BRAINA JOURNAL OF NEUROLOGY

A network centred on the inferior frontalcortex is critically involved in levodopa-induceddyskinesias

Antonio Cerasa,1,* Giacomo Koch,2,3,* Giulia Donzuso,1 Graziella Mangone,1

Maurizio Morelli,4 Livia Brusa,5 Mario Stampanoni Bassi,2,3 Viviana Ponzo,2 Silvia Picazio,2

Luca Passamonti,1 Maria Salsone,1 Antonio Augimeri,1 Carlo Caltagirone2 andAldo Quattrone1,4

*These authors contributed equally to this work.

Levodopa-induced dyskinesias are disabling motor complications of long-term dopamine replacement in patients with Parkinson’s

disease. In recent years, several alternative models have been proposed to explain the pathophysiological mechanisms underlying

this hyperkinetic motor disorder. In particular, our group has shed new light on the role of the prefrontal cortex as a key site of

interest, demonstrating that, among other areas, the inferior frontal cortex is particularly characterized by altered patterns of

anatomical and functional changes. However, how neural activity varies depending on levodopa treatment in patients with dys-

kinesias and whether the reported prefrontal abnormalities may have a critical role in dyskinesias is debated. To answer these

questions we performed independent functional magnetic resonance imaging and repetitive transcranial magnetic stimulation

studies. In the first experiment we applied resting state functional magnetic resonance imaging on 12 patients with Parkinson’s

disease with levodopa-induced dyskinesias and 12 clinically matched patients without dyskinesias, before and after administration

of levodopa. Functional connectivity of brain networks in the resting state was assessed in both groups. We chose the right inferior

frontal cortex as the seed region given the evidence highlighting the role of this region in motor control. In a second experiment, we

applied different forms of repetitive transcranial magnetic stimulation over the right inferior frontal cortex in a new group of

dyskinetic patients who were taking a supramaximal dose of levodopa, to verify the clinical relevance of this area in controlling the

development of hyperkinetic movements. The resting state functional imaging analysis revealed that in patients with levodopa-

induced dyskinesias connectivity of the right inferior frontal cortex was decreased with the left motor cortex and increased with the

right putamen when compared to patients without levodopa-induced dyskinesias. This abnormal pattern of connectivity was

evident only during the ON phase of levodopa treatment and the degree of such alteration correlated with motor disability.

The repetitive TMS experiments showed that a session of continuous but not intermittent or sham theta burst stimulation applied

over the inferior frontal cortex was able to reduce the amount of dyskinesias induced by a supramaximal single dose of levodopa,

suggesting that this area may play a key role in controlling the development of dyskinesias. Our combined resting state functional

magnetic resonance and transcranial magnetic stimulation studies demonstrate that pathophysiological mechanisms underlying

levodopa-induced dyskinesias may extend beyond the ‘classical’ basal ganglia dysfunctions model, including the modulation per-

formed by the neural network centred on the inferior frontal cortex.

1 Consiglio Nazionale delle Ricerche, IBFM, 88100, Catanzaro, Italy2 Non-Invasive Brain Stimulation Unit, Neurologia Clinica e Comportamentale, Fondazione Santa Lucia IRCCS, Rome, Italy3 Stroke Unit, Policlinico Tor Vergata, Rome, Italy4 Universita degli studi ‘Magna Graecia’, Clinica Neurologica, 88100, Catanzaro, Italy5 UOC Neurologia, Ospedale Sant’Eugenio, Rome, Italy

doi:10.1093/brain/awu329 BRAIN 2014: Page 1 of 14 | 1

Received June 13, 2014. Revised September 3, 2014. Accepted September 30, 2014.

� The Author (2014). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.

For Permissions, please email: journals.permissions@oup.com

Brain Advance Access published November 19, 2014by guest on N

ovember 20, 2014

Dow

nloaded from

Correspondence to: Antonio Cerasa,

Neuroimaging UNIT, IFBM-CNR, Germaneto (CZ), Italy

E-mail: a.cerasa@unicz.it

Correspondence may also be addressed to: Giacomo Koch, Non-Invasive Brain Stimulation Unit, Neurologia Clinica e

Comportamentale, Fondazione Santa Lucia IRCCS, Rome, Italy, e-mail: g.koch@hsantalucia.it

Keywords: Parkinson’s disease; levodopa-induced dyskinesias; resting state functional MRI; TMS; inferior frontal cortex

Abbreviations: AIMS = Abnormal Involuntary Movement Scale; IFC = inferior frontal cortex; LID = levodopa-induced dyskinesia;SMA = supplementary motor area; STN = subthalamic nucleus; TBS = theta burst stimulation; TMS = transcranial magnetic stimu-lation; UPDRS = Unified Parkinson’s Disease Rating Scale

IntroductionThe classical model of basal ganglia dysfunctions occurring

in Parkinson’s disease highlights the pathological imbalance

between the ‘direct’ and ‘indirect’ pathways in patients with

levodopa-induced dyskinesias (LIDs) (Obeso et al., 2000).

Although this model has been reinforced and modified to

extend knowledge about the interplay between striatal

nuclei and frontal cortex, its application has been shown

to offer an inadequate description of clinical and experi-

mental findings in patients with Parkinson’s disease (Obeso

et al., 2000; Morgante et al., 2006; Jenner, 2008). For this

reason, in recent years several alternative models have been

proposed to better understand the pathophysiological

mechanisms underlying motor fluctuations in Parkinson’s

disease (Carta et al., 2007; Cenci and Konradi, 2010;

Ahmed et al., 2011).

Our recent neuroimaging studies have shed new light on

the pathophysiological mechanisms underlying LIDs

(Cerasa et al., 2011, 2012, 2013a, b, 2014). Indeed, we

demonstrated that patients with Parkinson’s disease with

LIDs are characterized either by dysfunctional coupling be-

tween the right inferior frontal cortex (IFC) and the sup-

plementary motor area (SMA) during a simple motor task

(Cerasa et al., 2012) or by an abnormal increase of grey

matter volume in the IFC (Cerasa et al., 2011). This latter

finding has been confirmed using different population

(Cerasa et al., 2013a) and neuroimaging metrics (Cerasa

et al., 2013b), and has already raised an interesting scien-

tific debate on the toxic effects of levodopa on brain

morphometry (Vernon and Modo, 2012; Cerasa et al.,

2014) as well as on the hypothetical role of this area as

a new target for brain stimulation useful to decrease the

severity of LIDs (Cerasa and Quattrone, 2014a; Obeso and

Strafella, 2014).

The IFC is part of a well-known neural network involved

in response inhibition (Aron and Poldrack, 2006). Previous

structural and functional MRI studies have demonstrated

that this network is right-lateralized and comprised of the:

IFC, SMA/preSMA, primary motor cortex (M1), subthala-

mic nucleus (STN), and striatum (Aron et al., 2007; Li

et al., 2008; Neubert et al., 2010). The idea proposed by

our group, discussed by Aron and Obeso (2012), is that

LIDs may rely on a reduced inhibitory control mediated by

the IFC. What was clearly demonstrated by Aron et al.

(2004) and confirmed by other authors (Forstmann et al.,

2008; Hampshire et al., 2010; Neubert et al., 2010) was

that the ability to control motor responses depends critic-

ally upon the right IFC, which is embedded in a three-way

fronto–basal–ganglia network: the direct, indirect and

hyperdirect pathways (Aron et al., 2014). The direct path-

way is mainly involved in facilitating the initiation and

execution of voluntary movements, while the indirect path-

way has been demonstrated to implement and regulate re-

sponse inhibition (Nambu, 2004; Aron, 2007).

Furthermore, the hyperdirect pathway has been shown

convey powerful direct excitatory effects from the frontal

cortex (including preSMA and IFC) to the STN by passing

the striatum (Nambu et al., 2002) providing a quick route

for influencing the tonic inhibition of the basal ganglia

output pathways for achieving inhibition of inappropriate

or prepotent responses (Jahfari et al., 2011, Cerasa and

Quattrone 2014b).

Overall, what emerged from these studies was that

striato-frontal dysfunctions characterizing Parkinson’s dis-

ease, as well as the neural mechanisms regulating motor

inhibition, might share similar neural pathways. Indeed,

although idiopathic Parkinson’s disease is primarily a dis-

order of response initiation characterized by an excessive

motor inhibition (i.e. akinesia, bradykinesia), LIDs are

clearly a clinical expression of disinhibition of movement.

The presence of deficits in inhibitory control on tasks

requiring inhibition of prepotent motor response has been

widely reported in patients with Parkinson’s disease (Obeso

et al., 2011; Jahanshahi, 2013) and generally considered as

a ‘cognitive’ consequence of the disease. Nonetheless, the

idea that the neural networks subserving motor inhibition

may influence the induction of LIDs is a new intriguing

hypothesis.

In this multimodal work, we performed two independent

resting state functional MRI and theta burst stimulation

(TBS) experiments. In the first study, we aimed at revealing

levodopa-induced changes in the resting state functional

connectivity occurring in patients with Parkinson’s disease

with LIDs. As there are no functional MRI studies investi-

gating pathological changes in the brain of patients with

LIDs during the ON phase of levodopa treatment, we

investigated the presence of different brain connectivity

2 | BRAIN 2014: Page 2 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from

patterns comparing patients with Parkinson’s disease with

and without LIDs. In particular, we were interested in

investigating the pattern of connectivity with the right

IFC. Blood oxygen level-dependent functional MRI signal

recorded while the subject lies at rest with eyes closed rep-

resents an important tool for understanding brain disorders

in patients with Parkinson’s disease (Wu et al., 2011;

Esposito et al., 2013; Kahan et al., 2014). Brain regions

with similar functions have been shown to display func-

tional connectivity during rest, which reflects the presence

of direct and indirect anatomical pathways (Horn et al.,

2014). Next, in a second study using an independent

group of patients with Parkinson’s disease with LIDs, we

tested the reliability of repetitive transcranial magnetic

stimulation (TMS) over the right IFC to evaluate whether

the expected altered connectivity centred on the right IFC

may be considered as a consequence of LIDs or whether

this has a critical role on the aberrant neural mechanisms

contributing to the induction of involuntary movements,

thus improving motor disabilities. Fundamentally, we

hypothesized that: (i) levodopa replacement therapy influ-

ences the communication between the right IFC and brain

regions involved in the primary motor pathways (motor/

premotor cortices, basal ganglia, STN and cerebellum) in

patients with LIDs; and (ii) that IFC TBS has an antidyski-

netic effect in patients with Parkinson’s disease with LIDs,

possibly through the modulation of cortico-cortical path-

ways originating from this area.

Subjects and methods

Subjects

Patients with diagnoses of idiopathic Parkinson’s disease ac-cording to the United Kingdom Parkinson’s Disease SocietyBrain Bank criteria (Hughes et al., 1992) were enrolled inthis study. Inclusion criteria were: (i) a minimum 6-monthduration of levodopa treatment; (ii) the presence or absenceof peak-dose LIDs following an acute levodopa test observedby the examining neurologist on the occasion of the last visit;LIDs were defined as 425% of waking hours (item 32 ofUPDRS 5 2) and bothersome (item 33 5 2); (iii) stable medi-cation dose for 4 weeks; (iv) no evidence of vascular brainlesions, brain tumour or brain atrophy; (v) no evidence ofdementia or depression symptoms according to DSM-IV cri-teria; (vi) no use of antidepressant, anxiolytic or antipsychoticdrugs that could affect cerebral blood flow; (vii) right-handedness; (viii) absence of chronic medical conditions(heart disease, hypertension, or diabetes), and of contraindica-tions for MRI and TBS; (ix) no excessive movement artefactsduring functional MRI experiment; and (x) ability to toleratethe withdrawal of dopaminergic medication before functionalMRI session.

According to these criteria, we enrolled two independentgroups, one for the functional MRI and one for the TBS ex-periment. The characteristics of the patients enrolled for theneuroimaging (primary group) and TBS (secondary group) stu-dies are shown in Table 1.

For the functional MRI study 66 case patients with clinicaldiagnoses of Parkinson’s disease were consecutively selectedamong the outpatients of the Institute of Neurology of the“Magna Graecia” University of Catanzaro, Italy. Among thisgroup of treated patients with Parkinson’s disease, 34 weredyskinetic and 32 were non-dyskinetic. Dyskinetic and non-dyskinetic patients were then individually matched by a com-puter-generated program according to their disease duration(�16 months). A total sample of 12 pairs was then availablefor the analysis. All dyskinetic patients exhibited mild peak-dose choreic dyskinesia. Dyskinesias were predominately dis-tributed in the upper body or lower extremities with hyperkin-etic movements mostly affecting the right side of the body(84%). All patients with Parkinson’s disease were treatedwith levodopa with additional dopamine agonists(Supplementary Table 1). Before the OFF and ON phase func-tional MRI sessions, patients underwent clinical assessments,including Hoehn and Yahr staging, Unified Parkinson’sDisease Rating Scale (UPDRS) and Abnormal InvoluntaryMovement Scale (AIMS).

For the TBS experiments, 11 patients with Parkinson’s dis-ease suffering from disabling peak-dose dyskinesias after levo-dopa ingestion were enrolled at Santa Lucia Foundation inRome. Eight patients were enrolled for the main experimentin which TBS protocols were delivered over the right IFC andeight (five of whom participated in the main experiment) wererecruited for a control experiment in which TBS protocolswere applied over the left M1 (Table 1).

Informed consents and Ethics Committee approvals wereobtained for all patients enrolled in the two experiments, ac-cording to the Declaration of Helsinki.

Neuropsychological assessment

All patients enrolled for the functional MRI experiment com-pleted an extensive series of neuropsychological tests(Passamonti et al., 2013), which were administered by anexperienced clinical neuropsychologist blind to any otherresult. Neuropsychological evaluations were performed at thesame time as the MRI and neurological exams, evaluating thefollowing neuropsychological measures: (i) executive control(Frontal Assessment Battery, Modified Card Sorting Test andStroop task); (ii) attention and working memory (Digit SpanForward and Backward); (iii) short- and long-term verbalmemory (Rey Auditory Verbal Learning Test); (iv) verbal flu-ency and language comprehension (Word List Generation,Token Test); and (v) visual–spatial skills (Judgement of LineOrientation) and general cognitive status (Mini-Mental StateExamination). We further investigated the presence of impul-sivity trait and depressive and anxiety symptoms using theBarratt’s Impulsivity Scale, Beck Depression Inventory andHamilton Rating Scale Anxiety, respectively. Neuropsycholo-gical data are shown in Supplementary Table 2.

Resting state functional MRIexperiment

The resting-state functional MRI run lasted 57 min. Particularcare was taken to restrain the subjects’ movements with cush-ions and adhesive medical tape. All patients were scannedtwice in the same morning immediately before (practical

Disrupted cortico-cortical connectivity in PD with LIDs. BRAIN 2014: Page 3 of 14 | 3

by guest on Novem

ber 20, 2014D

ownloaded from

‘OFF phase’: 12 h after last medication) and �60 min after(‘ON phase’) drug administration, consistent with the expectedtime peaks of LIDs. In accordance with a previous imagingstudy (Pavese et al., 2006), we calculated the approximatetiming of LIDs onset based on patient’s personal medicationhistory. This prescan evaluation was performed for 1 weekbefore conducting the functional MRI experiment. After over-night withdrawal of anti-parkinsonian medications, patientswere examined the morning before and after taking their es-tablished dopaminergic therapy. This procedure helped us todefine the average individual times taken by each dyskineticpatient to switch from an OFF to an ON state. This informa-tion was subsequently used on the day of the experiment tostart functional MRI acquisition at the time immediatelybefore the expected commencement of LIDs. To better evaluatethis condition (and to avoid motion artefacts), patients werealso continuously monitored by clinicians (G.D. and G.M.)inside the scanner room. When the expected timing of patients’responses to levodopa treatment was reached, the functionalMRI session was started (Supplementary Video 1). A physicianremained inside the scanner room to visually observe the emer-gence of involuntary movements. If dopaminergic levelsreached the threshold for triggering peak-of-dose dyskinesias,MRI measurements were immediately stopped. None of thepatients with Parkinson’s disease with LIDs reached thepeak-of-dose dyskinesias during functional MRI session.

The levodopa treatment was executed either at the first orsecond functional MRI session in a counterbalanced orderacross patients. In the OFF phase, individuals were askednot only to interrupt levodopa medication, but also dopamineagonists and/or amantadine drugs.

Resting state functional MRIacquisition and preprocessing

Scanning was performed on a 3 T scanner (Discovery MR750General Electric) with an 8-channel head coil. A gradient-echoecho planar (EPI) T2*-weighted pulse sequence (repetition time/

echo time: 2000/25 ms; number of axial slices = 39; field ofview = 24 mm; thickness/gap = 3/0.8 mm; matrix size = 96 � 96)was used. During the functional scan, subjects were asked tosimply stay motionless, awake and relaxed with their eyesclosed; no visual or auditory stimuli were presented at any timeduring functional scanning.

Data of each functional MRI scan contained 200 timepoints. All functional images were preprocessed using a toolthat is part of the FMRIB’s Software Library (http://www.fmrib.ox.ac.uk/fsl MELODIC). Images were slice-time cor-rected and aligned to the Montreal Neurological Institute’s(MNI) 152 standards space using linear registration with 12degrees of freedom. Afterwards, the images were spatiallysmoothed using a Gaussian kernel of 8 mm full-width athalf-maximum. In addition, high-pass temporal filtering wasapplied to remove very low-frequency drifts and physiologicalhigh-frequency noise (Biswal et al., 1995). Finally, imagingdata were then temporally concatenated across subjects tocreate a single 4D data set for each subject.

Each subject’s head motion parameters were examined. Nosubject’s data set had 42 mm maximum translation in x, y, orz, or 1 mm of maximum rotation about three axes. As correl-ation analysis is sensitive to gross head motion, we character-ized the head motion as an index representing the peakdisplacement that was determined by using a previously vali-dated method applied on patients with Parkinson’s disease(Wu et al., 2011). Specifically, we calculated the Euclidian dis-tance travelled by each subject’s head from the first to the lastscan (i.e. combining the movements in the three axes into onevalue), separately for translation (x, y and z direction) and forrotation (pitch, roll and yaw). Then, a two-way ANOVA testwas performed on the peak displacement to examine between-group differences in head motion. The interaction effects ofgroup (patients with LIDs versus patients without LIDs) andlevodopa phase (OFF versus ON) did not reveal significantdifferences either for the six motion parameters (translationand rotation) or for the calculation of Euclidian distance(Supplementary Fig. 1). However, we detected some significantdifferences investigating the main effects of levodopa phase.

Table 1 Demographic and clinical data of Parkinson’s disease patients enrolled for the functional MRI (primary

group) and the repetitive TMS experiment (secondary group)

Primary group Secondary group

Variables PD with LIDs PD without LIDs P-values PD with LIDs

Gender (female/male) 3/9 4/8 0.98a 6/5

Age (years) 68.1 � 7.4 67.1 � 7.3 0.81b 69 � 8.25

Age at onset (years) 61 � 7.5 61.8 � 7.1 0.81b 59.6 � 9.29

Disease duration (years) 7 � 2.1 5.5 � 3.6 0.2b 11 � 5.2

Hoehn and Yahr Stage 2 (2– 2–5) 2 (2–3) 0.41c 2 (2– 2–5)

UPDRS-ME OFF drug 29.4 � 7.5 26.4 � 4.7 0.27b 25.3 � 6.4

UPDRS-ME ON drug 22.6 � 7.4 19.7 � 4.5 0.26b 20.2 � 8.1

Mean dose levodopa (mg/die) 536.4 � 231.4 383.4 � 99.4 0.11b 790.6 � 253.3

AIMS 12 � 2.9 - - 7.9 � 1.8

MMSE 26.1 � 1.8 25.8 � 1.6 0.86b 25.9 � 1.7

Data are given as mean values � SD, or median values (range) when appropriate.aChi-square test.bTwo-sample t-test.cMann-Whitney test.

PD = Parkinson’s disease; UPDRS-ME = Unified Parkinson’s Disease Rating Scale Motor Examination; MMSE = Mini-Mental State Examination; AIMS = Abnormal Involuntary

Movement Scale.

4 | BRAIN 2014: Page 4 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from

Specifically, all patients with Parkinson’s disease in the ONphase moved more than patients in the OFF phase, both forthe translation (z direction; P = 0.003) and for rotation (pitchparameter; P = 0.002).

Functional connectivity analysis

We defined the right IFC as region of interest. The MNI co-ordinates for this seed were derived from our past functionalMRI study (x: 42; y: 32; z: 12) where we detected alteredactivity in LID patients during motor task (Cerasa et al.,2012). Next, a 5-mm sphere around this coordinate was cre-ated and used to extract data for functional connectivity ana-lysis in the resting state. A seed reference time course wasobtained by averaging the time courses. Correlation analysiswas carried out between the seed reference and the wholebrain in a voxel-wise manner. In the correlation analysis, thesix head motion parameters were included as nuisance covari-ates. We did not apply global signal regression to avoid theintroduction of artificial negative correlations that may resultfrom removal of the global signal (Murphy et al., 2009; Wuet al., 2011). The individual correlation coefficient (r) mapswere converted into z-maps by Fisher’s Z-transform to im-prove the normality of the data (Press et al., 1992; Loweet al., 1998). The individual z-value was entered into arandom effect one-sample t-test in a voxel-wise manner to de-termine brain regions showing significant connectivity witheach ‘seed’ in the resting state within each group. Then, toassess the main/interaction effects of group (Parkinson’s dis-ease with LIDs versus patients without LIDs) and levodopaphase (OFF versus ON) in connectivity with the IFC, asecond-level ANOVA model was used.

In the current study, positive connectivity means that thespontaneous signal fluctuations in brain networks are in-phase with the fluctuations observed in the correspondingseed region; whereas negative connectivity means that thespontaneous signal fluctuations are antiphase related with thefluctuations observed in the corresponding seed region (Foxet al., 2005). A family-wise error (FWE) corrected thresholdof P5 0.05 within regions of interest was used for eitherwithin or between group results. Based on previous findings,we decided to use the left IFC, bilateral SMA/preSMA, M1,STN, thalamus, striatum and cerebellum as regions of interestgiven their critical engagement in LIDs-related pathologicalmechanisms (Rascol et al., 1998; Brooks et al., 2000; Obesoet al., 2000; Cerasa et al., 2014) and motor inhibition process-ing (Aron et al., 2014). In addition, to explore whether theinteractions of brain networks correlate with the disease sever-ity, a correlation analysis of functional connectivity resultsagainst the AIMS scores was performed in patients withParkinson’s disease with LIDs.

Theta burst stimulation experiments

In this experiment patients had fasted and were in withdrawalof therapy since the previous night (CAPIT). After 12 h ofovernight medication withdrawal, patients received 125% oftheir usual morning levodopa equivalent dose as immediaterelease levodopa/carbidopa (Koch et al., 2009). The assessmentin each video-recorded session consisted of a complete UPDRSIII and AIMS assessment. The degree of LIDs was assessedindividually in neck, trunk, right upper limb, right lower

limb, left upper limb and left lower limb and scored as follows:0, none; 1, slight; 2, mild; 3, moderate; and 4, severe. Overallseverity was expressed by the sum of scores in all regions(0–24). Following levodopa administration, UPDRS III andAIMS were calculated every 15 min for 1 h (t0, t15, t30, t45,t60). A complete evaluation lasted �3–4 min at t0. In threeseparate sessions performed 1 week apart continuous TBS,intermittent TBS or sham continuous TBS were applied overthe right IFC, immediately after levodopa administration. Theorder of the sessions was pseudo-randomized across patients.Two blinded raters (L.B. and M.S.B.), experts in the field ofmovement disorders, rated videotapes independently to pro-vide AIMS and UPDRS scores. Both video-raters were blindedto repetitive TMS, and the scores were generated after a con-sensus was reached comparing the individual scores. To ac-count for the variability of levodopa effect, particularly onLIDs, the three sessions were performed in consecutive weeksat the same hour of the day, and in the same pharmacologicalconditions.

A MagStim Super Rapid magnetic stimulator (MagstimCompany) connected with a figure of 8 coil with a diameterof 70 mm was used to deliver repetitive TMS over the scalpsite corresponding to the right IFC. The magnetic stimulus hada biphasic waveform with a pulse width of �300 ms. Duringthe first phase of the stimulus, the current in the centre of thecoil flowed toward the handle. For continuous TBS, threepulse bursts at 50 Hz were repeated every 200 ms for 40 s(Huang et al., 2005) over the right IFC (600 pulses). For inter-mittent TBS, three pulses bursts at 50 Hz were repeated at arate of 5 Hz; there were 20 trains of 10 bursts given at 8-sintervals (600 pulses in total). Sham stimulation was deliveredthrough a 7 cm figure of 8 focal coil angled at 90� with onlythe edge of the coil resting on the scalp. In every protocol,repetitive TMS were delivered at 80% of the active motorthreshold (Huang et al., 2005). Active motor threshold forthe first dorsal interosseus was tested over the right motorcortex. The coil position for the right IFC was defined andmonitored throughout the experiment with the SofTaxicNavigator system (Electro Medical Systems). The positionwas precisely determined using individual structural MRI tolocalize the sites and adjusted with respect to individualsulcal landmarks in each participant. To target the IFC thecoil was positioned over the caudal portion of the pars oper-cularis of the inferior frontal cortex, with the handle pointingforward. The mean Talairach coordinates used in the presentstudy to target the IFC were x = 50.9 � 3.1; y = 29.6 � 14.5;z = 19.4 � 3.3.

The IFC TMS paradigm was completed by an additionalcontrol experiment useful to test for the specificity of theabove described results by applying the same protocol overthe left M1 in a sample of eight patients with Parkinson’sdisease with LIDs. For the left M1 TBS the coil was positionedover the hotspot of the right first dorsal interosseous (Huanget al., 2005).

Statistical analysis

Statistical analyses of functional MRI data were performedwith STATISTICA Version 6.0 (www.statsoft.com).Assumptions for normality were tested for all continuous vari-ables by using the Kolmogorov–Smirnov test. In the first ex-periment, all variables were normally distributed, except for

Disrupted cortico-cortical connectivity in PD with LIDs. BRAIN 2014: Page 5 of 14 | 5

by guest on Novem

ber 20, 2014D

ownloaded from

Hoehn and Yahr, and Modified Card Sorting Test scores.Otherwise, in the second experiment clinical scores were notnormally distributed. The ANOVA, unpaired t-test, Mann–Whitney U-test and �2 were used appropriately. All statisticalanalyses had two-tailed alpha levels of 50.05 for definingsignificance. For the repetitive TMS experiment, non-paramet-ric Wilcoxon tests were applied on mean AIMS and UPDRSscores for each session. For all statistical analyses, a P-value of50.05 was considered to be significant.

Results

Functional connectivity of the rightIFC in each group

Before levodopa intake during the resting state, functional

MRI analysis showed that the right IFC of patients with

Parkinson’s disease with LIDs showed significant positive

connectivity with the left IFC, preSMA, bilateral putamen,

bilateral thalamus and bilateral STN and had significant

negative connectivity with the bilateral anterior cerebellum

(one sample t-test, P50.05, FWE corrected; Table 2).

Interestingly, in the ON phase the overall pattern of con-

nectivity remained the same, except for the left M1. Indeed,

levodopa replacement modulated specifically the connectiv-

ity between the right IFC and the left M1, which resulted to

be negatively correlated.

In the OFF phase, the right IFC of patients with

Parkinson’s disease without LIDs showed a different pat-

tern of connectivity from that detected in LIDs patients.

Indeed, significant positive connectivity was only detected

with the left IFC, SMA, right putamen and bilateral STN,

together with significant negative connectivity with the bi-

lateral anterior cerebellum. Interestingly, levodopa replace-

ment produced an overall enhancement of functional

connectivity. Indeed, apart from the regions described

above, in the ON phase patients without LIDs showed add-

itional significant positive connectivity with the preSMA,

bilateral M1, and left putamen, while the connectivity

with the right cerebellum disappeared.

Functional connectivity of the rightIFC: the main/interactive effects ofgroup and levodopa phase

Investigating the main effect Group (independent of levo-

dopa phase) as well the main effect of Levodopa phase

(independent of group), no significant functional change

was detected. In other words, the connectivity of the right

IFC was similar between Parkinson’s disease with or with-

out LIDs or between OFF and ON phases. When we inves-

tigated the interaction effects, two significant clusters

emerged. In patients with LIDs during the ON phase, the

right IFC showed significant negative connectivity with the

left M1 (MNI local maxima: x = �6, y = �28, z = 78;

F = 19.24; PFWE-ROI = 0.01) and positive connectivity with

the right putamen (MNI local maxima: x = 28, y = 4, z = 8;

F = 16.44; PFWE-ROI = 0.02). This pattern of connectivity

was totally inverted in patients with Parkinson’s disease

without LIDs (Fig. 1).

Correlation analysis

Correlation analysis between connectivity t-score and the

AIMS scores in brain areas that showed significant connec-

tion with the right IFC revealed significant relationships

with the same brain regions detected in the ANOVA ana-

lysis. In fact, in patients with LIDs during the ON phase we

detected a negative correlation in the left M1 (MNI local

maxima: x = �12, y = �24, z = 80; T-value = 4.15; PFWE-

ROI = 0.03) and a positive correlation in the bilateral puta-

men (MNI local maxima: x = 12, y = �10, z = �8; T-

value = 4.82; PFWE-ROI = 0.002; MNI local maxima:

x = �10, y = �12, z = �6; T-value = 4.21; PFWE-

ROI = 0.005 for the right and left hemispheres, respectively).

Moreover, extending this analysis to other regions of inter-

est, we found an additional significant association between

increasing motor disability and increasing brain connectiv-

ity of the IFC with the bilateral STN (MNI local maxima:

x = 16, y = 10, z = �2; T-value = 6.82; PFWE-ROI = 0.01;

MNI local maxima: x = �10, y = 8, z = �4; T-

value = 5.83; PFWE-ROI = 0.03 for the right and left hemi-

spheres, respectively) (Fig. 2).

Theta burst stimulation data

The procedure (Fig. 3) was well tolerated by all subjects.

The mean patients’ active motor threshold taken with the

MagStim Super Rapid magnetic stimulator was

46.2 � 8.5% of maximum stimulator output. Mean

UPDRS and AIMS score did not differ at baseline in the

three (continuous TBS, intermittent TBS, sham) conditions,

moreover the analysis of AIMS and UPDRS scores for each

body segment separately showed highly reproducible distri-

bution in each subject in the three sessions.

Wilcoxon Test revealed that motor abilities scored by the

mean UPDRS section III were not modified by any TBS

condition (Fig. 4A), whereas following continuous TBS,

but not intermittent TBS, global AIMS scores were

decreased in comparison with sham continuous TBS. This

was evident at t30 (4.50 � 1.30 versus 6.33 � 1.42;

P = 0.04) (Fig. 4B) following levodopa administration.

When analysing LIDs separately for left and right body

side we did not find significant differences (all P’s4 0.05)

(Fig. 4C and D). Finally, when TBS trains were applied

over the left M1 we did not find any effect on AIMS and

UPDRS scores (Fig. 4E and F).

DiscussionThe present study was designed to determine whether the

neural network centred on the right IFC plays a critical role

6 | BRAIN 2014: Page 6 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from

Table 2 Brain ‘target’ regions showing significant changes in connectivity with the right inferior frontal cortex (seed)

during either OFF or ON phases of levodopa treatment or in the two Parkinson’s disease groups (with/without

LIDs), independently (one-sample t-tests).

Brain regions Connectivity direction PFWE-values T local maxima MNI coordinates

x y z

Parkinson’s disease with LIDs in OFF phase

Left IFC Positive 50.001 11.20 �40 34 6

Pre-SMA Positive 0.04 6.36 0 32 48

SMA n.s. n.s. n.s.

Left M1 n.s. n.s. n.s.

Right M1 n.s. n.s. n.s.

Left putamen Positive 50.001 12.29 �30 4 �2

Right putamen Positive 0.008 7.38 22 12 8

Left thalamus Positive 0.01 6.46 �8 �6 �2

Right thalamus Positive 0.02 6.06 16 �10 4

Left STN Positive 50.001 6.26 �10 �8 �6

Right STN Positive 0.005 4.18 14 �14 4

Left CB Negative 0.02 7.61 �18 �44 �12

Right CB Negative 0.006 8.79 8 �40 �16

Parkinson’s disease with LIDs in ON phase

Left IFC Positive 0.001 9.34 �44 36 8

Pre-SMA Positive 0.01 7.31 6 38 46

SMA n.s. n.s. n.s.

Left M1 Negative 0.03 6.93 �15 �25 71

Right M1 n.s. n.s. n.s.

Left putamen Positive 0.04 12.29 �10 8 �4

Right putamen Positive 0.01 6.7 18 12 �2

Left thalamus Positive 0.04 5.7 �8 �6 0

Right thalamus Positive 0.04 5.47 10 �6 2

Left STN Positive 0.006 4.06 �12 �10 �4

Right STN Positive 0.003 4.47 12 �10 �8

Left CB Negative 0.01 7.99 �18 �54 �34

Right CB Negative 0.01 7.87 20 �58 �38

Parkinson’s disease without LIDs in OFF phase

Left IFC Positive 0.006 7.57 �42 36 8

Pre-SMA n.s. n.s. n.s.

SMA Positive 0.03 6.49 12 18 46

Left M1 n.s. n.s. n.s.

Right M1 n.s. n.s. n.s.

Left putamen n.s. n.s. n.s. n.s. n.s. n.s.

Right putamen Positive 0.004 8.08 22 22 �8

Left thalamus n.s. n.s. n.s.

Right thalamus n.s. n.s. n.s.

Left STN Positive 0.006 4.07 �10 �12 �6

Right STN Positive 0.005 4.22 12 �12 �4

Left CB Negative 0.01 8.14 �10 �70 �12

Right CB Negative 0.005 9.07 16 �53 �12

Parkinson’s disease with LIDs in ON phase

Left IFC Positive 0.004 7.93 �42 36 8

Pre-SMA Positive 0.006 8.32 �8 44 18

SMA Positive 0.02 6.71 2 22 44

Left M1 Positive 0.03 8.29 �16 �33 70

Right M1 Positive 0.01 8.73 15 �32 72

Left putamen Positive 0.002 12.29 �22 18 0

Right putamen Positive 0.003 8.4 18 16 �4

Left thalamus n.s. n.s. n.s.

Right thalamus n.s. n.s. n.s.

Left STN Positive 0.003 4.49 �10 �12 �6

Right STN Positive 0.001 5.19 12 �12 �4

Left CB Negative 0.009 8.27 �20 �36 �20

Right CB n.s. n.s. n.s.

CB = cerebellum; n.s. = not significant.

Disrupted cortico-cortical connectivity in PD with LIDs. BRAIN 2014: Page 7 of 14 | 7

by guest on Novem

ber 20, 2014D

ownloaded from

in the pathophysiological mechanisms of LIDs. To this end,

we performed two different experiments employing two in-

dependent groups of Parkinson’s disease with the aim: (i) to

explore the dopaminergic-induced modulation of the rest-

ing state neural connectivity in patients with Parkinson’s

disease with and without LIDs following levodopa admin-

istration; and (ii) to determine the influence of TBS proto-

col applied over the IFC in reducing hyperkinetic motor

disorders. Our functional MRI study revealed that patients

with LIDs are characterized by altered patterns of func-

tional connectivity between the right IFC and the contro-

lateral M1 and the ipsilateral putamen, which emerged

when switching from the OFF to the ON phase of levodopa

treatment. Although this first set of findings provides a sig-

nificant advance in understanding the neural basis of LIDs,

a fundamental conundrum persists: do these abnormal

functional changes have a critical role in the neural net-

work contributing to the manifestation of these involuntary

movements or should these only be considered as a conse-

quence of having long-term hyperkinetic movements (Aron

and Obeso, 2012)? To answer this critical question we per-

formed an additional neurophysiological experiment evalu-

ating the effect of brain stimulation over this prefrontal

region in a group of dyskinetic patients who took a supra-

maximal dose of levodopa. Hypothetically, the lack of sig-

nificant effects in ameliorating motor symptoms should

confirm the latter hypothesis. Instead, in accordance with

our a priori hypothesis, patients with LIDs showed a re-

duction of motor disability (AIMS scores) induced by TBS

stimulation. Thus, our multimodal neurophysiological

study clearly demonstrated that a network centred on the

right IFC is critically involved in LIDs.

The new pathophysiologicalhypothesis of LIDs: functional MRIevidence

The first piece of functional MRI data relates to the pres-

ence of reduced communication between the IFC and the

M1 in patients with Parkinson’s disease with LIDs after

levodopa medication (Fig. 1). Although, the significance

of negative correlations between brain regions in functional

connectivity analysis remain unclear, it has been suggested

that they arise between functional systems with apparently

opposite goals or may reflect antagonistic influences of one

network on another (Fox et al., 2005; Kelly et al., 2008).

With this in mind, we speculate that this pattern of altered

connectivity characterizing patients with Parkinson’s dis-

ease with LIDs in the ON phase may reflect a pathological

inhibitory control of the IFC over the primary motor

output pathway. Consistent with this view, a series of

neurophysiological studies have investigated the meaning

of the inhibitory influence of the right IFC over the M1

Figure 1 Seed-based analysis of the functional connectivity with the right IFC (blue circle on the left) in the resting state. We depict a significant

(FWE5 0.05) Group � Levodopa phase interaction effect in the connectivity between the right IFC with the left primary motor cortex and right

putamen. In particular, we find a reduced communication between the right IFC and the left motor cortex in the patients with Parkinson’s disease

with LIDs during the ON phase that is coupled with an increase in connectivity with the right putamen. Otherwise, levodopa treatment in patients

with Parkinson’s disease without LIDs would seem to restore and normalize the functional connectivity of the IFC. The colour bar represents

F-statistics. Dysk = patients with Parkinson’s disease with levodopa-induced dyskinesias; NoDysk = patients with Parkinson’s disease without

levodopa-induced dyskinesias; BOLD = blood oxygen level-dependent; IFC = functional connectivity.

8 | BRAIN 2014: Page 8 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from

during action control in healthy subjects (Aron et al., 2007;

Verbruggen and Logan, 2008; Buch et al., 2010; Neubert

et al., 2010; Vossel et al., 2011). Overall, inhibitory effects

have mainly been reported during stopping of motor re-

sponses (i.e. Go/NoGo tasks), whereas during a normal

action selection, a facilitatory effect of the right IFC on

the left M1 was reported (van Campen et al., 2013).

Although we have investigated patients in a condition of

rest, the fact that levodopa administration increased the

connectivity between the IFC and the M1 in patients with-

out LIDs characterized by excessive motor inhibition

(i.e. bradykinesia), suggests the presence of facilitatory ef-

fects or normalization of cortical functional connectivity. In

contrast, the decreased communication between the IFC

and M1 in patients where drug treatment induces disinhib-

ition of movements might indicate a tentative role of the

IFC to inhibit the M1 activity and output. Another aspect

of this finding is worth noting, namely the altered

interhemispheric (right versus left) communication between

the IFC and the M1. Several TMS and functional MRI

studies demonstrated that response inhibition is imple-

mented via functional and synchronized activities in the

IFC-SMA/preSMA-striatum-M1 pathway (Swann et al.,

2009; Neubert et al., 2010; Zandbelt et al., 2013). In

fact, converging lines of evidence confirmed that the right

IFC exerts motor control over the M1 either via cortico-

basal ganglia pathways including the striatum (Aron and

Poldrack, 2006), or via cortico-cortical pathways through

the SMA/preSMA (Zandbelt et al., 2013). Moreover, a

recent work combining TMS with DTI provided direct

evidence that there are multiple routes involved in right

IFC/left M1 interactions during action inhibition: a subcor-

tical route (right IFC!globus pallidus!STN!M1) being

used at longer latencies and a direct cortical route

(IFC!M1) at shorter latencies (Neubert et al., 2010).

However, what clearly emerged from these studies was

Figure 2 IFC functional connectivity relates to individual difference in AIMS scores. This figure demonstrates that the degree of

communication (t-scores, y-axis) between the right IFC with the left M1, bilateral putamen and STN, correlates with increasing motor disability

(AIMS scores, x-axis) during the ON phase in LIDs patients. Correlation line and confidence intervals are shown in red.

Disrupted cortico-cortical connectivity in PD with LIDs. BRAIN 2014: Page 9 of 14 | 9

by guest on Novem

ber 20, 2014D

ownloaded from

that the inhibitory control exerted by the right IFC over the

M1 is bilateral. It is possible that our current functional

MRI data showed an altered interhemispheric (right versus

left) communication between the IFC and the M1 because

of predominant hyperkinetic movements on the right side

of the body among patients with LIDs.

Another intriguing piece of evidence was the increased

connectivity between the right IFC and the right putamen

in patients with LIDs relative to patients without LIDs.

Several TMS and DTI studies demonstrated tight connec-

tions between these regions. King et al. (2012) found that

some specific fibre tracts are correlated to motor perform-

ance during a response inhibition task: tracts connecting

the pre-SMA/SMA with the striatum and other tract linking

the IFC with the striatum/STN. Moreover, Zandbelt et al.

(2013) demonstrated that repetitive TMS over the right IFC

and SMA was associated with shorter stop-signal reaction

time and increased M1 deactivation together with increased

right putaminal activation during a stop inhibition func-

tional MRI task. For this reason, consistently with our

previous hypothesis, we speculate that the increased com-

munication between the right IFC and the putamen in pa-

tients with LIDs might reflect a (failed) tentative of the

motor inhibition network to modulate another fronto-stri-

atal network (direct pathway?) involved in neurodegenera-

tive processes of LIDs.

The new pathophysiologicalhypothesis of LIDs: transcranialmagnetic stimulation evidence

We tested different inhibitory and facilitatory forms of IFC

TBS in comparison with sham stimulation. We found a

selective decrease in the development of LIDs after a supra-

maximal dose of levodopa that was evident after right IFC

continuous TBS, but not after intermittent TBS. The con-

tinuous TBS protocol is known to induce long-term depres-

sion-like after effects when applied over the primary motor

cortex (Huang et al., 2005). Therefore we can hypothesize

Figure 3 Overview of the stimulation paradigm. (A) The mean location for IFC coil positioning. (B) Schematic representation of experimental

procedure. In each session repetitive TMS was applied in different conditions immediately after L-DOPA ingestion. The assessment in each session

consisted of a complete UPDRS and AIMS calculated every 15 min for 1 h (t0, t15, t30, t45, and t60) after L-DOPA ingestion. For continuous TBS,

three pulse bursts at 50 Hz were repeated every 200 ms for 40 s (Huang et al., 2005) over the right IFC (600 pulses). For intermittent TBS, three

pulses bursts at 50-Hz were repeated at a rate of 5 Hz; there were 20 trains of 10 bursts given at 8-s intervals (600 pulses in total). Sham

stimulation was delivered through a 7 cm figure of 8 focal coil angled at 90� with only the edge of the coil resting on the scalp. cTBS = continuous

TBS; iTBS = intermittent TBS.

10 | BRAIN 2014: Page 10 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from

that the reduction of LIDs observed in our sample of pa-

tients with Parkinson’s disease could involve similar neuro-

physiological mechanisms within the IFC-related cortical

circuits. With this perspective, the reduction of LIDs after

inhibitory trains of continuous TBS is in line with previous

evidence suggesting that during motor fluctuations there is

an overexcitability of the cortical motor network (Rascol

et al., 1998; Brooks et al., 2000). This overactivity of the

motor cortical regions has been demonstrated to be (par-

tially) counteracted by the application of inhibitory repeti-

tive TMS trains over the SMA (Koch et al., 2005, 2009;

Brusa et al., 2006). The fact that we demonstrated a similar

effect also modulating the activity of the IFC further

confirmed the evidence that the IFC and SMA regions act

together for modulating the activity of motor circuits (Aron

et al., 2014). However, we also provided evidence on the

lack of anti-dyskinetic effects when inhibitory repetitive

TMS trains were applied over the M1 in patients with

LID. This is not surprising as a previous pilot study

(Wagle-Shukla et al., 2007) did not demonstrate a clear

benefit in improving dyskinesias when repetitive TMS was

applied over the M1. For this reason, considering the

abovementioned SMA TMS experiments (Koch et al.,

2005; Brusa et al., 2006) together with previous neuroima-

ging functional/structural studies (Cerasa et al., 2011,

2012, 2013a) describing neural abnormalities in the IFC

Figure 4 TBS Results. Top: Mean UPDRS (A) and AIMS (B) scores in the group of patients with Parkinson’s disease with LIDs tested in three

different conditions: sham stimulation, intermittent TBS and continuous TBS over the right IFC. Middle: Distribution of the AIMS scores for left

(C) and right (D) body. Bottom: Mean UPDRS (E) and AIMS (F) scores detected when TBS trains were applied over the left M1. Error bars

indicate SEM. *P5 0.05. cTBS = continuous TBS; iTBS = intermittent TBS; rIFC = right IFC.

Disrupted cortico-cortical connectivity in PD with LIDs. BRAIN 2014: Page 11 of 14 | 11

by guest on Novem

ber 20, 2014D

ownloaded from

and SMA (but not in the M1), we may conclude that these

prefrontal regions are critical hubs in the brain networks

underlying LIDs mechanisms.

On the basis of the current findings we may extend the

classical vision of LIDs pathophysiology proposing that

these motor symptoms are influenced by altered function-

ality of the neural network deputed to control of motor

inhibition. Within this model, the above described over-

activity of the frontal cortical motor network in LIDs

could lead to a disruption of the coherent signalling related

to the control of involuntary movement. IFC continuous

TBS could reduce LIDs directly by decreasing the increased

state of IFC activity through long-term depression-like

mechanisms or indirectly by rebalancing the altered pattern

of connectivity observed in the resting state functional MRI

experiment. However, it is important to note that TBS of

the IFC not only induces a local modulation of the neural

activity, but also affects the most prominent cortical con-

nections of the stimulated areas, causing widespread

changes in network connectivity (Gratton et al., 2013).

Hence it is likely that IFC TBS could have induced a cas-

cade of neurophysiological changes in the connectivity with

all brain regions involved in primary motor circuit and

functionally related to the activity of the right IFC, such

as: preSMA/SMA, M1, striatum and cerebellum.

Future applications of functionalneuroimaging in patients withParkinson’s disease with LIDs

It is important to note that MRI-based methods have in-

trinsic limitations when applied on patients with LIDs,

namely the contamination of functional data due to

motion artefacts. Although we have addressed this issue

using several strategies, we acknowledge that our procedure

to investigate the ON state of patients with LIDs immedi-

ately before the expected time peaks of plasma levodopa

levels (necessary to avoid motion artefacts) might introduce

between-subject variability in clinical characteristics. For

this reason, we have started a new protocol for improving

the investigation of functional changes following levodopa

ingestion in patients with LIDs using functional near infra-

red spectroscopy. This technique has a higher ecological

validity and is unique in visualizing cortical brain activation

during gross movements, and may be useful to detect

in vivo neural changes underlying the emergence of LIDs.

We retain that with this neuroimaging method we could

provide unquestionable evidence on the role of the IFC in

pathophysiological mechanisms of LIDs.

Future applications of TMS protocolin patients with Parkinson’s diseasewith LIDs

It is possible that the above described alterations of the

neural network deputed to the control of movement

inhibition could imply a decrease of the inhibitory top-

down control of the right IFC on the controlateral M1.

This hypothesis could be tested by adopting multifocal

TMS protocols in the context of movement inhibition

tasks (Koch et al., 2006) aimed at assessing eventual

task-related changes in the inhibitory influences of the

IFC on M1 in patients with LIDs. Moreover, combined

TMS/EEG protocols (Veniero et al., 2013) could measure

eventual changes in the coherence of oscillatory activity

related to movement inhibition. The effects of IFC continu-

ous TBS could be associated with changes in local beta-

band coupling, which may constitute a risk factor for the

development of motor response fluctuations or dyskinesias

(Stoffers et al., 2008). Finally, the clinical impact of pro-

longed sessions of continuous TBS when applied over the

right IFC in patients with Parkinson’s disease with LIDs

remains to be established. We recently reported that con-

tinuous TBS of the lateral cerebellum induced a reduction

of LIDs that lasted up to 1 month after the end of a 2-week

period of daily stimulation (Koch et al., 2009). It is possible

that longer sessions of continuous TBS of the IFC could

induce similar long-lasting beneficial effects; hence, further

studies could be important to show if the clinical improve-

ment is associated with changes in the structural and func-

tional network centred on the right IFC.

ConclusionIn sum, we provide novel evidence that patients with

Parkinson’s disease with LIDs are characterized by func-

tional abnormalities of a cortical–subcortical network

centred on the right IFC and that non-invasive stimulation

of this area counteracts the development of LIDs after levo-

dopa administration. Although neuroimaging and electro-

physiological measures have been conducted on different

groups of patients, our multimodal findings enlarge current

knowledge on the pathophysiological mechanisms underly-

ing LIDs, moving from classical striato-motor circuitry to-

wards a broader cortico–cortical network. We retain that

this study will open new scenarios on the development of

therapeutic protocols for inducing persistent beneficial clin-

ical effects in patients with Parkinson’s disease with LIDs.

FundingThis study was supported by grants of MIUR (Ministero

Universita e Ricerca; PON 01_01180) to Prof. Aldo

Quattrone and by grants of the Italian Ministry of Health

to Dr Giacomo Koch (GR-2009-1591859).

Supplementary materialSupplementary material is available at Brain online.

12 | BRAIN 2014: Page 12 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from

ReferencesAhmed I, Bose SK, Pavese N, Ramlackhansingh A, Turkheimer F,

Hotton G, et al. Glutamate NMDA receptor dysregulation in

Parkinson’s disease withdyskinesias. Brain 2011; 134 (Pt4): 979–86.

Aron AR. The neural basis of inhibition in cognitive control.

Neuroscientist 2007; 13: 214–28.

Aron AR, Robbins TW, Poldrack RA. Inhibition and the right inferior

frontal cortex. Trends Cogn Sci 2004; 8: 170–7.

Aron AR, Poldrack RA. Cortical and subcortical contributions to stop

signal response inhibition: role of the subthalamic nucleus.

J Neurosci 2006; 26: 2424–33.

Aron AR, Behrens TE, Smith S, Frank MJ, Poldrack RA. Triangulating

a cognitive control network using diffusion-weighted magnetic res-

onance imaging (MRI) and functional MRI. J Neurosci 2007; 27:

3743–52.

Aron AR, Obeso J. Is executive control used to compensate for invol-

untary movements in levodopa-induced dyskinesia? Mov Disord

2012; 27: 339–40.

Aron AR, Robbins TW, Poldrack RA. Inhibition and the right inferior

frontal cortex: one decade on. Trends Cogn Sci 2014; 18: 177–85.

Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity

in the motor cortex of resting human brain using echo-planar MRI.

Magn Reson Med 1995; 34: 537–41.

Brooks DJ, Piccini P, Turjanski N, Samuel M. Neuroimaging of dys-

kinesia. Ann Neurol 2000; 47 (Suppl 1): S154–8.Brusa L, Versace V, Koch G, Iani C, Stanzione P, Bernardi G, et al.

Low frequency rTMS of the SMA transiently ameliorates peak-

dose LID in Parkinson’s disease. Clin Neurophysiol 2006; 117:

1917–21.

Buch ER, Mars RB, Boorman ED, Rushworth MFS. A network cen-

tered on ventral premotor cortex exerts both facilitatory and inhibi-

tory control over primary motor cortex during action

reprogramming. J Neurosci 2010; 30: 1395–401.

Carta M, Carlsson T, Kirik D, Bjorklund A. Dopamine released from

5-HT terminals is the cause of L-DOPA-induced dyskinesia in par-

kinsonian rats. Brain 2007; 130: 1819–33.

Cenci MA, Konradi C. Maladaptive striatal plasticity in L-DOPA-

induced dyskinesia. Prog Brain Res 2010; 183: 209–33.Cerasa A, Messina D, Pugliese P, Morelli M, Lanza P, Salsone M,

et al. Increased prefrontal volume in PD with levodopa-induced dys-

kinesias: a voxel-based morphometry study. Mov Disord 2011; 26:

807–12.

Cerasa A, Pugliese P, Messina D, Morelli M, Gioia MC, Salsone M,

et al. Prefrontal alterations in Parkinson’s disease with levodopa-

induced dyskinesia during fMRI motor task. Mov Disord 2012;

27: 364–71.

Cerasa A, Salsone M, Morelli M, Pugliese P, Arabia G, Gioia CM,

et al. Age at onset influences neurodegenerative processes underlying

PD with levodopa-induced dyskinesias. Parkinsonism Relat Disord

2013a; 19: 883–8.

Cerasa A, Morelli M, Augimeri A, Salsone M, Novellino F, Gioia MC,

et al. Prefrontal thickening in PD with levodopa-induced dyskin-

esias: new evidence from cortical thickness measurement.

Parkinsonism Relat Disord 2013b; 19: 123–5.

Cerasa A, Fasano A, Morgante F, Koch G, Quattrone A. Maladaptive

plasticity in levodopa-induced dyskinesias and tardive dyskinesias:

old and new insights on the effects of dopamine receptor pharma-

cology. Front Neurol 2014; 95: 49.

Cerasa A, Quattrone A. May Stimulation of the PreSMA Become a

new therapeutic target for PD patients with Levodopa-induced

Dyskinesias? Brain Stimul 2014a; 7: 335–6.

Cerasa A, Quattrone A. May hyperdirect pathway be a plausible

neural substrate for understanding the rTMS-related effects on PD

patients with Levodopa-induced Dyskinesias? Brain Stimul 2014b;

7: 488–9.

Esposito F, Tessitore A, Giordano A, De Micco R, Paccone A,

Conforti R, et al. Rhythm-specific modulation of the sensorimotor

network in drug-naive patients with Parkinson’s disease by levo-

dopa. Brain 2013; 136 (Pt 3): 710–25.

Forstmann BU, Jahfari S, Scholte HS, Wolfensteller U, van den

Wildenberg WP, Ridderinkhof KR. Function and structure of the

right inferior frontal cortex predict individual differences in response

inhibition: a model-based approach. J Neurosci 2008; 28: 9790–6.

Fox MD, Snyder AZ, Vincent JL, Corbetta M, Van E, Raichle ME.

The human brain is intrinsically organized into dynamic, anti-corre-

lated functional networks. Proc Natl Acad Sci USA 2005; 102:

9673–8.

Gratton C, Lee TG, Nomura EM, D’Esposito M. The effect of theta-

burst TMS on cognitive control networks measured with resting

state fMRI. Front Syst Neurosci 2013; 30; 7: 124.

Hampshire A, Chamberlain SR, Monti MM, Duncan J, Owen AM.

The role of the right inferior frontal gyrus: inhibition and attentional

control. Neuroimage 2010; 50: 1313–9.

Horn A, Ostwald D, Reisert M, Blankenburg F. The structural-func-

tional connectome and the default mode network of the human

brain. Neuroimage 2014, in press; doi: 10.1016/

j.neuroimage.2013.09.069.

Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta

burst stimulation of the human motor cortex. Neuron 2005; 45:

201–6.

Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical

diagnosis of idiopathic Parkinson’s disease: a clinico-pathological

study of 100 cases. J Neurol Neurosurg Psychiatry 1992; 55:

181–4.

Jahanshahi M. Effects of deep brain stimulation of the subthalamic

nucleus on inhibitory and executive control over prepotent responses

in Parkinson’s disease. Front Syst Neurosci 2013; 7: 118.

Jahfari S, Waldorp L, van dem Wildenberg WPM, Scholte HS,

Ridderinkhof KR, Forstmann BU. Effective connectivity reveals im-

portant roles for both the hyperdirect (fronto-subthalamic) and the

indirect (fronto-striatal-pallidal) fronto-basal ganglia pathways

during response inhibition. J Neurosci 2011; 31: 6891–9.Jenner P. Molecular mechanisms of L-DOPA-induced dyskinesia. Nat

Rev Neurosci 2008; 9: 665–77.Kahan J, Urner M, Moran R, Flandin G, Marreiros A, Mancini L,

et al. Resting state functional MRI in Parkinson’s disease: the impact

of deep brain stimulation on ‘effective’ connectivity. Brain 2014;

137 (Pt 4): 1130–44.Kelly CAM, Uddin LQ, Biswal BB, Castellanos FX, Milham MP.

Competition between functional brain networks mediates behavioral

variability. Neuroimage 2008; 39: 527–37.King AV, Linke J, Gass A, Hennerici MG, Tost H, Poupon C,

Wessa M. Microstructure of a three-way anatomical network pre-

dicts individual differences in response inhibition: a tractography

study. Neuroimage 2012; 59: 1949–59.Koch G, Brusa L, Caltagirone C, Peppe A, Oliveri M, Stanzione P,

et al. rTMS of supplementary motor area modulates therapy-

induced dyskinesias in Parkinson disease. Neurology 2005; 65:

623–5.

Koch G, Franca M, Del Olmo MF, Cheeran B, Milton R, Alvarez

Sauco M, et al. Time course of functional connectivity between

dorsal premotor and contralateral motor cortex during movement

selection. J Neurosci 2006; 26: 7452–9.

Koch G, Brusa L, Carrillo F, Lo Gerfo E, Torriero S, Oliveri M, et al.

Cerebellar magnetic stimulation decreases levodopa-induced dyskin-

esias in Parkinson disease. Neurology 2009; 73: 113–9.

Li CS, Yan P, Sinha R, Lee TW. Subcortical processes of motor re-

sponse inhibition during a stop signal task. Neuroimage 2008; 41:

1352–63.

Lowe MJ, Mock BJ, Sorenson JA. Functional connectivity in single

and multislice echoplanar imaging using resting-state fluctuations.

Neuroimage 1998; 7: 119–32.

Disrupted cortico-cortical connectivity in PD with LIDs. BRAIN 2014: Page 13 of 14 | 13

by guest on Novem

ber 20, 2014D

ownloaded from

Morgante F, Espay AJ, Gunraj C, Lang AE, Chen R. Motor cortexplasticity in Parkinson’s disease and levodopa-induced dyskinesias.

Brain 2006; 129 (Pt 4): 1059–69.

Murphy K, Birn RM, Handwerker DA, Jones TB, Bandettini PA. The

impact of global signal regression on resting state correlations: areanti-correlated networks introduced? Neuroimage 2009; 44:

893–905.

Nambu A, Tokuno H, Takada M. Functional significance of the cor-

ticosubthalamo-pallidal ‘hyperdirect’ pathway. Neurosci Res 2002;43: 111–7.

Nambu A. A new dynamic model of the cortico-basal ganglia loop.

Prog Brain Res 2004; 143: 461–6.Neubert FX, Mars RB, Buch ER, Olivier E, Rushworth MFS. Cortical

and subcortical interactions during action reprogramming and their

related white matter pathways. Proc Natl Acad Sci USA 2010; 107:

13240–5.Obeso JA, Rodrıguez-Oroz MC, Rodrıguez M, Lanciego JL, Artieda J,

Gonzalo N, et al. Pathophysiology of the basal ganglia in

Parkinson’s disease. Trends Neurosci 2000; 23 (Suppl 10): S8–19.

Obeso I, Wilkinson L, Casabona E, Bringas ML, Alvarez M,Alvarez L, et al. Deficits in inhibitory control and conflict resolution

on cognitive and motor tasks in Parkinson’s disease. Exp Brain Res

2011; 212: 371–84.

Obeso I, Strafella AP. Boosting response inhibition neural networkwith rTMS may improve dyskinesias in Parkinson’s disease. Brain

Stimul 2014; 7: 336–7.

Passamonti L, Salsone M, Toschi N, Cerasa A, Giannelli M,Chiriaco C, et al. Dopamine-transporter levels drive striatal re-

sponses to apomorphine in Parkinson’s disease. Brain Behav 2013;

3: 249–62.

Pavese N, Evans AH, Tai YF, Hotton G, Brooks DJ, Lees AJ, et al.Clinical correlates of levodopa-induced dopamine release in

Parkinson disease: a PET study. Neurology 2006; 67: 1612–17.

Press WH TS, Vetterling WT, Flannery BP. Numerical recipes in C.

Cambridge, UK: Cambridge University Press; 1992.Rascol O, Sabatini U, Brefel C, Fabre N, Rai S, Senard JM, et al.

Cortical motor overactivation in parkinsonian patients with

L-dopa-induced peak-dose dyskinesia. Brain 1998; 121 (Pt 3):527–33.

Stoffers D, Bosboom JL, Wolters ECh, Stam CJ, Berendse HW.

Dopaminergic modulation of cortico-cortical functional connectivity

in Parkinson’s disease: an MEG study. Exp Neurol 2008; 213:191–5.

Swann N, Tandon N, Canolty R, Ellmore TM, McEvoy LK, Dreyer S,

et al. Intracranial EEG reveals a time- and frequency-specific role for

the right inferior frontal gyrus and primary motor cortex in stoppinginitiated responses. J Neurosci 2009; 29: 12675–85.

van Campen AD, Neubert FX, van den Wildenberg WP,

Ridderinkhof KR, Mars RB. November 12, 2013. Paired-pulse tran-scranial magnetic stimulation reveals probability-dependent changes

in functional connectivity between right inferior frontal cortex and

primary motor cortex during go/no-go performance. Front Hum

Neurosci 2013; 7: 736.Veniero D, Ponzo V, Koch G. Paired associative stimulation enforces

the communication between interconnected areas. J Neurosci 2013;

33: 13773–83.

Vernon AC, Modo M. Do levodopa treatments modify the morph-ology of the parkinsonian brain? Mov Disord 2012; 27: 166–7.

Verbruggen F, Logan GD. Response inhibition in the stop-signal para-

digm. Trends Cogn Sci 2008; 12: 418–24.

Vossel S, Weidner R, Fink GR. Dynamic coding of events within theinferior frontal gyrus in a probabilistic selective attention task. J

Cogn Neurosci 2011; 23: 414–24.

Wagle-Shukla A, Angel MJ, Zadikoff C, Enjati M, Gunraj C,Lang AE, Chen R. Low-frequency repetitive transcranial magnetic

stimulation for treatment of levodopa-induced dyskinesias.

Neurology 2007; 68: 704–5.

Wu T, Long X, Wang L, Hallett M, Zang Y, Li K, Chan P. Functionalconnectivity of cortical motor areas in the resting state in

Parkinson’s disease. Hum Brain Mapp 2011; 32: 1443–57.

Zandbelt BB, Bloemendaal M, Hoogendam JM, Kahn RS, Vink M.

Transcranial magnetic stimulation and functional MRI reveal cor-tical and subcortical interactions during stop-signal response inhib-

ition. J Cogn Neurosci 2013; 25: 157–74.

14 | BRAIN 2014: Page 14 of 14 A. Cerasa et al.

by guest on Novem

ber 20, 2014D

ownloaded from