Quantification of indirect pathway inhibition by the adenosine A2a antagonist SYN115 in Parkinson disease

Abstract

Adenosine A(2a) receptor antagonists reduce symptom severity in Parkinson disease (PD) and animal models. Rodent studies support the hypothesis that A(2a) antagonists produce this benefit by reducing the inhibitory output of the basal ganglia indirect pathway. One way to test this hypothesis in humans is to quantify regional pharmacodynamic responses with cerebral blood flow (CBF) imaging. That approach has also been proposed as a tool to accelerate pharmaceutical dose finding, but has not yet been applied in humans to drugs in development. We successfully addressed both these aims with a perfusion magnetic resonance imaging (MRI) study of the novel adenosine A(2a) antagonist SYN115. During a randomized, double-blind, placebo-controlled, crossover study in 21 PD patients on levodopa but no agonists, we acquired pulsed arterial spin labeling MRI at the end of each treatment period. SYN115 produced a highly significant decrease in thalamic CBF, consistent with reduced pallidothalamic inhibition via the indirect pathway. Similar decreases occurred in cortical regions whose activity decreases with increased alertness and externally focused attention, consistent with decreased self-reported sleepiness on SYN115. Remarkably, we also derived quantitative pharmacodynamic parameters from the CBF responses to SYN115. These results suggested that the doses tested were on the low end of the effective dose range, consistent with clinical data reported separately. We conclude that (1) SYN115 enters the brain and exerts dose-dependent regional effects, (2) the most prominent of these effects is consistent with deactivation of the indirect pathway as predicted by preclinical studies; and (3) perfusion MRI can provide rapid, quantitative, clinically relevant dose-finding information for pharmaceutical development.

Figures

Figure 1.

Study design (see Materials and Methods, Study Protocol for important details). a, Top, Schematic showing the temporal relationship of each subject's scan days to his or her taking active study drug and placebo. b, Bottom, Schematic showing each subject's participation on each of his or her two scan days. LD, Levodopa.

Figure 2.

a, Statistical image (color) representing the |t| values from the contrast (rCBF on SYN115 60 mg, b.i.d.) < (rCBF on placebo), laid over the study-specific MRI template image (grayscale). The right side of the brain is displayed on the right side of the coronal image and the bottom of the axial image. The color t image shows all voxels for which |t| ≥ 4.03 (corresponding to p < 0.001, uncorrected). The orthogonal hairlines in each cross section identify the location of the other two cross sections and cross at the voxel in right thalamus with the highest |t| in the whole image (19.51; 11 df). b, Dose-dependent decrease in rCBF with SYN115 in the right and left thalamus VOIs defined as described in Materials and Methods. c, d, The same image and conventions described in a, with different cross sections chosen to display other peak |t| values. c, Hairlines cross at the peak |t| value from cluster 2. d, Hairlines cross at the local peak that defined VOI 3 from cluster 1 (in precuneus). For coordinates and anatomical descriptions, see Table 1 and supplemental Table 3 (available at www.jneurosci.org as supplemental material).

Figure 3.

The pharmacokinetic variability across subjects was exploited to create a population dose–response curve for the rCBF change in the right dorsomedial thalamus VOI. For additional details, see Results, Quantifying SYN115 potency for decreasing rCBF as a function of plasma concentration.

Figure 4.

Expected effects in the indirect pathway of giving an A2a antagonist. See also Discussion, Physiological interpretation of results, and supplemental Table 7 (available at www.jneurosci.org as supplemental material). GPe, Globus pallidus, pars externa; GPi, globus pallidus, pars interna.

Figure 5.

Regional CBF responses to levodopa and to specific dopamine receptor agonists from previous studies. Note the midbrain response in all studies and the thalamus and posterior cingulate responses in the D2 and D3 agonist studies. a, Levodopa/carbidopa in PD and healthy controls (Hershey et al., 1998). b, Levodopa/carbidopa in advanced PD (Black et al., 2005). c, Levodopa/carbidopa in baboon (Hershey et al., 2000). d, Dopamine D1 agonist SKF82958 in baboon (Black et al., 2000). e, Dopamine D2 agonist U91356a in baboon (Black et al., 2002a). f, Dopamine D3 > D2 agonist pramipexole in baboon (Black et al., 2002b). c and d are from additional, unpublished analyses of the data collected for the publications named above.