RADIONUCLIDE VENOGRAPHY (RNV) IN LOWER EXTREMITY VENOUS DISEASE

Robert E. Henkin, James S. T. Yao, James 1. Quinn Ill, and John J. Bergan Northwestern University McGaw Medical Center, Chicago, illinois

Radionuclide venography (RNV) of the pel Those thrombi which occur above the calf are via and lower extremity veins using ttmTc-albu. most likely to embolize (1 ,3) . There is indirect sup mm microspheres was performed in 53 patients port for this conclusion in a recent study of 47 pa with pulmonary emboli or venous disease. In tients with laboratory evidence of calf vein throm 25 patients who underwent both contrast and bosis, none of whom had evidence of emboli (4). RNV studies, there were no false-positive studies These observations led us to evaluate radionuclide and one false-negative. In this series the accu venography (RNV) as a diagnostic modality for racy of RNV exceeded that of Doppler alone. detection of lower extremity venous disease. When the results of these two methods were Previous studies by Webber with DDmTc@macroag@ combined, false-negative studies were elimi gregated albumin (MAA) demonstrated the ability nated. Over half the patients with pulmonary of this agent to localize thrombi when static images emboli and clinically normal lower extremities were made after injection into pedal veins (5) . How had abnormal RNV studies. The exact method ever, false-positives were reported. of thrombus localization is not known but both Rosenthall investigated both static and dynamic mechanical and electrostatic factors may be in imaging of lower extremity veins using 9omTc@MAA. volved. RNV has been shown to be a safe, accu The results of the dynamic studies were encouraging rate, and easily performed examination which but there were no consistent findings on static imag is particularly useful in evaluating patients with ing (6,7). pulmonary emboli. We felt that 9omTchuman albumin microspheres (HAM) * might prove useful in both static and dynamic venous imaging and evaluated this agent Improvements in diagnostic techniques have dem in the diagnosis of venous thrombosis and occlusion onstrated pulmonary embolism to be more frequent above the calf. than previously suspected. This disease accounts for MATERIALS AND METHODS many deaths in the postoperative period, is the sin Patients were referred for RNV from the medical gle greatest factor in mortality in childbirth, and has and surgical services of Northwestern Memorial Hos a high incidence in protracted medical illnesses (1). pital, most commonly for symptoms of lower cx The source of pulmonary emboli is no mystery. tremity venous disease or pulmonary emboli. All Autopsy studies have shown 50—80% of hospitalized patients were evaluated clinically prior to radionu patients have lower extremity venous thrombi. Three clide venography by one of the investigators (JSY fourths of these thrombi were not evident clinically or JiB ) . The majority of patients (50 of 53 ) were during life (2) . Since so many thrombi are silent, studied by Doppler ultrasound using a Directional other methods of diagnosis had to be developed. Doppler unit Model 806t as described by Yao (8). Those most commonly employed are contrast venog Received July 13, 1973; revision accepted Oct. 15, 1973. raphy, Doppler ultrasound, and the radioiodinated For reprints contact: Robert E. Henkin, Dept. of Nuclear Medicine, Northwestern Memorial Hospital, Superior Street fibrinogen uptake. A combination of technical and & Fairbanks Court, Chicago, Ill. 60611. diagnostic limitations make the routine performance a 3M Co. of these tests uncommon in a genera! hospital. t ParksElectronics.

Volume 15, Number 3 171 @ .. . @2-@

HENKIN, YAO, QUINN, AND BERGAN

Patients were positioned under the camera so that the iliac veins and lower inferior vena cava were in the field of view. These veins were imaged by injec tion of 1—1.5mCi of DomTc@HAMinto each leg @ 4@_ I simultaneously followed by a 3—5-cc heparinized saline flush. If one iliac vein had abnormal transit (early or delayed filling) that leg was studied first as outlined below. If both iliac veins filled normally, the left femoral vein in the upper and mid-thigh was imaged next by unilateral injection of 0.5—0.75 mCi of oomTc@HAM followed by heparinized saline flush. The left pop liteal vein was similarly evaluated after placing a 5TCo marker lateral to the patella. The calf veins

A B

:‘

@,

*@ . *

FIG. 1. Contrast(A)andradionuclid.studies(B)of pelvic veins in patient with left iliofemoral occlusionof 6-months dura tion. Arrows identify corresponding collateral veins on both studies.

Radionuclide venography was performed using a Nuclear-Chicago Pho/Gamma III or Pho/Gamma HP scintillation camera equipped with either a 4,000- hole or 16,000-hole (high-sensitivity) collimator. Microspheres were prepared from kits, according to directions, so that the final radioactive concentra tion was approximately 4 mCi/mi. FIG.2. (A)Radionuclidevenogramof rightiliacveinin pa Patients were placed supine on a cart with a float tient with acute right lower extremity edema. RNV was interpreted ing top. Two tourniquets 1—2in. apart were placed as extrinsic pressure (arrow) and incomplete obstruction. (B) Re peat RNV 6 days later demonstrates development of collateral 2 in. above each ankle. Butterfly needles (19—25 circulation. Original area of extrinsic pressure is still present. gage) with attached three-way stopcocks were in (C) Contrast venogram immediately following B demonstrates soft tissue mass compressingright iliac vein (arrows). Biopsy of mass serted into a dorsal vein of each foot. yielded diagnosis of reticulum cell sarcoma.

172 JOURNAL OF NUCLEAR MEDICINE RADIONUCLIDE VENOORAPHY

major reasons for referral, those patients with clini cal phlebothrombosis, as expected, had a majority of positive scans (25 of 31) (Table 3) . However, 53% of patients (9 of 17) with pulmonary emboli of unknown origin (no clinical signs of venous dis ease) had positive RNVs (Table 4). Positive agreement of Doppler ultrasound and RNV was seen in 39 of 50 cases (78%) (Table 5). There was one false-negative RNV and 6 false negative Doppler studies. Four patients had insuf ficient followup data to confirm definitely venous disease. A tabulation of the abnormalities noted on RNV FIG.3. (A)Contrastvenogramdemonstratingincompleteob struction and thrombus formation (arrow) in left femoral vein. by side and site was made (Table 6). (B) Rt'IV dynamic study of same area delin.ates more clearly the There were I6 sites at which hot spots were iden collateral veins. (C) Delayed RNV image demonstrates hot spot (Arrow) at site of thrombus. tified in the group of patients who underwent both

above the tourniquets were imaged in like manner. A similar set of images was then obtained for the TABLE 1. TOTAL RNV RESULTS right leg. Diagnosis Patients

The elapsed time from injection to the first appear Positive ance of activity in the field of view was recorded. 12Equivocal6Total53Negative35 Images were recorded simultaneously on either Pola roid or 70-mm photographic film and digital video tape. At the conclusion of the study, a four-view camera lung scan was performed. If contrast venography was to be performed immediately, the needles were TABLE 2. PATIENTS EVALUATED WITH RNV left in place with a heparin lock and the patient sent STUDIESDiagnosisAND CONTRAST to the x-ray department. ContrastPositive RNV

Contrast venography was performed by simultane 18 19 ous and sequential injections of Renografin-60 via Negative 6 6 needles placed at RNV or needles placed in a similar 0TotalEquivocal 1 manner. Injections were made with the patient tilted 25TABLE 25 at a 30—45-degangle, feet down. Tourniquets were not employed. This method is essentially the same as that described by Thomas (9). PATIENTS REFERRED FOR SYMPTOMS RESULTS OF3.LOWERDISEASERNVEXTREMITYVENOUS All venograms were classified at the time of inter PatientsPositivediagnosis pretations into one of the following categories: (A) 25 normal; (B) suspicious (equivocal)—increased tran Negative 2 sit time, nonspecific flow abnormalities, or deep 4TotalEquivocal venous “hotspots―which faded by five mm post 31 injection; (C) abnormal—deep venous occlusion with or without collaterals or “hotspots―seen at injection and persisting more than 5 mm postinjection. Calf vein images were not evaluated diagnostically for TABLE4. PATIENTSREFERREDFOREVALUATION OF PULMONARYEMBOLIOF UNKNOWNORIGIN this study. Fifty-three patients were evaluated by RNV using RNV diagnosis Patients the above classifications (Table 1) . Twenty-five of Positive 9 these patients were further evaluated by contrast Negative 6 venography. There was complete agreement in 24 Equivocal 2 or 25 studies (96%) (Table 2). Total 17 When patient data were analyzed according to the

Volume 15, Number 3 173 HENKIN, YAO, QUINN, AND BERGAN

Doppler ultrasonography, a simple bedside test for TABLEBYDOPPLER5. PATIENTSSTUDIED venous occlusion, failed to identify a case of partial ANDRNV venous occlusion as well as several cases of chronic Final diagnosis Patients venous disease. The precise anatomical location of False-positiveDoppler0False-positive thrombi by this method was not as accurately deter RNV0False-negative mined as by RNV. Doppler6False-negative Because of its low energy, 251-fibnnogen is in RNV1Agreement andRNV39Insufficientbetween Doppler capable of imaging the deep venous structures of diagnosis4Total50data for the leg and only point counting can be employed. A disadvantage of this technique is the 24-hr delay before definitive diagnostic results can be obtained. The greatest accuracy of the fibrinogen test appears to be in diagnosing calf vein thrombosis. As one TABLE6.OFPOSITIVESITESATRNVTABULATIONOFLOCATION moves more proximally in the leg, the accuracy de creases due to increased background (10). No.No.posi Radionuclide venography fills the gap between the posi less specific noninvasive tests, radioiodinated fibrin Side tiveSite tive ogen uptake, Doppler ultrasound, and the more ac only7IVC5LeftRight curate but invasive technique of contrast venogra only15Iflocs12Bilateral9Femorals18IVC4Popliteals15Total— phy. The data presented above demonstrate RNV to have an accuracy comparable to contrast studies visualized deep (96% agreement) and to exceed the accuracy of 35No circulation6Totalvenous Doppler studies alone. sites56 The RNV examination requires 30 mm or less to perform. A knowledge of the deep venous anatomy of the legs and pelvis is essential for accurate inter pretation of the images. Since only major vessels are TABLE7. RELATION HOT visualized and interpretation of calf pathology is not TO OCCLUSIONINOFPROVENSPOTSCASES routinely made at present, skill in recognizing the No.withouthotsiteswithNo. sites diagnostic patterns is achieved quickly. spotshot spots The radiopharmaceutical of choice for these stud Total occlusion ies should fulfill several requirements. First, it should 0Normals06Incomplete occlusion5 38 not recirculate. Recirculation from the necessary mul tiple injections increases background, making inter pretation of the study difficult. Secondly, the radio contrast and radionuclide venography. These results pharmaceutical should be readily available and easily are tabulated for patients with documented, partial, prepared. Finally, it should be a suitable agent for and complete deep venous occlusion (Table 7). lung imaging in view of the possibility of emboli in this group of patients. DISCUSSION Two radiopharmaceuticals satisfy these criteria: The need for a simple, low-risk, easily interpretable 9DmTcMAA and OØmTC..HAMEarlier studies with test for lower extremity venous occlusive disease is MAA cited above suggest that HAM might be a evident. Of the currently available diagnostic tests, more suitable agent for the following reason : if only contrast venography can be relied upon for mechanical factors alone were responsible for throm accurate localization of proximal venous disease. The bus localization with MAA, then it would be diffi currently available techniques have shortcomings. cult to explain the incidence of false-positives pre Contrast venography can lead to venous damage viously noted (5). if contrast media is not adequately flushed out after An electrostatic attraction (or repulsion) may oc the study. The risk of hypersensitivity reaction in cur between clot, vein wall, and radiopharmaceutical herent in the intravenous injection of any contrast as suggested by Webber (1 1 ) . Radiopharmaceuticals medium accompanies this procedure as well. In our with different shapes should have different surface experience, patient acceptance of contrast venogra charge distributions. Webber's experimental data fur phy compared with RNV is quite low. It may there ther demonstrate decreased in vivo thrombus uptake fore be quite difficult to convince patients to undergo of HAM as compared with MAA. We elected to in necessary, repeat contrast studies. vestigate HAM in the hope of a decreased incidence

174 JOURNAL OF NUCLEAR MEDICINE RADIONUCLIDE VENOGRAPHY of false-positive studies. This indeed has been the RNVs place in the evaluation of patients with case. venous occlusive disease is still being defined. We The accumulation of microspheres at the site of currently advocate the use of this technique in pa a thrombosis appears from our data to be somewhat tients with pulmonary emboli and in patients who variable. These data suggest that the stage in the evo are clinically suspected of having lower extremity lution at which a thrombosis is studied may affect venous disease. microsphere localization. In summary, while the precise mechanism of Those patients with incomplete obstruction were thrombus localization is not known, vascular occlu symptomatic and studied during the acute phase sion, collateral circulation, and thrombi are accu whereas those patients with total occlusion were often rately identified with minimal discomfort and risk studied after thrombosis was established. to the patient by radionuclide venography. One of our criteria for diagnosis of venous dis ease has been the presence of collateral circulation. ACKNOWLEDGMENT We have been particularly impressed with the vis ualization of collateral venous channels in the pelvis. This work was supported in part by grant-in-aid #C72-72 from the Chicago Heart Association. As a rule visualization is better on the radionuclide than the contrast study. The explanation of this observation appears to be REFERENCES twofold. The long imaging time of RNV results in 1. KAKKAR VV, HOWE CT, NIcoLAuDns AN, et al: Deep an integrated image being obtained. This image is vein thrombosis of the leg—is there a “highrisk―group? essentially a composite instead of representations of AmlSurg 120:527—530,1970 contrast distribution at one or several instants in 2. SECKER-WALKERR, POTCHEN EL : Radiology of venous thrombosis—current status. Radiology 101 : 449—452,1971 time conventionally obtained at contrast venography. 3. Kxsu@xit VV, NIcoi@mEs AN, RENNEY JTG, et a): Secondly, it is not necessary to penetrate the heavy 1-125 labelled fibrinogen test adapted for routine screen bony pelvis during RNV. The presence of the bony ing for deep-vein thrombosis. Lancet I : 540—542,1970 pelvis decreases the image contrast and, therefore, 4. Waxy R, MAURERB, SHILLINOFORDJ: Prophylactic image quality on x-ray studies. anticoagulant therapy in the prevention of calf-vein throm RNV is of special value in those patients with pul bosis. N Engl I Med 288: 815—817,1973 monary emboli of undetermined origin. Fifty-three 5. WEBBER MM, RESNICK LH, VICTERY WK, et al: percent of the patients studied in this group had ab Thrombosis scanning: its reliability and usefulness. I Nucl normal RNVs. It is perhaps in this group of patients Med 13:476—477,1972 that RNV can have its greatest clinical utility by 6. ROSEN-mALLL: Combinedinferior venacavography, iliac venography, and lung imaging with @mTcalbumin identifying those patients in need of more extensive macroaggregates. Radiology 98: 623—626,1971 evaluation and therapy. 7. ROSENTHALLL, GREYSONND: Observationson the A followup lung scan is often performed in a pa use of “mTc albumin macroaggregates for detection of tient with known emboli. Injection of microspheres thrombophiebitis. Radiology 94: 413—416,1970 via the pedal veins at the time of followup scan pro 8. YAO ST, GxuaMos C, HOBBS JT: Detection of proxi vides additional important clinical information at ma) vein thrombosis by doppler ultrasound flow detection extremely low risk. method. Lancet I : 1—4,1972 Analysis of the transit time data is under way at 9. THOMASML: Phiebography.Arch Surg 104: 145— 151, 1972 present. These data are not easily obtained by any 10. BROWSE NL: The 1-125 fibrinogen uptake test. Arch other means. Preliminary observations indicate that Surg 104:160—163,1972 there is a group of patients whose sole abnormality is 11. WEBBER MM, VICTERY WK, CRAGIN MD: Demon an increase in transit time. The exact clinical signifi stration of thrombophlebitis and endothelial damage by win cance of this abnormality is not known at present. tiscanning. Radiology 100: 93—97,1971

Volume 15, Number 3 175