j nucl med-1988-kern-181-6

7
 ositron emission tomography PET)isadiagnostic imaging technique that is capable of estimating tissue metabolism, noninvasivel y 1-2). Glucose metabolism has been evaluated in brain tumors with PET usingthe positron emitting glucose analog fluorine-18-2-fluoro 2-deoxy-r-glucose FDG) 3—6).Poorly-differentiated, rapidly growing cerebral neoplasms have higher ratesof glycolysis than well-differentiated, slowly growing tu mor s 3—6).Since the relationship between rate of glucose metabolism and tumor biology appears to apply tomanytumors 7-11) PET measurements ofglucose utilization rat e GUR) may be an useful method of noninvasively categorizinggradeofmalignancy. The PET-FDG technique has been used to image certain somatic tumors, including breast arcinoma  12) and metastatic liver cancer 13). Blood flow and oxygen utilization in human extremity sarcoma has been determined with PET and oxygen-15 methods  14 . To date, no report has app ared regarding PET imaging of human extremity musculoskeletal tumors ReceivedDec.31,i986;revisionacceptedJuly30,1987. Forreprintscontact:S. M. Larson,MD, NuclearMedicine DepCC,Bldg10,Rm lC4OI,9000RockvillePike,Bethesda, M d . . Present address: Medicina Nucleare—Istituto di Scienze Ra diologiche,2ndMedicalSchool,Naples,Italy. with FDG. This study was performed with the goal of imaging human extremity musculoskeletal tumors us ing PET and FDG. We correlated glucose utilization rate asmeasured by PET with histopathologic grading of these tumors. The results obtained support the use ofPET asanoninvasive method ofpreoperative assess mentofhuman musculoskeletaltumors. MATERIALS AND METHODS PatientPopulation Five patients with musculoskeletal tumors referred to the National Cancer Institute or to an orthopedic oncologist  M.M.)wereenteredintothestudyafterobtaininginformed consent.Theiragesrangedbetween 12and63 yr;fourwere male, and one was female.Severalpatients had small mci sionalbiopsiesperformed atother hospitalspriorto referral for definitive treatment, and other patients had no tissue diagnosispriortoscanning.Therewerenoclinicallyevident changes in the tumor site in those patients who underwent incisional biopsy prior to imaging. All patients underwent PETscanningpriortotumorresection.Theclinicaldataon thesefivepatients ispresented inTable 1. Positron EmissionTomography PatientswerestudiedunderIND 23195 (S. M. Larson, MD,Sponsor).Patientswerekeptfastingfor6 hrpriorto scanning. PET studies were performe using the ECAT II 1 8 1 olum e 29 †¢ um be r 2 †¢eb ru ary1988 MetabolicImagingofHuman Extremity Musculoskeletal Tumors by PET Kenneth A.Kern, Arturo Brunetti, Jeffrey A.Norton, Alfred E.Chang, Martin Malawer, Ernest Lack, Ronald D. Finn, Steven A. Rosenberg, and Steven M.Larson Surgery Branch, NC , DCT; Nuclear Medicine Department, CC; and Anatomic Pathology Branch, NIH, Bethesda, Maryland Themeasurementf gluco seutilization ate GUR by positronemissionomographyPET using18F-2-fiuoro-2-deoxy-D-glucose FDG is avaluablemethodtoassessthegradeof malignancyof brain tumors. We have designed a feasibilitytrial to determine whether PET couldbeusedtoimageandpredicthegradeofmalignancyfhumanextremity musculoskeletaltumors. Five patients with extremity tumors four soft-tissue tumors and one osteogenictumor werestudied.PeakandmeanapparentOURsweredeterminednthe tumorregion. Alltum ors weresubseq ue ntlyesected andg rad ed in a stan dar dashionusin g theNCIgradingsystem.Peakapparent GUR5rangedfrom3.3mg/iOOg/minto15.2mg/i00 g/min,with the highestvaluesfoundinthe highgradetumors.Althoughthenumberof patientsstudiedwassmall, a goodcorrespondenceasshownbetweenGURsand histopathologicgrading.OurresultsindicatethatPETcanbeusedtoimageandevaluatethe metabolicactivity of humanmusculoskeletaltumors. J NuciMed29:181—186, 1988 only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal use  jnm.snmjournals.org Downloaded from

Upload: citradl

Post on 06-Oct-2015

215 views

Category:

Documents


0 download

DESCRIPTION

j

TRANSCRIPT

  • ositron emission tomography (PET) is a diagnosticimaging technique that is capable of estimating tissuemetabolism, noninvasively (1-2). Glucose metabolismhas been evaluated in brain tumors with PET using thepositron emitting glucose analog fluorine-18-2-fluoro2-deoxy-r-glucose (FDG) (36).Poorly-differentiated,rapidly growing cerebral neoplasms have higher rates ofglycolysis than well-differentiated, slowly growing tumors (36). Since the relationship between rate ofglucose metabolism and tumor biology appears to applyto many tumors ( 7-11), PET measurements of glucoseutilization rate (GUR) may be an useful method ofnoninvasively categorizing grade of malignancy.

    The PET-FDG technique has been used to imagecertain somatic tumors, including breast carcinoma(12) and metastatic liver cancer (13). Blood flow andoxygen utilization in human extremity sarcoma hasbeen determined with PET and oxygen-15 methods(14). To date, no report has appeared regarding PETimaging of human extremity musculoskeletal tumors

    ReceivedDec. 31, i986; revisionaccepted July 30, 1987.For reprintscontact: S. M. Larson,MD, Nuclear Medicine

    Dep CC, Bldg 10, Rm lC4OI, 9000 Rockville Pike, Bethesda,Md.

    . Present address: Medicina NucleareIstituto di Scienze Ra

    diologiche,2nd MedicalSchool,Naples,Italy.

    with FDG. This study was performed with the goal ofimaging human extremity musculoskeletal tumors using PET and FDG. We correlated glucose utilizationrate as measured by PET with histopathologic gradingof these tumors. The results obtained support the useofPET as a noninvasive method of preoperative assessment of human musculoskeletal tumors.

    MATERIALS AND METHODS

    Patient PopulationFive patients with musculoskeletal tumors referred to the

    National Cancer Institute or to an orthopedic oncologist(M.M.)wereentered into the study after obtaininginformedconsent. Their ages rangedbetween 12 and 63 yr; four weremale, and one was female. Several patients had small mcisional biopsies performed at other hospitals prior to referralfor definitive treatment, and other patients had no tissuediagnosis prior to scanning. There were no clinically evidentchanges in the tumor site in those patients who underwentincisional biopsy prior to imaging. All patients underwentPET scanningprior to tumor resection.The clinicaldata onthese five patients is presented in Table 1.

    Positron Emission TomographyPatients were studied under IND #23195 (S. M. Larson,

    MD, Sponsor).Patients were kept fasting for 6 hr prior toscanning. PET studies were performed using the ECAT II

    181Volume29 Number2 February1988

    Metabolic Imaging of Human ExtremityMusculoskeletal Tumors by PETKenneth A. Kern, Arturo Brunetti, Jeffrey A. Norton, Alfred E. Chang,Martin Malawer, Ernest Lack, Ronald D. Finn, Steven A. Rosenberg, andSteven M. Larson

    Surgery Branch, NC!, DCT; Nuclear Medicine Department, CC; and Anatomic PathologyBranch, NIH, Bethesda, Maryland

    Themeasurementofglucoseutilizationrate(GUR)bypositronemissiontomography(PET)using18F-2-fiuoro-2-deoxy-D-glucose(FDG)is a valuablemethodto assessthe gradeofmalignancy of brain tumors. We have designed a feasibility trial to determine whether PETcouldbeusedto imageandpredictthegradeofmalignancyofhumanextremitymusculoskeletal tumors. Five patients with extremity tumors (four soft-tissue tumors and oneosteogenictumor)werestudied.PeakandmeanapparentOURsweredeterminedin thetumorregion.Alltumorsweresubsequentlyresectedandgradedina standardfashionusingtheNCIgradingsystem.PeakapparentGUR5rangedfrom3.3 mg/iOOg/minto 15.2mg/i00g/min,with the highestvaluesfoundin the highgradetumors.Althoughthe numberofpatientsstudiedwassmall,a goodcorrespondencewasshownbetweenGURsandhistopathologicgrading.Our resultsindicatethat PETcanbe usedto imageandevaluatethemetabolic activity of human musculoskeletal tumors.

    J Nuci Med 29:181186,1988

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from

  • PETApparent

    glucoseutilizationrate(mg/PatientAge

    (yr),1009/mm)no.sexLocationSize (cm)Pathologic diagnosisPeakMean163,M(L)migh12x14Malmgnantfibroushis

    tiocytoma, Grade Ill15.189.10222,

    M(L) Thigh1 1 x 15Malignant fibroushistiocytoma, Grade II6.05

    4.50324,

    MR)t Thigh8 x 9Myxoid liposarcomaGradeI4.41

    3.24438,

    M(L) Femur7 x 13Cystic giantcelltumor3.702.76512,F(R)CaIf6x11Plexiformneurofibroma

    (benign)3.352.64.

    L =Left.tA = Right.

    TABLE IGlucose Utilization Rates in Extremity Tumors

    scanner (EG & 0 ORTEC, now CTI, Oak Ridge, TN). ECATII has 66 sodium iodide crystals arranged in a hexagon,producing one slice per scan. The PET scans were acquired inmedium resolutionacquisition mode, usinga medium resolution reconstruction filter. In this mode, the inplane resolution is 1.7 cm and the slice thickness is 2 cm.

    All tumors were large relative to the inplane and Z-axisresolution of the ECAT-Il scanner (Table 1). The location ofthe tumor was determined by inspection and palpation in allcases. The midpoint of the tumor was obtained, a line wasdrawn across the tumor, and two parallel lines were drawn at2 cm above and 2 cm below the midpoint. Using the horizontal laser line, the ECAT was positioned at the level of thecephalad slice. Before scanning was begun, the patients position was stabilized in relationship to the scanning bed usingsand-bags (out of field of view of scanner) and tape. Inaddition, it was determined that outlining the patient's bodycontour on an immobilized sheet fixed to the scanning bedwas a usefulmethod of maintaining patient position duringthe collection of data. In general, three planes were imaged:at the midpoint,at 2 cm abovethis point, and at 2 cm belowthe mid-plane slice. Transmission scans for attenuation correction were obtained with a gallium-68 source (1O@counts)in each patient studied prior to the injection of FDG.

    PET scanswereacquiredbeginning35 mm after i.v. injection of 5 mCi of FDG into an arm vein. In a typicalemissionscan ofthe tumor region, @@-0.5l.Ox lO@counts were collectedper image, over 200-600 sec of acquisition time. Care wastaken to ensure that the patient did not change positionbetween transmission and emission scans.

    Using a heating pad to warm the dorsum of the handopposite to the injection site, arterializedvenous blood was

    sequentially drawn during the study. After injection of theFOG, 25 to 30 blood samples of 0.5 cc each were drawn at30-sec intervals for the first 5 mm; at 1-mm intervals beginning510mm postinjection; at 5-mm intervals beginning 1020mm postinjection; and at 10-mm intervals thereafter.

    The apparentglucoseutilizationrate (OUR)wascalculatedusing the Sokoloff model (15), and an operational equationthat includes a k4or dephosphorylationcorrection term(16). The uptake of EDO in tumor was measured with thePET scannerat 3565mm postinjection.The integralof thetime-activity FDO plasma curve was used as the input function. Valuesfor the rate constantsand lumpedconstantcomefrom Phelps et al. (1 7). The lumped constant (LC) and graymatter transfer rate constants (Ks) used in the calculationswere: LC = 0.4 18, K1 = 0. 1020, K2 = 0. 1090, K3 = 0.0620,K4 =0.0068.

    The mean and peak OURS were obtained using a region ofinterest (ROI) program available on the ECAT-II system(ORSAY ROIprogram).A variableROIwas usedthatencompassed the entire cross-sectional area of the tumor on thetransaxial image. In all cases, the tumor was readily demarcated from surrounding tissue. Care was taken to ensure thatthe edgesof the ROI werejust within the boundary of theapparent active metabolic zone of tumor. The mean OURwas computed as the average value on a pixel-by-pixel basiswithin the total region. The peak value was the maximumpixelvalue in that region.All of the availableimagingplaneson each patient were analyzed, and the peak values reportedin Table 1and Figure4 arethe highestvalues obtained amongthe available slices. The mean value reported is the meanvalue in the tumor region acquired from the transaxial slicecontaining the highest pixel value within the tumor ROI.

    182 Kern, Bruneth, Nortonet al TheJournalofNuclearMedicine

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from

  • ASAL

    A

    B

    6-01NC! 50GLUCOSE

    J44 4.8 RATE4 . 6RCNRCL4.44.13.93.73.53.33.12.9R 2.72.52.32.11.91.7I .5

    NIH NUCLEAR MEDIC INE I.31.1

    FIGURE 1Patient 3: Low-grade sarcoma of theright proximal thigh (myxoid liposarcoma, grade I).A: The CT scanshows the tumor as a relatively lowdensity area anterior to the femur. B:In thePETimagethetumor(arrow)shows inhomogeneous uptake ofFDGwithpeakapparentGURof4.41mg/i00 g/min and mean apparentGUR 3.24 mg/i 00 g/min. The levelofthePETimageisslightiysuperiorcompared to the CT image (b = bladder).

    IBecause of the large size of these tumors (see Table 1), theROl was alwayslargein relationshipto the resolutionof theimaging equipment.

    SurgeryAll patients underwent resection of tumor based upon

    preoperative pathologic diagnosis determined by biopsy. PETresults were not used to determine the need for operation orits extent. After resection tumors were graded in a standardfashion using the NCI grading system (18). The pathologistsdid not know the results of PET prior to histologic grading ofthe resected tumor.

    RESULTSAll tumors accumulated FDG and were evident as

    hotspotsin the images. PET images are shown inFigures 13,along with the corresponding computedtomographic (CT) scans in three patients. CT scans areshown with the left-right orientation reversed to enableeasy comparison with the PET images.

    The peak and mean GUR for the five tumors wasdetermined. These values are listed in Table 1. Peakapparent OUR varied between 3.3 mg/l00 g/min and

    . 15.2 mg/i g/min. Mean apparent GUR varied be

    tween 2.6 mg/lOO g/min and 9.1 mg/lOO g/min, andthe rank order was identical for the five patients. Therewas good correspondence between increasing peak andmean GUR and increasing grade ofmalignancy. Figure4 is a scatter pli of peak OUR compared with gradeof tumor. There was a step-wise correspondence between grade oftumor and increasing OUR. Because ofthe small numbers in this preliminary study, we did notperform formal correlation analysis.

    DISCUSSIONPET determinations of OUR have been used to

    predict the biologic behavior ofcertain types of cerebralneoplasms. For example, DiChiro et al. have imaged

    Volume29 Number2 February1988 183

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from

  • A0 0L A A

    4-016 . [email protected] PATE6 . [email protected]

    ;T 3.9

    3.7: 3.4

    2.92.6

    @; 2.4

    2.1

    J31

    FIGURE 2Patient2: Intermediate-gradesaxcoma of the left proximal thigh (malignant fibrous histiocytoma,gradeII). A: The CT scanshowsa largetumor mass in the antero-medial regionof the leftthigh.B: In the PETscan the tumor mass is shown as anarea of relatively high FDG uptakewith a peak apparent GUR of 6.05mg/iOO g/min and mean apparentGUR of 4.50 mg/i00 g/min.Utiletracer uptake is observed at the leveloftherightthigh,consistentwiththefact that normal resting muscle hasextremely low glycolitic actMty.

    NCI

    L R

    NIH NUCLEAR MEDICINEB

    over 300 cases ofbrain tumors and have shown a strongcorrelation of tumor grade with OUR (4,6,19). Additionally, Patronas and DiChiro have reported that OURreflects the prognosis in patients with giiomas (5). Patients with tumors exhibiting high GUR by PET havea much lower mean survival time than patients withtumors showing low GUR (5).

    Theoretically, musculoskeletal tumors (particularlysarcomas) are excellent candidates for FDG-PET imaging, since glucose uptake rates in extremity sarcomasare known to be high (20). In two of the tumors fromthe present study, glucose 6-phosphatase enzymatic activity was measured, and found to be very low (KernKA, Norton JA, personal communication). The deoxyglucose method as originally described by Sokoloff(15) is valid in principle under conditions where glucose

    6-phosphatase activity is low during the evaluationperiod.

    We are aware that the values for OUR obtained inthis study are approximate, since the lumped constantand the transfer rate constants for the operational equation are not known for human musculoskeletal tumors.The values of tumor OUR reported here are estimatesbased upon the lumped constant (LC) and the kinetictransfer constants of gray matter (k's) in the same waythat Di Chiro et al. have used LC and gray matter k'sfor the determination of GUR in cerebral tumors (6).Such estimates ofGUR, although admittedly imperfect,appear to serve as a useful index of tumor metabolicactivity in human brain tumors. Our preliminary resultssuggest that a similar set of approximations may beused to compute estimates of OUR that may correlate

    1 84 Kern, Bruneth, Norton et al TheJournalofNudearMedicine

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from

  • NC I

    J184

    FIGURE 3Patient 1: High grade sarcoma of theleft proximal thigh (malignant fibroushistiocytoma, grade Ill). A: The CTscan shows a large tumor mass inthe antero-medial region of the leftthigh.B: InthePETscanthe lateralportion of the tumor mass shows avery high FDG uptake with a peakapparentGMR of i 5.18 mg/i00 g/mmand mean apparent GMR of 9.i 0mg/i00 g/min.

    with important differences in the tumor biology ofhuman musculoskeletal tumors. As in cerebral giiomas,histopathologic grading is the single most important

    . prognostic factor in patients with soft-tissue musculo

    skeletal tumors (21).This is the first reported study of PET-FDO imaging

    of human extremity musculoskeletal tumors. This feasibiity study indicates that these tumors accumulateFDO and are readily imaged with good contrast bypositron tomography. Further prospective studies areplanned to assess the clinical contribution of PET studies in the assessment ofhuman musculoskeletal tumors.

    TumorGrade ACKNOWLEDGMENTSFIGURE 4Scatter plot of apparent peak glucose utilization rates@ The authorsthank Dr. Giovanni Di Chiro for his valuableversustumorgrade.Glucoseutilizationratesareincreased commentsandadviceindesigningthisstudy.in tumors of higher malignant grade. A. Brunettiis a recipientof a FogartyFellowship.

    RL

    14 . 9GLUCOSE14.2 RATE13 . 4RCMRGL12.711.911.110.4

    9.68.98.17.46.65.95.1

    :.@ 4.3

    2.82.1

    @ 1.30.6

    [email protected]:@@T@@@

    NIH NUCLEAR MEDICINE

    A

    B

    TUMOR METABOUC RATES ACCORDING TO TUMOR GRADE

    Gred. III0

    @ 16

    C.2 GradII0@@.

    .@ C

    @ED @,@ orB.nign

    (3@5E

    @ [email protected].

    Volume29 Number2 February1988 185

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from

  • REFERENCES accumulation of 2-deoxy-2-['8flfluoro-D-glucose inliver metastases from colon carcinoma. J Nuci Med1982;23:11331137.

    14. Brownell OL, Kairento A-L, Swartz M. Positron emission tomography in oncologythe MassachusettsGeneral Hospital experience. Semin Nucl Med 1984;14:324341.

    15. Sokoloff L, Reivich M, Kennedy C, et al. The [4C]Deoxyglucose method for the measurement of localcerebral glucose utilization: theory, procedure, andnormal values in the conscious and anesthetized albinorat.JNeurochem 1977;28:897916.

    16. Brooks RA. Alternative formula for glucose utilizationusing labelled deoxyglucose. J Nucl Med 1982;23:538539.

    17. Phelps ME, Huang S-C, Hoffman El, et al. Tomographic measurement of local cerebral glucose metabolic rates in humans with [8fl2-fluoro-2-deoxy-i>glucose: validation of a method. Ann Neurol 1979;5:371388.

    18. Costa J, Wesley RA, Gladstein E, et al. The gradingof soft tissue sarcomas. Results of clinicohistopathologic correlation in a series of 163 cases. Cancer 1984;53:530541.

    19. DiChiro 0, Brooks RA, Bairamjan D, et al. Diagnosticand prognostic value ofpositron emission tomographyusing [8flfluorodeoxyglucose in brain tumors. In:Reivich M, Alavi A, eds. Positron emission tomography, New York: Alan Lisa, 1985.

    20. Norton JA, Burt ME, Brennan MF. In vivo utilizationofsubstrate by human sarcoma-bearing limbs. Cancer1980;45:29342939.

    21. Rosenberg SA, Suit HD, Baker LH. Sarcomas of softtissue. In: Devita VT, Heilman S. Rosenberg SA, eds.Cancer, principles and practice of oncology, secondedition. Philadelphia: J.B. Lippincott, 1985:1256.

    22. Abe Y, Matsuzawa T, Fujiwara T, et al. Assessmentof radiotherapeutic effects on experimental tumorsusing 8F-2-fluoro-2-deoxy-D-glucose. Eur J Nucl Med1986; 12:325328.

    1. Reivich M, Kuhi D, Wolf A, et al. The [8F]fluorodeoxyglucose method for the measurement of localcerebral glucose utilization in man. Circ Res 1979;44:127137.

    2. Phelps ME, Mazziotta JC, Schelbert HR, eds. Positronemission tomography and autoradiography: principlesand applications for the brain and the heart. NewYork: Raven Press, 1986.

    3. DiChiro 0, Brooks RA, Patronas NJ, et al. Issues inthe in vivo measurement of glucose metabolism ofhuman central nervous system tumors. Ann Neurol1984;15(Suppl):Sl38Sl46.

    4. DiChiro 0, DeLaPaz RL, Brooks RA, et al. Glucoseutilization ofcerebral gliomas measured by [8F]fluorodeoxyglucose and positron emission tomography.Neurology1982;32:13231329.

    5. Patronas NJ, DiChiro 0, Kufta C. Prediction of survival in glioma patients by means ofpositron emissiontomography. JNeurosurg 1985; 62:816822.

    6. DiChiro 0. Brain imaging of glucose utilization incerebral tumors. In: Sokoloff L, ed. Brain imagingand brainfunction. New York: Raven Press; 1985.

    7. Warburg 0. The metabolism oftumors. London: Constable&Co., 1930.

    8. Weber 0. Enzymology ofcancer cells (part I). NEJM1977; 296:486541.

    9. Weber 0. Enzymology of cancer cells (part II). NEJM1977;296:541551.

    10. Weber 0. Biochemical strategy ofcancer cells and thedesign of chemotherapy: O.H.A. Clowes memoriallecture.CancerRes1983;43:34663492.

    11. Weinhouse S. Oxidative metabolism of neoplastic tissues.Adv CancerRes1955;3:269325.

    12. Beaney RP. Positron emission tomography in thestudy of human tumors. Semin Nucl Med 1984;14:324341.

    13. Yonekura Y, Benua RS, Brill AB, et al. Increased

    1 86 Kern, Brunetti, Norton et al TheJournalof NuclearMedicine

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from

  • 1988;29:181-186.J Nucl Med.

    A. Rosenberg and Steven M. LarsonKenneth A. Kern, Arturo Brunetti, Jeffrey A. Norton, Alfred E. Chang, Martin Malawer, Ernest Lack, Ronald D. Finn, Steven

    Metabolic Imaging of Human Extremity Musculoskeletal Tumors by PET

    http://jnm.snmjournals.org/content/29/2/181This article and updated information are available at:

    http://jnm.snmjournals.org/site/subscriptions/online.xhtmlInformation 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 1988 SNMMI; all rights reserved.

    only. by Indonesia: J of Nuclear Medicine Sponsored on February 17, 2015. For personal usejnm.snmjournals.org Downloaded from