Quantitative analysis of scatter- and attenuation-compensated dynamic single-photon emission tomography for functional hepatic imaging with a receptor-binding radiopharmaceutical.

A new method for quantitative liver study was developed using the tracer technetium-99m diethylene triamine penta-acetic acid-galactosyl human serum albumin (99mTc-GSA), an analog ligand of the asialoglycoprotein receptor, which is a hepatocyte surface receptor specific for galactose-terminated glycoproteins. For quantitative dynamic single-photon emission tomographic (SPET) studies, attenuation compensation using transmission computed tomography (TCT) and the triple energy window (TEW) scatter compensation method were evaluated. As the TCT source, we used an uncollimated multi-tube source with the TEW scatter compensation method. To verify the accuracy of cross-calibrated SPET values as compared with measured radioactivities, we performed SPET of a cylindrical water pool phantom which contains seven hot rods filled with different concentrations of 99mTc activities, simulating the scan conditions in human studies. The results of the phantom studies showed good linearity and accuracy of the SPET values, with R2=0.993 and a regression line of y=0.941x+5.48. From the analysis of a kinetic model based on a one-compartment model, focussing on the initial stage of several minutes after 99mTc-GSA injection and taking the physiological expression presented in a three-compartment analysis into account, we introduced the Rutland equation (Patlak plot) in the 99mTc-GSA study by which the overall and regional effective hepatic blood flow (EHBF) and hepatic blood pool volume were determined. Preliminary clinical evaluations were performed for four normal male subjects (23-35 years of age) and one patient. Forty sequential 30-s dynamic SPET acquisitions were obtained for a period of 20 min following the intravenous injection of 99mTc-GSA with venous blood sampling at 10 min. After scatter compensation, the SPET images were reconstructed with attenuation compensation using an attenuation map obtained from TCT. The average normal value for the total EHBF was 468+/-83 ml/min and that for the hepatic blood pool volume, 777+/-123 ml. Functional images of the distribution of regional values of EHBF (ml/min/voxel) and hepatic blood pool volume (ml/voxel) were also generated corresponding to the original SPET images. The EHBF images showed regional liver function, higher in the right lobe than the left lobe in the normal cases, and the heptic blood pool volume images showed the distribution of intensified high values along major vascular structures. Receptor imaging with 99mTc-GSA using the Rutland method and dynamic SPET with scatter and attenuation compensation is an effective technique that allows the evaluation of total and regional hepatic functional parameters (EHBF, hepatic blood pool) in vivo.