With Fluorine-18-Fluorodeoxyglucose And

Assessment of Disease Severity in Parkinsonism with Fluorine- 18-Fluorodeoxyglucose and PET

David Eidelberg, James R. Moeller, Tatsuya Ishikawa, Vijay Dhawan, Phoebe Spetsieris, Thomas Chaly, Wffliam Robeson, J. Robert DaM and Donald Margouleff

L@mnenLc ofNeumlo@, Medicine and Research, North Shore Univei@y Hospital/Cornell UniVeiYity Medical College, Manhasset, New York,- Department ofPsychiat,y, New York State P@sychiat,ic Institute, Colwnbia College of Physicians and Suigeons, New YorlÃ§New York

(2â€”4),FDOPA/PET provides a useful imaging marker of Auonne-18-fluorodeoxyglucose(FDG) and PET have been the PD process. However, kinetic PET measurements of usedto identifyanabnormalregionalmetaboliccovariancepat striatal K@FDare technically demanding, and have limited terninParldnson'sdisease(PD).Toexaminethepotentialuseof practicalapplicabilityin a broad clinical setting (5). More thiscovanancepatternas a metabolicimagingmarkerfor PD, over, striatal K@ is a neurochemical indicator of under we describethe TopographicProfileRating(TPR),whichis a lying nigrostriatal dopaminergic pathology which is in methodforcalculatingsubjectscoresforthispatterninindividual PD patients.We then assessthe relationshipbetweenthese sensitive to changes in a pharmacologic state. As such, metabolicmeasuresandobjectiveindependentdiseaseseverfty this technique is generally not suitable for assessing the ratings.Methods:TwoindependentgroupsofPDpatientswere effects of antiparkinsonian pharmacotherapeutic inter studiedwWiFDG-PEr.GroupA cOnsiStedof23patients(mean ventions. age602 Â±12.2;meanHoehnandYahrstages2.4Â±I .3)and In contrast to striatal K@, PET measurements of re GroupB had14patients(meanage49.0Â±12.1; meanHoehn gional brainmetabolism (rCMRG1c)with â€˜8F-fluorodeoxy andYahrstage32 Â±I .2).Theregionalcerebralmetabolicrates glucose (FJXI) may provide a more versatile alternative as forglucose(rCMRGIC)inallpatientsineachgroupweremea sured.WA wasusedtocalculatesubjectscoresforthedisease an imaging marker for the diagnosis and assessment of PD. relatedcovariancepatternon a patient-by-patientbasis.Re In a recent FDG-PET study of a combined group of 22 PD suits: Inboth PD patientgroups, subject scores correlatedwith patients and 20 age-matched normals, we used the Scaled HoehnandYahrdiseasesevedtyratings(GroupA: r = 0.59, Subprofile Model (SSM) to identify a significant disease p < 0.004;GroupB:0.57,p < 0.04),quantitativeratingsfor related regional metabolic covariation pattern or topo bradykinesia(GroupA r = 0.63,p < 0.002;GroupB:r = 0.61, graphicprofile(6). This proffle(TopographicProfile1)was p < 0.03),rigidity(GroupA: r = 0.59,p < 0.004;GroupB: r = characterized by lentiform and thalamic hypermetabolism 0.59,p < 0.04),butnotwfthtremor.Conclusion:Thesefindings covarying with hypometabolism of the medial and lateral indicatethat regionalmetaboliccovariancepatternsare robust imagingmarkersof diseaseseverity.FDG-PETmaybe useful premotorcortical areas as well as the parieto-occipital as ciinicaltyin assessingparldnsonlandisabilityand diseasepro sociation areaâ€”a metabolic topography consistent with gression. experimental animal models of parkinsonism (7,8). We found that individualsubject scores for this profile, repre KeyWords: P&dnson's disease; glucose metabolism;positron emissiontomography;fluorodeoxyglucose senting the quantitative contribution of the proffle to the patient's overall brainmetabolism, correlated significantly J NucIMed1995;36:378-383 with independent objective clinical ratings of disease se verity. TopographicProffle1 correlatedclosely with a sim ilar disease related covariance pattern identified by us pre viously in an earlier SSM analysis of a separate sample of arkinson's disease (PD) is characterized by a presynap PD patients and normals studied with a different tomo tic nigrostriatal dopamine dysfunction. This abnormality may be quantified in living patients with â€˜8F-fluorodopagraph(2,6). In this study we examined the potential clinical applica (FDOPA) and PET (1,2). Because striatal uptake rate con stunts for FDOPA (K@FD)correlate with objective ratings of bility of Topographic Profile 1 as an imaging marker of PD. disease severity as well as mgral dopaminergic cell counts We used a modificationof SSM to calculate subject scores for this topographicprofiledirectly from individualpatient FDG-PET data. Calculated subject scores for Topographic ReceivedApr.14,1994;revisIonacceptedAug.8, 1994. Profile 1 correlated with parkinsonian disability in each of Foronrreepondenceor repdntscord@ Dr.E@berg,Departmentof Neurol two discrete patient samples studied with tomographs of ogy.NorthShoreUnWersltyHospit@lComatUnWersityMedk@ulCollege,300Corn [email protected];NY,11030. differentspatial resolution.

378 TheJournalofNudearMedicineâ€¢Vol.36â€¢No.3 â€¢March1995 TABLE I RegionalandGlobalMetabolicRatesfor Glucose(mg/rrHnhlOOg)

Normal(n=20)PD1(n=11)PD2(n=12)Mean (s.d.)Mean(s.d.)Mean(s.d.)Cerebellum7.46

(1.06)5.88(0.75)5.94(1.16)Pons5.05 (0.87)3.83(0.66)3.78(0.79)[email protected] (1.17)3.97(0.74)3.81(1.01)Caudate7.91 .02)5.84(1.31)Lentiform8.61 (1.25)6.04(1 .28)6.62(1.32)Thalamus7.74 (1.19)6.75(1 .39)5.74(1.06)Med. (1.09)6.15(1 .14)3.94(0.93)4.04(1.00)[email protected] (1 .09)4.90(1.39)Operculum8.70temporal7.52 (1.32)5.43(1 .51)5.64(1.40)Sup. (1.38)6.41(1 .27)6.37(1.42)Med.temporal9.02 (1.50)6.@(1 (1.26)5.73(1.19)5.14(1.42)L.at.frontal7.96 .30)5.@(1.50)Occipital10.16frontal8.71 (1.44)6.51(1 (1.51)7.41(1.11)7.04(1.79)Parieto-occlpftal9.76 (1.50)7.65(1.17)6.45(1.76)lnf. .10)5.82(1.57)Paracentral7.81parietal8.48 (1.28)624(1 (1.08)5.70(1.19)5.13(1.15)Whole .08)5.84(1.28)PD1brain8.30 (1.19)6.28(1

= mildParkinson'sdisease(Hoehnand Yahr I-Il) and PD2= moderateâ€”advancedParklnson's dIsease(Hoehnand Yahr lllâ€”V).

MATERIALS AND METHODS erence to the orbitomeatal (OM)line. A cylindrical tube filledwith Patients and PET 680ewas placedinthe fieldofview to provideinternalcalibration foreachslice.Allstudieswereperformedwiththesubject'seyes Group A. We studied 23 classical PD patients (13 male and 10 open in a dimly lit room and minimal auditory stimulation. The female; age 60.2 Â±12.2yr; mean Hoehn and Yahr score 2.4 Â±1.3) FDG-PET technique employed to calculate rCMRG1c has been (H&Y) (9). A diagnosis of PD was made if the patient had pure parkinsonism without a history of known causative factors, such described by us in detail (413). as encephalitisor neuroleptictreatment,didnot haveearlyde Region of interest (ROl) analysis was performed on 256 x 256 mentia, supranucleargaze palsy or ataxiabut did have a convinc PETreconstructionsusinga SUN microcomputer(490SPARC ing response to levodopa. In all patients, family historywas neg Sewer; Sun Microsystems, Mountain View, CA) with Scan/VP ative for neurodegenerative illnesses. In all patients T2-weighted Software(14). We defined26 (13per hemisphere)standardized MRI (echotime 80 msec; repetitiontime 1,500fieldstrength corticaland subcorticalgraymatter ROIsand two cerebellarand 1.OT)disclosed normal basal ganglia signal; cortical and subcor two brainstemROIS.ROISwere definedinteractive'yon recon tical atrophy was absent. These patientswere selected with a wide structedPET slicesby visualinspectionwith referenceto a stan range of clinical involvement. Eleven patients had mild parkin dardneuroanatomicalatlas(15)andMRI.To reducepartialvol sonism (H&Y Iâ€”I!),and 12 patients had moderate-severe disabil umeeffects,wecalculatedpeakrCMRGlcvaluesby averagingthe ity (H&YIll-IV). ThesePD patientsareentirelydifferentfrom upper20%ofROlpixelvalues(16).Wheneveranatomicalregions the 22 patientswho were employed by us for the identificationof straddledcontiguousPET slices, rCMRG1cwas calculatedby TopographicProfile1 (6). weightingcomponentROl values by the numberof thresholded Patientsfastedovernightpriorto FDO-PETscanning.In all pixels on each slice. Mean rCMRG1c data for Group A patients patients, antiparkinsonianmedicationswere discontinuedat least and20 normalsubjects(meanage 47.0 Â±17.1yr) aregivenin 12hrbefore PETwas conducted. At the timeofthe PETstudy, all Table1. patientswere ratedquantitativelyaccordingto the H&Y scale and Group B. We studied 14 other classical PD patients (13 male, 1 the Unified Parkinson Disease Rating Scale (UPDRS 3.0) (10). female; mean age 49 Â±12yr; mean.H&Y score 3.2 Â±1.2). The PETstudieswereperformedusingtheSuperpett3000tomograph clinicalcharacteristicsof these patientshavebeen presentedpre (Scanditroni,q Essex, MA) at North Shore University Hospital, viously (6). These patients were entirely differentfrom the 22 Manhasset,NY. The perfonnancecharacteristicsof this instru patients used in the identification ofTopographic Profile 1 (6), and ment have been described elsewhere (11). This four-ringBaF2 fromthesubsequent23patientsconstitutingGroupA described time-of-flight,whole-body tomograph acquires 14PET slices with above.ThesePDpatientswerestudiedwhileonantiparkinsonian Z-axis translation. Each slice is 8 mm thick and reconstructed medications with FDG-PET using the PC4600 PET tomograph with a transaxialresolutionof8 mm (FWHM).Ethicalpermission (slicethickness10mm;transaxialresolution10mmFWHM)at for these studies was obtained from the Institutional Review MemorialSloan Kettering Cancer Center, New York, NY (17). Board of North Shore University HospitatlCornell University Allpatientswere ratedquantitativelyat the timeofthe PETstudy Medical College. Written consent was obtained from each subject accordingtotheH&Yscaleandamodifiedversionofthe Cornell followinga detailedexplanationof theprocedures. UCLA scoring system in which bradykinesia, rigidity and tremor PatientswerepositionedinthescannerusingtheLaitinenste wereeachratedon a composite0â€”4scale(2,18).TheFDG-PET reoadapter (12) with three-dimensional laser alignment with ref procedures for rCMRG1ccalculationhave been previouslyde

Parkinson'sDiseaseAssessmentwithFDG-PETâ€¢Eideibergat Si. 379 @ : :.@ . : :

@ FIGURE1. Isometricdisplay(6,14)of â€˜@@., - r the regionweightsforTOpOgraphiCProfile1 @___J onrepresentativetransversebrainsliceac quiredappro@dmately40, 60 and 70 mm abovetheo@mea@ hne.The Insert(bw erleftyindicatesthepositionsofthe midhnes ofthesetransverseslicesona standardized parasagittaltwo-dimensionaldisplay.Asia @ five hypermetaboUsm(cc@orscale) of the . â€˜ . - I lentiformnucisiand thalamusis shownon thebottomslice(upperleft).Relafivehypo metabolism(greyscale)ofthe lateralfrontal, inferiorpaiietalandparieto-occipitalareasis evidenton the upperslices(right).

scribed (2). To facilitatethe comparisonof rCMRG1cdata ob the topographicprofile.The resultingregressioncoefficient quan tamed for Group A and Group B patients, we used similar ROl tifies the differencein topographicprofileexpression between the placement procedures for both sets of PET studies. However, unknown subject and the reference, i.e., the relative subject because of the lower resolution of the tomograph used for the score. In this study, the calculated subject scores were used to GroupB studies,a compositeROl includingbothcaudateand predict disease severity on an individual basis. lentiform nuclei was adopted for those analyses. The remaining The particularrCMRG1cdatasetused as a referenceis not ROIsdefinedforbothGroupA andB PETscansoccupiedanal criticalprovidedthat the same referenceis used for intersubject ogous positions relative to the OM line. Mean rCMRG1cdata for comparisonsorlongitudinalstudiesof individualsubjects.Inthis these 14 PD patients also have been previously reported (2). study we performed the analysis using Topographic Profile 1 regionweights(6), andGroupA andB rCMRG1cdatasets.We Topographic PrOfileRating: Rationale and chose as reference the mean rCMRG1cvalues from the combined ComputatIonal Procedure sample of 42 PD patients and normals used to identify Topo In previousstudieswe used SSM to identifydisease-related graphicProfile1in theoriginalSSManalysis(6). Thisreference topographiccovariance profiles from combined rCMRGIcdata was selectedto providea commonbaselinefor subjectscores sets obtained from patients and controls, blind to group member derived in the original SSM analysis and in the subsequent TPR ship (2,419). The mathematicalpropertiesof this model and its analyses. statistical assumptions have been described elsewhere (19â€”21).In Regionalmeasuresof glucosemetabolismandcalculatedsub contrast to combined group analyses for the identification of sig ject scores for Topographic Proffle 1were correlated with clinical nificant topographic profiles, SSM computational procedures can ratings in each group of PD patients, using Pearson product mo also be used to calculate subject scores for a predetermined to ment correlation coefficients. The following clinical parameters pographic profile on a patient by patient basis. We refer to this were usedin thesecorrelations:H&Yratingsof overalldisease clinical applicationof SSM as TopographicProfileRating(TPR). severity and individual ratings for bradykinesia, rigidity and WeselectedTopographicProfile1 (Fig.1)as a potentialimaging tremor.AllstatisticalanalyseswerecarriedoutusingSAS. markerfor diseaseseverityin PD (6), andused it to calculate subject scores for this profile from the individual patient rCMRG1c data.This procedurewas conducted separatelyforeach of the PD RESULTS patients comprisingboth GroupA and B. In the TPR computa tionalprocedure,subject scores were calculatedfor GroupA and Glucose Metabolism B patientsin nearlyan identicalfashion.BecauseTopographic Absolute rCMRG1cand whole-brainnormalizedrCMR Profile 1was determined on the same tomograph as that employed Glc values did not correlatewith clinical ratingsin eitherof in the GroupA studies, these patientswere analyzed using sepa the two patient groups. rateregionweightsforthecaudateandlentiformnuclei.Because ofthe lowerresolutionofthetomographemployedintheGroupB studies, composite basal ganglia ROl data were used with aver Topographic PrOfile RatIng agedcaudate-lentiformproffleregionweights. GroupA. Calculated subject scores for these 23 patients TPR providesa measureof the degreeto which an individual correlated with individual H&Y disease severity ratings subject expresses any given topographic proffle in their brain (r = 0.59, p < 0.004; Fig. 2A, left) andwith UPDRS ratings FDG-PET(rCMRGIc)data. This determination is made relative to for rigidity(r = 0.59, p < 0.004) andbradykinesia(r = 0.63, a referencerCMRG1cdatasetobtainedeitherfromanotherpatient p < 0.002; Fig. 2A, right);subject scores did not correlate ornormalcontrolsubject,orfrommeanrCMRGIcdataacquired significantly with tremor scores (r = 0.24, ns). across a populationof normals,patientsor both. In TPR, rCMR Glc data (from ROIs or individual pixels) from the unknown Group B. Calculated subject scores for these 14 patients subjectandthereferencearelog-transformedandsubtracted.The correlated significantlywith H&Y scores (r = 037, p < subject score of the unknown relative to the reference is deter 0.04; Fig. 2B, left), and with quantitative ratings for rigidity minedby using a multivariatelinearregressionanalysisto predict (r = 0.59, p < 0.04) and bradykinesia(r = 0.61, p < 0.03; theregionaldifferencesfromthecorrespondingregionweightsof Fig. 2B, right),but not tremor (r = 0.28, ns).

380 TheJournalofNudearMedicineâ€¢Vol.36â€¢No.3 â€¢March1995 GroupA diographic equation for Group A metabolic calculations (611,13). @ r- Is r.O.63 Additionally, clinical measurements of disease severity .v p(0004' a p@aOo2 in the two groups differedin the following ways: the phar macologic states of the patients in the two groups differed: @4@tf:i@@ I all GroupA patients were off antiparkinsonianmedication at the time of PET, while the majority(12 of 14) Group B 0 0.03 0.06 0.05 0.12 0.15 0.18 0 0.03 0.08 0.00 0.12 0.15 0.18 S@d Score SubjectScore patients had PET scans on medication; differentPD rating scales were used for clinical assessment in the two patient groups: the complete motor UPDRS was used in rating the

Group B Group A patients, while an abbreviated modification of the Cornell-UCLA scale (18) was employed in the Group B 8 ratings. In spite of these technical differences, calculated 4 p(O.03 @, 3 Topographic Profile 1 subject scores accounted for similar 2 variance in quantitative clinical scores in both groups (R2 35â€”40%).These results indicate that TPR subject scores

0 0.030.0S0.0$0.120.180.18 o 0J3 0.08 0.08 0.12 0.18 0.18 are a robust marker of disease severity relatively indepen Subject Score S(_@ Score dent of PET instrumentation,scanning protocols and cm ical rating scales. Equally important, this metabolic mea FiGURE2. CorralationinPDpatIentsbetweensubjectscoresfor TOpOgraphiCProfileI and standardizedclinicalrathgs of disease sure correlates with clinical disability in both medicated severity(left H&Y scores;right quantitatIvebredyidnesiaratings). and unmedicatedstates, suggesting its potential applicabil GroupA andGroupB correlationsappearin Figures2A andB, ity in the assessment of antiparkinsonianpharmacotherapy respectively.In bothpatientgroups,significantpositivecorrelations with PET. Indeed, levodopa infusion studies in parkinso werefound betweenthe calculatedsub@ scoresand the dk*al nian primates and PD patients have revealed significant diseaseseverityratings. reductions in basal ganglia and thalamic hypermetabolism with treatment(22,23). This suggests the potential applica bility of Topographic Profile 1 subject scores as useful DISCUSSION markers of successful pharmacologic intervention with In this report, we demonstrated that regional metabolic likely decreases in subject scores in patients undergoing data obtained through FDG-PET may be used as an imag dopaminergic therapy. In addition, we found that the mag nitude of clinical correlations with the calculated subject ing markerof parkinsoniandisease severity. Topographic scores was similar to that reported by us for analogous Proffle 1 is characterized by abnormal increases in basal correlations with striatal K@ (3). Moreover, as in our ganglia and thalamic metabolism and associated with pre earlier studies of clinical metabolic correlations in PD, we motor and parietal cortical metabolic reduction. Accord found no relationshipbetween subject scores for disease ingly, it can be used as a metabolic probe to quantify related topographic profiles and tremor ratings (2,6). We precisely its representation in the functional brain images attribute this to potential nondopaminergic mechanisms of individualsubjects. Thus, SSM methods are not limited subserving parkinsonian tremor (24), as supported by the to the identification of biologically relevant topographic absence of correlation between tremor ratings and striatal proffles, but may have a clinical application in the objective K@FD(2,3). gauging of the severity of parkinsonism and related disor Our findings suggest that FDOPA and FDG-PET tech ders. niques may yield comparable results as imaging markers of Under differentscanningconditions, TPRproducedsim parkinsonian disease severity. FDOPAIPET methods are ilar correlations between subject scores and independent costly, technically demanding and have been limited to few measures of clinical disability. Specifically, the Group A centers (5). Apart from the specialized radiochemistry nec and Group B FDC}-PFTscans differed from one another in essary for production (25), this technique also requires several important ways: (a) The tomographs used in the long kinetic scanning protocols, which may be uncomfort two groups were of markedly different spatial resolution able for patientswith advanced disease. Althoughmethod and camera design: that used in the GroupA studies was a ological simplifcations have been described and validated whole-body scanner of higher spatial resolution than the (3,26,27), the most accurate correlations with disease se brain dedicated instrument used for the Group B studies; verity can be achieved with arterialblood sampling and and (b) The scanning protocol for image functionalization high pressure liquid chromatography (HPLC) (3). Addi (Le., converting raw counts to metabolicvalues on a pixel tionally, precise quantitation of nigrostriatal dopaminergic by-pixel basis) was different in the two patient groups: function with FDOPA/PET may be hamperedby endoge individually calculated kinetic rate constants (k1-k3) were nous aminoacids (28) as well as concurrentlyadministered used to calculate rCMRG1c in Group B scans (2), while dopaminergic agents, their metabolites or both (29). These mean population rate constants were used in the autora technical limitations may be less constraining in other ap

Parkinson'sDiseaseAssessmentwithFDG-PErâ€¢EkielbergatSi. 381 proaches to the quantification of nigrostriatal function, the ParkinsonDisease Foundation.We thank Dr. Abdel such as the use of â€œ@I(1R)-2j.3-carbomethoxy-3@-(4-ioBelakhlef,Mr. ClaudeMargouleffandMs.JanieDillforhelpwith dophenyl)tropane] ([â€˜@I]@-CIT)with SPECT, for the imag the PET studies;Mr. RalphMatacchierifor cyclotron supportand ing of presynapticdopaminereuptakesites (28). In spite of Ms. Debra Segal for manuscript preparation. We acknowledge these limitations, direct mgrostriatalimaging is a useful, Dr. David A. Rottenberg's important contribution to the Group B straightforward method of demonstrating the presence of PETstudies(2). presynaptic lesions in living patients. These imaging meth ods have been proven useful in early disease detection (4) REFERENCES and in the evaluation of fetal tissue implants(31,32). They may also be appropriate for longitudinal studies of the 1. EidelbergD.PETStudiesinParkinsonism.InCedarbaumJM,GancherSI, eds.Neurologicclinics10.Philadelphia:W.B. Saunders;1992:421-433. naturalhistoiy of disease progression (33,34) and the com 2. EldelbeigD, MoellerJR. DhawanV. et al. Themetabolicanatomyof parative long-term assessment of neuroprotective strate Parkinson's disease: Complementaiy â€˜8F-fluorodeoxyglucoseand @F. gies. They are, however, unsuited for assessing the func [email protected])i@oid1990;5:203- 213. tional consequences of modifications in pharmacologic 3. lakikawa5, DhawanV. ChalyI, etal.inputfunctionsforâ€˜8F-fluorodopa state, as would be desirable in testing the efficacy of new quantitationin parkinsonism:comparativestudiesandclinicalcorrelations. dopaminergic agents for the treatment of PD. I NuciMed 1994;35:955â€”963. 4. Snow83, IooyamaI, McOeerEG, et al. HumanPET [â€˜8Fjfluorodopa Our results indicate that TPR applied to FDG-PET data studiescorrelatewith dopaniinecellcountsandlevels.Ann Neurol1993; has utility comparable to quantitative FDOPAIPET imag 34:324â€”330. ing for most of the clinical indications listed above. As an 5. Sawle GV. The detection ofpreclinical Parkinson's disease: what is the role of positron emission tomography? Move DLconI 1993;8:271â€”277. indirect functional marker of the nigrostriatal dopamine 6. EidelbergD, MoellerJR.DhawanV. et al.Themetabolictopographyof system, this technique can be used to determine the integ parkinsonism.I Cei@bBlcodFlowMetab1994;14:783-801. rity of this system at baseline and to assess progression 7. cmssmanAR. A hypothesisonthepathophysiolo@ca1mechanismsthat underlielevodopaâ€”ordopamineagonist-induceddyskinesiain Parkinson's over varying time periods. Mild PD is characterized by disease:implicationsfor futurestrategiesin treatment.MovDison!1990;5: clinical and metabolic asymmetries (2,6,9); therefore, SSM 100-108. analysis of left-right differences in regional glucose metab 8. PalomboE, PorrinoLI, BankiewiczKS, et al. Localcerebralglucose utilization in monkeyswith hemiparkinsonisminducedby intracarotid infu olism may be used for early diagnosis. In an SSM analysis sion of the neurotoxin MFfP. I Neurosci 199010:860-869. of 10 H&Y Stage I PD patients and 10 age-matched nor 9. HoehnMM, YahrMD. Parkinsonism.Onset,progressionandmortality. mals, we demonstrated that subject scores for a significant Neurology 1967;17:21â€”25. 10.FahnS, EltonRL, UPDRSDevelopmentCommittee.UnifiedParkinson covariance pattern of regional asymmetries discriminated diseaseratingscale.In: FahnS, MarsdenCD,CaineD, GoldsteinM, eds. early patients from age-matchednormalswith an accuracy Recent deve@entr in ParWjison'sdisease, volume 2. FloralPark,New comparable to striatal K@ (35). FDO-PET and TPR anal Jersey: Macmillan; 1987;293-304. 11.RobesonW, DhawanV, lakikawaS,etal.SuperPETF3000tline.of.fligJit ysis may have the added utility of differentiatingatypical tomograph: optimization of factors affecting quantification. IEEE Tmns drug-resistantpatients from their drug-responsivecounter NuciSci 1993;40:135-142. parts (1,6, 13), and perhaps of objectively gauging the effi 12.HarizMI, ErikssonAT. Reproducibilityofrepeatedmountlagof a nornn @rasiveCTIMRI stereoadapter.App! Newophys@I 1986;49:336-347. cacy of antiparkinsonian therapies. 13.EidelbergD, TakikawaS. MoellerJR,Cta).Striatalhypometabolismdis In conclusion, Topographic Profile 1 is a useful, robust tinguishesstriatonigral degenerationfrom Parkinson's disease.Ann Neurol imaging marker of the parkinsoman disease process. Be 199333:518â€”527. 14.SpetsierisP,DhawanV, TakikawaS,MargouleffD, EidelbergD. Imaging cause of the greater availability of FDO-PET in clinical cerebralfunction.IEEECo,nputerGnzphicsand4pplications1993;13:15- research centers, metabolic brain imaging with TPR tech 26. niques may be a useful alternative to the more technically 15.TalairachJ, TournouxP. Coplanarstereot&*atlasofthehumanbrain. New York: Thieme Medical Publishers, Inc, 1988. demanding FDOPA/PET methods. Recent simplified 16.RouenbergDA,MoellerJR.StrotherSC,DhawanV, SergiML. Effectsof methods for the estimation of rCMRG1c further increases percent thresholding on the extraction of â€˜8F.fluorodeoxyglucosepositron the applicabilityof FDC}-PETtechniques for the study of emission tomographic region of interest dat@I Cei@bBk@sdFkrw Metab 1991;11:A83-A88. large clinical cohorts (36). To the extent that regional ce 17.KearfottK.J,CarrollLR. Evaluationoftheperformancecharacteristicsof rebral blood flow and glucose metabolism are coupled in the PC4600positron emission tomograph.I ComputAssirt Tomogr 1984; PD (37), TPR may also be applicableto cerebralperfusion 8:502â€”513. 18.CedarbaumJ,HoeyM, McDowellFH. A double-blindcrossovercompar mapping with other widely available imaging modalities ironof SinemetCR4andstandardSinemet25/100in patientswith Parkin such as SPECT. Additionally, these functional imaging son's disease and fluctuating motor performance. I Neural Neurosuig techniques may have analogous clinical applicability in the Psych 198952:207â€”212. 19.MoellerJR,StrotherSC,SidtiSJJ,RottenbergDA. Thescaledsubprofile study of other neurodegenerative disorders including the model: a statistical approach to the analysis of functional patterns in dementias. positron emission tomographic data. I Ce@bBlood Flow Metab 1987;7: 649â€”658. 20. Moeller JR, Strother SC. A regionalcovariance approachto the analysisof ACKNOWLEDGMENTS functional patterns in positron emission tomographic data. I Ce@bBlood Thiswork was supportedby grantsfromthe ParkinsonDisease Flow Metab 1991;11:A121â€”A135. 21. SackeimHA, ProhovnikI, MoellerJR,MayeuxR,SternY, DevanandDP. Foundation and the Dystonia Medical Research Foundation. D.E. Regionalcerebral blood flow in mood disorders. II. Comparison of major is a faculty fellow of the Parkinson Disease Foundation and the depressionand Alzheimer's disease.JNuclMed 1993;34:1090-1101. United ParkinsonFoundation.T.I. is a VeolaT. Kerr fellowof 22. Pomno LI, Burns RS, CraneAM, et aL Local cerebral metabolic effectsof

382 The Journal of Nudear Medicineâ€¢Vol. 36 â€¢No. 3 â€¢March 1995 L-dopa therapy in 1-methyl-4-phenyl-1,2,3,6tetrahydropyridinc-induced 30. HoshiH, KuwabaraH, LegerG, OimmingP, GunmanM, OjeddeA. parkinsonism in monkeys. P@vcNatlAcad Sci USA 1987;84:5995-5999. 6.[18flfiuoro-L.dopametabolismin li@nghumanbrain:acomparisonof six 23. BlessR,BlinJ, MiletichR,etal.Levodopa-reducedglucosemetabolismin analyticalmethods.I CerabBloodFlowMetab1993;13:57-69. strla@dothalamo-corticodrcolt inParkinso&sdisease[Abstract].Neu 31. SawleGV, BIOOmfIeIdPM, BjorklundA, Cta!.Transplantationof fetal ?ulogy 1991;41:359. dopamineneuronsin Parkinson'sdisease:PET[18F]6.L.fluorodopastudies 24. BernheimerH, Birkmayerw,Hornykiewicz0, JellingerK, SeitelbergerF. intwopatientswithputaminalimplants.AnnNeural1992;31:166-173. Brain dopamineand the syndromesof Parkinson and Huntingron: clinical, 32. FreedCR, BreezeRE, RosenbergNL, et aLSurvivalof Ã¼npbantedfetal morphologicaland neurochemicalcorrelations.I Neurosci1973;20:415- dopaininecellsandneurobogicimprovement12-46moaftertransplantation 455- for PD.NEngIJMed 1992327:1549-1555. 25. LuxenA, MiltonP,BidsGI, etaLRemote,semiautomatedproductionof 33. BhattNIH,SnowBJ,MartinWRW,et ab.PETsuggeststhattherateof @[18flfluoro-L-dopaforhumanstudieswithPET.4pplRadiathot1990@, progressionof idiopathicparkinsonisinis slow.Ann Neural 1991;29:673- 41:275â€”281. 6Th 26. BrooksDJ,SalmonEP,MathiasCJ,et al.Therelationshipbetweenboo 34. SawbeGV,TurjanskiN, BrooksDJ,etaLTherateofdiseaseprogressionin motordisability,autonomicdysfunctionandthe integrityof the striatal Parkinson'sdisease:PElfindingsinpatientsreceivingmedicaltreatmentor dopaminergicsystemin patientswith multiplesystematrophy,pureauto followingfetalmesencephalictransplantation[Abstract].Neurology1992; nomicfailure,andParkinson'sdisease,studiedwithPEF.Brain1990@,113: 42:295. 1539â€”1552. 35. EidelbergD, ModllerJR.DhawanV. IshikawaI, PrzedborskiS.FahnS. 27. EidebbergD, lakikawa S, DhawanV, Ct aL Striatalâ€˜8F-dopauptake: Detectionof earlyParkinson'sdiseasewith â€˜8F-fluorodeoxygiucoseand absenceof anagingeffect.I Cei@bBloodFlowMetab1993;13:881-888. positronemissiontomography[Abstract].INucI Med 1994:1OP. 28. InnisRB,SeibybJP,ScanleyBE, et al. SPECFimagingdemonstratesboss 36. TakikawaS,Dhawanv, RobesonW, etaLNOninVasiVequantitativeFDG ofstriataldopaminetransportersinParkinsondisease.F@ucNat/AcedSd PET studies using an estimated input function derived from a population USA1993;9@liu1%5-11%9. arterialbboodcurve.Radiology1993;188:131-136. 29. LeendursKL, PoeweWH,PalmerAJ,Ctal.Inhibitionof L-(18F]fluorodopa 37. OtsukaM, Ich@raY, HosokawaS, etal.Striatalbloodflow,glucoseme uptakeintohumanbrainbyaminoacidsdcinonstratedbyPET.AWZNeWU1 tabolismandâ€˜8F-dopauptake:differencein Parkinson'sdiseaseandatypi 1986;20:258â€”262. calparkinsornsm.JNeurolNeumsurPsych199154:898-904.

Parldnson'sDiseaseAssessmentw@iFDG-PETâ€¢EklelbergatSi. 383