ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol, 11, No. 3, 251-257, 1997 
Differentiating between multiple system atrophy and Parkinson's disease by positron emission tomography with 18F-dopa and 18F-FDG 
Makoto OTSUKA,* Yasuo KUWABARA,** Yuichi ICHIYA,*** Sinichi HOSOKAWA,**** Masayuki SASAKJ,** Tsuyoshi YOSHIDA,** Toshimitsu FUKUMURA,** Motohiro KATO***** and Kouji MASUDA** 
*Department of Radiology. Medical Institute of Bioregulation. Kyushu University Departments of **Rediology, ****Neurology, and *** **Clinical Neurophysiology, Faculty of Medicine, Kyushu University ***Department of Radiology. National Kyushu Cancer Center 
Both the striatal 18F-dopa uptake and brain glucose metabolism were studied by PET with 6-L-[18F]fluorodopa (FD) and [18F]fluorodeoxyglucose (FDG) in 9 patients with multiple system atrophy (MSA) and 15 patients with idiopathic Parkinson's disease (PD). Five of the 9 MSA patients were diagnosed as having olivopontocerebellar atrophy, whereas 2 had striatonigral degeneration and 2 demonstrated Shy-Drager syndrome. The FD uptake ratios to the occipital cortex in the MSA patients at 120 min after the administration of FD were 2.07 +- 0.31 (mean +- SD) and 1.96 +- 0.29 in the caudate and the putamen, respectively, and decreased compared to those in the controls (2.72 +- 0.11, 2.71 +- 0.10). The same ratios in the PD patients were 2.07 +- 0.36 and 1.74 +- 0.24, respectively, which also decreased, but the decreased uptake in the putamen was more prominent. The caudate-putamen index (CPI) (%), which was calculated by a formula based on the difference in the uptakes in the caudate and putamen divided by the caudate uptake, indicated 5.6 +- 4.6 in the MSA patients and 14.8 +- 5.4 in the PD patients. The CPI for all PD patients was more than 7.0, which was the mean + 2SD for the controls, but the CPI for 3 MSA patients was more than 7.0 (accuracy: 88%). The glucose metabolic rates for each region in the PD patients showed no difference from the normal controls. The frontal and the temporal cortical glucose metabolism and the caudate, the putaminal, the cerebellar and the brainstem glucose metabolism in the MSA patients decreased significantly in comparison to those in the controls. But, as the glucose metabolic rates in such regions of each patient overlapped in the two groups, the accuracy of the FDG study for differentiation was lower than that of the FD study. The putaminal glucose metabolic rates, for example, in 3 PD patients were less than 6.8 (mg/min/100 ml), which was the mean - 2SD for the controls, while those in 3 MSA patients were more than 6.8 (accuracy: 75%). In addition, the combination of these two methods slightly improved the accuracy. The glucose metabolism is useful for evaluating the regional metabolic activity of the brain, and the FD study, which is specific to the dopamine system, seems to be more useful for differentiating between MSA and PD. 
Key words: multiple system atrophy; Parkinson's disease; 18F-dopa uptake; glucose metabolism 
INTRODUCTION 
MULTIPLE SYSTEM ATROPHY (MSA) is clinically characterized by a poorly L-dopa responsive akinetic-rigid syn-
Received April 28, 1997, revision accepted July 4, 1997. 
For reprint contact: Makoto Otsuka, M.D.. Department of Radiology, Medical Institute of Bioregulation. Kyushu University, 4546 Tsurumihara, Beppu 874, JAPAN. 
drome, and includes striatonigral degeneration (SND), olivopontocerebellar atrophy (OPCA) and Shy-Drager syndrome (SDS) in its spectrum. Its pathology is distinct from idiopathic Parkinson's disease (PD), and demonstrates neuronal degeneration in the absense of inclusion bodies in the caudate and putamen, the substantia nigra, cerebellar Purkinje cells, inferior olives, and the intermediolateral columns of the spinal cord.1,2 In spite of the clear pathological differences between MSA and PD, these akinetic-rigid syndromes can cause diagnostic difficulties in their early clinical stages.3 
Positron emission tomography (PET) studies with [18F]fluorodeoxyglucose (FDG) have demonstrated differences between MSA and PD cases. Striatal hypometabolism in cases of SND has been reported in comparison with the normal glucose metabolism in cases of PD.4,5 We previously reported that the glucose metabolism decreased in the frontal, temporal and parietal cortices, as well as in the caudate and putamen, the cerebellum and the brainstem in patients with MSA, but it was preserved in patients with PD.6 Nevertheless, it does not seem to be easy to differentiate the two akinetic-rigid syndromes based on the glucose metabolic values because of the relatively large inter-patient variations.6 PET studies with 6-L-[18F]fluorodopa (FD) also have been used to evaluate the nigrostriatal dopaminergic function, and it has also been reported that the FD uptake patterns of MSA and PD patients were different. The FD uptake decreased substantially in the putamen, but it was relatively spared in the caudate nucleus in patients with PD7-11 in comparison with the homogeneously decreased FD uptake in the caudate and putamen in patients with MSA.10 We previously reported that patients who were tentatively diagnosed as atypical parkinsonism including SND, SDS and pure akinesia also showed uniformly reduced FD uptakes in both the caudate and the putamen,11 but it has yet to be elucidated which PET study with FD or FDG is more useful in differentiating these two akinetic-rigid syndromes. 
In this study, we performed PET studies with both FD and FDG in patients with MSA or PD in order to evaluate which study appears to be more useful in differentiating between these two akinetic-rigid syndromes. 
MATERIALS AND METHODS 
Nine patients with MSA were studied. Five of them were diagnosed as having OPCA, while 2 had SND and 2 demonstrated SDS (Table 1). All patients had poorly or non-L-dopa-responsive akinetic-rigid syndrome. Five patients with OPCA were diagnosed based on a modification of Greenfield's classification12 and were referred to as cases #6, #8, #9, #10 and #11 in our previous paper. 13 Two patients presented progressive limb rigidity and bradykinesia associated with an early impairment of balance and gait. Their clinical findings were compatible with SND.14 They were also referred to as cases #12 and #15 in another previous paper.ll The other two patients had autonomic failure and cerebellar ataxia. Their clinical findings were compatible with SDS,1 Case 8 was described as case #13 in our previous paper.11 
Fifteen patients with PD were also studied. All patients had L-dopa-responsive akinetic-rigid syndrome without any evidence of cerebellar ataxia, pyramidal sign, or autonomic failure (Table 2). They were referred to as cases #1 , #2, #3, #4, #5, #6, #7, #8, #9, #11, #12, #13, #14, #16 and #17 in our previous paper.15 None of the PD patients were demented. 
Eight normal volunteers with a mean age of 50.1 (SD l3.8) years underwent FDG studies, and 8 normal volunteers with a mean age of 41.1 (SD 17.3) years underwent FD studies. None of the volunteers had any neurological deficits. Normal brain CT scans were obtained from volunteers over 50 years old. 
PET was performed with the HEADTOME III (Shimadzu Corp., Japan) at a spatial resolution of 8.2 mm full width at half maximum (FWHM).16 A transmission scan with a 68Ge/68Ga ring source was obtained for a correction of attenuation. Five slices were taken in the orbitomeatal line (OM) + 2 cm, + 3.5 cm, + 5 cm, + 6.5 cm and + 8 cm planes by single scanning. The regions-of-interest (ROIs) were visually placed or outlined by referring to the MRI images as described previously.11,17 The regions of the cerebellum and brainstem were placed in the OM + 2 cm plane, the region of the temporal cortex was placed in the OM + 3.5 cm plane, the regions of the frontal cortex and thalamus were placed in the OM + 5 cm plane, the region of the occipital cortex was placed in the OM + 6.5 cm plane, and the region of the parietal cortex was placed in the OM + 8 cm plane, respectively. The regions of the caudate nucleus and putamen were outlined in the OM + 5 cm plane on the glucose metabolic images and then transferred to the FD images. 
The regional cerebral metabolic rate for glucose was measured with FDG. A single injection of 6.3 to 9.3 mCi (230 to 340 MBq) of [18F]fluorodeoxyglucose was administered intravenously and arterial blood samples were obtained. The glucose metabolic rates were determined from the emission scan data obtained between 63 and 71 minutes post injection and blood curve data by using the model of Sokoloff et al.18 and Phelps et al,19 which was later modified by Brooks.20 
The FD study was performed by a single injection of 2.9 to 6.4 mCi (110 to 240 MBq) of 6-L-[18F]fluorodopa. The PET data were obtained between 112 and 127 minutes post injection (with a mean time of 120 minutes post injection). The ratio of the caudate or the putamen to the occipital cortex at 120 minutes was evaluated, since we previously showed that the ratio method at 120 minutes was comparable to the graphical analysis for estimating the presynaptic dopaminergic function.21 The caudateputamen index (CPI) (%), which was calculated by a formula based on the difference in the uptakes of the caudate and putamen divided by the caudate uptake, was also evaluated.15 
All subjects were examined with their eyes open during the PET studies. The glucose study and the FD study were performed with intervals of two to 23 days in each patient. All subjects received a full explanation of the diagnostic procedure and gave their informed consent. 
The mean values for the bilateral side in all cases were evaluated for the glucose metabolism, since the values for glucose metabolism did not decrease (or increase) in either the prominently affected side or the opposite side. On the other hand, since the FD uptake values on the opposite side of the brain for the prominently affected body side decreased more than on the other side, the contralateral brain side in the cases demonstrating asymmetry in their symptoms and the mean of the bilateral side in the other cases showing no asymmetry in their symptoms were evaluated for the FD uptakes. The statistical analyses were done by one factor ANOVA with the Newman-Keuls multiple comparison test. All PET studies were performed after clinical diagnosis, and no diagnostic changes were attributed to the PET results. 
RESULTS 
The ratios of caudate and the putaminal FD uptake to the occipital cortex in the patients with MSA and PD decreased in comparison to the normal controls (Table 3). The CPI in the patients with PD significantly increased in comparison to both the MSA patients and the controls. The CPI for all MSA and PD patients are shown in Fig. 1 . The CPI for all PD patients were more than 7.0, which was the mean + 2SD for the controls and is shown as a dotted line in Fig. 1 . If the patients with a CPI of less than 7.0 are classified as MSA cases and those with a CPI of more than 7.0 are classified as PD cases, then three MSA patients would thus have been misclassified, and the total accuracy would therefore be 88%. 
The glucose metabolic rates for each region in the PD patients showed no difference from those for the normal controls. The frontal and temporal cortical glucose metabolic rates and the caudate, the putaminal, the cerebellar and the brainstem glucose metabolic rates in the MSA patients significantly decreased in comparison to the controls. In addition, the glucose metabolic rates in the caudate and putamen, the brainstem and the cerebellum in the patients with MSA were significantly decreased in comparison to the patients with PD (Table 4). Since the glucose metabolic rates in both the brainstem and the cerebellum seem to be mainly due to the atrophy of such regions,6 the glucose metabolic rates in the caudate and the putamen thus seem to be more reliable for differentiating between MSA and PD. The glucose metabolic rates in the caudate and in the putamen for all the patients are shown in Figs. 2a and 2b, respectively. The glucose metabolic rates in the putamen in 3 MSA patients were more than 6.8 (mg/min/100 ml), which was the mean -2SD for the controls and are shown as a dotted line in Fig. 2b, but the same rates in 3 PD patients were less than 6.8. If the two groups are thus classified on the basis of the 6.8 value, these 6 patients would be misclassified, and the total accuracy would be 77%. If the differentiation line between the two groups is placed at a value of 6.2, which is shown by the solid line in Fig. 2b, then 5 MSA patients would be misclassified, and the total accuracy would be 79%. The caudate glucose metabolic rates for the patients in the two groups overlapped more closely than the putaminal glucose metabolism, and the total accuracy of the differentiation between the two groups was thus either the same or worse (Fig. 2a). 
The correlation between the CPI and the putaminal glucose metabolism is plotted in Fig. 3. The patients in the two groups still overlapped, but if the differentiation line between the two groups is drawn as the solid line in Fig.3, only one PD and one MSA patients would be misclassified, and then the total accuracy would be 92%. 
DISCUSSION 
MSA showed a homogeneously reduced FD uptake in the caudate and the putamen, but the reduced FD uptake in the putamen was more prominent than in the caudate in patients with PD in this study. These findings are consistent with previous PET reports10,ll and the pathological findings showing that the dopaminergic neurons in the ventrolateral portion of the substantia nigra projecting to the putamen decreased more than the neurons in the dorsomedial portion of the substantia nigra projecting to the caudate nucleus in patients with PD.22 FD studies can allow for the visualization of the capacity of dopamine storage in the caudate and putamen in vivo, and have been shown to correlate with both the dopamine cell counts in the substantia nigra and the dopamine levels in the caudate and putamen.23 PET studies showed a more reduced FD uptake in the putamen than in the caudate in patients with PD, and this finding was compatible to the relatively preserved dopamine level in the caudate head in autopsy studies.24-26 The greater cell loss of the substantia nigra in patients with SND than in patients with PD in autopsy studies22 might have caused the homogeneously reduced FD uptake in the caudate and putamen in patients with SND, whereas the degree of cell loss was still relatively preserved in the dorsomedial portion of the substantia nigra which is connected to the caudate nucleus.22 
In this study, the CPI seems to be able to differentiate between MSA and PD patients. All PD patients demonstrated a high CPI and only three MSA patients with a high CPI were misclassified and the total accuracy for the differentiation between the two groups was therefore 88%. The reason why the three MSA patients had a high CPI is uncertain, but since the three MSA patients with a high CPI were diagnosed as having OPCA and another OPCA patient also had a moderate CPI (6.70), it might thus have been due to the fact that they were patients with OPCA. 
Decreased glucose metabolic rates were observed in the caudate and putamen, in the frontal and the temporal cortices, in the brainstem and in the cerebellum in the MSA patients in the present study. De Volder et al.4 and Eidelberg et al.5 also reported that glucose metabolism in the caudate and putamen was decreased in SND patients, which is consistent with neural loss and gliosis involving the striatum in histopathological examinations in SND.14,27.28 The glucose metabolic rates for each region in the PD patients thus showed no difference in comparison to the normal controls. Since the frontal and the temporal cortical glucose metabolic rates in the PD patients slightly decreased compared to the controls, but the change was not statistically significant, the decreased glucose metabolic rates in only the caudate and putamen, the brainstem and the cerebellum in the patients with MSA were significantly different from the patients with PD. Moreover, as the glucose metabolic rates in the brainstem and in the cerebellum seem to mainly be due to the atrophy of such regions,6 the glucose metabolic rates in the caudate and the putamen therefore seem to be more reliable for differentiating between MSA and PD. The glucose metabolic rates in the putamen in each patient of the two groups overlapped to such a degree that the accuracy of the putaminal glucose metabolic rates in differentiating between the two groups was lower than that obtained with the CPI. 
Although critics may say that the clinical stages of PD patients were lower than those of MSA patients in this study, the CPI was not correlated with the clinical stage in the previous paper15 and MSA patients with higher clinical stages would show more decreased metabolism4,5 than PD patients with lower clinical stages, so that differentiation between the two groups by means of glucose metabolism should be easier, but a future study with a greater number of patients is still called for. The age range of the patients in this study was relatively large. The patient's age, however, does not seem to affect the results, since the FD uptake was reported no decline with age29 and the cerebral glucose metabolism was also reported no decline in the carefully screened healthy subjects.30,31 
Striatal receptor binding studies for differentiating between SND and PD have also been reported32-34 and some of them may possibly be even more sensitive tools, but such studies have so far only been performed at a small number of institutes. 
In conclusion, our findings suggest that glucose metabolism is useful for evaluating the regional total metabolic activity of the brain, but the FD study, which is specific to the dopamine system, seems to be more useful for differentiating between MSA and PD. In addition, the combination of these two methods may also potentially improve the overall accuracy. 
ACKNOWLEDGMENT 
The authors thank Dr. Brian T. Quinn for editorial assistance. This study was partly supported from the Japanese Ministry of Health and Welfare and from the Japanese Ministry of Education, Science and Culture. 
REFERENCES 
1. Spokes EGS, Bannister R, Oppenheimer DR. Multiple system atrophy with autonomic failure. Clinical, histological, and neurochemical observations on four cases. J Neurol Sci 43: 59-82, 1979. 
2. Kume A, Takahashi A, Hashizume Y, Asai J. A histometrical and comparative study on Purkinje cell loss and olivary nucleus cell loss in multiple system atrophy. J Neurol Sci 101: 178-186, 1991. 
3. Quinn N. Multiple system atrophy-the nature of the beast. J Neurol Neurosurg Psychiatry, special supplement 78-89, 1989. 
4. DeVolder AG, Francart J, Laterre C, Dooms G, Bol A, Michel C, et al. Decreased glucose utilization in the striatum and frontal lobe in probable striatonigral degeneration. Ann Neurol 26: 239-247, 1989. 
5. Eidelberg D, Takikawa S, Moeller JR, Dhawan V, Redington K, Chaly T, et al. Striatal hypometabolism distinguishes striatonigral degeneration from Parkinson's disease. Ann Neurol 33: 518-527, 1993. 
6. Otsuka M, Ichiya Y, Kuwabara Y, Hosokawa S, Sasaki M, Yoshida T, et al. Glucose metabolism in the cortical and subcortical brain structures in multiple system atrophy and Parkinson's disease: a positron emission tomographic study. J Neurol Sci 144: 77-83, 1996. 
7. Nahmias C, Garnett ES, Firnau G, Lang A. Striatal dopamine distribution in Parkinsonian patients during life. J Neurol Sci 69: 223-230, 1985. 
8. Martin WRW, Stoessl AJ, Adam MJ, Ammann W, Bergstrom M, Harrop R, et al. Positron emission tomography in Parkinson's disease: glucose and dopa metabolism. Adv Neurol 45: 95-98, 1986. 
9. Gamett ES, Lang AE, Chirakal R, Firnau G, Nahmias C. A rostrocaudal gradient for aromatic acid decarboxylase in the human striatum. Can J Neurol Sci 14: 444-447, 1987. 
10. Brooks DJ, Ibanez V, Sawle GV, Quinn N, Lees AJ, Mathias CJ, et al. Differing patterns of striatal 18F-Dopa uptake in Parkinson's disease, multiple system atrophy, and progressive supranuclear palsy. Ann Neurol 28: 547-555, 1990.
11. Otsuka M. Ichiya Y, Hosokawa S, Kuwabara Y. Tahara T, Fukumura T, et al. Striatal blood flow, glucose metabolism and 18F-Dopa uptake: difference in Parkinson's disease and atypical Parkinsonism. J Neurol Neurosurg Psychiatry 54: 898-904, 1991 . 
12. Adams RD, Victor M. Syndrome of progressive ataxia. In: Principles of Neurology. 4th ed. New York: McGraw-Hill, pp. 946-952, 1989. 
13. Otsuka M, Ichiya Y. Kuwabara Y, Hosokawa S. Akashi Y, Yoshida T, et al. Striatal 18F-Dopa uptake and brain glucose metabolism by PET in patients with syndrome of progressive ataxia. J Neurol Sci 124: 198-203, 1994. 
14. Fearnley JM, Lees AJ. Striatonigral degeneration: clinicopathological study. Brain 113: 1823-1842, 1990. 
15. Otsuka M, Ichiya Y. Kuwabara Y. Hosokawa S. Sasaki M, Yoshida T, et al. Differences in the reduced 18F-Dopa uptakes of the caudate and the putamen in Parkinson's disease: correlations with the three main symptoms. J Neurol Sci 136: 169-173, 1996. 
16. Kanno I. Miura S. Yamamoto S. Iida H, Murakami M, Takahashi K, et al. Design and evaluation of the positron emission tomograph: HEADTOME III. J Comput Assist Tomog 9: 931-939, 1985. 
17. Otsuka M. Ichiya Y. Kuwabara Y, Miyake Y. Tahara T, Masuda K, et al. Cerebral blood flow, oxygen and glucose metabolism with PET in progressive supranuclear palsy. Ann Nucl Med 3: 111-118, 1989. 
18. Sokoloff L, Reivich M, Kennedy M. DesRosiers MH, Patlak CS. Pettigrew KD, et al. The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: Theory, procedure and normal values in the conscious and anesthetized albino rat. J Neurochem 28: 897-916, 1977. 
19. Phelps ME, Huang SC, Hoffman EJ. Selin C. Sokoloff L, Kuhl DE. Tomographic measurements of local cerebral glucose metabolic rate in humans with [18F]2-fluoro-2-deoxy-D-glucose: Validation of method. Ann Neurol 6: 371-388, 1979. 
20. Brooks RA. Alternative formula for glucose utilization using labeled deoxyglucose. J Nucl Med 23: 538-539, 1982. 
21. Otsuka M, Ichiya Y. Kuwabara Y. Fukumura T, Sasaki M, Masuda K. Evaluation of the ratio method compared with graphical analyses for estimating nigrostriatal function in human 18F-dopa PET studies with or without carbidopa. Nucl Med Com 14: 862-867, 1993. 
22. Goto S. Hirano A, Matsumoto S. Subdivisional involvement of nigrostriatal loop in idiopathic Parkinson's disease and striatonigral degeneration. Ann Neurol 26: 766-770, 1989. 
23. Snow BJ, Tooyama I, MacGee EG, Yamada T, Calne DB, Takahashi H, et al. Human positron emission tomographic [18F]Fluorodopa studies correlate with dopamine cell counts and levels. Ann Neurol 34: 324-330, 1993. 
24. Bemheimer H. Birkmayer W, Hornykiewicz O, Jellinger K, Seitelberger F. Brain dopamine and the syndromes of Parkinson and Huntington. Clinical, morphological and neurochemical correlations. J Neurol Sci 20: 415-455, 1973. 
25. Kish SJ, Shannak K. Hornykiewicz O. Uneven pattern of dopamine loss in the striatum of patients with idiopathic Parkinson's disease. Pathological and clinical implications. N Engl J Med 318: 876-880, 1988. 
26. Gibb WRG. Lees AJ. Anatomy, pigmentation, ventral and dorsal subpopulations of the substantia nigra, and differential cell death in Parkinson's disease. J Neurol Neurosurg Psychiatry 54: 388-396, 1991 . 
27. Adams RD, van Bogaert L, van der Eecken H. Striato-nigral degeneration. J Neuropathol Exp Neurol 23: 584-608, 1964. 
28. Gibb WRG. Neuropathology of Parkinson's disease and related syndromes. Neurol Clin 10: 361-376, 1992. 
29. Eidelberg D. Takikawa S, Dhawan V, Chaly T. Robeson W, Dahl R, et al. Striatal 18F-DOPA uptake: Absence of an aging effect. J Cereb Blood Flow Metab 13: 88 1-888, 1993. 
30. Duara R. Margolin RA. Robertson-Tchabo EA, London ED, Schwartz M. Renfrew JW, et al. Cerebral glucose utilization, as measured with positron emission tomography in 21 resting healthy men between 21 and 83 years. Brain 106: 761-775, 1983. 
31. DeLeon MJ, George AE, Ferris SH, Christman DR, Fowler JS, Gentes CI, et al. Positron emission tomography and computed tomography assessments of the aging human brain. J Comput Assist Tomog 8: 88-94, 1984. 
32. Brooks DJ, Salmon EP, Mathias CJ, Quinn N, Leenders KL, Bannister R, et al. The relationship between locomotor disability, autonomic dysfunction, and the integrity of the striatal dopaminergic system in patients with multiple system atrophy, pure autonomic failure, and Parkinson's disease, studied with PET. Brain 113: 1539-1552, 1990. 
33. Shinotoh H, Inoue O, Hirayama K, Aotsuka A, Asahina M, Suhara T, et al. Dopamine Dl receptors in Parkinson's disease and striatonigral degeneration: a positron emission tomography study. J Neurol Neurosurg Psychiatry 56: 467-472, 1993.
34. Bum DJ, Mathias CJ, Quinn N, Marsden CD, Brooks DJ. Striatal opiate receptor binding in Parkinson's disease and multiple system atrophy: 11C-diprenorphine study. Neurology 43 (suppl): S454, 1992. 
35. Hoehn MM, Yahr MD. Parkinsonism: onset, progression, and mortality. Neurology 12: 427-442, 1967.