Résumé
Background Whether perfusion 'mismatch' modifies the effect of endovascular thrombectomy (EVT) in patients with large core ischemic stroke remains uncertain.Methods We conducted a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-compliant systematic review and meta-analysis of randomized controlled trials (RCTs) enrolling patients with large infarcts and reporting outcomes by perfusion mismatch status, defined as a mismatch ratio >= 1.2 and a mismatch volume >= 10 mL. A Mantel-Haenszel fixed-effect model was used to pool ORs for three 90-day outcomes: independent ambulation (modified Rankin Scale (mRS) score 0-3, primary outcome), good outcome (mRS score 0-2), severe disability or death (mRS score 5-6). We tested for interaction to assess whether the effect of thrombectomy differed according to mismatch status.Results Three RCTs met the inclusion criteria, totaling 863 patients, of whom 113 (13%) had no perfusion mismatch (47% from LASTE). There was no significant interaction between mismatch status and EVT efficacy across any of the outcomes. In the mismatch subgroup, EVT significantly improved outcomes compared with medical therapy (mRS 0-3: OR=2.14, 95% CI 1.57 to 2.90; mRS score 0-2: OR=3.22, 95% CI 2.14 to 4.85; mRS score 5-6: OR=0.66, 95% CI 0.50 to 0.89; all I-2 = 0%). In the no-mismatch subgroup, OR values were not significant for both mRS score 0-3 (OR=2.32, 95% CI 0.84 to 6.38) and mRS score 5-6 (OR=0.73, 95% CI 0.34 to 1.57); OR for mRS score 0-2 was neutral (OR=1.00, 95% CI 0.15 to 6.55).Conclusions No interaction was found between mismatch status and EVT effect. However, in patients without mismatch, effect estimates were non-significant for EVT. Additional data are needed to better assess treatment benefit in this subgroup.