SF.MINARS IN NEUROLOGY-VOLUME 9. NO. 4 DU:£MBER 1989

Positron Emission Tomography in Psychiatric andNeuropsychiatric Disorders

R. I Dolan, M.D* and K.I Friston, MA, (Canlab)t

Positron emission tomography (PET) has a po-tentially unique position in the study of psychiatricand neuropsychiatric disorders. The mechanismsby which regional pathologic changes disrupt nor-mal psychologic functioning and the relationshipsbetween mental events and brain physiology arcnow open 1.0 empirical study. Although progresshas been made in revealing the functional anatomyof human cognition, PET has not, as yet, clarifiedthe parhologic processes underlying the major PS)'·ch iatric disorders. This failure may mean that theprocesses mediating psychiatric disorders arc be-yond 1 he resolution of PET or, alternatively, thatthe design of PET studies of psychiatric disordersis not sufficiently refined.

The primary focus of PET studies up to thepresent has been the investigation of neurologicdisorders. In the application of PET to psychiatricresearch it is important to bear in mind essentialdifferences between psychiatric and neurologicdisorder-s. Psychiatric illness involves the disorgan-ization of internal experience as opposed to thebreakdown of primary sensorimotor function. Ar-guably PET methodologic strategies that haveproved fruitful in the study of neurologic disor-ders. where there are fixed pathologic features,may be inappropriate to the study of psychiatricdisorders. In psychiatric disorders, with no fixedpathologic features, abnormalities are likely to be"functional," involving disturbed neural integra-tion. Therefore novel st.rat.egies or augmentationof existillg methods will be necessary in the inves-tigation of these disorders.

The choice of appropriate PET methods for

data acquisition and analysis ill studies ofatric disorders has still to be established.normal responsivuy of psychiatric patientstochologic and physical stressors makesconditions critical. It can be argued thatstate metabolic scans should be augmenstudies during the performance of cogniti\tor during specific phannacologic manipMost PET studies reponed to dale havequired under resting conditions and thismatially explain the lack of consistent finding>absence of standardized approaches to PET'analysis in circumstances where little isknocerning regional pathologic characteristics ~thcr drawback and has led to an excess reliratio data. This type of approach can oftenleading because such relative values havshown to be highly sensitive to ambient coThe seemingly arbitrary region of interesttion reponed in the literature emphasizesfor standardized methods of region of'definition or the exploration of alternanpreaches.

As well as issues of study design, eqmust be applied to patient selection andization. Homogeneity of the patient groupdarnental in psychiatric studies. Lack ofneity can render relationships betweengroups and PET measures extremely elushomogeneity is difficult to achieve usingsectional assessments, which are standardch iatric research. Such methods need lobemented by either longitudinal assessmentsuse of genetic or other such biologic mark

=Nanonal Hospit a 1 and Academic Dcparuneru of Psychiau-v. Royal Free Hospital Medical School.London. United King-domtMRC Cyclotron Unit. Hammersmith Hospital. and Department of Psychiatry. Cbariug Cross andWestminster Medical Schools. London, United Kingdom.

Reprint requests: Dr. Dolan. AcadUl1ic Depart.mcut of Psychiatry. Royal Free Hospital MedicalSchool, London. U K .

Copyright © 1989 hy Thieme Medical Publishers. Inc .. 381 Park Avellue South, New York, NY10016. All rights reserved.

YCHIATRY-DOLAN, FRISTON

gandremitting nature of many psychiatricemphasizes the need for appropriatestudiesin relation to clinical state.

, articleexamines finLlings from PET stud-tientswith psychiau ic and neuropsychiai-ers and highlights areas of current andterest, Most studies have concerned theychiatricdisorders, schizophrenia and af-

rders, and consequently the greaterthisarticle will address findings in these

SCHIZOPHRENIA

phrenia is the most prevalent and dis-}'choticdisorder and is characterized byageof onset, symplOms such as halluci-delusions, disordered thinking (positives),and impoverished motivation and feel-tivesymptoms). A genetic predisposition

. hed,although beyond this there is muchyregarding euology, which may be mul-

Schizophrenia is defined phenomcnologi-d maysubsume a number of clinical syn-

studiesof schizophrenia dale from 1980according lO the group studied and theused. The most e-xtensively used PEThave been measurements of cerebral

W, metabolism, and reccptor binding inbycomparison with controls. Patients have. led according la standardized diagnos-

and this approach assumes regional'c features common to clinically definedesof schizophrenia. No such relationshipionalpathologic changes has been found.e early studies of regional cerebra I bloodBF) in schizophrenia using xenon inhala-niques suggested a relative decrease inregions1-' Despite rnethodologic prob-

riIh these pioneering studies, associationsbetweenabnormal patterns of' per-fusion• IC symptoms. ln general, patients witht frontal rCBF tended la be the most,mute, and indifferent. This pattern of

talilYtherefore is a finding antedatingremainsimportant and controversial.frontality, a relative reduction in pre-

corticalblood now or metabolism, has beenportedusing PET.' -7 Despite this, thereuingcontroversy regarding its reproduci-

ificity,sensitivity to the confounding ef-rugs,and dependence on cognitive stale

s.Three studies lIsing deoxyglucose aslic-tracer,which yielded contradictory re-beconsidered.

DeLisi et al" studied patients with chronicschizophrenia and matched controls. The patientsmet strict DSM-1I criteria tor schizophrenia andwere medication free for at least 2 weeks prior tothe study. Patients had significanLly lower anteriorto posterior ratios (eight of the nine patients hadratios less than 1). Cerebral atrophy as determinedby CAT scan was not associated with this aberrantmetabolic pattern. Cur et apuo in two communica-tions described abnormalities of the subcortico-conical metabolic gradient in schizophrenic pa-tients who were medication free for at least a week.The duration of illness varied from 3 lO 17 years.No evidence for hypofrontality was found. Finally,Szechtman et al!' in a controlled study examinedwhether neuroleptic treatment duration influencedthe regional distribution of metabolism in patientsmeeting Research Diagnostic Criteria (RDC) forschizophrenia. The patient group was dichoto-mized according to treatment duration. Bothgroups had a greater anterior to posr.erior ratiothan control subjects, although this was less evidentin the group with the longest neuroleptic expo-sure.

As in much other work on schizophrenia, thebasic findings with PET are often not reproduced.Explanations for these apparently inconsistent re-sults include the confounding effects of treatmentand illness duration as well as differences in scan-ning procedure and in image analysis. An impor-tant consideration is the relative preponderance ofpositive and negative symptoms in the groups stud-ied, which may reflect differing etiologies andpathologics. In this context Liddle" has describedthree subsyndromes of schizophrenia which in-clude a syndrome of "psychomotor poverty" char-acterized by negative symptoms (Ilat affect, pov-erty of speech and spontaneous movement), Basedon a comparison of signs and symptoms in focalbrain lesions, it is suggested that this syndrome isassociated with impaired dorsolateral prefrontalconical (DLPFC) function. The prediction that hy-pofrontality is associated with negative symptoms,within any patient group, has received indepen-dent support From PET studies, Dclisi et al13 re-poned that the only significant correlations be-tween relative hypofrontality and symptom ratingswere for emotional withdrawal, disorientation,distracubility, and helplessness or hopelessness,Kishomoto et al'" discriminated among three dis-tinct types of metabolic pattern in chronic schizo-phrenic patients. Hypofrontal patients tended toshow flat, blunted affect and a hypoparietal group,delusions and hallucinations. Because diagnosticcriteria place an emphasis on delusions and hallu-cinations (positive symptoms). psychomotor pov-erty syndromes (negative symptoms) are likely lO

SE~IINARS IN NEUROLOGY VOLUME 9. NUMBER 4 DECEMBER I

be undcrselcctcd in some study proiocols andtherefore hvpofrontality will be a variable findingacross SI ud ics.

The possible relationships between negativesymptoms, hypofrontality, and putative abnormal-ities of the mesocortical dopaminergic system havebeen addressed from a number of perspectives.Animal studies using autoradiography have shownincreased frontal and anterior cingulate metabolicresponse 10 the doparninergic agonist apomor-phine.!" Corresponding dopaminergic challenge inhumans has yet 10 be established, although initialstudies have been reponed. woikin et al'" reportdecreased frontal. temporal, and striatal glucosemetabolism in schizophrenic and control subjectsfollowing d-amphetamine (0.5 mg/kg orally). Cer-aud et al'" report a reversible hemodynamic hypo-frontality in yonng schizophrenic patients. Hypo-froruality was seen in chronic patients whosedisease had evolved over more than 2 years. andthis paLtcrn disappeared during exacerbation ofthe symptoms. In a subgroup who had not beentreated for seve-ral weeks a weak dose of a dopa-minergic agonist restored near-normal Frorualiry.The investigators conclude, "This [dopamine hy-persensitivity] may reflect either the role of neu-roleptic washout or a prirnir ive dopaminergic de-piction as proposed by some authors in the chronicform of schizophrenia."!" This is consistent with astudy of patients. characterized by short durationof illness, before and after medication.!" In a med-ication-free state, patients had asymmetric basalganglia uptake. left greater than right. Followingdopamine recepior blockade consequent upontrcauuem. this asymmetry was abolished. btu a leftprefrontal cortical reduction in glucose uptake be-came evident.

Basal ganglia changes in dopamine pharrna-cology have been reponed from postmortem stud-ies of schizophrenia patients, although the Findingsarc confounded by possible drug effects. Increasein D2 receptoi s in unmedicated schizophrenic pa-tients has partial support from the PET literature.\Vong et at,"' lIsillg displacement of lie N-rnethyl-spiperonehy nnlabelcd haloperidol, studied 02 re-cepior densny in normal, drug-naive, and treat-cd schizoplu euic subjects. A three-compartmentmodel with irreversible radioligand binding was as-sumed. Significant increases in receptors were re-poned for both schizophrenic groups, with the in-crease being more evident in drug-treated patients.There is continuing debate over the reproducibil-ity of these findings20 and criticisms of met hods ordata analysis have been expressed."

Making I'ET a behaviorally or pharmaoologi-cally specific technique is the object of activationparadigms. Bchavioral specificity can be achieved

by using PET in conjunction with cognitive orsorimotor activation. Regional deficits measuwith PET can be seen as a common concomitantclinical and neuropsychologic symptoms. Thsproach depends on first establishing the fundianatomy of relevant conical and subcorticalgions in normal subjects. Weinbcrger et al22

reported findings that imply physiologic dys!tion of the prefrontal cortex in response to acific cognitive challenge, the Wisconsin CardUsing the xenon-133 inhalation technique.reponed that at rest patients had a relativenot absolute reduction in dorsolateral prefrnow. During the cognitive activation, the pa'showed evidence of impaired augmentation ofregional now. These results have yet to be indedently reproduced, and the likelihood thatparticular complex task involving a visual.abstand motor component is regionally specificsuoversial.

Other, more simple, activation study dhave been reponed. Cohen et al23 studied cefunction during an auditory discriminationdesigned to emphasize sustained attention. Arcct relationship was found between metabolicin the prefrontal cortex and accuracy of pemance. In schizophrenics lower flow was founthe prefrontal cortex, which was unrelated toperformance. warkcntin et al,~1 again usinganon inhalation measurement of reBF. usedabal fluency task as a cognitive challenge. Themarked effect of the activation in normal suwas seen in Ihe left prefrontal area, but inschizophrenic group this increase was auenuThe investigators conclude, "The controvegarding frontal lobe dysfunction in schizophis related 10 whether these areas are fundichallenged or not." Using 11C-deoxyglucosePET tracer, Volkow et al25 studied pauentschronic schizophrenia using a smooth pursuitracking activation task. Both at baseline anding activation, patients displayed absolute analive frontal hypometabolism. Patients withlive symptoms had lower frontal, temporal.parietal metabolic rates compared with thosepositive symptoms across both conditions. Sicant between-condition differences weresolely in the positive group. During taskmance. significant negative correlationsfrontal lobe metabolism and symptom ratinpeared. This study therefore illustrates thethat activation studies may increase not onspecificity, but also the sensitivity of PET.

biology has not been addressed by PETies of schizophrenia. Berman et al2fi haveextheir xenon inhalation technique, in eonjtwith Wisconsin Card Sort, to studies of

twins concordant and discordant forrenia. Their findings suggest that theysiologicdefect in prefrontal function isfor the schizophrenic twin in a discordant

hermore, lifetime neuroleptic medication1 explain differences in hypofrontality in

ncordant pairs. This finding, which needs'on with PET, may suggest that hypofron-notdirectly mediated by the genotype.conclusion, PET studies of schizophreniaa pattern of hypofrontality in some gmupstsdiagnosed as schizophrenic. Hypofron-

hestcorrelated with a clinical psychomotorsyndrome with predominantly negatives.The functional state ar. the time of scan-greatlychange the nature of the findings,is best exemplified by studies using acti-radigms. Much work needs to be done

e and develop activation strategies bothand pharmacologic. Their combined use

he possibility of linking abnormal neuro-· 'on,behavioral impairment, and regional

icchanges in schizophrenia.

AFFECTIVE DISORDERS

etiologyof affective disorders is still un-The most widely accepted viewpoint positsIdisturbances of neural networks in er-it-

· centersconcerned with mood regulation.Ineural transmission in monoaminergic

shas been implicated by the advent of po-. epressant drugs. No specific monoami-stem abnormality in affective disordersbeen identified using indirect methods.

ure has frequently been attributed to theirectmeasures of brain function in vivo.pared with studies of schizophrenia,relativelyfew PET SIudies of affective pa-

d most of those reponed have focused onurememof cerebral blood flow (CRF) and

sm, Early studies used the xenon inhala-ique.These will be discussed as a prelirn-

a consideration of PET studies.iesof CBF in affective disorders, using

on inhalation technique, have suggested· ed decreases in CRr in depressed pa-athewet aJ27 studied 13 young patients

Age·and sex-matched control group. ThemetRDC for depression and had a medi-

hour period of 2 weeks. A generalizedincerebral blood flow was reponed in thecompared with control subjects. An in-tionship was observed between bloodan index of illness severity. The princi-

·Igswere replicated by a number of sepa-

rate investigators, who likewise reported decreasedCB F in depressed corn pared with control sub-jects.28.29 Uytdenhoef et apo reported changes of adifferent kind in a study of unipolar and bipolarpatients, the latter while in remission. Significantleft frontal hyperperfusion and right posterior hy-poperfusion were seen in the depressive comparedwith control subjects. This pattern was not ob-served in the euthymic bipolar patients, suggestingthat alterations in rCBF in depression might bestate dependent. These findings were not repli-cated in a number or other reports. Gustafson etapt compared a heterogeneous group of patients,including dcpressives. with normal controls andfailed to find differences across groups. Otherstudies of depressed and manic subjects32 and de-pressed patients alone" failed to find differencesfrom controls.

A single study of depressed patients, using thexenon technique, has used an activation paradigm.Patients were studied in the resting stale and dur-ing performance of both a verbal and a spatialtask." Although no overall differences were ap-parent between groups during rest, differencesemerged in the rCBF patterns of depressive andcontrol subjects during cognitive activation. Thedepressed female patients had higher than normalflow in all states, whereas the depressed male pa-tients had lower resting now that became normalduring cognitive activation.

The earliest study of affcctivety ill patients us-ing PET compared regional cerebral metabolicrate of glucose (rCMRglu) values in schizophrenic,affective disorder, and control subjects using asomatosensory paradigm (subjects received lIsecelectric shocks to the right forearm) which at-tempted to control for the ambient state across sub-jeers." The affective patients showed a reducedanteroposterior gradient in glucose metabolismcompared with control subjects. The order of mag-nitude of these changes is indicated by an inves-tigation that demonstrated a similarity in themetabolic rates for glucose between patients withdepression and those with multi-infarct demen-tia.:\6 Only a single report to date has failed to findaltered cerebral metabolism in depression." Nodifferences were found in cerebral glucose utiliza-tion in a heterogeneous group of chronic psychi-atric patients, including six chronic depressives,and normal control subjects. The srnall sample sizeand the chronic population studied limit and com-plicate the interpretation of this study.

In contrast to these preliminary reports a de-tailed series of studies of affective patients usingISF deoxyglucose as a metabolic tracer has been re-ported.3x-1u In the first of these, unipolar and bi-polar patients were studied under a variety of con- 333

I

SEMINARS IN NEUROLOGY VOLUME 9, NUMBER 4 DECEMBER

ditions. The unipolar group were scanned in adrug-free baseline state, following administrationof methylphenidate and when euthymic on follow-up. The bipolar patients were studied either in amanic or depressed phase and when euthyrnic onfollow-up. A n interesting observation was that thebipolar depressed patients in the resting state hadlower hemispheric metabolic rates for glucose thaneither control subjects. unipolar depressed or bi-polar manic patients. A subgroup of unipolardepressed patients were identified who hadmetabolic asymmeu-ies. most prominent in the pas-tcroinferior frontal areas. In this lauer group apositive clinical response to methyphcnidate wasassociated with normalization of metabolic aSYIIl-rneuies, whereas an absent clinical response was as-soci.ucd with persistence of asymmetries.

In a subsequent communication on an cn-larged series of patients, including unipolar de-prcssives, bipolar depressives, a bipolar mixedgroup, a manic group, and a control group, the bi-polar depressed and the mixed bipolar patientswere reported as displaying lower supratentorialwhole-brain glucose metabolic rates than any ofthe comparison groups." Patients with unipolardepression had significantly lower metabolic ratiosof caudate nucleus to whole hemisphere valuesthan either the other patient groups or control sub-jeers. The data further suggested, like the xenonstudies, that global hypometabolism in the de-pressed bipolar patients was state dependent be-cause the metabolic rates "normalized" when pa-tients were euthymic or in a manic phase.

The possibility of a common pathophysiologicbasis for the behavioral expression of depressionwas investigated in a study that examined cerebralme-tabolic rates in a heterogeneous group of de-pressives, patients with obsessive compulsive dis-orders with or without major depression, and nor-mal control subjects.4o A significant lowering ofmetabolic rates for glucose in the left dorsalanterolateral prefrontal cortex was reported in pa-tients with primary depression, unipolar and bi-polar, by comparison with other groups. Similar re-sults, although less marked, were obtained for theright dorsal anterolateral prefrontal cortex. Obses-sive compulsive patients with depression displayedlower metabolic values in the same regions as pa-uerns with primary depression. Glucose metabolicr.ues in the left anterolateral prefrontal cortex cor-related significantly with severity ratings of depres-sion. Following clinical response to medication. themetabolic pattern became normal. The investiga-tors concluded that the findings suggested a leftanterolateral prefrontal cortex abnormality in ma-jor depression.

in conclusion, PET, and early studies using xe-

non, of affecuvely ill patients are consistent ining decreased cerebral metabolism, greatest'inferior frontal region, during the depressedNo theoretical framework has been articulaccommodate these findings with conventheories of monoaminergic Cunction in deprThe findings have considerable face validity"mood changes are a frequent concomitant oftal lobe pathologic states. Although the findidecreased frontal metabolism are reminithose reponed in schizophrenic patients, theanatomic resolution of early PET techniquesnot exclude subtle focal differences betweennosologically distinct patient groups.

OB SESSIONAL DISORDERS

Obsessional compulsive disorders (among the most disabling non psychotic psycdisorders and occur in pure form or asphenomena in other psychiatric disorders,ularly depression, and in primary neuroorders. This association with neurologic .and the similarities between intrusive idOCD and intrusive motor acts, such as tics.to the hypothesis that OCD may be secondysfunction in basal ganglia or relatedThis hypothesis is now amenable to directby PET.

Baxter et al" compared metabolicglucose in 14 patients with OCD, normalsubjects, and patients with primary deprsignificant increase in the metabolic rate inorbital gyrus, a nonsignificant increase inorbital gyrus. and a bilateral increase inmetabolism, specific to patients withobserved. Changes in the caudate to hemetabolic ratio exclusive to medication-rOeD patients were also reported. In astudy, OCD patients with and withoudepression were compared with normalsubjects. Metabolic rates in the QCD pa .depression were significantly lower in [heterolateral prefrontal cortex compared \\'patients without depression, a metabolisimilar to that seen in patients withdepression. A single study of five patiGilles de la Tourene syndrome, a condniinvariably associated with obsessional phhas also reported increased metabolism inganglia.12

PET studies therefore support the hthat obsessional disorders are associatedtional changes in the basal ganglia andlobes. Because OCD is responsive to ps,

macologic treatment it would be of inter-blishthe effects such intervenuons mightcerebral metabolism. The findings if cs-would have a major impact on the noso-

tus of QeD, which has tradiuonally beenas a neurotic disorder with a presumed

gicetiology.

XIETYAND PANIC DISORDERS

ecurrent, discrete, episodic, and spontane-xieryattacks arc the principal features ofdisorder (PD). Its nosology is controversialgenerallyis considered as a form of anxiety,gh there is a viewpoint that PD is a sepa-osologicenuiy with unique pathophysiologicensues. A feature of PO is that it may bekedin susceptible patients by sodium lactate

'on.e relationship between panic disorder and

: has been investigated using PET.13 Cere-flow was measured in patients with a

, of panic disorder and control subjects.ups of patients with PD were identified,

l', patients with posiuv c and negative lactate. Whole brain and hemispheric CBF was

redas well as specific brain regions, includ-e hippocampus and parahippocampal gyrus.

differenceswere found between groups forbrain or hemispheric CBF, bUL the left LO

CBF ratio differed in the parahippocarnpalofthe lactate responders. In an extension of

study,with the addition of further patientsPD and control subjects. the finding was rep-

and in addition asyrnmetries of blood vol-andoxygen metabolism were found. The pa-

iswithPD also had higher whole brain oxygenism."

Therelationship between anxiety, panic dis-r, and normal anucipatory anxiety was in-·gatedin a study 01" subjects at rest and dur-sodiumlactate infusion." During periods of

induced anxiety. there were increases inin the anterior temporal poles, insular (or-

claustrum,lateral putamen, the vicinity of theiorcclliculus, and left. anterior cerebellar vcr-There was no increase in rCBF in the non-

. andcontrol subjects. A separate study mea-!CBF in subjects while they anticipated a

fulelectric shock.":' In these subjects anticipa-anxietywas associated with an increase in CBf

temporal poles.Boththese lauer studies suggest a commonwayfor the expression of anxiety, involvinguresin the temporal poles, and relate well r.o

recent conceptualizations of the pathophysiologicbasis of anxiety."6 Studies of patients with PD pro-vide intriguing evidence that this condition has abiologic basis separate from generalized anxiety.

NEUROPSYCHIATRIC DISORDERS

Psychiatric morbidity, both psychotic and non-psychotic, is common in primary neurologic illness.The application of PET r.o the study of such neu-ropsychiatric disorders should greatly extend ourknowledge of brain-bchavior relationships andclarify pathophysiologic mechanisms mediatingprimary psychiatric disorders. The mechanisms bywhich focal brain insults result in ncuropsychologicimpairment is open to study. Few PET studies, withnotable exceptions, have extended their bounda-ries beyond either a primary ncurologic or psychi-atric perspective. A full review of PET in neuro-psychiatric disorders is beyond the scope of thisarticle: two studies will be considered as illustrativeof the potential application of PET in neuropsy-chiatric research.

A wide range of clinical manifestat.ions are rec-ognized following isolated infarcts of the thalamus,including verbal and visual mcmory impairment,aphasia, neglect, and behavioral change. A clini-copathologic perspective might suggest that theseimpairments reflect a critical role for the thalamusin these diverse psychologic functions. However,findings from PET investigations of such patientswith thalamic lesions provide a different viewpoint,suggesting that neuropsychologic impairment re-lates to remote physiologic effects rather than thedirect pathologic effects of lesions. Baron et al'? ina comhined PET and ncuropsychologic study often patients with single isolated unilateral lesionsof the thalamus reponed that nine of ten displayedsignificant and widespread ipsilateral conical hy-pometabolism remote from the lesion. These pa-tients also displayed a range of neuropsychologicimpairments with a marked trend for deficits LO berelated LO degree of conical hypometabolism.

Depression is a common accornpaniment ofParkinson's disease and may frequently antedatethe onset of motor impairment. Its relationship tothe disorder is controversial and has been concep-tualized as a psychologic reaction to the motor dis-ability. Maybcrg et al-t!lhave investigated the rela-tionship between cerebral glucose metabolismand mood disorder in parkinsonian patients. De-pressed parkinsonian patients had a significantdecrease in rCMRglu in the orbital and inferiorprefrontal cortex compared with nondepressedpatients and control subjects. A significant corrc- :335

SEMINARS IN NEUROLOGY VOLUME 9, NUMBER 4 DECEMBER

lation emerged between ratings of mood severityand rCMRglu in these regions. It is tempting tospeculate that the decreased metabolism was re-lated to loss of extrastriatal ascending monoami-nergic projections, since destruction of these path-ways in primates is associated with decreasedfrontal metabolism.

SUMMARY

PET is potentially the most powerful tool yetavailable for the direct. in vivo investigation of thebiologic basis of psychiatric and neuropsychiatricdisorders. The fulfillment of its potential rests onthe development of methodologies and study de-sign appropriate to psychiatric disorders. To dale,findings in both schizophrenia and affective dis-order, using protocols largely based on resting statedata acquisition, suggest altered regional metabo-lism. These approaches need to be extended, par-ticularly by the application of protocols that utilizePET to obtain longitudinal data under controlledexperimental situations. In two conditions tradi-tionally ascribed to psychologic causes, QeD andPD, there is intriguing evidence of specific biologicabnormalities, which, if confirmed, would lead toa fundamental revision of their nosologic status. Inneuropsychiatric disorders PET findings, althoughpreliminary in nature, offer an alternative para-digm to traditional clinicopathologic correlationsby suggesting that clinical impairments relate tophysiologic effects at sites distant from structurallesions.

REFERENCES

I. lngvar OH. Franzen G. Distribution of cerebral activityin chronic schizophrenia. Lancet 1974:2: 1484-6

2. lngvar OH. Franzcn C. Abnormalities of cerebral bloodnow and metabolism in chronic schizophrenia. AC(;lPsychiatr Scand 1974;50:425-62

3. Smith MR. Brodie JD. Positron emission tomographicstudies in schizophrenia: a review. In: Trtmblc M. ed.New brain imaging techniques and psychopharmacol-ogy. Oxford: Oxford University Press. 1986;35-48

4. Widen I. Bcrgstrom M. Blomqvist C. et "I. Glucose me-tabolism in patients with schizophrenia=-emissioncomputed tomography measurements with ltCcg+u-cose.J Cr-reb Blood Flow Metab 1981;1(Suppl 1):455-6

Farkas T, wolf AP.Jaeger J, et al. Regional brain glucosemetabolism in chronic schizophrenia. Arch Cell Psy-chiatry I!J84;4 I :293-300

wolkin A. Jaeger J, Brodic] O. Cl at. Persistence of met-abolic abnormalities in chronic schizophrenia as deter-mined bv positron emission tomography. Arn] Psy-chiatry 1<185;142:56'1-71

Wolkin A, Angrist B. Wolf A, et at. Low frontal glu-cose utilization ill chronic schizophrenia: a replicationstudy. Arn ] Psychiatry 1988;145:251-3

Dcl.isi LE, Holcomb HH, Cohcn RM, et al. Positron

5,

6,

7.

33(; 8

emission tomography in schizophrenic patient!and without neuroleptic medication. J CereaFlow Metab 1985;5:201-6

9. Cur RE, Resnick SM, Alavi A, et at. Rcgionalbraianon in schizophrenia I. A positron emissionArch Cen Psychiatry 1987;44: 119-25

10. Cur RE, Resnick SM, Cur RC, et al. Regionalbraa(ion in schizophrenia 11. Repeated evaluationpositron emission tomography. Arch Cen1987:44: 126-9

11. Szechtman H, Nahmias C. Carueu S, et al. Effectofrolcpucs on altered cerebral ~Iucose me~.schizophrenia. Arch Gen Psychiatry 1~88;4;):J.

12. Liddle PF. The symptoms of chronic schizophrenia:examination of the positive negative dichoroarPsychiatry 1987;151:145-51

13. DeLisi LE, Buchsbaum MS. Holcomb IIH,etal.correlates of decreased anteroposterior metatxirdierus in positron emission tomography (PEllofophrenic patients. AmJ Psychiatry 1985;142:i

14. Kishimoto H, Kuahara H, Oh no S, et al. Threeof chronic schizophrenia idcuuficd using 11positron emission tomography. Psychiatry Ra.21:285-92

15. McCulloch J, Saraki !lE, McCulloch MC, et al.tribution of alterations in energy metabolisminbrain produced by apomorphine, Brain Res243:67-80

16. Wolkin A, Angrist B, Wolf A, et al. Effects ofamine on local cerebral metabolism in noschizophrenic subjects as determined by positronsion tomography. Psychopharmacology (Bed)I241-6

!7. Ceraud C, Arne Bes MC, Cucf A, Bes A. Reeof" haemodynamic hypofrontality in schizophCcreb Blood Flow Mctab 1987:7:9-12

18. Widen I, Greitz T, Linon JE, et al. PET studieselcose metabolism in patients with schizophrenia..1983;4 :550-2

19. Wong OF. wagner HN, Tune LE, et al. Positrontomography reveals elevated 02 dopamine .in drug naive schizophrenics. Science I986;2:W·63

20. Farde L, Wiescl FA, Hall H. et al. No 02 reccrease in PET study of schizophrenia. ArchGetchiatry 1987;44;671

21. Zeeberg BR. Gibson RE, Rcba RC. ElevatedD!mine rcceptors in drug naive schizophrenic,1988:239:789-91

22. Weinherger DR. Berrnan KF, Zec RF. Pb(dysfunction of the dorsolateral prefrontalschizophrenia: 1 Regional cerebral blooddence. Arch Cen Psychiatry 1986;43: 114-24

23. Cohen RM. Semple WE. Gross M. Cl al. Dysfuprefront a1 subsu-ate of sustained attention inphrenia. Life Sci 1987:40:2031-9

24. warkcmin S, Nilsson A. Risbcrg ] . Karlson$.Afrontal lobe activation ill schizophrenia.] CmiIFlo,,· Metab 1989:9(Suppl 1):S354

25. Volkow XD. Wolf AP, Van-Ocldcr I). et al. Phtlogical correlates of metabolic activity in 18with chronic schizophrenia. Am J Ps}"chiatl"l'144;151-8

26. Berman KF, Torrey EF. Daniel DC, weinbergerfrontal cortical blood now in monozygoticUi

cordant and discordant for schizophrenia. JBlood Flow Metab 1989;9(Suppll):S591

27. Mathcw RJ, Stir+ing-Meyer J, Francis DJ, eta!'blood now in depression. Am J Psychiatry137:1449-50

28. Warren LR, Butler RW, Katholi CR, et al. Focalin cerebral blood now prod uced by rnoneuryiduring a mental mathematics task in normalapressed subjects. Brain Cogn 1984;3:71-85

29. Sackeim HA, Prohovnik I, Aptcr ~~,et (11.Regi

bral bfectsGTUZI

iry, atUytden

rebration i

Oustafflow

Silfvendepr46:2

Coldst.dineorde

Cur Rpsyc695-

Buchslpost.phre1984

Kuh1 I

blood flow in affective disorders; baseline and d-iem of rrearmeru. In: Takabashi R, Flor-Henry 1',GnnelierJ, Niwa F, (eds): Cerebral dynamics, lateral-ily,andpsychopachoJogy. New York: Elscvier. 1987

denhoef P, Ponelacgc P, J~1CqUyJ. Cl at. Regional cc-JtbralbJood flow and l.ueralizcd hemispheric dysfunc-enin depression. BI J Psychian-y 1987;143. 128-32

afson L, Risberg J, Silfverskiold P. Cerebral bloodIlow in dementia and depression. Lancer 19MI; 1:275

erskicld P,RisbergJ. Regional cerebral blood now indepression and mania. Arch Ceu Psychiatry 1989:46:253-59

stein P, Brown GC, Welch K. et al. Age-related de-cline in reBF in schizophrenia and major affective dis-order.J Cereb Blood Mcrab 1985:5(Suppl 1):203-"1

RE,Skolnick BE, Cur RC, er al. Brain IUIIClion in, iatric disorders. Arch Cen Psychiau-v 1984:41:• 9baum MS, DeLisi I.E, l Iolcomb HII, et al. Antero-(trior gradients in cerebral glucose USt; in schizo-

phreniaand affective disorders. Arch Cen Psychiatry1914:41:1159-66

DE,Meuer EJ, Riq{c.· WHo Paucrns of cerebral glu-cse utilizauon in depression, multiple infarct derncn-uaand Alzheimer's di-a-ase. Res Pub! Assoc Res NcrvlderuDis 1985;63:211-~W

gAS, Mener EJ, Rit'gc WH, Kuhl DE: ComparisonofPETmeasurementof local brain glucose metabolismand CAT measurement 01 brain atrophy in chronicschizophrenia and depression. Am J Psychiatry1986:143:175-80

er LRJr, Phclps ME. Mazziona j C. et al. Cerebraleetabolic rates for glucose in mood disorders. Archten Psychiatry 1985;42:441-7

39. Schwartzj M, Baxtcr LR,Jr, Mazzioua jC. et al. The dif-Icrenual diagnosis or depression. JAMA 1987;258:1368-74

40. Baxtcr LR,.Jr, Scbwartz j M, Phclps ME. cl al. Reductionof prefrontal cortex glucose metabolism common tothree lypes of depression. Arch Cen Psychiatry IY89;'16:243-50

41. Baxter LR, .II', Phelps ME, xtazzioua j C. et ••I. Local ce-rebral glucose metabolic rates in obsessive-compuh,ivcdisorder. ARh Cen Psychiatry 1987;44:211-8

42. Chase TN, Foster NI., Fedio I', et al. Gilles de la Tourcuesyndrome: studies with the fluurine-18-I<lbekd Fluoro-dcoxygluccse positron emission tomographic method.Ann NeuruI1984:15(Suppl):SI75

43. Rciman E~l. Raichle 1\1E, Butler FK, et OIL A focal abnor-malil)' in panic disorder, a severe form 01 an xietv. Na-ture 1984:310:683-5

14. Reiman Elo.l. The study of panic disorder w.ing positronemission tomography. Psychiatr Dev 1987;5:63-78

45. Rciman EM, Raichlc ME, Robins E. lnvojvemen t of tem-poral poles in pathological and normal forms of anxi-ClY. .J Cereb Blood Flow Metab 1989;9(Suppll);S5S0

46. Gray JA. The 1H:!uropsycholog)' of anxiety, Oxford: Clar-endon, 1982

47. Baron JC, Amona RD, Pamauo P, Cl al. Effects of tha-lamic stroke on energy metabolism 01 the cerebral cor-tex. A positron tomography studv in man. Brain1986: 109: 1243-59

48. Maybcrg J-IS, Scarkstcin SE. Bolduc P, et al. Fromal lobeh),pomelaholism in depressed, uondemcrucd patientswith Parkinson's disease. J Cereb Blood Flow Mctab1989:9(Suppl 1):346

:l:l7