Cerebral glucose metabolism in five patients with Lennox-Gastaut syndrome

Original Articles

Cereb Patients

Glucose Metabolism in Five Lennox-CcastautSyndrome

Kazuie Iinuma, MD*, Kazuhiko Yanai, MD *, Toshiro Yanagisawa, MD*, Noboru Fueki, MD*, Keiya Tada, MD*, Masatoshi Ito, MDt, Taiju Matsuzawa, MDt, and Tatsuo Ido, PhD~

Regional cerebral glucose metabolic rates were estimated by positron emission tomography, in parallel with electroencephalography and cranial computed tomography in 5 patients with Lennox-Gastaut syndrome. The 5 patients, 3 boys and 2 girls, ranged in age from 10-15 years. Computed tomography revealed no gross abnormalities. Each patient received 2-5 mCi of 2-(lSF)-fluoro-2-deoxy-D-glucose (IBF-FDG) intravenously. Averaged cerebral glucose metabolic rates were reduced in each cerebral region as compared with controls. Unilateral hypometabolism was present in 4 patients: one in the inferior frontal gyrus as well as the posterior portion of the superior temporal gyms; one in the inferior frontal gyms; one in the posterior portion of the superior temporal gyms; and one demonstrated diffuse hemispheric hypometabolism including the inferior frontal and posterior portion of the superior temporal gyms. The side of hypometabolism was the same as the epileptogenic focus on the eleco troencephalogmm. No focal changes were demonstrated on the electroencephalogram of a patient whose positron emission tomography revealed hemispheric hypometabolism. Hypometabolism of the inferior frontal and posterior portion of the superior temporal gyrus may relate to the possible pathogenesis of Lennox-Gastaut syndrome. Positron emission tomography has the potential to reveal a latent focal or lateralized abnormality in some patients with nonlocalized electroencephalographic changes.

occurrence in preschool-age children. Seizure types include axial tonic, atonic, atypical absence, and myoclonic attacks. EEG characteristics consist of 1.5-2.5 Hz diffuse spike-and-wave complexes, with or without localized epileptiform activities, and background slowing. Recognized etiologic factors are congenital and hereditary abnormalities, perinatal insults, cerebral infections, and trauma. Diffuse brain damage is indicated by seizure characteristics, intellectual impairment, EEG abnormalities, and etiologies; however, the occasional appearance of localized epileptogenic changes on EEG and the occurrence of focal seizures imply the possibility of underlying focal brain damage. Positron emission tomography (PET) with fluorodeoxyglucose provides three dimensional imaging of glucose metabolism. This recently developed method has been expected to be a powerful tool for detecting a latent focus or for clarifying an expected abnormality in various types of epilepsy. PET studies, however, of secondary generalized epilepsy such as LGS are rare in the literature [1-3]. The purpose of this study is to evaluate the regional metabolic rate for glucose in LGS and its correlation with EEG findings. Methods

Introduction Lennox-Gastaut syndrome fiGS) is characterized by intractable seizures, specific electroencephalographic (EEG) features, poor intellectual functioning, and its

This study includes 5 patients with LGS, 3 boys and 2 girls, whose ages ranged from 10-15 years (mean: 13 years). All EEGs demonstrated diffuse, irregular slow spike-and-wave activity. Three patients experienced no localized discharges; two patients exhibited localized spikes or polyspikes as well as diffuse slow spike-and-wave activity. Clinical features are summarized in Table 1. LGS was defined by the following criteria: (1) The presence of at least two types of seizures (atonic seizures, tonic spasms, myoclonic seizures, and atypical absence); (2) Mental retardation; and, (3) Abnormal interictal EEGs with diffuse irregular slow spikeand-waves with or without localized discharges. Fluorine-18 was produced at the Cyclotron and Radioisotope Center, Tohoku University and 18F-FDG was prepared using a fully automated synthesis system [4]. The FDG method for determining regional cerebral metabolic rate for glucose (rCMRglc) in humans [5] is similar to the *4C-deoxyglucose autoradiographic method [6]. Each patient received 2-5 mCi of lsFDG intravenously. Arterial blood samples were obtained serially at timed intervals until the end of the

From the *Department of Pediatrics; Tohoku University School of Medicine; tDepartments of Radiology and Nuclear Medicine; Institute for Tuberculosis and Cancer; ~?Cyclotron and Radioisotope Center; Tohoku University; Sendal,Japan.

Communications should be addressed to: Dr. linuma; Department of Pediatrics; Tohoku University School of Medicine; 1-1 Seiryo-machi; Sendai 980, Japan. Received September 11, 1986; accepted December 29, 1986.

Iinuma K, Yanai K, Yanagisawa T, Fueki N, Tada K, Ito M, Matsuzawa T, Ido T. Cerebral glucose metabolism in five patients with Lennox-Gastaut syndrome. Pediatr Neurol 1987; 3:12-8.

12

PEDIATRIC NEUROLOGY

Vol. 3 No. 1

Table 1.

Patients' clinical profiles Presumed

Age* at

Etiologic

Seizure Type

Factors

at Onset

Padent

Age*/Sex

PET

1

2/M

10

None

Myoclonic

Seizure Type at Report

CT

Tonic,

N

EEG

rCMRglc

Diffuse slow spike-

R inf front

atypical

wave, focal spike

decrease

absence,

R anterior

GTC 2

3/M

12

None

Tonic

Tonic, atonic,

N

atypical

Diffuse slow spikewave

Diffusely decreased

absence 3

4/M

13

Head trauma

Tonic, atonic

Tonic, atonic,

N

GTC

Diffuse slow spikewave

at 4 yrs

L hemisphere decrease (inf front and post sup temp)

4

5/F

14

None

Atypical absence

Atonic, atypical

Mild VD

Diffuse slow spikewave

L post sup temp decrease

Diffuse atypical

R inf front

absence 5

2/F

15

None

Tonic, atypical absence

Tonic,

N

atypical

slow spike-wave,

R post sup

absence

focal polyspike

temp decrease

bursts in R frontocentral *Age (years) Abbreviations: GTC

= Generalized tonic-clonic

Inffront

= Inferior frontal gyrus

L

= Left

N Post sup temp

= Normal = Posterior superior temporal gyrus

R

= Right

VD

= Ventricular dilatation

procedure from a catheter placed in the radial artery. The sample withdrawal schedule was as follows: every 20 seconds for the first minute; every 30 seconds until 3 minutes postinjection; every 2.5 minutes until 10 minutes postinjection; every 10 minutes until the end of the procedure. The samples were centrifuged to provide plasma for measurement of radioactivity and glucose. The fluorine-18 radioactivity and glucose concentration in the samples were measured with well-type gamma counter and glucose oxidase methods, respectively. Forty-five minutes after the *sF-FDG injection, two PET slices at 4 and 5 or 6 cm over the orbito-meatal level were scanned for 10 or 15 min using an ECAT II (Ortec) with a spatial resolution of 17 mm. The PET images were reconstructed and corrected by transmission image before injection. Regional cerebral metabolic rate for glucose in units of mg/100 gm/min was calculated from brain radioactivity, time of testing,

plasma radioactivity, and plasma glucose level with the predetermined lumped constant value of 0.42. Because no normal values of rCMRglc in childhood and adolescence have been reported in the literature, mean rCMRglc of the unaffected side in patients with partial epilepsy was used for control values of rCMRglc [7]; this criterion excluded our patients. These values appear to be a more suitable control in a study of epileptics because patients with partial epilepsy also were administered antiepileptic drugs. Regional cerebral metabolic rate for glucose of the following gyri or regions were calculated in two slices: superior frontal gyrus, inferior frontal gyrus, precentral gyrus, anterior portion of the superior temporal gyms, posterior portion of the superior temporal gyrus, calcarine region, caudate nucleus, lenticular nucleus, thalamus, and hippocampal region. If a region appeared in one or more adjacent slices, a weighted mean rCMRglc was calculated by the following formula, where i = 1, 2 . . . . . equals each pickcell's number, (rCMRglc) i equals the regional

Iinuma et al: PET in Lennox-Gastaut Syndrome

13

10

-G )NTROL

F

T

O

C

Figure 1. Regional cerebral metabolic rate for glucose in each brain region averaged over both hemispheres in 5 patients with LGS (black columns) and in the 7 age-matched controls (white columns). The rCMRglc is expressed as mg of glucose/1 O0 gm brain tissue/min in the regions: F=frontal cortex; T=temporal cortex; O=occt~oital cortex; and C =caudate and putamen. The asterisk indicates statistical significance 60 < 0.001) between LGS and controls. Error bars represent one standard deviation.

MEAN BRAIN REGION

CASE

CASE 5 1

REGIONAL

'

I

5

CEREBRAL

2

CASE 5

METABOLIC

RATE

FOR

CASE

3 10

GLUCOSE

~mg 'OOg

4

r11.~

CASE

5

5

I

J.

5 IU

SUP F R O N T A L GYRUS

=lmm!B==mlap

INF FRONTAL GYRUS PRECENTRAL GYRUS ANT PART SUP TEMP POST PART SUP TEMP CALCARINE REGION CAUDATE NUCLEUS LENTICULAR NUCLEUS

THALAMUS

HIPPOCAMP

II

LEFT RIGHT • P 005

Figure 2. Mean regional cerebral metabolic rate for glucose in each cerebral region and gyrus in 5 patients with LGS (black columns: left, white columns., right). The asterisk indicates statisttcal significance between both hemispheres (p < 0.05). Error bars represent one standard devtation.

14

PEDIATRIC NEUROLOGY

Vol. 3 No. 1

L

R

O M + 40

L

OM + 60

Figure3. FDG PET images of Patient 3. Note the left hemispheral hypometabolism for glucose, especially the left postern'ortemporal region.

cerebral metabolic rate for glucose of each pickcell; N is the number ofpickcells within each region of interest: E (rCMRglc) i N These regions were determined by comparison with anatomical sections of CT scans of each patient.

Results The clinical features and results of the EEG, CT, and PET of each patient are summarized briefly in Table 1. When the values of rCMRglc were averaged over both hemispheres in all patients, they were reduced in all brain regions including the caudate when compared with those of controls (Fig 1). The difference between the LGS group and the controls was statistically significant ( p <0.001). In order to delineate a localized or asymmetric hypometabolism in each gyms or portion of cerebral hemispheres, Figure 2 was produced according to the formula presented in the data analysis. Figure 2 depicts weighted mean values for rCMRglc in grey matter of the individual lobes and nuclei of the left and right hemispheres with LGS. In Patient 3, relatively widespread hypometabolism of glucose was observed over the left hemisphere, most strikingly in the left inferior frontal gyrus and in the

posterior portion of the superior temporal gyms (Fig 3). In Patient 1, unilateral hypometabolism was observed in the right inferior frontal gyms and in the posterior portion of the superior temporal gyms. In Patient 4, unilateral hypometabolism was observed in the left posterior portion of the superior temporal gyms. In Patient 5, a hypometabolic region was observed in the right inferior frontal gyms and in the right posterior portion of the superior temporal gyms (Fig 4). In Patient 2, the metabolic rates of glucose were symmetric and low in all calculated grey matter areas (below 5 mg / 100 gm/min). It is noteworthy that unilateral bypometabolism of the inferior frontal gyms and the posterior portion of the superior temporal gyms were most commonly observed in 3 of the 5 patients. In two patients whose EEGs demonstrated localized epileptic discharges, the side of the EEG abnormality matched the side of hypometabolism of glucose. Of the other 3 patients whose EEGs did not demonstrate localized changes, 2 revealed unilateral hypometabolism of glucose.

Discussion Recently PET has been developed as a tool to

linuma et al: PET in Lennox-Gastaut Syndrome

15

R

OM + 40

L

/?

OM + 50

Figure 4. lnterictal FDG-PET images oJ Patient 5 demonstrating the/oca/izedlow metabolic rate jbr g/ucose m the nght poster~br portt'on of the superior temporal gyrus and in the small area of the right infertorfrontal gyrus.

evaluate cerebral function in epilepsy [8-12] in addition to electrophysiologic studies. The PET studies of epilepsy have mainly contributed to the assessment of epileptic foci in partial seizures by using the 2-( 1" F)fluoro-2-deoxy.D-glucose (FDG) method. In partial epilepsy, Enget et al. [8] reported that although the side of the interictal glucose hypometabolic zone disagreed with the epileptic focus revealed by individual scalp and depth EEG-recorded ictal and interictal epileptiform activity, the site of focal hypometabolism paralleled the epileptic focus determined by the combined results of all electrophysiologic studies and was epileptogenic. Engel et al. [9] also evaluated pathologic findings of temporal lobe tissue from 25 patients with complex partial seizures who had been studied previously with interictal FDG-PET. Nineteen of 22 patients with hypometabolic zones had corresponding focal pathologic findings. Engel et al. concluded that the hypometabolic zone determined by FDG-PET agreed with the pathologic brain lesion, although the size of the hypometabolic zone was generally larger than the area of pathologic involvement. Only a few PET studies of generalized epilepsy have been reported [13,14]. In the report of Theodore et al.

16

PEDIATRIC NEUROLOGY

Vol. 3 No. 1

[13], none of the 9 patients, ages 20-38 years, demonstrated focal or lateralized hypometabolism with clinical absence or generalized seizures. Engel et al. [14] estimated rCMRglc before and during petit mal absence. According to their report, although ictal scan revealed a diffuse hypermetabolic rate for glucose compared to interictal scan, interictal scan revealed normal rCMRglc. Neither interictal or ictal scans demonstrated a localized or lateralized hypoor hypermetabolic zone. FDG-PET studies of secondary generalized epilepsy such as LGS have been reported [1-3]. Gur et al. [1] reported 2 patients with LGS with temporal lobe unilateral hypometabolism. In 1 patient whose seizures were controlled after corpus caltosotomy, temporal lobe metabolism became symmetric. They concluded that a temporal lobe focus in 2 patients with Lennox-Gastaut syndrome was suggested. On the other hand, Theodore et al. [2] reported that FDG-PET studies on 5 patients with LGS demonstrated generalized hypometabolism, most prominent in the frontal and temporal regions in 3 patients and no identifiable abnormalities in 2 patients. No regions of hypometabolism were detected. Chugani et al. [3] reported variable FDG-PET scans of LGS, namely metabolic patterns including unilateral focal, unilateral

diffuse, bilateral multifocal, and bilateral diffuse abnormalities. They concluded that metabolic abnormalities were not necessarily in the temporal lobes; this conclusion contrasted with that of Gur et al. [1]. In our study, 1 patient demonstrated bilateral diffuse, 1 patient demonstrated unilateral diffuse, and 3 patients demonstrated unilateral focal hypometabolism. Our results also revealed variable findings of FDG-PET scans of LGS; however, these findings were characterized by two abnormalities. One was diffuse hypometabolism compared to our controls; the other was unilateral localized hypometabolism in either the posterior portion of superior temporal gyrus, the inferior frontal gyms, or both. Our patients did not previously have localized brain damage or partial seizure disorders. Only Patient 3 had a mild head trauma 6 months before the onset of seizures. The seizures, however, were of the LGS type from the onset, and his CT was normal. LGS has been presumed to be a characteristic epileptic syndrome of idiopathic and/or symptomatic generalized epilepsy [15]. Diffuse slow spike-and-wave complexes were characteristic and essential findings on EEG of LGS [16]. Generalized abnormalities on FDGPET, especially diffuse hypometabolism as observed in Patient 2, may be one of the main features in LGS. The regional metabolic rate for glucose does not demonstrate asymmetry between the two hemispheres in healthy subjects [17], although such clear evidence has not been confirmed in a younger age group less than 20 years of age. Therefore, the asymmetric rCMRg[c observed in our study is significant. As observed in partial seizures [9], the hypometabolic zone of glucose possibly relates to an epileptic focus in secondary generalized epilepsy. If there is a possibility of localized abnormality of glucose metabolism in LGS, either the posterior portion of superior temporal gyrus or the inferior frontal gyms may be a likelier candidate according to the present study. The report by Gut et al. [I] may partly support this hypothesis, although they only mentioned temporal lobe hypometabolism. In our study, the side of EEG localization matched the hypometabolic side in PET if the EEG had focal abnormalities. Even two patients' EEGs that did not indicate localization, did demonstrate unilateral hypometabolism. These findings suggest that diffuse electrical discharges observed in scalp EEG may not be primary, and a primarily pathologic lesion in LGS may be localized, and masked by generalization of the vast discharges. FDG-PET study can reveal a latent focal or lateralized abnormality in some patients with nonlocalized EEG changes. Gastaut et al. [16] reported that in some instances the spike-and-wave predominance was such as to suggest a temporal focus; in the course of repeated recordings a well-defined temporal focus was

observed, ishikawa et al. [18] reported a patient with LGS who had a porencephalic cyst in the left temporal lobe. Their patient exhibited marked improvement in clinical seizures and EEG after surgical treatment. Angelini et al. [19] also reported a patient with LGS who had a parietotemporal astrocytoma and was treated successfully by tumor removal. These reports suggest that the temporal lobe strongly relates to the pathogenesis of LGS. This work was supported in part by a Grant from the Ministry of Health and Welfare of Japan, 1984 (84-05-23) and 1985 (85-01-23).

References

I1] Gur RC, Sussman NM, Alavi A, et al. Positron emission tomography in two cases of childhood epileptic encephalopathy (Lcnnox-Gastaut syndrome). Neurology 1982;32:1191-4. [2] Theodore WH, Brooks RD, Patronas N, et al. Positron emission tomography in the Lennox-Gastaut syndrome Neurology 1984;34(Suppl 1): 106-7. [3] Chugani H, Engel J Jr, Mazziotta JC, Phelps ME. '"F-2fluorodeoxyglucose positron emission tomography in medically refractory childhood epilepsy. Neurology 1984;34(Suppl 1): 107. [4] Iwata R, ldo T, Takahashi T, Nonma M. Automated synthesis system for production of" 2-deoxy-2-Lsf]uoro-D-glucosc with computer control, lntJ Appl Radiat lsot 1984;35:445-34. [5] Phelps ME, Huang SC, Hoffman EJ, Selin C. Sokoloff L, Kuhl DE. Topographic measurement of local cerebral glucose metabolic rate in humans with 18-F-2-fluoro-2-glucosc: Validation of method. Ann Neurol 1979;6:371-88. [6] Sokoloff L, Reivich M, Kennedy C, et al. The (14C) deoxyglucos¢ method for the measurement of lo~al cerebral glucose utilization: Theory, procedure and normal values m the conscious and anesthetized albino rat.J Neurochem 1977;28:897-916. [7] Yanai K, Iinuma K, Tada K, c t a l . Regional cerebral metabolic rate for glucose in subacute sclerosing pancncephalitis. Eur J Pediatr (in press). [8] EngelJJr, Kuhl DE, Phelps ME. Crandall PH. Comparative localization of epileptic loci in partial epilepsy by PET and EEG. Ann Neurol 1982;12:529-37. [9] Engel J Jr, Brown WJ, Kuhl DE, Phelps ME, Mazziotta J r . Crandall P|l. Pathological findings underlying focal temporal lobe hypometabolism in partial epilepsy. Ann Neurol 1982:12:518-28. [10] Chase TN, Brooks RA, DeLaPaz RL, et al. Positron emission tomographic studies of epilepsy, brain tumor, and Alzhcimer's disease. Psychopharmacol Bull 1982; 18(3):3-6. [11] Bernardi S, Trimble MR, Frackowiak RSJ, Wise RJS,Jones T. An interictal study of partial epilepsy using positron emission tomography and the oxygen-15 inhalation technique. J Neurol Neurosurg Psychiatry 1983;46:473-7. [12] Gallhofer B, Trimble MR, Frackowiak R, GibbsJ,Jones T. A study of cerebral blood flow and metabolism in epileptic psychosis using positron emission tomograpby and oxygen. J Neurol Neurosurg Psychiatry 1985;48:201-6. [13] Theodore WH, Brooks R, Margolin R, et al. Positron emission tomography in generalized seizures. Neurology 1985; 35:684-90. [14] Engel J Jr, Lubens P, Kuhl DE, Phelps ME. Local cerebral metabolic rate for glucose during petit mal absences. Ann Neurol 1985;17:121-8. [15] Commission on the Classification and Terminology of the International League Against Epilepsy: Proposal for classification of epilepsies and epileptic syndromes. Epilepsia 1985 ;26:268-78.

linuma et al: PET in Lennox-Gastaut Syndrome

17

[16] Gastaut H, Roger J, Soulayrol R, Tassinari CA, Regis H, Dravet C. Childhood epileptic encephalopathy with diffuse slow spike-waves (otherwise known as "petit mal variant") or Lennox syndrome. Epilepsia 1966;7:139-79. [17] Duara R, Grady C, Haxby J, et al. Human brain glucose utilization and cognitive function in relation to age. Ann Neurol 1984;16:702-13.

[18] lshikawa T, Yamada K, Kanaya M, et al. A case of LennoxGastaut syndrome improved remarkably by surgical treatment of a porencephalic cyst: A consideration on the generalized corticoreticular epilepsy. No To Hattatsu 1983;15:356-65. [19] Angelini L, Broggi G, Riva D, Solero CL. A case of LennoxGastaut syndrome successfully treated by removal of a parictotemporal astrocytoma. Epilepsia 1979;20:665-9.

Erratum

In the article on hyperammonemia by Breningstall (Breningstall GN. Neurologic syndromes in hyperammonemic disorders. Pediatr Neurol 1986;2:253-62), the last four reference citations were inadvertently omitted during production. They are as follows: [24] Hantis DJ, Yang B, Wolf B, Snodgrass PJ. Dysautonomia in an infant with secondary hyperammonemia due to propionyl coenzyme A carboxylase deficiency. Pediatrics 1980;65:107-10. [25] I'Iillman RE, Keating JP, Williams JC. Biotin responsive propionic acidemia presenting as the rumination syndrome. J Pediatr 1978;92:439-41. [26] Dubowitz V. The floppy infant. 2nd ed. Philadelphia: JB Lippincott, 1980;107-8. [27] Flannery DB, Hsia YE, Wolf B. Current status of hyperammonemic syndromes. Hepatology 1982;2:495 -506.

18

PEDIATRIC NEUROLOGY

Vol. 3 No. 1