Selective serotonin reuptake inhibition modulates response inhibition in Parkinson's disease

Abstract

Impulsivity is common in Parkinson's disease even in the absence of impulse control disorders. It is likely to be multifactorial, including a dopaminergic 'overdose' and structural changes in the frontostriatal circuits for motor control. In addition, we proposed that changes in serotonergic projections to the forebrain also contribute to response inhibition in Parkinson's disease, based on preclinical animal and human studies. We therefore examined whether the selective serotonin reuptake inhibitor citalopram improves response inhibition, in terms of both behaviour and the efficiency of underlying neural mechanisms. This multimodal magnetic resonance imaging study used a double-blind randomized placebo-controlled crossover design with an integrated Stop-Signal and NoGo paradigm. Twenty-one patients with idiopathic Parkinson's disease (46-76 years old, 11 male, Hoehn and Yahr stage 1.5-3) received 30 mg citalopram or placebo in addition to their usual dopaminergic medication in two separate sessions. Twenty matched healthy control subjects (54-74 years old, 12 male) were tested without medication. The effects of disease and drug on behavioural performance and regional brain activity were analysed using general linear models. In addition, anatomical connectivity was examined using diffusion tensor imaging and tract-based spatial statistics. We confirmed that Parkinson's disease caused impairment in response inhibition, with longer Stop-Signal Reaction Time and more NoGo errors under placebo compared with controls, without affecting Go reaction times. This was associated with less stop-specific activation in the right inferior frontal cortex, but no significant difference in NoGo-related activation. Although there was no beneficial main effect of citalopram, it reduced Stop-Signal Reaction Time and NoGo errors, and enhanced inferior frontal activation, in patients with relatively more severe disease (higher Unified Parkinson's Disease Rating Scale motor score). The behavioural effect correlated with the citalopram-induced enhancement of prefrontal activation and the strength of preserved structural connectivity between the frontal and striatal regions. In conclusion, the behavioural effect of citalopram on response inhibition depends on individual differences in prefrontal cortical activation and frontostriatal connectivity. The correlation between disease severity and the effect of citalopram on response inhibition may be due to the progressive loss of forebrain serotonergic projections. These results contribute to a broader understanding of the critical roles of serotonin in regulating cognitive and behavioural control, as well as new strategies for patient stratification in clinical trials of serotonergic treatments in Parkinson's disease.

Keywords: Parkinson’s disease; citalopram; functional MRI; response inhibition; serotonin.

Figure 1

(A) Task and stimuli. (B) Design and drug. PD = Parkinson’s disease; DA = dopamine.

Figure 2

In behaviour, citalopram reduced SSRT (left) and NoGo error rate (right) in relatively more advanced disease. The drug-induced changes of SSRT (ΔSSRT) and NoGo-error (ΔNoGo-error) negatively correlated with UPDRS-motor score (P < 0.05).

Figure 3

(A) In functional MRI, controls showed greater stop-related (Stop-Signal > Go, orange) and NoGo-related activations (NoGo > Go, green) in the right inferior frontal gyrus (P < 0.05 corrected). Activation overlap is in red. Coordinates are in MNI space. (B) The stop-related right inferior frontal gyrus activation was significantly weaker in Parkinson’s disease-placebo (PD-PLA) than in controls (disease effect, P < 0.05 corrected). Colour scale indicates t-values. (C) Citalopram’s enhancement of right inferior frontal gyrus activation in successful stops positively correlated with UPDRS-motor score (P < 0.05). (D) The drug-induced changes of SSRT (ΔSSRT) and NoGo-error (ΔNoGo-error) respectively related to right inferior frontal gyrus activation changes in successful Stop-Signal trials (ΔStop-Signal-activation) and those in NoGo trials (ΔNoGo-activation).

Figure 4

Citalopram’s effect on response inhibition was also related to the strength of structural connectivity (fractional anisotropy, FA) between the frontal cortex and striatum, showing a negative correlation of the drug-induced SSRT change (ΔSSRT) with fractional anisotropy of the skeleton voxels in the right internal capsule (mean centred). Thresholded at P < 0.05 TFCE-corrected.