Neural correlates of dysexecutive functions in Parkinson's disease

Background: It is yet unclear whether cognitive dysfunction in Parkinson's disease is caused mainly by striatal dysfunction, altered striatocortical outflow or prefrontal dysfunction. Reports on dopaminergic modulation of cognitive functions in these patients are conflicting. We used fMRI to investigate the neural basis of impaired executive functions in medicated Parkinson's disease patients.

Methods: 12 patients (duration of illness: 5.9 +/- 3.8 years, Hoehn and Yahr stage: 2 +/- 1) and 12 age, sex and handedness matched control subjects (mean age 61 +/- 7 years) were studied using fMRI. Dementia and depression were excluded and patients under regular anti-parkinsonian medication. We used a computerised version of the Wisconsin-Card Sorting Test (WCST): a stimulus card had to be assigned to 1 of 4 reference cards by button press (sorting dimension: color, form, or number). 3 different conditions with step-wise reduction of task demands (A>B>C) were contrasted versus baseline (bl). SPM2 was used for data processing and analysis, group-analysis based on a random effects model (uncorrected p<0.001). Performance measures were number of errors, error types, reaction times and the number of dimensions achieved.

Results: Patients showed more perseverative errors (p<0.062), and achieved significantly fewer dimensions compared to controls (p<0.03), indicating set-shifting deficits. Simple working memory operations were intact. fMRI-data showed the network underlying the performance of the WCST (A-bl), consisting in bilateral activations of frontoparietal regions. With decreasing task demands (A>B>C), control subjects showed a successive reduction of activation mainly in right lateral and mesial prefrontal cortex, while patients showed a failure in 'economising' the cortical response, which was associated with altered pallidal activation.

Group-comparison (condition A) showed significant hypoactivation of right inferior frontal cortex (s. fig. 1) in the patient group. Right inferior frontal cortex has previously been related to inhibitory functions. Dysfunction of this region may underlie set-shifting deficits due to the failure to inhibit a once acquired response pattern.

Conclusion: Hypoactivation in right inferior frontal cortex may underlie set-shifting deficits in Parkinson's disease and may be less responsive to dopaminergic therapy than working memory functions. This suggests that selective cognitive functions may be differentially modulated by dopaminergic therapy.

Figure 1: Group comparison of controls>patients for condition A. Patients show significant hypoactivation in right inferior frontal cortex. Plotted are the mean percent signal changes (blue: patients, violet. controls) across condition. Random effects model, p_u <0.001.