Prediction of Cerebrovascular Reserve Capacity by Computed Tomography Perfusion Using 320-Row Computed Tomography

Abstract

Acetazolamide loading has been the "gold standard" for evaluating cerebrovascular reserve capacity (CVRC). However, life-threatening side effects of acetazolamide have recently been reported. The aim of the study was to identify alternative methods for evaluating CVRC. We reviewed 6 patients who underwent both computed tomography perfusion (CTP) imaging and xenon CT (XeCT) imaging with and without acetazolamide loading during the same periods. The data were obtained as volume data using 320-row CT and applied to the automated region of interest-determining software and converted to standardized images. Correlations between CVRC and CTP parameters were analyzed by Pearson correlation coefficient analysis, and simple regression was used to assess the relationship between the data. When statistically significant, correlation between CVRC and any CTP data is identified, and cutoff points for CVRC 30% and 10% were calculated with receiver operating characteristic curves. Of 4 CTP parameters evaluated, statistically significant correlations were observed between time to peak (TTP) by CTP and CVRC (P < .0001, r = -.7228) calculated from XeCT. The regression line using CVRC as outcome variable (y) and using TTP as predictor variable (x) was y = -9.062x + 140.1. The cutoff value for the TTP for CVRC less than 10% was 12.56 seconds (sensitivity of 86% and specificity of 85%) and that for CVRC less than 30% was 9.34 seconds (sensitivity of 77% and specificity of 96%). TTP calculated from CTP data correlated well with the CVRC calculated from XeCT data. These results suggest that TTP calculated from CTP could be used to estimate CVRC in patients with occlusive cardiovascular disease.