Comparison of liver perfusion parameters studied with conventional extravascular and experimental intravascular CT contrast agents

Rationale and Objectives To compare liver perfusion parameters obtained by using an extravascular contrast agent and a blood-pool agent. Materials and Methods Fifteen rabbits were imaged with a continuous 40-second single-slice computed tomography acquisition after a bolus injection of contrast agent (physiologic bolus duration 4–5 seconds, extravascular iohexol, n = 7; experimental nanoparticulated blood-pool agent WIN8883, n = 8). Time-density curves were generated for the aorta, portal vein, and liver. From the curves, arterial, portal, and total blood flows and hepatic perfusion index (HPI, arterial-to-total perfusion ratio) were determined by using two commonly applied fundamentally different analyzing methods: the single-compartment model and the peak gradient (PG) method. Also, the gamma variate fitting method was used. Results By using the single-compartment model, the obtained HPI and total blood flow were 0.14 ± 0.04 and 2.29 ± 0.40 (mL/min/mLtissue) for WIN8883, and 0.15 ± 0.06 (P = .54) and 4.60 ± 1.14 (mL/min/mLtissue) (P = .0002) for iohexol, respectively. With the PG, HPI and total blood flow were 0.15 ± 0.08 and 1.27 ± 0.24 (mL/min/mLtissue) for WIN8883, and 0.20 ± 0.06 (P = .12) and 2.11 ± 0.25 (mL/min/mLtissue) (P = .00002) for iohexol, respectively. With the blood pool agent, similar contrast enhancement to the conventional agent was achieved with about 36% reduced dosage of iodine per body weight (mg I/kg). Conclusions HPI was found to be quite insensitive to different contrast agent types and analyzing methods. However, the arterial, portal and total liver blood flow values strongly depend on contrast agent type and modeling method.