he technique of transcatheter therapeutic embolization with Ivalon has been in use since 1975 (1-5).Ivalon is made of polyvinyl alcohol that is an inert,nonbiodegradable substance. Ivalon particles aresponge-like in shape with jagged edges and large interstitial spaces (1). Because the Ivalon particles are radiolucent and on the order of 1 mm in size, they cannotbe directly visualized by fluoroscopy.

Inadvertent pulmonary embolization and reflux migration of Ivalon particles are two serious and potentially fatal complications of therapeutic embolizationthat may go unheeded. When the Ivalon particle size issmaller than the tumor vasculature, internal shuntingof Ivalon particles through the tumor may occur resulting in inadvertent pulmonary embolism. When preferential blood flow to the tumor is inadequate to carrythe Ivalon particles into the tumor, reflux migration ofIvalon particles may occur. A recent report has docu

ReceivedDec. 12, 1988;revisionaccepted Mar. 6, 1989.Forreprintscontact:StevenA. Sirr,HennepinCountyMedical

Center,701ParkAve.South,Minneapolis,MN55415.

mented two fatalities from inadvertent pulmonary embolism during therapeutic embolization oflvalon (6).

Dynamic scintigraphic detection of radiolabeled Ivalon particles using a portable gamma camera positionedin the angiographic suite permits precise localization ofradiolabeled Ivalon particles. Intermittent repositionings of the gamma camera head over the tumor, thechest, and distal to the tumor while constantly viewingthe persistence scope of the gamma camera permitsdynamic monitoring of inadvertent pulmonary embolism and reflux migration of radiolabeled Ivalon partides. When the distance from the image intensifier ofthe fluoroscopic tower to the patient's chest is large, thegamma camera head can be left in position over thepatient's chest thereby permitting continuous monitoring of possible inadvertent pulmonary embolism ofradiolabeled Ivalon particles.

In 1986, Jack et al. described a technique for radiolabeling Ivalon particles with technetium-99m sulfurcolloid ([99mTc]SC) ( 7). In his radiolabeling method,[99mTc]SCwas first produced in the standard mannerand then Ivalon particles were subsequently labeled byheating them in a suspension with the [@mTc]SC. The

1399Volume30 •Number 8 •August1989

An Improved Radiolabeling Techniqueof Ivalon and Its Use for Dynamic Monitoringof Complications During TherapeuticTranscatheter EmbolizationSteven A. Sirr, Timothy K. Johnson, David D. Stuart, Warren R. Stanchfleld,John F. Cardella, Rene P. duCret, and Robert J. Boudreau

Department ofRadiology, Division ofNuclear Medicine, Hennepin County Medical Center;Department ofRadiology, Metropolitan Medical Center; and Department of Radiology,Division ofNuclear Medicine, University ofMinnesota Hospitals, Minneapolis, Minnesota

Transcatheter embolization by Ivalonparticles for treatment of arteriovenous malformationshas been an accepted therapeutic technique for many years. We describe a new and efficientradiolabeling technique of lvalon particles using [@“Tc]suiturcolloid. Continuous and dynamicmonitoringof injected radiolabeled Ivalonparticles is made possible by viewingfriepersistence scope of a portable gamma camera whose head is positioned over the patientundergoing therapeutic embolization. Therefore, if inadvertent pulmonary embolism or refluxmigration of radiolabeled Ivalon partides has occurred, the angiographer is immediately awareof this potentially serious or fatal complication and can take corrective action. We describetwo patients, each with an artenovenous malformation, who had therapeutic embolizationwith radiolabeled Ivalon particles, one resulting in reflux migration and the other resulting ininadvertent pulmonary embolism.

J Nucl Med 30:1399—1404,1989

mechanism of labeling of Ivalon by [99mTcJSC,whoseparticle size is

TABLEIInvitro Percent LabelingEfficiencyfor [@“TcJSCIvalonParticles

in Normal Saline for the Modified RadiolabelingTechnique and the Technique of Jack etal.Time

Percent labelingefficiency±s.e.(hoursafter

MOdifiedlabeling Jack labelinglabeling) technique technique

K21 .025hr1LIVER

(85%)

@—@--a

. Numbers in parenthesis indicate number of incubations. Data

has been corrected for physicaldecay of @Tc.

0 0.110±0.015(8)1.7 0.098 ±0.010(5)2.5 0.100±0.007(3)4.4 0.095±0.008(5)5.2 0.095 ±0.015(3)5.5 0.090±0.015 (5)

positioned close to the angiographic table. A sterile plasticdrape was placed over the gamma camera head to ensuresterility of the field. The persistence scope of the portablegammacamerawasseton highintensity.In order to diminishexposureof radiation to the angiographerand technologistsby the radioactive Ivalon particles, the particles were surrounded by a 3-mm lead shield.

Afterthe initial embolization of[@mTc]SCIvalon particlesto the AVM using a 9 French catheter, the gamma cameraheadwasperiodicallyrepositionedbetweenthe lowerextremity and the chest permitting monitoring ofrefiux embolizationto the distal lowerextremityand for inadvertent pulmonaryembolism. During most of the embolization procedure, thegamma camera head was left in position over the chest permifting continuous monitoring ofinadvertent pulmonary embolism.At no time during the entire embolizationprocedurewas a solitary focus of radioactivity activity seen within thelungs. A repeat angiogramperformed after therapeutic embolization demonstrated obliteration of much of the tumorbulk.

After the completion of therapeutic embolization, the patient wastaken to nuclearmedicineto confirmour impressionthat no pulmonary emboli or refiux migration occurred. Awhole-bodyscintigramand gammacameraimagesofthe rightknee showed numerous [@mTc]ScIvalon particles withinthe AVM.No [99mTc]SCIvalonparticleswereseenwithinthelungs or distal to the knee. Next, a lung perfusion scan using2.5 mCi [@Tc]MAA was performed which was normal, alsoconfirming that no Ivalon particles embolized to the lung.

The patient returnedto the hospital after 1 yr because ofincreasing knee pain and swelling. A femoral artery angiogramrevealed at least four small clusters of AVM which had increased in size from the post-therapeutic embolism angiogramof I yr earlier. A repeat therapeutic embolization [@mTcJSCIvalonwasthen performedusinga 9 Frenchcatheter.

The portablegamma camera with a LEAP collimator waspositioned next to the angiographic table. The gamma camerahead, coveredwith a sterile plastic drape, was initially positioned over the chest to monitor for inadvertent pulmonaryembolism of the first injected Ivalon particles. No focus ofabnormal lung activity was noted and the gamma camerahead was then repositioned over the AVM where multiple fociofincreased activity where seen within the AVM. The gamma

0.047 ±0.006(8)0.028 ±0.010(8)

0.020 ±0.012(8)

SPLEEN(8%)

FIGURE 1Summary of MABDOS dosimetry model for [@rc]SCIvalonparticles used for therapeutic transcatheter embolization.

[@mTcJSCfrom [99mTc]SCIvalon particles in solutionsof normal saline, ionic, or nonionic contrast.

CASE REPORTS

Case IA 37-yr-oldwoman with known congenitalhemangioma

of the distal thigh and upper calf of the right leg complainedof increasing pain and swelling with ulceration of her rightleg. An angiogram performed prior to therapeutic embolization demonstratesa largeAVMsuppliedby multiplebranchesof both the superficial and deep femoral arteries. A lungperfusion scan using 2.5 mCi [@mTc]MAAperformed —8hrbefore therapeutic embolization was normal. Prior to therapeutic embolization of the AVM, 1.0 g of sterile Ivalonparticles 1.0 mm in size were radiolabeled with 104 mCi[99mTc]pertechnetate using our modified radiolabeling tech

nique. The [99mTc]SCIvalon particles, total activity of 11.0mCi, were then placed in normal saline and carried to theangiographicsuite.

A portablegamma camera(Picker Dyna Mo) with a LEAPcollimator was maneuvered into the angiographic suite and

1401Volume 30 •Number 8 •August1989

AVM

WHOLEBODY!BONEMARROW

(7%)

Adrenals59.6Marrow red23.2 AVM3449.8Bladderwall3.70th tissmuse41.5Bone

total69.3Ovaries9.5GIstomwall32.8Pancreas72.7GI

si19.9Skin37.5Gluliwall31.9Spleen341.6GIIliwall5.7Testes7.7Kidneys52.9Thyroid2.5Liver470.8Uterus

nongrv9.1Lungs31.8Total body85.2

TABLE 2MABDOSDosimetry Data for Patient 1

CumulativeOrgan Dose (mrad)

cameraheadwasthen repositionedover the lungswhilemoreradioactive Ivalon particles were injected. At no time duringthe entire procedure was a focus of increased activity seenwithin the lungs. However, after embolization of one of thesmall AVM clusters, the gamma camera head was repositionedover the lower extremity distal to the AVM where multiplefoci of increased activity were immediately seen on the persistence scope by the angiographer and technologists. Refiuxmigration of injected Ivalon particles was then diagnosed andthe catheter was then immediately repositioned. RemainingAVMclusterswereembolizedwithoutevidenceofinadvertentpulmonary embolism or more refiux migration. During embolization, the gamma camera head was frequently repositioned over the chest and distal to the AVM. At no timeduring therapeutic embolization was a solitary focus of increased activity detected with the lungs.

In order to document the reflux migration of [@mTc]SCIvalonparticles,imagingofthe lowerextremitywas performedwithin the nuclear medicinedepartment -@‘-8hr after comple

tion of the embolization procedure (Fig. 2). The gammacamera imageverifiesthe multiple foci of activity in the calfand foot seeninitiallyon the persistencescopeas representingrefiux migration.

The MABDOS dosimetry for Patient 1 is shown in Table2.

Case 2A 45-yr-oldwomanwitha knownpelvicAVMcomplained

of worsening discomfort. An angiogram demonstrated a verylargeAVM involvingthe right pelvisincludingthe right buttock. Using the modified radiolabeling technique, 1 g of 1.0-mm-sized Ivalon particles was labeled yielding 8.5 mCi of[99mTc]SCIvalon particles. The portable gamma camera wasmaneuvered into the angiographic suite and the gamma camera head was covered with a sterile plastic drape. With thegamma camera head positioned over the patient's chest, several l.0-mm-radiolabeled Ivalon particles were injectedthrough a 9 French catheter. An intense solitary focus ofradioactivity in the lung was immediately seen on the persistence scope of the gamma camera. A 60-sec portable gammacamera imagewasthen taken for documentation of inadvertent pulmonary embolism (Fig. 3). The catheter was repositioned and embolization proceededwithout complication. Anangiogram performed at the termination of the embolizationprocedure showed decrease ofAVM size.

At no time during the therapeutic embolization of eitherCase 1 or 2 with radiolabeled Ivalon particles was activitydiffuselyseen within the catheter deliveringthe Ivalon partides. No thyroidal activity was noted by gamma camera ineither Patient 1 or 2.

DISCUSSION

Ivalon particle transcatheter embolization has beenused in a wide variety oftherapeutic procedures including embolization to control acute hemorrhage, tumorembolization, and embolization of arteriovenous malformations (12—18). In order to achieve a satisfactorytherapeutic endpoint, it is not uncommon for the angiographer to use many hundreds or even thousands ofIvalon particles. Dynamic monitoring of[@mTc]SCIvaion particles with a portable gamma camera duringtherapeutic embolization permits continuous and precisc localization of each [99mTc]SCIvalon particle. Potentially serious or life-threatening complications such

FIGURE 2Gamma camera image of the calf (right lateral projection)-@-8hr after the embolization procedure confirming thatreflux migration of [@“TcJSCIvalon particles to the calf,ankle, and foot had occurred.

1402 Sin',Johnson,Stuartetal The Journal of Nuclear Medicine

0•

4

..@ p

.@

I@ /FIGURE 3A60-sec anterior scintigraphicimagefrom portable gamma camera headpositioned over the chest. Note solitary focus of radioactivity from a single 1.0-mm [@Tc]SC colloid Ivalonparticle in the apex of the right lung.

as inadvertent pulmonary embolism and reflux migration may be immediately detected.

Unlabeled Ivalon particles are very small and radiolucent and therefore they are impossible to detect byfluoroscopy. Recently, a commercial Ivalon impregnated with barium became available for transcatherembolization (Ingenor, Paris, France). However, because of overlying soft tissue, bone, and vasculature,the small barium impregnated Ivalon particles may goundetected by fluoroscopy. In addition, there is a moreserious problem—this commercial Ivalon preparationcontains @@.-80,000particles from 4 to 50 nm in size,which may result in pulmonary embolism, as has beendemonstrated by Repa et al. They have concluded thatembolization with barium impregnated Ivalon is dangerous and may be fatal (6).

Our modified radiolabelingtechnique differs fromthe method previously described by Jack et al. becausethe [99mTclSC formation and the Ivalon labeling proceed simultaneously in a single reaction vessel. Thissimplifies the preparation of [@mTcJSC Ivalon particlesand results in a 2.3 times higher labeling efficiency. Wepostulate that this increase in labeling efficiency is theresult of physical adsorption of [@mTc]SC not only onthe exposed external surface ofeach Ivalon particle, but

also within the interstitial spaces ofeach particle. Whilethe overall labeling efficiency is not high for eitherradiolabeling technique, a 2.3 times increase of labelingefficiency yields many more [9@Tc]SC Ivalon particlesfor a given amount of [99mTcjpertechnetate.This isclinically important because our modified radiolabelingtechnique produces more [99mTc]SCIvalon particles tobe used by the angiographer who is performing thetherapeutic embolization. We have demonstrated that[99mTc]SC Ivalon particles are stable in normal saline,ionic, and nonionic contrast media and may be keptfor several hours prior to therapeutic embolization without significant loss of the radiolabel. Compared to theJack radiolabeling technique, the modified techniquetakes less time and requires only one transfer therebydecreasing the likelihood of microbial contamination.

Because of the relatively small attenuation for the140-keV photons of@mTc in lung parenchyma and thenearly complete lack of background activity from thepatient at the time of therapeutic embolization, thedetection of inadvertent pulmonary embolism of radiolabeled Ivalon particles using a modern portablegamma camera is very sensitive.

The method of use for the radiolabeled Ivalon partides depends upon the total number of radiolabeled

1403Volume30 •Number 8 •August1989

Roentgenol1975;125:609—616.2. Castaneda-Zuniga WR, Sanchez R, Ampiatz K. Ex

perimental observations on short and long-term effectsof arterial occlusion with Ivalon. Radiology 1978;126:783—785.

3. Zollikofer C, Castaneda-Zuniga WR, Gaffiani C, Rysavy JA, Formanek A, Amplatz K. Therapeutic blockade of arteries using compressed Ivalon. Radiology1980; 136:635—640.

4. Herrera M, Rysavy J, Kotula F, Castaneda-ZunigaWR, Amplatz K. Ivalon shavings: technical considerations of a new embolic agent. Radiology 1982;144:638—640.

5. Jack CR, Forbes G, Dewanjee MK, Brown ML, Earnest F. Polyvinyl alcohol sponge for embolotherapy:particle size and morphology. AJNR 1985; 6:595—597.

6. Repa I, Moradian GA, Tadavarthy MS, et al. Twofatal complications with Ivalon embolization using acommercially available compound. Radiology 1988;169(P),260.

7. Jack CR, Dewanjee MK, Brown ML, Forbes G, Chowdhury S. Radiolabeied polyvinyl alcoholparticles: apotential agent to monitor embolization procedures.mtj RcAApplInstrum1986;13:235—243.

8. Johnson TK, Vessella RL. A generalized dosimetryschema for tumor preferential uptake of monoclonalantibodies in radionuclide immunotherapy [Abstract]. JNuclMed 1987;28(suppl 4):680.

9. Johnson TK. MABDOS:a program for the estimationof dose resulting from the administration of labeledmonoclonal antibodies for radioimmunotherapy [Abstract].MedPhys 1987; 14:456.

10. Johnson TK. MABDOS: a generalized program forinternal radionucide dosimetry. Comput Meth ProgramsBiomed 1988;27:159—167.

11. Summary of current radiation dose estimates to humans with various liver conditions from Tc-99m-sulfur colloid. MIRD Dose Estimate Report No. 3. JNuciMed 1975;l6:108A-B.

12. Castaneda-Zuniga WR, Lehnert M, Nath PH, Zoffikofer C, Verlazquez G, Ampiatz K. Therapeutic embolization of facial arteriovenous fistulae. Radiology1979; 132:599—602.

13. Bhansali 5, Wilner H, Jacobs JR. Arterial embolization for control ofbieeding in advanced head and neckcarcinoma. J Laryngol Owl 1986; 100:1289—1293.

14. Carrasco CH, Wallace 5, Charnsangavej C, Papadopoulos NE, Patt YZ, Mavligit GM. Treatment ofhepatic metastasis in ocular melanoma. Embolizationofthe hepatic artery with polyvinyl sponge and cisplatin.JAMA 1986;255:3152—3154.

15. Tisnado J, Cho SR, Beachiey MC, Margolius DA.Transcatheter embolization of angiodysplasia of therectum. Report of a case. Acta Radiol (Diagn) 1985;26:677—680.

16. Schwartz DN, Keilman RM, Cacayorin ED. Treatment of a lingual hemangioma by superselective embolization. Arch Otolaryngol Head Neck Surg 1986;112:96—98.

17. Scialfa G, Scotti G. Superselective injection of polyvinyl alcohol microemboli for the treatment of cerebral arteriovenous malformations. AJNR 1985; 6:957—960.

18. Widlus DM, Murray RR, White RI, et al. Congenitalarteriovenous malformations: tailored emboiotherapy.Radiology1988;169:511—516.

particles available for use by the angiographer and theextent of the AVM.If only a fewhundred radiolabeledIvalon particles are produced, the “hot―particles canbe embolized intermittently between a series of “cold―Ivalon particles. Ifa large number ofradiolabeled Ivalonparticles are produced, most, if not all ofthe embolizedIvalon particles can be radiolabeled. A small AVMrequires a relatively smaller number of Ivalon particlesto be embolized for a successful outcome.

A potential problem ofan artifactual “hotspot―existsfrom the common practice of many angiographers toshake off the last droplet of fluid from the syringe tipbefore injection. If a radiolabeled Ivalon particle isaccidentally flipped onto the sterile cloth sheet coveringthe patient, the potential for misdiagnosing inadvertentpulmonary embolism or reflux migration exists. Angiographers should be warned not to shake off the lastdroplet from the syringe tip.

We believe that the postulated mathematic modelused for estimation ofdosimetry for the kinetics is validfor thrombosed radioactive Ivalon particles embolizedto an AVM. This approach makes the assumption thatthe rate at which the sulfur colloid leaches off embolizedradiolabeled Ivalon particles within the AVM is thesame as the in vitro dissociation rate. Other than thefraction of seconds needed for the Ivalon particle tomove through the vessel feeding into the AVM, the invitro environment and the thrombosed environmentembedding the radioactive Ivalon particles embolizedwithin an AVM are roughly equivalent.

The dosimetry of [99mTcJSCIvalon embolization iscomparable to many other diagnostic examinations innuclear medicine and radiology. The typical radiationdose from fluoroscopy is -@-5rad/min of fluoroscopictime. Therapeutic embolization procedures are commonly 10 or more minutes in fluoroscopic length.Therefore, the added radiation burden to the patientfrom [99mTc]SCIvalon particles appears reasonable.Also, we believe that the added radiation risk is balanced by the benefit ofdetecting the potentially seriousor fatal complications of inadvertent pulmonary embolism and reflux migration ofradiolabeled Ivalon partides.

ACKNOWLEDGMENTS

The authors thank Sharon Coulter, CNMT, Kimberly Ziccarelli, CNMT, Dana Walker, CNMT, Kraig Schuster,CNMT, and Julianne Peterson, CNMT for their technicalcontributions. Cray X/MP computing time was providedcourtesyof Cray Research,Inc.

REFERENCES

1. Tadavarthy SM, Moller JH, Amplatz K. Polyvinylalcohol (Ivalon)—a new embolic material. Am J

1404 Sirr,Johnson,Stuartetal The Journal of Nuclear Medicine