Abstract 31: Ultraearly Hematoma Growth: Multicenter External Validation of the Adjustment of Intracerebral Hemorrhage Volume by Onset-to-imaging Time

Abstract

Background: The impact of baseline intracerebral hemorrhage (ICH) volume on hematoma growth (HG) and clinical outcome may vary widely depending on the onset-to-imaging time (OIT). We previously reported that the adjustment of initial ICH volume by OIT (coined as ultraearly HG, uHG) is a powerful tool for outcome prediction in acute ICH. We aimed to validate our previous findings in a multicenter external cohort and to assess the relationship between uHG and the CTA spot-sign. Methods: The PREDICT study was a prospective, observational cohort study of consecutive ICH patients <6 hours. Patients underwent baseline and 24-hour CT scans, and CTA for the blinded detection of spot sign. uHG was defined as the relation between baseline ICH volume/OIT, HG as hematoma expansion >33% or >6 mL, early neurologic deterioration (END) as increase ≥4 points in the NIHSS score or death at 24 hours, and poor outcome as mRS score >2 at 3 months. Results: Two hundred and thirty-seven patients were included in this study. Median baseline ICH volume was 14.5 (6[[Unable to Display Character: –]]30.4) mL, median OIT 135 (85.5[[Unable to Display Character: –]]199) minutes, and median uHG 6.5 (2.5-14.3) mL/h. The spot sign was present in 31.2% of patients. uHG was 2.7-fold higher in spot sign-positive patients (11.1 [5.7-17.7] mL/h vs. 4.1 [1.9-12.3] mL/h, P<0.001). uHG >4.7 mL/h improved the sensitivity-specificity of both baseline ICH volume >10 mL and spot sign in the prediction of HG (73.9%-57.5% vs. 68.1%-54.3% and 48.9%-73.9%), 90-day mortality (82%-56.3% vs. 78%-53.5% and 45.8%-70.8%), and poor outcome (72.9%-80.4% vs. 69.8%-70.6% and 38%-75%), respectively. The median hematoma expansion at 24 hours among spot-positive patients was 3.2 mL in uHG <5 mL/h group, 4.1 mL in uHG 5-10 mL/h group, and 4.8 mL in uHG >10 mL/h group (P<0.001). In adjusted multivariate analyses uHG independently predicted HG (OR 1.08, 95% CI 1.04-1.12), END (OR 1.06, 95% CI 1.02-1.10), 90-day mortality (OR 1.07, 95% CI 1.02-1.11), and poor outcome (OR 1.12, 95% CI 1.02-1.23). Conclusions: These results validate uHG as a powerful predictor of outcome in acute ICH. uHG is significantly higher in spot sign patients, improves the accuracy of baseline ICH volume and spot sign in the prediction of HG and clinical outcome, and independently predicts HG, END, 90-day mortality, and poor outcome.