DUSK
Endovascular Versus Medical Management in Distal Medium Vessel Occlusion Stroke: The DUSK Study
Year of Publication: 2024
Authors: Mahmoud H. Mohammaden, Lorena Souza Viana, Hend Abdelhamid, ..., Raul G. Nogueira
Journal: Stroke
Citation: Stroke. 2024;55:1489–1497. DOI: 10.1161/STROKEAHA.123.045228
Link: https://www.ahajournals.org/doi/10.1161/STROKEAHA.123.045228
PDF: https://www.ahajournals.org/doi/reader/1...EAHA.123.045228
PDF: Clinical Question
Does endovascular treatment (EVT) improve clinical outcomes compared to medical management (MM) in patients with isolated distal medium vessel occlusion (DMVO) strokes?
Bottom Line
EVT did not demonstrate significant outcome differences compared to MM in patients with isolated DMVO, reinforcing clinical equipoise. However, subgroup analysis suggested a potential benefit of EVT for excellent outcome (mRS 0–1) in patients with moderate-to-severe strokes (NIHSS ≥8).
Major Points
- No significant difference in 90-day mRS ordinal shift between EVT and MM in both multivariable and IPTW models.
- Rates of good (mRS 0–2) and excellent (mRS 0–1) outcomes were not significantly different between groups.
- Successful reperfusion (mTICI ≥2B) was achieved in 89.4% of EVT patients.
- Subgroup analysis showed EVT was associated with higher likelihood of excellent outcome in NIHSS ≥8 patients (aOR 3.00, 95% CI 1.69–5.32; P-interaction=0.001), surviving Bonferroni correction.
- Safety measures (sICH and 90-day mortality) were comparable between EVT and MM in multivariable analysis.
- IPTW model unexpectedly showed lower sICH with EVT (aOR 0.46, P=0.013), likely due to large weights assigned to MM patients with sICH.
- 6 out of 9 sICH events in the MM group occurred in patients who did not receive IV thrombolysis.
- 40.8% of overall cohort received IV thrombolysis.
- EVT group had higher baseline NIHSS (10 vs 6), higher age (69.4 vs 66.4), and longer time to treatment (340 vs 249 minutes).
- Findings reinforce clinical equipoise and support ongoing randomized trials in DMVO.
Design
Study Type: Retrospective analysis of prospectively collected data (multicenter observational cohort)
Randomization:
Blinding: Unblinded; mRS assessed by experienced clinicians in an unblinded manner
Enrollment Period: January 2017 to June 2021
Follow-up Duration: 90 days
Centers: 7
Countries: USA, Spain
Sample Size: 321
Analysis: Multivariable ordinal logistic regression for mRS shift, binary logistic regression for dichotomized outcomes, inverse probability of treatment weighting (IPTW). Adjusted for age, stroke etiology, baseline NIHSS, time from LKN to imaging/treatment, occlusion site, baseline blood glucose, previous stroke, and IV thrombolysis. Bonferroni correction applied for subgroup analyses (P<0.0017). R Software v4.3.1.
Inclusion Criteria
- Isolated DMVO stroke on initial CTA: MCA-M3/M4, ACA-A2/A3, or PCA-P1/P2
- Baseline mRS score ≤2
- Last known normal (LKN) to imaging/treatment ≤24 hours
- Follow-up data available at 90 days
Exclusion Criteria
- Primary proximal large vessel occlusion
- Multivessel occlusion
- Missing baseline mRS scores
- Missing clinical variables
- PCA-P3 segment occlusion (all treated with MM)
- ACA-A1 occlusion (variability in classification)
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| N | 142 | 179 |
| Age - mean ± SD | 66.4 ± 13.5 years | 69.4 ± 13 years |
| Female | 59.2% | 65.9% |
| HTN | 76.1% | 81% |
| HLD | 41.5% | 37.4% |
| Diabetes | 39.4% | 33.5% |
| AF | 24.6% | 31.3% |
| Current smoker | 26.1% | 23.5% |
| Etiology - Cardioembolic | 41.5% | 53.1% |
| Etiology - Large vessel disease | 8.5% | 9.5% |
| Etiology - ICAD | 7% | 9.5% |
| Etiology - Others | 43% | 27.9% |
| Initial blood glucose - median [IQR] | 66 [59–76] mg/dL | 128 [108–151] mg/dL |
| Previous stroke | 11.3% | 16.8% |
| Baseline NIHSS - median [IQR] | 6 [3–11] | 10 [7–16] |
| IV thrombolysis | 40.8% | 41.3% |
| MCA-M3 occlusion | 26.8% | 34.1% |
| MCA-M4 occlusion | 2.1% | 4.5% |
| ACA-A2 occlusion | 10.6% | 13.4% |
| ACA-A3 occlusion | 4.2% | 2.2% |
| PCA-P1 occlusion | 20.4% | 28.5% |
| PCA-P2 occlusion | 35.9% | 17.3% |
| Time to imaging/treatment - median [IQR] | 249 [120–629] min | 340 [200–694] min |
| General anesthesia | 16.2% | |
| Successful reperfusion (mTICI ≥2B) | 89.4% |
Arms
| Field | EVT | Control |
|---|---|---|
| Intervention | Mechanical thrombectomy (technique and devices at operator discretion) ± IV thrombolysis if eligible | Standard medical treatment including IV thrombolysis if presenting within 4.5 hours with no contraindications |
| Duration |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Ordinal shift in 90-day mRS (degree of disability) | Primary | Median mRS 3 [2–5] | Median mRS 3 [2–5] | IPTW: P=0.110; Multivariable: P=0.556 | |
| Good outcome (mRS 0–2 at 90 days) | Secondary | 86/142 (60.6%) | 96/179 (53.6%) | IPTW aOR 1.32 (95% CI 0.97–1.80); Multivariable aOR 1.39 (95% CI 0.80–2.45) | IPTW P=0.075; Multivariable P=0.244 |
| Excellent outcome (mRS 0–1 at 90 days) | Secondary | 56/120 (46.7%) | 62/163 (38%) | IPTW aOR 1.32 (95% CI 0.94–1.85); Multivariable aOR 0.96 (95% CI 0.53–1.74) | IPTW P=0.098; Multivariable P=0.884 |
| 24-hour NIHSS score | Secondary | Median 3 | Median 5 | P=0.03 | |
| Symptomatic ICH | Adverse | 9/142 (6.3%) | 10/179 (5.6%) | Multivariable aOR 0.57 (95% CI 0.21–1.58); IPTW aOR 0.46 (95% CI 0.24–0.85) | Multivariable P=0.277; IPTW P=0.013 |
| 90-day Mortality | Adverse | 13/142 (9.2%) | 31/179 (17.3%) | Multivariable aOR 1.73 (95% CI 0.80–3.90); IPTW aOR 1.20 (95% CI 0.78–1.85) | Multivariable P=0.173; IPTW P=0.395 |
Subgroup Analysis
EVT had a significant treatment effect on excellent outcome (mRS 0–1) in patients with baseline NIHSS ≥8 (aOR 3.00, 95% CI 1.69–5.32) vs NIHSS <8 (aOR 0.84, 95% CI 0.51–1.40; P-interaction=0.001), surviving Bonferroni correction. A trend favoring EVT for good outcome in NIHSS ≥8 (P-interaction=0.008) did not survive Bonferroni correction. Posterior circulation DMVO showed higher mortality trend (aOR 1.87) vs anterior (aOR 0.73; P-interaction=0.034), but lost significance after Bonferroni correction. No definite treatment effect modification by age, IV thrombolysis, time to treatment, occlusion site, or mismatch volume for primary outcome.
Criticisms
- Retrospective, non-randomized design with potential selection bias.
- Unblinded outcome assessment (mRS).
- Time of imaging acquisition unavailable in EVT group; puncture time used as surrogate.
- Study not powered to detect small treatment differences.
- Small sample size limited power for subgroup and safety analyses.
- CTP parameters not included in IPTW model (>50% missing in each group).
- No core laboratory for central adjudication of occlusion site or reperfusion rates.
- Cognitive outcomes not assessed, particularly relevant for posterior circulation DMVO.
- mRS may not capture all DMVO-related deficits (cognitive, visual impairments).
- Decision to perform EVT vs MM was not standardized across centers.
- Baseline imbalances between groups (NIHSS, age, time to treatment) despite IPTW adjustment.
Subgroup Analysis
Significant benefit of EVT was observed for excellent outcome (mRS 0–1) among patients with NIHSS ≥8 (aOR 3.00; 95% CI 1.69–5.32; P_interaction=0.001). No significant treatment effect modification was detected in other subgroups, including IV thrombolysis, age, occlusion site, and imaging mismatch volume.
Funding
None
Based on: DUSK (Stroke, 2024)
Authors: Mahmoud H. Mohammaden, Lorena Souza Viana, Hend Abdelhamid, ..., Raul G. Nogueira
Citation: Stroke. 2024;55:1489–1497. DOI: 10.1161/STROKEAHA.123.045228
Content summarized and formatted by NeuroTrials.ai.