contact Aspiration (ADAPT) | Control |
|---|---|---|
| Intervention | First-line contact aspiration thrombectomy technique | First-line stent retriever thrombectomy technique |
| Duration | Single procedure | Single procedure |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Successful revascularization (mTICI 2b/3) at the end of all endovascular procedures | Primary | 83.1% (stent retriever) | 85.4% (aspiration) | 1.2 | 0.53 |
| Functional independence (mRS 0-2) at 90 days | Secondary | 50.0% | 45.3% | OR 0.83 (0.54-1.26) | 0.38 |
| Puncture-to-revascularization time (min, median) | Secondary | 45 | 38 | 0.10 | |
| Rescue therapy used (no significant difference between groups) | Secondary | 23.8% | 32.8% | OR 1.57 (0.99-2.47) | 0.05 |
| Symptomatic ICH | Adverse | 6.5% | 5.3% | NS | |
| 90-day mortality | Adverse | 19.2% | 19.3% | NS |
Subgroup Analysis
No significant interaction by IV tPA treatment, site of occlusion (ICA vs M1 vs M2), or clot burden score in prespecified subgroups. No heterogeneity by clot length in unplanned analysis.
Criticisms
- Primary endpoint was technical (revascularization) rather than clinical (90-day mRS)
- Superiority design powered to detect 15% difference; not designed for noninferiority or equivalence
- Cannot exclude smaller but clinically significant difference between groups
- Rescue therapy at operator discretion is potential source of bias
- All participating centers highly experienced in both techniques; may limit generalizability
- Degree of randomness limited by stratified permuted randomization with fixed blocks of 4
- Not powered for subgroup analyses
- Open-label design with only endpoint blinding
- 24-hour NIHSS assessed by unblinded treating physician
Funding
Sponsored by Fondation Ophtalmologique Adolphe de Rothschild. Unrestricted research grant provided by Penumbra, Alameda, California. Funders had no role in study design, conduct, data analysis, or manuscript preparation.
Based on: ASTER (JAMA, 2017)
Authors: Bertrand Lapergue, Raphael Blanc, Benjamin Gory, ..., for the ASTER Trial Investigators
Citation: JAMA. 2017;318(5):443-452
Content summarized and formatted by NeuroTrials.ai.