_______________________________________________________________

_______________________________________________________________ Report Information from ProQuest30 June 2015 10:08_______________________________________________________________

Document 1 of 1 Functional brain activity and presynaptic dopamine uptake in patients with Parkinson's disease andmild cognitive impairment: a cross-sectional study Author: Ekman, Urban; Eriksson, Johan; Forsgren, Lars; Mo, Susanna Jakobson; Riklund, Katrine; Nyberg, Lars ProQuest document link Abstract: Many patients with Parkinson's disease have mild cognitive impairment (MCI). Deficits in executivefunctions and working memory suggest dysfunctional frontostriatal brain circuitry. We aimed to assess brainresponses during a working memory task in a cohort of newly diagnosed drug-naive patients with Parkinson'sdisease with and without MCI. Participants were recruited within a prospective cohort study of incident patients with idiopathic parkinsonism,including Parkinson's disease. Between Jan 1, 2004, and April 30, 2009, all physicians in the Umeå catchmentarea were requested to refer all individuals with suspected parkinsonism to the Department of Neurology atUmeå University. Included patients fulfilled the UK Parkinson's Disease Society Brain Bank clinical diagnosticcriteria for Parkinson's disease. Control individuals were matched on the basis of age and sex with the first 50patients included in the study. Participants who scored 1·5 SDs or more below the population mean on at leasttwo cognitive measures were diagnosed with MCI. The primary outcome measures were functional MRI blood-oxygen-level-dependent signal and SPECT presynaptic uptake. Functional MRI was done during a verbal two-back working memory task. Presynaptic dopamine SPECT was done to assess presynaptic striataldopaminergic system integrity. Event-related transient analyses of functional MRI data were done for the wholebrain and for frontostriatal regions of interest, and semi-quantitative SPECT analyses were done for striatalregions of interest. Compared with controls (n=24), patients with Parkinson's disease (n=77) had under-recruitment in an extensivebrain network including bilateral striatal and frontal regions (p<0·001). Within the Parkinson's disease group,patients with Parkinson's disease and MCI (n=30) had additional under-recruitment in the right dorsal caudatenucleus (p=0·005) and the bilateral anterior cingulate cortex (p<0·001) compared with patients with Parkinson'sdisease without MCI (n=26). In patients with Parkinson's disease and MCI, SPECT uptake in the right caudatewas lower than in patients with Parkinson's disease without MCI (p=0·008) and correlated with striatal functionalMRI blood-oxygen-level-dependent signal ( r =0·32, p=0·031). These altered brain responses in patients with Parkinson's disease and MCI suggest that cognitive impairmentis linked to frontostriatal dysfunction. Swedish Medical Research Council, Swedish Parkinson Foundation, Swedish Parkinson's Disease Association,Umeå University, Kempe Foundation, Foundation for Clinical Neuroscience at Umeå University Hospital,Västerbotten County Council (ALF), King Gustaf V's and Queen Victoria's Freemason Foundation, Knut andAlice Wallenberg Foundation, and Swedish Brain Power. Full text:

Patients with Parkinson's disease(n=77)

Controls(n=24)

p valuePatients withParkinson's disease

WithMCI(n=30)

WithoutMCI(n=26)

pvalue

Men 46 (60%)12(50%)

..22(73%)

10(38%)

..

Table 1 - Demographic and clinical characteristics, including performance in neuropsychological tests Data are mean (SD) or number (%). p values are one-tailed except for age. Neuropsychological data representtransformed scaled t values (population mean t =50). The degrees of freedom were between 92 and 99 for thecomparisons between the Parkinson's disease group and the control group and between 49 and 54 for thecomparison between patients with Parkinson's disease with and without MCI. MCI=mild cognitive impairment.WCST=Wisconsin card sorting test. BVMT-R=brief visuospatial memory test, revised.

Women 31 (40%)12(50%)

..8(27%)

16(62%)

..

Age (years) 67·6 (9·8)67·9(6·8)

0·8767·8(9·6)

68·3(8·9)

0·82

Education (years) 10·1 (4·4) 12·2 (4) 0·0428·6(3·2)

10·7(3·5)

0·011

Unified Parkinson's disease ratingscale part 3, motor

24·3 (10·2)0·1(0·5)

<0·000127·1(10·4)

22·0(10·1)

0·034

Montgomery-Åsberg depressionrating scale

3·9 (3·3)0·6(1·8)

<0·00013·7(3·4)

3·9(3·4)

0·43

Mini-mental state examination 29·1 (1)29·2(0·9)

0·3528·8(1·1)

29·5(0·9)

0·011

WCST total errors 44·0 (12·3)50·5(11)

0·01237·7(12·1)

50·7(9·9)

<0·0001

WCST persevere 49·5 (13.3)48·5(10)

0·3844·3(14)

53·3(10·5)

0·0075

Fluency (animal) 52·3 (13·3)58·6(11·2)

0·01948·7(13·1)

54·9(14·3)

0·049

Trail making test, part B 40·8 (10·7)50·6(6·7)

<0·000133·7(10·7)

47·7(5·9)

<0·0001

Digit span back 41·7 (10·5)48·9(13.5)

0·004036·9(8·8)

49·0(9·6)

<0·0001

BVMT-R free recall 44·5 (10·3)53·0(8·6)

0·000538·0(10·7)

50·1(7·8)

<0·0001

BVMT-R delayed recall 46·9 (10·5)56·2(6·7)

<0·000141·2(10·6)

51·7(7·5)

<0·0001

Free and cued selective remindingtest, free recall

46·2 (10·2)50·2(8·9)

0·04841·5(10·3)

50·8(5·9)

<0·0001

Boston naming scale 51·6 (13·7)51·5(8·6)

0·4448·7(15·7)

55·5(11·5)

0·038

Judgment of line orientation test 53·5 (8·1)55·9(6·5)

0·09452·5(7·9)

55·0(6·4)

0·11

Side Peak (x, y, z) t kSubcortical

Striatum R 24, 4, 16 5·21 550

Striatum L -10, 2, -8 5·13 404

Frontal Precentral sulcus L-34, -22,56

6·31

407 Precentral sulcus R46, -6,50

5·00

675Dorsolateralprefrontal cortex

R36, 56,24

4·73

120 Insula L-22, 36,8

4·61

98Anterior cingulatecortex

L-14, 46,26

3·96

35 White matter L-26, -2,30

3·91

38 Eye fields L-22, 16,64

3·72

34 Orbitofrontal cortex L-34, 32,0

3·70

33Anterior cingulatecortex

R 2, 20, 28 3·33

11 Parietal Postcentral gyrus L-36, -38,70

5·45 183Supramarginalgyrus

R62, -20,40

3·54 34 CerebellumHemisphere andvermis

R

12, -50, -20 5·05 207Hemisphere andvermis *

L R

-16, -48, -24 4·94 1083 Crus 1 R

30, -74, -26 3·51 17 Crus 1 R

38, -48, -34 3·41 13 OccipitalCalcarine sulcus

L -14, -94, -6 4·70 144Lingualgyrus

Table 2 - Functional MRI blood-oxygen-level-dependent signals in patients with Parkinson's disease comparedwith controls The coordinates x, y, z refer to the anatomical location in Montreal Neurological Institute space for the peakvalues. p<0·001, k ≥10. R=right. L=left. t =t values. k =number of voxels.

Table 3 - Blood-oxygen-level-dependent signal differences in patients with Parkinson's disease with mildcognitive impairment versus those without p values are based on peak value within the clusters. L=left. R=right. k =number of voxels. t =t values.

L -36, -88, -16 3·73 22Inferiorsulcus

R 34, -94, -8 3·48 11Temporal

White matter L -50, -36, -10 3·85 43

Fusiform gyrus L -52, -52, -18 3·68 53

Superior gyrus L -54, -30, 12 3·50 12

Cingulate ventricle R 14, -22, -18 3·43 25

White matter R 28, -14, 24 3·36 10

Side Peak (x, y, z) t p value k

WithoutMCIversuswithMCI

Anterior cingulate cortex * L R 10, 40, 30 3·85 0·00017 57

White matter and ventricle R 22, 30, 20 4·00 0·00011 71

R 18, 4, 28 3·67 0·00029 29

WithMCIversuswithoutMCI

Parahippocampus R 22, 8, -24 4·04 <0·0001 18

Parietal superior lobe R 28, -66, 60 3·61 0·00035 10

Side Centre (x, y, z) t p value k Caudate

L -18, 8, 14 0·68 0·25 15 Caudate

R 14, -8, 18 2·68 0·005 62Putamen

L -24, 8, 0 0·95 0·17 128Putamen

Table 4 - Region of interest analyses in patients with Parkinson's disease without mild cognitive impairmentversus those with mild cognitive impairment p values are based on the effects of the total cluster size. L=left. R=right. t =t value. k =number of voxelsdefining the region of interest. Introduction Parkinson's disease is characterised by motor symptoms, but cognitive deficits are also common.1-3 Severalcognitive domains are affected,1-5 with deficits consistently reported for measures of executive function, episodicmemory, and working memory that rely on frontostriatal brain circuits5-10 and dopaminergic neurotransmission.5,7

Correspondingly, findings from imaging studies have shown altered functional activity in the striatum in patientswith Parkinson's disease compared with that in healthy individuals during task-set shifting,6 manipulation,7 andupdating.8

Between 20% and 40% of patients with Parkinson's disease have mild cognitive impairment (MCI) in the earlyphase of the disease.4 MCI is defined as a cognitive deficit commonly quantified as a performance level 1-2 SDsbelow the population mean in one or more cognitive domains.11 Patients with Parkinson's disease and MCI havean increased risk of developing dementia compared with patients without MCI.1,5,9 Patients with Parkinson'sdisease and MCI might experience more pronounced alterations in the frontostriatal circuit than those withoutMCI, and phasic hypoactivity has been noted in the striatum and the prefrontal cortex of patients withParkinson's disease with impairments in executive function compared with cognitively non-impaired patients. 10

However, functional brain responses have not been studied in patients with Parkinson's disease and MCI. Thecognitive dysfunctions associated with Parkinson's disease are gaining clinical importance because of therelative success of therapeutic approaches in the treatment of motor symptoms, but knowledge of theneuropathology underlying cognitive impairment remains insufficient. 5 We therefore examined brain responsesduring a working memory task and presynaptic dopamine striatal integrity during the resting state in apopulation-based cohort of newly diagnosed drug-naive patients with Parkinson's disease with or without MCI.We hypothesised that frontostriatal under-recruitment occurs in patients with Parkinson's disease comparedwith healthy individuals and in patients with Parkinson's disease with MCI compared with those without MCI. Methods Participants Participants were recruited within the newly diagnosed parkinsonism in Umeå (NYPUM) project, which is alongitudinal population-based cohort study of incident patients with idiopathic parkinsonism, includingParkinson's disease.2 Between Jan 1, 2004, and April 30, 2009, all physicians in the Umeå catchment area(about 142 000 inhabitants) were asked to refer all patients with suspected parkinsonism to the Department ofNeurology at Umeå University. All referred patients underwent a standardised clinical examination by aneurologist specialised in movement disorders (LF). Patients who fulfilled the UK Parkinson's Disease SocietyBrain Bank clinical diagnostic criteria for Parkinson's disease were included in the study. We excluded patientson antiparkinsonian drugs (ie, dopaminergic and anticholinergic drugs) and those diagnosed with dementia;participants with functional MRI (fMRI)-related issues (ie, unsatisfactory picture quality caused by participantsmoving within the scanner during acquisition or other scanner-specific technical complications), those whoanswered correctly less than 55% of questions in the working memory task, and participants with a depressionrating score of over 17 (suggestive of major depression according to the Montgomery-Åsberg depression ratingscale) 12 were also excluded. Healthy control individuals were recruited via advertisement in the localnewspaper and were matched on the basis of age and sex with the first 50 patients included in the study.

R 30, 0, -8 -0·14 0·45 252

Frontaldorsolateralcortex

Control individuals had no history of neurological disorders, had a neurological examination with normalfindings, and had normal dopamine uptake as assessed by dopamine transporter binding with 123 I-N-(omega)-fluoropropyl-2-beta-carbomethoxy-3-beta-(4-iodophenyl) nortropane SPECT (123 I-FP-CIT). This study was approved by the ethics committee of the Faculty of Medicine at Umeå University, Umeå,Sweden. Written informed consent was obtained from all participants before inclusion in the study. Procedures MCI was diagnosed according to recently proposed consensus criteria established by the Movement DisorderSociety's commissioned task force.11 The cognitive domains examined were executive function, attention andworking memory, episodic memory, language, and visuospatial function. The neuropsychological assessmentsand test scores are listed in table 1. The raw cognitive test scores were converted into standardised age-matched t values (population mean t =50) according to the normative age-matched test data. Participants whoscored 1·5 SDs or more below the mean value for at least two cognitive measures were diagnosed with MCI.Because only one cognitive measure in the language and visuospatial domains was available in the test battery,Parkinson's disease with MCI was diagnosed on level 1 criteria (ie, subtyping not possible). In agreement with arecent meta-analysis, 4 no subjective measures were part of our MCI classification. Participants who scoredbelow the normative age-matched mean value (1·5 SDs or more) on one cognitive measure were defined ashaving Parkinson's disease with intermediate cognitive impairment. Dementia was diagnosed using consensuscriteria, 13 which include both subjective measures (clinical interview) and objective measures (cognitiveimpairment in at least two cognitive domains with scores over 2 SDs below normative age-matched t values). Patients were followed up for 12-60 months and the diagnosis according to the UK Parkinson's Disease SocietyBrain Bank clinical diagnostic criteria for Parkinson's disease was reassessed and confirmed at the latest follow-up. Neuropsychological assessments, fMRI, and 123 I-FP-CIT SPECT were done within 1-2 months of initialdiagnosis, and all patients were drug naive for dopaminergic medication at the time of the examinations.Primary outcome measures were fMRI blood-oxygen-level-dependent (BOLD) signal and SPECT presynapticdopamine uptake. fMRI analyses were done using two different scanners: a 1·5 T Philips Intera scanner and a 3 T Philips Achievascanner (both from Philips, Eindhoven, Netherlands). Both scanners collected T2 * -weighted images withsingle-shot gradient-echo echoplanar imaging for BOLD imaging. Acquisition parameters for the 1·5 T scannerwere repetition time (TR) 3000 ms (33 slices), echo time (TE) 50 ms, flip angle 90°, field of view 22×22 cm,64×64 matrix, and 4·40 mm slice thickness (voxel size 3·44×3·44×4·40 mm). Five dummy scans were donebefore image acquisition to avoid the effects of signal saturation. Acquisition parameters for the 3 T scannerwere TR 1500 ms (31 slices), TE 30 ms, flip angle 70°, field of view 22×22 cm, 64×64 matrix, and 4·65 mm slicethickness (voxel size 3·44×3·44×4·65 mm). Ten dummy scans were done before image acquisition. Duringscanning, participants did a verbal two-back working memory task in which they needed to maintain and activelyupdate relevant information (a noun). 8 The participants received instructions and practised on the two-back taskbefore scanning for about 10 min. They were asked to respond "yes" (right index finger) when the wordmatched the one presented two items earlier (two-back), and "no" when it was different (left index finger), usingMRI-compatible keypads (Lumitouch reply system, Lightwave Medical Industries, Burnaby, BC, Canada).Behavioural performance was recorded using E-prime 1.1 (Psychology Software Tools, Sharpsburg, PA, USA).In the 1·5 T scanner, single words (nouns) were presented for 2·5 s each. The inter-stimulus intervals were 2-20s in a random fashion. Four task blocks (63 s per block) were interleaved with baseline blocks (21 s) and eachtask block consisted of eight trials. In the 3 T scanner, single words (nouns) were presented for 1·5 s each. Theinter-stimulus intervals were 3 s. Four task blocks (51 s per block) were done interleaved with baseline blocks(21 s). Each task block consisted of 15 trials. During the baseline condition, participants were instructed to donothing apart from keep their gaze fixed on a small circle displayed at the centre of the screen (in bothscanners). All fMRI assessments were done unmasked to clinical information and diagnosis.

Presynaptic dopamine integrity was examined by 123 I-FP-CIT SPECT.14 SPECT imaging was done byspecialists in nuclear medicine and radiology (SJM and KR) masked to any clinical information or diagnosis andwas analysed as described previously.14 In summary, 123 I-FP-CIT SPECT was done 3 h after receiving thyroidprotection and an intravenous injection bolus dose of a mean of 187 MBq. Statistical analysis Preprocessing and statistical analyses of fMRI data were done with SPM 8 (Wellcome Department of ImagingNeuroscience, London, UK). Preprocessing included slice timing correction, realignment to the first image in theseries, unwarping, normalisation, and smoothing applying an isotropic Gaussian filter kernel by a full with at halfmaximum of 8 mm. To assess group-specific anatomical brain differences, all participants were initiallynormalised to the Montreal Neurological Institute echoplanar imaging template. In a second step, normalisationwas done to group-specific means for control individuals, patients with Parkinson's disease without MCI, andpatients with Parkinson's disease and MCI. Finally, the mean-normalised brain templates were projected to thetotal mean of all participants. Event-related transient fMRI responses were modelled as regressors containing delta functions thatrepresented onsets of word stimuli. The regressors were convolved with a canonical haemodynamic responsefunction. Model estimations from each individual were input into a second-level group-by-scanner factorialdesign to assess possible scanner and protocol differences. 15 Unified Parkinson's disease rating scale(UPDRS) part 3 motor scores for each individual were also included as covariates in the model comparingpatients with Parkinson's disease with and without MCI. Confounding effects of scanner and motor scores wereevaluated by assessing their regional overlap with reported group differences and by F tests on scanner ormotor score main effects or group-by-scanner or group-by-motor score interactions. The statistical threshold in the fMRI analyses was set to p<0·001 (uncorrected) and number of voxels (k ) ten orgreater for the whole-brain analysis. A region of interest (ROI) analysis was done to compare findings frompatients with Parkinson's disease with and without MCI. Five ROIs were defined, four in the striatum (bilateralcaudate and putamen) and one in the right dorsolateral prefrontal cortex. The striatal ROIs were defined bycombining (inclusive masking) the activation pattern derived from comparing patients with Parkinson's diseasewith control individuals (voxels that displayed scanner effects were excluded) with anatomical information ofstriatal structure (automated anatomical labelling). 16 The right dorsolateral prefrontal cortex ROI was definedfrom the activation contrast between patients with Parkinson's disease and control individuals. The results of thecomparison of patients with Parkinson's disease with and without MCI were Bonferroni corrected for number ofROIs (p<0·01). fMRI behavioural analyses were done with scanner as a covariate (p<0·05). Allneuropsychological and experimental test values were analysed as one-tailed t tests because of a-priorihypotheses. Semi-quantitative analysis of the SPECT image data was done using ROIs in the bilateral caudate andputamen, and a background reference ROI was applied to the occipital cortex. Uptake ratio was calculated bydividing the uptake in the striatal subregions by that in the reference region. Role of the funding source The sponsors of the study had no role in the study design, data collection, data analysis, data interpretation, orwriting of the report. The corresponding author had full access to all the data in the study, and all coauthors hadfinal responsibility for the decision to submit for publication. Results 77 patients with Parkinson's disease were included in the study, of whom 33 (43%) were diagnosed with MCI(figure 1). Within the Parkinson's disease and MCI group, 21 participants scored below the criteria (ie, ≥1·5 SDsbelow the normative age-matched test data) on at least one executive function measure. The correspondingnumbers of participants for the other cognitive domains were as follows: attention or working memory, 25;episodic memory, 21; language, seven; and visuospatial function, two. To increase homogeneity in the MCI

group, three participants who were diagnosed with MCI on the basis of decline in only two cognitive measures,of which one was within the language or visuospatial domain, were excluded from the fMRI analyses (figure 1).18 participants without MCI who scored below criteria on one cognitive measure were excluded in the maincomparison between patients with Parkinson's disease with and without MCI (figure 1). In control analyses,these 18 participants were included in the group without MCI. One of 25 control individuals fulfilled the MCIcriteria and was therefore excluded. Thus, the final groups consisted of 77 patients with Parkinson's diseaseand 24 control individuals. Of the 77 patients with Parkinson's disease, 30 patients with MCI were comparedwith 26 patients without MCI (table 1 and figure 1). Of these, 19 patients with Parkinson's disease and MCI, 15patients with Parkinson's disease without MCI, 13 patients with Parkinson's disease and intermediate cognitiveimpairment, and 17 control individuals were examined on a dual-head hybrid gamma camera system (InfiniaHawkeye, General Electric, Milwaukee, WI, USA) as described previously. 14 Because of equipmentreplacement, the remaining cases were examined on another gamma camera system (triple-head; GeneralElectric). As shown in table 1, the control individuals scored significantly better on most cognitive measures than didpatients with Parkinson's disease. Patients with Parkinson's disease had fewer correct answers than controls (F1,98=4·0, p=0·024) on the scanner working memory task. Patients with Parkinson's disease and MCI hadstatistically significantly reduced performance on all cognitive measures compared with that of those withoutMCI, except for the judgment of line orientation test, and had fewer correct answers on the scanner workingmemory task compared with both control individuals ( F 1,51=4·88, p=0·02) and patients with Parkinson'sdisease without MCI (F 1,53=3·36, p=0·04). Neither measures of reaction time (all p>0·26) nor the number ofbutton presses (all p>0·23) were significantly different in any of the group comparisons. Also, there was nosignificant difference in proportion of yes or no answers between groups (p=0·79). For fMRI analyses, 41 participants in the Parkinson's disease group did the verbal two-back working memorytask while in the 1·5 T scanner and 36 while in the 3 T scanner. The corresponding numbers for the groups withParkinson's disease with and without MCI were 16 and 14 in the 1·5 T scanner and 14 and 12 in the 3 Tscanner, respectively. In the control group, 12 participants were assessed in each scanner. Compared withcontrols, patients with Parkinson's disease had significant under-recruitment of the right dorsolateral prefrontalcortex, the bilateral primary and premotor cortices, the occipital cortex, and the cerebellum when undertakingthe two-back task (table 2 and figure 2). Additionally, a lower BOLD signal was noted for the Parkinson'sdisease group in bilateral striatal regions compared with the control group. When control individuals werecompared with patients with Parkinson's disease with or without MCI separately, a similar group effect wasnoted, but there was a more pronounced frontostriatal reduction in those with MCI (appendix). Patients withParkinson's disease did not have a higher BOLD signal than control individuals in any part of the brain (p>0·01).For the comparison between control individuals and patients with Parkinson's disease (4262 significant voxels),the main effect of scanner (757 voxels) and group-by-scanner interactions (31 voxels) overlapped (18·5%) withgroup differences, mainly in cerebellar regions, the occipital cortex, the ventral striatum, and frontal motorregions. In the whole-brain analysis, patients with Parkinson's disease and MCI had significant under-recruitment of thebilateral anterior cingulate cortex compared with those without MCI (table 3 and figure 3). The anterior cingulatecortex region was significantly recruited in both control individuals and patients with Parkinson's disease withoutMCI, but weakly and non-significantly recruited in patients with Parkinson's disease with MCI (figure 3). Whenthe 18 patients with Parkinson's disease with intermediate cognitive impairment were included in theParkinson's disease without MCI group and compared with the Parkinson's disease with MCI group, the anteriorcingulate cortex effect was attenuated but remained significant (p=0·00074). Additionally, there was significantunder-recruitment in the right parahippocampus and right superior parietal cortex in patients with Parkinson'sdisease without MCI compared with those with MCI (table 3). There was no overlap between scanner or motor

score effects and the comparisons between the Parkinson's disease with and without MCI groups, even at aliberal threshold (p<0·01). Quantification of the anterior cingulate cortex response for the 1·5 T and 3 Tscanners separately showed that the magnitudes of BOLD signal change were similar (Cohen's d for 1·5 T, 0·70and 3 T, 1·54; p=0·38). In the ROI analysis, there was under-recruitment of the right dorsal caudate in patients with Parkinson's diseasewith MCI compared with those without MCI (p=0·005; figure 3 and table 4). A plot of the effect revealed astepwise response such that the right caudate response was maximum in control individuals, intermediate inpatients with Parkinson's disease without MCI, and lowest and non-significant in those with Parkinson's diseaseand MCI (figure 3). When the 18 patients with Parkinson's disease and intermediate cognitive impairment wereincluded in the Parkinson's disease without MCI group and compared with the Parkinson's disease with MCIgroup, the effect in the caudate was attenuated but significant (p=0·024). Additionally, there was some evidenceof under-recruitment in patients with Parkinson's disease and MCI in the right dorsolateral prefrontal cortex(p=0·033) but not in any other ROI (all p>0·17). To further verify that the differences between patients withParkinson's disease with and those without MCI were driven by the cognitive demands of the working memorytask and not by motor impairment, we compared groups matched (p=0·95) on (UPDRS part 3) motor scores(Parkinson's disease with MCI group mean 25·8 [SD 8]; Parkinson's disease without MCI group 25·6 [7·7];figure 3) as well as number of participants per group (n=18), and scanner model. The results persisted (rightcaudate p=0·001; anterior cingulate cortex p=0·00074). Quantification of the right caudate response for the 1·5T and 3 T scanners separately revealed similar magnitudes of BOLD signal change (Cohen's d for 1·5 T, 1·08and 3 T, 0·52; p=0·86). Patients with Parkinson's disease had significantly lower presynaptic dopamine uptake in all striatal regionsthan control individuals (p<0·0001). Uptake was lower in the right caudate in patients with Parkinson's diseaseand MCI than in those without MCI (p=0·008) and correlated with BOLD signal in the right caudate ( r =0·32,p=0·031; figure 4). Discussion The main finding of this study was the pronounced working-memory-related under-recruitment of the right dorsalcaudate nucleus, the bilateral anterior cingulate cortex, and, more weakly, the right dorsolateral prefrontal cortexin patients with Parkinson's disease with MCI compared with those without MCI (panel). Furthermore, patientswith Parkinson's disease and MCI had lower presynaptic 123 I-FP-CIT binding in the right caudate than thosewithout MCI, and binding correlated with caudate BOLD signal alterations. Moreover, compared with controlindividuals, patients with Parkinson's disease had lower scores on the working memory task and under-recruitment of an extensive brain network including bilateral striatal and frontal regions. 6-8 Taken together, theseresults are in agreement with previous reports showing that frontostriatal regions are implicated in the updatingof working memory,17 that binding of raclopride to dopamine D2 receptors is raised during the updating ofworking memory,18 and that training for updating leads to increased striatal BOLD signal as well as dopaminebinding.17,18

The caudate is a key region in the updating of phasic working memory,6-8,17 and reduced dopamine integrity inthe caudate nucleus is associated with impaired working memory performance and executive deficits in patientswith Parkinson's disease.19,20 Additionally, the anterior cingulate cortex has an important role in high-levelcognitive processing,21 and alterations in the anterior cingulate cortex have been reported in patients withParkinson's disease both for those with MCI and those with dementia. The alterations have been associatedwith reduced grey matter volumes,22,23 dopaminergic functioning,24 and higher Lewy body densities.25 Thus, ourresults show the importance of the caudate and the anterior cingulate cortex in the updating of working memoryand their associated functional alterations in Parkinson's disease with cognitive impairment. The dorsal caudate nucleus is one of the main output targets of the nigrostriatal tract,26 and abnormalmodulation in nigrostriatal dopaminergic circuits is associated with cognitive impairments in early stages of

Parkinson's disease.5,19,27 In the present study, we confirmed that deficits in nigrostriatal dopaminergictransmission in relation to altered BOLD signal are associated with cognitive impairment. The alteration in thecaudate might have secondary effects on the anterior cingulate cortex and the dorsolateral prefrontal cortexbecause of the functional connectivity between these regions. 28 Also, abnormal modulation of mesolimbicdopaminergic transmission might in part explain altered frontal recruitment. Both the nigrostriatal and themesocortical dopamine systems are involved in and interact during working memory processing,27 and themesocortical dopaminergic system facilitates working memory through direct inputs to the prefrontal cortex.19

Uptake reductions in both the mesocortical and nigrostriatal dopamine systems are evident at later stages ofcognitive decline (in patients with Parkinson's disease dementia).24 Thus, an abnormally modulateddopaminergic pattern related to the updating of working memory might in part account for the functionalalterations reported in patients with Parkinson's disease and MCI. Diverse alterations in neurochemistry might cause cognitive impairment in early Parkinson's disease.5

Dysfunctions in cholinergic and norepinephrinergic systems have been proposed to be involved in high-ordercognitive dysfunctions and might to some extent have contributed to the present results. Even though thematched group analyses support the notion that the differences in findings between patients with Parkinson'sdisease with and without MCI were driven by cognitive rather than motor processes, we cannot fully rule out thatother aspects related to Parkinson's disease (eg, mood or anxiety) could have affected our findings. Increased metabolism has previously been reported in patients with Parkinson's disease and MCI.29 Theincreased BOLD signal in patients with MCI, compared with that in those without MCI, in the parahippocampusand the parietal cortex suggests larger attentional demands and reliance on declarative memory systems inpatients with Parkinson's disease and MCI, because of frontostriatal dysfunction. However, such potentialcompensatory effects would be better assessed with longitudinal follow-up. 30 The absence of BOLD signalincreases in patients with Parkinson's disease compared with control individuals might be because of the earlystage of disease development (newly diagnosed), such that compensatory activation has not yet developed.Furthermore, the significant involvement of the caudate in the working memory task might have restrictedcortical overactivity, as previously suggested. 6

We acquired data on two different scanners, which is a limitation of the present study. The detection ability insubcortical regions is limited in the 1·5 T scanner.31 However, similar scanner distributions within the groupsprobably reduced the effect of the scanner factor on our results. Also, control analyses of scanner effects didnot suggest overlap with the reported differences between the group of patients with MCI and the group withoutMCI. A further limitation was the inability to compare subgroups of patients with Parkinson's disease and MCIbecause of limitations in the neuropsychological battery, which might have identified different structural andfunctional correlates of subgroup-specific cognitive deficits. Future studies could differentiate MCI subtypes byincreasing the number of neuropsychological measures (ie, at least two measures in each domain). In conclusion, we identified altered striatal functional recruitment in patients with Parkinson's disease comparedwith control individuals, along with BOLD signal reduction in motor areas and in the right dorsal caudate nucleusduring a working memory task. Furthermore, we found decreased BOLD signals in the right caudate and frontalcortex in patients with Parkinson's disease and MCI compared with those without MCI, together with impairedpresynaptic function in the right caudate. The population-based cohort design of this cross-sectional study mightenable generalisation of the results to the general Parkinson's disease population, and the fMRI andpresynaptic dopamine alterations reported add to the knowledge of the neural basis of MCI in Parkinson'sdisease. This online publication has been corrected. The corrected version first appeared at thelancet.com/neurology onOctober 15, 2012 Contributors UE collected, analysed, and interpreted the data. JE analysed and interpreted the data. LF designed the study

and collected the data. SJM collected, analysed, and interpreted SPECT data. KR designed the study andinterpreted the SPECT data. LN designed the study and interpreted the data. UE, JE, and LN wrote themanuscript. All authors reviewed and edited the manuscript. Conflicts of interest We declare that we have no conflicts of interest. Acknowledgments We thank the staff at the Umeå Center for Functional Brain Imaging, and our collaborators in the NYPUMproject. Supplementary Material Research in context Systematic review We searched Google Scholar and PubMed (up to December, 2011), with the search terms "Parkinson'sdisease", "cognitive impairment", "MCI", "fMRI", "SPECT", and "working-memory". We restricted our search toreports in English. We also searched the reference lists within identified reports. Cohort studies and randomisedcontrolled trials were prioritised. No previous prospective study assessing functional MRI (fMRI) and SPECT inpatients with Parkinson's disease and mild cognitive impairment (MCI) was identified. Interpretation We did a large-scale population-based fMRI and presynaptic dopamine SPECT study in drug-naive patientswith early Parkinson's disease with and without MCI. The population-based cohort design might enablegeneralisation of the results to the general Parkinson's disease population. This study provides new informationon the neural dysfunction that underlies cognitive impairment in Parkinson's disease. We identified decreased(blood-oxygen-level-dependent) BOLD signals in patients with Parkinson's disease and MCI in frontostriatalcircuits compared with patients with Parkinson's disease without MCI. Furthermore, we noted a significantassociation between striatal dopamine transporter binding and fMRI signal change, which provides new supportto the notion that impaired striatal functioning plays a key part in Parkinson's disease with MCI. Even though ourfindings suggest that the reported effects were driven by cognitive demands and not by motor severity, wecannot fully reject that other aspects related to Parkinson's disease could have affected the findings. References Williams-Gray CH, Foltynie T, Brayne CEG, Robbins TW, Barker RA: Evolution of cognitive dysfunction in anincident Parkinson's disease cohort . Brain 130 1787-1798, 2007. Elgh E, Domellöf M, Linder J, Edström M, Stenlund H, Forsgren L: Cognitive function in early Parkinson'sdisease: a population-based study . Eur J Neurol 16 1278-1284, 2009. Muslimovic D, Post B, Speelman JD, Schmand B: Cognitive profile of patients with newly diagnosed Parkinsondisease . Neurology 65 1239-1245, 2005. Aarsland D, Bronnick K, Williams-Gray C: Mild cognitive impairment in Parkinson disease: a multicenter pooledanalysis . Neurology 5 1062-1069, 2010. Kehagia A, Barker RA, Robbins TW: Neuropsychological and clinical heterogeneity of cognitive impairment anddementia in patients with Parkinson's disease . Lancet Neurol 9 1200-1213, 2010. Monchi O, Petrides M, Mejia-Constain B, Strafella AP: Cortical activity in Parkinson's disease during executiveprocessing depends on striatal involvement . Brain 130 233-244, 2007. Owen AM: Cognitive dysfunction in Parkinson's disease: the role of frontostriatal circuitry . Neuroscientist 10525-537, 2004. Marklund P, Larsson A, Elgh E: Temporal dynamics of basal ganglia under-recruitment in Parkinson's disease:transient caudate abnormalities during updating of working memory . Brain 132 336-346, 2009. Emre M: Dementia associated with Parkinson's disease . Lancet Neurol 2 229-237, 2003. Lewis SJG, Dove A, Robbins TW, Barker RA, Owen AM: Cognitive impairments in early Parkinson's disease

are accompanied by reductions in activity in frontostriatal neural circuitry . J Neurosci 23 6351-6356, 2003. Litvan I, Goldman JG, Tröster AI: Diagnostic criteria for mild cognitive impairment in Parkinson's disease:Movement Disorder Society Task Force guidelines . Mov Disord 27 349-356, 2012. Montgomery SA, Åsberg M: A new depression scale designed to be sensitive to change . Br J Psychiatry 134382-389, 1979. Emre M, Aarsland D, Brown R: Clinical diagnostic criteria for dementia associated with Parkinson's disease .Mov Disord 22 1689-1707, 2007. Mo SJ, Linder J, Forsgren L, Larsson A, Johansson L, Riklund K: Pre- and postsynaptic dopamine SPECT inthe early phase of idiopathic parkinsonism: a population-based study . Eur J Nucl Med Mol Imaging 37 2154-2164, 2010. Stonnington CM, Tan G, Klöppel S: Interpreting scan data acquired from multiple scanners: a study withAlzheimer's disease . Neuroimage 39 1180-1185, 2008. Tzourio-Mazoyer N, Landeau B, Papathanassiou D: Automated anatomical labelling of activations in SPM usinga macroscopic anatomical parcellation of the MNI MRI single-subject brain . Neuroimage 15 273-289, 2002. Dahlin E, Neely AS, Larsson A, Bäckman L, Nyberg L: Transfer of learning after updating training mediated bythe striatum . Science 320 1510-1512, 2008. Bäckman L, Nyberg L, Soveri A: Effects of working-memory training on striatal dopamine release . Science 333718- 2011. Rinne JO, Portin R, Ruottinen H: Cognitive impairment and the brain dopaminergic system in Parkinsondisease: [18F] fluorodopa positron emission tomographic study . Arch Neurol 57 470-475, 2000. Marié RM, Barré L, Dupuy B, Viader F, Defer G, Baron JC: Relationships between striatal dopaminedenervation and frontal executive tests in Parkinson's disease . Neurosci Lett 260 77-80, 1999. Duncan J, Owen AM: Common regions of the human frontal lobe recruited by diverse cognitive demands .Trends Neurosci 23 475-483, 2000. Lee JE, Park H-J, Song SK, Sohn YH, Lee JD, Lee PH: Neuroanatomic basis of amnestic MCI differs in patientswith and without Parkinson disease . Neurology 75 2009-2016, 2010. Nagano-Saito A, Washimi Y, Arahata Y: Cerebral atrophy and its relation to cognitive impairment in Parkinsondisease . Neurology 64 224-229, 2005. Ito K, Nagano-Saito A, Kato T: Striatal and extrastriatal dysfunction in Parkinson's disease with dementia: a 6-[18F]fluoro-L-dopa PET study . Neurology 125 1358-1365, 2002. Kövari E, Gold G, Herrmann FR: Lewy body densities in the entorhinal and anterior cingulate cortex predictcognitive deficits in Parkinson's disease . Acta Neuropathol 106 83-88, 2003. Bernheimer H, Birkmayer W, Hornykiewicz O, Jellinger K, Seitelberger F: Brain dopamine and the syndromes ofParkinson and Huntington. Clinical, morphological and neurochemical correlations . J Neurol Sci 20 415-455,1973. Sawamoto N, Piccini P, Hotton G, Pavese N, Thielemans K, Brooks DJ: Cognitive deficits and striato-frontaldopamine release in Parkinson's disease . Brain 131 1294-1302, 2008. Postuma RB, Dagher A: Basal ganglia functional connectivity based on a meta-analysis of 126 positronemission tomography and functional magnetic resonance imaging publications . Cereb Cortex 16 1508-1521,2006. Huang C, Mattis P, Perrine K, Brown N, Dhawan V, Eidelberg D: Metabolic abnormalities associated with mildcognitive impairment in Parkinson's disease . Neurology 70 1470-1477, 2008. Nyberg L, Salami A, Andersson M: Longitudinal evidence for diminished frontal cortex function in aging . ProcNatl Acad Sci USA 107 22682-22686, 2010. Nyberg L, Larsson A, Eriksson J, Birgander R, Sundström T, Riklund Ahlström K: Comparing 1.5T and 3T boldfMRI imaging of finger tapping with familiar and novel sequences . Schaller B: Neuroscience imaging research

trends New York, NY Nova Science Publishers. 161-172, 2007. AuthorAffiliation Lars Nyberg, Johan Eriksson, Urban Ekman. Department of Integrative Medical Biology, Umeå University,Umeå, Sweden Katrine Riklund, Lars Nyberg, Urban Ekman, Susanna Jakobson Mo. Department of Radiation Sciences,Diagnostic Radiology, Umeå University, Umeå, Sweden Lars Forsgren, Urban Ekman. Department of Pharmacology and Clinical Neuroscience, Umeå University,Umeå, Sweden Katrine Riklund, Lars Nyberg, Johan Eriksson, Urban Ekman. Umeå Center for Functional Brain Imaging, UmeåUniversity, Umeå, Sweden; Correspondence to: Mr Urban Ekman, Department of Integrative Medical Biology,Umeå University, SE-901 87 Umeå, Sweden MeSH: Aged, Brain -- metabolism, Brain -- radionuclide imaging, Cross-Sectional Studies, Female, Humans,Magnetic Resonance Imaging, Male, Middle Aged, Mild Cognitive Impairment -- metabolism, Mild CognitiveImpairment -- radionuclide imaging, Parkinson Disease -- metabolism, Parkinson Disease -- radionuclideimaging, Prospective Studies, Brain -- physiopathology (major), Dopamine -- metabolism (major), Mild CognitiveImpairment -- physiopathology (major), Parkinson Disease -- physiopathology (major), Presynaptic Terminals --metabolism (major) Substance: Dopamine; Publication title: The Lancet Neurology Volume: 11 Issue: 8 Pages: 679-87 Publication year: 2012 Publication date: Aug 2012 Year: 2012 Publisher: Elsevier Limited Place of publication: London Country of publication: United States Publication subject: Medical Sciences--Psychiatry And Neurology ISSN: 14744422 CODEN: LANCAO Source type: Scholarly Journals Language of publication: English Document type: DIS, Journal Article DOI: http://dx.doi.org/10.1016/S1474-4422(12)70138-2 Accession number: 22742929 ProQuest document ID: 1139220637 Document URL: http://search.proquest.com/docview/1139220637?accountid=34643

Copyright: © 2012 Elsevier Ltd Last updated: 2014-08-16 Database: ProQuest Nursing & Allied Health Source

_______________________________________________________________ Contact ProQuest Copyright 2015 ProQuest LLC. All rights reserved. - Terms and Conditions