PDF: EVT Beyond 24h Kobeissi
(2023)Objective
Pooled outcomes of endovascular thrombectomy performed more than 24 hours after last known well in acute ischemic stroke — systematic review and meta-analysis of 7 observational studies (n=569) evaluating feasibility, safety, and functional outcomes.
Study Summary
• Primary outcome: 90-day mRS 0-2 achieved in 32.0% (95% CI 24.7-40.2%) — comparable to late-window (6-24 h) trial populations.
• TICI 2b-3 reperfusion in 81.9% (95% CI 78.5-84.9%); TICI 3 in 45.3% — technical success rates are preserved in the very late window.
• Symptomatic ICH rate 6.8% and 90-day mortality 27.2% — safety profile similar to published late-window trials.
• Significant heterogeneity in functional outcomes (I²=66%), reflecting variable patient selection (imaging criteria, collateral status, tissue-window vs time-window).
• All included studies were observational, without control arm — establishes feasibility but not efficacy vs medical management beyond 24 hours.
Intervention
Systematic review and meta-analysis of published studies (2018-2023) of EVT in acute ischemic stroke from large vessel occlusion presenting >24 hours after last known well. Pooled frequencies of clinical and procedural outcomes using random-effects generalized linear mixed models.
Inclusion Criteria
Original observational studies reporting EVT for acute ischemic stroke with LVO presenting beyond 24 hours after last known well; ≥10 patients per study; reported 90-day mRS, TICI, sICH, or mortality outcomes.
Study Design
Arms: Single-arm pooled analysis — EVT beyond 24 h (no comparator)
Patients per Arm: 569 patients across 7 studies (sample sizes 21-185 per study)
Outcome
• TICI 2b-3: 81.9% (95% CI 78.5-84.9%); I²=0%
• TICI 3: 45.3% (95% CI 36.6-54.4%); I²=62%
• sICH: 6.8% (95% CI 4.3-10.7%)
• 90-day mortality: 27.2% (95% CI 22.9-31.9%); ENI 36.9%; END 14.3%
Clinical Question
What are the pooled rates of functional independence, reperfusion, symptomatic ICH, and mortality after EVT for acute ischemic stroke presenting more than 24 hours after last known well?
Bottom Line
In a pooled analysis of 7 studies (569 patients) of EVT in the very late window (>24 h, mean 46.2 h to puncture), 32.0% of patients achieved 90-day functional independence (mRS 0-2), TICI 2b-3 reperfusion was achieved in 81.9%, symptomatic ICH rate was 6.8%, and 90-day mortality was 27.2% — suggesting EVT is feasible and reasonably safe in highly selected patients beyond 24 hours, though RCTs are needed to define who benefits.
Major Points
- Systematic review and meta-analysis registered with PROSPERO (CRD42023386868); PRISMA 2020 reporting
- 7 observational studies (6 prospective database, 4 multicenter) meeting inclusion criteria
- 569 patients; mean age 68.9 y; 55% male; mean NIHSS 13.6; mean ASPECTS 7.9
- Mean last known well to puncture time: 46.2 hours (95% CI 32.4-65.9)
- Only 4.9% received IV thrombolytics (outside lytic window)
- Primary outcome — 90-day mRS 0-2: 32.0% (95% CI 24.7-40.2%); I²=66%
- TICI 2b-3 reperfusion: 81.9% (95% CI 78.5-84.9%); TICI 3: 45.3%
- Symptomatic ICH: 6.8% (95% CI 4.3-10.7%); 90-day mortality: 27.2% (22.9-31.9%)
- Early neurological improvement: 36.9%; early neurological deterioration: 14.3%
- Risk of bias: 6 studies rated good (Newcastle-Ottawa), 1 fair
- Heterogeneity reflects variable selection — imaging criteria, collaterals, tissue-window vs time-window
- Conclusions: EVT feasible and reasonably safe beyond 24 h in selected patients; RCTs needed
- Informed subsequent RCTs — MR CLEAN-LATE, RESCUE-Japan LIMIT, and ongoing trials for extended window
Study Design
- Study Type
- Systematic review and meta-analysis of observational studies
- Randomization
- No
- Blinding
- Not applicable (review)
- Sample Size
- 569
- Follow-up
- 90 days (for mRS, mortality)
Primary Outcome
Definition: 90-day functional independence (mRS 0-2)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| No control | 32.0% | - (24.7-40.2%) | I²=66%, heterogeneity P=0.01 |
Limitations & Criticisms
- No randomized comparator — cannot establish causal benefit over medical management in this window
- Publication/selection bias — only centers with favorable experience likely published their very-late-window cohorts
- Significant heterogeneity in functional outcomes (I²=66%) limits precision of the point estimate
- Variable imaging selection criteria (CTP, MRI-DWI, collateral score) not harmonized across studies
- Small total sample (N=569) across 7 studies; subgroup analyses underpowered
- Only 4.9% received IV thrombolytics; generalizability to patients who could be treated with lytics is limited
- Most studies from experienced high-volume centers — results may not generalize to community practice