SELECT2 Trial: Large Core Thrombectomy
In anterior LVO patients with a large ischemic core (ASPECTS 3–5 or core ≥50 mL) within 24 hours of last known well, does endovascular thrombectomy improve the mRS distribution at 90 days compared with medical management?
Sarraj et al. (NEJM 2023;388(14):1259–1271) · doi:10.1056/NEJMoa2214403 · 352 patients
Population
Included
- Anterior LVO (ICA or M1)
- Age ≤85
- Pre-stroke mRS 0–1
- Large ischemic core: ASPECTS 3–5 OR perfusion core ≥50 mL
- Within 24 hours of last known well
Clinical
- Adults (18–85 years) with final acute ischemic stroke diagnosis
- NIHSS score ≥ 6
- Last known well to groin puncture or medical management: 0–24 hours
- Pre-stroke mRS of 0–1
- Eligible for thrombectomy or medical management
- Signed informed consent obtained
- Protocol follow-up compliance commitment
- Anticipated life expectancy ≥ 3 months
Neuroimaging
- Large vessel occlusion (ICA or MCA-M1) via MRA or CTA
- Large infarct-core on: ASPECTS 3–5, CT perfusion (rCBF<30% ≥50cc), or MRI-DWI (ADC<620 ≥50cc)
Excluded
- ASPECTS ≥6 (use small-core EVT evidence)
- Pre-stroke mRS ≥2
- Age >85
- Significant mass effect on baseline imaging
Clinical
- Unable to undergo CT angiography/perfusion
- Co-morbid conditions confounding assessments
- Thrombolytic treatment >4.5 hours post-onset
- Thrombolytic at 3–4.5 hours with age >80, anticoagulant use, diabetes with prior stroke, NIHSS >25, or >1/3 MCA involvement
- Current participation in other investigational studies
Neuroimaging
- ASPECTS 6–10 AND core <50cc
- ASPECTS ≤2
- Intracranial tumor, hemorrhage, or AVM
- Significant mass effect with midline shift
- Flow-limiting carotid dissection or aortic dissection
- Intracranial stent precluding device deployment
- Multiple vascular territory occlusions
- Established infarct with cerebral edema
Source: ClinicalTrials.gov NCT03876457· Retrieved 2026-06-08
Primary Outcome — Ordinal mRS Shift
Ordinal-shift design: Primary is generalized OR for ordinal mRS shift (gOR 1.51, 95% CI 1.20–1.89). The mRS 0–2 binary shown below (functional independence) is a secondary outcome; frame as "less disability on average", not "independence restored" — only 20% achieve mRS 0–2.
mRS 0–2 at 90 Days (secondary)
Study Arms
- Agent
- Mechanical thrombectomy (stent retrievers, aspiration devices, or both; various manufacturers)
- Route
- Endovascular
- Frequency
- Single procedure
- Duration
- Expected to begin within 24 h of stroke onset
- Co-interventions
- IV thrombolysis (alteplase or tenecteplase) for patients first assessed within 4.5 h of onset; standard medical care per AHA/ASA, ESO, and Australia/NZ Stroke Foundation guidelines (BP management, critical care, rehabilitation); decompressive hemicraniectomy for severe swelling per local practice
Large-core selection: ASPECTS 3–5 on NCCT or core ≥50 mL (rCBF<30% via RAPID, or ADC<620 on DWI); no upper core-volume limit. Tandem and isolated cervical ICA occlusions allowed. 1:1 adaptive minimization (Sarraj NEJM 2023 p.1261).
- Agent
- Standard medical management (no thrombectomy)
- Route
- Medical
- Duration
- Through hospitalization + outpatient follow-up
- Co-interventions
- IV thrombolysis (alteplase or tenecteplase) for patients first assessed within 4.5 h of onset; standard medical care per AHA/ASA, ESO, and Australia/NZ Stroke Foundation guidelines; decompressive hemicraniectomy for severe swelling per local practice
Both arms received guideline-based medical care; IV thrombolysis administered identically in both arms when within 4.5 h. Substantive standard-of-care control, not sham (Sarraj NEJM 2023 p.1261).
Safety
Symptomatic ICH
0.6%
1.1%
0.6% EVT vs 1.1% medical (NS). Lower than ANGEL-ASPECT (6.1% vs 2.7%). SELECT2 had a notably cleaner safety profile.
90-day mortality
38.4%
41.5%
38.4% vs 41.5% (NS). Mortality is high in large-core patients overall; EVT does not worsen it.
Trial Design
Type
- Randomized open-label international trial
- Large core selection (ASPECTS 3-5 or Core ≥50ml)
- 1:1 allocation (Thrombectomy vs. Medical)
Timeline
Enrolled 2019-2022
N
352
Enrollment
352 patients (178 EVT / 174 medical) internationally; age ≤85. Stopped early at 2nd interim for efficacy. Published NEJM 2023;388(14):1259–1271.
ClinicalTrials.gov
NCT03876457Bedside Pearl
For anterior LVO with large ischemic core (ASPECTS 3–5 or core ≥50 mL) within 24h, EVT shifts disability one step lower on average (gOR 1.51). Functional independence (mRS 0–2) is uncommon at 20% (vs 7% medical). Frame as "less disability" not "independence restored". NNT 7.7 is from the secondary mRS 0–2 outcome.
See also