Introduction
Compared to the PET tracer [11C]DPN, the F-18-labelled homologue 6-O-(2-[18F]fluoroethyl)-6-O-desmethyldiprenorphine ([18F]FDPN) has a longer half-life while maintaining similar pharmacologic properties. Thus, [18F]FDPN allows for more experimental flexibility and applications at centers without an on-site cyclotron. Here, we wished to determine the ideal duration of [18F]FDPN dynamic PET studies.
Methods
[18F]FDPN PET scans were acquired in eight healthy male volunteers (mean age 36.8 years, range 33–40 years) using a Siemens/CTI ECAT EXACT HR+ scanner (Knoxville, TN, USA) in 3D mode. Images were acquired over 120 minutes with arterial blood and metabolite sampling. Each subject's dynamic dataset was realigned and normalized to a ligand-specific template in SPM2 (Wellcome Dept. of Cognitive Neurology, London, UK). Binding kinetics were quantified under protocols of different lengths in different VOIs defined using the MARSBAR toolbox (marsbar.sourceforge.net). We analyzed amygdala, caudate, putamen, thalamus, cingulate, frontal, and occipital VOIs. VOI extraction and kinetic analyses were performed in PMOD Medical Imaging Program, version 2.5 (PMOD Group, Zurich Switzerland). To evaluate the quantification of [18F]FDPN binding kinetics under protocols of different lengths (40, 60, 80, 90, 100, and 120 minutes), a 1 tissue or 2 tissue compartment model was fit to the VOI data with a measurement error variance defined to be equal to the average VOI concentration divided by the frame length.
Results
The VOI data in the group of eight volunteers indicate that compared to the full 120 minute acquisition, DV values can be estimated within a 10% error range when reducing the acquisition time to 90 minutes. While larger VOIs show around a 10% bias at 60 minutes, smaller VOIs show enhanced bias of up to approximately 18%. The correlation analyses of the DV estimates between the full and the shortened acquisition protocols indicate that a good replication of DV values, both at low and high binding, is attained with shortened schedules of 100 and 90 minutes. At 80 minutes the correlation is characterized by more variability and a trend towards increased deviation in high binding regions compared with low binding regions. This trend is further augmented in the 60 and 40 minute protocols, which results in a substantial underestimation of the DV in the 40 minute protocol.
Conclusions
The evaluation of [18F]FDPN distribution volumes under protocols of different lengths shows that a 90 minute protocol results in less then 10% bias compared to the full length protocol in all regions analyzed. This acquisition time represents a reasonable cut-off between enhanced parameter precision with longer sampling and increased noise and uncertainty with shorter sampling. However, if an experimentalist is interested in adopting an even shorter protocol, the increased bias in DV must be weighted against the benefits of a shorter protocol.