FALSE.POSITIVETHROMBOSCINTIGRAMRESULTING FROM LYMPHEDEMA-A ROENTGENPATHOLOGICAL MODELJames D. Collins, MD, Lawrence W. Bassett, MD, Harold D. Snow, DVM, Nancy A. Ross, BA,and Thomas Patin

Abnormal thromboscintigrams were observedin patients with lymphatic obstruction. Thissyndrome was reproduced by surgically ligat-ing the lymphatic drainage of the lower ex-tremity of a dog prior to thromboscintigraphy.

Thromboscintigraphy is a benign screeningprocedure for thrombophlebitis.1-4 Thromboscin-tigraphy (radionuclide-clot localization) dependson particle entrapment (usually technetium labeledmacroaggregates of albumin) in areas of fibrindeposition or endothelial damage to diagnose clotformation. On isotope scans, this entrapment re-

From the UCLA School of Medicine, Department of Radiol-ogy, Los Angeles, California. Presented at the NationalMedical Association's 86th Annual Convention and Scien-tific Assembly, July 23, 1981. Requests for reprints shouldbe addressed to Dr. James D. Collins, Department ofRadiology, UCLA School of Medicine, Los Angeles, CA90024.

sults in localized abnormal accumulations of theisotope. Any factor that interrupts the flow of theradionuclide results in abnormal accumulations ofthe isotope.

In 1975, the senior author used thromboscintig-raphy for bipedal lymphangiogram to establishradionuclide perfusion in dogs.5 In August 1976,Grollman et al6 reported that while thromboscin-tigraphy is highly sensitive for detection of throm-bophlebitis, the false-positive results are alsohigh. These false positives have been attributed toseveral causes including venous stasis, endothelialabnormalities, and cellulitis.7

In several cases of lymphedema with lym-phangiograms demonstrating obstructions of thelymphatics, false-positive thromboscintigramswere obtained. The venograms were entirely nor-mal. These cases were clinically suspected of hav-ing thrombophlebitis involving the legs. The lym-phatic abnormalities of lymph stasis, perivascularstaining, and dermal back flow were demonstratedon the subsequent lymphangiograms. As the resultof this observation, an experiment was designed toContinued on page 878

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 78, NO. 9, 1986 875

FALSE-POSITIVE THROMBOSCINTIGRAM

Continued from page 875

.. 44

.... :.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.........

1* t i r1

4K~~~~~~~~~4Figure A k L ;+

Figure 1. A 21-year-old woman with recurrentepisodes of right lower extremity swelling. Athromboscintigram performed after injection oftechnetium 99m labeled macroaggregates of al-bumin in both feet shows abnormal accumulationsof isotope throughout the right lower extremity, in-dicating possible thrombophlebitis. Venogramswere normal

determine if a false-positive thromboscintigramcould be created in an animal model.

The purpose of this paper is to present a repre-sentative clinical example of thrombophlebitis andto use an animal model to determine the possiblecause of false-positive thromboscintigram.

CASE REPORTA 21-year-old woman had swelling and pain in

the left leg. A clinical examination resulted in adiagnosis of thrombophlebitis. She was treated

Figure 2. A lymphangiogram reveals absence of thedeep lymphatic nodal structures and perivesicularaccumulations of lymphangiogram contrast agent(Ethiodol). Same patient as shown in Figure 1

with heparin for 10 days. The swelling decreased,but never totally resolved. One month after com-pleted therapy, she returned with recurrent leg

878 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 78, NO. 9, 1986

FALSE-POSITIVE THROMBOSCINTIGRAM

swelling and pleuritic chest pain. Despite a nega-tive lung scan, she again was treated with heparinand discharged. In another month, she again hadleg swelling, was treated with heparin, and dis-charged. A month later, she was admitted for simi-lar complaints, and as thrombophlebitis was con-sidered likely, she was discharged. She continuedto complain of swelling in her left leg. A throm-boscintigram showed abnormal accumulation ofisotope throughout the popliteal and calf region,indicating possible thrombophlebitis (Figure 1).The venogram, however, was normal. Finally, alymphangiogram revealed congenital absence ofdeep lymph nodes, atrophy of normal lymphaticdrainage, and perivascular accumulation of con-trast material in the left lower extremity (Figure 2).

METHOD AND MATERIALSAn experiment was performed on a 32-kg dog to

verify that a false-positive thromboscintigramcould result from a surgically induced lymphaticobstruction. To establish a normal baseline, athromboscintigram, phlebogram, and lymphan-giogram were performed. The baseline studieswere again repeated after surgical preparation ofthe animal model. All studies were normal.A description of the experimental procedures is

given in the following sections.

LymphangiogramBefore the procedure, the dog was anesthetized

by an intravenous injection of 18 mL of a 5 percentsolution of sodium thiamylal (Surital). Approx-imately 30 minutes later, 0.5 mL of methylene blueand a 1 percent lidocaine hydrochloride solution(Xylocaine) was injected into the subcutaneoustissue of both hind feet. The animal was then pre-pared, draped, and a superficial cutdown was per-formed over the dorsum of both feet. The lymphat-ics were then isolated, cannulated with a Via-monte type needle and 5 mL of Ethiodol was in-jected bilaterally. Routine radiographs of the hindlegs, pelvis, and abdomen were obtained on com-pletion of the study and at the 24-hour interval.

ThromboscintigramA technetium-sulfur suspension was made after

a modified Patton procedure. Into a clean, dry,sterile, stoppered serum vial was placed 3.5 mL ofpertechnetate solution in saline from a sterilemolybdenum-technetium 99m generator. To thiswas added 1.0 mL of a 10 mg/mL solution ofsodium thisulfate and 1.0 mL of IN HCI. All so-lutions were sterile and pyrogen free. The vial wasvented by inserting a 22-gauge needle and placedin boiling water bath for 3'/2 minutes, then cooledin running water for 3'/2 minutes. One mL of a1:10 dilution of human serum (HSA) was added tothe vial, followed by the addition of 2 mL of ace-tate buffer. The vial was then placed in boilingwater and agitated for about 30 seconds. Mac-roaggregates were obtained in the size of approx-imately 50 microns.

The 99Tc labeled macroaggregates were then in-jected into the veins of the hind legs of the dog.Images were recorded on an Anger camera. Uponclearing of the injected macroaggregates, the dogwas maintained in the same position and the ven-ogram was performed.

VenogramThe animal, still in the same position, was in-

jected with 50 mL (25 mL in each leg) of Reno-grafin-76. The injection was made into the samecatheters through which the isotope was intro-duced. The entire procedure was performed underfluoroscopic control with spot films taken of eachleg following the injection.

To create a lymphatic obstruction, most of theafferent vessels to the right inguinal lymph nodeswere surgically occluded by a standard laparot-omy.

Surgical Tying Off of LymphaticsA standard laparotomy approach was per-

formed with careful attention not to harm the vas-cular structures. The lymphatics stained withmethylene blue were then surgically tied. A stand-ard surgical closure of the wound was then ac-complished. Following surgery, gross measure-

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 78, NO. 9,1986 879

FALSE-POSITIVE THROMBOSCINTIGRAM

TABLE 1. STUDIES OF SURGICALLY INDUCEDLYMPHATIC OBSTRUCTIONS IN DOGS

Lymphan- Thrombo-giogram scintigram Venogram

Baseline studies Normal Normal NormalPostoperative Abnormal Abnormal Normal3 Months on right on right

Postoperative Abnormal Abnormal Normal6 Months on right on right

Postoperative Abnormal Abnormal Normal1 Year* on right on right

*Animal model 1, right hind limb 3/4 in larger than leftlimb, while animal model 2 had right hind limb 1/2 inlarger than left limbNote: One of three canines died at 6 months

ments of the paw, ankle, and thigh were taken attwo-week intervals.

RESULTSImaging of the animal model one week after the

operation revealed an abnormal thromboscinti-gram and lymphangiogram. The venogram wasentirely normal. The thromboscintigram revealedan accumulation of the isotope in the venousdrainage of the right hind limb (Figure 3). Thelymphangiogram demonstrated obstruction of theEthiodol contrast by circumvention of the forwardflow in the right hind limb (Figure 4). On subse-quent studies incomplete drainage of the lymphat-ics was again demonstrated (Table 1).

Figure 3. The lymphatic drainage of a dog's right lower extremity has been ligated. A thromboscintigramimmediately after the injection of technetium-labeled aggregates in the dorsal veins of the feet of the dogshowed abnormal accumulations of isotope in the right lower extremity (left)

880 JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 78, NO. 9,1986

FALSE-POSITIVE THROMBOSCINTIGRAM

9

Figure 4. A lymphangiogram using contrast agentdemonstrates circumvention of contrast material tothe obturator lymphatics on the right lower extrem-ity (the surgical ligation of lymphatic drainage isindicated by the arrow)

The venous drainage was normal on thefollow-up venograms for over one year. The hindlimb on the right (operative side) measured largerthan the left side (Table 1). The entire experimentwas repeated with two other canines with the sameresults.

DISCUSSIONThe results demonstrate that obstructive

changes to the flow of lymph in the extremitiesmay render a false impression of a positive throm-boscintigram. The senior author has observed var-

ious stages of edema in patients presenting withswollen extremities. The subcutaneous tissues inthese patients were in various stages of thickeningand entrapment with scar tissue. The capillariesand lymph channels were often encased by thefibrous tissue within the extremity. The prolifera-tion of tissue resulting in swelling may be thecause of a false-positive thromboscintigram. Cer-tainly, other causes as mentioned contribute tofalse studies. The intention here was to test onlythe theory that edema may be a differentialpossibility.

Since the presentation of this paper, Henze etal8 used one of the surviving canines to test a newagent for lymphoscintigraphy. The 99Tc labeleddextran selectively absorbed by the lymphaticsdisplayed the obstruction as reported. It is there-fore recommended that in those patients present-ing with swelling of an extremity who may havelymph obstruction, a lymphangiogram, thrombo-scintigram, or lymphoscintigram be considered.

AcknowledgmentResearch for this paper was supported partially by

funds from the Committee on Research of the AcademicSenate, Los Angeles Division.

Literature Cited1. Webber MM, Bennett LR, Cragin M, Webb R Jr.

Thrombophlebitis-Demonstration by scintiscanning. Ra-diology 1969; 92:620-623.

2. Webber MM, Victery W, Cragin MD. Demonstrationof thrombophlebitis and endothelial damage by scinti-scanning. Radiology 1971; 100:93-97.

3. Duffy CG, D'Auria A, Brien TG, et al. New radioisotopetest for detection of deep venous thrombosis in the legs. BrMed J 1973; 1:712-714.

4. Webber MM, Pollak EW, Victery W. Thrombosis de-tection by radionuclide particle (MAA) entrapment. Corre-lation with fibrogen uptake and venography. Radiology1974; 11 1:645-650.

5. Steckle RJ, Furmanski S, Collins JD, Poe N. Radio-nuclide perfusion lymphangiography. A follow-up study. BrRadiol 1978; 51 :35-40.

6. Grollman JH Jr, Webber MM, Gomes A. Phlebog-raphy and radionuclide-clot localization in the lower ex-tremities. Radiol Clin North Am 1976; 14:371-385.

7. Rosenthall L, Greyson ND. Observations on the useof 99mTc albumin macroaggregates for detection ofthrombophlebitis. Radiology 1970; 94:413-416.

8. Henze E, Shelbert HR, ColUns JD, et al. Tc-99mlabeled dextran: A new radiopharmaceutical for radionu-clide lymphangiography. J NucI Med 1982; 3:823-829.

JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION, VOL. 78, NO. 9, 1986 881