Abstract WMP99: Hematoma Retraction in Spontaneous Intracerebral Hemorrhage: Frequency, Functional Significance and Relation to Perihematomal Edema

Abstract

Background: The natural history of spontaneous intracerebral hematomas (ICH) is for variable degree of early hematoma retraction (HR), which may determine perihematomal brain edema (PHE), but such inter-relationships have not been rigorously studied. Methods: A pooled analysis of 760 computed tomography (CT) scans within 24±12 hours and 1197 CT scans within 72±12 hours of diagnostic CT (0-6 hours from symptom onset), from non-surgical ICH patients included in the Virtual International Stroke Trials Archive (VISTA), and INTERACT I, CLEAR III and MISTIE II clinical trials. Multivariable logistic regression was used to assess independent associations of HR and good 90-day functional outcome (modified Rankin scale [mRS] <3). Results: Median (IQR) diagnostic ICH volume was 11.4 (6.1-22.8) mL. HR occurred at defined thresholds in 5.8% (3mL), 2.4% (6mL) and 2.8% (33%), and in 10.3%, 4.0% and 4.6%, at 24 and 72 hours, respectively. Median retraction volumes at 24/72 hours were -0.9[-2.2, -0.3] and -1.2[-2.9, -0.5) mL, respectively. Between 13.2% and 24% of patients met common hematoma expansion definitions. Clot retraction volume at 24/72 hours was significantly correlated with diagnostic ICH volume (Spearman’s rho: -0.65/-0.64; p<0.001), ICH score, baseline NIHSS and higher baseline serum glucose, but inversely correlated with 24/72 hour PHE volume after adjustment for ICH volume (p=0.02; p=0.006, respectively). There was a negative correlation between clot retraction and good outcome; after adjustment for age, admission Glasgow coma scale (GCS), ICH and IVH volumes: each 1 mL smaller clot retraction at 24 hours was associated with almost 2.5-fold increase in odds of good outcome (95%CI:1.15- to 5.24) for clot retractions between 2 and 5 mL. There was no significant association with good outcomes for clot retractions less than 2 mL, after adjustment for severity indicators. No patient with clot retraction >5mL had good outcome (N = 20). Conclusion: HR is uncommon and continues to increase over first 72 hours after ICH. Greater HR is associated with more severe ICH, but with less PHE volume, and predicts a poor outcome. Further study is required to better understand these findings.