Aim: The purpose of this study was to develop a minimally invasive procedure to derive an arterial input function (AIF) in rats through tail artery blood sampling for pharmacokinetic modeling in preclinical PET molecular imaging studies. The procedure involved a microvolumetric blood counter (μBC) and a correction to compensate for delay and dispersion of the automatic blood sampling.
Materials and methods: AIFs were simultaneously obtained from femoral and tail arteries in rats, manually and using a μBC, after 18F-FDG injection (n=6) in order to compare the shape of the AIFs and the kinetic analysis results at equilibrium and after implementation of a dispersion correction method. These AIFs were used to estimate the myocardial metabolic rate of glucose (MMRG). AIFs were also obtained from a single withdrawal site by three methods to confirm accurate MMRG values: manual tail artery (n=5), μBC tail artery (n=5), and μBC femoral artery (n=3).
Results: Simultaneous withdrawal at equilibrium results in similar AIF shapes and influx rate constants (Ki) from Patlak analysis (P>0.05). Manually withdrawn and dispersion-corrected μBC AIFs in the simultaneous experiment did not reveal statistically different shapes and constants (K1, Ki) from a three-compartment kinetic analysis, regardless of the withdrawal methods or sites (P>0.05). Kinetic analysis of the three single-site blood sampling methods yielded similar MMRG (one-way ANOVA; Patlak, P=0.52; three-compartment, P=0.10).
Conclusion: Both minimally invasive manual withdrawal and dispersion-corrected μBC-based blood sampling in the tail artery are reliable methods for deriving AIFs for pharmacokinetic follow-up studies in the same animal.