Comparison of CT Perfusion and Digital Subtraction Angiography in the Evaluation of Delayed Cerebral Ischemia

Abstract

Delayed cerebral ischemia (DCI) is a devastating condition that occurs secondary to aneurysmal subarachnoid hemorrhage (A-SAH). The purpose is to compare computed tomography perfusion (CTP) and digital subtraction angiography (DSA) for determining DCI in A-SAH. A retrospective study of A-SAH patients admitted at our institution between December 2004 and December 2008 was performed. CTP and DSA were obtained at days 6-8 after aneurysm rupture. Both qualitative and quantitative analyses of CT perfusion deficits were performed. DSA was categorized as presence or absence of vasospasm. The reference standard for determining DCI was based on clinical deterioration or infarction on CT or MRI. The test characteristics of CTP and DSA were calculated and their graphs of conditional probabilities were constructed using Bayesian analysis. Fifty-seven patients were included; 79% (45/57) had DCI. Seventy percent (40/57) had CTP perfusion deficits; 80% (36/45) of the DCI and 33% (4/12) of no DCI patients. Sixty-three percent (36/57) had DSA demonstrating vasospasm; 73% (33/45) of the DCI and 25% (3/12) of no DCI patients. Quantitative analysis of the CTP data revealed a significant difference in cerebral blood flow values for the DCI (29.4 mL/100 g/minute) and no DCI groups (40.5 mL/100 g/minute, P = .0213). The sensitivity, specificity, and positive and negative predictive values for CTP were 0.80 (95% CI 0.68-0.92), 0.67 (95% CI 0.40-0.93), 0.90 (95% CI 0.82-0.96), 0.47 (95% CI 0.27-0.62), and for DSA were 0.73 (95% CI 0.60-0.86), 0.75 (95% CI 0.50-0.99), 0.92 (95% CI 0.82-0.98), and 0.43 (95% CI 0.26-0.53), respectively. CTP and DSA have similar test characteristics and Bayesian analysis for determining DCI in A-SAH patients.