48ournal ofNeurology, Neurosurgery, and Psychiatry 1997;62:458-463

Psychiatric profiles and patterns of cerebral bloodflow in focal epilepsy: interactions betweendepression, obsessionality, and perfusion relatedto the laterality of the epilepsy

E Bettina Schmitz, John Moriarty, Durval CMichael R Trimble

Costa, Howard A Ring, Peter J Ell,

AbstractObjectives-In a study of patients withfocal epilepsy the hypothesis was exploredthat different measurements of psy-chopathology are related to specific dis-tributions of cerebral perfusion.Methods-Forty patients had SPECTperformed with 9hnTc-HMPAO. In addi-tion, patients received a psychiatric evalu-ation with the following psychiatricquestionnaires: the Beck depressioninventory, the Leyton obsessionalityinventory, the Bear-Fedio questionnaire,and the social stress and support inter-view. Patients were analysed in twogroups according to the laterality of theepilepsy. Nine patients were excludedbased on poor quality scans (n = 1),unlateralised epilepsy (n = 4), and left orambidextrous handedness (n = 4).Results-There were no overall differ-ences between the left and right epilepsygroups on measures of psychopathology.Associations were found between scoreson some of the rating scales and regionalcerebral blood flow. Specifically, forpatients with left sided epilepsy, higherscores on the Beck depression inventorywere associated with lower contralateraltemporal and bilateral frontal perfusion,and higher occipital perfusion. Forpatients with right sided epilepsy higherscores on the Leyton obsessionality inven-tory were associated with increased per-fusion in ipsilateral temporal, thalamic,and basal ganglia regions and bilateralfrontal regions.Conclusion-The results do not supportthe notion that lateralised epileptogeniclesions are associated with different levelsof depression, obsessionality, or personal-ity traits. They support the view thatcertain psychopathological symptom pat-terns are related to specific regional dys-functions depending on the laterality of ahemispheric lesion.

(7 Neurol Neurosurg Psychiatry 1997;62:458-463)

Keywords: focal epilepsy; laterality; HMPAO-SPECT;depression; obsessive-compulsive disorder

Behavioural changes in patients with epilepsyhave been identified for many years. Theserange from depression and anxiety' to psy-

choses,2 and may include some specific per-sonality problems that have variously beenreferred to as the interictal personality disorderof epilepsy and the Gastaut-Geschwind syn-drome.3 A specific association between thesesyndromes and temporal lobe epilepsy hasbeen claimed.45 This has not been easy toestablish, however, and psychopathology islikely also to reflect such factors as seizureseverity, global cerebral damage, medicationeffects, and adequacy of psychosocial sup-ports.The personality syndrome of epilepsy has

been particularly controversial and, althoughBear and Fedio6 developed their questionnairespecifically to monitor behavioural problemsassociated with temporal lobe disorders, it hasbeen criticised as being a general measure ofpsychopathology,7 with little specificity.

It has become possible to explore brain-behaviour relations in neuropsychiatric disor-ders using new brain imaging techniques.Disordered cerebral function has beendescribed in schizophrenia, depression, andobsessive-compulsive disorder using both PETand single photon emission computed tomog-raphy (SPECT).4 In these states, mesial tem-poral regions as well as the frontal lobes andbasal ganglia are often highlighted. A goodexample is the work of Baxter et al,8 who haveidentified changes in frontal-basal ganglia cir-cuits in patients with obsessive-compulsivedisorder. These changes have been confirmedby others4 and, furthermore, are altered withtreatment.9 10

Interictal SPECT in patients with focalepilepsies reliably show hypometabolism andreduced cerebral blood flow associated withthe site of the epileptogenic focus."I 12 How-ever, we and others have emphasised the wide-spread nature of perfusion changes seen inassociation with temporal lobe epilepsy, pre-dominantly in limbic frontal regions.'2 13 Thesemay be related to functional deafferentation,interictal inhibitory activity, postictal depletionof substrates, or other phenomena. '" Epilepsyis a cerebral disorder with more widespreaddisturbance of cerebral function than oftenappreciated, which may be reflected in a widevariety of behaviours, from a seizure, to anautomatism, to interictal cognitive and behav-ioural symptoms. It is possible that the distri-bution of localised hypometabolism isassociated with the interictal cognitive andbehaviour states.

In this study we explore some of these rela-

Abteilung furNeurologie, VirchowKlinikum, Humboldt-Universitot, Berlin,GermanyE B SchmitzDepartment ofNeuropsychiatry,Institute ofNeurology,Queen Square,London, UKJ MoriartyH A RingM R TrimbleInstitute ofNuclearMedicine, UCLMedical School,London, UKD C CostaP J EllCorrespondence to:Dr Bettina Schmitz,Abteilung fur Neurologie,Virchow KlinikumAugustenburger Platz 1,13533 Berlin, Germany.Received 22 November 1994and in final revised form9 December 1996Accepted 10 December 1996

458

on March 6, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.62.5.458 on 1 M

ay 1997. Dow

nloaded from

CBF and psychopathology in epilepsy

tions further, using epilepsy as a model andSPECT as our imaging method. Our mainhypotheses are: (1) affective symptoms areassociated with decreased perfusion in fron-totemporal regions; (2) obsessive compulsivebehaviour is associated with increased perfu-sion in frontotemporal and basal gangliaregions; (3) personality symptoms identifiedby the Bear-Fedio questionnaire are associatedwith decreased perfusion in the temporal lobe;and (4) social variables are not associated withlocalised cerebral perfusion.

In addition to temporal, frontal, and basalganglia regions we also measured blood flow inthe parietal and occipital cortex. These wereused as control regions and we predicted find-ing no associations between perfusion and psy-chopathology in these regions. We investigatedpatients with left and right sided epilepsiesseparately and predicted that perfusion pat-terns would relate to psychopathology regard-less of the laterality of the epilepsy.

MethodsPATIENTSWe investigated 40 consecutive patients whowere referred to the National Hospital forNeurology and Neurosurgery, London forassessment and treatment of chronic focalepilepsies. One patient was subsequentlyexcluded as a result of a parainjection of theisotope (resulting in a very poor quality scan).The clinical diagnosis of epilepsy was made byexperienced epileptologists according to theInternational Classification of Epilepsies andEpileptic Syndromes.'4 None of the patientswere being investigated specifically for psychi-atric illness, although some had had previousepisodes of psychopathology. Patients aredescribed in more detail elsewhere."2 The pro-ject was approved by the ethics committee ofthe National Hospital for Neurology andNeurosurgery.

LATERALITY OF EPILEPSYThe laterality of the epilepsy was establishedon the basis of EEG and MRI findings.Seventeen patients had ictal EEGs in additionto interictal EEGs. All patients had MRI(Siemens, 1.5 Tesla). The laterality of EEGand MRI findings were independently deter-mined by investigators who were blind to theSPECT and psychiatric data. Handedness wasassessed with the handedness research ques-tionnaire. I5

SPECT STUDIESWe used 99mTc-HMPAO SPECT (mean dose606 MBq). Patients were scanned interictallywith their eyes closed using a triple detectorSPECT system (GE Neurocam, high resolu-tion collimators, 128 projections, 64 x 64matrix, total acquisition time 28 minutes).Slices of 9 mm thickness were reconstructed incoronal, sagittal, and transverse planes using afiltered back projection method. The trans-verse planes were specially oriented along theaxis of the temporal lobe. The resolution ofthe scanner at FWHM is 9-2 mm.'6

Scans were quantitatively analysed by anobserver who was blind to clinical data andpsychiatric findings. For quantification weused a region of interest method; 4 cm2 boxregions of interest (5 x 5 pixel, 3-6 cm' voxel)were placed in the following regions: mesialtemporal, lateral temporal, orbitofrontal, dor-solateral frontal, thalamus and parietal cortexon coronal slices, basal ganglia and occipitalcortex on horizontal slices, and frontopolarcortex on sagittal slices. Regions were identi-fied visually from the SPECT with referenceto an atlas of brain anatomy.'7 To allow inter-subject comparisons, for all regions of interestwe estimated an activity index (AI):

AI =Mean counts per pixel (MCPP)

MCPP of six extratemporalcortical regions

The six extratemporal cortical regions wereright and left dorsolateral frontal cortex, rightand left parietal cortex, and right and left occip-ital cortex

PSYCHIATRIC QUESTIONNAIRESAll patients were given the following question-naires on the day of the SPECT: the Beckdepression inventory, the Leyton obsessionalinventory, the Bear-Fedio questionnaire, andthe social stress and support interview.The Beck depression inventory is a 21 item

self rating scale with four severity ratings foreach item. Each item describes a specific behav-iour manifestation of depression.'8 The Leytonobsessional inventory is a comprehensive ratingscale that evaluates both state and trait obses-sive-compulsive symptoms and consists of 68questions. We only considered the total (stateplus trait) score for the purpose of this analy-sis.19 Average scores of 18 (SD 10) have beenreported in normal populations with a tendencyto score slightly lower on repeat testing.Obsessional patients score much higher (44(SD 1 1))."' The Bear-Fedio questionnaire is aself administered 18 item questionnaire relatingto issues thought to reflect the interictal person-ality syndrome of epilepsy, such as religiosity,hypergraphia, circumstantiality, viscosity, para-noia, and altered sexuality.6 There are no stan-dardised scores for this scale. The social stressand support interview2" is a semistructuredinterview to rate social stress and support, con-sisting of six sections which are each given rat-ings between -2 and +2. We included this as a"control" questionnaire because there are nohypotheses with respect to localised abnormali-ties relating to social problems. All patientsunderwent formal psychological testing andboth verbal and performance IQs were esti-mated.

MEASURES OF ASSOCIATIONBecause laterality of the epilepsy focus itself islikely to affect measures of perfusion and mayalso influence psychopathology, patients weredivided into those whose epilepsy could be lat-eralised to the left or the right. For clarity,patients in whom the epilepsy could not clearlybe lateralised were excluded as well as patients

459

on March 6, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.62.5.458 on 1 M

ay 1997. Dow

nloaded from

Schmitz, Moriarty, Costa, Ring, Ell, Trimble

Table 1 Scores on psychiatric rating scales in patients with right and left sided epilepsy

All patients Right epilepsy Left epilepsyMean (SD, range) Mean (SD, range) Mean (SD, range) P value*

BDI 10-8 12 6 8 3 0-38(13-1, 0-71) (16 7, 0-71) (4 9, 2-17)

LOI 22-7 25-6 18-5 0 09(11-31, 2-54) (13 4, 2-54) (5 8, 7-26)

BFQ 4 0 4-7 3-2 0 21(3-1, 0-12) (3-7, 0-12) (2-1, 0-8)

SSSI 1.1 1-5 0 5 0 33(2 79, -6-+5) (2-5, -4-+5) (0 3, -6-+4)

VIQ 92 4 93-1 91 5 0 78(13 8, 70-123) (14-2, 70-123) (13 9, 70-107)

PIQ 93 3 92-1 94-8 0-68(16 5, 64-122) (17-0, 64-122) (16 5, 67-113)

*Unpaired t test, two tailed, test for equality.BDI = Beck depression inventory; LOI = Leyton obsessionality inventory; BFQ = Bear-Fedioquestionnaire; SSSI = social stress and support interview; VIQ = Verbal IQ; PIQ =performance IQ.

Table 2 Patients with left sided epilepsy (n = 13): correlations between regional bloodflow (activity index) and scores on psychiatric rating scales

ROI BDI LOI BFQ 5SSI

Mesial temporal Right -0-19 -0-19 0-41 -0-11Lateral temporal Right -0-62* -0-15 0-16 0 50Orbitofrontal Right 0-37 0-23 0-43 0 11Frontopolar Right -0 80** -0 01 0-26 0-51Dorsolateral frontal Right -0-73** -0-35 0.33 0.06Basal ganglia Right - 004 0-03 0-04 - 0 04Thalamus Right -0-12 0-06 0 14 0 03Parietal Right -0 36 0 06 0 53 -0 20Occipital Right 0-60* 0-08 -0-20 0-32Mesial temporal Left -0-08 0-29 0 30 0 17Lateral temporal Left -0-12 0-35 0-29 0-42Orbitofrontal Left -0 8 0-06 0-28 0 32Frontopolar Left -0-89*** 0 07 0 21 0-36Dorsolateral frontal Left -0 71* -0 30 -0-25 0 09Basal ganglia Left -0-19 0-16 0-04 -0 21Thalamus Left -0-02 0-13 -0-01 0-12Parietal Left 0 30 0-39 -0-08 - 0-46Occipital Left 0.62* 0-01 -0 40 0 17

*P < 0-05; **P < 0-01; ***P < 0-001; Pearson's correlation coefficient, two tailed.

who were not right handed. The Pearson'sproduct moment correlation coefficient wasthen calculated for each selected cerebral regionand each measure of behaviour. All statisticalanalyses were two tailed. Because of theexploratory nature of these studies, we did notapply strict Bonferroni corrections. However,to compensate for multiple testing, the signifi-cance level was set at 0 01, although a statististi-cal trend at a level of 0 05 is indicated in thetables. The SPSS/PC+ computer programme(v4) was used for these calculations.

ResultsThe EEG was non-localising in seven cases.Of the remaining patients, 31 had a fron-

Table 3 Patients with right sided epilepsy (n = 18): correlations between regional bloodflow (activity index) and scores on psychiatric rating scales

ROI BDI LOI BFQ SSSI

Mesial temporal Right 0-25 0-56* 0-22 -0 40Lateral temporal Right 0-11 0-65** 0-20 -0-13Orbitofrontal Right 0-06 0.62** 0-19 -0 30Frontopolar Right 0-24 0 43 0-28 -0 33Dorsolateral frontal Right 0-14 0-17 0-20 -0-13Basal ganglia Right -0-08 0.54* 0-06 -0 07Thalamus Right 0-21 0.56* 0-14 -0 15Parietal Right 0-31 0-11 -0-32 -0-25Occipital Right 0-24 0 00 0-06 0-24Mesial temporal Left 0-26 0 30 0-26 -0-22Lateral temporal Left 0-10 0-27 0-21 -0-15Orbitofrontal Left -0-03 0.58* 0 35 -0 07Frontopolar Left 0-23 0.50* 0 35 -0-28Dorsolateral frontal Left 0 0 14 0-19 -0 35Basal ganglia Left 0 03 0 46 -0 12 0.19Thalamus Left 0-15 0-45 0-08 0 04Parietal Left 0-18 -0 23 -0-15 -0-17Occipital Left 0 33 -0-16 -0-02 0-44

*P < 0-05; **P < 0-01; Pearson's correlation coefficient, two tailed.

totemporal focus and one had an occipitalfocus. Brain MRI was normal in sevenpatients, 28 had hippocampal pathology, threehad temporal pathology without hippocampalinvolvement, and one had an extratemporalstructural abnormality.

In 35 patients the epilepsy could be later-alised to one side based on abnormalities onEEG, MRI, or both. In two patients, who weresubsequently excluded, EEG and MRI indi-cated opposite sides. Another two patientswho had negative or non-lateralising findingsin both EEG and MRI were excluded. We alsoexcluded two patients who were left handedand two patients who were ambidextrous.Eighteen patients with right sided epilepsy and13 patients with left sided epilepsy remainedfor further analyses.The mean age of onset of epilepsy was 10-6

(SD 6-9, range 1-26) years, and the mean ageat investigation was 29-1 (SD 9-2, range16-56) years. Thirteen patients were male, 18female. All patients were taking antiepilepticdrugs. None was receiving medication for psy-chiatric indications.

Table 1 gives the scores on the rating scales.There was no relation between the scores andsex, age of onset of epilepsy, or age at investi-gation. Scores did not significantly differenti-ate patients with right or left sided epilepsy.

There was a significant positive correlationbetween scores on the Leyton obsessionalityinventory and the Bear-Fedio questionnaire(r = 053, P < 0.01). Comparing the right:left ratios of blood flow in patients with rightversus patients with left sided epilepsy therewas significant ipsilateral hypoperfusion of themesial temporal lobe (t test for unequality,P < 0 05).

MEASURES OF ASSOCIATION (TABLES 2 AND 3)Patients with left sided epilepsy showed a neg-ative association between scores on the Beckdepression inventory and contralateral lateraltemporal lobe perfusion and bilateral frontalperfusion (highly significant for the left fron-topolar cortex), and a positive associationbetween scores on the Beck depression inven-tory and occipital cortex perfusion.

Patients with right sided epilepsy showed apositive association between scores on theLeyton obsessionality inventory and ipsilateraltemporal lobe, thalamic and basal ganglia per-fusion, and bilateral frontal perfusion.

There were no significant correlationsbetween scores on both the Bear-Fedio ques-tionnaire and the social stress and supportinterview, and regional cerebral blood flow.

DiscussionMETHODOLOGICAL CONSIDERATIONSThe scanner used by our group is an up todate, triple headed camera, with high resolu-tion. In epilepsy we have shown that this hasgood specificity with regard to laterality of theseizure focus.'2 With respect to quantificationof perfusion it must be acknowledged that aci-tivity indices can only approximate absoluteblood flow.'2 Our method of region of interest

460

on March 6, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.62.5.458 on 1 M

ay 1997. Dow

nloaded from

CBF and psychopathology in epilepsy

placement has been shown to have validity andreliability in past investigations.22

In presenting the data, we have relied oncorrelation analysis, and draw attention toprofiles of perfusion found in association withvarious rating scale scores. Although multipleanalyses have been performed, strictBonferroni corrections have not been applied,as we are primarily interested at this stage inexploring possible patterns of interactions andcorrelations, rather than individual localisa-tion. Thus it is conceivable that some of ourapparently significant correlations could havearisen by chance, although not all of them. Ofcourse, there is no way of knowing which onesmight and which ones might not be spurious.

Epilepsy has been used by us as a biologicalmodel of cerebral dysfunction with which toexplore brain-behaviour relations. Ourpatients had various seizure types, and syn-dromes, and were not selected because theyhad. However, the psychiatric illnesses are

common in epilepsy, and a high frequency ofreported symptoms may be expected in thispopulation. As these patients did not have a

systematic psychiatric evaluation, we cannotexclude the possibility that some of them mayhave satisfied diagnostic criteria for a psychi-atric syndrome. Although further evaluation oftheir mental states from a clinical viewpointwould have been useful, our intention was tostudy interrelations between levels of perfu-sion and psychiatric symptoms, as opposed tospecific diagnostic categories.The range of scores on the rating scales

allow us to explore the associations betweenparticular behavioural syndromes in epilepsyand patterns of regional blood flow. For exam-ple, although only two of our patients scoredgreater than 30 on the Beck depression inven-tory (indicating a severe degree of depression),almost half of them (n = 17) scored greaterthan 10, suggesting at least milder degrees ofmood disturbance not amounting to psychi-atric disorder. In three comparable studies onfocal epilepsy, the mean Beck depressioninventory scores were lower than in our study,being 6, 8, and 9 respectively.'32324 There areno comparable data with respect to scores onthe Leyton obsessionality inventory and scoreson the social stress and support interview. Asnoted above, there are few data available on

the Bear-Fedio questionnaire, and in the origi-nal paper mean data are not provided.

LATERALITY OF EPILEPSY, MOOD, ANDREGIONAL PERFUSIONFor the depression scores, a relation withregional perfusion abnormalities was foundwith the left sided epilepsies only, the higherthe ratings for depression, the lower the perfu-sion in frontal areas bilaterally. This finding ofan association between mood and frontal dys-function is in keeping with the data from otherswho have used neurological models to explorethe genesis of affective symptoms. For exam-ple, the results of Starkstein snd Robinson25 instroke patients, and of Ring et al26 in patientswith Parkinson's disease emphasise an associa-tion between affect and frontal areas. It also

supports data from patients with affective dis-order and no underlying neurological lesions,in whom several groups have noted frontalhypoperfusion.2728

In epilepsy, Hermann et al 24 noted a specificrelation between levels of perseveration, sug-gestive of frontal lobe involvement, anddepression, but only in patients with a leftsided focus. These findings were confirmedand extended in a recent study by Seidenberg etal.23An association between left hemispheric

lesions and depressive symptoms has beennoted previously. Specifically, in relation toepilepsy, many-for example, Mendez eta129-but not all investigators have found thisrelation. In a study of depression and epilepsy,Robertson et al I were unable to find any spe-cific links between depression and the localisa-tion of the epilepsy focus. We found similarlevels of self reported depression for patientswith left and right unilateral epilepsy.

This is in accordance with the findings ofSeidenberg et al23 who suggested that specificcortical regions in both hemispheres asymmet-rically are involved in depression. Yamaguchiet al,30 in a study on blood flow and depressionin poststroke patients, found that severity ofdepression was inversely correlated with bloodflow in the parieto-occipital regions of theright hemisphere and in anterior temporalregions of the left hemisphere. Heller3' andRobinson et al 32 have discussed further the evi-dence supporting the possibility of two distinctlateralised neural systems influencing moodstate, a left frontal and a right posterior sys-tem. The studies of Hermann et al 24 andSeidenberg et al23 were unable to investigatethis hypothesis because they only used theWisconsin card sorting test as a measurementfor frontal dysfunction. We included posteriorregions in our analysis, and are able to test thehypothesis of an additional posterior moodregulating system which may be involved inpatients with right hemispheric lesions. Ourdata do not support these suggestions, as wefound no hypoperfusion changes in theseregions.

Ross and Rush33 suggested that endogenousdepressions are predominantly modulated bythe right hemisphere, whereas the more cogni-tive non-endogenous depressions are modu-lated by the left hemisphere. The Beckdepression inventory is heighly weightedtoward cognitive aspects of depression ratherthan the syndrome of major depression.Frontal lobe hypoperfusion in this contextmay reflect the effects of cortical rather thansubcortical dysfunction, leading to cognitiverather than vegetative manifestations."3 Thusone explanation of the lack of any localisedperfusion pattern in our right sided epilepsygroup might be on account of our use of theBesk depression inventory, as opposed toanother method of assessing affective symp-toms. In addition, prominent endogenous andvegetative features, which may be related moreto diencephalic abnornmalities, cannot be stud-ied by present HMPAO-SPECT or PET tech-nologies.

461

on March 6, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.62.5.458 on 1 M

ay 1997. Dow

nloaded from

Schmitz, Moriarty, Costa, Ring, Ell, Trimble

The relative occipital hyperperfusion, seenonly in patients with left sided pathology, isconsistent with several findings that a posteriorshift in perfusion may occur in depression.34

LATERALITY OF EPILEPSY, OBSESSIVECOMPULSIVE BEHAVIOUR, AND REGIONALPERFUSIONDespite the reporting of single cases35 36 a spe-cific link between epilepsy and obsessive-com-pulsive disorder has not been shown.However, it is known that several neurologicalstates are clinically associated with a range ofobsessive-compulsive disorder-like phenom-ena, perhaps better described as obsessive-compulsive behaviour. These includeencephalits lethargica, Sydenham's chorea,and Parkinson's disease.374 In descriptions ofthe personality of patients with epilepsy, someobsessional traits have been identified, andsome of these are included in the Bear-Fedioquestionnaire. We found a significant associa-tion between subject scores on the the Leytonobsessionality inventory and the Bear-Fedioquestionnaire.

For our perfusion data, the most obviousassociations with the Leyton obsessionalityinventory were right temporal and bilateralfrontal, with additional associations being seenwith the measurements of right thalamic andbilateral basal ganglia perfusion. These find-ings are of particular interest in view of thePET findings in obsessive-compulsive dis-order,8 3839 which include frontal, caudate, andthalamic hyperperfusion. It is important tonote that the changes in cerebral perfusionassociated with the Leyton obsessionalityinventory do not directly reflect the pathologyof the seizure focus, as interictal scans inpatients with focal epilepsies show reducedcerebral blood flow associated with the site ofthe epileptogenic focus.'2Our finding that the links between perfusion

and obsessive-compulsive disorder wererestricted to the right epilepsy group wereunexpected but nevertheless intriguing. It doesnot seem that the measure of obsessionality islinked to the presence of an epilepsy focus onthe right side, as we found no difference in ourscores for patients with right or left sidedlesions. Rather, our data support the hypothesisthat specific cerebral perfusion patterns areseen in association with our measure of obses-sionality. We found this association only inthose patients whose epilepsy had a right sidedfocus. Our data do show a clear distinctionbetween the perfusion changes linked with theLeyton obsessionality inventory and thosewith the Beck depression inventory. The firstreflect mainly hyperperfusion, the secondhypoperfusion.

PERSONALITY AND SOCIAL VARIABLES ANDPATTERN OF PERFUSIONOur study failed to show significant associa-tions between regional blood flow and thetotal score on the Bear-Fedio questionnaire.This finding is consistent with some previousstudies which rejected Bear and Fedio's originalfindings that there is a specific constellation of

personality traits related to temporal lobepathology.7 40We did not find specific perfusion patterns

associated with the social stress and supportinterview. As this was deliberately chosen as acontrol questionnaire, containing items whichseemed unlikely to be related to the localisa-tion of the epileptogenic lesion, the absence ofany significant correlations strengthens ourpositive findings for the measures of psy-chopathology.

ConclusionsStudies of psychopathology in the absence of aneurological illness have increasingly consid-ered symptoms or syndromes as opposed todiagnostic categories for example, the associ-ations between psychomotor retardation andabnormalities of dorsolateral prefrontal cortexperfusion described by Bench et al.27 Our datasupport this approach. The finding in ourstudy of different patterns of blood flow inpatients with epilepsy arising from differentsites, and the fact that these patterns can beassociated with selective psychopathologicalsymptoms, is of special interest to those inter-ested in behavioural neurology. In particularwe emphasise that perfusion changes seen dis-tant from a focal site of pathology, whichreflect on distributed networks of abnormality,may be closely linked to the changes in behav-iour reported in patients with neurologicallesions, including epilepsy.The relations we found between depressed

mood and frontal hypoperfusion and betweenobsessionality and frontal hyperperfusion addto the accumulating data supporting a promi-nent role for frontal brain regions in the regula-tion of mood and behaviour.4 Our data alsosupport the reported associations betweenobsessive-compulsive disorder and cerebralcircuits involving frontal-subcortical distrib-uted pathways.

1 Robertson MM. The organic contribution to depressive ill-ness in patients with epilepsy. 3rournal of Epilepsy 1989;2:189-230.

2 Trimble MR. The psychoses of epilepsy. New York: RavenPress, 1991.

3 Trimble MR. Interictal psychoses of epilepsy. In: SmithDB, Treiman DM, Trimble MR, eds. Advances in neurol-ogy. New York: Raven Press, 1991:143-52.

4 Trimble MR. Biological psychiatry. 2nd ed. Chichester: JohnWiley, 1996.

5 Waxman SG, Geschwind N. The interictal behaviour syn-drome in temporal lobe epilepsy. Arch Gen Psychiatry1975;144: 1053-6.

6 Bear DM, Fedio P. Quantitative analysis of interictalbehaviour in temporal lobe epilepsy. Arch Neurol 1977;34:454-67.

7 Mungas D. Interictal behaviour abnormality in temporallobe epilepsy. A specific syndrome or non-specific psy-chopathology? Arch Gen Psychiatry 1982;39: 108-i1.

8 Baxter LR, Schwartz JM, Maziotta JC, et al. Cerebral glu-cose metabolic rates in obsessive compulsive disorder.Am JPsychiatry 1988;145:1560-3.

9 Hoehn-Saric R, Pearlson GD, Harris GJ, Machlin SR,Camargo EE. Effects of fluoxetine on regional blood flowin obsessive compulsive patients Am Psychiatry1991;148: 1243-5.

10 Baxter LR, Schwartz JM, Bergman JM, et al. Caudate glu-cose metabolic rate changes with both drug and behav-iour therapy for obsessive compulsive disorder. Arch GenPsychiatry 1992;49:681-9.

11 Duncan R, Patterson J, Hadley DM, et al. CT, MR, andSPECT imaging in temporal lobe epilepsy. J NeurolNeurosurg Psychiatry 1990;53:1 1-5.

12 Schmitz B, Costa DC, Jackson DG, Moriarty J, Duncan JS,Trimble MR, Ell PJ. Optimised interictal HMPAO-SPECT in the evaluation of partial epilepsy. Epilepsy Res1995;21:159-67.

462

on March 6, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.62.5.458 on 1 M

ay 1997. Dow

nloaded from

CBF and psychopathology in epilepsy

13 Bromfield EB, Altshuler L, Leiderman DB, et al. Cerebralmetabolism amd depression in patients with complexpartial seizures. Arch Neurol 1992;49:617-23.

14 Commission on Classification and Terminology of theInternational League Against Epilepsy. Proposal forrevised classification of epilepsies and epileptic syn-dromes. Epilepsia 1989;30:389-99.

15 Annett M. The binomial distribution of right, mixed andleft handedness. QJExp PsycholA 1967;19:327-33.

16 Kouris K, Jarritt PH, Costa DC, Ell PJ. Physical assess-ment of the GEICGR Neurocam and comparison with asingle rotating gamma-camera. Eur J7 Nucl Med 1992;9:236-42.

17 Tailarach J, Tournoux P. Co-planar stereotaxic atlas of thehuman brain. Stuttgart: Thieme Verlag, 1988.

18 Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J.An inventory for measuring depression. Arch GenPsychiatry 1961;4:561-71.

19 Snowdon J. A comparison of the written and postbox formsof the Leyton obsessional inventory. Psychol Med 1980;10:165-70.

20 Cooper J. The Leyton obsessional inventory. Psychol Med1970;1:48-64.

21 Jenkins R, Mann AH, Belsey E. The background, designand use of a short interview to assess social stress andsupport in research and clinical settings. Soc Sci Med1981;15:195-203.

22 Ring HA. Functional imaging in neuropsychiatric disorders[MD thesis]. London: University of London, 1993.

23 Seidenberg M, Hermann B, Noe A, Wyler AR. Depressionin temporal lobe epilepsy: interaction between lateralityof lesion and Wisconsin card sort performance.Neuropsychiatry, Neuropsychology, and BehaviouralNeurology 1995;8:81-7.

24 Hermann BP, Seidenberg M, Haltiner A, Wyler AR. Moodstate in unilateral temporal lobe epilepsy. Biol Psychiatry1991;30:1205-18.

25 Starkstein SE, Robinson RG. Affective disorders and cere-bral vascular disease. Br7 Psychiatry 1989;154:170-82.

26 Ring HA, Bench C, Trimble MR. Depression inParkinson's disease. BrJ Psychiatry 1994;165:333-9.

27 Baxter LR, Schwartz JM, Phelps ME, et al. Reduction ofprefrontal cortex glucose metabolism common to three

types of depression. Arch Gen Psychiatry 1989;46:243-50.28 Bench CJ, Friston KJ, Brown RG, Frackowiak RS, Dolan

RJ. Regional cerebral blood flow in depression: relation-ship to clinical dimensions. Psychol Med 1993;23:579-90.

29 Mendez MF, Taylor JL, Doss RC, Salguero P. Depressionin secondary epilepsy: relation to lesion laterality. JfNeurol Neurosurg Psychiatry 1994;57:232-3.

30 Yamagushi S, Kobayashi S, Koide H, Tsunematsu T.Longitudinal study of regional cerebral blood flowchanges in depression after stroke. Stroke 1992;23:1716-22.

31 Heller W. The neuropsychology of emotion: developmentalpatterns and implications for psychopathology. In: SteinWL, Leventhal B, Trabasso T, eds. Psychological and bio-logical approaches to emotion. Hillsdale, NJ: LawrenceErlbaum Associates, 1990.

32 Robinson RG, Starr LB, Price TR. A two year longitudinalstudy of mood disorders following stroke. BrJ7 Psychiatry1984;144:256-62.

33 Ross ED, Rush J. Diagnosis and neuroanatomical corre-lates of depression in brain-damaged patients. Arch GenPsychiatry 1981;38: 1344-54.

34 Bench CJ, Friston KJ, Brown RG, Scott LC, FrackowiakRS, Dolan RJ. The anatomy of melancholia-focalabnormalities of cerebral blood flow in major depression.PsycholMed 1992;22:607-15.

35 Kettl, PA, Marks IM. Neurological factors in obsessive-compulsive disorder. Two case reports and a review ofthe literature. BrJ7Psychiatry 1986;149:315-9.

36 Kroll L, Drummond LM. Temporal lobe epilepsy andobsessive compulsive symptoms. J7 Nerv Ment Dis 1983;181:457-8.

37 Cummings JL. Frontal-subcortical circuits and humanbehaviour. Arch Neurol 1993;50:873-80.

38 McGuire PK, Bench CJ, Frith CD, et al. Functionalanatomy of obsessive-compulsive phenomena. Br JPsychiatry 1994;164:459-68.

39 Rauch SL, Jenike MA, Alpert NM, et al. Regional CBFmeasured during symptom preoccupation in OCD using015-labelled CO2 and PET. Arch Gen Psychiatry 1994;51:62-70.

40 Rodin E, Schmaltz S. The Bear-Fedio personality inventoryand temporal lobe epilepsy. Neurology 1984;34:591-6.

463 on M

arch 6, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.62.5.458 on 1 May 1997. D

ownloaded from