DEVT
(2021)Objective
To determine whether endovascular thrombectomy alone is noninferior to IV alteplase plus thrombectomy in anterior circulation LVO stroke within 4.5 hours.
Study Summary
• Endovascular thrombectomy alone (n=116) vs IV alteplase + thrombectomy (n=118) in anterior circulation LVO within 4.5 hours.
• Primary outcome (mRS 0–2 at 90d): 54.3% vs 46.6%; difference +7.7% (1-sided 97.5% CI −5.1% to ∞); P for noninferiority = 0.003 — noninferiority met.
• Excellent outcome (mRS 0–1): 37.9% vs 31.4% (NS).
• Successful reperfusion (eTICI 2b–3): 88.5% vs 87.2% (NS).
• Symptomatic ICH (Heidelberg): 6.1% vs 6.8% (NS). Any ICH: 21.7% vs 32.5%.
• 90-day mortality: 17.2% vs 17.8% (NS).
• Clot migration: 17.7% vs 23.9%. Procedure complications: 30.1% vs 41.0%.
• Trial stopped early after first interim analysis (234/970 enrolled) for efficacy.
• Primary outcome (mRS 0–2 at 90d): 54.3% vs 46.6%; difference +7.7% (1-sided 97.5% CI −5.1% to ∞); P for noninferiority = 0.003 — noninferiority met.
• Excellent outcome (mRS 0–1): 37.9% vs 31.4% (NS).
• Successful reperfusion (eTICI 2b–3): 88.5% vs 87.2% (NS).
• Symptomatic ICH (Heidelberg): 6.1% vs 6.8% (NS). Any ICH: 21.7% vs 32.5%.
• 90-day mortality: 17.2% vs 17.8% (NS).
• Clot migration: 17.7% vs 23.9%. Procedure complications: 30.1% vs 41.0%.
• Trial stopped early after first interim analysis (234/970 enrolled) for efficacy.
Intervention
Endovascular thrombectomy alone vs. IV alteplase (0.9 mg/kg) followed by thrombectomy. Patients randomized within 4.5 hours of onset.
Inclusion Criteria
Age ≥18; acute ischemic stroke; intracranial ICA or M1 occlusion on CTA/MRA; eligible for IV alteplase within 4.5 hours; randomization ≤4h15m from onset; pre-stroke mRS 0–1.
Study Design
Arms: Thrombectomy Alone vs. Alteplase + Thrombectomy
Patients per Arm: Thrombectomy: 116, Combined: 118
Outcome
• Primary (mRS 0–2 at 90d): 54.3% vs. 46.6%, difference 7.7%, met noninferiority margin.• mRS 0–1: 37.9% vs. 31.4%.• Median mRS: 2 vs. 3.• Successful reperfusion (eTICI 2b–3): 88.5% vs. 87.2%.• Mortality at 90d: 17.2% vs. 17.8%.• Symptomatic ICH (Heidelberg): 6.1% vs. 6.8%.• Any ICH: 21.7% vs. 32.5%.• Serious adverse events: 25.9% vs. 22.0%.• Procedure complications: 30.1% vs. 41.0%.
Bottom Line
Among 234 patients with proximal anterior circulation LVO stroke eligible for IV alteplase within 4.5 hours, endovascular thrombectomy alone achieved functional independence (mRS 0-2 at 90 days) in 54.3% vs 46.6% with combined alteplase plus thrombectomy — meeting the prespecified noninferiority margin of -10% (difference 7.7%; 1-sided 97.5% CI, -5.1% to infinity; P for noninferiority = 0.003), with no significant differences in symptomatic ICH (6.1% vs 6.8%) or 90-day mortality (17.2% vs 17.8%).
Major Points
- DEVT was a multicenter, randomized, open-label, blinded-endpoint noninferiority trial conducted at 33 stroke centers in China, enrolling 234 of a planned 970 patients before early termination for efficacy.
- The trial enrolled patients aged ≥18 years with ICA or M1 occlusion confirmed on CTA or MRA who were eligible for IV alteplase within 4.5 hours; randomization had to occur within 4 hours 15 minutes of onset.
- Primary outcome (mRS 0-2 at 90 days): 54.3% (63/116) in thrombectomy-alone vs 46.6% (55/118) in combined treatment; difference +7.7% (1-sided 97.5% CI, -5.1% to infinity); P for noninferiority = 0.003, crossing the prespecified efficacy boundary.
- No significant difference in symptomatic ICH by Heidelberg classification: 6.1% (7/115) vs 6.8% (8/117); difference -0.8% (95% CI, -7.1% to 5.6%). Asymptomatic ICH was numerically lower in thrombectomy-alone: 15.7% vs 25.6% (difference -10.0%; 95% CI, -20.3% to 0.3%).
- 90-day mortality was similar: 17.2% (20/116) vs 17.8% (21/118); difference -0.5% (95% CI, -10.3% to 9.2%). Any intracranial hemorrhage: 21.7% (25/115) vs 32.5% (38/117).
- Successful reperfusion on final angiogram (eTICI 2b-3): 88.5% (100/113) vs 87.2% (102/117). Reperfusion on 48-hour CTA/MRA: 97.0% (96/99) vs 93.1% (94/101).
- Pre-thrombectomy reperfusion on initial angiogram (post hoc): only 1.7% (2 patients) in thrombectomy-alone vs 2.6% (3 patients) in combined group — suggesting alteplase provided minimal prethrombectomy benefit.
- Clot migration occurred less frequently in thrombectomy-alone: 17.7% (20/113) vs 23.9% (28/117). Procedure-associated complications overall: 30.1% (34/113) vs 41.0% (48/117).
- Median door-to-needle time in combined arm: 61 minutes (IQR 49-81), comparable to DIRECT-MT (59 min) but longer than Western trials, reflecting China's informed-consent requirement before alteplase.
- Trial stopped early by DSMB after first planned interim analysis (194 patients, 20% of maximum sample) on May 12, 2020; thrombectomy-alone group showed 54.64% vs 47.42% functional independence (difference 7.2%; z=2.4042, P noninferiority=0.008), crossing prespecified efficacy boundary (z=2.35826, P=0.009).
Study Design
- Study Type
- Multicenter, randomized, open-label, blinded-endpoint (PROBE design), noninferiority trial
- Randomization
- Yes
- Blinding
- Open-label treatment; blinded central outcome assessment by 2 independent mRS-certified neurologists via video/voice recording; blinded imaging core laboratory; blinded independent clinical events committee for adverse events
- Sample Size
- 234
- Follow-up
- 90 days (final follow-up July 22, 2020)
- Centers
- 33
- Countries
- China
Primary Outcome
Definition: Functional independence at 90 days (mRS 0-2)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 55/118 (46.6%) | 63/116 (54.3%) | 1.36 (0.82-2.28) | 0.003 (noninferiority) |
Limitations & Criticisms
- Trial stopped early at 234 of planned 970 patients (24.1% enrolled); early termination risks overestimation of treatment effects, though a highly conservative prespecified Pocock analog stopping rule was used
- Noninferiority margin of -10% was not derived from the minimal clinically important difference or fixed-margin methodology, and no consensus exists for this threshold in the stroke field
- Asian-only population (33 centers in China) with high intracranial atherosclerotic disease prevalence (24.1% vs 19.5% ICAD) limits generalizability to Western populations
- Median door-to-needle time of 61 minutes (IQR 49-81) is longer than most Western thrombectomy trials, potentially underestimating alteplase benefit when given faster (e.g., EXTEND-IA-TNK with tenecteplase showing better prethrombectomy reperfusion)
- Informed consent required before IV alteplase in China (financial pre-payment), creating unavoidable delay that may have biased against bridging therapy
- M2 MCA occlusions explicitly excluded (which respond better to IVT), potentially selecting a population where alteplase adds less benefit
- Pre-thrombectomy reperfusion rate was very low in both arms (1.7% vs 2.6%), suggesting limited opportunity for alteplase to avert thrombectomy in this cohort
- Only Asian populations studied; DIRECT-MT and SKIP trials also limited to Asian patients, restricting generalizability to Western stroke populations with different ICAD burden
Citation
JAMA. 2021;325(3):234-243. doi:10.1001/jama.2020.23523