Dual-radiotracer translational SPECT neuroimaging. Comparison of three methods for the simultaneous brain imaging of D2/3 and 5-HT2A receptors.

PURPOSE: SPECT imaging with two radiotracers at the same time is feasible if two different radioisotopes are employed, given their distinct energy emission spectra. In the case of 123I and 125I, dual SPECT imaging is not straightforward: 123I emits photons at a principal energy emission spectrum of 143.1-179.9 keV. However, it also emits at a secondary energy spectrum (15-45 keV) that overlaps with the one of 125I and the resulting cross-talk of emissions impedes the accurate quantification of 125I. In this paper, we describe three different methods for the correction of this cross-talk and the simultaneous in vivo [123I]IBZM and [125I]R91150 imaging of D2/3 and 5-HT2A receptors in the rat brain. METHODS: Three methods were evaluated for the correction of the effect of cross-talk in a series of simultaneous, [123I]IBZM and [125I]R91150 in vivo and phantom SPECT scans. Method 1 employs a dual-energy window (DEW) approach, in which the cross-talk on 125I is considered a stable fraction of the energy emitted from 123I at the principal emission spectrum. The coefficient describing the relationship between the emission of 123I at the principal and the secondary spectrum was estimated from a series of single-radiotracer [123I]IBZM SPECT studies. In Method 2, spectral factor analysis (FA) is applied to separate the radioactivity from 123I and 125I on the basis of their distinct emission patterns across the energy spectrum. Method 3 uses a modified simplified reference tissue model (SRTMC) to describe the kinetics of [125I]R91150. It includes the coefficient describing the cross-talk on 125I from 123I in the model parameters. The results of the correction of cross-talk on [125I]R91150 binding potential (BPND) with each of the three methods, using cerebellum as the reference region, were validated against the results of a series of single-radiotracer [123I]R91150 SPECT studies. In addition, the DEW approach (Method 1), considered to be the most straightforward to apply of the three, was further applied in a dual-radiotracer SPECT study of the relationship between D2/3 and 5-HT2A receptor binding in the striatum, both at the voxel and at the regional level. RESULTS: Average regional BPND values of [125I]R91150, estimated on the cross-talk corrected dual-radiotracer SPECT studies provided satisfactory correlations with the BPND values for [123I]R91150 from single-radiotracer studies: r = 0.92, p < 0.001 for Method 1, r = 0.92, p < 0.001 for Method 2, r = 0.92, p < 0.001, for Method 3. The coefficient describing the ratio of the 123I-emitted radioactivity at the 125I-emission spectrum to the radioactivity that it emits at its principal emission spectrum was 0.34 in vivo. Dual-radiotracer in vivo SPECT studies corrected with Method 1 demonstrated a positive correlation between D2/3 and 5-HT2A receptor binding in the rat nucleus accumbens at the voxel level. At the VOI-level, a positive correlation was confirmed in the same region (r = 0.78, p < 0.01). CONCLUSION: Dual-radiotracer SPECT imaging using 123I and 125I-labeled radiotracers is feasible if the cross-talk of 123I on the 125I emission spectrum is properly corrected. The most straightforward approach is Method 1, in which a fraction (34%) of the radioactivity emitted from 123I at its principal energy spectrum is subtracted from the measured radioactivity at the spectrum of 125I. With this method, a positive correlation between the binding of [123I]IBZM and [125I]R91150 was demonstrated in the rat nucleus accumbens. This result highlights the interest of dual-radiotracer SPECT imaging to study multiple neurotransmitter systems at the same time and under the same biological conditions.

A single-scan protocol for absolute D2/3 receptor quantification with [123I]IBZM SPECT.

PURPOSE: Molecular imaging of the D2/3 receptor is widely used in neuropsychiatric research. Non-displaceable binding potential (BPND) is a very popular quantitative index, defined as the product of the receptor concentration (Bavail) and the radiotracer affinity for the receptor (1/appKd). As the appKd is influenced by parameters such as the endogenous neurotransmitter dynamics, it often constitutes a confounding factor in research studies. A simplified method for absolute quantification of both these parameters would be of great interest in this context. Here, we describe the use of a partial saturation protocol that permits to produce an in vivo Scatchard plot and thus estimate Bavail and appKd separately, through a single dynamic SPECT session. To validate this approach, a multi-injection protocol is used for the full kinetic modeling of [123I]IBZM using a 3-tissue compartment, 7-parameter model (3T-7k). Finally, more "classic" BPND estimation methods are also validated against the results of the 3T-7k. METHODS: Twenty-nine male rats were used. Binding parameters were estimated using the 3T-7k in a multi-injection protocol. A partial saturation protocol was applied at the region- and voxel-level and results were compared to those obtained with the 3T-7k model. The partial saturation protocol was applied after an adenovirus-mediated D2 receptor striatal overexpression and in an amphetamine-induced dopamine release paradigm. The Simplified Reference Tissue Model (SRTM), the Logan's non-invasive graphical analysis (LNIGA) and a simple standardized uptake ratio (SUR) method were equally applied. RESULTS: The partial saturation experiments gave similar values as the 3T-7k both at the regional and voxel-level. After adenoviral-mediated D2-receptor overexpression, an increase in Bavail by approximately 18% was observed in the striatum. After amphetamine administration, a 16.93% decrease in Bavail (p<0.05) and a 39.12% increase (p<0.01) in appKd was observed. BPND derived from SRTM, LNIGA and SUR correlated well with the Bavail values from the 3T-7k (r=0.84, r=0.84 and r=0.83, respectively, p<0.0001 for all correlations). CONCLUSION: A partial saturation protocol permits the non-invasive and time-efficient estimation of Bavail and appKd separately. Given the different biological phenomena that underlie these parameters, this method may be applied for the in-depth study of the dopaminergic system in translational molecular imaging studies. It can detect the biological variations in these parameters, dissociating the variations in receptor density (Bavail) from affinity (1/appKd), which reflects the interactions of the receptor with its endogenous ligand.