Background:
Efficacy and safety of endovascular thrombectomy (EVT) in large vessel occlusion (LVO) patients with mild deficits is unclear.
Methods:
Pooled cohort of pts with mild deficits (NIHSS<6) due to (ICA, M1, M2) LVO from EXTEND IA TNK I & II RCTs and prospective data from 12 centers (US, AUS, NZ, Canada, Spain) from 1/2013 to 2/2020 was divided into medical management (MM) vs EVT. All pts had baseline CT, CTA, CTPRAPID software estimated ischemic core and mismatch. Pts stratified into with or without target profile (≥1cc core / mismatch ratio ≥ 1.8 / mismatch volume ≥ 15cc). Primary outcome- excellent (90 day mRS 0-1); Secondary- mRS shift, safety (sICH, neuro-worsening, mortality).
Results:
Of 371 pts, 189 (51%) had EVT. Time LKW to EVT center: EVT 165 (70- 416) vs MM 200 (72-564) min, p=0.35 were similar. EVT pts had larger perfusion lesions (51 cc (23-86) vs 30.1 (5, 65), p<0.001), higher NIHSS 4 (2-5) vs 3 (2-4), p=0.009), less IV tPA (30% vs 41%, p=0.044), more M1s (44% vs 29%, p<0.001). 93 pts (25%) had target profile, of whom 60% had EVT. Of 278 without target profile, 48% had EVT. Among all pts, excellent outcomes and mRS distribution were similar (EVT 63.5% vs MM 59.1%, aOR 1.55, 95%, p=0.16) and (adj cOR 1.44, p=0.16) Fig 1A. EVT had worse safety; sICH (6% vs 0%, p=0.002); neuro-worsening (19% vs 3%, p<0.001) and mortality (5% vs 1%, p=0.06). With target profile, EVT associated with more excellent outcomes (66% vs 49%, aOR 4.44, 95% CI 1.04-18.95, p=0.04), shift to better outcomes (adj cOR 2.9, 95% CI 1.03-7.91, p=0.04) Fig 1B. Safety was similar; sICH 2% vs 0%, p>0.99, neuro-worsening 17% vs 6%, p=0.30) and mortality 5% vs 3%, p>0.99). Without target profile, excellent outcomes were similar without a shift, Fig 1C. Safety was worse with EVT: sICH 8% vs 0%, p=0.001; neuro-worsening 20% vs 3%, p<0.001).
Conclusion:
EVT was not associated with improved outcomes in patients with mild deficits; safety was worse. However, EVT was safe and associated with improved outcomes in target profile patients.