Iodine-123-N-(3-iodopropene-2-yl)-2beta-carbomethoxy-3beta-( 4-chlorophenyl) tropane (123I-IPT) is a new dopamine transporter ligand that selectively binds the dopamine reuptake sites. Transporter concentrations have been known to decrease in Parkinson's disease patients. The purpose of this study was to evaluate the usefulness of IPT as an imaging agent for measuring changes in transporter concentrations in Parkinson's disease. METHODS IPT labeled with 6.78 +/- 0.67 mCi 123I was injected intravenously as a bolus into eight normal controls (mean age 41 +/- 12 yr) and 17 Parkinson's disease patients (mean age 55 +/- 9 yr). Dynamic SPECT scans of the brain were then performed for 5 min each over 120 min on a triple-headed gamma camera equipped with medium-energy collimators. Regions of interest were drawn on the middle set of the image at the level of the basal ganglia (BG) for each subject. Time-activity curves were generated for the left BG, right BG and occipital cortex (OCC). The empirical ratios between BG-OCC and OCC, which represent specific-to-nonspecific binding ratios, were computed at various time points. The statistical parameter k3/k4 was estimated by two methods: a variation of the graphic method that derives the ratio of ligand distribution volumes (R[V]) and the area ratio method (R[A]), in which the ratio is calculated from the areas under the specific and nonspecific binding activity curves. RESULTS The mean (BG-OCC)/OCC ratio for normal controls (3.07 +/- 0.73) was significantly higher than that for Parkinson's disease patients at 115 min (1.10 +/- 0.56) (p = 2.76 x 10[-5]). The mean R(V) and R(A) for normal controls were 2.06 +/- 0.27 and 1.50 +/- 0.15, respectively. The mean R(V) and R(A) for Parkinson's disease patients were 0.78 +/- 0.31 and 0.65 +/- 0.24, respectively. Both R(V) and R(A) for normal controls were significantly higher than those for Parkinson's disease patients (p values for R(V) and R(A) were 1.91 x 10(-8) and 3.46 x 10(-10), respectively). The R(V) has linear relationships with both R(A) and (BG-OCC)/OCC ratio at 115 min. The R(V) has a higher correlation (r = 0.99) with R(A) than it does with (BG-OCC)/OCC (r = 0.93). CONCLUSIONS The R(V), R(A) and (BG-OCC)/OCC for Parkinson's disease patients were clearly separated from those of normal controls, and they may be useful outcome measures for clinical diagnosis. The simplest (BG-OCC)/OCC ratio, requiring a single late time point, could be useful in clinical situations, whereas R(V) or R(A) is preferred when the dynamic data are available. The findings suggest that 123I-IPT is a useful tracer for diagnosing Parkinson's disease and studying dopamine reuptake sites.