Psychiatry Research: Neuroimaging 123 (2003) 65–79

Rightward cerebral asymmetry in subtypes ofschizophrenia according to Leonhard’s classification and to DSM-IV: a structural MRI study

Paulo C. Salletaa , Helio Elkis , Taniaˆ´ M. Alves a , Jose R. Oliveira aa , Erlei Sassi , Claudio Campi de Castrobaa , Geraldo F. Busatto , Wagner F. Gattaz, *

aDepartment of Psychiatry, Faculty of Medicine, University of Sao˜ Paulo, Rua Ovidio Pires de Campos syn, Sao˜ Paulo CEP 05403-010, Brazil bDepartment of Radiology, Faculty of Medicine, University of Sao˜ Paulo, Rua Ovidio Pires de Campos syn, Sao˜ Paulo CEP 05403-010, Brazil

Received 25 February 2002; received in revised form 16 October 2002; accepted 12 November 2002

Abstract

Although well documented, brain structural abnormalities in schizophrenia are non-specific, and morphometric parameters show significant overlap between patients and healthy controls. Such inconsistencies in neuroimaging findings could represent different levels of severity along a single pathogenic process or distinct clinical and etiopathological psychoses within a schizophrenic spectrum. The aim ofthe present study was the investigation of distinct brain abnormalities in different subtypes of schizophrenia. Forty patients were classified according to DSM- IV and Leonhard’s classifications. Psychopathology was assessed by the Positive and Negative Syndrome Scale (PANSS) and the Negative Symptom Rating Scale (NSRS). Patients were compared to 20 healthy volunteers on volumetric measures ofcerebral structures (hemisphere, hippocampus and planum temporale) and ventricular–brain ratio (VBR) obtained by magnetic resonance imaging. Patients showed rightward asymmetry ofcerebral hemispheres and increased VBR. Rightward asymmetry correlated with severity ofnegative symptoms and prevailed in the systematic forms of Leonhard, suggesting a distinct pattern of left hemisphere abnormality in this subgroup of psychoses. Increased VBR values showed a single normal distribution in the subgroups, indicating that ventricular enlargement is not restricted to a subgroup but is present to a certain degree in all cases. ᮊ 2003 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Schizophrenia; Magnetic resonance imaging; Morphometry; Subtyping; Symptoms

1. Introduction is associated with brain structural abnormalities. The most robust findings are enlargement of lateral In the last decades, cumulating evidence from and third ventricles, decreased brain volume, vol- neuroimaging studies has shown that schizophrenia umetric reduction in temporal lobe and limbic structures, and decreased volumes ofsubcortical *Corresponding author. Tel.: q55-11-3062-9029; fax: q55- structures such as the caudate and thalamus in 11-3083-6588. neuroleptic-naive patients (Harrison, 1999; Shen- ( ) E-mail address: [email protected] W.F. Gattaz . ton et al., 2001). 0925-4927/03/$ - see front matter ᮊ 2003 Elsevier Science Ireland Ltd. All rights reserved. PII: S0925-4927Ž03.00020-9 66 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79

With reference to the timing of brain changes, 1999) or the psychoses ofthe schizophrenia spec- ventricular enlargement and cortical volume reduc- trum constitute distinct clinical and etiopathologi- tion are already present in first-episode patients cal groups (Tsuang and Faraone, 1995; Franzek (Gur et al., 1998; Zipursky et al., 1998), and the and Beckmann, 1998). absence ofgliosis and other neurodegenerative Karl Leonhard (1904–1988) developed his clas- abnormalities (Harrison, 1997) favors the neuro- sification of psychoses based on previous works developmental hypothesis ofschizophrenia. More ofWernicke (1900) and Kleist (1934) that empha- controversial is the question ofwhether such sized the association between particular forms of changes are static (Jaskiw et al., 1994; Vita et al., schizophrenia and abnormalities ofbrain systems, 1997), progressive (Gattaz et al., 1981; DeLisi et in analogy to the anatomofunctional substrate of al., 1997) or a combination ofboth (Gur et al., neurological degenerative disorders (Leonhard, 1998). 1995). In the case ofneurological degenerative Using MRI, standardized measurements have disorders, even when anatomopathological meth- been developed to identify other putative brain ods do not allow the identification of an organic structural abnormalities in schizophrenic patients. substrate, it is possible to identify a regular clinical The planum temporale (PT) is believed to consist syndrome related to the affected neurological sys- ofthe association auditory cortex, which is related tem (e.g. the pyramidal tract syndrome in spastic to the integration and processing oflanguage palsy). In the case ofschizophrenia, the anatomic (Shapleske et al., 1999). The most right-handed substrates are unknown or very imprecise, but we healthy individuals have the left PT larger than can recognize defect state syndromes (‘Defektzu- the right PT, reflecting the physiologic asymmetry stand Syndrome’) composed ofregular and specif- ofregions involved in language processing ic symptomatological complexes (‘Symptom- (Geschwind and Levitsky, 1968). Several MRI verbande’¨ ), which point to differently disturbed studies with schizophrenic patients have demon- ‘psychic systems’. However, in comparison with strated symmetry or reversed asymmetry ofthe PT neurological disorders, these systems are connected (Kwon et al., 1999; Hirayasu et al., 2000). Others with more extended functional areas and, hence, have found associations between reversed asym- structural deficits should be too diffuse in brain metry ofthe PT and severity ofpsychotic symp- areas to be easily detected (Leonhard, 1970). Such toms (Rossi et al., 1994; Flaum et al., 1995; Petty statements are particularly pertinent to the so- et al., 1995), raising the hypothesis that some called systematic schizophrenias (SS), which schizophrenic symptoms could result from dis- according to Leonhard show insidious onset, pro- turbed functional lateralization of brain areas relat- gressive course and well-delimited symptomatolo- ed to language. Nevertheless, several studies gy. On the other hand, non-systematic forms (NSS) reported negative findings, either with reference to show a multiplicity ofsymptoms, frequently the hypothesis ofreversed asymmetry (Kulynych including elements characteristic ofother psycho- et al., 1995; O’Leary et al., 1995; Frangou et al., ses, such as cycloid and manic-depressive psycho- 1997) or to the correlation with symptomatology ses and other non-systematic forms. In contrast to (DeLisi et al., 1994; Kleinschmidt et al., 1994; SS, NSS forms present periods of crisis intermin- Barta et al., 1997). gled with periods ofpartial remission, and bipolar Indeed, those morphometric changes are non- characteristics at the level ofcertain functionssuch specific, showing significant overlap between as affect, formal thought and psychomotor schizophrenia and other neuropsychiatric condi- functions. tions, and even for the most robust findings there A previous MRI study involving endogenous are a number ofstudies reporting negative results. psychoses defined according to Leonhard’s classi- Such inconsistencies raise the question whether fication showed that SS had reduced temporal lobe ‘schizophrenia’ is a single disease with different volumes compared with NSS (Serfling et al., levels ofseverity along a pathogenetic continuum 1995). According to the low genetic risk and high (Goldberg and Weinberger, 1995; Andreasen, incidence ofmaternal infectionsobserved in sys- P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 67 tematic forms (Franzek and Beckmann, 1996; daily dose of811 mg (S.D. "425)(American Stober¨ et al., 1994, 1997), Serfling and colleagues Psychiatric Association, 1997). Exclusion criteria (1995) interpreted the finding of volumetric reduc- were: (a) age lower than 18 or greater than 60 tion in the temporal lobes as evidence ofexoge- years; (b) history ofsubstance dependence; and nous damage acting on brain development. (c) history ofhead trauma, degenerative neurolog- It is possible that the inconsistencies in neuroim- ical disorders and previous treatment with steroid aging findings reflect the heterogeneity of the medication. A control group of20 healthy volun- studied populations. As far as psychopathology is teers, recruited from the community, was selected concerned in the search for more homogeneous using the same exclusion criteria above (based on subgroups, we predicted that a diagnostic system the SCID-IyP interview). A written informed con- grounded also in the longitudinal course ofsymp- sent was obtained from all subjects before partici- toms (Leonhard’s classification) would provide pation. Local research and ethics committees more validity than the cross-sectional criteria approved the study. adopted in the diagnostic decision ofDSM-IV Current symptom severity was measured using subtypes. Therefore, we performed the present the Positive and Negative Syndrome Scale study to investigate commonly reported brain mor- (PANSS)(Kay et al., 1987) and the Negative phologic abnormalities associated with schizophre- Symptom Rating Scale (NSRS)(Iager et al., nia in discrete subgroups ofschizophrenic patients 1985). Handedness was assessed with the Hand- classified according to Leonhard, compared to the edness Inventory (Briggs and Nebes, 1975). All DSM-IV subgroups. the patients were receiving stable doses ofanti- psychotics at the time ofthe study. Clinical ratings 2. Methods were obtained by one psychiatrist (T.M.A.) blind to the MRI results. 2.1. Subjects and clinical assessment 2.2. MRI acquisition and volumetric measurements Forty schizophrenic patients were diagnosed according to DSM-IV (American Psychiatric Asso- Structural MRI scans ofthe entire brain in both ciation, 1994), based on the Portuguese version of patients and controls were obtained using a 1.5 T the Structured Clinical Interview for DSM-IV— Philips Gyroscan S15-ACS scanner, with T1-FFE Patient Edition (SCID-IyP)(First et al., 1996). weighed continuous coronal slices 1.2 mm thick, Patients were classified in DSM-IV subtypes based FOVs240, matrix 256=256. Coronal images on their clinical current state. were oriented in planes perpendicular to the ante- The classification according to Leonhard’s cri- rior-to-posterior commissural axis. Images were teria is based on a careful examination of the transferred to a SUN Workstation and measure- association ofsymptoms along the longitudinal ments were manually performed using software course, which reduces considerably the value of (Gyroview—HR 2.1) that also permitted recon- cut-off evaluations at one simple point in time and struction in axial and sagittal planes. This was renders the use ofstructured questionnaires not important to ensure accuracy in delineation of entirely reliable (Leonhard, 1990, 1995). The diag- regions ofinterest, especially ofthe planum nostic assessment ofour patients was based on temporale. several interviews with a structured questionnaire Measurements were performed in the coronal (Ban, 1982) as well as a review ofhospital charts. plane: hippocampus and planum temporale were Subjects were consecutively admitted from the measured at every 1.2-mm continuous slices; and specialized outpatient clinic at the Institute of ventricles and brain volumes at a distance of3.6 Psychiatry, University ofSao˜ Paulo, a psychiatry mm. Volumes were calculated multiplying the sum tertiary referral center. Patients were taking typical ofareas by the thickness ofslices. Absolute values (12 patients) or atypical antipsychotics (28 were corrected using the algorithm: corrected val- patients), with a mean chlorpromazine-equivalent uesabsolute value=brain volumey1000. Meas- 68 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 urements were performed blind to the group status 2.2.4. Brain volume (BV) ofall subjects. Brain volume was obtained from the sum of the volumes ofhemispheres (Hem). In the middle 2.2.1. Hippocampus (H) segment of the brain, a line traced from the inferior The anterior border was designed as the first point ofthe third ventricle to the curvature ofthe slice showing the pes hippocampus, separated parahippocampal gyrus excluded part ofthe mid- superiorly from the amygdala by the alveus brain, pons and cerebellum. For reasons explained (included), laterally by the temporal horn oflateral elsewhere (Sallet et al., submitted), our measure- ventricle, and lateroinferiorly by the grayywhite ments did not include the temporal pole, but only interface between the subiculum and the white the segment ofthe temporal lobe continuous to matter ofthe parahippocampal gyrus. Along the the remainder telencephalon, as seen in the coronal hippocampal axis, the medial border was arbitrarily plane. An asymmetry coefficient between hemi- designated as a perpendicular line in the middle spheres was calculated using the algorithm ofthe parahippocampal–subiculum curvature. The ACHemsw(RHemyLHem)y0.5=(RHemq posterior border was at the point where the crus LHem)x=1000. ofthe fornixappears in total extension. Anatomic landmarks were based on the study ofNaidich et 2.2.5. Reliability ( ) al. 1987 . To ensure consistent delineation, structures were always measured by the same operator (P.C.S.) ( ) 2.2.2. Planum temporale PT blind to diagnosis. Two other investigators (J.R.O The anterior border was defined according to and E.S.) carried out a number ofmeasurements ( ) Pfeifer’s criteria Von Economo and Horn, 1930 to assess inter-rater reliability. The results were: as the transverse sulcus lying behind the anterior right hippocampus (RH)(ICCs0.85; Ps0.000), ( ) transverse gyrus Heschl’s gyrus . In case ofmore LH (ICCs0.84; Ps0.000); right planum tempor- than one Heschl’s gyrus, Beck’s intermedious ale (RPT)(ICCs0.84; Ps0.000), LPT (ICCs ( sulcus was identified its origin is not posterior to 0.91; Ps0.000); right VBR (ICCs0.98; Ps ) the insula ofReil and measurement began from 0.000), and LVBR (ICCs0.98; Ps0.000). The the sulcus immediately posterior to it. The poste- ICC coefficient was calculated according to the rior border was defined as the caudal end of the methods described by Bartko and Carpenter ( ) horizontal ramus ofthe Sylvian fissure SF . The (1976). inferior boundary was the interface between the gray and white matter ofthe superior temporal gyrus. An asymmetry coefficient of the planum 2.3. Statistical analysis temporale was calculated using the algorithm pro- posed by Galaburda et al. (1987): ACPTs(RPTy The statistical analysis was conducted using LPT)y0.5=(RPTqLPT). Positive values indicate SPSS version 10.0 for Windows, 1999. Between- rightward asymmetry; negative values indicate left- group differences were assessed using parametric ward ‘normal’ asymmetry. (t-test, one-way ANOVA) and non-parametric tests (Mann–Whitney, Kruskal–Wallis), depending on 2.2.3. Ventricular–brain ratio (VBR) the normality distribution. Secondly, we used a Ventricular volumes were measured at every 3.6 general linear model controlling for brain volume mm in coronal slices. The VBR was obtained with and sociodemographic variables (gender, age, the algorithm: VBRsventricular volumeybrain handedness and years ofeducation ) and with volume=100. Right and left VBR were calculated Bonferroni correction for multiple comparisons. separately in order to investigate interhemispheri- Correlations between neuroimaging measure- cal differences (lateral ventricles included temporal ments and rating scale scores were performed first horns and third ventricle was divided by a mean by bivariate correlation (Spearman rho), and sec- line). ondly, with partial correlation controlling for P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 69 sociodemographic differences. The significance (RHem)LHem) than the control group (GLM: level was set at two-tailed P-0.05. Fs4.6, d.f.s1, 54, Ps0.036). Mean brain and hemisphere volumes ofpatients and controls were 3. Results not significantly different. Systematic schizophrenias showed an asymme- 3.1. Demographic and clinical findings (Table 1) try coefficient of hemispheres significantly higher ( ) ) ( s Patients and healthy controls were not signifi- RHem LHem than healthy controls F 3.6, s s ) cantly different in terms of gender, handedness d.f. 2, 53, P 0.035 , while non-systematic forms and years ofeducation, but patients were generally and healthy controls showed similar asymmetry ( ) older than controls (Ps0.09). coefficients Table 2 . From the 40 schizophrenic patients, according to DSM-IV subtypes, 27 were classified as para- 3.2.2. Hippocampus noid, eight as disorganized and five as residual. In schizophrenic patients, the volume ofthe left Considering the unequal distribution in the sub- hippocampus was significantly smaller than that groups, for the sake of statistical power in the ofthe right hippocampus (paired t-testsy3.6; comparison between them, we subdivided DSM- d.f.s39; Ps0.001); this difference was not ( s ) IV subtypes into paranoid n 27 and non-para- observed in healthy controls. ( s ) noid n 13 . Paranoid subtype patients were There were no other significant differences, ( generally younger than normal controls Kruskal– either between patients and normal controls or 2s s ) Wallis: x 4.6, P 0.1 and, when compared with between subgroups, in terms ofasymmetry coef- non-paranoid patients, were being treated with ficients and volumes of the hippocampus. relatively lower doses ofantipsychotics (tsy1.9, Ps0.067), and had lower duration ofillness (ts y3.6, Ps0.001), higher age at onset (ts5.7, P- 3.2.3. Planum temporale 0.000) and lower scores in negative symptoms PT volumes and asymmetry coefficients showed (Table 1). no significant differences between groups. According to Leonhard’s classification, 18 patients were defined as suffering from non-sys- 3.2.4. Ventricles tematic schizophrenias (NSS)(16 affective para- Compared to normal controls, schizophrenic phrenia, 1 cataphasia and 1 periodic catatonia) and patients showed bilaterally enlarged coronal VBR, 22 as systematic schizophrenias (SS)(13 para- even with demographic variables as covariates ) phrenia, 8 hebephrenia and 1 catatonia . For the (Fs7.1, d.f.s1, 54, Ps0.01). Asymmetry coef- same reason mentioned above, we have limited ficients of VBR were not significantly different statistical analysis to NSS and SS subgroups. NSS between groups. and SS were statistically similar in terms ofage, Compared with healthy controls, both paranoid gender, education, duration ofillness and chlor- (ANOVA: Fs3.9, d.f.s2, 57, Ps0.018) and non- promazine-equivalent doses ofantipsychotics, but paranoid (Ps0.022) subgroups ofDSM-IV schiz- SS patients included the four left-handers, showed ( s s ) ophrenia showed increased VBR. However, lower age at onset ofdisease t 2.7, P 0.009 controlling for demographic variables, only the and showed higher scores in negative symptoms, paranoid subgroup retained a statistically signifi- as assessed by the NSRS (Table 1). cant difference (Fs3.5, d.f.s2, 53, Ps0.037). 3.2. Group differences from healthy controls (Table Both forms defined by Leonhard showed signif- 2) icantly increased VBR compared to healthy con- trols, but after controlling for demographic 3.2.1. Brain and hemisphere volumes variables, both non-systematic (Fs3.5, d.f.s2, Schizophrenic patients showed an asymmetry 53, Ps0.076) and systematic forms (Ps0.085) coefficient of hemispheres significantly higher showed only trends towards increased VBR. 70 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 1 2 383 ) 3 3.5 9.0 8.3 9.4 8.6 12.6 25.1 7.1 day. ) ) ) " y 22 " " " " " " " " " s 18:4 16:6 5:17 n 0.009 16.5 17.4 s P Antipsychotics ) * ( 2.7, 1 2 s 475 864 )( 4 7.5 7.6 32.7 7.24.3 16.2 8.58.1 22.1 10.3 35.7 21.9 75.0 6.4 17.7 )( )( )( " 18 " " " " " " " " " -Test t s 747 18:0 8:10 7:11 n ( ( Leonhard’s classification 35.6 15.2 19.8 17.6 10.3 1. 4 2 5 1 3 6 S.D. chlorpromazine-equivalent mg 124 )( " 3.4 8.6 8.27 1.3 10.3 9.9 8.3 11.4 35.3 25.5 70.8 7.0 17.9 " 13 " " " " " " " " " systematic schizophrenias. )( s s ) 988 4:9 n 21.9 14.6 24.0 21.9 0.1 and subgroups according to DSM-IV and Leonhard ) s P 40 s n 4.6, ( s 2 0.014 x ( s P 2.6, 4 ; doses are expressed as mean ) 2 3 5 6 sy 396 )( 3.4 7.5 6.9 7.7 36.9 4.2 11.7 35.5 22.7 77.4 6.2 6.7 17.9 7.4 )( " 27 " " " " " " " " " -Test t s non-systematic schizophrenias; SS 8:19 n DSM-IV 32.6 Paranoid( Non-paranoid NSS SS 1. Kruskal–Wallis , schizophrenic patients s ) 20 s n ( 0.1. NSS F 0.001 2. 0.067 0.045 0.002 ) P s 0.000 s s s clozapine and risperidone ( P P P P - 0.094 P 3.6, 1.9, 2.1, 4.4, s P 5.7, sy s sy sy sy 1 3 8.2 426 726 ) 1.7, 8.4 12.9 4.211.5 19.6 35.6 12.5 71.0 8 9.2 10.5 6.7 17.7 8.8 18.1 )( " " and atypical 40 " " " " " " " " -Test -Test -Test -Test -Test ) t t t t t s sy 12:28 n Z 2. 3. Schizophre- 34.0 4. 5. 6. ( 1 )( 10.8 88 20 " " s n 11 16:4 34:4 25:2 11:2 12:8 24:16 16:11 8:5 31.5 Healthy ( )( -values are reported only for differences with haloperidol and chlorpromazine ) ) P ( ) R:L ) ( years S.D. ) ( years ( M:F " ) ( day 811 years y ( typical:atypical years mg illness ( ( Means Education Handedness Duration of15.8 Age at onset 18.0 Gender PANSS general 35.6 PANSS total 73.1 *Antipsychotics: Table 1 Comparison ofdemographic and clinical data between healthyDemographic controls and clinicalAge controls nic patients Statistic 1. Mann–Whitney data included typical NSRS 14.2 PANSS positive 17.8 PANSS negative 20.2 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 71 ’s 0.7 0.8 1.0 1.0 0.7 0.9 0.010 0.037 0.063 0.076 0.085 0.007 0.022 0.037 0.018 0.009 " " " " " " s s s s s s s s s s P P P P P P P P P P 1 3 2 3 2 3 2 3 2 3 VBR 0.29 2.18 0.14 2.29 0.31 2.32 0.18 2.14 0.25 2.22 0.30 1.59 . Volume values are " " " " " right and left hippo- " ) s 0.03 ACPT y Bonferroni general linear model controlling ( ) 0.45 0.01 0.51 0.09 0.51 0.03 0.43 0.05 0.54 0.46 0.04 3 not significant. ( " " " " " " s LPT 0.54 1.99 0.56 1.88 0.46 1.88 0.63 1.98 0.61 1.98 0.64 1.92 " " " " " " RPT ventricular-brain ratio; ns 0.38 1.99 0.44 2.09 0.41 1.97 0.39 2.08 0.40 1.93 0.40 2.02 s " " " " " " adjusted for multiple comparisons ) LH 1000. asymmetry coefficient of hemispheres; R and L H y s 0.39 3.15 0.44 3.24 0.43 3.22 0.39 3.13 0.35 3.29 0.41 3.18 ANOVA with least significant difference; " " " " " " ) 2 ( RH ns nsns ns ns ns nsns ns ns ns ns ns ns ns brain volume -test; t = ) 4 3.32 3.6 3.45 4 3.40 3.9 3.36 4 3.31 4.2 3.35 1 ( " " " " " " 0.036 0.053 0.035 0.021 0.072 s s s s s 0.2 1.1 P P ns P P P 1 3 2 3 2 3 ACHem y y asymmetry coefficient of PT; VBR s absolute volume 60 53 63 0.5 72 1.7 76 1.8 69 0.9 0.020 0.047 s " " " " " " s s 0.1 are reported: right and left hemisphere; ACHem P P 2 3 LHem - s ) gender, age, handedness and years ofeducation ( 61 526 50 558 62 547 74 505 74 539 68 534 0.036 " " " " " " two-tailed s ( P ns 2 3 RHem -values 120 526 102 555 125 548 146 508 149 543 136 535 P " " " " " " 0.072 0.026 s s S.D. P P 2 3 BV ns ns nsnsns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns " brain volume; R and L Hem right and left planum temporale; ACPT s s ) ) ) BV 18 22 s s 40 n n : ( ( s ) n ( ) ) ) 20 27 13 s s s Leonhard n n n ( Abbreviations: controls ( ( ( schizophr. schizophr. Values indicate mean patients Non-systematic 1052 classifications Table 2 Comparison ofmorphometric parameters between normal controls vs.Subjects schizophrenia and normal controlsHealthy vs. subtypes according to DSM-IV and Leonhard 1113 DSM-IV: ParanoidNon-paranoid 1096 Forms 1013 Systematic 1083 campus; R and L PT Schizophrenic 1069 for brain volumecorrected and for demographic brain variables volume differences: corrected volume 72 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79

3.3. Correlational analysis: ROI measurements vs. 3.3.4. Ventricular–brain ratio (VBR) clinical symptoms (Table 3) and vs. demographic VBR did not show significant correlations with variables psychopathological scores. In the total sample (ns 60), age correlated positively with VBR (rs0.4, s ) 3.3.1. Asymmetry coefficient of hemispheres P 0.002 , with controls showing a more pro- ( s s ) (ACHem) nounced effect r 0.6, P 0.015 than schizo- ( s s ) By partial correlation controlling for demo- phrenic patients r 0.25, P 0.14 . graphic variables, increased asymmetry coefficient ofhemispheres (RHem)LHem) showed signifi- 4. Discussion cant positive correlations with symptom scores of PANSS-Negative (blunted affect, poor contact, 4.1. Asymmetry coefficient of hemispheres difficulty with abstract thinking and poor sponta- neous speech) and NSRS (disorientation, deficit Schizophrenic patients showed increased right in motion and speech content)(Spearman corre- asymmetry ofcerebral hemispheres when com- lation coefficients are reported in Table 3). pared with healthy controls. Such findings proved to be more robust in systematic schizophrenias, ( ) while non-systematic forms and DSM-IV sub- 3.3.2. Hippocampus H groups had asymmetry coefficients statistically Reduction of left hippocampus correlated signif- similar to normal controls. Furthermore, rightward icantly with symptom severity (total score of ( - ) asymmetry ofhemisphere volumes LHem PANSS and PANSS positive . Within the positive RHem) correlated with negative symptoms, sug- symptoms, severity ofdelusions correlated with gesting that the latter might be mainly related to reduction in LH, and hallucinatory behavior cor- left hemisphere abnormalities. Such findings are related with bilateral reduction in the hippocampus. in agreement with several studies involving neu- With reference to negative symptoms, partial cor- ropsychological testing, brain electrical activity, relation showed an association between total score positron emission tomography and biochemistry, ofPANSS-negative and reduced LH, but control- indicating left hemisphere abnormalities in schizo- ling for demographic variables showed only an phrenic patients (for a review, see Gruzelier, 1999; association between reduced left hippocampus and Petty, 1999). the subitem social withdrawal ofPANSS-negative. ( s ) In both groups together n 60 we found sig- 4.2. Hippocampus nificant positive correlations between years of education and RH (Spearman rs0.33, Ps0.013) ( s s ) Our findings involving hippocampus and symp- and LH r 0.29, P 0.032 . Within groups, such tom severity seems to constitute an additional clue correlation proved to be relatively more robust in ( s s s to left hemisphere abnormalities. Reduced left healthy controls RH: r 0.37, P 0.16; LH: r hippocampus showed a significant correlation with 0.38, Ps0.14) than in schizophrenic patients (RH: ( ) s s s s ) symptom severity total scores ofPANSS , mainly r 0.25, P 0.15; LH: r 0.13, P 0.45 . due to positive symptoms (delusions and halluci- natory behavior). In the literature, there are reports 3.3.3. Planum temporale (PT) ofbilateral reduction in hippocampus related with Partial correlations controlling for demographic the severity ofpositive symptoms (Bogerts et al., variables showed significant positive correlations 1993) and the disorganized syndrome (Fukuzako between volume ofleftPT and the subitems et al., 1996). However, Chua et al. (1997) found hallucinatory behavior, social withdrawal and ster- an association between disturbed formal thinking eotyped thinking (PANSS). Deficits in memory and bilateral increase in limbic structures, and and judgmentydecision ofNSRS showed positive other studies found no significant correlation correlations with reversed asymmetry ofPT between psychopathology and volume ofmesi- (LPT-RPT). otemporal structures (Whitworth et al., 1998; Gur P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 73 0.006 0.018 s s P P 0.43; 0.37; sy sy ns ns r r a b a b ACPT ns ns 0.038 0.034 0.020 0.016 0.040 0.036 s s s s s s P P P P P P 0.34; 0.35; 0.38; 0.38; 0.33; 0.33; sq sq sq sq sq sq ns r r r r r r a b a b a b a b LPT RPT nsnsnsns nsns ns ns ns ns ns ns ns nsns ns ns ns ns 0.040 0.046 0.016 0.047 0.036 0.013 0.040 0.047 0.050 0.040 0.029 s s s s s s s s s s s P P P P P P P P P P P 0.34; 0.33; 0.39; 0.32; 0.34; 0.39; 0.33; 0.32; 0.32; 0.33; 0.35; sy sy sy sy sy sy sy sy sy sy sy r r r r r ns r ns r r r r r a b a b a b a b a b a b a b LH nsnsnsns nsns ns ns ns ns ns ns ns ns ns ns ns ns 0.041 0.016 0.015 0.038 0.029 0.036 0.049 0.017 0.016 s s s s s s s s s P P P P P P P P P 0.34; 0.39; 0.38; 0.33; 0.35; 0.34; 0.31; 0.38; 0.38; sy sy sy sy sy sy sy sy sy ns ns ns r ns r r r r r r r r a b a b a b a b a b a b a b RH ns nsnsns nsns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 0.024 0.023 0.050 0.022 0.040 0.035 0.039 0.035 0.009 s s s s s s s s s P P P P P P P P P 0.38; 0.38; 0.32; 0.38; 0.34; 0.34; 0.33; 0.33; 0.43; sq sq sq sy sq sy sq sq sq r r r r r r r ns r r a b a b a b a b a b a b ACHem ns ns ns ns ) ) ) ) total score total score total score total score thinking ( ( ( ( ambitendency control ns disorganization ns thinking speech behavior scales Table 3 Correlation between morphometric parameters and symptom scores assessedRating by scales PANSS and NSRS PANSS total PANSS positive 1. Delusions2. Conceptual3. Hallucinatory ns6. Suspiciousness ns ns PANSS general ns6. Depression ns 13. Volitive ns 14. Poor impulse16. ns Social withdrawal ns ns ns Negative symptom PANSS negative 1. Blunted ns affec 3. Poor contact 4. Social withdrawal5. ns Difficulty in abstract ns 6. ns Poor spontaneous7. Stereotyped ns ns ns ns ns ns ns 74 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 s 0.046 0.049 0.020 s s s P P P 0.33; 0.33; 0.37; sq sq sq r r r a b a b ACPT right and left planum temporale; ACPT s LPT ns partial correlation controlling for brain volume and 0.005 ) s b P ( ; ) 0.45; sq r a b RPT nsns ns ns ns ns Spearman rho ( 0.050 0.034 s s P P 0.0002. right and left hippocampus; R and L PT 0.31; 0.34; F s P sy sy ns ns r r a b a b LH bivariate correlation ) a ( RH nsns nsns nsns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 0.045 0.049 0.024 0.048 0.007 s s s s s P P P P P index and symptom score by: y 0.34; 0.33; 0.38; 0.33; 0.42; asymmetry coefficient of hemispheres; R and LH sq sq sq sq sq r r r r ns r s a b a b a b a b ACHem ) ACHem decision ns ns y (Continued) total score ( Correlations between volume Abbreviations: demographic variables. If Bonferroni-corrected, the significance should be at asymmetry coefficient of PT. Table 3 Rating scales NSRS 1. Speech content2. Judgment ns 3. Memory5. Orientation 7. Motivation ns8. Motion ns ns ns ns ns ns P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 75 et al., 2000). Our results are partially in agreement reversal ofthe physiological asymmetry in schiz- with those reported by Bogerts et al. (1993), albeit ophrenia, mainly due to a volumetric reduction of restricted to the left hippocampus and at a lower the LPT (Kwon et al., 1999; Hirayasu et al., level ofcorrelation. In fact,the analysis ofthese 2000). On the other hand, the studies ofBarta et results must be interpreted with caution since the al. (1997) and Frangou et al. (1997) reported correlation coefficients are relatively modest negative findings. With reference to psychopathol- (Spearman rho approx. y0.34 and P-0.05) and ogy, Barta et al. (1997) found a significant corre- the elevated number ofcorrelations increases the lation between reduction ofthe LPT and severity risk offalse–positive results. ofthought disorder, and Kwon et al. (1999) However, although bilateral volumetric reduc- reported an association between reduced LPT and tion ofthe hippocampus constitutes a well-docu- the item suspiciousnessypersecution ofthe PANSS. mented finding in schizophrenia (Nelson et al., However, Hirayasu et al. (2000) found no signif- 1998), we found no significant differences between icant correlations between volume ofthe PT and schizophrenic patients (or subgroups according to symptom scores ofthe BPRS. Our results were Leonhard and DSM-IV) and healthy controls. somewhat inconsistent with the literature, given On the other hand, we found an intriguing that hallucinatory behavior and social withdrawal correlation between bilateral volume ofthe hip- (PANSS) showed correlations with increased LPT pocampus and years ofeducation: subjects with volumes. On the other hand, reversed asymmetry higher degrees ofschooling showed bilaterally ofthe PT correlated with deficitsin memory and larger hippocampi. Indeed, some studies have judgementydecision (NSRS). However, on the already reported positive correlations between grounds ofmultiple correlations and increased risk years ofeducation and volume offrontal structures oftype-I error, it seems more prudent to avoid (DeMyer et al., 1988; Andreasen et al., 1990), and inferences other than the inconsistency between Gur et al. (2000) found positive correlations morphometry ofPT and psychopathology in our between volume ofhippocampus and performance study. in memory in both schizophrenic and control groups. Bogerts et al. (1993) suggested that hip- 4.4. Ventricular–brain ratio (VBR) pocampal reduction may be related to lesser schooling as an effect of disease and, therefore, In the present study schizophrenic patients matching subjects for years of education could showed larger VBR compared to healthy controls, introduce a bias in terms ofselecting patients with a finding largely reported in the literature (Elkis less severe syndromes and controls with poorer et al., 1995). Only in the control group was age cognitive performance. Our results argue against correlated with enlargement ofthe VBR. Patients this hypothesis given that positive correlations showed no significant correlation between VBR between hippocampal volumes and schooling were and age or duration ofillness, suggesting that stronger in the control group. Therefore, more than ventricular enlargement in schizophrenic patients a mere pathophysiological effect of disease, the does not progress, which is not in line with our reported hippocampal reduction in schizophrenia earlier CT findings (Gattaz et al., 1981). could represent, at least in part, an effect of Enlargement of VBR was non-specific and dif- different schooling between patients and healthy fuse among the diagnostic subtypes. However, controls. controlling for demographic factors, only the par- anoid subtype ofDSM-IV (Ps0.037) showed 4.3. Planum temporale (PT) significant increased VBR. With reference to symptomatology, however, no symptom ofthe PT morphometry showed no significant differ- paranoid dimension seemed to be related to ences between groups, either within DSM-IV or increased VBR. Such results are more supportive within Leonhard’s subgroups. Two offourvolu- ofthe hypothesis that ventricular enlargement is metric studies involving the PT found a significant present as a continuum in the schizophrenic psy- 76 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 choses (Marsh et al., 1999). Our results also do ofstructural abnormalities with respect to the not support the early hypothesis that prevalence of hippocampus, planum temporale and VBR. negative symptoms is related to ventricular Our study had some limitations. Not all the enlargement (Crow, 1985; Andreasen et al., 1990; subtypes ofschizophrenia were represented in the Gur et al., 1994). study (e.g. the catatonic subtype), and the relative- ly small number ofsubjects prevented a more 4.5. Differences between subgroups according to specific analysis of DSM-IV subtypes and Leon- Leonhard and DSM-IV hard subforms. On the other hand, the symptom- based subtypes ofDSM have a very modest stability along the course ofillness. Patients ini- Leonhard’s classification allowed a better dis- tially diagnosed as having one subtype may display tinction than DSM-IV with reference to the inter- a combination ofsymptoms that changes over time hemispheric asymmetry coefficient. The systematic to another subtype (Goldberg and Weinberger, forms of Leonhard brought together the cases with 1995), thus challenging the stability ofthe diag- relatively smaller left than right hemispheres, a nosis. In this sense, by means ofthe observation finding that was associated with increased severity ofsymptom associations along the longitudinal ofnegative symptoms on the PANSS and the course, Leonhard’s classification seems to provide NSRS. Given that neither doses nor categories of more stability on diagnosis over time. antipsychotics were significantly different between SS and NSS patients, such interhemispheric differ- ences seem to be independent ofmedication. This 5. Conclusion finding is also in agreement with the Leonhardian observations that systematic schizophrenias show Schizophrenic patients showed rightward asym- an early onset ofillness and a poorer prognosis metry ofthe cerebral hemispheres and increased compared with the non-systematic forms (Leon- VBR in comparison with the control group. Right- hard, 1995). Furthermore, in line with these prop- ward asymmetry ofhemispheres correlated with ositions, several studies ofbrain electrical activity severity ofnegative symptoms and was more have shown that the dislocation ofthe P300 field pronounced in the systematic forms of Leonhard, map to the right is associated with a more homo- suggesting a distinct pattern oflefthemisphere geneous subgroup ofnuclear psychoses. Thus, abnormality in this subgroup ofpsychoses. dysfunction of left-hemispheric neuronal genera- Schizophrenic patients showed the left hippo- tors has been related to the absence ofacute campus to be significantly smaller than the right episodes, more severe residual symptoms and poor hippocampus, and the volume ofthe lefthippo- outcome (Heidrich and Strik, 1997; Strik et al., campus correlated inversely with severity ofsymp- 1993, 1996; Strik and Fallgatter, 1998). Although toms, which provides an additional clue for left we have not investigated obstetric complications hemisphere abnormalities in schizophrenia. in our study, a possible explanation for the right- Although more pronounced in the paranoid sub- ward cerebral asymmetry observed in systematic type, increased VBR seems to follow a single schizophrenias could be related to the different normal distribution in the subgroups, indicating maturational timing between the right and left that, at least in terms ofthis parameter, ventricular hemispheres, which renders the left hemisphere enlargement is not restricted to a subgroup but is more vulnerable to potential insults in the second present to various degrees in all cases. trimester ofpregnancy (Bracha, 1991; Petty, With reference to the initial hypothesis that 1999). Leonhard’s classification could provide a reliable On the other hand, with the exception ofthe instrument to distinguish groups that are biologi- asymmetry coefficient of hemispheres, neither the cally more homogeneous, our finding of rightward subtypes ofDSM-IV nor the formsofLeonhard asymmetry in the systematic forms seems promis- seemed to be able to discriminate different patterns ing and deserves further investigation. P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 77

Acknowledgments DeLisi, L.E., Hoff, A.L., Neale, C., Kushner, M., 1994. Asymmetries in the superior temporal lobe in male and female first-episode schizophrenic patients: measures of the The authors thank Prof. Helmut Beckmann and planum temporale and superior temporal gyrus by MRI. his team at the University ofWurzburg,¨ Germany Schizophrenia Research 12, 19–28. (Dr Pfuhlmann, Dr Jabs and Dr Bartsch) for DeLisi, L.E., Sakuma, M., Tew, W., Kushner, M., Hoff, A.L., helpful advice and orientations in Leonhard’s clas- Grimson, R., 1997. Schizophrenia as a chronic active brain sification and Prof. Julio Cesar Rodrigues Pereira process: a study ofprogressive brain structural change for help in statistical analysis. This work was subsequent to the onset ofschizophrenia. Psychiatry Research: Neuroimaging 74, 129–140. ˜ supported by a grant from the ‘Fundacao¸ de DeMyer, M.K., Gilmor, R.L., Hendrie, H.C., DeMyer, W.E., Amparo a`˜ Pesquisa do Estado de Sao Paulo’ Augustyn, G.T., Jackson, R.K., 1988. Magnetic resonance (FAPESP-Brazil; project 98y10487-2) to Dr Sallet. brain images in schizophrenic and normal subjects: influence ofdiagnosis and education. Schizophrenia Bulletin 14, References 21–32. Elkis, H., Friedman, L., Wise, A., Meltzer, H.Y., 1995. Meta- analyses ofstudies ofventricular enlargement and cortical American Psychiatric Association, 1994. DSM-IV: Diagnostic sulcal prominence in mood disorders: comparison with and Statistical Manual ofMental Disorders, 4th edition. controls or patients with schizophrenia. Archives ofGeneral American Psychiatric Press, Washington, DC. Psychiatry 52, 735–746. American Psychiatric Association, 1997. Practice guideline for First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1996. the treatment ofpatients with schizophrenia. American ‘Entrevista Clınica´ Estruturada para o DSM-IV-Transtornos Journal ofPsychiatry 154 (4), 1–63. do Eixo I. Edicao¸˜ para Pacientes (SCID-IyP–2.0)’. Trans- Andreasen, N.C., Ehrhardt, J.C., Swayze, V.W., Alliger, R.J., lated by Tavares, M., University ofBrasılia,´´ Brasılia. Yuh, W.T.C., Cohen, G., Ziebell, S., 1990. Magnetic reso- nance imaging ofthe brain in schizophrenia: the pathophy- Flaum, M., O’Leary, D.S., Swayze, V.W., Miller, D.D., Arndt, siologic significance of structural abnormalities. Archives S.V., Andreasen, N.C., 1995. Symptom dimensions and brain ofGeneral Psychiatry 47, 35–44. morphology in schizophrenia and related psychotic disor- Andreasen, N.C., 1999. A unitary model ofschizophrenia: ders. Journal ofPsychiatric Research 29 (4), 261–276. Bleuler’s ‘fragmented phrene’ as schizencephaly. American Frangou, S., Sharma, T., Sigmudsson, T., Barta, P., Pearlson, Journal ofPsychiatry 56, 781–787. G., Murray, M., 1997. The Maudsley Family Study 4. Ban, T.A., 1982. Chronic schizophrenias: a guide to Leonhard’s Normal planum temporale asymmetry in familial schizo- classification. Comprehensive Psychiatry 23 (2), 155–169. phrenia. A volumetric MRI study. British Journal ofPsy- Barta, P.E., Pearlson, G.D., Brill II, L.B., Royall, R., Mc- chiatry 170, 328–333. Gilchrist, I.K., Pulver, A.E., Powers, R.E., Casanova, M.F., Franzek, E., Beckmann, H., 1996. General environment inter- Tien, A.Y., Frangou, S., Petty, R.G., 1997. Planum temporale action in schizophrenia: season-of-birth effect reveals etio- asymmetry reversal in schizophrenia: replication and rela- logically different subgroups. Psychopathology 29, 14–16. tionship to gray matter abnormalities. American Journal of Franzek, E., Beckmann, H., 1998. Different genetic back- Psychiatry 154 (5), 661–667. ground ofschizophrenia spectrum psychoses: a twin study. Bartko, J.J., Carpenter, W.T., 1976. On the methods and theory American Journal ofPsychiatry 155 (1), 76–83. ofreliability. Journal ofNervous and Mental Disease 163 Fukuzako, H., Fukuzako, T., Hashiguchi, T., Hokazono, Y., (5), 307–317. Takeuchi, K., Hirakawa, K., Ueyama, K., Takigawa, M., Bogerts, B., Lieberman, J.A., Ashtari, M., Bilder, R.M., Kajiya, Y., Nakago, M., Fujimoto, T., 1996. Reduction in Degreef, G., Lerner, G., Johns, C., Masia, S., 1993. Hippo- hippocampal formation volume is caused mainly by its campus-amygdala volumes and psychopathology in chronic shortening in chronic schizophrenia: assessment by MRI. schizophrenia. Biological Psychiatry 33, 236–246. Biological Psychiatry 39, 938–945. Bracha, H.S., 1991. Etiology ofstructural asymmetry in Galaburda, A.M., Corsiglia, J., Rosen, G.D., Sherman, G.F., schizophrenia: an alternative hypothesis. Schizophrenia Bul- 1987. Planum temporale asymmetry, reappraisal since letin 17 (4), 551–553. Geschwind and Levitsky. Neuropsychologia 25 (6), Briggs, G.G., Nebes, R.D., 1975. Patterns ofhand preference 853–868. in a student population. Cortex 11 (3), 230–238. Gattaz, W.F., Kasper, S., Kohlmeyer, K., Beckmann, H., 1981. Chua, S.E., Wright, I.C., Poline, J.-B., Liddle, P.F., Murray, Die kraniale computertomographie in der schizophreniefor- R.M., Frackowiack, R.S.J., Friston, K.J., McGuire, P.K., schung. Fortschritte der Neurologie und Psychiatrie 49, 1997. Grey matter correlates ofsyndromes in schizophrenia. 286–291. British Journal ofPsychiatry 170, 406–410. Geschwind, N., Levitsky, W., 1968. Human brain: left–right Crow, T.J., 1985. The two-syndrome concept: origins and asymmetries in temporal speech region. Science 161, current status. Schizophrenia Bulletin 11, 471–486. 186–187. 78 P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79

Goldberg, T.E., Weinberger, D.R., 1995. A case against sub- Left planum temporale reduction in schizophrenia. Archives typing in schizophrenia. Schizophrenia Research 17, ofGeneral Psychiatry 56, 142–148. 147–152. Leonhard, K., 1970. Biopsychologie der Endogenen Psycho- Gruzelier, J.H., 1999. Functional neuropsychophysiological sen. Hirzel, Leipzig. asymmetry in schizophrenia: a review and reorientation. Leonhard, K., 1990. Differenzierte diagnostik der endogenen Schizophrenia Bulletin 25 (1), 91–120. psychosen unter anlehnung an einem symptomenkatalog. Gur, R.E., Mozley, D., Shtasel, D.L., Cannon, T.D., Gallacher, Psychiatrie und Neurologie Medical Psychologie 42, F., Turetsky, B., Grossman, R., Gur, R.C., 1994. Clinical 136–145. subtypes of schizophrenia: differences in brain and CSF Leonhard, K., 1995. Aufteilung der Endogenen Psychosen und volume. American Journal ofPsychiatry 151, 343–350. ihre Differenzierte Atiologie.¨ 7th edition. Thieme, Stuttgart, Gur, R.E., Cowell, P., Turetsky, B.I., Gallacher, F., Cannon, T., pp. 1995. Bilker, W., Gur, R.C., 1998. A follow-up magnetic resonance Marsh, L., Lim, K.O., Hoff, A.L., Harris, D., Beal, M., Minn, imaging study ofschizophrenia. Archives ofGeneral Psy- K., Faustman, W.O., Csernansky, J.G., Sullivan, E.V., Pfef- chiatry 55, 145–152. ferbaum, A., 1999. Severity of schizophrenia and magnetic Gur, R.E., Turetsky, B.I., Cowell, P.E., Finkelman, C., Maany, resonance imaging abnormalities: a comparison ofstate and V., Grossman, R.I., Arnold, S.E., Bilker, W.B., Gur, R.C., Veterans hospital patients. Biological Psychiatry 45, 49–61. 2000. Temporolimbic volume reductions in schizophrenia. Naidich, T.P., Daniels, D.L., Haughton, V.M., Williams, A., Archives ofGeneral Psychiatry 57, 769–775. Pojunas, K., Palacios, E., 1987. Hippocampal formation and Harrison, P.J., 1997. Schizophrenia and its dementia. In: Esiri, related structures oflimbic lobe: anatomic-RM correlation. M.M., Morris, J.H. (Eds.), The Neuropathology ofDemen- Part I. Surface features and coronal sections. Part II. Sagittal tia. Cambridge University Press, Cambridge, pp. 385–397. sections. Radiology 162, 747–761. Harrison, P.J., 1999. The neuropathology ofschizophrenia. A Nelson, M.D., Saykin, A.J., Flashman, L.A., Riordan, H.J., critical review ofthe data and their interpretation. Brain 1998. Hippocampal volume reduction in schizophrenia as 122, 593–624. assessed by magnetic resonance imaging: a meta-analytic Heidrich, A., Strik, W.K., 1997. Auditory P300 topography study. Archives ofGeneral Psychiatry 55, 433–440. and neuropsychological test performance: evidence for left O’Leary, D.S., Kesler, M., Mosnik, D.M., Tranel, A.P., Hurtig, hemispheric dysfunction in schizophrenia. Biological Psy- R., Hichwa, R., Andreasen, N.C., 1995. Regional cerebral ( ) chiatry 41, 327–335. blood flow rCBF within Heschl’s gyrus and the planum temporale in schizophrenic patients and controls during Hirayasu, Y., McCarley, R.W., Salisbury, D.F., Tanaka, S., language tasks: an MRI and PET study (abstract). Schizo- Kwon, J.S., Frumin, M., Snyderman, D., Yurgelun-Todd, D., phrenia Research 2, 94. Kikinis, R., Jolesz, F.A., Shenton, M.E., 2000. Planum Petty, R.G., Barta, P.E., Pearlson, G.D., McGilchrist, I.K., temporale and Heschl gyrus volume reduction in schizo- Lewis, R.W., Tien, A.Y., Pulver, A., Vaughn, D.D., Casa- phrenia. A magnetic resonance imaging study offirst- nova, M.F., Powers, R.E., 1995. Reversal ofasymmetry of episode patients. Archives ofGeneral Psychiatry 57, the planum temporale in schizophrenia. American Journal 692–699. ofPsychiatry 152, 715–721. Iager, A.-C., Kirch, D.G., Wyatt, R.J., 1985. A Negative Petty, R.G., 1999. Structural asymmetries ofthe human brain Symptom Rating Scale. Psychiatry Research 16, 27–36. and their disturbance in schizophrenia. Schizophrenia Bul- Jaskiw, G.E., Juliano, D.M., Goldberg, T.E., Hertzman, M., letin 25 (1), 121–139. Urow-Hamell, E., Weinberger, D.R., 1994. Cerebral ventric- Rossi, A., Serio, A., Stratta, P., Petruzzi, C., Schiazza, G., ular enlargement in schizophreniform disorder does not Mancini, F., Casachia, M., 1994. Planum temporale asym- progress. A seven-year follow-up study. Schizophrenia metry and thought disorder in schizophrenia. Schizophrenia Research 14, 23–28. Research 12, 1–7. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The Positive and Sallet, P.C., Elkis, H., Alves, T.M., Oliveira, J.R., Sassi, E., ( ) Negative Syndrome Scale PANSS for schizophrenia. Castro, C.C, Busatto, G.F., Gattaz, W.F. Reduced cortical Schizophrenia Bulletin 13, 261. folding in schizophrenia: an MRI morphometric study. Kleinschmidt, A., Falkai, P., Huang, Y., Schneider, T., Furst,¨ Submitted for publication. G., Steinmetz, H., 1994. In vivo morphometry ofplanum Serfling, R., Lossner,¨ A., Schreier, J., 1995. MRI studies into temporale asymmetry in first-episode schizophrenia. Schiz- morphological cerebral deficits in patients diagnosed schizo- ophrenia Research 2, 9–18. phrenic according to Leonhard. In: Beckmann, H., Neumar- Kleist, K., 1934. Gehirnpatologie. Thieme, Leipzig. ker, K.-H. (Eds.), Endogenous Psychoses: Leonhard’s Kulynych, J.J., Vladar, K., Bantie, B.D., Jones, W.D., Wein- Impact on Modern Psychiatry. Ulstein Mosby, Berlin, Wies- berger, D.R., 1995. Normal asymmetry ofthe planum baden, pp. 221–229. temporale in patients with schizophrenia. British Journal of Shapleske, J., Rossell, S.L., Woodruff, P.W.R., David, A.S., Psychiatry 166, 742–749. 1999. The planum temporale: a systematic, quantitative Kwon, J.S., McCarley, R.W., Hirayasu, Y., Anderson, J.E., review of its structural, functional and clinical significance. Fischeri, I.A., Kikinis, R., Jolesz, F., Shenton, M.E., 1999. Brain Research Reviews 29, 26–49. P.C. Sallet et al. / Psychiatry Research: Neuroimaging 123 (2003) 65–79 79

Shenton, M.E., Dickey, C.C., Frumin, M., McCarley, R.W., Tsuang, M.T., Faraone, S.V., 1995. The case for heterogeneity 2001. A review ofMRI findingsin schizophrenia. Schizo- in the etiology ofschizophrenia. Schizophrenia Research phrenia Research 49, 1–52. 17, 161–175. Stober,¨ G., Franzek, H., Beckmann, H., 1994. Schwanger- Vita, A., Dieci, M., Giobbio, G.M., Tenconi, F., Invernizzi, G., schaftsinfektionen bei muttern¨ von chronisch schizophrenen. 1997. Time course ofcerebral ventricular enlargement in Die bedeutung einer differenzierten nosologie. Nervenarzt schizophrenia supports the hypothesis ofits neurodevelop- 65, 175–182. mental nature. Schizophrenia Research 23, 25–30. Stober,¨ G., Kocher, I., Franzek, E., Beckmann, H., 1997. First- Von Economo, C., Horn, L., 1930. ber Windungsrelief, Masse und Rindenarchitetektonik der supratemporalflache,¨ ihre trimester maternal gestational infections and cycloid psycho- individuellen und ihre seitenunterschiede. Zeitschrift fur¨ sis. Acta Psychiatrica Scandinavica 95, 319–324. Neurologie und Psychiatrie 130, 678–757. Strik, W.K., Dierks, T., Franzek, E., Stober,¨ G., Maurer, K., Wernicke, C., 1900. Grundriss der Psychiatrie in Klinischen 1993. P300 asymmetries in schizophrenia revisited with Vorlesungen. Thieme, Leipzig. reference-independent methods. Psychiatry Research: Neu- Whitworth, A.B., Honeder, M., Kremser, C., Kemmler, G., roimaging 55, 153–163. Felber, S., Hausmann, A., Wanko, C., Wechdorn, H., Aich- Strik, W.K., Dierks, T., Kulke, H., Maurer, K., Fallgatter, A.J., ner, F., Stuppaeck, C.H., Fleischhacker, W.W., 1998. Hip- 1996. The predictive value ofP300-amplitudes in the course pocampal volume reduction in male schizophrenic patients. ofschizophrenic disorders. Journal ofNeural Transmission Schizophrenia Research 31, 73–81. 103, 1351–1359. Zipursky, R.B., Lambe, E.K., Kapur, S., Mikulis, D.J., 1998. Strik, W.K., Fallgatter, A.J., 1998. Neurophysiologische befun- Cerebral gray matter volume deficits in first episode psy- de bei endogenen psychosen. Psycho 24, 365–372. chosis. Archives ofGeneral Psychiatry 55, 540–546.