study was to employ a specific physiological activationparadigm that had been examined previously using PETand to determine whether the predicted local activationeffect could be detected and whether the magnitude ofactivation was comparableto the effect observedwith PET.For this purpose, we used a finger-thumb opposition task,which results in a metabolic activation of the pre- andpost-central gyrus ofthe contraiaterai cerebral cortex (12-16).

MATERIALS AND METHODS

SubjectsTwelve control subjects were recruited from relatives and

friendsassociatedwith the local branchofthe Aizheimer'sdiseasesociety. They included four men and eight women with a meanage of 63 yr (s.d. 3.5, range 59—68yr). Patients were examinedusing the Annett Handedness Scale, scored according to Briggs& Nebes(17).TheywereassessedneuropsychologicallyusingtheNational Adult Reading Test (18), the block design and digit

symboltestsof the WechslerAdult IntelligenceScaleR (19) andthe Wechsler Memory Scale-R (20) to ascertain that no subjectshowed evidence of cognitive or psychomotor impairment. Allsubjectsgaveinformedwrittenconsentaccordingto a procedureapprovedby the local ethics committee.

Imaging ProcedureSubjects were imaged with a single slice multi-detector dedi

cated head scanner(Multi-X 810, StrichmanMedical EquipmentInc., Boston, MA) after two repeated injections of 250 MBq of99mTc@xame@ime,as described in detail elsewhere (11). Withthe use of the intermediate 572 whole collimators, the in-sliceresolution ofthe scanner is given as 7.5 mm (FWHM) by 15 mm

slice thickness. The sensitivity of the instrument was measuredas 520 cps/l kBuJml (11). An indwelling intravenous catheterwas inserted in an arm vein 15—30mm before the injection. Forthe injection, subjectswere positioned comfortably on the imaging table with eyes patched and ears unplugged. A metronomewas calibratedto 90 bpm and activated for 20 beats before and 5mm after start ofthe injection (bolus injected over 5 sec). Duringthis time, the subject was instructed not to move in rhythm tothe beat, and ambient nose was kept to a minimum. The subject'shead was then placed in a moulded headrest, positioned with thehelp oftwo crossed iight beams, and fixed with two pressure padsover the zygomatic arches. Five slices were acquired parallel tothe orbito-meatal line at +55, +65, +71, +77 and +83 mm. The

The studyexaminesthe sensitivityof a regionof interestapproachto detect functionalchangesin brain metabolismwith SPECT and split-dose @Tc-exametazimeby replicatinga simplehandmovementexperimentpreviouslyearnedoutwith PET. Regionaluptakeof @‘Tc-exametazimewas determined in 12 healthy controls before and during a thumb-digitopposition task. Analysis of regional uptake was carried outblind to the hand used in the opposition task and showed asignfficant unilateral activation effect in a pericentral region ofinterest with opposite results in left- and right-handed activation. The maximum contralateral increase in tracer uptakewas16% beforeand26% aftercorrectionforbackdiffusion.Thisis in goodagreementwithpreviousresultsemployingabsolute cerebral blood flow determination with PET andconfirmstheusefulnessof @Tc-exametazimeSPECTfor theexaminationof functionalmetabolicchanges.

J NucIMed 1992;33:1637-1641

echnetium-99m-exametazime was introduced on thebasis that it gives an estimate of regional cerebral bloodflow (rCBF) (1). Its quantitative properties in the estimation of rCBF at rest have been determined against severalother methods, such as ‘33Xe(2), ‘50-iabeiedCO2positronemission tomography (PET) (3-5), ‘8F-iabeiedfluoromethane PET (6) and radioiabeied microspheres (7).These studies have shown a monotonic, but non-linear,relationship between uptake of 99mTc@exametazime andrCBF.

In addition to the measurement of resting rCBF,singlephoton emission tomographywith @mTc@exametazimecanbe employed to examine changes in rCBF during physiological, psychological or pharmacological activation (8-10). We have recently reported a split-dose technique forthe study ofsuch activation effects, using a repeatinjectionof 250 MBq 99mTc@xame@imeapplied in the same imaging session, which yielded good reproducibility of resultsin repeat baseline studies (11). The purpose ofthe present

ReCeiVedJan.16,1992;revisionacceptedApr.9, 1992.For reprints contact: Dr. K.P. Ebmaler, MRC Brain Metab@m 1k@it,Royal

Edinburgh Hospital, Momingside Park, Edinburgh, Scotland EH1O5HF.

1623HandMovementsandso@@@Tc@ExametazimeSPECT•Ebmeieretal

Unilateral Voluntary Hand Movement andRegional Cerebral Uptake of Technetium-99m-Exametazime in Human Control SubjectsKlaus P. Ebmeier, Catherine L. Murray, Nadine J. Dougall, Ronan E. O'Carroll, and Guy M. Goodwin

MRC Brain Metabolism Unit, Royal Edinburgh Hospital, Morningside Park, Edinburgh, Scotland

by on December 1, 2018. For personal use only. jnm.snmjournals.org Downloaded from

Scan@ x 5min/sliceJ @i@2.5min/sI@]

c MstronomeAct@vat@on@ M.trono@;@j andFIngr@n_flon 4@l Movement

Tracer

Practice t 4@Isessions

subjects were then instructed to perform a “fingeroppositiontask,―touchingeachoffour fingerssuccessivelywith their thumb(12). They had to perform this movement in rhythm with themetronome, one half movement, i.e., touch or extend, per beatand moving briskly from one extreme position to the other.Subjectswerealternativelyallocatedto right-and left-handmovement conditions. Two short training sessionswere performedabout 10 mm after the first injection and again, after the firstscan, immediatelybeforethe secondinjection.Oncethe task wasperformedaccurately,the subjectwaspositionedcomfortably,allmuscle groups not involved in the movement task were relaxed,and the secondinjectionof@mTc@exametazimewasgiven.Movements werestartedaftera signalgiven by the experimenterlightlytouching the subject'sforearm 10 beats afterthe metronome wasstarted. The second intravenous bolus of tracer (250 MBq in 5sec) was given after an additional 10 beats. Movements werestopped by a second touch signalfrom the experimenter3 mmafter the injection. For the remaining 2 mm, conditions wereidentical to baseline. Subjects were repositioned and the samesliceswereacquiredas for the baselinecondition.The acquisitiontime for the second scan was 2.5 mm per slice, half the timerequired for the baseline scan. We routinely use two half-dosesinstead of a larger dose for the second (activation) injection,becausethe potentialgain in signal-to-noiseratio affordedby thelatter is offsetin psychiatricpopulationsby increasedmovementartifacts during the longer acquisition time of the first (lowerdose)scan.The designof the activationand imagingprocedureis illustratedin Figure 1. Lipophiicity of the tracer after reconstitution was measured in six subjects at baseline as 85% (s.d.,6.1%)and for the activation scans as 87%(s.d., 3.2%).The meandifference in lipophilicity between scans was 2% with a 95%confidence interval from —5%to 10%. The radiation dose foreach subjectwasapproximately8 mSv.

Data AnalysisA template was drawn outlining the slice hemispheres and a

pericentralregion ofinterest (ROI) in the shape of parallelogram.The wholetemplatewasfittedto the outlineof the hemispheresdefined after removing the lower 40% of the intensity spectrum.Theparallelogramhaditsanteriorboundaryapproximately30%and its posteriorboundary60%ofthe distancefrom the anteriorto the posterior pole of the brain. It was chosen to include thepre- and post-centralgyms in a standard brain atlas (Fig. 2).Whenappliedto thethreeslicesat 71, 77 and83 mm abovetheorbito-meatal (OM) line, the mean area of the pericentral ROlwas 13 cm2(s.d. 2.1) left and 13.4 cm2(s.d.2.1) right.The interrater reliability of the measurement of mean regional countdensitiesin this ROl wasdeterminedby normalizingthe regionalcount densities to the whole slice, calculating the difference ofmeasures between two raters (NJD and KPE) and expressingthese differences as a percentage of the mean of the two raters.

The inter-ratererror, definedas twicethe s.d. ofthese percentagedifferencevalues,amounted to 5%for the pericentralROI.Therewas no systematic differencebetween the two raters,so that 95%ofall inter-raterdifferencesweresituatedbetween—5%and +5%.This means that in a comparison of two groups of six subjects,differencesin group means of more than 3.2% can be detectedwith confidence. Whole brain activity was calculated from hemisliceactivitiesofthe three slicesat 71, 77 and 83 mm. The ROlto-wholebrain ratio(R)wascalculatedforROIsin all threeslices.R valueswerecorrectedforflow-dependentbackdiffusionusing

FIGURE 1. Protocolfor tracerinjectionand activationprocedures.

the Lassen algorithm as suggested by Inugami et al. (4) andWoodset al. (8):

Rcorr aR/(l + a —R),

where a 2. The net uptake during the activation procedure(secondinjection)was calculatedby subtractingtime-correctedcounts measured during the first scan from the total countsmeasuredduring the second scan, as describedelsewhere (11).

ROI-to-wholebrain ratios before and after Lassencorrectionwere compared between the two groups of patients with left- andright-handed activation using multivariate analysis of variancewith: (1) left/right side of brain, (2) three slice levels and (3)baseline/activation scans as three within-subjects factors. Thismethod takes into account all repeatedmeasuresof varianceandco-variances and represents a relatively conservative omnibustest for the presence of an activation effect. Crucial to the detection of activation would be a significant interaction between sideof activation, side of the brain and condition. This derives fromthe hypothesis that increases in rCBF were expected in one ormore slices for the side of the brain contralateralto hand movement only during the activation condition.

Becausethe level ofthe slice showing an activation effect couldnot be predicted a priori, uptake ratios were averaged post hocacross three slices and a right-leftindex was calculated as 200 x(R—L)/(R+L)to determinean effectsize.One-tailedt-testswereused to compare indices between subjects with right- and leftsided hand movement. Afterthese tests had been conducted blindto the side ofactivation, a more specific examination ofthe locusofactivation effect was carried out by selecting the slice with the

FIGURE2. Configurationof the ternplatewfthtwo hemislice and two pencentral ROls (theunderlying atlasslice is reprinted

@th pemiiss@nfromRef.21).

1624 TheJournalof NuclearMedicine•Vol.33 •No.9 •September1992

by on December 1, 2018. For personal use only. jnm.snmjournals.org Downloaded from

TABLE1DescriptiveandNeuropsychologicalScreeningVariablesTestMeanRangeHandedness

(Briggs&Nebes,22.820-241975)Years

ofeducation15.912—21NART-lQ(Nelson&Willison)119.3111—127Block

design(WAIS-R)(Wechsler,13.311—171981)Digit

symbol(WAIS-R)12.29—15Attention/Concentration(WMS-R)114.271—130(Wechsler,

I987)Generalmemory@WMS-R)109.692—133Delayedrecall(WMS-R)1 19.693—138

Right-HandedactivationLeft-Handed activationt (d.f.= 10)p(one-tailed)No

LassencorrectionBaseline3.2(2.1)1.4(1.6)—0.670.52Activation—2.1(2.9)7.4(2.4)2.530.03Lassen

correctionBaseline4.7(3.1)2.2(2.4)—0.630.54Activation—3.3(4.4)1 1.3 (3.6)2.560.03

might have received some tactile and proprioceptive input.As predicted, left-handed activation results in an increasein uptake to the right side (i.e., a high index), whereasright-handed activation reverses the baseline right asymmetry towards the left (index < 0).

To determine the effect of size of contralateral rCBFincreases, it was assumed that the pericentral ROl showingthe highest contralateral increase in uptake from baselineto activation scan included the brain area with specificfunctional activation due to hand movements. Table 3shows the corresponding percentage increases before andafter Lassen corrections. The maximum increase in perfusion and, therefore, the putative locus of activationtended to be in the highest slice, 83 mm above the OMline, although in a few cases, maximum changes wereobserved up to 12 mm lower (Table 3).

The anatomy of such a peak effect in one subject (no.2, Table 3) is illustrated in Figure 3. Slices at 71, 77, and83 mm above and parallel to the OM line are depictedbefore and during left hand actiyation, in combinationwith the pixel by pixel subtraction image.

DISCUSSION

The results show that the distribution of 99mTcxametazime can be used to illustrate dynamic changes in rCBFin neocortex which parallelchanges in neuronal activationassociated with physiological stimuli (12—15,24, 25). As agroup, subjects with right-hand movement showed increased uptake to the left pericentral cortex of slices 7 1, 77and 83 mm above the OM line. Subjects with left-handedactivation showed an opposite effect. While the level ofslices and the location of the ROl was predetermined byreference to a previous PET study (12), measurementswere made blind to the site of activation. In addition, theomnibus muitivariate analysis of variance did not rely onthe determination of areas with peak change as used inWoods et al's. (8) study of visual stimulation, thus furtheravoiding the possibility of spuriously positive results.

The magnitude of effect produced in motor and proprioceptive sensory activation was determined post-hoc andresulted in a percentage increase very similar to that reported by Colebatch et al. (12) who used a quantitative

greatest increase in mean uptake in the pericentrai ROl of thecontralateralhemisphereand by estimating the maximum size ofthe activation.

RESULTS

The subjects were all right-handed;their neuropsychological test results (Table 1) indicated that none showedcognitive deterioration.

That increasesin rCBFoccurredin the pericentralROIsof the hemispheres contralateral to the activation onlyduring the activation condition was supported by the Fvalue of the appropriate omnibus test (side of activationx sideofbrain x activationcondition—F,,,0= 7.37,p =0.022). The results were identical if uptake ratios werecorrected for backdiffusion at higher flow rates (F,,,0 =7.77, p = 0.0 19).

Our conservative estimate ofthe lateralshift in rCBFofpericentral ROIs was based on the average uptake of tracerin pericentral ROIs across the three slices. Mean valueswere calculated separately for left- and right-handed activation groups. These are shown in Table 2 during baselineand activation scans with and without Lassen correction.Although at baseline both groups showed an asymmetrytowards the right as predicted from the literature (22),there was a nonsignificant trend for the right-left index tobe smaller in the left hand than the right-hand activationgroup. This suggestsrelativelygreateruptake contralateralto the side of the cannulated, nonactivated side, which

TABLE 2Riqht-Left Indices Calculated from Uptake to Pericentral ROls Averaged Across Three Slices

Right-Leftindex(s.e.)

HandMovementsand @Tc-ExametazimeSPECT•Ebmeleret al 1625

by on December 1, 2018. For personal use only. jnm.snmjournals.org Downloaded from

Mean %lncrease in Tracer Uptake in the ROl and Sloe withthe Highest ContralateralIncrease%lncreaseSide

ofcorticalafterLassenSubjectno.activation Slice%lncreasecorrection1R

+71mm21362R+71mm13223R+83mm—1—24R+83mm19295R+83mm21326R+83mm34567L+77mm29498L+77mm—2—29L+83mm142110L+83mm172711L+83mm132012L+83mm1625Mean(s.d.)16(10)26(17)

At a@ , a,

@ @t‘@‘B

@ a.@ fiii\ s@@ (@J@(&@@t@ •Q,r@.:.@

tazime method does not allow resolution of such smalleffects, or because the effect is actually less robust than thePET report suggests. With the use of larger ROIs, anyactivation effect is likely to disappear due to backgroundnoise. The solution to this problem may be a pixel-bypixel subtractionprocedureto remove backgroundactivityand to identify functional hot spots. With such a methodit will again be essential to identify the anatomical localization of activated areas using parallel MRI scans or astereotactic reference. Finally, the additional use ofa threedimensional reconstruction procedure (e.g., Neuro 900,Strichman Medical Equipment Inc.) or a multiple-headedrotating gamma camera might have increased transaxialresolution and the likelihood ofdetecting smaller changes.

CONCLUSION

Our results suggest that the measurement of regionaluptake of 99mTc..@xame@imeprovides a sensitive methodfordetecting cerebralcorrelatesofphysiological activation.The order of magnitude of the effect observed compareswell with the results derived from ‘@OPET and ‘33XeSPECT studies. While PET studies employing ‘@Oor ‘tFwill remain restricted to few centers and relatively cooperative patients and controls, and ‘33XeSPECT studieshave certain drawbacksdue to their poorer spatial resolution,@ SPECT is easily available for theexamination ofmetabolic abnormalities in more disturbedpatient populations. First results suggest that simple physiological activation procedures might accentuate differences in tracer uptake between patients and controls (15).Such methods, extended to neuropsychological activation(26) offer a potential means for identifying the functionalabnormalities in psychiatric disorders and the action ofdrugs for their treatment.

ACKNOWLEDGMENTS

The authors thank the WellcomeTrust for financial supportand Norma Brearleyfor assistancewith the manuscript.

integral/dynamic method ofCBF determination with t5()labeled CO2 (23). The calculation of percentage increasein uptake using a correction for backdiffusion resulted invalues slightly in excess of those reported with PET (21%)(12). Guenther et al. (15) reported a similar experimentusing opening and closing of the right hand and ‘33XeSPECT with a tomographic scanner. These authors reportan increaseofrCBF in the left motor cortex ofeight healthycontrols of 26%.

In two subjects (nos. 3 and 8, Table 3), no contralateralactivation effect could be observed. This is likely due touncontrolled variability in task performance, e.g., an inadvertent and unobserved activation of other musclegroups which could have lead to an overall increase oftracer uptake and, therefore, a masking of any unilateralchanges in regional uptake into sensory-motor cortex.

In addition to primary sensory-motor activation, Colebatch et al. (12) also report increased blood flow in supplementary motor cortex. An attempt to fit a small ROlover this area in our scans did not yield a significantactivation effect. This may be because the @“@Tc-exame

FIGURE3. Demonstrationofpre-andpost-central gyrus activation dunng thefingeroppositiontask. (A)Restingbaselinescan. (B) Scan during left-handed activation, transaxial slices from 71 mm aboveand parallel to the OM line at 6 mm intervals for subject 2 (Table 3).

1626 The Journal of Nuclear Medicine@ Vol. 33 •No. 9@ September1992

by on December 1, 2018. For personal use only. jnm.snmjournals.org Downloaded from

cerebral blood flow during voluntary arm and hand movements in humansubjects. JNeurophysiol 199l;423: 1392—1401.

13. Roland PE, Larsen B, LessenNA, Skinhoj E. Supplementarymotor areaand other cortical areas in organization of voluntary movements in man.JNeurophysio/1980;43:l18—136.

14. Roland PE, Skinhoj E, Lassen NA, Larsen B. DifFerentialcortical areas inman in organization of voluntary movements in extrapersonal space. INeurophysioll980;43:137—l50.

15. Guenther W, Moser E, Mueller-Spahn F, von Oefele K, Bud U, HippiusH. Pathologicalcerebral blood flow duringmotorfunction in schizophrenicandendogenousdepressedpatients.BiolPsychiatryl986;2l:889—899.

16. Raichle ME. Circulatory and metabolic correlates of brain function innormal humans. In: Mountcastle VB, ed. Handbook ofphysiology, thenervoussystem. New York: Am PhysiolSee; 1987:643-674.

17. BriggsGG, Nebes RD. Patternsofhand preferencein a student population.Cortex l975;l 1:230—238.

18. Nelson HE, Willison JR. Supplement to the national adult reading testmanual: restandardisazion of the NART against the WAIS-R. Windsor@

NFERNelson; 1991.19. Wechsler D. Wechs/er adult intelligence scale—revisedtest manual. New

York:PsychologicalCorp.; 1981.20. Wechsler D. Wechs/er memory scale revised—testmanual. New York:

Psychological Corp.; 1987.21. Talairach J, Zilkha G, Tournoux P, et al. At/as d@anatomiestereotactique

duze/encephale.Paris:Masson;1988.22. PrOdreka I, Goldenberg G, Baumgartner C, et al. HMPAO brain uptake

in youngnormal subjects: genderdifferencesand hemispheric asymmetries.I Cereb BloodF/ow Metab 1989;9:S202.

23. Lammertsma AA, Cunningham VJ, Deiber MP, Combination of dynamicand integral methods for generating reproducible functional CBF images.J CerebBloodF/owMetab l990;l0:675-686.

24. Olesen J. Contralateral focal increase ofcerebral blood flow in man duringarm work. Brain 197l;94:634—646.

25. lngvar DH. Patterns ofbrain activity revealed by measurements of regionalcerebral blood flow. In: Ingvar DH, Lassen NA, eds. Brain Work—thecouplingoffunction. metabolismand bloodfiow in the brain.Copenhagen:Munksgaard; 1975:397—413.

26. Rubin P, Holm S, Friberg L, et al. Altered modulation of prefrontal andsubcortical brain activity in newly diagnosed schizophrenia and schizophreniformdisorder.ArchGenPsychiatry199l;48:987—995.

HandMovementsand @Tc-ExametazimeSPECT@ Ebmeleret al 1627

REFERENCES

1. Neirinckx RD, Canning LR, Piper IM, et al. Technetium-99m-d,l-HMPAO; a new radiopharmaceutical for SPECT imaging of regionalcerebral blood flow perfusion. JNuc/Med l987;28:191—202.

2. Andersen AR, Friberg H, Lassen NA, Kristensen K, Neirinckx RD. Serialstudies of cerebral blood flow using @‘Tc-HMPAO:a comparison with‘“Xe.Nuc/Med Commun 1987;8:549.

3. Yonekura Y, Nishizawa S. Mukai 1, et al. SPECT with Tc-HMPAOcompared with regional cerebral blood flow measured by PET: effects oflinearisation. J Cereb B/cod F/ow Metab l988S82-S89.

4.InugamiA,KannoI,UemuraK,etal.LinearisationcorrectionofTclabelledHMPAOimagein termsofregionalCBFdistribution:comparisonto C―02inhalation steady-state method measured by positron emissiontomography. J Cereb B/cod Flow Metab l988;8:S52—S60.

5. Gemmell HG, Evans NTS, Season JAO, et al. Regional CBF imaging aquantitativecomparisonof@Tc-HMPAO SPEC1@withC―O@PET.JNuc/Med1990;3l:l595—l600.

6. Heiss W-D, Herholz K, POdreka I, Neubauer I, Pietrzyk U. Comparisonof [@Tc]HMPAO SPECT with [‘8F]fluoromethanePET in cerebrovasculardisease.J CerebBloodFlowMetab 1990;lO:687—697.

7. Pupi A, Bisi 0, Sciagra R, Santoro 0, De Cristofaro MTR, Formiconi AR,Meldolesi U. The comparison ofbrain distribution ofHMPAO and microspheres in humans. I Cereb Blood F/ow Metab l989;9:S4l 1.

8. Woods sw, Hegeman IM, Zubal 10, et al. Visual stimulation increasestechnetium-99m-HMPAO distribution in human visual cortex. J NuclMed l99l;32:2l0—2l4.

9. Shedlack IU, Hunter R, Wyper D, McLuskie R, Fink G, Goodwin GM.The panern of cerebral activity underlying verbal fluency shown by splitdose single photon emission tomography (SPET) in normal volunteers.Psycho/Med 199l21:687—696.

10. Hunter R, Wyper DJ, Patterson 3, Hansen MT. Goodwin GM. Cerebralpharmacodynamics of physostigmine in Alzheimer's disease investigatedusing @mTc@HMPAOSPET imaging. BrJ Psychiat l991;l58:35 1—357.

11. Ebmeier KP, Dougall NJ, Austin M-P, et al. The split-dose technique forthe study ofpsychological and pharmacological activation with the cerebralblood flow marker @“Tc-exametazimeand single photon emission cornputed tomography (5PECT): reproducibility and rater reliability. hit IMethodsPsychialRes 199l;l:27—38.

12.ColebatchJG,PassingharnRE,FristonKJ,FrackowiackRSJ.Regional

by on December 1, 2018. For personal use only. jnm.snmjournals.org Downloaded from

1992;33:1623-1627.J Nucl Med.   Klaus P. Ebmeier, Catherine L. Murray, Nadine J. Dougall, Ronan E. O'Carroll and Guy M. Goodwin  Technetium-99m-Exametazime in Human Control SubjectsUnilateral Voluntary Hand Movement and Regional Cerebral Uptake of

http://jnm.snmjournals.org/content/33/9/1623This article and updated information are available at:

  http://jnm.snmjournals.org/site/subscriptions/online.xhtml

Information about subscriptions to JNM can be found at:  

http://jnm.snmjournals.org/site/misc/permission.xhtmlInformation about reproducing figures, tables, or other portions of this article can be found online at:

(Print ISSN: 0161-5505, Online ISSN: 2159-662X)1850 Samuel Morse Drive, Reston, VA 20190.SNMMI | Society of Nuclear Medicine and Molecular Imaging

is published monthly.The Journal of Nuclear Medicine

© Copyright 1992 SNMMI; all rights reserved.

by on December 1, 2018. For personal use only. jnm.snmjournals.org Downloaded from