Serotonin 2A receptor agonist binding in the human brain with [¹¹C]Cimbi-36

Abstract

[(11)C]Cimbi-36 was recently developed as a selective serotonin 2A (5-HT(2A)) receptor agonist radioligand for positron emission tomography (PET) brain imaging. Such an agonist PET radioligand may provide a novel, and more functional, measure of the serotonergic system and agonist binding is more likely than antagonist binding to reflect 5-HT levels in vivo. Here, we show data from a first-in-human clinical trial with [(11)C]Cimbi-36. In 29 healthy volunteers, we found high brain uptake and distribution according to 5-HT(2A) receptors with [(11)C]Cimbi-36 PET. The two-tissue compartment model using arterial input measurements provided the most optimal quantification of cerebral [(11)C]Cimbi-36 binding. Reference tissue modeling was feasible as it induced a negative but predictable bias in [(11)C]Cimbi-36 PET outcome measures. In five subjects, pretreatment with the 5-HT(2A) receptor antagonist ketanserin before a second PET scan significantly decreased [(11)C]Cimbi-36 binding in all cortical regions with no effects in cerebellum. These results confirm that [(11)C]Cimbi-36 binding is selective for 5-HT(2A) receptors in the cerebral cortex and that cerebellum is an appropriate reference tissue for quantification of 5-HT(2A) receptors in the human brain. Thus, we here describe [(11)C]Cimbi-36 as the first agonist PET radioligand to successfully image and quantify 5-HT(2A) receptors in the human brain.

Figure 1

Representative [¹¹C]Cimbi-36 positron emission tomography (PET) brain images from a healthy volunteer at baseline and after ketanserin pretreatment. PET images show radioligand uptake from 40 to 120 minutes and scaled to standardized uptake values (SUVs; unit: g/mL), and images are superimposed on a magnetic resonance imaging (MRI) template of the brain.

Figure 2

Representative [¹¹C]Cimbi-36 time-activity curves in plasma and brain tissue from a healthy volunteer. Radioactive concentrations in tissue and plasma are normalized to injected doses per kg bodyweight to attain standardized uptake values (SUVs). Brain tissue time-activity curves are fitted with one- and two-compartment models (1TCM and 2TCM).

Figure 3

Radiometabolism of [¹¹C]Cimbi-36. (A) Time course of the percentage [¹¹C]Cimbi-36 parent compound fraction and radiometabolites measured in human arterial plasma. Points represent mean±s.e.m. of 29 subjects, and the solid black indicates a two-exponential fit to the mean. (B) Representative radiochromatogram of human plasma sample taken 20 minutes after [¹¹C]Cimbi-36 injection.

Figure 4

Correlations between outcome measures V_T (A, B) and nondisplaceable binding potential (BP_ND) (C, D) quantified with linear (B, D) and nonlinear (A, C, and E) models. Reference outcome measure was obtained with the two-tissue compartment model (2TCM) fitting five parameters (K₁, k₂, k₃, k₄, V_b). In V_T estimations, eight regions are analyzed from each of the 29 subjects: striatum, thalamus, frontal cortex, parietal cortex, somatosensory cortex, occipital cortex, temporal cortex, and cerebellum. Cerebellum was used as a reference region in BP_ND estimations, and with 2TCM, BP_ND was calculated as (V_T^target−V_T^cerebellum)/V_T^cerebellum. Solid black lines show linear regressions and the regression equations are displayed in each plot. Dotted gray lines represent the line of identity. GA, graphical analyses; MRTM0, multi-linear reference tissue model; NIGA, noninvasive GA; SRTM, simplified reference tissue model.

Figure 5

[¹¹C]Cimbi-36 V_T derived from the two-tissue compartment model (2TCM) at baseline (n=5) and after per oral ketanserin pretreatment (10 mg, n=2 or 20 mg, n=3). ****P<0.0001 refers to comparison of V_T in a two-way ANOVA with Bonferroni-corrected multiple comparisons test.

Figure 6

Individual occupancy plots of the effect of ketanserin on [¹¹C]Cimbi-36 binding in the human brain.