Journal of Computer Assisted Tomography 7(6):946-953, December © 1983 Raven Press, New York Hypometabolic Cortical Lesions in Tuberous Sclerosis with Epilepsy: Demonstration by Positron B. Szelies, K. Herholz, W.-D. Heiss, A. Rackl, G. Pawlik, R. Wagner, H. W. Ilsen, and Emission Tomography K. Wienhard Abstract: Four patients with a well-established diagnosis of tuberous sclerosis and grand mal type epileptic seizures as their principal clinical symptom were examined by conventional surface electroencephalography (EEG), X-ray com- puted tomography, and positron emission tomography (PET) using the ['8F]- 2-fluoro-2-deoxyglucose method. The interictal EEG showed various abnor- malities of poor localizing value, but no focal epileptic discharges. X-ray com- puted tomography demonstrated subependymal calcifications in all cases, al- though cortical lesions were found only twice. However, in the PET images of each patient one or two localized cortical foci with a metabolic rate for glucose more than 40% lower than in the respective contralateral region were clearly delineated. It may be assumed that those hypometabolic areas repre- sent the epileptogenic cortical tubers, which are characteristic of the disease but usually cannot be detected in vivo by other methods. Index Terms: Epi- lepsy—Tuberous sclerosis—Cerebral metabolism—Emission computed to- mography. Diagnosis of tuberous sclerosis (TS, Bourneville— Pringle disease) is easily established if the typical triad—adenoma sebaceum, mental deficiency, and epileptic seizures (1)—is present. However, in formes frustes typical skin lesions are often missing (2-6) and further diagnostic procedures must be performed to demonstrate characteristic focal alter- ations of the brain tissue presenting as cortical tubers and calcified subependymal nodules (7). In X-ray computed tomography (XCT) the latter cal- cifications—if visible at all—are most conspicuous. Frequently, they cannot be detected despite typical clinical symptoms including seizures (8), although they may be found in some relatives of TS patients not showing any of the typical manifestations of the disease (5,9). The cortical tubers, however, which gave the disease its name, usually are of normal density in XCT and can be seen only in the few cases in which they are calcified (10). These tubers probably represent the epileptogenic foci character- From the Max-Planck-Institut fiir neurologische Forschung, Ostmerheimer Strasse 200, 5000 KéIn 91 (Merheim), Federal Re- public of Germany. Address correspondence and reprint re- quests to Dr. Heiss. 946 istic of TS and, therefore, their detection may not only be of diagnostic value but also guide therapy, e.g., towards surgical removal of the lesion. According to Kuhl et al. (11) and Engel et al. (12,13), epileptogenic foci can be detected by pos- itron emission tomography (PET) because those pathological brain regions usually exhibit a de- creased uptake of [!8F]-2-fluoro-2-deoxyglucose (FDG) in the interictal state. The present study was performed to investigate the incidence of such hy- pometabolic cortical lesions in patients with TS pre- senting with epileptic seizures as their principal clinical symptom. MATERIALS AND METHODS Four patients with firmly established TS were ex- amined by PET. In each case diagnosis was based ona typical history, clinical examination, XCT (Sie- mens Somatom 2000, GE CT/T 8800), and conven- tional mono- and bipolar surface electroencepha- lographic (EEG) recordings with electrodes placed according to the international 10/20 system. The routine described by Reivich et al. (14) and Phelps (15) was used to measure local cerebral metabolic HYPOMETABOLISM IN TUBEROUS SCLEROSIS 947 rate of glucose (LCMRGN: approximately 5 mCi of FDG was injected intravenously, and arterialized blood was sampled from the vein of a heated hand at successively increasing intervals. from the time of injection to the end of the recording period. Plasma FDG and glucose concentrations were de- termined and entered into the memory of the com- puter (VAX 11/780, DEC) connected with the tomo- graphic device. Cross-sectional positron images of the brain were obtained with a four-ring/seven slice system (Positron Camera PC 384-7B, Scanditronix) designed by Eriksson et al. (16) starting 30 min after FDG injection. This device has a spatial resolution of 8 mm FWHM in 13 mm slices. The FDG activity in the brain was first recorded for 10 min with the lowest ring positioned in the canthomeatal (CM) plane, yielding further tomograms at 13.5, 27, 40.5, 54. 67.5, and &1 mm above the CM line. Subse- quently, for another 10 min run, the patient's head was moved 7 to 20 mm out of the gantry to obtain the activity distribution in sections centered be- tween the first set of parallel planes. That way the entire brain was scanned. From the local cerebral concentration of '*F and the plasma concentration curve of FDG. the LCMRGI was calculated according to the equation derived by Sokoloff et al. (17), using the average values of rate constants and the lumped constant given by Phelps (15). The implemented software permitted display of metabolic maps in units of pmol glucose/l00 g/min. or the extraction of av- erage LCMRGI values from selected cortical re- gions of interest (ranging in size between 1.9 and 18.0 cm? in the plane of section). To minimize in- terindividual variability the metabolic rates of vi- sually suspect regions were compared with those of corresponding contralateral areas and a metabolic index (MI} = LCMRGI suspect/LCMRGI contra- lateral was computed for each pair. CASE REPORTS Case 1 A9-year-old girl developed episodes of motor agitation during sleep at the age of 7: She cried and struggled with her arms and legs. Then. breathing heavily, she usually became stiff all over, During these seizures no tongue bite or welling ever occurred. Over the next year, frequency and duration of episodes increased and often series of seizures were observed during a single night and oceca- sionally also during the day. Therefore. anticonvulsant medication was initiated and the epileptic manifestations ceased. When the child was referred to the neurological hospital for clinical evaluauion, she complained about im- pairment of concentration and memory. loss of perfor- mance at school, and behavioral disturbances. No herit- able diseases were reported in the family. Clinical findings included adenoma sebaccum. sha- green patches on the trunk, depigmented nevi on the legs. There were no focal neurological signs or mental retar- dation. The EEG revealed bilateral temporal theta-delta slowing. The XCT demonstrated ventricular dilatation and multiple calcifications in the ventricular walls (Fig. la). The PET showed focal hypometabolism in the left frontal cortex (Fig. 1b). Case 2 A 28-year-old woman was known to have suffered from sudden loss of consciousness and muscle tone at an av- erage attack rate of one every month from early child- hood to the age of 16. These episodes. lasting from sec- onds to minutes, were taken as manifestations of a cir culatory insufficiency. and the patient was not seen by a neurologist until she was 28 years old. At that time she was referred to the neurological hospital because of a single grand mal type epileptic seizure. Four years earlier a benign renal hamartoma had been diagnosed. which over the following 3 years had turned into a lipoangio- leiomyosarcoma necessitating nephrectomy. Highly vas- cularized tumors of the other kidney also were described. One year after her first neurological hospitalization. pul- monary cystic Myomatosis caused pneumothorax on two occasions, There was no family history of heritable dis- eases. Clinical findings included adenoma sebaceum, sha- green patches in the dorsolumbar region, depigmented nevi on both thighs, mild focal neurologic disturbances on the right side. but no mental retardation. The EEG revealed intermittent left precentrotemporal theta slowing. and. following provocation by sleep withdrawal, marked left precentrotemporal dysrhythmia, The XCT demonstrated some frontal cortical and cerebellar atrophy, subependymal calcifications localized in the walls of the body of the lateral ventricles near the foramen of Monro, and blurred areas of slightly reduced density in the right temporal white matter. The PET showed focal hypometabolism in the left parietal cortex, Case 3 A 22-year-old man, who had been delivered by forceps, suffered his first grand ma! seizure at the age of 16, at which time drug treatment was started. Two years later. during a change of anticonvulsant medication, a series of epileptic seizures occurred. At that time an interictal EEG showed a right temporoparietal theta-delta focus with sharp and slow waves. A right carotid angiogram was normal, although a pneumoencephalogram demon- strated dilatation of the lateral ventricles, more marked on the left side. The XCT revealed a hypodense area in the right frontotemporal region. After 2 more years and another series of epileptic seizures, the diagnosis of TS was established. The patient’s father had died of kidney failure due to a fibroleiomyosarcoma at the age of 38. Clinical findings included adenoma sebaceum. but no neurologic or mental deficits. The EEG revealed a right lemporoparietal theta-delta focus without epileptic poten- tials. The XCT demonstrated multiple subependymal periventricular calcified nodule. calcifications in the right frontal subcortical white matter and in the interhemi- spheric fissure an irregular, patchy area of mixed density in the left frontoparietal cortex and adjacent white matter (Fig. 2a). The PET showed hypometabolic foci in the right frontal and left frontoparietal cortex (Fig. 2b). FA Camput Assist Lomoyr Vol > Ne 6. 98s -Ay payeauljap Ajaeajo @ MOUS | ase jo saBew! 13d Buipuodsaiog :q ‘a[d14j}UaA [B19]; ‘ujwB QOL ow ul uaal6 aie (ajeas Aai6 aduaiajas == yHis ay) yO Apog ay) ui suo ysabue) ayy SuoediyI9/29 se|NpoOU ajdijinw pue uoyeyelIp sejN aas) Sayes TNOGRIaW (8 ‘ON 9dI|S PESYMOIe) x9}O9 JeJUOL 19] 9} U] SNDOJ DIjoqeyawod = -d14JUaBA MOYS aURIA [EIBAWOUIUED ay] BADGE WW £9 O} EL WO4j | BSED JO SI DX ‘Be “Lb “Old Gy 4 qe HYPOMETABOLISM IN TUBEROUS SCLEROSIS 949 FIG. 2. a: XCTs of Case 3 from 20 to 54 mm above the canthomeatal plane show multiple subependymal calcifications. a calcified nodule in the right frontal subcortical white matter (open arrow), calcifications in the anterior interhemispheric fissure. and an irregular, patchy area of mixed density (partially calcified), in the left frontal lobe (solid arrows). b: Corresponding PET images of Case 3 show a large hypometabolic area in the right frontal cortex (open arrowhead) and a smaller one in the left frontoparietal cortex (solid arrowhead). Metabolic rates (see reference grey scale) are given in wmol 100 g min PCumput Wsvat Lemos Val 7 No 6 1988 950 B. SZELIES ET AL. Case 4 A 38-year-old man had been operated on for an aneu- rysm of the right anterior cerebral artery after his first generalized epileptic seizure at the age of 4'/2. Fourteen years later, when repeated seizures occurred, antiepi- leptic medication was started. A pneumoencephalogram at that time showed a dilatation of the right lateral ven- tricle due to postsurgical atrophy. At 22 years the patient noticed nodular lesions in his face and in the nail bed of his fingers. Histological examination led to the derma- tological diagnosis of Pringle disease. After a series of epileptic seizures at the age of 33, the patient was ad- mitted to the neurological hospital. Clinical findings included an adenoma sebaceum, Ko- enen tumors of the nails, depigmented nevi, and café-au- lait spots, but no neurological or mental deficits. The EEG revealed severe generalized paroxysmal dys- rhythmia and a right temporal theta-delta focus. The XCT showed dilatation of the lateral ventricles, more marked on the right side, as well as two calcified nodules close to the body of the lateral ventricles. Five years later a clinical reexamination because of re- peated seizures failed to show any changes. The EEG revealed theta-delta dysrhythmia over the left hemisphere intensified by hyperventilation. The XCT showed ven- tricular dilatation, more marked on the right side: calci- fications in the lateral walls of both lateral ventricles; and localized right frontal atrophy following aneurysm sur- gery. The PET demonstrated hypometabolic areas in the tight frontal and left parieto-occipital cortex. RESULTS Pertinent data on the four patients are summa- rized in Table 1. At the time of the PET study the interictal surface EEG showed various uncharac- teristic abnormalities, including slow wave activity of variable degree and dysrhythmic changes in all four patients. However, no typical epi’’, ‘c dis- charges were recorded and the observed a. -mal- ities were of poor localizing value. Previous at- tempts to demonstrate an epileptic EEG focus in the drug-free interictal state had equally failed in all but one patient (Case 3) In each case XCT showed the characteristic le- sions of TS (3,6,10,18-21). Multiple calcified no- dular lesions adjacent to the ependyma of the body and the frontal horns (Cases | and 3) or in the sub- cortical white matter (Case 3) were found. Only one patient (Case 3) exhibited a partially calcified cor- tical lesion. Furthermore, XCT revealed dilatation of the ventricles in Cases | and 4, and cortical atrophy and an area of lower density in the temporal white matter in Case 2. In each of the four patients at least one localized cortical area of markedly decreased glucose metab- olism could be demonstrated by PET. All regions with metabolic indices of less than 0.6 were located in the frontal (Fig. 1) or parietal cortex. Only three of them showed some correspondence with a lesion J Comput Assist Tomogr, Vol. 7, No. 6, 1983 visible in XCT: the left frontoparietal cortical area of mixed density in Case 3 (Fig. 2) appeared smaller than the hypometabolic area in the PET image, and the same patient’s right frontolateral hypometabolic focus was exactly superimposed on a subcortical nodular calcification. A frontal hypodense area in the XCT of Case 4 probably corresponded with scar formation following surgery; PET determination of glucose metabolism in this area gave an MI of 0.52. DISCUSSION Interictal PET with FDG demonstrated one or two hypometabolic cortical foci in all of our patients with TS in whom epileptic seizures were the prin- cipal symptom. These abnormalities were rather conspicuous although XCT revealed normal cor- tical structures in all hypometabolic areas except one, in which a calcified cortical tuber corre- sponding to a previously documented epileptic focus in EEG was found, and another, in which a tissue lesion was demonstrated as a sequel of sur- gery. All hypometabolic foci were clearly delineated and could easily be identified by visual inspection of the tomographic images. This quantitative ap- proach was substantiated by qualitative evaluation; in all pathological areas side-to-side differences of LCMRGI exceeded 40%, whereas for normal cor- tical regions an average difference of 6.5% and a maximum value of 18.7% were found (Table 2). All the described hypometabolic areas were lo- cated in the frontal or parietal cortex, the most common site of cortical tubers in TS (7). Those tubers exhibit a disorganization of the cortical ar- chitecture with a marked decrease in the overall density of cells, which are extremely variable in size and shape, often appearing as glial monster cells with astrocytic features (22). Consequently, the total number of neurons consuming significantly more energy than glial cells is markedly reduced compared with the surrounding normal cortical areas, which may largely explain the observed focal hypometabolism. Furthermore, neuronal activity may also be reduced by abnormal neuroglial junc- tions suggesting aberrant synapse formations as ob- served in cortical tubers (23). These peculiarities of defective cortical development in TS form the mor- phological basis of epileptogenic foci, often char- acterized by loss of neurons, glial proliferation, and abnormalities in the dendritic and synaptic struc- tures (24). From the above characteristics of cortical tubers it follows that both altered morphology and depres- sion of functional activity may give rise to those changes in metabolism which manifest themselves as hypometabolic lesions in the interictal PET. It is this combination of possible causes that may ex- plain why morphological changes either are not vis- HYPOMETABOLISM IN TUBEROUS SCLEROSIS YS] TABLE 1. Stammary of patient data PET Average Localization metabohe rate Manifestations of hy pometabolhic mn contralateral of tuberous cortical Metabolic cortical region Case age (years) sex sclerosis focus index (umol 100 ¢ mind 19k Grand mal epilepsy. Left frontal 0.32 Vo skin lesions 220F Grand mal epilepsy. Lett panetal 0.89 MSs mild focal neurological signs on the right. skin lesions, kidney tumors, pulmonary cystic myomatosis 322M Grand mal epilepsy. Right tronto- O44 AUR skin lesions lateral Left fronto- 0.0 36.6 parietal 43RM Grand mal epilepsy. Right frontal 82 amd shin lesions Left paneto- (1.56 2.2 occipital XCT Abnormality corresponding to cortical Other Interictal REG Case age (years) sex PET focus abnormalities abnormalities 19F — Ventricular dilatanon, Bilateral temporal muluple nodular theta-delta slowing. subepends mi : etfigations and areas of lower density 2 — brontal basal cerebral Marked left precentral cort nd ebellar temporal dysrhythmia atrophy, two nodular following provocation subepends mal cal- by sleep withdrawal cifications, right temporal white matter area of lower density 322M Nodular calcification Multiple nodular Right temporoparietal in frontal white calcifications around the theta-delta focus matter adjacent ventricles and in to PET focus the interhemispheric fissure Area of lower — density and small irregular ring of caleification im corte and adjacent white matter 438M Postsurgery defect Internal hydrocephalus _ and localized and bilateral nodular cortigal atrophy Subependy mal cal- — elficanion at the level Lett hemisphere theta- of the body of the delta dysrhythmia in- lateral ventricles tensified by hyperventilation ible in XCT or are much smaller in size than in PET. The latter observation is in good agreement with findings of Engel et al.. who noticed in 19 of 22 cases of temporal lobe epilepsy with focal histo- pathologic lesions that the zone of hypometabolism usually Was much larger than the actual lesion area even if artifactual influences. e.g... partial volume effects. were considered (13). Determination of absolute values of LCMRGI was restricted to obviously normal cortical regions, since it depends on a variety of assumptions as to the value of rate constants and the lumped constant TCompat Wnt Domecr Vol 7 Ne 0 Lust 952 B. SZELIES ET AL. TABLE 2. Comparison of cortical glucose utilization in the two hemispheres Metabolic rate for glucose in normal cortical regions of right cerebral hemisphere Metabolic rate for glucose in normal cortical regions of left cerebral hemisphere Side-to-side differences of the metabolic rate in corresponding regions of normal cerebral cortex Case no. (mean + SD) (mean + SD) (% of larger value, mean + SD) 1 37.1 + 6,36 38.0 + 6.25 5.3 + 3.33 2 25.9 + 2.72 25.0 + 2.49 6.8 + 3.60 3 36.6 + 3.44 36.7 + 3.91 5.6 + 3.43 4 28.7 + 3.90 30.6 + 3.90 8.2 + 5.87 Total 32.7 + 6.09 33.2 + 6.14 6.5 + 4.41 (14,15,25). According to Hawkins et al., the rate constants are different in pathological tissue (26), and the same may be true for the lumped constant (27). However, if regional rate constants were de- termined by dynamic PET in pathological and normal areas (28), no significantly different results were encountered. As demonstrated by Kuhl et al. (11) and Engel et al. (12,13), epileptogenic lesions that may be suited for surgical removal can be identified as hypome- tabolic areas by PET using the FDG method. Re- sults from the present study indicate that this tech- nique is useful also in the detection of cortical tubers, which are a likely cause of seizures in pa- tients with TS. Moreover, those tubers may slowly increase in size. Their metabolic activity corre- sponds to that determined in low-grade astrocy- tomas (22 + pmol/100 g/min) (29). Their early de- tection would, therefore, serve two purposes: the long-term follow-up of potentially malignant tu- morous cortical lesions and the detection of epilep- togenic foci that can successfully be treated by sur- gery. In a study reported by Perot et al. (30) exci- sion of cortical tubers cured epilepsy in three of six TS patients, and substantially reduced the seizure rate in the other cases. Removal of a calcified sub- ependymal nodule, on the other hand, did not result in any improvement. In conclusion it would appear that PET exami- nations of TS not only are of great value for the understanding of the pathogenesis of symptomatic epilepsy, but also can provide a better rationale for medical or surgical therapy, prognosis, and genetic counseling, particularly in the formes frustes of the disease. Acknowledgment: The XCTs were kindly provided by Dr. Schmitz-Draeger and Dr. Hoeffken. REFERENCES 1. Vogt H. Zur Diagnostik der tuberésen Sklerose. Z Erforsch Behandl jugendl Schwachsinns 1908;2:1-12. 2. Nevin NC, Pearce WG. Diagnostic and genetical aspects of tuberous sclerosis. J Med Genet 1968;5:273-80. 3. Martin GL, Kaisermann D, Wiegler D, Amorosi ED, Nadel J Comput Assist Tomogr, Vol, 7, No. 6, 1983 14, 15. 19. 20. 21. H. Computer-assisted cranial tomography in early diagnosis of tuberous sclerosis. JAMA 1976,235:2323-5. . Pampiglione G, Moynahan EJ, The tuberous sclerosis syn- drome: clinical and EEG studies in 100 children. J Neurol Neurosurg Psychiatry 1976;39:666-73. . Burck U, Lagenstein I. CAT scan findings in clinically healthy relatives of tuberous sclerosis patients. Eur J Pe- diatr 1980;135:117-8. . Nevin NC. Computer-assisted cranial tomography in tub- erous sclerosis. Dev Med Child Neurol 1980;22:379-90. . Donegani G, Grattarola FR, Wildi E. Tuberous sclerosis Bourneville disease. In: Vinken PJ, Bruyn GW, eds. Hand- book of clinical neurology, Vol. 14. The phakomatosis. Am- sterdam:North Holland, 1972:340-89. . Murphy JV, D'Souza BJ, Haughton VM. CT scans and tu- berous sclerosis. JAMA 1976;236:1115. . Rott HD, Huk W, Miiller E. Zerebrale Verkalkungen ohne Krampfanfalle bei einem Patienten mit tuberdéser Sklerose. Nervenarzt 1982;53:54-S. . Houser OW, McLeod RA. Roentgenographic experience at the Mayo Clinic. In: Gomez MR, ed. Tuberous sclerosis. New York: Raven Press, 1979:27-53. . Kuhl DE, Engel J, Phelps ME, Selin C. Epileptic patterns of local cerebral metabolism and perfusion in humans. De- termined by emission computed tomography of '*FDG and NH. Ann Neurol 1980;8:348-60. . Engel J Jr, Kuhl DE, Phelps ME, Mazziotta JC. Interictal cerebral glucose metabolism in partial epilepsy and its re- lation to EEG changes. Ann Neurol 1982;12:510-7. . Engel J Jr, Brown WJ, Kuhl DE, Phelps ME, Mazziotta JC, Crandall PH. Pathological findings underlying focal temporal lobe hypometabolism in partial epilepsy. Ann Neurol 1982; 12:518-28. Reivich M, Kuhl D, Wolf A, et al. The ('8F) fluorodeoxy- glucose method for the measurement of local cerebral glu- cose utilization in man. Circ Res 1979;44:127-37, Phelps ME. Positron computed tomography studies of ce- rebral glucose metabolism in man: theory and application in nuclear medicine. Semin Nucl Med 1981.11:32-49. . Eriksson L, Bohm C, Kesselberg M, et al. A four ring pos- itron camera system for emission tomography of the brain. IEEE Trans Nucl Sci 1982;NS-29:539-43. . Sokoloff L, Reivich M, Kennedy C, et al. The 'C-deoxy- glucose method for the measurement of local cerebral glu- cose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 1977;28:897-916. . Gomez MR, Mellinger JF, Reese DF. The use of comput- erized transaxial tomography in the diagnosis of tuberous sclerosis. Mayo Clin Proc 1975;50:553-6. Kazner E, Lanksch W, Steinhoff H. Cranial computerized tomography in the diagnosis of brain disorders in infants and children. Neuropddiatrie 1976;7:136-73. Garrick R, Gomez MR, Houser OW. Demyelination of the brain in tuberous sclerosis. Mayo Clin Proc 1979;54:685-9. Huk W, Rott HD. Computertomographische Befunde bei der a HYPOMETABOLISM IN TUBEROUS SCLEROSIS tuberosen Sklerose Bournesille-Pringle. In: Spranger J. Tolksdorf Meds. Alinische Genetih in der Padiatric. 2. Symposium in Mainz. Stuttgart: Thieme. 1980: 14-8. Critchley Mo Earl CJC. Tuberous sclerosis and allied con- ditions, Brain 1932;86: 311-46, Trombles IK. Mirra SS. Ultrastructure of tuberous sclerosis, cortical tuber and subependymal tumor. Ana Neurol POST Q 174-81. Woodbury DM. Kemp JW. Initiation propagation and arrest of seizures. In: Mrsulia BB. Rakic LIM, Klatzho E, Spatz M. eds. Pathophysiology of cerebral enerey metabolivm. New York: Plenum Press. 1977:313-S1 Phelps ME. Huang SC. Hoffman EJ. Selin ©. Sokoloff L.. Kuh! DE. Tomographic measurement of local cerebral glu- cose metabolic rate in humans with (E-1812-fuore-2-deory - d-glucose: validation of method. Ana Neurol 1979.6: 37 1-88. 26 27. 0 ys3 Hawkins RA, Phelps ME. Huang SC. Kuhl DE. Effect of ischentia on quantification of local cerebral glucose meta- bohe rate in mao. J Cereh Blood Flow Metab (9X11: 37-51 Crane PD. Pardridge WM. Braun 0D. Nyerges AM. Old- endort WH. The interaction of transport and metabolism on brain glucose utilization: a reevaluation of the lumped con- stant. J Neurochem (98136: 1601-4 . Wienhard K. Pawlik G. Eriksson b. et al) Kinetic constants of cerebral glucose metabolism in pathological conditions. J Cereh Blood Flow Metubol 983.MSuppl lava —§ ~ Di Chiro G, DeLaPas RE. Brooky RA. et al) Glucose utth- gation of cerebral ghomas measured by CF Mluorodeoys- glucose and positron emission tomography. Neurology ONY) 1982232:1423-9 Perot Po Weir B. Rasmussen T. Tuberous sclerosis, Arch Neurol 1966: 18:498 -S06, FD Comput Avast Lomour Vel 7 Ne 6 198d