Hemicraniectomy for Dominant vs Nondominant Middle Cerebral Artery Infarction : A Systematic Review and Meta- Analysis

Objectives: Decompressive hemicraniectomy decreases mortality and severe disabil-ity from space-occupying middle cerebral artery infarction in selected patients. However, attitudes towards hemicraniectomy for dominant-hemispheric stroke have been hesitant. This systematic review and meta-analysis examines the associa-tion of stroke laterality with outcome after hemicraniectomy. Materials and methods: We performed a systematic literature search up to 6th February 2020 to retrieve original articles about hemicraniectomy for space-occupying middle cere-bral artery infarction that reported outcome in relation to laterality. The primary outcome was severe disability (modified Rankin Scale 4-6 or 5-6 or Glasgow Out -come Scale 1-3) or death. A two-stage combined individual patient and aggregate data meta-analysis evaluated the association between dominant-lateralized stroke and (a) short-term (< 3 months) and (b) long-term (> 3 months) outcome. We per -formed sensitivity analyses excluding studies with sheer mortality outcome, sec -ond-look strokectomy, low quality, or small sample size, and comparing populations from North America/Europe vs Asia/South America. Results: The analysis included 51 studies (46 observational studies, one nonrandomized trial, and four randomized controlled trials) comprising 2361 patients. We found no asso-ciation between dominant laterality and unfavorable short-term (OR 1.00, 95% CI 0.69-1.45) or long-term (OR 1.01, 95% CI 0.76-1.33) outcome. The results were unchanged in all sensitivity analyses. The grade of evidence was very low for short -term and low for long-term outcome. Conclusions: This meta-analysis suggests that patients with dominant-hemispheric stroke have equal outcome after hemicraniec-tomy compared to patients with nondominant stroke. Despite the shortcomings of the available evidence, our results do not support withholding hemicraniectomy based on stroke laterality.