afferent intravascular injection of the radiopharmaceutical (8).In a previous study, we used a two-compartment model tocalculate HEE of @Tc-mebrofeninin the dog. The forward(k1 ) and reverse (k2) rate constants were determined by asimple graphical method using linearized time-activity data(9—11). We compared this graphical method with more complex direct numerical solutions to the model using a nonlinearleast-squares algorithm (8). We found no significant differencebetween the two methods of calculating HEE from directafferent injection (8). The graphical method used was chosenbecause of its simplicity in calculating HEE. HEE determinedfrom afferent vascular injection is a direct measure of first-passhepatic extraction but is limited to research application becauseof the invasiveness of achieving direct afferent vascular access.

The term hepatic extraction fraction (HEF) traditionally hasbeen used to describe the hepatic extraction after deconvolutional analysis of data obtained from intravenous injection(1,4—6).The normal HEF of @mTc@mebrofeninin humans isgenerally considered to be 100% (4). In normal dogs, HEF of

@@Tc-mebrofeninis 92%—100%. In normal dogs, there is nosignificant difference in HEF and HEE (12).

HEF of 99mTcmebrofenin using deconvolutional analysishas been reported as a quantitative measure of hepatocytefunction (1,2,5,6,13,14). With hepatocellular disease, hepaticextraction is expected to decrease proportionally to the severityof liver insult (1,4—6). Numerous articles have been writtenthat show the usefulness of HEF in evaluating patients withliver disease (1—4,6,7,14—18).The purpose ofthis study was todetermine if the hepatic extraction of v9mTcmebrofenin couldbe correlated to histopathologic changes in the liver. Wecompared the hepatic extraction of 99mTc@mebrofenin,measured directly from an afferent vascular injection, to quantitative histopathology in dogs with acute toxic injury. In addition,we compared the hepatic extraction of @mTc@mebrofenintochanges in serum biochemistry analysis.

MATERIALS AND METhODSTen conditioned, adult mongrel dogs (6 female, 4 male; body

weight range 14.5—29.1kg; mean 21.39 kg) were obtained from thelaboratory animal section of the University of Tennessee, Collegeof Veterinary Medicine. The dogs were housed in facilities at theUniversity of Tennessee that are accredited by the AmericanAssociation of the Accreditation of Laboratory Animal Care, andall protocols were approved by the University of TennesseeInstitutional Animal Care and Use Committee. Preliminary dataincluded a physical examination, serum biochemistry analysis andfasting bile acid concentrations. Portal scintigraphy was performedto exclude the presence of a portosystemic shunt that would haveinvalidated the HEE data (19—21).

The radiopharmaceutical kits containing 45 mg mebrofenin(Choletec, Squibb Diagnostic, Princeton, NJ) were reconstituted

Extraction of the hepatobiliary radiopharmaceutical @Tc-N-(3-bromo-2,4,6-trimethyacetanilide) iminodiacetic acid (mebrofenin;Choletec, Squibb Diagnostic, Princeton, NJ) by the liver may beused as an index of hepatocellular function. The hepatic parenchymel cells extract mebrofenin from the blood using the same activetransport mechanism as bilirubin. Methods In this study, weinduced hepatocellular disease by administering a hepatotoxic drugand compared the hepatic extraction efficiency (HEE@,measureddirectly from an afferent injectkn of @rc-mebrofenin,to quantitath/e histopathology and to serum biochemistry analysis. ResultsThe baseline HEE was 95.9% ±2.71% (mean ±s.d.). Dogs thatwere affected by the hepatotoxic drug had reduced HEE. HEEcorrelated well to the severity of histologic lesions (r = —0.83,p =0.003). HEE also correlated well to the increases in the actMties ofalanine arninotransferase (ALT;r = —0.85,p = 0.002) and aspartateaminotransferase (AST; r = —0.89,p =

.@1oo

@ 020L.111111110040Ti me I60

80n Seconds 40 60 100

Ti me i n Seconds

SeverityNo.CriteñaMild1096-10%

of hepatic parenchymalcellsdemonstrated@alonModerate21

1%—25%of hepatic parenchymalcellsdemonstratedmelonMarked326%-75%

of hepatic parenchymalcellsdemonstratedlesionSevere476%-100%

of hepatic parenchymalcellsdemonstrated lesion

U)C

4-'

C

0U

— Liver

..-. Heart

FIGURE1.Graphshowingnetcountdenst@esfromliverandheartROlsforfirst90 sec after injectionof @Tc-mebrofeninin mesentenc veinof normaldog. Note that the majorityof radioactivityremained in liver,with littleradioactivityseen withinheart.

using the manufacturer's recommendations. The maximum recommended quantity of pertechnetate (100 mCi in 4 ml sterile eluate)was used to reconstitute the kits to keep the actual milligramquantity of mebrofenin that was injected as low as possible.Radiochemical purity was assessed using two ascending chromatographic systems: a silica gel-impregnated instant thin-layer chromatography paper developed with distilled water and a silicaacid-type instant thin-layer chromatography paper developed with20% NaC1.

The dogs were premedicated with acepromazine (PromAce, FortDodge Laboratories, Fort Dodge, IA) at a dose of 0.05 mg/kgintramuscularly and butorphanol tartrate (Torbugesic, Fort DodgeLaboratories, Fort Dodge, IA) at a dose of 0.5 mg/kg intramuscularly. General anesthesia was achieved with isoflurane (IsoFlo,Solvay Animal Health, Inc., Mendota Heights, MN) through maskinduction and was maintained after tracheal intubation. A ventralmidline laparotomy was performed and a segment ofjejunum wasisolated where a 19-gauge catheter was placed within a mesentericvein. The end of the catheter was exteriorized and the segment ofbowel was replaced into the abdomen. The incision was closed ina routine manner.

The dogs were placed in a right lateral recumbent view over alarge-field-of-view gamma camera (37 Tube ZLC gamma camera,Searle Nuclear, Des Plaines, IL) fitted with a low-energy, generalpurpose, parallel-hole collimator. A dynamic frame-mode acquisition using a 64 x 64 X 16 matrix size was controlled by animaging computer (Gamma 600 Imaging Computer, StrichmanMedical, Medfield, MA) and was initiated simultaneously withinjection of the radiopharmaceutical into the mesenteric catheter.The frame rate was 1 frame per second for 2 mm, then 1 frame perminute for an additional 13 mm. After the end of the scintigraphicprocedure, the mesenteric catheter was removed and the animalrecovered from anesthesia.

The HEE of the 99mTcmebrofenin was determined from thedynamic acquisition. Regions of interest (ROIs) were drawn overthe liver and heart. The counts from the first few seconds (beforebolus arrival) were subtracted from the remaining frames toremove background activity. With the background corrected,counts per pixel from the liver and heart ROIs were plotted tocreate time-activity curves (Figs. 1 and 2). The heart time-activitydata were used to represent blood-pool activity, and the livertime-activity data were used to represent hepatocyte activity plusthe portion of the blood pool in the liver vascular space. Atwo-compartment model was used to determine HEE, as previouslydescribed (8).

At least 2 days after the baseline hepatobiliary scintigraphy, ahepatotoxic drug was given to the dogs to induce liver disease.

FiGURE2.GraphshowingnetcountdensitiesfromliverandheartROlsforfirst90 sec afterinjectionof @c-mebrofenininmesentenc veinofdog withacute livertoxicity.Note that a large portionof radioactivityis in heart.

Thiacetarsamide (Caparsolate, Sanofi Animal Health, OverlandPark, KS) was injected intravenously at a dose of 3.3 mg/kg every12 hr for 2 days in 2 dogs and at a dose of 6.6 mg/kg every 12 hrfor 1 day in the remaining dogs to induce liver disease. Twelvehours after the last thiacetarsamide administration, a catheter wasplaced into a mesenteric vein as before and the hepatobiliary scanand analysis were repeated as described previously. Serum biochemistry analysis was repeated just before the second hepatobiliaryscan.

After completion ofthe second hepatobiliary scan, the dogs wereeuthanized by an overdose of a barbiturate. Liver specimens fromthe right lateral, left lateral and quadrate lobes were placed in 10%buffered formalin and processed for light microscopy. Sectionswere cut at 6-sm thickness and stained with hematoxylin-eosinstain. Each liver section was evaluated for the presence of thefollowing lesions: hepatocellular degeneration, hepatocellular necrosis, inflammation and hemorrhage. Each lesion was given ascore based on the percentage of cells involved using the criteria inTable 1. A total score for each dog was determined by adding thescores for each type of liver lesion.

Statistical AnalysisPearson product-moment correlation was performed to deter

mine the linear relationship of hepatic extraction of @mTc@mebro@fenin to the numerical score from the quantitative histopathology ofthe liver sections. Pearson product-moment correlations wereperformed to compare hepatic extraction of 9@―Tc-mebrofeninandthe numerical score from the quantitative histopathology of theliver to various serum biochemistry analysis values. All tests had alevel of significance of p < 0.05.

RESULTSAll dogs were considered normal based on serum biochem

istry analysis, fasting bile acid concentrations and portal scmtigraphy at the beginning of the study. The dogs received

TABLE ICriteria Used to Grade Severity of Histopathoiogical Lesions

HEPATIC Ex@n@i@ciIoN EFFICIENCY OF TEc@m@ETIur@i-99M-MEBRoFENm@ •Daniel et al. 1287

4 8

20 3228FiGURE3. Dynamicimages(bottom)Obtainedafterinjectionof @rc-mebrofenininmesentencveininnormaldog.Timeinsecondsatwhicheachimagewas acquiredis listedbelowimage.Notethat majorityofactMtyremainsinlrver,indicatinghighhepaticextractionefficiency.SChematiCdrawing(top)showsorientationof images relativeto dog, withheart and liverlabeled.

injections with 8.89 ± 1.58 mCi (mean ±s.d.) of 9@°@Tclabeled to 4.25 ±1.66 mg (mean ±s.d.) of mebrofenin. Theradiochemical purity ofthe radiopharmaceutical was 93.52% ±2.82% (mean ±s.d.) of 9@Tc boundto mebrofenin.Thebaseline HEE was 95.9% ±2.71% (mean ±s.d.).

The dogs varied in their response to the thiacetarsamide.Three dogs showed no clinical illness (Group C, Table 2) andno abnormalities in their serum biochemistry analysis other thana mild increase in the activity of alanine aminotran@ferase(ALT) in one dog. In these dogs, the liver retained the vastmajority of the radiopharmaceutical during the first passage,with little activity observed in the cardiac blood pool (Fig. 3). Inthese dogs, the hepatic extraction of 99mTc@mebrofenin wasdetermined to be normal (>95%) and the histopathologicevaluation oftheir livers were essentially normal, having a totalhistologic score of 1.

Four dogs became severely clinically depressed with vomiting and diarrhea (Group A, Table 2). These dogs had markedincreases in the activities of ALT, aspartate aminotransferase(AST), alkaline phosphatase (Alk Phos), total bilirubin andfasting bile acid concentrations. These dogs had decreased liveruptake of 99mTc..mebrofemn with a large percentage of theradiopharmaceutical passing through the liver, resulting inprominent cardiac blood-pool activity (Fig. 4). The hepaticextraction of 9@°'Tc-mebrofeninwas markedly decreased(56.4% ± 18.0%, mean ± s.d.). Hepatocellular necrosis,hemorrhage and inflammation were the most notable histopathologic abnormalities. The remaining three dogs werenormal to mildly depressed (Group B, Table 2) and had serumbiochemistry analysis values that were intermediate between

the previously described groups. Their values of hepatic extraction of 99mlcmebrofenin were normal to slightly decreasedand their histopathologic scores were intermediate between thepreviously described groups.

When comparing all 10 dogs, there was a significant conelation of the HEF to quantitative histopatholo@y (r = —0.833,p = 0.003). As the hepatic extraction of 9@Tc-mebrofenindecreased, the histopathologic score of the liver lesions increased in a linear fashion (Fig. 5). There was good correlationbetween the hepatic extraction of @“Tc-mebrofeninand theserum biochemistry analysis (Figs. 6—9).The histopathologicscores also showed good correlation with the serum chemistries,but the correlation coefficient was not as high as it was for thehepatic extraction of 99mTc@mebrofenin (Table 3).

DISCUSSIONThe hepatic extraction of the iminodiacetic acid derivative

radiopharmaceuticals provides a useful marker of hepatocellular function. These radiopharmaceuticals can be used to evaluate two aspects of hepatocyte function: the dye-anion receptorfunction and excretion into the biliary tree. These agents enterthe hepatocyte by a carrier-mediated, nonsodium-dependent,organic anion pathway using a mechanism very similar to thatof serum bilirubin (3,15,22). The efficiency of hepatic parenchymal cells to extract the mebrofenin from the plasma decreases with hepatocellular injury (3). In this study, we foundthat the severity of the liver lesions, based on histopathologicabnormalities, was correlated with the decrease in hepaticextraction of 99mTcmebrofen@

Technetium-99m-mebrofenin also can be used to evaluate

1288 THEJouRH@OFNUCLEARMEDICINE•Vol. 39 •No. 7 •July 1998

HEEbaselineHEEafter thiacetarsemide Hlatopathologk

Group* (%) administration(%) score AlkPhos ALTASTBileacidsBilirubinA

96 145 98.0 35.7 11 44820,55013,290215.004.20A96 146 96.0 78.7 9 12513,0958,74510.600.60A96 024 97.0 51.0 9 1.19738,40020,200183.805.70A95 522 96.9 60.1 5 38915,9409,760168.004.00Mean

— 56.4 8.5 54021,99612,999144.353.63s.d.— 18.0 2.5 46011,3595,18191.282.15B

96 046 94.9 90.0 5 2404,9602,4405.700.20B93 022 90.1 84.9 5 896733—0.20B95 264 95.4 96.1 4 92 1.3951162.000.10Mean

— 90.3 4.7 1402,1418633.860.17s.d.— 5.61 0.6 86 2,5301,366—0.06C

95 051 99.7 96.4 1 6246152.500.10C95 259 95.6 96.5 1 531911030.00.0C96 021 96.9 94.9 1 6346152.500.10Mean

— 95.90 1 59 94441.670.07s.d.— 0.9 — 5 84511.440.06*Group

A animalsshowed severe clinicalsigns, Group B animalsshowed minimalclinicalsigns and Group Canimals showed noclinicalsigns.1@JkPhos = alkaline phosphatase; ALT = alanine aminotransferase; AST = aspartate aminotransferase.

TABLE 2Summary Data Following Thiacetarsamide Administration and Baseline Hepatic Extraction Efficiency (HEE)

excretory pathways into the biliary tree. Technetium-99m-mebrofenin does not undergo biotransformation during thetransit through the hepatocyte to the bile caniliculus (23,24).

The final excretory step involves transportation across thecanalicular membrane. The ability of the hepatocyte to excrete99mTcmebrofenin is decreased in cases of hepatocellular dys

20 24 28 32

FIGURE4. Dynamicimages(bottom)obtainedafterinjectionof @Tc-mthrofenininmesentencveinofdog withacutelhertodcity. lime inseconds at whicheach image was acquired is listed below image. Note that major portion of radioactivitypassed through liver, indk@atinglow hepatic extraction of

@“Tc-mebrofenin.Schematic drawing(top)shows orientationof images reIatR@eto dog, withheart and liverlabeled.

HEPATIC Exm@CTIoN EFFICIENCY OF TECHNETIUM-99M-MEBROFENIN •Daniel et al. 1289

30 40 50 60 70 80 90 100

Hepatic Extraction Efficiency

I 30 40 50 60 70 80 90Hepatic Extraction Efficiency

FastingHistopathologicscore ,@JkPhos ALT AST bileacid

,oJk Phos = alkaline phosphatase; ALT = alanine aminotransferase;AST= aspartate aminotransferase.

R = -0.833p = 0.00310

8

6

4

2

30 40 50 60 70 80 90 100

Hepatic Extraction Efficiency

cJ 200

:o 150

a@o:@@ 100@ a..@ 50

0

30 40 50 60 70 80Hepatic Extraction Efficiency

R = -0.967p < 0.001

.@

I

0

a

S S S

S

go 100

FiGURE8@Pk@tshowing relationshipbetween fasting bileacki concentrations and HEE.

tion and hemorrhage occur simultaneously with elevations ofliver-derived enzymes in the plasma, namely ALT, AST andAlk Phos. Elevation of ALT and AST indicate a loss ofparenchymal cell integrity but are innocuous and nonspecific indetermining the extent of hepatic involvement (5,27). In manycases, elevation of these liver leakage enzymes may not beassociated with clinically significant liver disease (26). Inaddition, the activity of AST and Alk Phos may be increaseddue to nonliver sources (26). The severity of liver diseasecannot be determined from these liver enzymes alone.

In our model of acute toxic liver disease, the liver enzymeelevations were profound in the affected dogs. Dogs with themost severe liver enzyme elevations had the greatest reductionin HEE and the highest histopathologic score. The liver enzymes were mild to moderately elevated in the dogs withnormal to slight decreases in HEE, suggesting that HEE haslesser sensitivity in mild acute liver injury (Table 1). Toxicinjuries to the liver can cause substantial increases in cellularpermeability. Our data suggest that mild to moderate increases

FiGURE9. Plot shovung relationshipbetween alkalinephosphatase andHEE

TABLE 3Resutts of the Pearson Product-Moment Correlation Comparing

Histopathologic Score with Serum Biochemistnes

FIGURE5. Plot showing relationshipbetween histopathologicscore fromliverbiopsyand HEE.

function but it is also decreased in cases of extrahepaticcholestasis (3,5, 7). Hepatic excretion was not measured in thisstudy because the experimental design stopped data acquisitionat 15 mm. Because the normal hepatic excretion T112in the dogis 19 mm, there was insufficient data for proper curve fitting ofthe excretory phase of the time-activity curve (8).

Determining the severity of liver injury is difficult, oftenrequiring the consideration of physical findings, serum biochemistry analysis and comparison of morphologic informationderived from ultrasound, CT or MRI with the functionalinformation from scintigraphy (11,25). A variety of liverdiseases have been described in which the functional andmorphologic lesions often coincide (26). Histologic lesions ofhepatocellular degeneration, hepatocellular necrosis, inflamma

20

R=-0.894

p< 0.00115

1200

@ 1000

@ R= -0.731

@ P•@@$@

30 40 50 60 70 80 90 100

Hepatic Extraction Efficiency

FiGURE6. Plotsho@ngrelationshipbetweenASTandHEE

R = -0.851p= 0.002

Correlationcoefficient—0.602—0.770—0.814—0.694pvalue0.0160.0090.0040.038

FIGURE7. Plot showingrelationshipbetween ALTand HEE.

1290 THEJOURNALOFNUCLEARMEDICINE•Vol. 39 •No. 7 •July1998

for the transport of 99mTcmebrofenin into the hepatocyte. Thedecreased hepatic extraction observed after thiacetarsamideadministration may be due to the alteration or dysfunction of thebinding proteins, such as the enzyme glutathione transferase,which are important to the function of the dye-anion receptor(22,31 ). Dogs with more severe histologic liver abnormalitieshad a more profound decrease in liver uptake of 99mTc@mebrofenin. In these dogs, a large percentage 99mTc@mebrofe@nm passed through the liver and showed up in the cardiac bloodpool. Dogs with normal liver histopathology and normal clinical pathology values had retention of the majority of 99mTc@mebrofenin by the liver similar to previous studies in normaldogs (8). The quantitative value, HEE, was significantly correlated to the severity of the histologic lesion. As the HEEdecreased, the histopathologic score of the liver increased. Theability of the liver to extract 9@Tc-mebrofenin appears to be agood predictor of the severity of hepatic parenchymal celldamage.

CONCLUSIONThe measurement of the hepatic extraction of 9@Tc-mebro

fenin is a useful index of hepatocellular function. The decreasein hepatic extraction of 99mTc..mebrofenin correlates well withthe severity of histologic changes in the liver and with thepermeability of the hepatocellular membrane in this acute toxicliver injury model. Further work is needed to compare thehepatic extraction of @‘Tc-mebrofeninto the severity ofacquired liver disease.

REFERENCES1. Lieberman DA, Brown PH, Krishnamurthy GT. Improved scintigraphic assessment of

severe cholestasis with the hepatic extraction fraction. Dig Dis Sci 1990;35:1385—1390.

2. Heyman S. Hepatic scintigraphy as a liver function test. JNuc/Med 1994;35:436—437.3. Krishnamurthy GT, Turner FE. Pharmacokinetics and clinical application of techne

tium-99m-labeled hepatobiliary agents. Semin Nuc/ Med I990;20:130—149.4. Doo E, KrishnamuiThy GT, Ekiem Mi, Gilbert 5, Brown PH. Quantification of

hepatobiliary function as an integral part of imaging with technetium-99m-mebrofeninin health and disease. J Nuci Med 1991;32:48—57.

5. Juni JE, Reichle R. Measurement of hepatocellular function with deconvolutionalanalysis: application in the differential diagnosis of acute jaundice. Radiology1990;177:171—175.

6. Brown PH, Juni JE, Lieberman DA, Krishnamurthy GT. Hepatocyte versus biliarydisease: a distinction by deconvolution analysis of technetium-99m IDA time-activitycurves. J NucI Med l988;29:623—630.

7. Lieberman DA, Krishnamurthy GT. Intrahepatic versus extrahepatic choleostasis.Gastroent 1986;90:734—743.

8. Daniel GB, Bah.r A, Dykes JA, DeNovo R, Young K, Smith GT. Hepatic extractionefficiency and excretion rate of technetium-99m-mebrofenin in dogs. J Nucl Med1996;37:1846—1849.

9. Huang SC, Phelps ME. Principles of tracer kinetic modeling in PET and autoradiography. In: Phelps ME, Mazziotta J, Schelbert HR. eds. Positron emission tomograpkvand autoradiography. New York: Raven Press; 1986:287—346.

10. Patlak CS, Blasberg RG, Fenstermacker JD. Graphical evaluation of blood-to-braintransfer constants from multiple-time uptake data. J Cereb Blood Flow Metabl983;3: 1—7.

11. Thie JA, Smith GT, Hubner KF. Total graphical analysis of dynamic PET data withidentification of up to four parameters [Abstract]. J Nuci Med 1993;34: l85P.

12. Bahr A, Daniel GB, DeNovo R, Young K, Menyman JL. Quantitative hepatobiliaryscintigraphy with deconvolutional analysis for the measurement of hepatic function indogs. Vet Radiol Ultrasound 1996;37:2 14—220.

13. Tagge EP, Campbell DA, Reichie R. Quantitative scintigraphy with deconvolutionalanalysis for the dynamic measurement of hepatic function. J Surg Res 1987;42:605—612.

14. Howman-Giles R, Moase A, Gaskin K, Uren R. Hepatobiliary scintigraphy in apediatric population: determination of hepatic extraction fraction by deconvolutionanalysis. J NucI Med 1993;34:2l4—22l.

15. Krishnamurthy 5, Krishnamurthy GT. Technetium-99m-iminodiacetic acid organicanions: review of biokinetics and clinical application in hepatology. Hepatologyl989;9: 139—153.

16. Gad MA, Krishnamurthy GT, Glowniak JV. Indentification and differentiation ofcongenital gallbladder abnormality by quantitative technetium-99m IDA cholescintigraphy.J NucIMed I992;33:431—434.

17. Ben-Haim 5, Seabold JE, Kao SC, Johnson J, Iran D, Brown BP. Utility of9@'Tc-mebrofenin scintigraphy in the assessment of infantile jaundice. C/in NucI Med1995;20:153—163.

in cellular permeability have only a minimal effect on thedye-anion receptor function. Further work is needed to determine if this disparity is a function of time. In addition, studiesare needed to evaluate the sensitivity and specificity of HEE inspontaneously acquired liver diseases. Severe hepatic dysfunction associated with end-stage cirrhosis can exist in the absenceof serum enzyme abnormalities (26). In this study, we foundthat HEE had a more linear relationship to serum biochemistryanalysis than did the quantitative histopathologic score. Quantitative hepatobiliary scintigraphy may be better able to gradethe severity of liver damage in cases of end-stage cirrhosis thanserum biochemistry analysis.

The more severely affected dogs had increased total bilirubinlevels. Elevated bilirubin is seen with hepatic parenchymaldisease but it also can be elevated due to pre- and posthepaticcauses, thereby decreasing its usefulness as an indicator of liverdisease. Bilirubin elevations can lag behind rapid changes inliver function or they may plateau in cases of severe disease. Inaddition, bilirubin levels are influenced by biliary tree patency,erythrocyte degradation and renal excretion of bilirubin. In ourstudy, we found a linear relationship between the serumbilirubin concentration and the hepatic extraction of 99mTc@mebrofenin. As the serum bilirubin concentrations rise, thehepatic uptake of iminodiacetic acid derivative radiopharmaceuticals is expected to decrease. Technetium-99m-mebrofeninis commonly used for quantitative hepatobiliary scintigraphybecause it is avidly taken up by the hepatic parenchymal cellseven in the face of profound hyperbilirubinemia (3). Technetium-99m-mebrofenin has greater resistance to displacement byserum bilirubin than other iminodiacetic acid derivatives (3,22—24,28,29). Studies in humans have shown that patients can havenormal hepatic extraction of 99mTc..mebrofenin with serumbilirubin concentrations between 4.9 and 8.2 mg/dl (4). Thehighest recorded bilirubin concentration in our dogs was 5.7mg/dl. Although we cannot rule out the possibility of competitive inhibition for the receptor as the cause of the decreasedhepatic extraction of 9@Tc-mebrofenin, it is more likely thatthe elevated bilirubin concentration and the decreased hepaticextraction of 99mTcmebrofenifl are both directly related to thedecrease in liver function.

Hepatocellular injury was created by the administration of ahepatotoxic agent. To minimize animal suffering, an agent anddose were chosen that would induce acute toxic injury to theliver. Thiacetarsamide is an arsenical agent that was formerlyused as an anthelmintic drug for treating canine dirofilariasis(heartworm disease). This drug has been largely replacedbecause of the high incidence of liver and kidney toxicity thatdeveloped within 24—48hr following its administration. Previous work indicated that the liver leakage enzymes peakedwithin 24—48hr of initiation of thiacetarsamide administrationat a dose of 3.3 mg/kg (1.5 X recommended dose) and within24—36hr at a doseof 6.6 mg/kg(3.0 X recommendeddose)(30). Initially, a dose of 3.3 mg/kg was chosen but was laterincreased to 6.6 mg/kg so that a greater percentage of dogswould manifest hepatic toxicity.

Thiacetarsamide is a Type II liver toxin, meaning that thelesions are nonpredictable and are dose and time independent(31 ). As expected, this agent produced varying degrees of liverinjury from no abnormalities to severe hepatocellular necrosis.The dogs also varied in their clinical response to the thiacetarsamide, ranging from no abnormalities to clinical illness withmarked increases in the activities of ALT, AST and bile acidconcentration. Arsenicals produce zonal to massive hepatocellular necrosis (31 ). This cytotoxic injury can result in adecreased function of the carrier-mediated mechanism needed

HEPATIC EXTRACTION EFFICIENCY OF TECHNETIUM-99M-MEBROFENIN •Daniel et al. 1291

18. Brunot B, Petras 5, Germain P. Vinee P. Constantinesco A. Biopsy and quantitativehepatobiliary scintigraphy in evaluation of liver transplantation. J NucI Med 1994;35:1321—1327.

19. Daniel GB, Bright RM, Monnet E, Ollis P. Comparison of per-rectal portal scintigraphy using @“technetiumpertechnetate to mesenteric injection of radioactive microspheres for quantification of portosystemic shunts in an experimental dog model. VetRadiol Ultrasound I990;3 1:I75—181.

20. Susumu 5, Kuroki T, Kurai 0, et al. Portal circulation by technetium-99m pertechnetate per-rectal portal scintigraphy. J NucI Med 1988;29:460—465.

21. Daniel GB, Bright RM, Ollis P. Shull R. Per rectal portal scintigraphy using @“Tcpertechnetate to diagnosis portosystemic shunts in dogs and cats. J Vet Intern Med1991;5:23—27.

22. Fritzberg AR. Advances in the development ofhepatobiliary radiopharmaceuticals. In:Fritzberg AR, ed. Radiopharmaceuticals: progress and clinical perspectives. BocaRaton, FL: CRC Press; 1986:90—I16.

23. Loberg MD, Cooper M, Harvey E. Development ofnew radiopharmaceutical based onN-substitution of iminodiacetic acid. J Nuci Med l976;17:633—638.

24. Loberg MD, Fields AT. Chemical structures of@―Tc-labeled N-(2,6-dimethylphenylcarbamoylmethyl) iminodiacetic acid (Tc-HIDA). mt J AppI Radiat iso 1978;29: I67—173.

25. O'Connor KW, Snodgrass PJ, Swonder JE, et al. A blinded prospective studycomparing four current noninvasive approaches in the differential diagnosis of medicalversus surgical jaundice. Gastroenterology I983;84: 1498—1504.

26. Center SA. Pathophysiology. laboratory diagnosis and diseases ofthe liver. In: EttingerSi, Feldman EC, eds. Textbook ofveterinary internal medicine, disease ofihe dog andcat. Philadelphia: WB Saunders; l995;1261—l3l2.

27. Feldman BF. Clinical pathology of the liver. In: Kirk RW, ed. Current veterinarytherapy VI!!. Philadelphia: WB Saunders; 1980:875—885.

28. Klingensmith wC, Fritzberg AR, Spitzer VM, Kuni CC, Williamson MR. Gerhold JP.Work in progress: clinical evaluation of c9mTc1@.@ethylbromo.IDAand 9'@―Tcdiisopropyl-IDA for hepatobiliary imaging. Radiology 1997; 146:181—184.

29. Nunn AD, Loberg MD, Conley RA. A structure-distribution relationship approachleading to the development of @‘@‘Tc-mebrofenin:an improved cholescintigraphicagent. J NucI Med l983;24:423—430.

30. Winograd H, Lorgue G, Keister DM. Thiacetarsamide toxicity in the normal dog. In:Soil MD, Coleman MW, Courtney CH, Jackson RF, Keough Ri, Knight DH, eds.Proceeding of the heartworm sYmposium 92. Batavia, IL: American HeartwormSociety; 1992:215—222.

3 1. Zimmerman HJ. Hepatotoxicity. New York: Appleton-Century-Crofts; 1978:47—90.

phaticopelvic fistulization, skipping of lymphatic chain, patency of thoracic duct and abnormal leg lymphatics (1—3).However, it requires tedious cannulation of lymphatics, isinvasive and is not readily reproducible. It also can result inlocal tissue necrosis, fat embolism to the lungs, hypersensitivityreaction and exacerbation of lymphedema by the contrastmaterial (4). Lymphoscintigraphy using 99mTc@sulfurmicrocolbid, antimony sulfide colloid, stannous phytate, rhenium sulfurcolloid, human serum albumin or dextran delineates the patternof lymphatic drainage, is fast and nontraumatic and has noknown side effects (5—6). Use of lymphoscintigraphy in investigating chyluna, chyloperitoneum and chylothorax has beenlimited to a few case reports (7—13). We retrospectivelyanalyzed lymphoscintigraphic studies to determine their efficacy in the investigation of chyluria, chyloperitoneum andchylothorax.

MATERIALS AND METhODSBetween June 1989 and May 1996, lymphoscintigraphy was

performed on 18 patients with chyluria (9 women, 2 men), chyluriaand chyloperitoneum (2 men), chyloperitoneum and chylothorax (3women, 2 men). Follow-up scans were obtained on 3 patients,giving a total of 21 lymphoscintigraphic studies. The averagepatient age was 38.0 yr (range 25—57yr), and average duration ofsymptoms was 4.7 yr (range 3 mo to 36 yr) for chyluna and 6.0 mo(range 10 days to 18 mo) for chyloperitoneum and chylothorax.Elephantiasis of the lower extremities was the most commonassociated symptom. Other associated symptoms and signs included hematuria, flank pain, dysuria, urinary frequency, oliguria,weight loss, fever, cough, anasarca, chest discomfort, dyspnea andenlarged inguinal lymph nodes. A previous history of filariasis wasknown in 7 of 13 patients with chyluria. Urine, peritoneal, pleuralor pericardial fluid were positive for fat droplets after Sudanstaining and negative for acid-fast bacilli, culture and cytology inall patients. Repeated thick-film night blood smears were alsonegative for microfilaria in all patients. Final diagnoses were made

Lymphoscintigraphy allows functional assessment of lymphatictransport and depiction of regional @,mphnodes, is fast and nontraumatic and has no known side effects. We retrospectively analyzed lymphoscintigraphic studies to determine their efficacy in theinvestigation of chyluria, chyloperitoneum and chylothorax. Methods: Twenty-one whole-body lymphoscintigrams using @rc-antimony sutilde colloki or dextran were acquired in 18 patients withchyluria, chyloperitoneum and/or chylothorax. The images werereviewed to assess the rate of tracer transport and number, size anddistributionof lymph vessels and nodes as well as the presence ofcollateral, fistula or lymph reflux. Results Lymphoscintigraphy wasnormal (5 of 11 patients) or showed lymphatic obstruction (6 of 11patients) in chyluria associated with filariasis. Lymphatic obstructionwas demonstrated in ch@rioperftoneumand/or chylothorax associated with liver cirrhosis (2 patients), postoperative (1 patient) orcongenital (1 patient) lymphatic dysplasia, inferior vena cava obstruction (1 patient) and nephrotic syndrome (1 patient). Enhancedlymph flow was seen in systemic lupus erythematosus (1 patient).Follow-up lymphoscintigrams showed patency of lymphovenousanastomosis (1 patient), improvement (1 patient) or no change (1patient) in lymphatic drainage aftertreatment. Conclusion: Lymphoscintigraphy can demonstrate abnormal lymphatic drainage in chyluria, chyloperitoneum and chybthorax. It is useful for selectingpatients for surgery and assessing the effect of treatment.Key Words chyluria; chyloperitoneum; chylothorax; lymphoscintigraphy

J Nucl Med I998 39.1292-1296

L ymphangiographyhasbeenthemainimagingmodalityininvestigating chyluria, chyloperitoneum and chylothorax. It isuseful for detecting abnormal retroperitoneal lymph nodes,leakage from dilated lymphatics, lymphoperitoneal and lym

ReceivedJun. 23, 1997;accepted Sep. 3, 1997.For correspondence or reprints contact: Margaret H. Pul, MD, Department of

RadlObgy, Ago Khan Un@ersftyMedical center, Stadium Rd., P0 Box 3500, Karachi74800, Pakistan.

1292 THEJOURNALOFNUCLEARMEDICINE•Vol. 39 . No. 7 •July 1998

Lympho scintigraphy in Chyluria, Chyloperitoneumand ChylothoraxMargaret H. Pui and Tian-Chao YuehDepartment ofNuclear Medicine, First Affiliated Hospital ofSun Yat-Sen University ofMedical Sciences, Guangzhou,People ‘sRepublic of China