Comparison of 4 methods for quantification of dopamine transporters by SPECT with [123I]IACFT.

Abstract

2Beta-carbomethoxy-3beta-(4-fluorophenyl)-n-(1-iodoprop-1-en -3-yl) nortropane (IACFT) is a highly selective ligand for dopamine transporter (DAT) sites in the striatum. Recent reports have described the basic kinetics, neurobiology, and imaging properties of [123I]IACFT. This report focuses on the structural (i.e., the ability to produce consistent binding estimates) validity of 4 methods to quantify striatal binding potential (BP) for IACFT. Seven healthy volunteers and 8 patients with Parkinson's disease were subjects for this study. Dynamic SPECT images and arterial blood samples were acquired during the 1.5-2 h after injection of 185-370 MBq [123I]IACFT. Plasma radioactivity was analyzed chromatographically to obtain metabolite-corrected arterial input functions. The k3/k4 ratio (BP) for striatal DAT sites was calculated by 4 methods. In the first method, tissue time-activity curves and metabolite-corrected arterial input functions were analyzed by a linear graphic method developed for reversible receptor ligands. The second method was also graphic; however, the occipital cortex time-activity curve was used as the input function. In the third method, the difference between the striatal and occipital cortex time-activity curves at secular equilibrium was taken to represent bound tracer, the occipital cortex time-activity curve was used to represent tracer in the free and nonspecifically bound state, and equilibrium receptor equations were used to determine BP. The fourth method used the occipital cortex time-activity curve to mathematically derive an input function for fitting the striatal time-activity curve and to determine BP. Analysis of the dynamic SPECT data by methods 1 and 2 resulted in highly linear plots (after approximately 15 min), supporting the reversibility of the tracer. A high linear correlation was found for BP determined by all 4 methods. ANOVA showed that methods 1-3 were indistinguishable; method 4 yielded lower BPs than did methods 1-3. These results show that BP can be estimated consistently using 4 different methods. This finding lends support to the modeling assumptions and provides methods suitable for clinical investigation.