Abstract WP339: Herniation Despite Decompressive Hemicraniectomy in Large Hemispherical Ischemic Stroke Patients

Abstract

Background: Despite timely decompressive hemicraniectomy (DHC), clinical failure or progressive herniation requiring a second decompression procedure and mortality has been reported in acute middle cerebral artery (MCA) strokes. Hypothesis: Our objective was to determine the stroke and surgical parameters measured on head CT scan that are associated with progressive herniation despite DHC in patients with large hemispheric MCA strokes. Methods: Retrospective chart review of all medical records and imaging features of patients with malignant hemispheric infarction who underwent DHC for cerebral edema from July 2010 to June 2015, was performed. Patients who died from postoperative hemorrhagic complications were excluded. Infarct volume was calculated using ABC/2 method on CT scans within 48 hours of symptom onset (Kostov et al, 2012, World Neurosurg). Radiologic parameters of the craniectomy bone flap (length, width, area) and brain volume protruding out of the skull (height and volume) were measured (Chung et al, 2011, Neurologist). Images were reviewed by a board certified neuroradiologist to determine whether the craniectomy bed was sufficiently centered on the stroke bed (Zweckberger et al, 2014, Cerebrovasc Dis) and the brain volume not included in the craniectomy bed. Groups were compared using Fisher exact test for categorical variables and T-test or Mann-Whitney U test for continuous variables, as appropriate. Results: Out of 41 patients who underwent DHC for cerebral edema (mean age 53.1 ±12, 48.7% females, 36.5% African Americans) 7 had progressive herniation leading to mortality. Radiographic parameters that were significantly different between both groups were presence of malignant edema (p=0.047), insufficient centering of the craniectomy bed on the stroke bed (p=0.03), large infarct volume not centered on the craniectomy bed (p=0.011), presence of anterior cerebral artery infarction (p=0.047), and smaller craniectomy length (p=0.05). There was a trend in protruding brain volume (p=0.056). Conclusion: Besides the craniectomy length, sufficient centering of the craniectomy over the stroke bed may be required to prevent progressive herniation.