Citalopram Enhances Action Inhibition Systems in Parkinson’s DiseaseZ Ye1, E Altena1, C Nombela-Otero1, C Housden1, H Maxwell1, T Rittman1, C Huddleston1, CL Rae2,

R Regenthal3, BJ Sahakian4, RA Barker1, TW Robbins1,4, JB Rowe1,2,4

1University of Cambridge, UK 2Medical Research Council Cognition and Brain Science Unit, UK 3University of Leipzig, Germany 4Behavioural and Clinical Neuroscience Institute, UK

Contact Dr RoweJames.Rowe@mrc-cbu.cam.ac.uk

Download poster (No.1189)ww4.aievolution.com/hbm1301/

Contact Dr Yezy250@medschl.cam.ac.uk

Follow the Rowe Labwww.brc.cam.ac.uk/principal-investigators/james-rowe/

Introduction

Impulsivity in Parkinson’s disease (PD) is highlighted by the severity ofimpulse control disorders (ICDs). However, subtler impulsivity iscommon even in the absence of ICDs and is likely to be multifactorial.In addition to dopaminergic ‘overdose’1,2 and structural changes3 inthe frontostriatal circuits for motor control, we propose that changesin serotonergic projections to the forebrain also exacerbate theimpairment of response inhibition. Enhancing central serotonintransmission with selective serotonin reuptake inhibitors (SSRIs)might therefore provide adjunctive treatment for behaviouralimpulsivity in Parkinson’s disease.

We investigated whether the SSRI citalopram reduces impulsivity andenhances the neural systems mediating response inhibition inParkinson’s disease. We studied two forms of inhibition: (1) restraint,using NoGo events; and (2) cancellation, in terms of the Stop-SignalReaction Time (SSRT). There is strong preclinical evidence fromanimal and human studies that serotonin plays an important role inregulating action restraint4. In contrast, the link between serotoninand action cancellation is less well established. We tested threespecific hypotheses.

Hypothesis1: PD impairs both action restraint and cancellation.

Hypothesis2: The effect of citalopram on behavioural performancedepends on patients’ disease severity.

Hypothesis3: The behavioural effect relates to the enhancement ofinferior frontal cortical activation following citalopram.

Methods

Participants

• PD patients (N=21): right-handed.

• Controls (N=20): right-handed; no history of neurological or psychiatric disorders.

Table 1: Demographic data, clinical features and neuropsychological measures (means, SDs and p values of two-sample t tests)

Items PD Control P (2-tailed)

Sex (M:F) 11:10 12:8 n.s.

Age 64 (8) 65 (6) n.s.

MMSE 29 (1) 29 (1) n.s.

Years of disease 11 (5) -- --

UPDRS-III motor 21 (8) -- --

Hoehn & Yahr 1.9 (0.4) -- --

Levodopa equivalent dose (LED, mg/day)

632.6 (310.6) -- --

Beck Depression Inventory

9.9 (5.5) 3.8 (3.9) < 0.001

Simple reaction time (ms) 294 (53) 314 (72) n.s.

Choice reaction time (ms) 353 (47) 392 (70) < 0.05

Drug and Design

• Double-blind randomised placebo-controlled crossover design

• Patients (sessions ≥6 days apart): 30mg citalopram (CIT) and placebo (PLA)

• Controls (one session): on drug

• Citalopram: SSRI that increases extracellular cortical serotonin 4 fold

Task and Stimuli

• An integrated Stop-Signal and NoGo paradigm with 360 Go trials (75%), 40 NoGo trials (8%) and 80 Stop-Signal trials (SS, 17%, with ~50% accuracy).

• Go: Subject responded to a left/right black arrow (1000ms) with their right hand.

• NoGo: Subjects withheld a response to a red arrow (1000ms) and auditory tone.

• SS: A button press was cued by the left/right black arrow but signalled to stop by a colour change (to red) and auditory tone after a variable delay. An online tracking algorithm maintained approx. 50% successful inhibition5.

Results

• Behaviour (Table 2 & Fig.1A): Disease effect (Hypothesis1: controlvs. PD-PLA) and drug effect (Hypothesis2: PD under CIT vs. PLA)were examined respectively with two-sample t-tests and repeated-measures ANOVAs with disease severity, age, LED, and plasmaconcentration of CIT as covariates.

• fMRI (Fig.1B): General linear models estimated brain responses tocorrect SS, Go and NoGo trials, commission errors (SS, NoGo andGo trials), and Go omission errors. Contrast maps of SS>Go andNoGo>Go were computed for each group and compared betweengroups (Hypothesis1). We focused on the RIFG which showedinhibition-related activations in controls.

• Behaviour-fMRI (Fig.1C): Parameter estimates (betas) were pooledfrom the RIFG for correlation tests between changes in behaviourand changes in frontal activation following citalopram(Hypothesis3).

Table 2: Disease effect on behavioural performance (means, SDs and pvalues of two-sample t tests)

Parameter Control PD-PLA PD-CIT P (1-tailed)

Go RT (ms) 532 (129) 554 (108) 555 (100) n.s.

SSRT (ms) 142 (44) 167 (50) 180 (75) 0.05

NoGo error 0.06 (0.13) 0.14 (0.13) 0.16 (0.18) < 0.05

Go error 0.08 (0.05) 0.14 (0.06) 0.13 (0.08) < 0.001

Fig.1

A. Behavioural effect of CIT (changes of SSRT and NoGo errors) depended on disease severity (p<0.05).

B. The stop-related RIFG activation was significantly weaker in PD-PLA than in controls (p<0.05 FWE-corrected).

C. CIT-induced changes of SSRT and NoGo error correlated with the changes of inferior frontal cortical activation (p<0.05).

Conclusion

• PD impairs both action restraint and cancellation.

• Effect of citalopram on behavioural impulsivity depends on the severity of PD and individual difference in inferior frontal cortical activation.

• This study indicates the need for patient stratification in clinical trials and serotonergic treatments of impulsivity in PD.

References and Funding

1. Voon et al. 2011. Curr Opin Neurol 24(4): 324-30.

2. Rowe et al. 2008. Brain 131(Pt 8): 2094-105.

3. Rae et al. 2012. Neuroimage 62(3): 1675-84.

4. Eagle et al. 2008. Psychopharmacology (Berl) 199(3): 439-56.

5. Chamberlain et al. 2007. Biol Psychiatry 62(9): 977-84.

This work was primarily funded by