DIRECT-SAFE
DIRECT-SAFE: A Randomized Controlled Trial of DIRECT Endovascular Clot Retrieval versus Standard Bridging Therapy
Year of Publication: 2022
Authors: Peter J. Mitchell, Bernard Yan, Leonid Churilov, ..., Stephen M. Davis
Journal: Journal of Stroke
Citation: Journal of Stroke 2022:24(1):57-64. https://doi.org/10.5853/jos.2021.03475
PDF: Clinical Question
Does direct endovascular clot retrieval (DIRECT EVT) improve clinical outcome in acute ischemic stroke with large vessel occlusion compared to standard bridging therapy (intravenous alteplase followed by EVT)?
Bottom Line
Direct endovascular clot retrieval (DIRECT EVT) was non-inferior to standard bridging therapy for functional independence at 90 days in patients with acute ischemic stroke and large vessel occlusion. DIRECT EVT also reduced the overall time to reperfusion, with similar rates of symptomatic intracranial hemorrhage, suggesting it is a safe and effective alternative to bridging therapy.
Major Points
- 529 patients with acute ischemic stroke and large vessel occlusion were randomized (265 to DIRECT EVT, 264 to bridging therapy).
- The trial was stopped early due to futility for superiority, but continued for non-inferiority.
- The primary outcome (functional independence [mRS 0-2] at 90 days) occurred in 55% in the DIRECT EVT group and 52% in the bridging therapy group (difference, 3%; 95% CI, -6 to 9; p=0.003 for non-inferiority).
- Overall reperfusion (mTICI 2b-3) was achieved in 91% in the DIRECT EVT group and 89% in the bridging therapy group (p=0.38).
- Median time from randomization to reperfusion was significantly shorter with DIRECT EVT (45 minutes vs. 59 minutes; p<0.001).
- Symptomatic intracranial hemorrhage occurred in 3% in both groups (difference, 0%; 95% CI, -2 to 2; p=0.96).
- 90-day mortality was 14% in the DIRECT EVT group and 16% in the bridging therapy group (difference, -2%; 95% CI, -7 to 3; p=0.45).
- Patients in the DIRECT EVT group experienced less mild bleeding (16% vs. 23%; p=0.03) and fewer serious adverse events (23% vs. 32%; p=0.02) compared to bridging therapy.
Design
Study Type: Randomized controlled trial (Prospective, multicenter, open-label, blinded outcome assessment, non-inferiority, parallel group)
Randomization: 1
Blinding: Blinded outcome assessment (mRS at 90 days assessed by a central adjudication committee unaware of treatment allocation).
Enrollment Period: September 2017 to December 2020 (stopped early)
Follow-up Duration: 90 days
Centers: 20
Countries: Australia, Vietnam, China
Sample Size: 529
Analysis: Intention-to-treat (ITT) analysis. Primary outcome (mRS 0-2 at 90 days) analyzed using a logistic regression model adjusted for stratification factors. Non-inferiority margin for the difference in mRS 0-2 was -10%. Secondary and safety outcomes were analyzed using logistic or linear regression as appropriate. P-values are 2-sided. Missing mRS data handled using imputation if needed. Statistical analysis performed using SAS, version 9.4.
Inclusion Criteria
- Adults (≥18 years old).
- Acute ischemic stroke due to large vessel occlusion in the anterior circulation (intracranial internal carotid artery, M1, or M2 segment of the middle cerebral artery).
- National Institutes of Health Stroke Scale (NIHSS) score of ≥6.
- Last known normal (LKN) to randomization time of ≤4.5 hours.
- Prestroke modified Rankin Scale (mRS) score of 0-1.
Exclusion Criteria
- Known intracranial hemorrhage on baseline imaging.
- Prior stroke resulting in mRS >1.
- Significant pre-existing disability (mRS >1).
- Planned mechanical thrombectomy for extracranial occlusions.
- Contraindication to intravenous thrombolysis.
- Participation in another stroke intervention trial.
- Pregnancy or breastfeeding.
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| Age, mean (SD), y | 71 (13) | 70 (14) |
| Male sex, n (%) | 146 (55) | 147 (56) |
| LKN to randomization, median (IQR), min | 165 (120-210) | 165 (120-210) |
| LKN to imaging, median (IQR), min | 100 (60-150) | 100 (60-150) |
| Baseline NIHSS, median (IQR) | 17 (14-20) | 17 (14-20) |
| Baseline ASPECTS, median (IQR) | 9 (8-10) | 9 (8-10) |
| Site of occlusion - ICA, n (%) | 49 (19) | 52 (20) |
| Site of occlusion - M1, n (%) | 194 (74) | 189 (72) |
| Site of occlusion - M2, n (%) | 19 (7) | 21 (8) |
| Atrial fibrillation, n (%) | 70 (27) | 64 (24) |
| History of stroke/TIA, n (%) | 24 (9) | 26 (10) |
| History of hypertension, n (%) | 170 (65) | 176 (67) |
| History of diabetes, n (%) | 72 (27) | 70 (27) |
| History of hyperlipidemia, n (%) | 82 (31) | 79 (30) |
| Current smoker, n (%) | 40 (15) | 42 (16) |
Arms
| Field | DIRECT EVT | Control |
|---|---|---|
| Intervention | Direct endovascular clot retrieval (EVT) without prior intravenous alteplase. Any EVT device (stent retriever or aspiration catheter) could be used at the discretion of the operator. | Standard bridging therapy: intravenous alteplase (0.9 mg/kg) followed by EVT. |
| Duration | 90 days | 90 days |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Functional independence at 90 days, defined as a modified Rankin Scale (mRS) score of 0 to 2. | Primary | 52% | 55% | 3.00% | 0.003 (for non-inferiority) |
| Overall reperfusion (mTICI 2b-3) | Secondary | 89% | 91% | 0.38 | |
| Symptomatic intracranial hemorrhage (sICH) | Secondary | 3% | 3% | 0.96 | |
| 90-day mortality | Secondary | 16% | 14% | 0.45 | |
| Any intracranial hemorrhage (ICH) | Secondary | 26% | 26% | 0.96 | |
| Time from randomization to reperfusion (minutes) | Secondary | 59 | 45 | <0.001 | |
| Symptomatic intracranial hemorrhage (sICH) | Adverse | 3% | 3% | 0.96 | |
| 90-day mortality | Adverse | 16% | 14% | 0.45 | |
| Mild bleeding | Adverse | 23% | 16% | 0.03 | |
| Serious adverse events | Adverse | 32% | 23% | 0.02 |
Criticisms
- The trial was stopped early due to futility for superiority, potentially limiting its ability to detect smaller, but clinically relevant, differences.
- The non-inferiority margin of -10% may be considered large by some, raising questions about whether direct EVT is truly comparable to bridging therapy for all aspects of outcome.
- The study was open-label, meaning patients and investigators were aware of the treatment assignment, which could introduce bias, though outcome assessment was blinded.
- The study did not mandate the use of advanced imaging (e.g., CT perfusion, MRI) for patient selection, relying primarily on noncontrast CT and CT angiography, which might affect generalizability to settings where more advanced imaging is standard.
- The trial did not specifically investigate the safety and efficacy of EVT alone versus IVT alone, as all patients in the control group eventually received EVT.
- The observed rates of ICH and mortality were similar between groups, but the study was not specifically powered to detect differences in these safety outcomes.
- The study had broad inclusion criteria across various stroke etiologies and patient characteristics, which may increase heterogeneity in the study population.
Subgroup Analysis
The interaction analysis for the primary outcome showed no significant subgroup interactions based on age, sex, baseline NIHSS, ASPECTS, stroke etiology, site of occlusion, and time from last known normal to randomization. There was no evidence of significant heterogeneity in the efficacy and safety outcomes across subgroups.
Funding
National Health and Medical Research Council (NHMRC) of Australia.
Based on: DIRECT-SAFE (Journal of Stroke, 2022)
Authors: Peter J. Mitchell, Bernard Yan, Leonid Churilov, ..., Stephen M. Davis
Citation: Journal of Stroke 2022:24(1):57-64. https://doi.org/10.5853/jos.2021.03475
Content summarized and formatted by NeuroTrials.ai.