Tracer input for kinetic modelling of liver physiology determined without sampling portal venous blood in pigs.

Abstract

Quantification of hepatic tracer kinetics by PET requires measurement of tracer input from the hepatic artery (HA) and portal vein (PV). We wished to develop a method for estimating dual tracer input without the necessity to sample PV blood. Pigs weighing 40 kg were given bolus doses of C(15)O (CO), 2-[(18)F]fluoro-2-deoxy-D-glucose (FDG), [(11)C]-methylglucose (MG), 2-[(18)F]fluoro-2-deoxy-D-galactose (FDGal) or H(2)(15)O(H(2)O). Tracer concentration 3-min time courses were measured in the femoral artery and PV by blood sampling. Blood flow was measured in the HA and PV using flow-meters. A model for transfer of tracer through the splanchnic circulation was used to estimate values of a tracer-specific model parameter β. Tracer-specific mean values of β were used to estimate tracer concentration time courses in the PV from the measured arterial concentration. A model-derived dual-input was calculated using the mean HA flow fraction (0.25) and validated by comparison of the use of the measured dual-input and a kinetic model with a fixed "true" K(1)(true), i.e. clearance of tracer from blood to liver cells. The rank order of the means of β was CO < FDG ≈ MG < FDGal < H(2)O, reflecting their different splanchnic mean transit times. Estimated K(1)(est) was not significantly different from "true" K(1)(true). The hepatic dual tracer input, which is of great importance for the assessment of processes such as transfer across the plasma-hepatocyte membrane or hepatic blood perfusion, can be well approximated in pigs without the necessity to sample PV blood and measure hepatic blood flow; only arterial blood sampling is needed.