Noninvasive quantification of cerebral blood flow using 99mTc-ECD and SPECT

The aim of this study was to develop a simple, noninvasive method for quantifying regional cerebral blood flow (rCBF) using 99mTc-ethyl cysteinate dimer (ECD) by a single SPECT scan and single venous sampling.

Using a three-compartment model, we introduced the regional brain fractionation index (BFI), Cb(Ts)/integral of 0-Ts Ca(tau)dtau [Ca(t), arterial input; Cb(t), brain activity]. Regional BFI obtained at the optimum time Ts (min) was converted to rCBF using an exponential function, which was obtained by analyzing the relationship between regional BFI and rCBF (= F) obtained by the standard 133Xe inhalation SPECT method. The integral of the concentration of 99mTc-ECD in arterial blood corrected for physical decay [Ca(t)] in BFI was estimated from a single venous blood sample obtained at the optimum time Tv using the regression line obtained by analyzing the relationship between the integral of Ca(t) and venous sample data. The data come from three groups of patients. The first group of patients (n = 16) underwent a complete 99mTc-ECD BFI study with measurement of Ca(t) and dynamic SPECT scanning, as well as a 133Xe inhalation study to measure rCBF The results were used to analyze the relationship between regional BFI and rCBF (obtained with 133Xe) and to determine the optimum time Ts for obtaining BFI. Data from the second group of patients (n = 15) were used to analyze the relationship between the integral of Ca(t) and venous sample data and to determine the optimum time Tv for one-point venous blood sampling. Finally, the third group of patients (8 patients, 10 studies) was used to validate the current method by comparing the results with 133Xe inhalation SPECT.

Regional BFI obtained at time Ts = 20 min showed good agreement (r = 0.907; a = 0.552, b = 0.962) with rCBF. The venous sample data obtained at time Tv = 6 min showed a good correlation (r = 0.988) with BFI. In comparing rCBF values thus obtained and those obtained by the 133Xe method, we found a good correlation (r = 0.917, slope = 1.01).

The proposed method has three advantages: (a) accurate quantification of rCBF without underestimation in the high flow range, (b) simplicity and noninvasiveness and (c) the ability to use any type of SPECT camera for the study.