Intracerebral Hematoma Morphologic Appearance on Noncontrast Computed Tomography Predicts Significant Hematoma Expansion.

BACKGROUND AND PURPOSE Hematoma expansion in intracerebral hemorrhage is associated with higher morbidity and mortality. The computed tomography (CT) angiographic spot sign is highly predictive of expansion, but other morphological features of intracerebral hemorrhage such as fluid levels, density heterogeneity, and margin irregularity may also predict expansion, particularly in centres where CT angiography is not readily available. METHODS Baseline noncontrast CT scans from patients enrolled in the Predicting Hematoma Growth and Outcome in Intracerebral Hemorrhage Using Contrast Bolus CT (PREDICT) study were assessed for the presence of fluid levels and degree of density heterogeneity and margin irregularity using previously validated scales. Presence and grade of these metrics were correlated with the presence of hematoma expansion as defined by the PREDICT study on 24-hour follow-up scan. RESULTS Three hundred eleven patients were included in the analysis. The presence of fluid levels and increasing heterogeneity and irregularity were associated with 24-hour hematoma expansion (P=0.021, 0.003 and 0.049, respectively) as well as increases in absolute hematoma size. Fluid levels had the highest positive predictive value (50%; 28%-71%), whereas margin irregularity had the highest negative predictive value (78%; 71%-85). Noncontrast metrics had comparable predictive values as spot sign for expansion when controlled for vitamin K, antiplatelet use, and baseline National Institutes of Health Stroke Scale, although in a combined area under the receiver-operating characteristic curve model, spot sign remained the most predictive. CONCLUSIONS Fluid levels, density heterogeneity, and margin irregularity on noncontrast CT are associated with hematoma expansion at 24 hours. These markers may assist in prediction of outcomes in scenarios where CT angiography is not readily available and may be of future help in refining the predictive value of the CT angiography spot sign.