ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 12, No. 2, 89-94, 1998 

Clinical significance of reduced cerebral metabolism in multiple sclerosis: A combined PET and MRI study 

Xiayan SUN,* Makoto TANAKA,* Susumu KoNDo,* Koichi OKAMOTO* and Shunsaku HIRAI** 
*Department of Neurology, Gunma University School of Medicine * *Tokyo Metropolitan Neurological Hospital 

Magnetic resonance imaging (MRI) in patients with multiple sclerosis (MS) has provided major insights into the disease's natural history, and many studies have focussed on possible correlations between MRI findings and the clinical manifestations of MS. In contrast, there are few reports on possible relationships between functional imaging data and cognitive function. The present study assessed the relationship between clinical presentation and combined anatomical and functional imaging data in MS. Twenty patients with definite MS underwent MRI and positron emission tomography (PET) to evaluate cerebral blood flow (rCBF) and oxygen metabolism (rCMO2). The relationships between these neuroimaging findings and clinical data, including the Expanded Disability Status Scale (EDSS), Mini-mental status scale, Hasegawa Dementia Scale and relapse time, were evaluated with Spearman's rank correlation coefficients. A general reduction in rCBF and rCMRO2 in the gray and white matter were found in the MS patients. EDSS was correlated with the number and size of the lesions on MRI and was negatively correlated with rCMRO2. A correlation between the decrease in rCMRO2 and the level of cognitive impairment was also found. The severity of cerebral hypometabolism was also related to the number of relapses. Morphological and functional findings obtained by MRI and PET are closely related to the clinical status in MS. Our results suggest that measurement of cerebral metabolism in MS has the potential to be an objective marker for monitoring disease activity and to provide prognostic information. 

Key words: multiple sclerosis, cerebral oxygen metabolism, magnetic resonance imaging, hypometabolism, prognosis 

INTRODUCTION 
THE NATURAL HISTORY of multiple sclerosis (MS) is not fully known, although anatomic imaging studies with magnetic resonance imaging (MRI) in MS patients have provided major insights into the disease.1-3 The possible correlation between the level of disability and the lesion burden detected on MRI indicates that this imaging method might serve as an objective clinical parameter for monitoring therapy,4-6 but there is pathophysiological 

Received October 1 , 1997, revision accepted December 22, 1997. For reprint contact: Makoto Tanaka, M.D.. Department of Neurology. Gunma University School of Medicine, Showa-machi, Maebashi. Gunma 371-85 11 , JAPAN. 

heterogeneity of lesions, so that all appear the same on conventional MRI.3 Investigation of the functional pathophysiological impairments in relation to the lesion load detected by MRI in MS patients should improve our understanding of the disease. A functional imaging study with positron emission tomography (PET) previously demonstrated generalized hypometabolism in the gray and white matter of MS patients.7 Another study with single photon emission computed tomography (SPECT) showed a regional reduction in the frontal lobes and the left temporal lobe.8 Both of these studies gave important information about metabolic or blood flow abnormalities of the brain in MS patients and showed that cerebral hypometabolism was well correlated with cognitive impairment, but no previous study has explored both quantitative functional imaging data and the clinical status of the patients. The present work was initiated to determine the relationship between brain hypometabolism and the clinical features of MS, as well as to reexamine correlations between the lesion load on MRI and cognitive function and phsical disability . 

MATERIALS AND METHODS 
Patients Twenty patients ( 10 women and 10 men) with clinically definite multiple sclerosis 9 (mean age +- SD; 38.6 +- 11.1 years) were studied after giving informed consent. The mean total disease duration was 108.2 +- 69.2 months. The clinical course of the patients is shown in Table 1 . Eighty percent of the patients had relapsing and remitting MS. Fourteen age-matched normal controls (43.6 +- 13.3 years) also participated in the PET study. PET and MRI were performed in all patients during remission. Disability was scored with Kurtzke's Expanded Disability Status Scale (EDSS).10 The mean EDSS score was 2.7 +- 1.4, indicating that this group of patients had mild to moderate disability. Cognitive function was assessed with the Mini-mental status scale (MMS)11 and the Hasegawa Dementia Scale (HDS; full score, 32.5 points).12 The mean MMS and HDS scores were 26.8 +- 2.8 and 27.9 +- 3.7, respectively, suggesting the existence of focal cognitive impairment. These scales were assessed just before the PET study. The number of relapses during follow-up after PET and MRI was determined in 16 patients with a follow-up period of over one year. The total number of relapses was deter-mined in 17 patients with a total disease duration of over one year. Relapse was defined as the appearance of a new symptom or deterioration of existing symptoms and signs for at least 24 hours, after excluding involvement of other diseases. The mean annual relapse rate during the total disease period and during follow-up was 0.43 and 0.52, respectively. 

Magnetic resonance imaging 
T2-weighted and proton MRI of the brain was performed in 16 patients with a Siemens system operating at 1.5 T (SE 2,400/90 and 2,400/20, 256 x 256 image matrix). Contiguous 7-mm thick slices of the whole brain were obtained from the foramen magnum to the higher convexity. The remaining four patients underwent T2-weighted and proton MRI with a 0.5 T Philips Gyroscan TS-NT (SE 2,500/90 and 2,500/20, 256 x 256 image matrix). A computerized image analysis system was used to quantify the total areas of lesions detected by MRI. Lesion borders were delineated on all axial slices displayed on a computer monitor with a mouse cursor and the lesion area per slice was calculated with an image analysis software. The total number of lesions was also counted in the same pictures, so that the total area and number of lesions were obtained for each patient. Confirmation of the results was done by another doctor independently. 

Positron emission tomography 
PET was performed with a PCT-HI (Hitachi, Japan), which has four rings with 128 bismuth germanate oxide detectors providing seven views per scan cycle.13,14 The best spatial resolution was 7.4 mm (full width at half maximum) at the center of the scanning field, and the axial resolution at the center was 16 mm (the slice thickness). Prior to the PET study. CT scanning (CT-HSF, Hitachi) was done to identify the anatomical structures in the PET views. The PET and CT scanners were located side by side and joined by rails on which the bed moved with the patient's position fixed, thus providing identical tomographic planes. The patient' s head was fixed to the headrest to obtain a tomographic plane parallel to the orbitomeatal line. The initial positioning and the absence of head tilt during the scan were ensured by crossed beams projected onto ink marks drawn on the patient's face. A 10-min transmission scan with germanium-68 and gallium-68 was performed for attenuation correction. Calibration factors between the PET scanner and the well counter were obtained in each study. The oxygen- 1 5 steady state technique was employed to measure regional cerebral blood flow (rCBF) and the regional cerebral metabolic rate of oxygen (rCMRO2).13,15 Oxygen-15 gas (750-1 , 100 MBq/min) and then oxygen-15 labeled carbon dioxide (350-550 MBq/min) were inhaled continuously while scan data were collected for 5-8 minutes. Blood samples were obtained 3 times during each scanning process from a cannula in the radial artery, and were used for determination of oxygen-15 radioactivity in whole blood and plasma as well as for blood cell counts and blood gas analysis. Correction for intravascular oxygen-15 labeled oxyhemoglobin was performed by means of bolus inhalation of oxygen- 15-labeled carbon monoxide. 15 
Image data were processed in a Hitachi image process-ing computer with system subroutines to reconstruct functional images of 128 x 128 pixels. These images were visually evaluated for any focal or regional reduction in rCBF and rCMRO2. Then regions of interest (ROIs) were set in the bilateral cerebral cortices and the cerebellar hemispheres according to the CT images.14 On each slice, cortical ribbons of frontal, temporal, parietal, and occipital lobe cortices, and outlines of the cerebellar hemispheres were traced with a track ball, avoiding the sinuses and cerebrospinal fluid spaces. The same ROIs were then superimposed on the rCBF and rCMRO2 tomographic planes identical to the CT image. The mean rCBF and rCMRO2 of the gray matter were calculated from the ROIs of all cerebral cortices, and were designated as G-CBF and G-CMRO2, respectively . The mean rCBF and rCMRO2 for white matter were obtained from the right and left centrum semiovale, and were designated as W-CBF and W-CMRO2 respectively .
 
Statistical procedures 
The mean age, rCBF, and rCMRO2 Values for the MS patients were compared with those of the controls and statistical differences were determined by Student's t-test. Correlations of PET values with EDSS, MMS, HDS and relapsing time were evaluated by Spearman's rank correlation coefficient (SRCC) analysis. Correlations between the quantitative MRI data and the above clinical parameters were also evaluated by SRCC analysis. 

RESULTS 
Magnetic Resonance Imaging Findings Sixteen of the 20 MS patients had one or more lesions in the cerebral white matter of varying severity, but 4 had no lesion. The brainstem was involved in 7 patients and the spinal cord in 9 with or without white matter involvement. The total lesion load in the brain was quantified in terms of lesion area and lesion number; the mean area of the total lesions was 10.9 d: 13.0 cm2, and the mean number of lesions was 10.9 ~ 15.2. There were no concomitant abnormal findings, including brain atrophy. 

Positron Emission Tomography Findings 
A generalized significant reduction in rCBF and rCMRO2 in the cerebral gray matter was found in the MS patients, with a 28% reduction in G-CBF and 27% decrease in G-CMRO2 as compared with the normal controls. These changes were paralleled by a reduction of 26% in W-CBF and 16% in W-CMRO2 (Figure 1 ). Similar changes were found in the rCBF and rCMRO2 of the cerebellar hemispheres in the MS patients (data not shown). Visual analysis of the gray matter of rCBF and rCMRO2 revealed no regional decrease in the patients compared with the normal controls (data not shown), indicating that diffuse pathophysiological impairment associated with white matter lesions was a general finding in the MS group, but one patient who had acalculia showed definite left temporoparietal lobar hypometabolism. 

Correlations between Imaging Data and Clinical Findings (Table 2) 
The relationship between imaging data and clinical findings was assessed by SRCC analysis. A strong negative correlation was found between the lesion load on MRI and the cognitive scores (SRCC = -0.677 and p < 0.01 for MMS; SRCC = -0.8 16 and p < O.OO1 for HDS). In addition, a positive correlation was found between gray matter metabolism and cognitive function (Figure 2). The number of lesions on MRI was correlated with the severity of clinical disability (SRCC = 0.542 and p < O.05). Further-more, the total lesion area on MRI was correlated with the EDSS score (SRCC = 0.486 and p < 0.05), and a coupled decrease in G-CBF and G-CMRO2 in the gray matter of MS patients was negatively correlated with the EDSS score (SRCC = -O.560 and p < 0.05), indicating a relation-ship between cerebral metabolism and clinical disability. The relationship between cerebral metabolism measured by PET and relapse time during follow-up (ranging from 21 to 83 months) or the total disease period (ranging from 22 to 248 months) was analyzed. G-CMRO2 was significantly correlated with the number of relapses both during follow-up and during the total disease period (Figure 3), but no relationship between G-CBF and re-lapse was found. Cognitive function, physical disability and relapse time were not correlated with W-CBF or W-CMRO2.
 
DISCUSSION 
MRI is clinically useful in the diagnosis of MS. Clinicopathological examination of demyelinated lesions with a high signal intensity on T2-weighted images in the brains of MS patients has indicated the reliability and validity of lesion measurement by MRI,16 Previous studies have extensively investigated the relationship between MRI lesions and the clinical status of MS patients, but because controversy has arisen as a consequence of the wide range of disease presentation and the limitations of MRI,17 a well-defined relationship between functional abnormalities of the MS brain and the extent and sites of lesions on MRI remains to be established. In the present study, PET confirmed the global reduction in cerebral rCBF and rCMRO2 in MS patients. This hypometabolism is unlikely to be explained solely by atrophy of the central white matter and cortical gray matter,7 since no atrophy of the brain was detected by MRI in our patients, but the frequent occurrence of plaques (especially in the periventricular area) was noted, so that white matter lesions appeared to influence cortical metabolism. A few previous studies have pointed out that white matter lesions can be associated with reduced cerebral metabolism.18 In MS, the white matter plaques represent myelin breakdown and gliosis.19 These lesions seem to disturb axonal transport and to exert a depressant effect on the cerebral cortex. The relapse of MS superimposed on a generalized disease process affecting multiple white matter regions would aggravate cerebral hypometabolism. This global (i.e., non-focal) reduction in brain metabolism, obviously recognized with visual inspection and shown in Figure l, suggests that cortical functional changes in MS are likely to be caused not only by the white matter lesions identified by MRI, but also by functional disturbance in the normal-appearing white matter. This is partly supported by Husted et al. who have shown biochemical alterations in MS lesions and normal-appearing white matter detected by magnetic resonance spectroscopic methods.20 Alternatively, cerebral metabolic hypoactivity detected by PET might represent a consequence of autoimmune changes in the brain that are independent of the visible lesion(s). The fact that no regional metabolism abnormality was found in the whole population agreed with the results reported by Brooks et al.7 In this study, a strong correlation was shown between the lesion burden (in terms of total lesion area and total lesion number) on T2-weighted MRI and cognitive scores. These findings extend previous results by demonstrating the usefulness of the screening tests for the evaluation of cognitive function in association with MRI data in MS.21-23 Employment of a rapid dementia scale as a routine test for MS should be practical and is clinically important since cognitive impairment occurs in 13%(; to 65% of MS patients.24 A correlation between lesion burden on MRI and the disability score was also found in this study, a result that conflicted with some previous reports.25 Differences in the selection of patients and technical considerations can readily explain these discrepancies. First, studies of a patient group more homogeneous in the clinical course seem to give clearer correlations in MS. In our patient group, 80% showed a relapsing-remitting course (Table 1 ). Another recent study has also reported that relapsing and remitting MS is likely to show a strong association between the MRI lesion load and the degree of disability.6 Second, the field strength of the MRI scanner seems to have a great influence upon the detection of plaques.18 Therefore, our MRI scanner (operating at 1.5 Tesla) may have been more accurate in detecting and quantitating the lesion load. The correlation between lesion burden and the degree of disability suggests that accumulated brain lesions in the white matter play a very important role in causing physical functional impairment in MS. In the present study, a correlation between cerebral cortical metabolism and the disability score was shown for the first time, as opposed to previous findings.7 It is difficult to compare our data quantitatively evaluating physical disability with the previous study that only graded the severity of locomotor dysfunction.7 More global and quantitative assessment of physical disability should contribute to a better understanding of the natural history of MS because of the wide variations in clinical presentation. We suggest that the white matter lesion(s) might be one of the factors which cause diffuse cerebral hypometabolism responsible for clinical disability. Another important finding of this study was that the decreased gray matter metabolism was related to relapse during follow-up and during the total disease period. This correlation between metabolic disturbance and relapse after PET (in the follow-up period) seems to be of clinical importance, suggesting that measurement of cerebral metabolism in MS has the potential to be an objective marker for providing prognostic information as well as monitoring disease activity. In contrast, there was no relationship between relapse and MRI data, although the reason for this was not clear. Reduced global cerebral metabolism indicates wide-ranging disturbance of neuronal function, which is probably caused by the disease process in the white matter. Therefore, the reduction in cortical metabolism may reflect functional derangement in the white matter which MRI cannot find out before relapsing of the disease occurs. These results suggest that PET might have the potential to be an objective marker for monitoring disease activity. We also found that there was no correlation between cerebral rCBF and the above clinical parameters, but the reason remains unclear. The direct correlation between cerebral hypometabolism and relapse time could be a reflection of a basic cerebral metabolic defect due to a number of factors, rather than a secondary effect of the decreased rCBF. Finally, cerebral blood flow and oxygen metabolism in the white matter, which seem to directly reflect the disease process, were not correlated with the clinical parameters. The ROIs for W-CBF and W-CMRO2 were placed in the centrum semiovale, which may show the demyelinating lesion(s) on MRI. Because of technical limitations, the ROI included the demyelinating lesion(s) in some patients and did not in others. Therefore, the W-CBF and W-CMRO2 may not have represented overall cerebral perfusion and metabolism in the white matter. Improved spatial resolution of the PET could reveal the relationship between cerebral blood flow and metabolism in the white matter and clinical status in MS patients. 

REFERENCES 
l. Lukes SA. Crooks LE. Aminoff MJ. Kaufman L. Panitch HS, Mills C, et al. Nuclear magnetic resonance imaging in multiple sclerosis. Ann Neurol 13: 592~501 , 1983. 
2. McDonald WI. The dynamics of multiple sclerosis. JNeurol 240: 28-36, 1993. 
3. Miller DH, Albert PS. Barkhof F, Francis G. Frank JA, Hodgkinson S, et al. Guidelines for the use of magnetic resonance techniques in monitoring the treatment of multiple sclerosis. Ann Neurol 39: 6-16, 1996. 
4. Filippi M. Horsfield MA, Morrissey SP, MacManus DG, Rudge P, McDonald WI, et al. Quantitative brain MRI lesion load predicts the course of clinically isolated syn-dromes suggestive of multiple sclerosis. Neurology 44: 635-641, 1994. 
5. Khoury SJ, Guttmann CR, Orav EJ, Hohol MJ. Ahn SS. Hsu L, et al. Longitudinal MRI in multiple sclerosis: correlation between disability and lesion burden. Neurology 44: 2120-2124, 1994. 
6. Filippi M, Paty DW. Kappos L. Barkhof F, Compston DAS, Thompson AJ, et al. Correlations between changes in disability and T2-weighted brain MRI activity in multiple sclerosis: a follow-up study. Neurology 45: 255-260, 1995. 
7. Brooks DJ, Leenders KL, Head G, Marshall J, Legg NJ, Jones T. Studies on regional cerebral oxygen utilisation and cognitive function in multiple sclerosis. J Neurol Neurosurg Psychiatry 47: 1182-1191 , 1984. 
8. Pozzilli C. Passafiume D, Bemardi S, Pantano P, Incoccia C, Bastianello S, et al. SPECT, MRI and cognitive functions in multiple sclerosis. J Neurol Neurosurg Psychiatry 54: 110-115, 1991. 
9. Poster CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 13: 227-231, 1983. 
10. Kurtzke JK. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 33: 1444-1452, 1983. 
11. Folstein MF, Folstein SE. McHugh PR. Mini-Mental State: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12: 189-198, 1975. 
12. Hasegawa K. The clinical assessment of dementia in the aged; a dementia screening scale for psychogeriatric patients. In Aging in the Eighties and Beyond, New York, Springer-Verlag, pp. 207-218, 1983.
13. Tanaka M, Kondo S, Hirai S, Ishiguro K, Ishihara T, Morimatsu M. Crossed cerebellar diaschisis accompanied by hemiataxia: a PET study. J Neurol Neurosurg Psychiatry 55: 121-125, 1992. 
14. Sun X, Tanaka M, Kondo S, Hirai S. Ishihara T. Reduced cerebellar blood flow and oxygen metabolism in spinocerebellar degeneration: a combined PET and MRI study. J Neurol 241: 295-300, 1994. 
15. Frackowiak RS. Pozzilli C, Legg NJ, DuBoulay GH, Marshall J, Lenzi GH, et al. Regional cerebral oxygen supply and utilization in dementia. A clinical and physiological study with oxygen-15 and positron tomography. Brain 104: 753-778, 1981. 
16. Estes ML, Rudick RA, Barnett GH, Ransohoff RM. Stereo-tactic biopsy of an active multiple sclerosis lesion: Immunocytochemical analysis and neuropathologic correlation with magnetic resonance imaging. Arch Neurol 47: 1299-1303, 1990. 
17. Filippi M, Campi A, Mammi S, Martinelli V, Locatelli T. Scotti G, et al. Brain magnetic resonance imaging and multimodal evoked potentials in benign and secondary progressive multiple sclerosis. J Neurol Neurosurg Psychiatry 58: 31-37, 1995. 
18. DeCarli C, Murphy DGM. Tranh M. Grady CL, Haxby JV, Gillette JA, et al. The effect of white matter hyperintensity volume on brain structure, cognitive performance, and cerebral metabolism of glucose in 5 1 healthy adults. Neurology 45: 2077-2084, 1995. 
19. Mattews WB. Pathology of multiple sclerosis. In McAlpine's Multiple Sclerosis: Mattews WB, Acheson ED, Batchelor JR, Weller RO (eds.), New York, Churchill-Livingstone, pp. 301-341, 1985. 
20. Husted CA, Goodin DS, Hugg JW, Maudsley AA, Tsuruda JS, deBie SH, et al. Biochemical alterations in multiple sclerosis lesions and normal-appearing white matter detected by in vivo 31P and 1H spectroscopic imaging. Ann Neurol 36: 157-165, 1994. 
21 . Beatty WW. Goodkin DE. Screening for cognitive impairment in multiple sclerosis: an evaluation of the mini-mental state examination. Arch Neurol 47: 297-301 , 1990. 
22. Anzola GP, Bevilacqua L, Cappa SF, Capra R, Faglina L, Farina E, et al. Neuropsychological assessment in patients with relapsing-remitting multiple sclerosis and mild functional impairment: correlation with magnetic resonance imaging. J Neurol Neurosurg Psychiatry 53: 142-145, 1990. 
23. Comi G. Filippi M, Martinelli V, Sirabian G, Visciani A, Campi A, et al. Brain magnetic resonance imaging correlates of cognitive impairment in multiple sclerosis. J Neurol Sci 115 (suppl): S66-S73, 1993. 
24. Peyser JM, Rao SM, LaRocca NG, Kaplan E. Guidelines for neuropsychological research in multiple sclerosis. Arch Neurol 47: 94¥97, 1990. 
25. Thompson AJ. Kermode AG, MacManus DG, Kendall BE, Kingsley DRE, Moseley IF, et al. Patterns of disease activity in multiple sclerosis: clinical and magnetic resonance imaging study. Br Med J 300: 631-634, 1990.