European Journal of

Nuclear Medicine Original article

Benzodiazepine receptors and cerebral blood flow in partial epilepsy Peter Bar tens te in 1, A lbe r t Ludo lph 2, O tmar Schobe r 1, Ge rha rd Lot tes 1, K l e m e n s S c h e i d h a u e r 1, Joachim Sciuk 1, and Hans -F r i ede r Beer 3

Department of Nuclear Medicine and 2 Department of Neurology, University of M(inster, Albert-Schweitzer-Strasse 33, D-4400 Mfinster, Federal Republic of Germany a Paul Scherrer Institut Villingen, Switzerland

Received 17 May and in revised form 20 August, t990

Abstract. It was the aim of this study to compare benzo- diazepine (Bz) receptor binding and cerebral perfusion in patients with partial epilepsy. Single photon emission tomography (SPET) studies with the flow-marker tech- netium 99m hexamethylpropylene amine oxine (99mTC- HMPAO) and with the 12q-labelled Bz-receptor ligand Ro 16-0154 (~23I-Iomazenil) were performed in 12 pa- tients with partial epilepsy, all with normal magnetic resonance imaging (MRI) and computed tomography (CT) scans. The SPET studies with 123I-Iomazenil were carried out 5 min and 2 h after injection. At 2 h the distribution of activity was very similar to the expected distribution of Bz-receptors in the human brain, known from positron emission tomography (PET) work and post-mortem studies. Early images showed a significant- ly higher tracer accumulation in the area of the basal ganglia, cerebellum, and naso-pharyngeal space. This finding is caused by non-specific binding and the contri- bution of the tracer in the blood pool in this phase. Also after 2 h p.i. of 123I-Iomazenil, 9 of the 12 patients showed a focal decrease of of Bz-receptor binding. Ten patients had focal flow abnormalities with 99mTC- HMPAO SPET. In 8 subjects impairment of flow was seen in sites of reduced *23I-Iomazenil uptake. 123I-Io- mazenil is suitable for Bz-receptor mapping. In this series of patients, Bz-receptor mapping with SPET seems to offer no advantage over 99mTc-HMPAO in the detection of epileptic foci.

Key words: Partial epilepsy SPET Benzodiazepine receptors

Eur J Nuel Med (1991) 18:111-118

Offprint requests to: P. Bartenstein

I n t r o d u c t i o n

In drug-resistant partial epilepsy, the removal of the fo- cus by surgery leads to an improvement of the disease in two-thirds of all patients (Ravagnati 1987). For suc- cessful surgery, however, the site of the epileptic focus needs to be known exactly. This is not easy even using electroencephalography (EEG), transmission computed X-ray tomography (CT) or magnetic reasonance imag- ing (MRI) (Sussman et al. 1984; Theodore et al. 1985). Even intracerebral EEG, an invasive method which car- ries a small risk of cerebral bleeding and infection, is inaccurate as a localizing method in 10% 20% of all cases (Lieb et al. 1981).

Nuclear medicine methods, such as the mapping of regional glucose metabolism with fluorodeoxy F18 glu- cose (18F-FDG) by position emission tomography (PET) or the imaging of cerebral blood flow with lipo- philic compounds like technetium 99m hexamethylpro- pylene amine oxine (99mTc-HMPAO) by single photon emission tomography (SPET), are beginning to be ap- plied to the diagnosis of epilepsy (Bonte et al. 1983; Ell and Costa 1989; Wienhard etal. 1989; Kuikka etal. 1990). More specific tracers should improve the diagnos- tic value of nuclear medicine in this clinical question.

Disturbance of the inhibitory system associated with gamma-aminobutyric acid (GABA) in the epileptic focus is a topic of current interest (Perry and Hansen 1981; Ribak 1985; Sherwin et al. 1986, 1988).

Reduction of benzodiazepine (Bz) receptor binding in epileptic loci has been reported by a Swedish group (Savic et al. 1988). They used the *1C-labelled Bz-ant- agonist Ro 15-1788 (Flumazenil) as receptor-ligand and positron emission tomography (PET). The Bz-receptor is part of a postsynaptic membrane receptor ionophore complex, which contains the GABA receptor and is therefore a suitable marker for the GABA-ergic synapsis (Olsen 5981).

© Springer-Verlag 1991

112

I t was the a im o f ou r s tudy to examine Bz- recep tor m a p p i n g wi th S P E T using Ro 16-0154 (e thyl -7- iodo-5 ,6- d ihyd ro -5 -me thy l -6 -oxo -4H- imidazo [1 ,5 - a ] [1 ,4 ] -benzo - zep ine -3 -ca rboxy la t e ; Iomazeni l ) , an 123I-labelled der iv- a t ive o f F l u m a z e n i l in pa t ien t s wi th pa r t i a l epi lepsy. Here i t was o f pa r t i cu l a r in teres t whe the r or no t there is a decrease o f r ecep to r b ind ing de tec tab le in the epi lep- tic focus. F u r t h e r m o r e , we w a n t e d to c o m p a r e the dis tr i - b u t i o n p a t t e r n o f ~23I-Iomazenil wi th the m a p o f cere- b ra l b l o o d f low o b t a i n e d by 99mTc-HMPAO S P E T in the same pat ien ts .

Patients and methods

Patients. Some 15 patients (5 female, 10 male) aged 20-65 years (mean 33 years) took part in the study. All patients had partial epilepsy with lateralised seizures, 13 had complex partial seizures, 2 had simple partial seizures. Secondary generalization was ob- served in 14 patients. Seizure frequency ranged from 5 times a year to 3 times weekly, and history of the disease between 1 and 11 years. Three patients (1 female, 2 male) with complex partial seizures had abnormal CT and/or MRI scans and were excluded subsequently form this study. The other 12 patients showed no morphological changes on CT or MRI.

In the case of a normal conventional EEG, recording included provocation methods like hyperventilation, photo-stimulation and sleep deprivation. All patients underwent at least 3 EEG record- ings.

At least 2 months before the study no anti-epileptic drugs inter- acting with the GABA/Bz-receptor were given. Two patients re- ceived no therapy, ten patients were on carbamazepine medication only, one patient was on phenytoin medication, and the remaining two were on a combined carbamazepine/phenytoin medication. The average dialy dosage of carbamazepine was 750 mg (600- 1200 rag); the daily dosage of phenytoin was 250-350 nag.

Methods. All patients underwent interictal SPET studies with 99mTc-HMPAO and 123I-Iomazenil during a period of between 2 and 7 days. The substances were injected slowly i.v., with eyes open, in a poorly lightened, quiet room.

Acquisition of the SPET study was started 15-30min after administration of 500 MBq 99mTc-HMPAO.

After intravenous injection of 185 MBq lz3I-Iomazenil in all patients, two SPET studies at 5 min and 2 h p.i. were performed. In three patients an additional SPET study 24 h p.i. was performed. The choice of acquisition times was based on PET studies of the pharmacokinetics of 11C-Flumazenil (Samson et al. 1985) and on animal data based on 123I-Iomazenil (Beer et al. 1990).

We measured in all studies 64 projections in a 64 x 64 matrix over 360 °, 20 s for each. This means a revolution time of about 25 rain. The gamma-camera used was a 400 ACT from General Electric with a STAR computer and standard SPET software. With a LEAP collimator and a 20% energy window, we achieved count rates of about 3000 cps for the 99mTc-HMPAO studies, 1000 cps for the 5 rain p.i. - and 700 cps for the 2 h p.i. - 123I-Iomazenil studies.

Reconstruction was carried out in transversal, coronal and sa- gittal slices. It was done with a Butterworth filter of order 4, a cut-off frequency of 0.4 cm ~ and a slice thickness of 12 mm. Slices were reorientated, so that the transversal ones were parallel to the meato-orbital line. We used a linear attenuation coefficient of 0.12 cm -~ for the ~23I-Iomazenil studies and of 0.13 cm -~ for

the 99mTc-HMPAO studies. Uniformity correction was done using a 30 million counts flood-field for each of the two isotopes. The spatial resolution of our configuration is between 17 and 20 ram, which was determined by measuring a Jaszcak phantom under conditions similar to those in the patient studies. To optimize the superimposition of corresponding slices from 99mTc-HMPAO and 123I-Iomazenil studies we stepped through the stack of slices using a three-dimensional programme and measured the extension of the slices in the desired directions. The data analysis was primary performed by qualitative visual interpretation. The images were interpreted independently by two nuclear medicine physicians and one neurologist without knowledge of the other results.

Afterwards, the ratio of focus/contralateral side was individual- ly calculated from coronal slices using small regions of interest (ROI). In addition, in 5 patients quantification of the 2 h study with 42 cortical ROI and a reference region in the white matter was carried out from transversal slices (Fig. 1).

The study was approved by the local ethical committee. In- formed written consent was obtained from each patient.

Results

In the ear ly 123I-Iomazenil s tudy, in a d d i t i o n to the very high u p t a k e in the ce rebra l grey m a t t e r and high u p t a k e in the cerebel lum, m o d e r a t e u p t a k e was obse rved in the a rea o f the basa l gang l ia in the whi te m a t t e r and the n a s o p h a r y n g e a l space. This indica tes a s ignif icant non- specific b ind ing a n d the c o n t r i b u t i o n o f the t racer in the b l o o d p o o l in this phase .

The s tudy 2 h p.i. showed a s imi lar d i s t r i bu t ion o f ac t iv i ty in the grey ma t t e r . In the cerebel lum, basa l gan- glia a n d whi te ma t t e r , there was a m a r k e d decrease o f ac t iv i ty in c o m p a r i s o n wi th the ear l ier s tudy (Figs. 2, 3). A t 24 h p.i. there was no s ignif icant ac t iv i ty left in the bra in , bu t despi te b lock ing wi th sod ium perch lora te , there was s ignif icant u p t a k e o f ac t iv i ty in the t hy ro id and s tomach . This indica tes tha t the l igand is largely d e i o d i n a t e d du r ing this t ime.

The three pa t ien t s exc luded f rom the s tudy because o f abno rma l i t i e s in CT and M R I studies h a d f indings c lear ly d is t inc t f r o m the o the r pa t ien ts . The first pa t i en t was a 25-year -o ld m a n s t rong ly suspected o f suffer ing f rom c omple x par t ia l , s econda ry genera l ized epilepsy. He had whi te m a t t e r lesions on M R i m a g e , bu t no abno r - mal i t ies on E E G , 99mTc-HMPAO SPET or 12aI-Ioma- zenil SPET scans. Af t e rwards , ( add i t iona l? ) mul t ip le sclerosis was conf i rmed . The second pa t ien t , a 65-year- o ld w o m a n wi th complex pa r t i a l seizures, showed on CT a n d M R I scans a m a r k e d ce rebra l a t r o p h y especia l ly o f the left t e m p o r a l lobe. As expected, a decrease o f ac t iv i ty on 99mTc-HMPAO S P E T a n d 123I-Iomazenil

SPET scans was found. The th i rd pa t ien t , an 38-year -o ld m a n wi th c omple x pa r t i a l seizures, had a gliosis in the left t e m p o r a l lobe con f i rmed by M R I examina t ion . E E G s tudy showed a c o r r e s p o n d i n g spike and wave focus, and a focal decrease o f ac t iv i ty on 99mTc-HMPAO SPET scan was observed. In teres t ingly , there was no s ignif icant decrease o f t racer a c c u m u l a t i o n in bo th 123I-Iomazenil s tudies (5 min and 2 h p.i.).

113

R

2

4

6

8 Fig. 1. Regions of interest (ROI) used for regional quantification of the 123I-Iomazenil single photon emission tomography (SPET)

i ~ !i i ̧ ̧ i;

; ! i ̧ ;

~AT~C B 3

!~; i ̧ ~ ! i i ; a . 8 a a o ~ a~ t

2 t l e

Fig. 2. Corresponding sagittal (top row) and transversal (bottom row) slices from left to right: technetium 99m hexamethyl propylene amine oxime (99mTC- HMPAO) SPET, 123I-Iomazenil SPET 5 min p.i. and 123I- Iomazenil SPET 2 h p.i. The ratios occipital cortex/cerebellum (1.06, 1.34, 1.86) and occipital cortex/basal ganglia (1.39, 2.06, 3.83) are calculated with the ROI technique. Decrease of activity is seen compared with the occipital cortex in the cerebellum and basal ganglia/white matter from the early to the delayed ~23I- Iomazenil SPET image

114

All 12 patients without detectable morphological le- sions had focal abnormalities documented with at least one method (EEG, 99mTc-HMPAO or az3I-Iomazenil SPET) (Table 1). Eight patients demonstrated focal ab- normalities detectable on EEG scan (spikes and waves, delta focus).

,.5 i !

I

1

• ~o median

o •

I • m e d i a n

1 e~ median

T c - 9 9 m - H MPAO 1-123-1omazeni l 1-123-1omazeni l 5 rain p.i. 2 h p. i.

Fig. 3. Ratio of activity in occipital cortex/cerebellum in 99mTc- HMPAO SPET study, 123I-Iomazenil SPET study 5 rain and 2 h p.i. (n = 5). There is an evident increase of this ratio in all patients when comparing early and delayed 123I-Iomazenil SPET study

Table 1. Details of study population

A focal decrease of regional cerebral blood flow was observed in 8 patients involving the cortical area as ex- pected by clinical and EEG findings. One patient (no. 8) with sensible jacksonian seizures and no pathological EEG finding had a focus on the 99mTc-HMPAO SPET study on the side contralateral to that clinically expected. Another subject (no. 3) had multiple, slightly pro- nounced loci on the 99mTc-HMPAO SPET scan, but normal E E G and 123I-Iomazenil studies. In two patients we observed no pathological findings with 99mTC- H M P A O SPET. With 123I-Iomazenil SPET 2 h p.i., sev- en foci were observed concordant with the 99mTC- H MP A O studies (Fig. 4). In addition three of these pa- tients (nos. 1, 2, 4) showed an increase of activity in cortical areas near the focus (Fig. 5). One patient (no. 12) with simple partial seizures accompanied by dis- turbance of the right visual field exhibited a decrease of activity in the left occipital cortex in the study 2 h p.i. This patient had a EEG focus left occipito-parietal, but no abnormalities on the 99mTc-HMPAO SPET study. In the patient with sensible jacksonian seizures on the 123I-Iomazenil study a finding was observed very similar to the 99mTc-HMPAO SPET result and discor- dant to clinical examination. Three subjects had no ab- normalities on 12aI-Iomazenil SPET scan. Generally the foci in the 123I-Iomazenil study seemed to be more ex-

No. Age/sex Clinical diagnosis Electro (kind of epilepsy) encephalogram

(focus)

99mTc_HMPAO 123 I-Iomazenil

5 min 120 min

6

7

8

9

10

11

12

1 25/F Complex-partial, Left temporal no generalization

2 29/M Complex-partial, No finding secondary generalized

3 39/M Complex-partial, No finding secondary generalized

4 54/M Complex-partial, Left temporal secondary generalized

5 21/M Complex-partial, Right temporal secondary generalized

23/F Complex-partial, Dysrhythmia, left secondary generalized accentuation

23/M Complex-partial, Right temporal secondary generalized

30/M Simple-partial, No finding secondary generalized

41/F Complex-partial, No finding secondary generalized

20/M Complex-partial, Left front•- secondary generalized temporal

36/F Complex-partial, Left temporal secondary generalized

20/M Simple-partial, Left occipito- secondary generalized parietal

Left temporal No finding Left temporal reduction, left parietal increase

Left temporal Left temporal Left temporal reduction, fronto-parietal increase

Right temporal, No finding No finding left parietal and frontal

Left temporal Left temporal Left temporal reduction, fronto-temporal increase

Right temporal Right temporal Right temporal

No finding No finding No finding

Right temporal, Right temporal Right temporal left parietal

Left fronto-temporal Left front•- Left fronto-temporal temporal

Left temporal (Left temporal) No finding

Left fronto-temporal, Left front•- Left fronto-temporal right temporal temporal

Left temporal Left temporal Left temporal

No finding No finding Left occipito-parietal

115

Fig. 4. A 36-year-old woman with complex partial seizures, has EEG delta focus left temporal. 99mTc-HMPAO SPET (top row) and 123I-Iomazenil SPET 2 h p.i. (bottom row) show decrease of activity in the left temporal cortex

Fig. 5. A 25-year-old woman with complex partial seizures show on EEG spikes and waves under photo- stimulation left temporal, on 99mTC-

HMPAO SPET decrease of activity left temporal. With 123I-Iomazenil SPET 2 h p.i., in addition to the decrease of activity left temporal, increase of activity in the left temporo-parietal cortex is seen

116

EEG

Tc-O9m-HMPAO 1-123-1omazenil

Fig. 6. Concordance of findings on EEG, 99mTc-HMPAO SPET and 123I-Iomazenil SPET 2 h p.i.

1.2

1.1 •

1 :

0 . 9 .......... i * * ...... m e d i a n

0 . 8

%

0 . 7

Fig. 7. Relative uptake of 123I-Iomazenil on SPET scan 2 h p.i. (focus/contralateral side) in the 12 patients without abnormalities on magnetic resonance imaging or computed tomography study

tended compared with those seen on 99mTc-HMPAO study. Figure 6 shows the concordance of the findings with all three methods.

The individual quantification with ROI focus/con- tralateral side on the ~23I-Iomazenil SPET scan 2 h p.i. showed a median ratio of activity of 0.90 (range 0.76- 1.11) (Fig. 7). If there was no visually detectable focus, a cortical region according to the findings on 99mTc- HMPAO or EEG study was chosen.

By quantification with multiple regions obtained in 5 patients, the foci were detected. Compared with indi- vidual quantification the ratio of focus/contralateral side was lower (Fig. 8).

D i s c u s s i o n

EEG recording methods are very specific for partial epi- lepsy but have a poor ability to locate the focus exactly (St•fan et al. 1987). Other modalities, using the electric activity of the brain-like magnetoencephalography (SQUID), are still in the experimental stage (Modena et al. 1982; Vieth 1987).

The data about the frequency of positive findings

Fig. 8. In regional quantification, significant decrease of activity compared with the contralateral side only in left temporal cortex (ROI 4, top left, focus/contralateral : 0.86)

with CT and MRI in partial epilepsy vary widely in the literature. For CT most authors report positive find- ings in about 3 0 % 4 0 % of the patients (Kuhl et al. 1980; Engel et al. 1982; Wyler and Bolender 1983). The frequency of positive MRI findings is reported by most authors to be about 70%-80% (Sussman et al. 1984; Theodore et al. 1985; Sperling et al. 1986; Stefan et al. 1987; Wienhard et al. 1989). In all studies comparing CT, MRI and 18F-FDG PET, the sensitivity of 18F- F D G PET was significantly higher than that of the other imaging modalities (Kuhl et al. 1980; Theodore et al. 1985; Sperling et al. 1986; Stefan et al. 1987). Most au- thors agree, that in the case of positive CT or MRI scans the diagnostic information is better compared with that from 18F-FDG PET study, because CT and MRI provide accurate anatomical data (Wienhard et al. 1989). So patients with normal MRI and CT scans, like our patient group, will best benefit from functional im- aging methods like PET and SPET.

l aF-FDG PET is, like the flow-marker 99mTc- HMPAO for SPET, not specific for epileptic changes of the brain. Bz-receptor mapping with PET and SPET tracers would perhaps allow a more specific detection of epileptic foci. Because PET is an expensive, not gener- ally available method, a SPET tracer would be of greater clinical relevance.

Savic et al. (1988) investigated 10 patients with the Bz-receptor ligand l aC-Flumazenil and PET. They, like in our study, examined only patients with partial epilep- sy who had normal CT and MRI scans and were re- stricted to carbamazepine and/or phenytoin medication. They reported a positive result in all cases.

A decrease of Bz-receptor binding in epileptic foci was found in 9 of our 12 patients. We therefore achieved a significantly lower sensitivity although we had a very

117

similar patient group. The difference is perhaps due to the low spatial resolution of our SPET system (17- 20 mm) compared with a modern PET system (5- 10 ram).

Doubling the activity and revolution time would have allowed us to use a reconstruction filter with a higher cut-off frequency (0.5 cm-~), which would have improved the resolution about 3 ram. However, we were restricted by our local ethical committee to a maximum dose of 185 MBq 1/3I-Iomazenil, and we chose not to use a longer revolution time to avoid uncontrolled side- effects caused by significant wash-out during acquisition.

With a multiheaded dedicated SPET system a maxi- mum spatial resolution of 7 mm is achievable, which is comparable with PET systems (Kimura et al. 1990). These dedicated systems are currently, like PET, also very expensive and only available in a few centers. It seems more useful to examine the clinical relevance of a tracer like 123I-Iomazenil with a routine SPET system.

Samson et al. (1985) and Shinotoh et al. (1986) ob- served in PET studies with 11C-Flumazenil (20-30 rain p.i.) a distribution of activity very close to that of Bz- receptors in the human brain known from post-mortem studies (Baestrup etal. 1977; Richards and M6hler 1984).

The best imaging time for 123I-Iomazenil is still under discussion (Hoell et al. 1989; Beer et al. 1990; Van Isselt et al. 1990; Verhoeff et al. 1990). A recent study on the pharmacokinetics of 123I-Iomazenil in healthy volunteers showed a significantly later Bma x and much slower washout in comparison with ~C-Flumazenil (Verhoeff et al. 1990). These authors found a very good correlation of activity distribution to the known Bz-re- ceptor density of human brain between 60 and 210 min p.i. This difference compared with l~C-Flumazenil can be explained by the higher lipophilicity of 1/3I-Iomazen- il. It is known that the kinetics of iodinated compounds is usually slower than that of fluorinated compounds (Katzenellenbogen et al. 1980; Liu et al. 1987).

Thus, 2 h p.i. should be a suitable time for mapping of the receptor distribution. Indeed, the activity distribu- tion in our 2 h study (Fig. 2) correlates very well with the results obtained at 30 rain p.i. in the PET studies mentioned above. This qualitative visual impression was confirmed by the relative quantification obtained in five typical patients (Fig. 1). The early study, in which the activity distribution reflects a mixture of specific binding and unspecific effects, did not provide any additional information.

In comparison with our study, Van Isselt et al. (1990) achieved slightly better results (15/17 positive) starting data acquisition at 30 rain p.i. of ~23I-Iomazenil. They investigated patients with drug-resistant partial epilepsy. Because of their drug resistance, a more severe stage of disease in their patient group is probable. This may explain the higher sensitivity. In their study they com- pared 123I-Iomazenil SPET with tSF-FDG PET and found no significant difference. This is very similar to

our findings, that detection of epileptic foci was not su- perior with 123I-Iomazenil compared with the unspecific flow-marker 99mTc-HMPAO.

The influence of blood flow on the distribution of lSF-FDG is known (Sokoloffet al. 1977). Further evalu- ation is needed to obtain accurate data about the effect of blood flow on the distribution of ~23I-Iomazenil.

An influence on the distribution of 99mTc-HMPAO or 123I-Iomazenil caused by the carbamazepine and phenytoin medication of our patients is not probable. Both drugs in therapeutic dosage do not have any known effect on cerebral blood flow. They are membrane stabi- lizing agents with no evidence of interference with the Bz-receptor (Goodman Gilman et al. 1985).

The relevance of the increased binding of activity in cortical areas adjacent to the focus observed in three of our patients still remains unclear. This finding may indicate the existence of an inhibitory reaction in the cortex surrounding the focus (Elger and Speckmann 1983).

In summary, it was possible to demonstrate that 123I- Iomazenil is a suitable ligand for Bz-receptor mapping with SPET. In most of our patients, an interictal de- crease of receptor binding in the epileptic focus was de- tectable. Nevertheless, the clinical relevance of this inves- tigation with a conventional SPET system for location of an epileptic focus seems questionable.

Acknowledgement. We would like to thank PD Dr. A. Bekier, St. Gallen, who has developed the regional quantification procedure for Fig. 1.

References

Baestrup C, Albrechtsen R, Squires RF (1977) High densities of benzodiazepine receptors in human cortical areas. Nature 269:702 704

Beer HF, Bl~iuenstein PA, Hasler PH, Delaloye B, Riccabona G, Bangerl I, Hunkeler W, Bonetti EP, Pieri L, Richards JG, Schu- biger PA (1990) In vitro and in vivo evaluation of iodine-123- Ro 16-0154: a new imaging agent for SPECT investigations of benzodiazepine receptors. J Nucl Med 31 : 1007-1014

Bonte FJ, Stokely EM, Devous MD, Homan RW (1983) Single photon tomographic study of regional cerebral blood flow in epilepsy. Arch Neurol 40:267-270

Elger CE, Speckmann EJ (1983) Penicillin induced foci in the motor cortex: vertical inhibition. Electroencephalogr Clin Neurophy- siol 56 : 604-622

Ell PJ, Costa DC (1989) Role of routine functional brain imaging in temporal lobe epilepsy (letter). Lancet I : 959

Engel J, Brown WJ, Kuhl DE, Phelps ME, Mazziotta JC, Crandall PH (1982) Pathological findings underlying focal temporal lobe hypometabolism in partial epilepsy. Ann Neurol 12:518- 528

Goodman Gilman A, Goodman LS, Rall TW, Murad F (1985) The pharmacological basis of therapeutics. Macmillan, New York

Hoell K, Deisenhammer E, Dauth J, Carmann H, Schubiger PA (1989) Imaging benzodiazepine receptors in the human brain

118

by single photon emission computed tomography (SPECT). Nucl Med Biol 16:759-763

Katzenellenbogen JA, Carlson KE, Heiman DF, Goswami R (1980) Receptor-binding radiopharmaceuticals for imaging breast tumors: estrogen-receptor interactions and selectivity of tissue uptake of halogenated estrogen analogs. J Nucl Med 21 : 550-558

Kimura K, Hashikawa K, Etani H, Uehara A, Kozuka T, Moriw- aki H, Isaka Y, Matsumoto M, Kamada T, Moriyema H, Tabu- hi H (1990) A new apparatus for brain imaging: four-head rotating gamma camera single-photon emission computed to- mograph. J Nucl Med 31:603-609

Kuhl DE, Engel J, Phelps ME, Selin C (1980) Epileptic patterns of local cerebral metabolism and perfusion in humans deter- mined by emission computed tomography of ~SFDG and ~3NHa. Ann Neurol 8:348-360

Kuikka JT, Lfinsimies E, Kuikka EO (1990) Quantitative analysis of 99mTc-HMPAO kinetics in epilepsy. Eur J Nucl Med 16:275- 283

Lieb JP, Engel J, Gevins AS, Crandall PH (1981) Surface and deep EEG correlates of surgical outcome in temporal lobe epi- lepsy. Epilepsia 22:515-538

Liu BL, Kung HF, Billings J, Blau M (1987) Comparison of (82Br)4-bromoantipyrine and (125i)4_iodoantipyrine: the kinet- ics of exchange reaction and biodistribution in rats. Nucl Med Biol 14:69-74

Modena I, Ricci GB, Barbarena S, Leoni R, Romani GL, Carelli P (1982) Biomagnetic measurements of spontaneous brain ac- tivity in epileptic patients. Electroencephalogr Clin Neurophy- siol 54: 622-628

Olsen RW (1981) The GABA postsynaptic membrane receptor- ionophore complex. Mol Cell Biochem 39:261-279

Perry TL, Hansen S (1981) Amino acid abnormalities in epilepto- genic foci. Neurology 31:872-876

Ravagnati L (1987) Stereotactic surgery for epilepsy. In: Wieser HG, Elger CE (eds) Presurgical evaluation of epileptics. Sprin- ger, Berlin Heidelberg New York, pp 350-363

Ribak CE (1985) Axon terminals of GABAergic chandelier cells are lost at epileptic foci. Brain Res 326 : 251-260

Richards JG, M6hler H (1984) Benzodiazepine receptors. Neu- ropharmacology 23:233-242

Samson Y, Hantraye P, Baron JC, Soussaline F, Comar D, Maziere M (1985) Kinetics and displacement of (11C)Ro 15-1788, a ben- zodiazepine antagonist, studied in human brain in vivo by posi- tron tomography. Eur J Pharmacol 110: 247-251

Savic I, Persson A, Roland P, Pauli S, Sedvall G, Widen L (1988) In vivo demonstration of reduced benzodiazepine receptor binding in human epileptic foci. Lancet I:863-866

Sherwin A, Matthew E, Blain M, Guevremont D (1986) Benzodia-

zepine receptor binding is not altered in human epileptogenic cortical foci. Neurology 36:1380-1382

Sherwin A, Robitaille Y, Quesney F, Olivier A, Villemure J, Leb- lanc R, Feindel W, Andermann E, Gotman J, Andermann F, Ethier R, Kish S (1988) Excitatory amino acids are elevated in human epileptic cerebral cortex. Neurology 38:920-923

Shinotoh H, Yamasaki T, Inoue O, Itoh T, Suzuki K, Hashimoto K, Tateno Y, Ikehira H (1986) Visualisation of specific binding sites of benzodiazepine in human brain. J Nucl Med 27:1593- 1599

Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M (1977) The (I'~C)- deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897- 961

Sperling MR, Wilson G, Engel J, Babb TL, Phelps M, Bradley W (1986) Magnetic resonance imaging in intractable partial epilepsy: correlative studies. Ann Neurol 20: 57-62

Stefan H, Pawlik G, B6cher-Schwarz HG, Biersack HJ, Burr W, Penin H, Heiss WD (51987) Functional and morphological ab- normalities in temporal lobe epilepsy: a comparison of interictal and ictal EEG, CT, MRI, SPECT and PET. J Neurol 234: 377- 384

Sussman MN, Scanlon M, Garfinkle W, Callahan M, Barry E, Katz RI, Harner N (1984) Magnetic resonance imaging in tem- poral lobe epilepsy: comparison with EEG and computerized tomography. Epilepsia 25 : 650-651

Theodore WH, Dorwart R, Holmes M, Porter RJ, DiChiro G (1985) Neuroimaging in epilepsy: comparison of PET, CT and MRI. Neurology 25s: 135

Van Isselt JW, van Bentum AME, van Huffelen AC, van Rijk PP (1990) The detection of epileptic foci with 123I-Iomazenil. Eur J Nucl Med 16 : 409

Verhoeff NPLG, Van Royen EA, Ell P J, Costa DC, Hasler PH, Schubiger PA (1990) Benzodiazepine receptor density in human brain measured in vivo with ~23I-Iomazenil SPECT. In: Ell PJ, Van der Schoot JB, Schmidt HAE (eds) Quantitative analy- sis in imaging and function. European Nuclear Medicine Con- gress, Amsterdam 1990. Schattauer Verlag, Stuttgart (in press)

Vieth J (1987) Magnetoencephalography and epilepsy. In: Wieser HG, Elger CE (eds) Presurgical evaluation of epileptics. Sprin- ger, Berlin Heidelberg New York, pp 118-128

Wienhard K, Wagner R, Heiss WD (1989) PET Grundlagen und Anwendungen der Positronen-Emissions-Tomographie. Sprin- ger, Berlin Heidelberg New York

Wyler AR, Bolender N (1983) Preoperative CT diagnosis of mesial temporal sclerosis for surgical treatment. Ann Neurol 13:59 64