Positron-emitting 38K, which has a half-life of 7.6 mm, isa suitable tracer for investigating potassium kinetics in theseorgans. Although several experimental and clinical studieshave reported on cardiac PET imaging (8—15),only a fewreports have been published concerning 38K kinetics in theheart (13,14), none regarding the kidney, and 1 preliminaryreport on the brain (16). We recently developed a novel wayof producing 38K (17) and a high-resolution cardiac PETsystem for imaging and quantitation of myocardial bloodflow in rabbits (18). The purpose of this study was toevaluate the kinetics and image quality of 38K PET inthe heart, kidney, and brain, based on our experimentalprocedures.

MATERIALS AND METhODS

Tracer Preparation38K was obtained as a saline solution of 38K@after the

“°Ar(p,3n)38Knuclear reaction. The target chamber (150 mm thick,96 mL, 20 bar) was filled with pure argon gas and irradiated with 40MeV of protons at a beam intensity of 6—6.5@iAfor 15 mm.Immediately after the bombardment, irradiated gas was releasedthrough a 5-mL water reservoir to trap 38Kin the gas phase. Thewater was then quickly introduced into the chamber, allowed toreflux for 1 mm to recover 38Kabsorbed on the inner surface of thetarget chamber, and collected in a sterilized vial containing 45 mgNaCl through an anion exchange column and a 0.22 pm Millexfilter. 38Khas a radionuclidic purity > 99.9%, which was determined by pure germanium detector. The chemical form of 38Kwasdetermined by using radio-ion chromatography as K@.All procedures were previously described by Nagatsu et al. (17).

Animal PreparationEighteen male Japanese white rabbits weighing 2.9—3.7kg were

studied. The animals were anesthetized with sodium pentobarbital(35 mg/kg) injected into the right marginal ear vein. The anesthetized state was maintained by a constant intravenous infusion ofsodium pentobarbital at 6 mg/kg/h, beginning 1 h after induction.The rabbits were ventilated by tracheotomy using a small-animalventilator (SN-480--5; Shinano, Tokyo, Japan). The settings of thisventilatorandtheoxygencontentof inspiredair wereadjustedtomaintain the blood gases in a physiologic range throughout theexperimental period. Arterial blood samples were analyzed frequently for arterial Pao,, PaCo@,and pH, with a Ciba-ComingpWblood gas analyser (Model 238; Chiron Diagnostics, Emeryville, CA). The rabbits were paralyzed with pancuronium

The pUrpOSeofthis study was to evaluate the kinetics and imagequality of positron-emitting “°K(half-life, 7.6 mm) and highresolution small-animal PET in the heart, kidney, and brain ofrabbits. Methods: Studies were performed with 18 closed-chestanesthetized rabbits at baseline and dunng infusions of adenosine (0.2 mg/kg/mm) and propranolol (0.5—1.0 mg/kg intravenously) using high-resolution small-animal PET. @Kwas injectedintravenouslyand dynamic PET imagingof the heart, kidney,orbrain was performed for 3 mm. Colored microspheres wereInjected into the left ventricle to measure organ blood flow.Arterial blood was withdrawn directly from the femoral artery,and, after the animals were killed, @Kactivities in each organwere measureddirectly with a well counter. Uptake of @Kwascalculated by dMding the @Kactivities in each organ by theintegral of the input function. The extraction fraction of @Kwasestimated by dividing the uptake of @Kin each organ by theorgan blood flow, measuredby microspheres.Results: The leftventricularmyocardiumand kidneywereclearlyvisualized,butthere was no visual @Kuptake in the brain. For the heart, kidney,and brain, respectively, average blood flow was 2.91 ±1.29,5.49 ±0.71 , and 0.57 ±0.11 mL/min/g, and the extractionfractionof @Kat baseline was 0.55 ±0.13, 0.48 ±0.13, and0.022 ±0004. The Renkin-Crone model fit the relation betweenmyocardialextractionand flow undera wide rangeof myocardialblood flow (r = 0.89). Conclusion: @Kis a suitable tracer fornoninvasively showing the potassium kinetics of the heart,kidney,and brainby PETimaging.KeyWords: @K;PET;myocardlum;kidney;brain

J NucI Med2000;41:763—769

mong the many radionucides of potassium analogs,201T1and @Rbare widely used for cardiac SPECT and PETimaging to noninvasively assess myocardial perfusion andmembrane integrity (1,2). Renal uptake of these tracers mayalso be used in the assessment of renal perfusion (3—5).However, brain uptake of potassium analogs is largelyprevented by the blood—brainbarrier (BBB) (6). Yet thesetracers are useful for detecting breakdowns in the BBB, suchas those caused by brain tumors or radiation injury (7).

Received Mar. 1, 1999; revIsion acceptedAug. 24, 1999.Forcorrespondence or reprints contact Katsuya Yoshida, MD, ThIrd Depart

mentoflntemalMedicine,ChibaUniversitySchoolofMedicine,1-8-1Inohana,Chuo-ku, Chiba, 260-8677, Japan.

38K PET IN RABBITS•Takami et al. 763

Uptakes and Images of 38Kin Rabbit Heart,Kidney, and BrainAkira Takami, Katsuya Yoshida, Hiroyuki Tadokoro, Shinobu Kitsukawa, Kazuhiro Shimada, Mikio Sato,KazutoshiSuzuki,YoshiakiMasuda,and ShujiTanada

Third Department oflnternalMedicine, Chiba University School ofMedicine, Chiba; and Division ofAdvanced Technologyfor MedicalImaging, Nationallnstitute ofRadiological Sciences, Chiba, Japan

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

bromide (0.3 mg/kg intravenously), with additional injections of0.15 mg/kg intravenously every 30-40 mm. Rectal body temperature was maintained at 38°C—39°Cwith a heating pad. Catheterswere inserted into the right femoral artery for collection ofmicrosphere reference samples and for obtaining arterial inputfunction, into the left femoral artery for the measurement of bloodpressure,and into the rightjugular veinfor the infusionof normalsaline. For the injection of 38K,catheters were inserted into the leftmarginal ear vein for heart and kidney imaging. For brain imaging,we chose the femoral vein as the tracer injection site to avoidartifacts from the high tracer activities in the marginal ear vein tothe brain image. The right carotid artery was cannulated for theinjection of 15 pm colored polystyrene microspheres (Dye-Trak;Triton Technology, Inc., San Diego, CA) through a polyethylenecatheter (0.87 mm inside diameter, 1.27 mm outside diameter;Natume, Tokyo, Japan) advanced into the left ventricle. Theanimals were heparinized (400 U/kg) after the cannulations.Arterial pressures and heart rates were monitored with a multichannd polygraph (Omniace RT3200N; NEC San-ei, Tokyo, Japan).The rabbits were then placed in a supine position within the animalPET device. The experimental protocols for this study wereapproved by the Animal Welfare Committee of the NationalInstitute of Radiological Sciences, Chiba, Japan.

PET

PET images were obtained using a small-animal PET device(SHR-2000; Hamamatsu Photonics K.K., Hamamatsu, Japan) thatprovided 7 transaxial slices simultaneously. The slices had atransaxial resolution of 3.0 mm full width at half maximum(FWHM) and were separated by 6.5 mm. Axial resolution was 4.8and 4.1 mm FWHM in direct and cross planes, respectively (19).

Initially, blank data for the correction ofdetection efficiency andthe transmission data for tissue attenuation correction were collected using a @Ge/@8Gasource. Subsequently, 38Kwas injectedintravenously as a 20-s slow bolus into the marginal ear vein forheart and kidney imaging and into the femoral vein for brainimaging. The injection dose was 56—140,83—140,and 182—222MBq for heart, kidney, and brain, respectively. Simultaneously, aninitial set of 6 10-s data frames was acquired, followed by 3 20-sand 2 30-s data. PET imaging (13 heart, 3 kidney, 3 brain) wasperformed for 3 mm.

Because there was little visual uptake of 38Kin the brain, weperformed FDG PET studies on a rabbit after 38K brain PETimaging to clarify the position of the brain. FIX) (186 MBq)injected intravenously at 12mm after the 38Kinjection, and, 50mmlater, the data were collected over 10 mm.

Each image was displayed as 180 X 180 pixels with a pixel sizeof 1.0 x 1.0 mm. Count losses at the high counting rates werecorrected by single photon signals, because count losses on the PETscanner occur when single photon signals enter the photomultiplieror positron-encoding circuit. All reconstructions were performedwithout electrocardiographic gate and corrected for physical decayof the tracer.

Experimental ProtocolsEach rabbit was allowed to stabilize for 20—30rain after

completion of the procedures. Approximately 1.0 X l0@ colored

microspheres were dispersed with a mechanical mixer and themimmediately injected into the left ventricle. Blood withdrawalbegan 15 Sbefore the microsphere injection from the femoral arteryat a constant rate of 1.5 mL/min to collect microspheres and obtain

arterial input function with BACC-4 (Hamamatsu Photomcs K.K.).Immediately after the injection of microspheres, 38K was administered as a 20-s slow bolus while the imaging sequence began. PETimages were them acquired. Blood sampling was stopped 2 mmafter the tracer injection. Low-molecular-weight dextran (mol wt40,000; Otsuka Pharmaceutical Co. Ltd., Tokushima, Japan) wasinfused into the ear vein at a rate of 1.5 mLlmin, concurrent withthe withdrawal of the blood sample, to prevent any significantdecrease in blood pressure during withdrawal. Intravenous infu

sions of adenosine and propranolol were given to obtain a widerange of myocardial blood flow. Eight rabbits were studied withoutany pharmacologic intervention. In 5 rabbits, myocardial hyperemia was induced with intravenous adenosimeat a concentration of0.2 mg/kg/mn infused for 6 mm. At 3 mm after the start ofademosine infusion, microspheres and 38Kwere administered. Toevaluate low flow, several doses of propranolol (0.5- to 1-mg/kgbolus intravenously) were injected into 5 rabbits. In total, 8 rest, 5ademosine, and 5 propranolol experiments were performed in 18rabbits. At 3 mm after 38Kinjection, the animals were killed duringdeep anesthesia with KC1solution, and organs (heart, kidneys, andbrain) were removed. The left ventricular myocardium, the rightand left renal cortices, and the right and left cerebrums weredissected. The organs were weighed, counted in a well counter(Mimaxi @j;Packard Instrument, Downers Grove, IL), and correctedfor radioactive decay.

The extraction of microspheres from blood and tissue sampleswas performed as described by Hale et al. (20). Regional bloodflow was calculated from the equation:

regional blood flow = Cm X Qr/Cr, Eq.l

where Cm is the total number of microspheres/g of myocardialtissue, Qr is the withdrawal rate of the reference blood sample(mL/min), and Cr is the total number of microspheres in thereference blood sample.

Arterial Input FunctionThe high-efficiency bismuth germanate (BOO) coincidence

detection system (BACC-4) was designed and built for applicationsin arterial blood sampling from small animals undergoing PET(18). The detection system uses 4 BOO detectors in a configurationto provide a small central opening with very high sensitivity forinsertion of small-volume, flow-through blood-sampling tubing.Blood was withdrawn from the right femoral artery through thetubing at a constant flow (1.5 mL/min) with a Harvard pump(Model 11; Harvard Apparatus, Millis, MA), which allowed thesimultaneous collection of colored microspheres. To calibrate thissystem to the well counter, serial blood samples were obtainedfrom the ascending aortic catheter immediately after 38Kintravenous injection in 5 rabbits under control conditions. A clippedcatheter was placed within the ascending aorta through the rightcarotid artery. The clip was released to collect 2 or 3 drops of blood.From each blood sample, 250 pL were themtransferred to a test tubeby micropipette. These samples then were measured with the wellcounter. Sampling intervals were 5 s during the first minute, every10 s for the next minute, and every 20 s for the last minute. Arterialblood was also withdrawn from the femoral artery by Harvardpump and measured with the BACC-4. The dead volume from thefemoral artery to BACC-4 was approximately 0.9 mL. The integralvalues of blood sampling data were compared with those determined with the BACC-4.

764 THE Joui@i. OF NUCLEARMEDICINE •Vol. 41 No. 4 . April 2000

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

SyStolIC BPDiaStoliC BPHeart rateRPP(mmHgxIntervention(mmHg)(mm Hg)(bpm)bpm)

@KOrganUptakes38Korgan uptakes (extraction X flow) were calculated by the

following equation:

Cm(t)EXF=@ , Eq.2

10 Ca(x)dx X g

where E is the extraction fraction 38 @;F is the organ blood flow(mL/min/g); Cm(t) and Ca(t) are the 38Kactivity concentrations inthe organs and arterial blood, respectively, at time t; and g is thespecific gravity of the organs (heart, 1.056 g/mL; kidney, 1.062

g/mL; and brain, 1.043 g/mL) (21). In this study, 38Korgan uptakeswere calculated during the first 3 mm of data acquisition (t = 3mm). Tracer concentrations in the organs were measured by wellcounter after the animals were killed. Arterial tracer concentrationswere also measured with the BACC-4.

Statistical AnalysisHemodynamic data were expressed as mean ±SD. All data

were analyzed by ANOVA. If a statistical significance wasobtained, we used Scheffé'scriteria for multiple comparison(22). A two-tailed P value < 0.05 was considered statisticallysignificant.

RESULTS

HemodynamicsThe hemodynamics recorded at baseline and during the

different pharmacologic interventions are summarized inTable 1. Compared with the control condition, adenosinecaused a decrease in rate pressure product (RPP) (P0.097). Propranolol caused a significant decrease in systolicand diastolic blood pressure, heart rate, and RPP.

Arterial input FunctionFigure 1A shows the time—activitycurves of the arterial

input function obtained with the well counter and BACC-4.The integral values obtained between 0 and 3 mm anddisplayed in scatter plots between the results of the well

TABLE 1Summary of Hemodynamic Data During

Pharmacologic Interventions

p

Time (sac)

x1o65

A x10 xiO$

I

B

4

I

0

Va 3.31 x 1O-2X + 5.1 x 10@r—0.995n—S

6 9

Integrated Blood ActivityWell Counter (counts/mI)

12

x10e

FIGURE 1. (A) Time—activitycurvesof arterialinputfunctionobtained by samplingdirectly from ascendingaorta. 0 = wellcounter@S = BACC-4. (B) Companson of integral values of @oodsampling data withthose determined withBACC-4.

counter and BACC-4 measurements are shown in Figure lB.Although a time shift and dispersion of the arterial inputfunction were observed with BACC-4 and not with serialsampling data, there was a good correlation between theintegral values of serial sampling and BACC-4 data (r =0.995).

PET ImagingFigure 2A showstypical serial imagesof the heart of 1

rabbit obtained after intravenous administration of 38K atcontrol condition. Good image quality was observed. At themidventricular level, the bolus transit through the bloodpools and lungs was visualized during the first 30 s.

Thereafter, the myocardial image was delineated. Figure 2B

Control(n=8)

Adenosine(n= 5)

Propranolol(n=5) 92±24t 67±16t

129±15 93±13 308±31 40,082±7,788

126 ±18* 80 ±11 264 ±59 32,090±4,528*

236 ±SOt 21,017 ±2,856t

*P< 0.05versuspropranolol.tP < 0.05versuscontrol.BP = blood pressure; APP = rate pressure product and was

calculatedas the productof systolicBPand heartrate.Valuesaremean±SD.

38K PET IN RABBITS•Takami et al. 765

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

A@*,*0-

1010-2020-3030-40,.4.40-5050-6060-80150-

180secBIIII

A•S.@•.0-1010

2020-3030-40.@-

. . S@‘

@ @4;

@40

5050-6060-80150l8OsecBI

III•0120

@me(sac)C‘@o12-@T.•.-@-S@•r.@‘1@S,t)

0 10 120T@*@

C

,-AS%s%,/@#4

IP'123

0 ISO

the renal cortical ROl in the same rabbit in Figure 3A. Figure3C shows the summed images of the kidney obtained 2—3mm after the administration of 38K in a control-conditionrabbit. 38Kyielded high-quality PET images of the kidney.

Figure 4 shows the coronal images of the same normalrabbit brain obtained with 38K and FDG at control conditions. It shows summed images obtained 50—60mm afterthe administration of FDC, when the brain was clearlyvisualized. However, no visual 38K uptake was seen in thebrain.

FIGURE 2. (A) SerialPET imagesobtainedafterintravenousinjectionof @Kat midventncularlevel in 1 rabbit.Lateralmyocardium is on right; anterior is uppermost; and septum is on left. (B)Myocardialtime—activitycurve in same rabbit.(C)Threecontiguous cross-sectional myocardial images obtained 2—3mm afteradministrationof@Kina rabbitundercontrolcondition.Uptakeoftracer is homogeneous, and contrast between myocardiumandboth blood and lung is high. A = anterior@IP = inferoposterior@L =lateral; S = interventncular septum.

shows a typical time—activitycurve of the myocardium fromthe myocardial region of interest (ROl) in the same rabbit inFigure 2A. Figure 2C shows 3 contiguous summed imagesof the left ventricular myocardium obtained 2—3mm afterthe administration of 38K in a control-condition rabbit. Themyocardial image exhibited homogeneous accumulation of38Kactivity.

Figure 3A showsa typical seriesof cross-sectionalPETimages of the left kidney of 1 rabbit, obtained after theintravenous administration of 38K under control conditions.The renal cortex was seen primarily during the first 30 s.Figure 3B is a typical time—activitycurve of the kidney from

FIGURE3. (A) SerialPET imagesobtainedafter intravenousinjection with @Kat level of mid left of rabbit kidney. (B)Time—activitycurve of same kidney. (C) Six contiguous crosssectional images of kidney obtained 2—3mm after administrationof @Kin a rabbitundercontrolcondition.

766 Tiii@Joui@i@ OF NUCLEARMEDICINE •Vol. 41 No. 4 •April 2000

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

RegionalOrganbloodflow

(mL/min/g)Uptake(mL/min/g)ExtractionfractionHeart2.91

±1.291.50 ±0.380.55 ±0.13Kidneys5.49±0.712.59 ±0.640.48 ±0.13Brain0.57±0.110.012 ±0.0020.022 ±0.004

38K

2

FDG

FIGURE 4. Comparisonof coronalPETimagesofbrainobtainedwith @KandFDGin same rabbit. Note that brain images wereclearly delineated by FDG PET (bottomrow),buttherewasnovisualuptakeof @Kin brain (top row).2 3

Comparison of @KUptake with Blood Flow Measuredwith Microspheres

Table 2 shows 38Kuptake in organs (mL'min/g), regionalblood flow (mL/min/g) obtained with colored microspheres,and extraction fraction in the heart, kidney, and brain undercontrol conditions.

Figure5A showsthe correlationbetween38Kmyocardialuptakeandmyocardialbloodflowmeasuredsimultaneouslywith microspheres. The curve shows that the relation isalmost linear for flows <2.5—3.0 mL/min/g and that increases in flow lead to smaller changes in 38Kuptake. FigureSB shows the conelation between 38Kmyocardial extractionand flow. Applying the Renkin-Crone model (23,24) to therelation between 38K myocardial uptake and flow, the bestequation fining our data for the relation between them was:

U=FXE=FX(l—e2@)[r0.89]. Eq.3

The relationship between renal uptake of 38K and renalcortical blood flow under control conditions and withpharmacologic intervention is shown in Figure 6. Renal 38Kuptake tends to rise as flow increases, although the pointswere slightly scattered compared with the relation betweenmyocardial 38K uptake and flow. There was little uptake of

TABLE 2Regional Blood Flow, Uptake, and Extraction Fraction

38K in the brain and this did not correlate with brain bloodflow.

DISCUSSION

Since Sapirstein (25) described the fractionation distribution of indicators using 42K, the initial uptake of the isotopesof potassium analogs has been used as a marker of regionalperfusion, particularly in the heart. We focused on thekinetics of the first 3 mm after tracer injection to assess thefeasibility of 38K as a PET imaging tracer in the heart,kidney, and brain. We estimated the quality of PET imagesand simultaneously measured the uptake and extraction ofthe tracer for these 3 organs. The results showed that the initialtracer uptake could be clearly visualized by dynamic PET in theheart and kidney. As expected, the uptake and extraction werehigh for both organs. Although brain uptake and extraction wereextremely low because of the BBB, the possibilities offered bythis technique in pathologic situationssuch as brain tumor andradiation necrosis indicate the need for further investigations.

The fitted curve (Fig. 5A) shows that the increase ofmyocardial 38K uptake was linear up to a flow of approximately 2.5—3.0mL/min/g and then increased more graduallyat high flows. The myocardial extraction fraction variedinversely with flow, and the Renkin-Crone model fit therelation between them (23,24). We have reported previouslyon the permeability X surface area (the PS product) of13N-ammonia in the heart in the same experimental preparation (18). The average value of 13N-ammonia, 2.7 mLlmin/g,was slightly higher than that of 38K, probably becauseseveral percent of total ammonia is lipid-soluble NH3.Bassingthwaighte et al. (26) recently observed that the initialuptake of thallium exceeded that of potassium, interpretingthis to mean that Na channels were certainly activated and

38K PET IN RABBITS •Takami et al. 767

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

A2.5

E

a•

Iaccount for much thallium transport compared with potassium. The comparison between the kinetics ofpotassium andits analogs is important. Additional experiments need to beperformed using potassium analogs such as 82Rband 201'flinthe same experimental preparation. 38Kand PET may also beuseful for the in vivo evaluation of the effects on myocardialpotassium uptake under various pathophysiologic statesaffecting potassium channels and Na-K adenosine triphosphatase. The averaged myocardial blood flow at baseline in rabbits isreported to be 1.7-2.4 mUmin/g (27), which compares favorablywith our data. However, these values are higher than those indogs and humans, and care should be taken in applying theresults ofthe rabbit studies to data from humans and dogs.

In this study, we measured cortical blood flow in thekidney, because it comprises 80%—90%of total renal bloodflow (28). The flow values measured with microspheres inrabbits in this study are comparable with those in differentanimal species and human studies summarized by Nitzscheet al. (29). The averaged extraction fraction of 38K in thekidney measured at 0.48 is close to the value of 0.53measured by Tamaki et al. (4) using dogs and 82Rb PET witha steady-state infusion model. Mullani et al. (5) reported thatapproximately 44.5% of the injected dose of 82Rbis excreted

II'

5.

3

@•3 ox

I 2 x ••x11 x

0

0

. 0

0 2 4 6 8 10RenalBloodFlow(mIImin/g)(microepheres)

FIGURE6. Relationshipbetween @Krenaluptake(extractionxflow)and actual blood flow measured with microspheres.

in the venous side and 9.75% in the ureter during the 150 s ofdata collection in a dog experiment. Thus, the extractionfraction is approximately 0.45, which represents an extraction fraction equivalent to that in our experimental preparation. Renal 38K uptake tends to rise as flow increases,although data were slightly scattered compared with therelation between myocardial 38K uptake and flow. In thisstudy, adenosine and propranolol were chosen as pharmacologic interventions, mainly for assessing the 38K kinetics inthe heart. In the kidney, however, renal blood flow is wellcontrolled by autoregulation and does not increase remarkably above baseline. Additional studies need to be performed for a variety of pathophysiologic states related todecreased renal blood flow, such as renal artery stenosis, toinvestigate 38Kkinetics in the kidney.

Potassium and its analogs cross the intact BBB slowly (6),so the initial uptake and extraction there were extremely lowcompared with those in the heart and kidney. Our meanbaseline 38K extraction of 0.022 was close to the value of0.021 measured by Brooks et al. (30) using 82Rb PET inhumans, and the mean baseline 38K uptake of 12 pL/min/gwas higher than that estimated by direct quantification of

@Rbuptake in the brain from normal rats (3.5 ± 0.3piLimin/g) as reported by Cserr et al. (31). Comparisonsbetween the values should be carefully considered, asdifferent species and experimental preparations were involved. Brain tumor uptake of potassium analogs wasextremely high compared with that in the healthy brain andwas a combined function of BBB permeability and cellviability. 201'fl is widely used for this purpose as a SPECTtracer (7). 82Rb also has the potential for this application(30,32—35). The kinetics of 38K in brain tumor should be

investigated further.

CONCLUSION

High-resolution PET for small animals was used to assessthe potential capability of positron-emitting 38Kfor imaging

0 0

.

5.U —F x (i-e@41'@)r—0.89n —18

0 2 4 6 8 10Myocardlal Blood Flow (mI/mln/g) (Microspheres)

0.8

0.6

0.4

0.2

-2.4/FC - 1-e

00 2

Myocardial Blood4 6 8 10Flow (ml/minlg) (microspheres)

FIGURE5. (A) Relationshipbetween3@Kmyocardialuptake(extractionx flow)andactualbloodflowmeasuredwithmicrospheres. 0 = Adenosine; •= controls; X = propranolol. (B)Relationshipbetween @Kmyocardialextractionfractionandblood flow measured with microspheres.

768 Tm@JouRN,@i OF NUCLEARMEDICINE •Vol. 41 •No. 4 •April 2000

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

14. Bol A, Melin JA, Wijns M, et al. In vivo assessment of myocardial perfusion andviability by kinetics of potassium-38 [abstract]. J NuclMed. 1994;35:77P.

15. Melon P0, De LandsheereC, Degueldre C, PetersIL, Kulbertus HE, PierardLC.Relation between contractile reserve and positron emission tomographic patternsof perfusion and glucose utilization in chronic ischemic left ventricular dysfunclion. JAm Coil Cardiol. l997;30: 1651—1659.

16. Duncan CC, Lambrecht RM, Bennett GW,RescignoA, Ment LR. Observations ofthe dynamics ofionic potassium-38 in brain. Stroke. 1984;15:145—148.

17. Nagatsu K, KuboderaA, Suzuki K. A novel way of producing an aqueous solution

of@K@via the @°Ar(p,3n)-process.AppiRadiatiso:. 1998;49:1505—1510.18. Shiinada K, Yoshida K, Tadokoro H, Ctal. High-resolution cardiac PET in rabbits:

imaging and quantitation of myocardial blood flow. J Nuci Med. 1998;39:2022—2027.

19. Watanabe M, Uchida H, Okada H, et al. Ahigh resolution PET for animal studies.

IEEE Trans Med imaging. l992;l 1:577—580.20. Hale SL, Alker KJ, Kioner RA. Evaluation of nonradioactive, colored micro

spheres for measurement of regional myocardial blood flow in dogs. Circulation.1988;78:428—434.

21. Webb Al, Weaver BM. Density of equine tissue at 37°C.Res VetSci. 1979;26:71—75.

22. Wallenstein S. Zucker CL, Fleiss JL. Some statistical methods useful incirculation research. Circ Res. 1980;47:1—9.

23. Renkin EM. Transport of potassium-42 from blood to tissue in isolated mamma

han skeletal muscles. Am J Physiol. 1959;197: 1205—1210.

24. Crone C. The permeability ofcapillaries in various organs as determined by use of

the indicator diffusion method. Acta PhysiolScand@ 1963;58:292—305.25. SapirsteinLA. Regionalbloodflow by fractionaldistributionof indicators.AmJ

Physiol. 1958;193:161—168.26. Bassingthwaighte lB. V/inkier B, King RB. Potassium and thallium uptake in dog

myocardium. JNucl Me€L1997;38:264-274.

27. BrewerNR, CroiseU. Physiology.In: ManningPJ,RinglerDH, NewcomerCE,eds. The Biology of the Laboratory Rabbit. 2nd ed. San Diego, CA: AcademicPress; 1994:63—64.

28. Knox FG, Spielman WS. Renal circulation. In: Shepherd if, Abboud FM, eds.

Handbook ofPhysioiogy: The Cardiovascular System. Bethesda, MD: AmericanPhysiologic Society; 1983:183-217.

29. Nitzsche EU, Choi Y, Killion D, et al. Quantification and parametric imaging ofrenal cortical blood flow in vivo based on Patlak graphical analysis. Kidney kit.1993;44:985—996.

30. Brooks DJ, Beaney RI@Lammertsma AA, et al. Quantitative measurement of

blood-brain barrier permeability using rubidium-82 and positron emission tomog

raphy. J Cereb Blood Flow Metab. 1984;4:535—545.31. Cserr HF, DePasquale M, Patlak CS. Volume regulatory influx of electrolytes

from plasma to brain during acute hyperosmolarity. Am I Physiol. 1987;253:F530—F537.

32. Valk PE, Budinger IF, Levin VA, Silver P. Gutin PH, Doyle WK. PET ofmalignant cerebral tumors after interstitial brachytherapy. Demonstration of

metabolic activity and correlation with clinical outcome. I Neurosurg. 1988;69:

830—838.33. Dhawan V, Jarden JO, MoellerJR, Strother SC, Rottenberg DA. Positron emission

tomographic measurement of blood-to-brain and blood-to-tumour transport of82Rb. II: clinical data and validation of technique. Phys MedBiol. 1989;34:1875—1794.

34. Zunkeler B, Carson RE, Olson J, et al. Hyperosmolar blood-brain barnerdisruption in baboons: an in vivo study using positron emission tomography andrubidium-82. I Neurosurg. 1996;84:494—502.

35. Roeckle U, Radu EW, Hausmann 0, Vontobel P. Maguire R1@Leenders KL.

Tracer transport and metabolism in a patient with juvenile pilocytotic astrocytoma. A PET study.JNeurooncol. l998;36;279—283.

heart, kidney, and brain. 38Kuptake was high for the heartand kidneys, and high-quality PET images were obtained 3mm after the tracer injection. Although 38K uptake in thebrain was extremely low because of intact BBB, it might beuseful for assessing BBB damage. 38K and PET wouldenable the noninvasive estimation of the kinetics of potassium in the heart, kidneys, and brain.

ACKNOWLEDGMENTS

The authors thank Keiji Shimizu, Tsuyoshi Kosugi, andHiroyuki Okada, Hamamatsu Photonics K.K., Hamamatsu,Japan, for their excellent assistance with PET imaging. Theyalso thank Kotaro Nagatsu and Makoto Takei for preparing

the 38K. This work was supported in part by the JapaneseSpecial Coordination Fund for the Promotion of Science andTechnology and by grants from the National CardiovascularCenter, Smoking Research Foundation, and KashiwadoMemorial Foundation of Japan.

REFERENCES1. Bonow RO, Dilsizian v. Thatlium-201 for assessment of myocardial viability.

SeminNuclMed@ l991;21:230—241.

2. Gould KL. Clinical cardiacpositron emissiontomography:state of the art.Circulation. 1991;84: 122—136.

3. Hurwitz GA, Powe JE, Wesolowski CA, MattarAG. Comparison ofll-201 renaluptake with Tc-99m DTPA angiorenography in patients with hypertension.Measures of renal asymmetry. Cliii NuclMed@ 1992;17:463—468.

4. Tamaki N, Rabito CA, Alpert NM, et al. Serial analysis of renal blood flow bypositron tomography with rubidium-82. Am I PhysioL 1986;25 1:H1024—H1030.

5. Mullani NA, Ekas RD. Marani S. Kim EE, Gould KL. Feasibility of measuringfirst pass extraction and flow with rubidium-82 in the kidneys. Am J Physiolimaging. 1990;5:133—140.

6. BetzAL. Transportof ions acrossthe blood-brainbarrier.FederationPmc.1986;45:2050—2054.

7. Bruneni A, Alfano B, Soricelli A, et al. Functional characterization of brainWmors: an overview of the potential clinical value. NucI Med Biol. 1996;23:699—715.

8. Lambrecht RM, Ham T, Gallagher BM, Wo1fAP, Ansari A, Atkins H. Cyclotronisotopes and radiopharmaceuticals-XXVffl. Production of potassium-38 formyocardial perfusion studies. in: JAppi Radio: iso:. 1978;29:667—671.

9. Myers WG, Bigler RE, Graham MC. PET tomography in studies of distributions

ofl.6-min potassium-38 in the dog heart. EurJNuc!Med@ 1984;9:272—277.10. Guillaume M, De Landsheere C, Rigo P. Czichosz R. Automated production of

potassium-38 for the study of myocardial perfusion using positron emissiontomography. App! Radio: iso:. 1988;39:97—107.

11. Duboc D, Kahan A, Maziere B, et al. The effect of nifedipme on myocardialperfusion and metabolism in systemic sclerosis. A positron emission tomographystudy. Arthritis Rheum. 1991;34: 198—203.

12. De Landsheere C, Mannheimer C, Habets A, et a!. Effect of spinal cordstimulation on regional myocardial perfusion assessed by positron emissiontomography. AmlCardiol. 1992;69:1143—1149.

13. Melon P0, Brihaye C, Degueldre C, et al. Myocardial kinetics ofpotassium-38 inhumans and comparison with copper-62-PTSM. JNuclMed. 1994;35:1116—1122.

38K PET IN RABBITS•Takami et al. 769

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from

2000;41:763-769.J Nucl Med.   Yoshiaki Masuda and Shuji TanadaAkira Takami, Katsuya Yoshida, Hiroyuki Tadokoro, Shinobu Kitsukawa, Kazuhiro Shimada, Mikio Sato, Kazutoshi Suzuki, 

K in Rabbit Heart, Kidney, and Brain38Uptakes and Images of

http://jnm.snmjournals.org/content/41/4/763This 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 2000 SNMMI; all rights reserved.

by on February 9, 2020. For personal use only. jnm.snmjournals.org Downloaded from