Supplementary Material for: Spot Signs in Intracerebral Hemorrhage: Useful for Identifying Patients at Risk for Hematoma Enlargement?

Background: Prognostic signs for the identification of patients with acute spontaneous intracerebral hemorrhage (SICH) prone to hematoma expansion are limited. Contrast extravasation (spot signs, SpS) on computed tomographic angiography (CTA) may be a promising method to predict hematoma expansion in acute SICH. However, prospective data on the predictive value of the SpS on hematoma expansion and clinical outcome are still limited. We aimed to investigate associations between the presence of SpS, hematoma expansion, and clinical outcome in acute SICH. Methods: A prospective observational study was performed between 08/2008 and 08/2011. Patients with SICH presenting within 6 h of symptom onset were included. Patients with secondary hematomas, purely intraventricular hematomas, incomplete CT evaluation, hematoma evacuation prior to follow-up brain imaging, and incomplete follow-up data and those who refused to give consent for data analysis were excluded. CT and CTA brain imaging were carried out in all patients at baseline. After 24 h, follow-up brain imaging was performed. Hematoma location, hematoma volume, and substantial hematoma expansion were documented. CTA images were evaluated by two investigators for the presence of SpS. In all positive SpS cases, images were additionally reviewed by a third rater to achieve consensus for interpretating contrast extravasation. Clinical outcome was measured by the modified Rankin Scale (mRS) at discharge and at 3 months. Results: In total, 101 patients [median age 73 years (interquartile range 60-79); male 61.4%] were included in the analysis. Median time from onset to CTA was 128 min (interquartile range 90-209 min); median initial National Institute of Health Stroke Scale score was 16 (8-21). SpS were detected in 27 patients (26.7%). Cohen's kappa for the presence of SpS was 0.606, indicating moderate agreement. SpS patients had significantly higher initial hematoma volumes than patients without SpS (36.0 vs. 14.39 ml, p = 0.005). Hematoma expansion was significantly more frequent in SpS patients (59.3 vs. 21.6%, p < 0.001) and associated with the presence of SpS in the univariate analysis (OR 5.273; 95% CI 2.047-13.584, p = 0.001) and in multivariable analysis adjusted for the initial hematoma volume (OR 4.678, 95% CI 1.781-12.288, p = 0.002). Sensitivity of SpS to predict hematoma expansion was 0.5, specificity was 0.84. The positive likelihood ratio for SpS to predict hematoma expansion was 3.136 (95% CI 1.649-5.967), the negative likelihood ratio was 0.595 (95% CI 0.414-0.854). No difference in 3-month clinical outcome was observed between patients with and without SpS (median mRS score 4 and 4, p = 0.457). Conclusions: The clinical value of SpS needs to be further explored. Future studies should particularly focus on structured training procedures to identify SpS and measure the time needed to precisely assess the presence of SpS and on the prevalence of SpS in consecutive intracerebral hemorrhage populations.