Cerebellar and lobar blood flow in schizophrenia: A perfusion weighted imaging study

Psychiatry Research: Neuroimaging 193 (2011) 46–52

Contents lists available at ScienceDirect

Psychiatry Research: Neuroimaging

j ourna l homepage: www.e lsev ie r.com/ locate /psychresns

Cerebellar and lobar blood flow in schizophrenia: A perfusion weightedimaging study

Marcella Bellani a, Denis Peruzzo b, Miriam Isola c, Gianluca Rambaldelli a, Cinzia Perlini a, Monica Baiano d,Roberto Cerini e, Nicola Andreone a, Marco Barillari e, Roberto Pozzi Mucelli e, Matteo Balestrieri f,Michele Tansella a, Alessandra Bertoldo b, Paolo Brambilla f,g,⁎a Department of Public Health and Community Medicine, Section of Psychiatry and Clinical Psychology, Inter-University Centre for Behavioral Neurosciences, University of Verona,Verona, Italyb Department of Information Engineering, University of Padova, Padova, Italyc Department of Medical and Morphological Research, Section of Statistics, University of Udine, Udine, Italyd Centre for Eating Disorders, ASSL 10 “Veneto Orientale”, Venice, Italye Department of Morphological and Biomedical Sciences, Section of Radiology, G.B. Rossi Hospital, University of Verona, Verona, Italyf DPMSC, Section of Psychiatry, Inter-University Centre for Behavioral Neurosciences, University of Udine, Udine, Italyg Scientific Institute IRCCS ‘E. Medea’, Udine, Italy

⁎ Corresponding author. Dipartimento di Patologia etale, Cattedra di Psichiatria, Policlinico Universitario, ViaTel.: +39 0432 55 9494; fax: +39 0432 54 5526.

E-mail address: [email protected] (P. Brambi

0925-4927/$ – see front matter © 2010 Elsevier Irelanddoi:10.1016/j.pscychresns.2010.12.010

a b s t r a c t

Article history:Received 17 January 2010Received in revised form 5 October 2010Accepted 13 December 2010

Keywords:NeuroimagingGadoliniumMRICerebral blood flowCerebellar blood flowSchizophrenia

It is still not clear whether brain hemodynamics plays a role in the functional and structural alterations inschizophrenia, since prior imaging studies showed conflicting findings. In this study we non-invasivelyexplored cerebral and cerebellar lobe perfusion in the largest population of participants with schizophreniathus far studied with perfusion-weighted imaging (PWI). Forty-seven participants affected by schizophreniaand 29 normal controls were recruited. PWI images were acquired following the intravenous injection of aparamagnetic contrast agent. Regional cerebral blood volume (CBV), blood flow (rCBF), andmean transit time(MTT) were obtained with the block-Circulant Singular Value Decomposition (cSVD) for frontal, temporal,parietal, occipital, and cerebellar lobes, bilaterally. Perfusion parameters were separately obtained for bothgray and white matter in each lobe. Subjects with schizophrenia showed no significant differences inperfusion parameters when compared with controls. Interestingly, inverse correlations between age at onsetand occipital, frontal and cerebellar MTT and between length of illness and frontal CBV were found. Preservedcerebral and cerebellar perfusion in our chronic population may in part be due to the effects of antipsychotictreatment which may have normalized blood volume and flow. Hypoperfusion in relation to chronicity,particularly in the frontal lobe, has been observed in accordance with earlier studies using positron emissiontomography.

Medicina Clinica e Sperimen-Colugna 50, 33100 Udine, Italy.

lla).

Ltd. All rights reserved.

© 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

There is consistent evidence, as shown by neuroimaging studies,that schizophrenia may be associated with several structural andfunctional abnormalities that may distinguish the brains of indivi-duals affected by schizophrenia from non-affected healthy controls(Agarwal et al., 2010). This may lead to impaired cortico-subcorticalconnectivity ultimately resulting in compromised higher cognitiveprocesses, such as attention andmemory processing (Andreasen et al.,1998; Schlösser et al., 2003; Laurens et al., 2005; Gur et al., 2007;Bellani et al., 2009a,b,c;Whitfield-Gabrieli et al., 2009). In this context,the vascular organization may play a role (Hanson and Gottesman,

2005). Indeed, frontal, parietal, temporal, and cerebellar metabolicand perfusion deficits have been observed in schizophrenia in studiesusing positron emission tomography (PET), single photon emissiontomography (SPECT), computed tomograpy, xenon Xe 133-inhalationtechnique and perfusion magnetic resonance imaging (MRI) studies(see for review Bachneff, 1996; Davidson and Heinrichs, 2003;Théberge, 2008). Nonetheless, contradictory findings have also beenreported with reduced (Steinberg et al., 1995; Andreasen et al., 1997;Crespo-Facorro et al., 1999; Kim et al., 2000; Malaspina et al., 2004;Kanahara et al., 2009), increased (Cohen et al., 1995; Andreasen et al.,1997; Loeber et al., 1999; Kim et al., 2000; Malaspina et al., 2004;Ortuño et al., 2006; Ben-Shachar et al., 2007; Fujimoto et al., 2007),and preserved perfusion (Gur et al., 1983; Loeber et al., 1999).

In this context, perfusion patterns have been explored indifferent groups of patients affected by schizophrenia, i.e. chronic,not treated, or recent onset individuals. Indeed frontal hypoperfu-sion was demonstrated in first episode patients, neuroleptic-naïve

47M. Bellani et al. / Psychiatry Research: Neuroimaging 193 (2011) 46–52

and also in nonmedicated schizophrenia patients, while increasedperfusion was more sporadic and with inconsistent regional braindistributions in the various studies (Andreasen et al., 1997; Novaket al., 2005; Wake et al., 2010). Flow deficits were also presentalso in nonmedicated chronic patients with schizophrenia (Kim etal., 2000; Scheef et al., 2010). In Kim and colleagues' studyneuroleptic free patients with chronic schizophrenia showed lowerblood flow in prefrontal areas and higher flow in thalamic andcerebellar regions than normal comparison subjects. The authorsargue that similarity between first-episode and chronically illpatients suggests that prefrontal hypoperfusion is not due to thechronicity of illness or long-term neuroleptic treatment. Scheef etal. (2010) found reduced blood flow in frontal, parietal lobes andanterior cingulate gyrus, while increased perfusion was found inseveral regions (i.e. cerebellum, thalamus and occipital lobe).Medicated schizophrenia patients had preserved perfusion at rest,and specific flow increase emerged only during task performancein Gur et al. (1983).

On the other hand, there is some evidence that antipsychoticsmay variably affect and normalize cerebral blood flow inschizophrenia, although the effects may be regionally and drugspecific (Miller et al., 1997; Lahti et al., 2003; Ertugrul et al.,2009). Most of the investigations on cerebral blood flow have beenconducted using SPECT and PET techniques. More recently, MRIperfusion methods have been applied. Indeed, perfusion-weightedimaging (PWI) allows to assess subtle brain perfusion changes anddoes not require the exposure to ionizing radiation (as in SPECTand PET). Dynamic susceptibility contrast enhanced MRI (DSC-MRI) (Edelman et al., 1990; Rosen et al., 1991) has increasinglybeen used for the measurement of cerebral perfusion in humans(Kaneoke et al., 1989; Théberge, 2008). The most importantparameters provided by DSC-MRI, which reflect cerebral hemody-namics at a capillary level, are the cerebral blood flow (CBF), thecerebral blood volume (CBV), and the mean transit time (MTT).More specifically, cerebral blood flow is the volume of blooddelivered in a given time to a given mass of brain tissue/voxel; thecerebral blood volume is the volume of intravascular blood in theregion of interest; the mean transit time is the average time that aparticle of tracer takes to traverse the capillary bed within theregion of interest (Théberge, 2008). In this study, we applied DSC-MRI and aimed to further investigate lobar and cerebellarvasculature in schizophrenia by segmenting, for the first time,gray and white matter in the largest population thus far studied toour best knowledge.

Table 1Sample's features.

Controls (N=29)

Age (years) 45.17±10.41Males/females 13/16Ethnicity CaucasianBody mass index (BMI) median (range) 23.57 (17.18–31.83)Education, low and higha 16/10Smokers/non-smokers 3/26Left handed/right handed 1/25GAF score, median (range) 72 (65–80)Age of onset, median (range) –

Length of illness, y, median (range) –

Number of prior hospitalizations, median (range) –

CPZ-eq for atypical antipsychotics –

CPZ-eq for typical antipsychotics median (range)BPRS Total scores, median (range) –

Positive symptoms, median (range)Negative symptoms, median (range)

BPRS = Brief Psychiatric Rating Scale; GAF = Global Assessment of Functioning scale; CPZ-a Low education included primary and middle school; while high education included hig

2. Materials and methods

2.1. Subjects

Forty-seven participants with DSM-IV schizophrenia (mean age±S.D.=36.98±11.43 years; 31 males, 16 females; all Caucasians) and29 normal controls (mean age±S.D.=45.17±10.41 years; 13 males,16 females; all Caucasians) were studied (Table 1). This samplerepresents a subgroup of the one analyzed in a previous work(Brambilla et al., 2007). Seven participants of the original samplewereexcluded for technical reasons (segmentation process, unsuccessfulcompletion in the post-processing phase). Participants were recruitedfrom the South-Verona Psychiatric Care Register (PCR) (Tansella andBurti, 2003; Amaddeo et al., 2009), a community-based mental healthregister. Participants' diagnoses were established using the ItemGroup Checklist of the Schedule for Clinical Assessment in Neuropsy-chiatry (IGC-SCAN). The IGC-SCAN is a semi-structured standardizedchecklist investigating a wide range of psychiatric symptoms, whichare included in 41 psychopathological item groups (Wing et al., 1990).It is worth noting that the Italian version of the SCAN was editedby our group (Tansella and Nardini, 1996) and that our investigatorsattended specific training courses held by official trainer in order tolearn how to administer the checklist. The IGC-SCAN was performedby trained raters who completed at least 10 IGC-SCAN assessmentswith an expert senior investigator. Successively, they had to befully reliable with the senior investigator, achieving similar diagnosesfor at least eight out of 10 IGC-SCANs. Moreover, the two raterswere compared for each psychopathological item group, in order toevaluate any major symptom inconsistencies. Also, we regularlyassured reliability of the IGC-SCAN diagnoses by holding consensusmeetings with treating psychiatrists and a senior investigator.Finally, the diagnosis of schizophrenia was confirmed by two staffpsychiatrists according to the DSM-IV criteria. Exclusion criteria forparticipants included comorbid psychiatric disorders, alcohol orsubstance abuse within the 6 months preceding the study, historyof traumatic head injury with loss of consciousness, epilepsy or otherneurological or medical diseases, including hypertension and diabe-tes. None of our participants had a history of electroconvulsivetherapy. All participants, except three, them were receiving antipsy-chotic medications and none of them was on any cardiovasculardrugs at the time of imaging, including ß-blockers or nitrates.Specifically, 20 participants were on olanzapine, seven on haloperidol,seven on clozapine, six on risperidone, three on chlorpromazine,and the remaining four on other antipsychotics such as quetiapine

Schizophrenia Participants (N=47) Statistics p

36.98±11.43 t=3.14 0.00231/16 χ2=3.28 0.07Caucasian25.34 (19.05–51.56) z=−2.46 0.01430/17 χ2=0.04 0.8525/22 χ2=14.15 b0.00018/38 χ2=2.79 0.0955 (25–70) z=−26.53 b0.000122 (15–58)12 (0–37)2 (0–15)327.15±137.2744.64 (21–150)42.50 (26–100)11 (5–29)10.50 (7–22)

eq = chlorpromazine equivalents; y = years.h school, college and university.

48 M. Bellani et al. / Psychiatry Research: Neuroimaging 193 (2011) 46–52

(N=1), thioridazine (N=1), zuclopenthixol (N=1), and fluphen-azine (N=1). Participants' clinical information was retrieved frompsychiatric interviews, the attending psychiatrist, and medical charts;also, clinical symptoms were identified using the Brief PsychiatricRating Scale (BPRS 24-item version) (Ventura et al., 2000). Symptomratings were assessed with the BPRS by trained research clinicalpsychologists and the reliability was achieved and monitored usingsimilar procedures as for the IGC-SCAN.

Control subjects were assessed for DSM-IV axis I disorders by usingthe SCID-IV non-patient version (SCID-NP). They had no history ofpsychiatric disorders among first-degree relatives, alcohol or sub-stance abuse, and current major neurological or medical illness,including hypertension and diabetes. They were subjects undergoingMR scanning for dizziness without evidence of central nervous systemabnormalities on the serial conventional MR images and on the pre-and post-contrast MR acquisitions, as reviewed by the neuroradiol-ogist (R.C.). Dizziness was due to benign paroxysmal positionalvertigo or to non-toxic labyrinthitis. Control individuals were scannedonly after a full medical history and general neurological, otoscopic,and physical examinations, being already completely recovered fromdizziness. Also, none of them were on medication at the time ofparticipation in the study, including drugs for nausea or vertigo.

This study protocol was approved by the biomedical EthicsCommittee of the Azienda Ospedaliera di Verona. All participantsand control subjects signed a written informed consent, after havingunderstood all issues involved in study participation.

2.2. MRI data acquisition

MRI scanswere acquiredwith a 1.5 T SiemensMagnetomSymphonyMaestro Class, Syngo MR 2002B. A standard head coil was used for RFtransmission and reception of the MR signal. All participants wereprovided with earplugs to reduce acoustic noise and their head wascomfortably placed in a head holder and maintained stable with foampads, thus minimizing movement artefact. T1-weighted images werefirst acquired to verify subject's head position and image quality(TR=450 ms, TE=14 ms, flip angle=90°, FOV=230×230, slicethickness=5 mm, matrix size=384×512, NEX=2). Successively,PD/T2-weighted images were obtained (TR=2500 ms, TE=24/121 ms, flip angle=180°, FOV=230×230, slice thickness=5 mm,matrix size=410×512, NEX=2), according to anaxial plane parallel tothe anterior–posterior commissure (AC–PC) to exclude the presence offocal brain lesions. Perfusion-weighted acquisitions, consisting of echo-planar imaging of T2-weighted sequence, were acquired in the axialplane parallel to the AC–PC line (20 sequential images for 60 repetitions,TR=2160 ms, TE=47ms, FOV=230×230, slice thickness=5 mm,matrix size=256×256, NEX=1, EPI factor=128) immediately before,during, and after injection of a bolus of gadopentetate dimeglumine-dietilen-tetra-penta-acetic acid (Gd-DTPA), a paramagnetic agent withintravascular space distribution. Contrast material (0.1 mmol/kg)administrationwas started after 4 s by power injector (Medrad SpectrisMR injector) through an 18- or 20-gauge angiocatheter through theright antecubital vein at a rate of 2.5 mL/s, followed immediately by25 mL of continuous saline flush. The same neuroradiologist (R.C.)controlled timing and accuracy of gadolinium administration for allparticipants and controls. Finally, post-contrast spin echo T1-weightedimages were obtained in the axial plane in order to exclude possiblefocal lesions (TR=448 ms, TE=14 ms, flip angle=90°, FOV=230×230, slice thickness=5 mm,matrix size=384×512, NEX=2). Normalcontrols and participants with schizophrenia completed the MRIsession without any apparent hyperventilation due to anxiety reaction.Indeed, in order to reduce any possible anxiety symptoms, fellowsfrom our research group carefully provided full information onMRI andpersonally accompanied subjects at the MR center, waiting for themuntil the end of the session.

2.3. Post-processing and image analysis

Raw echo-planar data of the images were first transferred to acommercial workstation and semi-automatically processed by usingin-house software written in MATLab developed by our engineers(version 7; The Mathworks Inc., Natick, MA).

The regions of interest (ROIs) corresponded to frontal, parietal,temporal, occipital cortices and cerebellar hemispheres, and werehand-drawn on the mean DSC volume (Fig. 1).

Frontal lobe (Fig. 1A): a linewas traced from the top to the bottom ofthe superior frontal sulcus. The first two slices including the orbito-frontal cortex and the second two slices including the prefrontal cortexwere excluded. We considered the fifth and sixth slices comprising thefrontal cortex.

Temporal lobe (Fig. 1B): in the first slice, in which the orbitofrontalcortex was visible, a line was traced at the level of the pre-occipitalscissure to identify the temporal lobe. Then, we followed temporallobe's shape along the antero-superior brain area.

Parietal lobe (Fig. 1C): the slice in which the ventricles and corpuscallosum were clearly detectable and the one in which only the inter-hemispheric schism was visible, but not the ventricles and corpuscallosum, were considered in our tracing. The right and the left sideswere defined by the post-central sulcus.

Occipital lobe (Fig. 1D): the slice in which for the first time the pre-occipital scissurewas detectable, tracing a line following that scissure toidentify the occipital lobe, and the following slice, were selected. Then,we followed occipital lobe's shape along the infero-posterior brain area.

Cerebellum (Fig. 1E): The first slice where the cerebellum is visible,excluding the vermis, and the following one were used for tracing.

All measurements were obtained manually by a trained evaluator,who was blind to the diagnosis of schizophrenia and to subjects'identity. The intra-class correlation coefficients (ICCs), which werecalculated by having two independent raters trace 10 scans, werehigher than 0.95 for both hemispheres. The ICC was computed onthe signal Contrast Enhancement (CE), which is defined as the ratiobetween the maximum signal drop due to the contrast agent and thesignal baseline before the contrast agent injection.

EachROIwasdivided into graymatter (GM) andwhitematter (WM)by using the Statistical Parametric Mapping toolbox, version 8 (SPM8).GM andWMmaps were obtained by segmenting the second volume oftheDSC-MRI sequence of each subject. The selected volume is a standardT2*-weighted image because it is acquired before the tracer injection.Segmentationwas achieved by using the default parameter setting, andthen a 0.85 threshold was applied on the resulting probability maps toobtain GM andWM binary masks. In the end, 20 ROIs were considered,i.e. left and right, GM and WM ROI for the frontal, parietal, temporal,occipital cortices and for the cerebellar hemispheres.

For each ROI, the mean MR signal was computed to quantify theROI perfusion parameters. Then, the CBF, CBV and MTT assessed onthe mean signal were assumed to characterize the whole ROI.

AIF was also determined in each subject by averaging the MR signalof a few voxels placed in the middle cerebral artery (MCA). A physicianmanually selected the arterial voxels in the MCA branch supplying thefrontal cortex.

Perfusion parameters were calculated on the basis of the principlesof tracer kinetics for nondiffusible tracers (Zierler, 1962, 1965). First ofall, the concentration curve was obtained from the MR signal S(t)following the relation:

C tð Þ = − 1TE

lnS tð ÞS0

� �

where S0 was computed by averaging the signal samples beforethe tracer injection (Østergaard et al., 1996a). Subsequently, a fit witha gamma-variate function was computed to eliminate the recirculation

Fig. 1. ROIs tracing in brain lobes and cerebellum. A = Frontal lobe, B = Occipital lobe, C = Parietal lobe, D = Temporal lobe, E = Cerebellum.

49M. Bellani et al. / Psychiatry Research: Neuroimaging 193 (2011) 46–52

as reported elsewhere (Porkka et al., 1991; Østergaard et al., 1996a,b;Benner et al., 1997).

CBVvalueswere computedas the ratio between thevoxelAreaUnderthe Curve (AUC) and the arterial AUC, as reported by Østergaard et al.

CBV =

∫∞

0

CROI tð Þdt

∫∞

0

CAIF tð Þdt

CBF values were computed by using the block-circulant SingularValue Decomposition (cSVD) method as reported by Wu et al. (Wuet al., 2003), and then MTT values were computed as the ratiobetween CBV and CBF (Østergaard et al., 1996a).

Currently, the most used deconvolution methods for clinicalstudies are the Singular Value Decomposition (SVD) and the block-Circulant Singular Value Decomposition (cSVD). SVD is historically thefirst and the most important deconvolution method proposed in theDSC-MRI context. However, the SDV deconvolution method issensitive to the tracer delay between the arterial and tissue signal.The cSVD method is the natural evolution of SVD and it has beenproposed to reduce the effect of the delay. cSVD results have beenshown to be insensitive to the presence of delay (Wu et al., 2003).Thus, for this reason cSVD has been used to estimate CBF. It is of notethat absolute quantification of CBF and CBV with PWI data is notstraightforward because of difficulties in absolute scaling of AIF andtissue concentration curves and normalization procedures are re-quired to determine absolute CBF and CBV values (Calamante et al.,

Table 3Cerebral blood volume (CBV) in healthy controls and schizophrenia participants.

Controls(N=29)

Schizophreniaparticipants(N=47)

Statistics(d.f.=1/71)

Cerebral blood volume (CBV) Mean S.D. Mean S.D. F p

White matterRight frontal lobe 0.174 0.062 0.201 0.171 0.089 0.766Left frontal lobe 0.178 0.063 0.201 0.155 0.254 0.622Right temporal lobe 0.199 0.070 0.193 0.080 0.352 0.555Left temporal lobe 0.201 0.061 0.198 0.081 0.481 0.490Right parietal lobe 0.148 0.057 0.148 0.060 0.427 0.516Left parietal lobe 0.167 0.071 0.158 0.058 0.846 0.361Right occipital lobe 0.155 0.061 0.162 0.073 0.11 0.918Left occipital lobe 0.172 0.075 0.177 0.072 0.027 0.869

Gray matterRight frontal lobe 0.321 0.120 0.371 0.295 0.353 0.554Left frontal lobe 0.314 0.104 0.369 0.310 0.174 0.678Right temporal lobe 0.348 0.102 0.368 0.154 0.194 0.661Left temporal lobe 0.322 0.096 0.361 0.145 0.214 0.645Right parietal lobe 0.301 0.094 0.319 0.130 0.109 0.742Left parietal lobe 0.292 0.105 0.311 0.125 0.119 0.731Right occipital 0.285 0.110 0.307 0.130 0.148 0.701Left occipital 0.272 0.099 0.301 0.131 0.001 0.978

GLM for repeated measures with age, BMI and smoking status as covariates wasperformed to compare CBV (F=0.178, d.f.=1/71, p=0.675). Values betweenparticipants with schizophrenia and normal controls. In the table the statistics of theunivariate GLM are reported.

50 M. Bellani et al. / Psychiatry Research: Neuroimaging 193 (2011) 46–52

2002; Chen et al., 2005; Sakaie et al., 2005). Unfortunately, there isno general consensus concerning the best absolute normalizationprocedure; thus, CBF and CBV values are usually obtained from PWIdata for less than a scale factor.

2.4. Statistical analyses

SPSS for Windows software, version 18.0 (SPSS Inc., Chicago), wasused to perform all statistical analyses, and the two-tailed statisticalsignificance level was set at pb0.05.

Data were tested for normal distribution using the Shapiro-Wilktest and for homogeneity of variances with the Levene test.Comparisons between groups for quantitative variables were per-formed by t-test or Mann-Whitney U-test, depending on the Shapiro-Wilk test results. Chi-square (χ2) tests were used to analysecategorical values; when assumptions for χ2 test were not verified,Fisher's exact test was used.

A General LinearModel (GLM) for repeatedmeasures (hemisphereand region of interest as repeated measures factor) was used tocompare cerebral blood volume (CBV), cerebral blood flow (CBF) andmean transit time (MTT) between schizophrenia participants andcontrol subjects with age, body-mass index (BMI), and smoker/non-smoker status as covariates.

The assumptions that the vector of the measures followed amultivariate normal distribution (Shapiro-Wilk test) and the variance–covariance matrices were circular in form (Mauchly's test) wereverified. Univariate GLM analysis was performed to compare CBV,CBF and MTT coefficients of left and right brain structures betweenparticipants with schizophrenia and control participants.

Pearson's or Spearman's correlation analyses were used to explorepossible associations between age, age of onset, length of illness,number of prior hospitalizations, chlorpromazine equivalents, BPRSscores and perfusion measures, depending on Shapiro-Wilk testresults. The Bonferroni correction was not used.

3. Results

None of the subjects reported any adverse reaction to rapid powerinjection of the contrast material.

Table 2Cerebral blood flow (CBF) in controls and schizophrenia participants.

Controls(N=29)

Schizophreniaparticipants(N=47)

Statistics(d.f.=1/71)

Cerebral blood flow (CBF) Mean S.D. Mean S.D. F p

White matterRight frontal lobe 0.018 0.007 0.020 0.014 0.048 0.826Left frontal lobe 0.018 0.006 0.020 0.012 0.076 0.784Right temporal lobe 0.019 0.007 0.019 0.006 0.050 0.823Left temporal lobe 0.018 0.006 0.018 0.005 0.001 0.976Right parietal lobe 0.014 0.006 0.015 0.004 0.071 0.791Left parietal lobe 0.015 0.007 0.015 0.004 0.000 0.998Right occipital lobe 0.016 0.006 0.017 0.006 0.637 0.427Left occipital lobe 0.017 0.008 0.017 0.006 0.153 0.697

Gray matterRight frontal lobe 0.031 0.011 0.034 0.019 0.137 0.713Left frontal lobe 0.030 0.011 0.034 0.021 0.125 0.725Right temporal lobe 0.035 0.011 0.036 0.011 0.018 0.893Left temporal lobe 0.034 0.011 0.037 0.011 0.711 0.402Right parietal lobe 0.032 0.011 0.033 0.010 0.009 0.925Left parietal lobe 0.033 0.013 0.034 0.011 0.018 0.894Right occipital 0.031 0.012 0.032 0.011 0.001 0.969Left occipital 0.030 0.012 0.032 0.011 0.046 0.813

GLM for repeated measures with age, BMI and smoking status as covariates wasperformed to compare CBF (F=0.002, d.f.=1/71, p=0.966). Values betweenparticipants with schizophrenia and normal controls. In the table the statistics of theunivariate GLM are reported.

Age was different in the two groups, as BMI, smoking status andGAF score. The first three variables were therefore used as covariatesin the analysis.

No evidence of significant difference in the perfusion parameters(CBF, CBV, and MTT) for the two groups was shown for lobes,cerebellar hemispheres, bilaterally and segmented for the white andgray matter (pN0.05) (Tables 2–5).

Spearman correlation analyses performed in the schizophreniagroup resulted in significant inverse relationships between age ofonset and MTT values for frontal gray matter (left r=−0.355,

Table 4Mean Transit Time (MTT) in healthy controls and schizophrenia participants.

Controls(N=29)

SchizophreniaParticipants(N=47)

Statistics(d.f.=1/71)

Mean transit time (MTT) Mean S.D. Mean S.D. F p

White matterRight frontal lobe 10.023 1.737 10.010 2.084 1.417 0.238Left frontal lobe 10.224 2.038 10.144 2.226 1.383 0.244Right temporal lobe 10.635 1.711 10.426 2.332 2.074 0.154Left temporal lobe 11.049 1.868 10.967 2.517 1.897 0.173Right parietal lobe 10.560 1.741 10.350 2.441 2.564 0.114Left parietal lobe 11.053 1.751 10.872 2.512 3.667 0.060Right occipital lobe 9.780 1.732 9.579 2.109 1.458 0.231Left occipital lobe 10.019 1.917 10.283 2.763 1.885 0.174

Gray matterRight frontal lobe 10.592 1.971 10.993 2.537 0.771 0.383Left frontal lobe 10.800 1.971 11.241 2.521 0.407 0.526Right temporal lobe 10.198 1.652 10.393 2.614 0.473 0.494Left temporal lobe 9.688 1.773 9.843 2.315 0.358 0.551Right parietal lobe 9.845 1.800 10.050 2.452 0.689 0.409Left parietal lobe 9.109 1.543 9.374 2.261 0.744 0.391Right occipital lobe 9.289 1.647 9.534 2.105 0.389 0.535Left occipital lobe 9.116 1.509 9.260 2.173 0.421 0.519

GLM for repeated measures with age, BMI and smoking status as covariates wasperformed to compare MTT (F=1.333, d.f.=1/71, p=0.252). Values betweenparticipants with schizophrenia and normal controls. In the table the statistics of theunivariate GLM are reported.

Table 5Cerebellum: CBF, CBV and MTT in healthy controls and schizophrenia participants.

Controls(N=29)

SchizophreniaParticipants(N=47)

Statistics(d.f.=1/71)

Mean S.D. Mean S.D. F p

Cerebellum blood flow (CBF)Right cerebellum white matter 0.019 0.007 0.020 0.006 0.017 0.896Left cerebellum white matter 0.020 0.008 0.021 0.006 0.003 0.954Right cerebellum gray matter 0.023 0.008 0.025 0.010 0.060 0.808Left cerebellum gray matter 0.024 0.009 0.025 0.008 0.490 0.486

Cerebellum blood volume (CBV)Right cerebellum white matter 0.198 0.077 0.199 0.087 0.598 0.442Left cerebellum white matter 0.200 0.084 0.205 0.095 0.302 0.584Right cerebellum gray matter 0.227 0.082 0.227 0.120 0.249 0.619Left cerebellum gray matter 0.225 0.093 0.244 0.099 0.007 0.933

Mean transit time (MTT)Right cerebellum white matter 10.452 1.691 10.065 2.708 1.986 0.163Left cerebellum white matter 10.034 1.317 9.942 2.668 1.271 0.263Right cerebellum gray matter 9.682 1.753 9.506 2.464 1.270 0.264Left cerebellum gray matter 9.309 1.506 9.410 2.232 0.268 0.606

GLM for repeated measures with age, BMI and smoking status as covariates wasperformed to compare CBF, CBV and MTT (F=1.333, d.f.=1/71, p=0.252). Valuesbetween participants with schizophrenia and normal controls. In the table the statisticsof the univariate GLM are reported.

51M. Bellani et al. / Psychiatry Research: Neuroimaging 193 (2011) 46–52

p=0.015; right r=−0.376, p=0.010), occipital lobes (left graymatter: r=−0.291, p=0.05; left white matter: r=−0.335;p=0.023, right gray matter: r=−0.344, p=0.019; right whitematter: r=−0.364, p=0.013) and cerebellar hemispheres (left:r=−0.303, p=0.041; right: r=−0.316; p=0.032). Moreover, thelength of illness inversely correlated with frontal gray matter CBV(left: r=−0.355, p=0.015; right: r=−0.376, p=0.010).

All the other correlations with clinical variables were notsignificant (pN0.05).

4. Discussion

This study showed no alterations of the lobar and cerebellarperfusion parameters in participants with chronic schizophreniacompared to controls. This is consistent with a previous work byGur et al. (1983) on medicated schizophrenia patients (N=15). Theyshow preserved perfusion at rest, compared to matched controls(N=25) and specific flow increase emerged only during taskperformance.

Our sample was composed by chronic treated participants andthere is evidence that neuroleptics influence cerebral blood flowInparticular, some studies have shown that the antipsychotics,particularly the atypical ones, improve neurocognitive performanceand normalize cerebral and cerebellar metabolism and blood flow inschizophrenia (Loeber et al., 2002; Davis et al., 2005; Molina et al.,2005; Novak et al., 2005; Lindenmayer et al., 2007). In Miller et al.(2001) haloperidol and risperidone were tested in a group ofmedication-free patients with schizophrenia. Haloperidol was foundto decrease frontal cerebral blood flow and increase the blood flow inthe basal ganglia, while risperidone did not affect basal ganglia andslightly reduced the frontal perfusion. Moreover, clozapine andhaloperidol were found to induce increase of cerebral blood flow inbilateral dorsolateral frontal cortex (Lahti et al., 2003; Ertugrul et al.,2009), suggesting that a normalization of cerebral blood flow may beinduced by means of neuroleptic use in schizophrenia patients.Therefore, we cannot exclude that chronic treatment with anti-psychotics may have normalized or preserved lobar and cerebellarperfusion in schizophrenia.

Interestingly, inverse correlations between age of onset andfrontal, occipital and cerebellar MTT and between length of illnessand bilateral frontal CBV were found, suggesting potential effects of

chronicity in leading to frontal hypoperfusion. This issue shouldfurther be investigated by comparing first-episode subjects withchronically ill subjects and by including participants with differentage of onset and length of illness.

Some of methodological factors may have limited the interpreta-tion of our findings. First, participants and controls were not 1:1matched for socio-demographical variables. However, such variableswere used as covariates in the statistical analyses and a large numberof participants suffering from schizophrenia were enrolled, providingadequate statistical power. Second, normal controls were selectedfrom individuals undergoingMR scanning for dizziness, but they werescanned only after a full medical history and general neurological,otoscopic, and physical examinations, being already completelyrecovered from dizziness and had no evidences of central nervoussystem abnormalities. Third, only visual inspection was used to detectmotion artifacts, and no pulse sequence immune to susceptibilityartifacts was used to control for Gadolinium susceptibility artifactspotentially limiting the detection of minor artifacts interfering withthe quantitation of subtle perfusion changes. However, this wouldhave affected perfusion parameters for both subjects with schizo-phrenia and normal controls, therefore minimizing the impact of suchlimitation.

No evidence of lobar and cerebellar altered perfusion was found inthis study. Nonetheless, frontal hypoperfusion related to chronicity,which is in accordance to prior PET studies. Preserved brain perfusionin both gray or white matter in our sample of participants withschizophrenia may in part be due to antipsychotic treatment whichmay have normalized cerebral blood volume and flow. Futureperfusion imaging studies should try to explore the effects ofantipsychotic treatment by longitudinally follow ill participants atfirst-episode.

Acknowledgements

This work was partly supported by grants from the AmericanPsychiatric Institute for Research and Education (APIRE), the ItalianMinistry for University and Research (PRIN n. 2005068874), and theItalian Ministry of Health (IRCCS “E. Medea”) to Dr. Paolo Brambilla.

References

Agarwal, N., Port, J.D., Bazzocchi, M., Renshaw, P.F., 2010. Update on the use of MR forassessment and diagnosis of psychiatric diseases. Radiology 255, 23–41.

Amaddeo, F., Burti, L., Ruggeri, M., Tansella, M., 2009. Long-term monitoring andevaluation of a new system of community-based psychiatric care. Integratingresearch, teaching and practice at the University of Verona. Annali dell' IstitutoSuperiore di Sanità 45, 43–53.

Andreasen, N.C., O'Leary, D.S.O., Flaum, M., Nopoulos, P., Watkins, G.L., Boles Ponto, L.L.,Hichwa, R.D., 1997. Hypofrontality in schizophrenia: distributed dysfunctionalcircuits in neuroleptic-naïve participants. Lancet 349, 1730–1734.

Andreasen, N.C., Paradiso, S., O'Leary, D.S., 1998. “Cognitive dysmetria” as an integrativetheory of schizophrenia: a dysfunction in cortical–subcortical–cerebellar circuitry?Schizophrenia Bulletin 24, 203–218.

Bachneff, S.A., 1996. Regional cerebral blood flow in schizophrenia and the local circuitneurons hypothesis. Schizophrenia Bulletin 22, 163–182.

Bellani, M., Fagnani, C., Brambilla, P., 2009a. Twin studies in psychotic disorders.Epidemiologia e Psichiatria Sociale 18, 195–199.

Bellani, M., Marzi, C.A., Brambilla, P., 2009b. Interhemispheric communication inschizophrenia. Epidemiologia e Psichiatria Sociale 18, 104–106.

Bellani, M., Perlini, C., Brambilla, P., 2009c. Language disturbances in schizophrenia.Epidemiologia e Psichiatria Sociale 18, 314–317.

Benner, T., Heiland, S., Erb, G., Forsting, M., Sartor, K., 1997. Accuracy of gamma-variatefits to concentration–time curves from dynamic susceptibility-contrast enhancedMRI: influence of time resolution, maximal signal drop and signal-to-noise.Magnetic Resonance Imaging 15, 307–317.

Ben-Shachar, D., Bonne, O., Chisin, R., Klein, E., Lester, H., Aharon-Peretz, J., Yona, I.,Freedman, N., 2007. Cerebral glucose utilization and platelet mitochondrialcomplex I activity in schizophrenia: a FDG-PET study. Progress in Neuro-Psychopharmacology and Biological Psychiatry 31, 807–813.

Brambilla, P., Cerini, R., Fabene, P.F., Andreone, N., Rambaldelli, G., Farace, P., Versace, A.,Perlini, C., Polizza, L., Gasparini, A., Gatti, R., Bellani, M., Dusi, N., Barbui, C., Nosè, M.,Tournikioti, K., Sbarbati, A., Tansella, M., 2007. Assessment of cerebral bloodvolume in schizophrenia: a magnetic resonance imaging study. Journal ofPsychiatric Research 41, 502–510.

52 M. Bellani et al. / Psychiatry Research: Neuroimaging 193 (2011) 46–52

Calamante, F., Gadian, D.G., Connelly, A., 2002. Quantification of perfusion using bolustracking magnetic resonance imaging in stroke: assumptions, limitations, andpotentials for clinical use. Stroke 33, 1146–1151.

Chen, J.J., Frayne, R., Smith, M.R., 2005. Reassessing the clinical efficacy of two MRquantitative DSC PWI CBF algorithms following cross-calibration with PET images.Physics in Medicine and Biology 50, 1251–1263.

Cohen, B.M., Yurgelun-Todd, D., English, C.D., Renshaw, P.F., 1995. Abnormalities ofregional distribution of cerebral vasculature in schizophrenia detected by dynamicsusceptibility contrast MRI. The American Journal of Psychiatry 152, 1801–1803.

Crespo-Facorro, B., Paradiso, S., Andreasen, N.C., O'Leary, D.S., Watkins, G.L., Boles Ponto,L.L., Hichwa, R.D., 1999. Recalling word lists reveal “cognitive dysmetria” inschizophrenia: a positron emission tomography study. The American Journal ofPsychiatry 156, 386–392.

Davidson, L.L., Heinrichs, R.W., 2003. Quantification of frontal and temporal lobe brain-imaging findings in schizophrenia: a meta-analysis. Psychiatry Research: Neuroima-ging 122, 69–87.

Davis, C.E., Jeste, D.V., Eyler, L.T., 2005. Review of longitudinal functional neuroimagingstudies of drug treatments in participants with schizophrenia. SchizophreniaResearch 78, 45–60.

Edelman, R.R., Mattle, H.P., Atkinson, D.J., Hill, T., Finn, J.P., Mayman, C., Ronthal, M.,Hoogewoud, H.M., Kleefield, J., 1990. Cerebral blood flow: assessment with dynamiccontrast-enhanced T2*-weighted MR imaging at 1.5 T. Radiology 176, 211–220.

Ertugrul, A., Volkan-Salanci, B., Basar, K., Karli Oguz, K., Demir, B., Ergun, E.L., Senturk, S.,Erbas, B., Cila, A., Ulug, B., 2009. The effect of clozapine on regional cerebral bloodflow and brain metabolite ratios in schizophrenia: relationship with treatmentresponse. Psychiatry Research: Neuroimaging 174, 121–129.

Fujimoto, T., Takeuch, K., Matsumoto, T., Kamimura, K., Hamada, R., Nakamura, K., Kato,N., 2007. Abnormal glucose metabolism in the anterior cingulate cortex inparticipants with schizophrenia. Psychiatry Research: Neuroimaging 154, 49–58.

Gur, R.E., Skolnick, B.E., Gur, R.C., Caroff, S., Rieger, W., Obrist, W.D., Younkin, D., Reivich,M., 1983. Brain function in psychiatric disorders. I. Regional cerebral blood flow inmedicated schizophrenics. Archives of General Psychiatry 40, 1250–1254.

Gur, R.E., Turetsky, B.I., Loughead, J., Snyder, W., Kohle, C., Elliott, M., RamapriyanPratiwadi, R., Ragland, J.D., Bilker, W.B., Siegel, S.J., Kanes, S.J., Arnold, S.E., Gur, R.C.,2007. Visual attention circuitry in schizophrenia investigated with oddball event-related functional magnetic resonance imaging. The American Journal of Psychiatry164, 442–449.

Hanson, D.R., Gottesman, I.I., 2005. Theories of schizophrenia: a genetic-inflammatory-vascular syntesis. Medical Genetics 6, 7.

Kanahara, N., Shimizu, E., Sekine, Y., Uchida, Y., Shibuya, T., Yamanaka, H., Hashimoto, T.,Asaka, T., Sasaki, T., Miyatake, R., Ohkami, T., Fukami, G., Fujisaki, M., Watanabe, H.,Shirayama, Y., Hayashi, H., Hashimoto, K., Asano, M., Iyo, M., 2009. Doeshypofrontality expand to global brain area in progression of schizophrenia?: across-sectional study between first-episode and chronic schizophrenia. Progressesin Neuropsychopharmacology and Biological Psychiatry 33, 410–415.

Kaneoke, Y., Furuse, M., Yoshida, K., Saso, K., Ichihara, K., Motegi, Y., 1989. Transferindex of MR relaxation enhancer: a quantitative evaluation of MR contrastenhancement. American Journal of Neuroradiology 10, 329–333.

Kim, J.J.,Mohamed,S., Andreasen,N.C.,O'Leary,D.S.,Watkins, G.L., Boles Ponto, L.L.,Hichwa,R.D., 2000. Regional neural dysfunctions in chronic schizophrenia studied withpositron emission tomography. The American Journal of Psychiatry 157, 542–548.

Lahti, A.C., Holcomb, H.H., Weiler, M.A., Medoff, D.R., Tamminga, C.A., 2003. Functionaleffects of antipsychotic drugs: comparing clozapine with haloperidol. BiologicalPsychiatry 53, 601–608.

Laurens, K.R., Kiehl, K.A., Ngan, E.T., Liddle, P.F., 2005. Attention orienting dysfunctionduring salient novel stimulus processing in schizophrenia. Schizophrenia Research15, 159–171.

Lindenmayer, J.P., Khan, A., Iskander, A., Abad, M.T., Parker, B., 2007. A randomizedcontrolled trial of olanzapine versus haloperidol in the treatment of primarynegative symptoms and neurocognitive deficits in schizophrenia. The Journal ofClinical Psychiatry 68, 368–379.

Loeber, R.T., Sherwood, A.R., Renshaw, P.F., Cohen, B.M., Yurgelun-Todd, D.A., 1999.Differences in cerebellar blood volume in schizophrenia and bipolar disorder.Schizophrenia Research 37, 81–89.

Loeber, R.T., Gruber, S.A., Cohen, B.M., Renshaw, P.F., Sherwood, A.R., Yurgelun-Todd, D.A.,2002. Cerebellar blood volume in bipolar participants correlates with medication.Biological Psychiatry 51, 370–376.

Malaspina, D., Harkavy-Friedman, J., Corcoran, C., Mujica-Parodi, L., Printz, D., Gorman,J.M., Van Heertum, R., 2004. Resting neural activity distinguishes subgroups ofschizophrenia participants. Biological Psychiatry 56, 931–937.

Miller, D.D., Rezai, K., Alliger, R., Andreasen, N.C., 1997. The effect of antipsychoticmedication on relative cerebral blood perfusion in schizophrenia: assessment with

technetium-99m hexamethyl-propyleneamine oxime single photon emissioncomputed tomography. Biological Psychiatry 41, 550–559.

Miller, D.D., Andreasen, N.C., O'Leary, D.S., Watkins, G.L., Boles Ponto, L.L., Hichwa, R.D.,2001. Comparison of the effects of risperidone and haloperidol on regional cerebralblood flow in schizophrenia. Biological Psychiatry 49, 704–715.

Molina, V., Gispert, J.D., Reig, S., Sanz, J., Pascau, J., Santos, A., Des, 2005. Cerebralmetabolic changes induced by clozapine in schizophrenia and related to clinicalimprovement. Psychopharmacology 178, 17–26.

Novak, B., Milcinski, M., Grmek, M., Kocmur, M., 2005. Early effects of treatment onregional cerebral blood flow in first episode schizophrenia participants evaluatedwith 99Tc-ECD-SPECT. Neuroendocrinology Letters 26, 685–689.

Ortuño, F., Mobreno-Iñiguez, M., Millán, M., Soutullo, C.A., Bonelli, R.M., 2006. Corticalblood flow during rest and Wisconsin Card Sorting Test performance inschizophrenia. Wiener Medizinische Wochenschrift 156, 179–184.

Østergaard, L., Weisskoff, R.M., Chesler, D.A., Gyldensted, C., Rosen, B.R., 1996a. Highresolution measurements of cerebral blood flow using intravascular tracer boluspassages. Part I: mathematical approach and statistical analysis. MagneticResonance in Medicine 36, 715–725.

Østergaard, L., Sorensen, A.G., Kwong, K.K., Weisskoff, R.M., Gyldensted, C., Rosen, B.R.,1996b. High resolution measurement of cerebral blood flow using intravasculartracer bolus passages. Part II: experimental comparison and preliminary results.Magnetic Resonance in Medicine 36, 726–736.

Porkka, L., Neuder, M., Hunter, G., Weisskoff, R.M., Belliveau, J., Rosen, B.R., 1991.Arterial input function measurement with MRI. Proceedings of ISMRM 10th AnnualMeeting, San Francisco, California, USA, p. 120.

Rosen, B.R., Belliveau, J.W., Aronen, H.J., Kennedy, D., Buchbinder, B.R., Fishman, A.,Gruber, M., Glas, J., Weisskoff, R.M., Cohen, M.S., Hochberg, F.H., Brady, T.J., 1991.Susceptibility contrast imaging of cerebral blood volume: human experience.Magnetic Resonance in Medicine 22, 293–299.

Sakaie, K.E., Shin, W., Curtin, K.R., McCarthy, R.M., Cashen, T.A., Carroll, T.J., 2005.Method for improving the accuracy of quantitative cerebral perfusion imaging.Journal of Magnetic Resonance Imaging 21, 512–519.

Scheef, L., Manka, C., Daamen, M., Kühn, K.U., Maier, W., Schild, H.H., Jessen, F., 2010.Resting-state perfusion in nonmedicated schizophrenic participants: a continuousarterial spin-labeling 3.0-T MR study. Radiology 256, 253–260.

Schlösser, R., Gesierich, T., Kaufmann, B., Vucurevic, G., Hunsche, S., Gawehn, J., Stoeter,P., 2003. Altered effective connectivity during working memory performance inschizophrenia: a study with fMRI and structural equation modelling. Neuroimage19, 751–763.

Steinberg, J.L., Devous, M.D., Moeller, F.G., Paulman, R.G., Raese, J.D., Gregory, R.R., 1995.Cerebellar blood flow in schizophrenic participants and normal control subjects.Psychiatry Research: Neuroimaging 61, 15–31.

Tansella, M., Burti, L., 2003. Integrating evaluative research and community-basedmental health care in Verona, Italy. The British Journal of Psychiatry 183, 167–169.

Tansella, M., Nardini, M. (Eds.), 1996. Schede di valutazione clinica in neuropsichiatria.SCAN 2.1. Il Pensiero Scientifico Editore, Rome.

Théberge, J., 2008. Perfusion magnetic resonance imaging in psychiatry. Top MagneticResonance Imaging 19, 111–130.

Ventura, J., Nuechterlein, K.H., Subotnik, K.L., Gutkind, D., Gilbert, E.A., 2000. Symptomdimensions in recent-onset schizophrenia andmania: a principal component analysisof the 24-item Brief Psychiatric Rating Scale. Psychiatry Research 97, 129–135.

Wake, R., Miyaoka, T., Kawakami, K., Tsuchie, K., Inagaki, T., Horiguchi, J., Yamamoto, Y.,Hayashi, T., Kitagaki, H., 2010. Characteristic brain hypoperfusion by 99mTc-ECDsingle photon emission computed tomography (SPECT) in participants with thefirst-episode schizophrenia. European Psychiatry 25, 361–365.

Whitfield-Gabrieli, S., Thermenos, H.W., Milanovic, S., Tsuang, M.T., Faraone, S.V.,McCarley, R.W., Shenton, M.E., Green, A.I., Nieto-Castanon, A., LaViolette, P., Wojcik,J., Gabrieli, J.D., Seidman, L.J., 2009. Hyperactivity and hyperconnectivity of thedefault network in schizophrenia and in first-degree relatives of persons withschizophrenia. Proceedings of the National Academy of Sciences of the UnitedStates of America 106, 1279–1284.

Wing, J.K., Babor, T., Brugha, T., Burke, J., Cooper, J.E., Giel, R., Jablenski, A., Regier, D.,Sartorius, N., 1990. SCAN. Schedules for clinical assessment in neuropsychiatry.Archives of General Psychiatry 47, 589–593.

Wu, O., Østergaard, L., Weisskoff, R.M., Benner, T., Rosen, B.R., Sorensen, A.G., 2003.Tracer arrival timing-insensitive technique for estimating flow in MR perfusion-weighted imaging using singular value decomposition with a block-circulantdeconvolution matrix. Magnetic Resonance in Medicine 50, 164–174.

Zierler, K.L., 1962. Theoretical basis of indicator-dilution methods for measuring flowand volume. Circulation Research 10, 393–407.

Zierler, K.L., 1965. Equations for measuring blood flow by external monitoring ofradioisotopes. Circulation Research 16, 309–321.