study compares the differential function as assessed by@Tc-MAG3and @Tc-DMSAas well as investigates the

sensitivityand specificityof@mTc@MAG3in the detectionof a parenchymalrenal defect.

MATERIALS AND METhODS

Fifty-nine children, between the ages of0.l-l4 yr (median 5.3yr), 28 boys and 31 girls, were referred 12—24wk after a urinarytract infection (UT!) for both a @mTC@DMSAand a @mTc@MAG3scan and an IRC. These two scans were undertaken w@ithin4 wkof each other. The @“Tc-DMSAscan wascarriedout accordingto the protocolof Gordonet a!. (8). A posteriorviewand twoposterior oblique views (250 K counts/per view) were acquiredby the gamma camera linked to a dedicatedcomputer in a 256x 256 matrix, 6 hr after the injection of @Tc-DMSA,the dosescaled on a body surface area basis to an adult dose of 100 MBq.Static imageswere recordedon a dedicatedcomputer and analyzed for focal defects. Differential function with backgroundsubtractionwasalsocalculated.

The dynamic 99mTc@@4@G3study was carried out with thechild supine on the gamma camera with a dynamic 20 mmacquisition in 10 sec frames. The first 34 children had a dosebased on an adult dose of8O MBq (Group 1)while the subsequent25 childrenhad a dose basedon an adult dose of 112 MBq(Group 2) scaled on a body surfacearea. Renal, perirenal andcardiac regions of interest (ROIs) were created and time-activitycurvesweregeneratedforall fiveROIs.Differentialrenalfunctionwascalculatedfromthe slopeofthe backgroundsubtractedrenalcurvesin the Patlak plot (9) as popularizedby Rutland (10) andvalidated at this institution (11), the analysis was undertakenbetween40—120sec followingthe injection. The theory of thePatlak plot as applied to an RO! of the whole kidney is wellestablished,however,its use on a pixel basis is new (12). Functional clearanceimagesof the kidneyswerecreatedon a pixelbypixel basis between 40—120sec using the Patlak plot. This required regroupingthe imagesinto 20 sec frames,compressingto64 by 64 matrix and smoothing the data. The curve from thecardiacRO! wasusedwith the individualcurvesgeneratedfromeach pixel to generate a Patlak plot, a functional image wascreatedwhereeach pixel is assigneda value proportionalto theslope of its Patlak plot. Those pixels which show progressiveaccumulation of isotope will exhibit higher values than thosewhich do not (12). The 25 children in Group 2 also had asummedimagecreatedovertheperiod40-120sec.

A kidney was considered abnormal if the differentialfunctionon @“Tc-DMSAwas<43% (8) and/or ifa focaldefectwasseen.

The presenceof focal renaldamage dictatesdifferentmanagementof a childwith urinarytract infection(UTI)comparedwith childrenwho have normal kidneys.Technetium-99m-dimercaptosuccanlc(DMSA) has a high sensitivity in the detection of a focal defect, and allows estimation of differentialfunction. The introduction of @rc-MAG3with high renalextractionsuggeststhat this may be usefulin childrenwithUTI but its role remains speculative. Fifty-nine children withpreviousUTIunderwentboth @‘Tc-DMSAandMAG3within4 wk of each other. DifferentialfUnCtiOnand assessment ofthe images were undertaken.There is close correlation(R2=0.97)between the differentialfunction.Analysisof the

@Tc-DMSAand functionalMAG3imagesshowedthat thefunctionalimagehada specificityof 88% anda sensitivityof88% in the detectionof a focalparenchymaldefect.Technetium-99m-MAG3 in the dinical sethng of a child with UTIallows accurate assessment of differential function and a highprobabilityof detectinga focalrenalabnormality.

J NucI Med 1992; 33:2090—2093

echnetium-99m-dimercaptosuccithcacidscan(DMSA)is established as the most sensitive method to detect arenal scar (1,2). The clinical importance ofa renal scar inpediatrics lies in the increased risk of developing systemichypertension (10%) in later years (3) and that refluxnephropathy is the second most common cause of childrenrequiring dialysis and transplantation (4). Indirect radionuclide cystography (IRC) has become established in manyinstitutions (5-7), the introduction of@Tc-mercaptoacetyltriglycine (MAG3) has further popularized this technique. In institutions where the IRC is carried out, information about renal function is availableat no additional“cost―by acquisition of a renogram. If focal renal abnormalities could be detected, then no additional examinationwouldbe required.Technetium-99m-MAG3requiresevaluation for its potential to assess differential renal functionand detect focal renal parenchymal abnormalities. This

ReceIved Sept. 24, 1991 ; revIsion accepted Jul.23, 1992.For reprkits contact: Isky Gordon, Department of Ra@ology, Hospftal for

Sick Children,Great Ormond St., London, EnglandWC1N 3JH.

2090 The Journalof NuclearMedicine•Vol. 33 •No. 12 •December 1992

Can Technetium-99m-MercaptoacetyltriglycineReplace Technetium-99m-DimercaptosuccinicAcid in the Exclusion of a Focal Renal Defect?IskyGordon, Peter J. Anderson, Mark F. Lythgoe, and Michael Orton

Department ofRadiology, Hospizalfor Sick Children and The Institute ofChild Health, London England

by on April 8, 2018. For personal use only. jnm.snmjournals.org Downloaded from

MAG3FunctionalClearanceImageNORMALABNORMALe

I4U)

0-I

4

0537m

4644

MAG3FunctIonalClearanceImageNORMALABNORMALC

.@

4U)

0-j

4

0z4850

zm4

The 99mTcDMSAimages,functionaland summed @mTc@MAG3images were reviewed independently without knowledge of theother investigation.For this comparison each kidney in eachinvestigationwasscoredaseithernormal,uncertain,or abnormal,those eight kidneys regarded as uncertain were excluded. There

was one child with a single kidney, leaving 110 kidneys foranalysisin these 59 children. In Group 1,there were63 kidneysfor analysis,whilein Group 2 there were47 kidneysfor analysis.The kidneysin Group 2 classifiedas abnormal werealso scoredfor the site of the focal defect. Defects were categorized asoccurringin either the upper pole, the lowerpole or multiple.

Statistical AnalysisSimplelinear regressionanalysiswasappliedusingthe princi

plcofleast squaresto the DMSAand MAG3differentialfunctiondata When analyzing the images to create the sensitivity andspecificity from the statistical decision matrix, a normal kidneywas classifiedas a negative result; therefore a normal @mTc@DMSAimageanda normal @“Tc-MAG3functionalimageis a“truenegative.―

RESULTS

DifferentialFunctionAnalysisThe percentage function of the left kidney in each child

has been compared using both radiopharmaceuticals. Thechild with a single kidney has been excluded (n = 58).Closecorrelationbetweenthe differentialfunctionof thesetwo investigations was found (R2 = 0.97 p = <0.001, Fig.1).

ImageAnalysisThe resultsof the DMSAimageshave been considered

asthe referencemethod in declaringthe kidney asnormalor abnormal. In the overall results (n = 110), sensitivitywas 88% and specificity 88% (Table 1). There was no

FIGURE1. Plotofdifferentialfunctionof @“Tc-DMSAagainstthat of @“Tc-MAG3,n = 58. The lineof regressionhas beendrawn.

TABLE IComparison Between @rc-DMSAand @Tc-MAG3

FunctionalImages(n = 110 Kidneys,Sensitivity= 88%,Specificity= 88%)

difference between those children who received the lowerdose of 99mTcMAG3 from those who received the higherdose. In group 1 (n = 63), sensitivitywas 88% and specificity 89%. In group 2 (n = 47), sensitivity was 88% andspecificity 86%.

Image Analysis of Kidneys with Normal DifferentialFunction (43%-57%)

Comparison ofthese 72 kidneys showed a sensitivity of79% and specificity of9 1% (Table 2). In group 1 (n = 52),the sensitivity was 80% and specificity 89%. In group 2,(n = 20) the sensitivity was 75% and specificity 94%.

Technetium-@,vm-MAG3Summed Images Comparedwith @rc-DMSAImages

Comparison between the summed 99mTcMAG3 imageand the 99mTcDMSA images in Group 2 gave a sensitivityof 96% and specificity of 39% (Table 3).

TABLE2Comparison Between @TC-DMSAand @Tc-MAG3

FunctionalImagesof 3 Childrenwith NormalDifferentialFunction (n = 72 Kidneys, Sensitivity 79%,

Specificity = 91%)

100

80

60

n = 58

p = 0.001

r@4=0.97

C@)04

‘@40U-0

20

2091Focal Renal Defects in @‘TcMAG3 and @“Tc-DMSAImages •Gordon et al

DFofDMSA

by on April 8, 2018. For personal use only. jnm.snmjournals.org Downloaded from

MAG3 SummedImageNORMALABNORMALa

4Cl)

0-J

4

0z914124

TABLE3Group 2—ComparisonBetween @Tc-DMSAand @TcMAG3 SummedImagesof 25 Children(n = 48 Kidneys,

Sensitivity = 96%, Specificity 39%)

FiGURE3.A@cMAG3 functionalclearance imageshowsa focal defect in the upperpole of the left kidney.Notetheslightbackgroundactlvity,especiallyintheregionof the liver.

tical compared to 99mTc..DTPA,yet provides high qualityimages and good statistics in the renal regions for dataanalysis (14,16). Children with UT! are referred for imaging, which currently includes a @mTc@DMSAscan, tosearch for a focal defect, and if the child is toilet trained,an IRC is performed using 99mTcMAG3. In such institutions, the images obtained during the early phase of thedynamic renogram have led clinicians to ask whether onecould not use this information to replace the @mTc@DMSAscan in the detection of a renal parenchymal defect.

Detailed analysis ofthe early phase ofthe @mTc@MAG3study is required ifa focal renal parenchymal abnormalityis to be excluded. Using the simple summed image, asensitivity of 60% was reported when compared to 99mTc@DMSA (1 7). This is the first clinical study to use a functional image representing clearance based on the Patlakplot. Theoretically, this image may increase the chance ofrevealing a focal defect. The image created is based on ananalysis pixel by pixel over the entire field of view of thegamma camera where each pixel is assigned a value proportional to the slope of the Patlak plot during the period40—120sec(12).

If the differential function is less than 43% on @mTc@DMSA, then the kidney may be classified as abnormal

FIGURE4.A@TcMAG3summedimage ShOWSa focaldefect in the upperpole of the left kidney. Note that thebackground activity is most markedover the region ofthe liver.

4

Analysisof FocalDefectsWhen the exact site ofthe focal defect was correlated in

the 22 kidneys which were abnormal on both DMSA andMAG3 from Group 2, two kidneys were excluded becausethere was global reduction in uptake of isotope with nofocal defect on either @mTc@DMSAor @mTc@MAG3.Theremaining 20 kidneys showed that more defects were seenon the 99mTcDMSA scan than on the functional 99mTcMAG3 image (Figs. 2—4).There was, however, good correlation with the exception of one kidney which showedan upper pole defect on the @mTc@MAG3functional imageand a lower pole defect on the 99mTc@DMSAimages (Table4).

DISCUSSIONIn pediatric practice, there is a need to attempt to reduce

the radiation exposure wherever possible. The introduction of 99mTcMAG3 for dynamic renal scan with a highplasma protein binding (90%) and a high renal extractionfraction (40%) results in 70% ofthe isotope in the bladderat 30 mm (13). 99mTcMAG3 is therefore an ideal agentfor use whenever drainage of the kidney requires investigation (14) or IRC is required (15). The high renal extraction allows the use of a smaller dose of radiopharmaceu

FIGURE2.A@'rcDMSAscanshowsafocaldefectintheupper pole of theleftkidney.

2092 The Journalof NuclearMedicine•Vol. 33 •No. 12 •December 1992

by on April 8, 2018. For personal use only. jnm.snmjournals.org Downloaded from

MAG3FunctionalClearanceImageUPPERLOWERMULTIPLEeUPPER302LOWER20MULTIPLE14

TABLE 4Group2—Comparisonof Site of FocalDefecton

DMSA and Functional @‘Tc-MAG3Images(n = 20)

59 children with 109 kidneys is encouraging. Group 2 had

a 25% increase in the amount of @Tc-MAG3injectedyet there is no significant difference between the sensitivityand specificity of these two groups. One possible explanation is that to achieve an improvement when the sensitivityand specificity are 88% may require a larger increase inthe renal count rate than that which resulted from theincreased activity injected in Group 2 of this study.

The results suggest that in routine clinical practice, a9@―Tc-MAG3scan will provide accurate differential function and a functional image will detect a focal parenchymalabnormality with a sensitivity and specificity of 88%. Theresults do not change significantly with a 25% increase inadministered activity.

ACKNOWLEDGMENTS

Mr. M. Lythgoeand Mr. M. Orton are supported by a grantfrom the WeltonFoundation.

REFERENCES

1. Merrick MV, Uttley WS, Wild SR. The detection ofpyelonephritic scarringin children by radioisotope imaging. Br JRadiol 1980;53:544—548.

2. Goldraich NP, Ramos OL, Goldraich 114.Urography versus DMSA scanin children with vesicoureteric reflux. Pediatric Nephrol 1989;3:l-5.

3. Still JL, and Cottom D. Severe hypertension in childhood. Arch Dis Child.1967;42:34—37.

4. Combined report on Regular Dialysis and Transplantation in Europa,X1X, 1988. NephrolDiaiysis Transplant 1989;4:l2.

5. Gordon I. Indirect radionucide cystography—the coming of age. NuciMed Commun l989;lO:457—548.

6. Gordon I, Peters AM, Morony S. Indirect radionucide cystography: asensitive technique for the detection of vesico-ureteral reflux. PediatrNephroll990;4:604—607.

7. Chapman 53, Chantler C, Haycock GB, Maisey MN, Saxton HM. Radionucide cystography in vesico-ureteric reflux. Arch Dis Childhood 1988;63:650—653.

8. Gordon I, Evans K, Peters AM, et al. The quantitation of@Tc-DMSA inpaediatrics NuclMedCommun l987;8:66l-667.

9. Patlak CS, Blasberg RG, FenstermacherJD. Graphical evaluation of bloodto brain transfer constants from multiple time uptake data. I Cereb BloodFlowMetab l983;3:l—7.

10. Rutland MD. A comprehensive analysis of renal DTPA studies@Theoryand normal values. NuciMed Commun 1985;6:ll-l5.

11. Kelleher JP, Anderson PJ, Gordon 1, Snell ME. Estimation of glomerularfiltration rate in the miniature pigby kidney uptake on the gamma camera.NuclMed Commun 199l;l2:8l7—822.

12. Lythgoe MF, Anderson PJ, Gordon I, Surman P3, Orton M, ToddPokropeck A. Can Tc-99m MAG3 replace Tc-99m DMSA in the exclusionofa renal scar? [Abstract]. Nuc/Med Commun l992;l3:369.

13. Taylor A, Eshima D, FItZberg AR, Christian PE, Kasina S. Comparison ofiodine-131-OIH and technetium-99m-MAG3 renal imaging in volunteers.JNuclMed l986;27:795—80l.

14. Jafn RA, Britton KE, Nimmon CC, et al. Technetium-99m MAG3, acomparison with iodine-l23 and iodine-13l orthoidohippurate in patientswith renal disorders. JNuclMed l98829:147-152.

15. Pickworth FE, Vivian GC, Franklin K, Brown EF. Tc-99m MAG3 inpediatric renal tract disease. Br JRadiol l992;65:2l—29.

16. Taylor A, Eshima D, Christian PE, Milton W. Evaluation of Tc-99mMAG3 with patients with impaired renal function. Radiology l987;162:365—369.

17. Muller Suur R, Haenadler L, Zachrisson I, Mesko L. Combination of Tc99m-MAG3 gamma camera renography and Tc-99m-DMSA scans for theevaluation ofacute pyelonephritis in children [Abstract]. Nucl Med Commun 1992;13:375.

(8). In this homogeneous group ofchildren with suspected

VUR and no evidence of obstruction, the results showclose correlation between the differential function as measured by @“Tc-DMSAand 99mTc@MAG3.This suggests thatthe kidney may be regarded as definitely abnormal if thedifferential function is <43% on @“Tc-MAG3.Furthermore monitoring of the differential function may then beaccurately achieved using @mTc@MAG3alone.

When the differential function is normal (43%—57%),there may be bilateral equal renal damage or two normalkidneys present. However, in this group the functionalimage has been shown to detect the focal abnormality witha decrease in the sensitivity (79%) but not change in thespecificity (88%). This suggests that even in this importantsubgroup of children, the functional image provides valuable additional information at no extra burden to thechild.

The correlation between results of the exact site of thefocal abnormality revealed more abnormalities on @mTc@DMSA scanthan on functional @mTc@MAG3image, anexpected result. The explanation is probably related to thedata acquisition since the images from the @mTc@DMSAscan are static images acquired on either a high resolutionor pin holecollimatorfor a relativelyprolongedperiodoftime, while the functional image is acquired on a highsensitivity collimator and created over the time period 40—120 sec in a 64 x 64 matrix.

The summed images were noisy, resulting in a highfalse—positiverate and giving this image a specificity of39% which was unacceptable in clinical practice. Since theonly additional information required to generate the functional clearance image (which the computer generates inunder 2 mm) is positioning of the cardiac ROl, there islittle point in generating only a summed image whenlooking for a focal parenchymal abnormality.

The high sensitivity 88% and specificity 88% in these

Focal Renal Defects in @“TcMAG3 and @“Tc-DMSAImages •Gordon et al 2093

by on April 8, 2018. For personal use only. jnm.snmjournals.org Downloaded from

1992;33:2090-2093.J Nucl Med.   Isky Gordon, Peter J. Anderson, Mark F. Lythgoe and Michael Orton  Technetium-99m-Dimercaptosuccinic Acid in the Exclusion of a Focal Renal Defect?Can Technetium-99m-Mercaptoacetyltriglycine Replace

http://jnm.snmjournals.org/content/33/12/2090This 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 April 8, 2018. For personal use only. jnm.snmjournals.org Downloaded from