- Email: [email protected]
Epilepsia partialis continua studied by PET
44
Epilepsy Res.. 9 (1991) 44-48 Eisevier
EPIRES 00401
Epilepsia partialis continua studied by PET
Marketa HajeklT2, Angelo Antonini2, Klaus L. Leenders’~2 and Heinz-Gregor ‘Neurology Depanmenl,
Wieserl
University Hospital, Ziirich and ‘Paul Scherrer Institute, ViIligen (Switzerland)
(Received 22 November 1990; accepted 12 December 1990) Key work
[18F]Fluordeoxyglucose (FDG)-PET; Epilepsia partialis continua; ‘Chronic encephalitis’ of Rasmussen
We report an [‘8F]fluordeoxyglucose (FDG)-PET study performed in an 1l-year-old girl with a 5-month history of epilepsia partialis continua (epc). Visual inspection of PET images showed a h~~eta~lic focus in the right central cortex and in the ipsilateral thalamus, which was confirmed by the absolute values of regional cerebral glucose metabolism (rCMRGlu). The thalamic hypermetabolism provides evidence for an involvement of thalamic nuclei in this ictally epileptic process, The scalp EEG revealed a theta-delta and sharp wave focus in the right Rolandic cortex at the same location as the hypermetabolic zone seen in PET. Simultaneously recorded EMG of the left tibialis anterior muscle showed regular jerks, time-locked to the sharp waves at the right central region, and myoclonic ‘storms’ during focal motor seizures. The results of the brain biopsy and the child8 clinical course led us to a diagnosis of ‘chronic encephalitis’ of Rasmussen.
INTRODUCTXON Epilepsia partialis continua (epc) is a rare syndrome with varying etiology and complex pathophysiological mechanisms. We report an f’“F]fluordeoxyglucose (FDG)-PET study in an llyear-old girl with a 5-month history of epc. The clinical history of our patient fulfills the criteria of the ‘chronic encephalitis’ syndrome, first described by Rasmussen’. CLINICAL HISTORY Family history, gestation, birth and psychomotor development of the patient were unremarkable Correspondence to: H.G. Wieser, Neurology Department, University Hospital, Frauenklinikstr. 26, CH-8091 Zurich (Switzerland). Tel.: +41-l-255 55 30; Fax: +41-l-255 44 29.
0920-1211/91/$03.50@ 1991 Elsevier Science Publishers B.V.
until the age of 9% years (February 1989). She then experienced her first seizure which started as a simple partial motor seizure of the left hand but progressed after a few seconds into a hemiconvulsive attack, with eyes and head turning to the ieft. Secondary generalization then occurred. Antiepileptic drug (AED) treatment with 800 mg carbamazepine resulted in a seizure-free period of 3 months. Thereafter, frequent simple partial motor seizures of the left hand occurred, also involving the left hand and arm at a later stage. From June 1989 onwards, Todd’s palsy of the left arm with increasing duration was observed after each seizure. In November 1989, an increase in seizure frequency occurred, leading to an epc of the left upper limb. One month later the jerks also spread to the left foot and progressively involved the entire leg. The myoclonic jerks were present during wakefulness and, with a reduced frequency, also
45
during sleep. AED treatment was modified in January 1990 to 40 mg Ospolot, 900 rng carbamazepine and 30 mg Clobazam, but the epileptic condition did not ameliorate. The patient was referred to our department in March 1990. She presented herself as a fully oriented, alert girl with severe paresis of the left arm and moderate paresis of the left leg. Irregular synchronous and as~chronous myocloni of the left extremities were present. Left focal motor seizures occurred without alteration of consciousness, with an interseizure interval of 5-10 min. DIAGNOSTIC
TESTS
In February 1989, after the first generalized seizure, CT scans with and without enhancement were normal. The EEG at this time showed a
slight right fronto-parietal theta focus. Ten months later, in December 1989, the native CT scan showed a hypodensity in the right centro-parietal region (Fig. l), and the n-weighted MRI images showed a hyperdensity (Tl-weighted images: hypodensity) of the right lobulus paracentralis and in the right Rolandic area (Fig. 1). In December 1989 a stereotactic brain biopsy of the right central area was done. Light and electron microscopy, and immunomorphological investigations, did not show evidence of a tumor or an inflammatory process. The only pathological finding was slight gliosis. Viral studies of blood and CSF were normal. A scalp EEG using the 10120 system combined with an EMG (using a superficial electrode attached above the left tibialis anterior muscle) was performed in March 1990 with the aim of a quanti-
Fig. 1. Top: Native CT scan performed 10 months after the first seizure. The scans (3 subsequent axial planes) display a hypodense zone in the right central area. Bottom: axial (left) and coronal (right) TICweighted MRI images (TR 3oo0, TE SO), obtained 12 months after the first seizure, show a hyperintense lesion in the right Rolandic cortex.
46 titative analysis. The recordings were performed using a Nihon Kohden 21 channel apparatus and were evaluated visually, as well as by off-line quantitative analysis. The scalp EEG showed a slowed background activity and a continuous theta-delta focus with sharp waves in the right central Rolandic region (Fig. 2). The simultaneously recorded EMG showed regular jerks which were time-locked to the EEG sharp waves of the right central area with a latency between 30 and 50 msec. There were also myoclonic ‘storms’ accompanied by the recruitment of rhythmic epileptic discharges and/or delta waves in the EEG (Fig. 2). POSITRON EMISSION TOMOGRAPHY An FDG-PET scan was performed using a CT1 (Siemens) tomograph (type 933). This scanner measures 7 contiguous planes (width 8 mm) simultaneously with an in-plane transaxial resolution of 8 mm. A dynamic scan procedure (48 min) was performed with the field of view parallel to and from 10 to 66 mm above the orbitomeatal line (OM line). In addition a static emission scan (5 min) was performed in a second position displacing the field of view 40 mm in the cranial direction.
Before tracer administration a transmission scan was performed in both positions using an external ge~~i~~gallium~ ring source. Through an arm vein 3.33 mCi [i*F]FDG was slowly infused over 3 min using an infusion pump. At the same time a series of consecutive scans with increasing scan duration was started. Twenty-three arterial blood samples were collected from an indwelling radial artery catheter (Teflon Cl.8mm) for determination of the blood plasma radioactivity and glucose con~entration. After image reconstruction 78 regions of interest (ROIs) were determined on a visual display unit according to a ~mi-standardized interactive method. Each ROI was characterized by its distance in mm from the OM line. The regional glucose utilization (rCMRGlu) was calculated according to an ‘autoradiographic’ method6 in all 78 ROIs. The autoradiographic method was chosen in order to obtain comparable values for the entire brain. Visual inspection of the PET images revealed 2 striking focal abnormalities located within the cerebral cortex (Fig. 3). In the plane cutting the brain 94 mm above the OM line, a marked hypometabolism was present in the right central area at the location of the hyperintense abnormality shown
Fig. 2. Bipolar 20 channel EEG and EMG recordings of the Ieft tibialis anterior muscle with regular muscle jerks (left) time-locked with a latency of 30-50 mseec to the sharp waves of the right central region (phase reversal is indicated by stars) and with a focal motor seizure accompanied by a myocfonic ‘storm’ (middle). The arrow indicates right central posterior Iocated trains of I%ec waves correlating with the EMG. EEG map (right) shows the theta-delta focus right central. The map displays the ratio of theta+delt~~p~a+beta power.
47
Fig. 3. Visual representation of [18F]FDG uptake in the patient’s brain. The left PET image 94 mm above the orbito-meatal line shows a hypometabolic zone in the right central area close to the midline, which, according to the brain biopsy, histopathclogically consists of gliosis. A distinct hypermetabolic focus is located in the central lateral cortex. The right PET image (62 mm above the orbito-meatal line) shows the hypermetabolic right thalamus.
by MRI. Centrally and laterally there was a prominent and clearly delineated hypermetabolic area. This focal zone of increased hypermetabolism closely corresponded to the site of the sharp wave focus. All other parts of the cortex did not present striking asymmetries of glucose uptake. Within the subcortical structures the right thalamus (62 mm bove the OM-line) showed a marked hypermetabolism (Fig. 3). Quantitative determinations of rCMRGlu values were made in 78 ROIs. Sideto-side differences of homologous ROIs were calculated and the percentage of asymmetry was expressed as [(ROI right-R01
left)/ROI left] x 100.
In Table I the rCMRGlu values of various brain regions and the percentage of side-to-side differences are given. According to the visual analysis, the maximal increase of rCMRGlu was within the marked hypermetabolic areas in the right central cortex and in the right thalamus. The side-to-side difference was +22% in the central cortex and +30% in the thalamus. The percentage differences of other brain regions were between -3.5% and i-9%. The mean of all ROIs was 23.1 + 8.4
~mol(lO0 ml)-’ min-’ in the right hemisphere and 22.5 + 7.9ymol(lOO ml)-’ min-’ in the left. DISCUSSION Reviewing the literature, we found 3 PET studies dealing with patients suffering from epclm3. In these studies one or more cortical or subcortical TABLE I Absolute rCMRglu values (firno (100 ml)-’ miti’) of some regions of interest (ROIs) and percentage differences of homologous ROIs [(ROI right-ROI 1eft)lROIleft] x 100 in right and left hemispheres Topography
Frontal cortex Rolandic cortex Parietal cortex Temporal cortex Occipital cortex Thalamus Caudatus Putamen White matter Cerebellum
rCMRGiu Right
Left
27.0 33.1 28.6 22.5 24.5 30.7 27.0 31.7 12.5 25.1
25.4 27.1 27.5 23.2 25.4 23.7 25.7 29.1 12.8 24.0
Asymmetry index (%) +6.6 +22.1 +4.0 -3.0 -3.5 +30.0 +5.0 +9.0 -2.3 +4.6
48 brain regions with increased metabolism were found. In no study, however, was as clear a relationship between the cortical focus and the ipsilatera1 thalamus reported as that found in our patient. Our findings of a strong involvement of the thalamus in an active epileptic process are in agreement with recently published interictal PET studies performed in patients suffering from temporal lobe (TL) epilepsy, in whom glucose uptake of the thalamus ipsilateral to the affected TL was significantly decreased4. The metabolic increase in both structures is presumably due to the reciprocal thalamocortical connections between the motor cortex and the ventrolateral and the intruding thalamic nuclei’. Despite higher thalamic asymmetries than those of the cortical focus, we assume a cortical origin of the epileptic discharges and a secondary activation of the thalamus. This hypothesis is supported by stereo-electroen~ephalographic (SEEG) recordings in another patient suffering from epc. Cortical discharges always preceded the thalamic ones, and stimulation of the ventral thalamic nuclei had no significant effect on the spontaneous myoclonic jerking”. From reports on the effects of thalamic surgery in such cases, however, no clear-cut conclusion can be drawn concerning a passive or an active role of the thalamus. In the latter case, the thalamus might sustain or even activate the myoclonus. REFERENCES 1 Cowan, J.M.A., Rothwell, J.C., Wise, R.J.S. and Marsden, CD., Electrophysiological and positron emission studies in a patient with cortical myoclonus, epilepsia partialis continua and motor epifepsy, .I. Neural. Neurosurg. Psychiatry, 49 (1986) 796-807. 2 Engel, J., Kuhl, D.E., Phelps, M.E., Rausch, R. and Nuwer, M., Local cerebral metabolism during partial seizures, Neurology, 33 (1983) 400-413. 3 Franck, G., Sadzot, B., Safmon, E., Depresseux, J.C., Gri-
sar, T., Peters, J.M., Guillaume, M., Quaglia, L., Deifore, D. and Lamotte, D., Regional cerebral blood flow and metabolic rates in human focal epilepsy and status epilepticus, Adv. Neural., 44 (1986) 935-948. 4 Henry,
T.R., Mazziotta, J.C., Engel, J., Christenson, P.D., Zhang, J.X., Phelps, M.E. and Kuhi, D.E., Quantifying interictal metabolic activity in human temporal lobe epilepsy, J. Cerebral Blood Flow Metab., 10 (1990) 748-757. 5 Jones, E.G., Wise, S.P. and Coulter, J.D., Differential thalamic relationship of sensory-motor and parietal cortical
Siegfried” performed, without postoperative clinical improvement, an extensive resection of the Rolandic and pre-Rolandic region in a patient with epc and Jacksonian fits caused by chronic encephalitis. A few weeks later, a stereotactic coagulation of the ventrolateral part of the thalamus was carried out in the same patient, after which the seizures stopped. In an extensive review on epc LGhler and Peters found 10 patients in whom a coagulation of the ventrolateral thalamic nucleus was performed. In only 4 patients, however, did such therapy result in clinical improvement’. In our patient, with a history and clinical signs suggesting ‘chronic en~eph~itis’ (Rasmussen) the histopathological findings of the brain biopsy did not reveal inflammatory signs or typical histopathological changes. The absence of inflammatory signs, however, does not rule out the presence of ‘chronic encephalitis’, as emphasized by Rasmussen in his recently published review of 48 operated patients with this syndrome’. The histopathologitally determined ‘gliosis’, located close to the hypermetabolic focus in PET, developed apparently during the ensuing course of the disease. ~nfortunateIy, we have no ultimate proof that this gliotic brain tissue or its boundaries trigger the epileptic discharges. In our patient resective surgery was discussed, but was refused by the girl’s parents despite further progressive deterioration with increased hemiparesis. fields in monkey, J. Camp. Neural., 183 (1979) 833-882. 6 Lammertsma, A.A., Brooks, D.J. and Frackowiak, R.S.J., Measurement of glucose utilisation with [“F]2-fluoro-2deoxy-D-glucose: a comparison of different analytical methods, .I. Cerebral Blood Flow Metab,, 7 (1987) 161-172. 7 Lohler, J. and Peters, U.H., Epilepsia partialis continua (Kozevnikov-Epilepsie), Fortschr. Neurol. Psychiatr., 42 (1974) 165-212. 8 Rasmussen, T., OIszewski, J. and Lloyd-Smith,
D., Focal seizures due to chronic localized encephalitis, Neurology, 8 (1958) 435-445. 9 Rasmussen, T. and Andermann, F., Update on the syndrome of ‘chronic encephalitis’ and epilepsy, Clin. Aspects Pediatr. Epilepsy, 56 (1989) 181-184. 10 Siegfried, J. and Bernoulli, C., Stereoelectroencephalographic exploration and cpilepsia partialis continua, Acra Neurochir. (SuppI.), 23 (1976) 183-191. 11 Wieser, H.G., Graf, H.P., Bernoulli, C. and Siegfried, J., Quantitative analysis of intracerebral recordings in epilepsia partialis continua, Electroencephalogr. Clin. Neurophysiol., 44 (1978) 14-22.
Epilepsy Res.. 9 (1991) 44-48 Eisevier
EPIRES 00401
Epilepsia partialis continua studied by PET
Marketa HajeklT2, Angelo Antonini2, Klaus L. Leenders’~2 and Heinz-Gregor ‘Neurology Depanmenl,
Wieserl
University Hospital, Ziirich and ‘Paul Scherrer Institute, ViIligen (Switzerland)
(Received 22 November 1990; accepted 12 December 1990) Key work
[18F]Fluordeoxyglucose (FDG)-PET; Epilepsia partialis continua; ‘Chronic encephalitis’ of Rasmussen
We report an [‘8F]fluordeoxyglucose (FDG)-PET study performed in an 1l-year-old girl with a 5-month history of epilepsia partialis continua (epc). Visual inspection of PET images showed a h~~eta~lic focus in the right central cortex and in the ipsilateral thalamus, which was confirmed by the absolute values of regional cerebral glucose metabolism (rCMRGlu). The thalamic hypermetabolism provides evidence for an involvement of thalamic nuclei in this ictally epileptic process, The scalp EEG revealed a theta-delta and sharp wave focus in the right Rolandic cortex at the same location as the hypermetabolic zone seen in PET. Simultaneously recorded EMG of the left tibialis anterior muscle showed regular jerks, time-locked to the sharp waves at the right central region, and myoclonic ‘storms’ during focal motor seizures. The results of the brain biopsy and the child8 clinical course led us to a diagnosis of ‘chronic encephalitis’ of Rasmussen.
INTRODUCTXON Epilepsia partialis continua (epc) is a rare syndrome with varying etiology and complex pathophysiological mechanisms. We report an f’“F]fluordeoxyglucose (FDG)-PET study in an llyear-old girl with a 5-month history of epc. The clinical history of our patient fulfills the criteria of the ‘chronic encephalitis’ syndrome, first described by Rasmussen’. CLINICAL HISTORY Family history, gestation, birth and psychomotor development of the patient were unremarkable Correspondence to: H.G. Wieser, Neurology Department, University Hospital, Frauenklinikstr. 26, CH-8091 Zurich (Switzerland). Tel.: +41-l-255 55 30; Fax: +41-l-255 44 29.
0920-1211/91/$03.50@ 1991 Elsevier Science Publishers B.V.
until the age of 9% years (February 1989). She then experienced her first seizure which started as a simple partial motor seizure of the left hand but progressed after a few seconds into a hemiconvulsive attack, with eyes and head turning to the ieft. Secondary generalization then occurred. Antiepileptic drug (AED) treatment with 800 mg carbamazepine resulted in a seizure-free period of 3 months. Thereafter, frequent simple partial motor seizures of the left hand occurred, also involving the left hand and arm at a later stage. From June 1989 onwards, Todd’s palsy of the left arm with increasing duration was observed after each seizure. In November 1989, an increase in seizure frequency occurred, leading to an epc of the left upper limb. One month later the jerks also spread to the left foot and progressively involved the entire leg. The myoclonic jerks were present during wakefulness and, with a reduced frequency, also
45
during sleep. AED treatment was modified in January 1990 to 40 mg Ospolot, 900 rng carbamazepine and 30 mg Clobazam, but the epileptic condition did not ameliorate. The patient was referred to our department in March 1990. She presented herself as a fully oriented, alert girl with severe paresis of the left arm and moderate paresis of the left leg. Irregular synchronous and as~chronous myocloni of the left extremities were present. Left focal motor seizures occurred without alteration of consciousness, with an interseizure interval of 5-10 min. DIAGNOSTIC
TESTS
In February 1989, after the first generalized seizure, CT scans with and without enhancement were normal. The EEG at this time showed a
slight right fronto-parietal theta focus. Ten months later, in December 1989, the native CT scan showed a hypodensity in the right centro-parietal region (Fig. l), and the n-weighted MRI images showed a hyperdensity (Tl-weighted images: hypodensity) of the right lobulus paracentralis and in the right Rolandic area (Fig. 1). In December 1989 a stereotactic brain biopsy of the right central area was done. Light and electron microscopy, and immunomorphological investigations, did not show evidence of a tumor or an inflammatory process. The only pathological finding was slight gliosis. Viral studies of blood and CSF were normal. A scalp EEG using the 10120 system combined with an EMG (using a superficial electrode attached above the left tibialis anterior muscle) was performed in March 1990 with the aim of a quanti-
Fig. 1. Top: Native CT scan performed 10 months after the first seizure. The scans (3 subsequent axial planes) display a hypodense zone in the right central area. Bottom: axial (left) and coronal (right) TICweighted MRI images (TR 3oo0, TE SO), obtained 12 months after the first seizure, show a hyperintense lesion in the right Rolandic cortex.
46 titative analysis. The recordings were performed using a Nihon Kohden 21 channel apparatus and were evaluated visually, as well as by off-line quantitative analysis. The scalp EEG showed a slowed background activity and a continuous theta-delta focus with sharp waves in the right central Rolandic region (Fig. 2). The simultaneously recorded EMG showed regular jerks which were time-locked to the EEG sharp waves of the right central area with a latency between 30 and 50 msec. There were also myoclonic ‘storms’ accompanied by the recruitment of rhythmic epileptic discharges and/or delta waves in the EEG (Fig. 2). POSITRON EMISSION TOMOGRAPHY An FDG-PET scan was performed using a CT1 (Siemens) tomograph (type 933). This scanner measures 7 contiguous planes (width 8 mm) simultaneously with an in-plane transaxial resolution of 8 mm. A dynamic scan procedure (48 min) was performed with the field of view parallel to and from 10 to 66 mm above the orbitomeatal line (OM line). In addition a static emission scan (5 min) was performed in a second position displacing the field of view 40 mm in the cranial direction.
Before tracer administration a transmission scan was performed in both positions using an external ge~~i~~gallium~ ring source. Through an arm vein 3.33 mCi [i*F]FDG was slowly infused over 3 min using an infusion pump. At the same time a series of consecutive scans with increasing scan duration was started. Twenty-three arterial blood samples were collected from an indwelling radial artery catheter (Teflon Cl.8mm) for determination of the blood plasma radioactivity and glucose con~entration. After image reconstruction 78 regions of interest (ROIs) were determined on a visual display unit according to a ~mi-standardized interactive method. Each ROI was characterized by its distance in mm from the OM line. The regional glucose utilization (rCMRGlu) was calculated according to an ‘autoradiographic’ method6 in all 78 ROIs. The autoradiographic method was chosen in order to obtain comparable values for the entire brain. Visual inspection of the PET images revealed 2 striking focal abnormalities located within the cerebral cortex (Fig. 3). In the plane cutting the brain 94 mm above the OM line, a marked hypometabolism was present in the right central area at the location of the hyperintense abnormality shown
Fig. 2. Bipolar 20 channel EEG and EMG recordings of the Ieft tibialis anterior muscle with regular muscle jerks (left) time-locked with a latency of 30-50 mseec to the sharp waves of the right central region (phase reversal is indicated by stars) and with a focal motor seizure accompanied by a myocfonic ‘storm’ (middle). The arrow indicates right central posterior Iocated trains of I%ec waves correlating with the EMG. EEG map (right) shows the theta-delta focus right central. The map displays the ratio of theta+delt~~p~a+beta power.
47
Fig. 3. Visual representation of [18F]FDG uptake in the patient’s brain. The left PET image 94 mm above the orbito-meatal line shows a hypometabolic zone in the right central area close to the midline, which, according to the brain biopsy, histopathclogically consists of gliosis. A distinct hypermetabolic focus is located in the central lateral cortex. The right PET image (62 mm above the orbito-meatal line) shows the hypermetabolic right thalamus.
by MRI. Centrally and laterally there was a prominent and clearly delineated hypermetabolic area. This focal zone of increased hypermetabolism closely corresponded to the site of the sharp wave focus. All other parts of the cortex did not present striking asymmetries of glucose uptake. Within the subcortical structures the right thalamus (62 mm bove the OM-line) showed a marked hypermetabolism (Fig. 3). Quantitative determinations of rCMRGlu values were made in 78 ROIs. Sideto-side differences of homologous ROIs were calculated and the percentage of asymmetry was expressed as [(ROI right-R01
left)/ROI left] x 100.
In Table I the rCMRGlu values of various brain regions and the percentage of side-to-side differences are given. According to the visual analysis, the maximal increase of rCMRGlu was within the marked hypermetabolic areas in the right central cortex and in the right thalamus. The side-to-side difference was +22% in the central cortex and +30% in the thalamus. The percentage differences of other brain regions were between -3.5% and i-9%. The mean of all ROIs was 23.1 + 8.4
~mol(lO0 ml)-’ min-’ in the right hemisphere and 22.5 + 7.9ymol(lOO ml)-’ min-’ in the left. DISCUSSION Reviewing the literature, we found 3 PET studies dealing with patients suffering from epclm3. In these studies one or more cortical or subcortical TABLE I Absolute rCMRglu values (firno (100 ml)-’ miti’) of some regions of interest (ROIs) and percentage differences of homologous ROIs [(ROI right-ROI 1eft)lROIleft] x 100 in right and left hemispheres Topography
Frontal cortex Rolandic cortex Parietal cortex Temporal cortex Occipital cortex Thalamus Caudatus Putamen White matter Cerebellum
rCMRGiu Right
Left
27.0 33.1 28.6 22.5 24.5 30.7 27.0 31.7 12.5 25.1
25.4 27.1 27.5 23.2 25.4 23.7 25.7 29.1 12.8 24.0
Asymmetry index (%) +6.6 +22.1 +4.0 -3.0 -3.5 +30.0 +5.0 +9.0 -2.3 +4.6
48 brain regions with increased metabolism were found. In no study, however, was as clear a relationship between the cortical focus and the ipsilatera1 thalamus reported as that found in our patient. Our findings of a strong involvement of the thalamus in an active epileptic process are in agreement with recently published interictal PET studies performed in patients suffering from temporal lobe (TL) epilepsy, in whom glucose uptake of the thalamus ipsilateral to the affected TL was significantly decreased4. The metabolic increase in both structures is presumably due to the reciprocal thalamocortical connections between the motor cortex and the ventrolateral and the intruding thalamic nuclei’. Despite higher thalamic asymmetries than those of the cortical focus, we assume a cortical origin of the epileptic discharges and a secondary activation of the thalamus. This hypothesis is supported by stereo-electroen~ephalographic (SEEG) recordings in another patient suffering from epc. Cortical discharges always preceded the thalamic ones, and stimulation of the ventral thalamic nuclei had no significant effect on the spontaneous myoclonic jerking”. From reports on the effects of thalamic surgery in such cases, however, no clear-cut conclusion can be drawn concerning a passive or an active role of the thalamus. In the latter case, the thalamus might sustain or even activate the myoclonus. REFERENCES 1 Cowan, J.M.A., Rothwell, J.C., Wise, R.J.S. and Marsden, CD., Electrophysiological and positron emission studies in a patient with cortical myoclonus, epilepsia partialis continua and motor epifepsy, .I. Neural. Neurosurg. Psychiatry, 49 (1986) 796-807. 2 Engel, J., Kuhl, D.E., Phelps, M.E., Rausch, R. and Nuwer, M., Local cerebral metabolism during partial seizures, Neurology, 33 (1983) 400-413. 3 Franck, G., Sadzot, B., Safmon, E., Depresseux, J.C., Gri-
sar, T., Peters, J.M., Guillaume, M., Quaglia, L., Deifore, D. and Lamotte, D., Regional cerebral blood flow and metabolic rates in human focal epilepsy and status epilepticus, Adv. Neural., 44 (1986) 935-948. 4 Henry,
T.R., Mazziotta, J.C., Engel, J., Christenson, P.D., Zhang, J.X., Phelps, M.E. and Kuhi, D.E., Quantifying interictal metabolic activity in human temporal lobe epilepsy, J. Cerebral Blood Flow Metab., 10 (1990) 748-757. 5 Jones, E.G., Wise, S.P. and Coulter, J.D., Differential thalamic relationship of sensory-motor and parietal cortical
Siegfried” performed, without postoperative clinical improvement, an extensive resection of the Rolandic and pre-Rolandic region in a patient with epc and Jacksonian fits caused by chronic encephalitis. A few weeks later, a stereotactic coagulation of the ventrolateral part of the thalamus was carried out in the same patient, after which the seizures stopped. In an extensive review on epc LGhler and Peters found 10 patients in whom a coagulation of the ventrolateral thalamic nucleus was performed. In only 4 patients, however, did such therapy result in clinical improvement’. In our patient, with a history and clinical signs suggesting ‘chronic en~eph~itis’ (Rasmussen) the histopathological findings of the brain biopsy did not reveal inflammatory signs or typical histopathological changes. The absence of inflammatory signs, however, does not rule out the presence of ‘chronic encephalitis’, as emphasized by Rasmussen in his recently published review of 48 operated patients with this syndrome’. The histopathologitally determined ‘gliosis’, located close to the hypermetabolic focus in PET, developed apparently during the ensuing course of the disease. ~nfortunateIy, we have no ultimate proof that this gliotic brain tissue or its boundaries trigger the epileptic discharges. In our patient resective surgery was discussed, but was refused by the girl’s parents despite further progressive deterioration with increased hemiparesis. fields in monkey, J. Camp. Neural., 183 (1979) 833-882. 6 Lammertsma, A.A., Brooks, D.J. and Frackowiak, R.S.J., Measurement of glucose utilisation with [“F]2-fluoro-2deoxy-D-glucose: a comparison of different analytical methods, .I. Cerebral Blood Flow Metab,, 7 (1987) 161-172. 7 Lohler, J. and Peters, U.H., Epilepsia partialis continua (Kozevnikov-Epilepsie), Fortschr. Neurol. Psychiatr., 42 (1974) 165-212. 8 Rasmussen, T., OIszewski, J. and Lloyd-Smith,
D., Focal seizures due to chronic localized encephalitis, Neurology, 8 (1958) 435-445. 9 Rasmussen, T. and Andermann, F., Update on the syndrome of ‘chronic encephalitis’ and epilepsy, Clin. Aspects Pediatr. Epilepsy, 56 (1989) 181-184. 10 Siegfried, J. and Bernoulli, C., Stereoelectroencephalographic exploration and cpilepsia partialis continua, Acra Neurochir. (SuppI.), 23 (1976) 183-191. 11 Wieser, H.G., Graf, H.P., Bernoulli, C. and Siegfried, J., Quantitative analysis of intracerebral recordings in epilepsia partialis continua, Electroencephalogr. Clin. Neurophysiol., 44 (1978) 14-22.