Modeling and analysis of PET studies with norepinephrine transporter ligands: the search for a reference region

The development of positron emission tomography (PET) ligands for the norepinephrine transporter (NET) has been slow compared to the development of radiotracers for others systems, such as the dopamine (DAT) or the serotonin transporters (SERT). The main reason for this appears to be the high nonspecific (non-NET) binding exhibited by many of these tracers, which makes the identification of a reference region difficult. With other PET ligands the use of a reference region increases the reproducibility of the outcome measure in test/retest studies. The focus of this work is to identify a suitable reference region or means of normalizing data for the NET ligands investigated.

We have analyzed the results of PET studies in the baboon brain with labeled reboxetine derivatives (S,S)-[(11)C]O-methyl reboxetine (SS-MRB), (S,S)-[(18)F]fluororeboxetine (SS-FRB) as well as O-[(11)C]nisoxetine and N-[(11)C]nisoxetine (NIS), and, for comparison, the less active (R,R) enantiomers (RR-MRB, RR-FRB) in terms of the distribution volume (DV) using measured arterial input functions.

(1) For a given subject, a large variation in DV for successive baseline studies was observed in regions with both high and low NET density. (2) The occipital cortex and the basal ganglia were found to be the regions with the smallest change between baseline (SS-MRB) and pretreatment with cocaine, and were therefore used as a composite reference region for calculation of a distribution volume ratio (DVR). (3) The variability [as measured by the coefficient of variation (CV) = standard deviation/mean] in the distribution volume ratio (DVR) of thalamus (to reference region) was considerably reduced over that of the DV using this composite reference region. (4) Pretreatment with nisoxetine (1.0 mg/kg 10 min prior to tracer) in one study produced (in decreasing order) reductions in thalamus, cerebellum, cingulate and frontal cortex consistent with known NET densities. (5) [(11)C]Nisoxetine had a higher background non-NET binding (DV) than the other tracers reported here with basal ganglia (a non-NET region) higher than thalamus.

The reboxetine derivatives show a lot of promise as tracers for human PET studies of the norepinephrine system. We have identified a strategy for normalizing DVs to a reference region with the understanding that the DVR for these tracers may not be related to the binding potential in the same way as, for example, for the dopamine tracers, since the non-NET binding may differ between the target and nontarget regions. From our baboon studies the average DVR for thalamus (n = 18) for SS-MRB is 1.8; however, the lower limit is most likely less than 1 due to this difference in non-NET binding.