Quantification of dopamine transporter density with [18F]FECNT PET in healthy humans.

INTRODUCTION

Fluorine-18 labeled 2β-carbomethoxy-3β-(4-chlorophenyl)-8-(2-fluoroethyl)nortropane ([(18)F]FECNT) binds reversibly to the dopamine transporter (DAT) with high selectivity. [(18)F]FECNT has been used extensively in the quantification of DAT occupancy in non-human primate brain and can distinguish between Parkinson's and healthy controls in humans. The purpose of this work was to develop a compartment model to characterize the kinetics of [(18)F]FECNT for quantification of DAT density in healthy human brain.

METHODS

Twelve healthy volunteers underwent 180 min dynamic [(18)F]FECNT PET imaging including sampling of arterial blood. Regional time-activity curves were extracted from the caudate, putamen and midbrain including a reference region placed in the cerebellum. Binding potential, BPND, was calculated for all regions using kinetic parameters estimated from compartmental and Logan graphical model fits to the time-activity data. Simulations were performed to determine whether the compartment model could reliably fit time-activity data over a range of BPND values.

RESULTS

The kinetics of [(18)F]FECNT were well-described by the reversible 2-tissue arterial input and full reference tissue compartment models. Calculated binding potentials in the caudate, putamen and midbrain were in good agreement between the arterial input model, reference tissue model and the Logan graphical model. The distribution volume in the cerebellum did not reach a plateau over the duration of the study, which may be a result of non-specific binding in the cerebellum. Simulations that included non-specific binding show that the reference and arterial input models are able to estimate BPND for DAT densities well below that observed in normal volunteers.

CONCLUSION

The kinetics of [(18)F]FECNT in human brain are well-described by arterial input and reference tissue compartment models. Measured and simulated data show that BPND calculated with reference tissue model is proportional to BPND calculated from the arterial input model.