Comparison of graphical analysis methods for VT estimation of a novel kappa opioid receptor tracer in humans

1184 Objectives To assess the utility of the relative equilibrium-based (RE) graphical analysis method [1] and RE plot with the Gjedde-Patlak correction (RE-GP) for estimation of total distribution volume (VT) for the novel kappa-opioid receptor antagonist tracer, [11C]LY2795050[3] because RE [1] developed VT image with high signal-to-noise. Methods Dynamic [11C]LY2795050 PET scans on the HRRT were acquired for 90 min in 4 healthy human subjects. Arterial blood samples were obtained and plasma input functions were constructed based on HPLC-determined parent fraction. Brain regional time-activity curves were fitted using the two-tissue compartment model (2TCM). Pixel-by-pixel and ROI based graphical methods were used for calculation of VT according to Zhou et al [1, 2]. Results RE plot results were highly dependent on the choice of relative equilibrium time (t*). Regional VT estimates from the RE plot increased with t*. Throughout the scan duration, the ratio of tissue to plasma concentration continuously increased in high binding regions (amygdala and anterior cingulate). In addition, the apparent t* varied by region and subject. By comparison, the RE-GP plot showed consistent and reliable VT estimates if time points with very low plasma activity concentration (Cp) were avoided. The correlation with 2TCM was high (y=0.979x+0.047, r2=0.98). VT images by the modified RE-GP method had high signal-to-noise. Conclusions The RE plot does not appear to be an appropriate method for calculating VT estimates of the new kappa-opioid antagonist ligand, [11C]LY2795050. This is probably due to the inability of the tracer to reach transient equilibrium within 90 min in all regions. The RE-GP plot appears to be valid for [11C]LY2795050 because the VT(RE-GP) method is robust to the choise of t* and because the VT(RE-GP) values are highly correlated with VT(2TCM)