Human [123I]5-I-A-85380 dynamic SPECT studies in normals: kinetic analysis and parametric imaging

[¹²³I]5-I-A-85380 ([¹²³I]5-IA) is a ligand specific for the α4β2 nicotinic acetylcholine receptor (nAChR) subtype. It was reported that distribution volume (DV) is an appropriate measurement for α4β2 nAChR density in baboon [¹²³I]5-IA dynamic SPECT studies. To evaluate the kinetics obtained from human [¹²³I]5-IA dynamic SPECT, three different methods were applied to region of interest (ROI) kinetics analysis as follows: A: a 2-compartment 3-parameters model, B: 3-compartment 5-parameter model, and C: same model as in B but fitting with a constraint of ratio of K₁/k₂e estimated by fitting occipital kinetics to model in A. The DV was estimated as DV= K₁/k₂ for A, and DV=(K₁/k₂)(1+k₃/k₄) for B and C. Nonlinear least squares fitting using Marquardt algorithm was performed for A-C. Akaike information criterion (AIC) and R² were used for evaluation of fitting results. A linear parametric imaging algorithm derived from model in method A was used for generating images of K₁ and DV. Four normal human [¹²³I]5-IA dynamic studies were evaluated by A-C. In each study, after a bolus of [¹²³I]5-IA (5∼10 mCi, specific activity>10 Ci/umol) was injected intravenously, a dynamic scanning with 20 acquisitions over 6 hour was started immediately on a Trionix TriadXLT scanner. 30-40 arterial plasma samples were taken during the study and HPLC was performed to determine the metabolite-corrected arterial plasma radioactivity as input function. The ROI regions (cerebellum, frontal cortex, occipital cortex, pones, and thalamus) were drawn on co-registered MRI images. Results show that A fits to all ROI kinetics very well (R²>0.99), and paired T-test for AIC showed there is no statistical significantly improvement of model fitting from B, C methods. The K₁ and DV images are of good image quality. Conclusion: A is best for ROI kinetic analysis and parametric imaging for DV and K₁.