Lower extremity DVT is a common disorder. Distribution of Acute Lower Extremity Deep Venous Thrombosis in Symptomatic and Asymptomatic Patients:

Transcription

1 Distribution of Acute Lower Extremity Deep Venous Thrombosis in Symptomatic and Asymptomatic Patients: Imaging Implications Steven C. Rose, MD, William J. Zwiebel, MD, Franklin J. Miller, MD The ability of noninvasive imaging modalities to di agnose lower extremity DVT depends, in part, on the anatomic location of the thrombus. To define the pattern of thrombus formation in symptomatic and asymptomatic high-risk patient populations, 172 con secutive lower extremity venograms were submitted to blinded, retrospective interpretation. Acute DVT was present in 59 venograms (34 symptomatic and 25 asymptomatic patients). Among symptomatic patients with acute DVT, 26 of 34 (76%) patients had an above-knee thrombus and only eight of 34 (24%) patients had a thrombus isolated to the calf. In comparison, only three of 25 (12%) asymptomatic patients with DVT had an above-knee thrombus and 22 of 25 (88%) patients had a thrombus isolated to the calf veins (most involving only one venous segment). Failure to examine the calf veins, particularly in asymptomatic patients, would result in missing at least half of patients with DVT. Alternatively, since all cases of iliac vein DVT extended into the femoropopliteal segment, failure to visualize the iliac veins is unlikely to miss patients with DVT. Our results suggest merit to routine examination of the deep femoral, anterior tibial, and particularly the soleal (but not the gastrocnemius) veins and also to use of an imaging technique to detect congenital duplications of the superficial femoral and popliteal veins. KEY woros: Deep venous thrombosis; Lower extremities; Noninvasive imaging modalities. Lower extremity DVT is a common disorder that may have serious consequences in the form of a pulmonary embolism or lower extremity chronic venous insufficiency. The distribu- ABBREVIATIONS DVI', Deep venous thrombosis Received july 23, 1993, from the Department of Radiology, Uni versity of Utah Medical Center, Salt Lake City, Utah, and the Veterans Administration Medical Center, Salt Lake City, Utah. Revised manuscript accepted for publication November 3, Address correspondence and reprint requests to Steven C. Rose, MD, Department of Radiology, Virginia Mason Clinic, P.O. Box 900, Seattle, WA tion of a thrombus in the deep venous system is important as it potentially influences the efficacy and techniques of imaging studies used to diagnose acute DVT. Older autopsy and venographic studies reported in the medical literature concerning the distribution of lower extremity DVT indicate the following: (1) acute DVT is more common in the deep calf veins than in the more proximal lower extremity deep veins, (2) the muscular veins of the calf may be the sole location of venous thrombosis, and (3) it is uncommon for an acute thrombus to be isolated to the popliteal, femoral, or iliac veins. t- 7 The existing reports of thrombus distribution are based principally on symptomatic patients. Little information is available concerning thrombus distribution in asymptomatic persons and how thrombus distribution may by the American Institute of Ultrasound in Medicine J Ultrasound Med 13: , /941$3.50

2 244 ACUTE LOWER EXTREMITY DEEP VENOUS THROMBOSIS J Ultrasound Med 13: , 1994 differ between these groups. In addition, existing reports provide only limited information concerning the isolation of a thrombus to specific segments of the deep venous system. To further address the issue of thrombus distribution in patients with lower extremity DVT, we tabulated in detail the location of thrombosis in patients studied with contrast venography. From this tabulation, and through correlation with the existing literature, we derived a clearer picture of thrombus distribution in general, the differences in thrombus patterns between symptomatic and asymptomatic patients, and the potential impact of thrombus location on imaging studies used to evaluate DVT. MATERIALS AND METHODS During a 26 month period, contrast venography was used to evaluate a total of 172 lower limbs (164 patients) for possible acute DVT. Lower extremity symptoms or signs possibly caused by DVT (lower extremity pain, swelling, or erythema) were present in 93limbs (87 patients). In the remaining 79limbs (77 patients), no symptoms existed, but the patients were at high risk for the development of DVT because they had had hip or knee replacement. These joint replacement patients were evaluated with venography as part of a study comparing prophylactic antithrombotic drug regimens. Of the symptomatic group of patients, 59 were men and 33 were women and the mean age was 54 years (range, years). Among the asymptomatic, high risk patients 1 55 were men and 22 were women and the mean age was 64 years (range, years). Contrast venograms were performed with the patient positioned supine on a radiographic fluoroscopy table tilted to approximately 45 to 60 degrees reverse Trendelenburg (head up). A total of 100 to 150 ml of contrast medium (Conray 43; Mallinckrodt, St. Louis, MO) was injected into a vein on the dorsum of the foot via a sheathed cannula (lnsyte; Deseret Medical, Sandy, UT). The adequacy of venous filling was monitored fluoroscopically. Three radiographs of the calf (lateral, anteroposterior, and 15 degree internal rotation) were obtained while tight tourniquets were applied at both the ankle and the knee. After release of the tourniquets, the three calf radiographs were repeated. Anteroposterior radiographs were obtained, centered over the knee, the thigh, and the hip, respectively. The final anteroposterior radiograph of the pelvis was obtained after the following maneuvers were conducted: (1) compression of the femoral veins, (2) sustained Valsalva maneuver, (3) return to a horizontal patient position, (4) elevation of the leg, and (5) release of femoral vein compression. After the venogram was completed, 1000 units of heparin was added to 250 ml of normalized saline solution and allowed to infuse via the foot venous access site. Each of the venograms was interpreted independently by two radiologists experienced with contrast venography. Discordant interpretations were infrequent and were resolved by consensus. Acute DVT was diagnosed by either of two findings: (1) the presence of an intraluminal defect seen on more than one view or (2) nonopacification of a venous segment that could not be attributed to the examination technique and was associated with either a "trailing edge" of a thrombus or small-caliber collateral veins. Chronic occlusive DVT was diagnosed by the presence of a nonopacified venous segment associated with large-caliber, serpiginous collateral veins but not with a filling defect or "trailing edge" of a thrombus. Venograms were considered negative for acute thrombosis under the following circumstances: (1) they were completely normal; (2) there were findings consistent with prior DVT (undulating course or irregular venous walls with an opacified central channel, large serpiginous collaterals, or both) but no changes suggestive of acute DVT. The location of a thrombus in each extremity was mapped with respect to the following venous segments; (1) common iliac vein, (2) external iliac vein, (3) common femoral vein, (4) deep femoral vein, (5) superficial femoral vein (a) proximal to the adductor canal and (b) within the adductor canal, (6) popliteal vein, (7) anterior tibial veins, (8) posterior tibial veins, (9) peroneal veins, (10) soleal sinuses, and (11) gastrocnemius veins. A thrombus was defined as isolated if it was entirely contained within only one of the above-named venous segments and as contiguous if it involved two or more segments in direct continuity (e.g., external iliac and femoral veins). A thrombus was defined as associated if it was present in two or more noncontiguous segments (e.g., soleal and anterior tibial veins). RESULTS One hundred seventy-two extremities (93 symptomatic and 79 asymptomatic) were available for venographic analysis. Fifty-nine extremities (34 symptomatic and 25 asymptomatic) exhibited acute thrombus. These 59 lower extremities were used to evaluate thrombus distribution. Among the 113 lower extremities without an acute thrombus, 99 were normal and 14 exhibited only chronic changes of distant episodes of thrombosis (11 from the symptomatic and three from the asymptomatic groups).

3 J Ultrasound Med 13:24~250, 1994 ROSE ET AL 245 Table 1: Distribution of a Thrombus* in 34 Symptomatic and 25 Asymptomatic Lower Extremities Thrombus Location Symptomatic Asymptomatic Iliac segment: Common iliac 3 0 External iliac 5 0 Femoral segment: Common femoral 6 0 Superficial femoral, 14 1 proximal Superficial femoral, 12 1 adductor canal Deep femoral 6 0 Popliteal segment 21 2 Calf Veins: Tibioperoneal trunk 19 1 Anterior tibial 15 5 Posterior tibial 23 6 Peroneal So leal Gastrocnemius 12 2 Jn most patients, DVT involved multiple segments. Regional Distribution of a Thrombus. The overall distribution of thrombus in the symptomatic and asymptomatic groups is shown in Table 1. The relative distribution of thrombus between the femoropopliteal and calf deep veins is given in Table 2. In the symptomatic group, an acute thrombus was present in the iliofemoropopliteal veins in 76% (26 of 34) of extremities with acute DVT. In contrast, an acute iliofemoropopliteal thrombus was present in only 12% (three of 25) of asymptomatic extremities with acute DVT. A thrombus was isolated to the calf veins in 24% (eight of 34) of symptomatic extremities with acute DVT and in 88% (22 of 25) of asymptomatic extremities with acute DVT. A thrombus was isolated to a single calf segment in 18% (six of 34) of symptomatic extremities with acute DVT and in 60% (15 of 25) of asymptomatic extremities with acute DVT. Table 2: Relative Incidence of Above-Knee and Calf Acute Thrombosis in Symptomatic and Asymptomatic Lower Extremities lliofemoropopliteal DVT CalfDVT With femoropopliteal DVT calf DVT to one calf segment Symptomatic Asymptomatic 26/34 (76%) 31/34 (91%) 23/34 (68%) 8/34 (24%) 6/34 (18%) 3/25 (12%) (100%) 3/25 (12%) 22/25 (88%) 15/25 (60%) Segment-by-Segment Distribution of a Thrombus The association of a thrombus in one venous segment with a thrombus elsewhere in the lower extremity is shown in Table 3. There were no cases of an isolated iliac vein thrombus. A thrombus occurred in the common or external iliac veins only in symptomatic extremities and always in continuity with a femoropopliteal thrombus. Femoropopliteal System In the femoral system (common, superficial proximal to the adductor canal, or deep femoral veins), 25 thrombi occurred in symptomatic extremities in continuity with popliteal segment thrombi. In one symptomatic extremity, a 2 em long thrombus was isolated to the deep femoral vein (Fig. 1). Only one proximal superficial femoral thrombus occurred in an asymptomatic extremity, and this was associated with calf DVT. The superficial femoral vein was duplicated in 18% (31 of 172) of all extremities studied (with or without a thrombus). In no case was a thrombus isolated to one member of a duplicated superficial femoral pair. Twelve thrombi occurred in the adductor segment of the superficial femoral vein in symptomatic extremities, and all were contiguous with a thrombus distally in the popliteal vein or more proximally in the superficial femoral vein. In one asymptomatic patient, a 0. 9 em long acute thrombus occurred in the adductor segment of the superficial femoral vein in association (but not in continuity) with an extensive calf vein thrombus. Twenty nonduplicated popliteal veins in the symptomatic cohort contained an acute thrombus. One additional symptomatic extremity contained a thrombus isolated to one branch of a duplicated popliteal system (Fig. 2). The overall incidence of popliteal duplication in all extremities (with and without acute DVT) was 36% (67 of 172). Infrapopliteal Conduit Veins In the anterior tibial system, a thrombus was contiguous or associated with a popliteal vein thrombus in 14 of 15 symptomatic extremities and in two of five asymptomatic extremities. A thrombus was isolated to the anterior tibial system in two asymptomatic extremities and was contiguous or associated with thrombi in other calf veins in the remainder. A posterior tibial vein thrombus was present in 23 symptomatic and 16 asymptomatic extremities. A thrombus in this segment was contiguous or associated with femoropopliteal DVT in 18 of 23 symptomatic extremities and two of 16 asymptomatic extrem-

4 246 ACUfE LOWER EXTREMITY DEEP VENOUS THROMBOSIS J Ultrasound Med 13: , 1994 Table 3: Relationship of a Thrombus in Individual Venous Segment with a Thrombus in Other Areas Location of Thrombus Iliac (common or external) Relationships Contiguous with femoral Associated with popl.iteal or calf Symptomatic Asymptomatic Femoral (common, deep; or proximal superficial) Superficioll femoral in adductor canal Popliteal Anterior tibial Posterior Tibial Peroneal Soleal Gastrocnemius Associated with popliteal or calf Contiguous with femoral or distal popliteal Associated with calf Contiguous proximal or d istal Associated proximal or distal Contiguous with poplileal Contiguous or associated with calf only Contiguous or associated with calf only Contiguous or associated with calf Isola led Contiguous or associated with calf conduit only Contiguous or associated with calf conduit 1 (deep) (duplicated)

5 J Ultrasound Med 13:24~250, ROSE ET AL Figure 1 deep femoral vein DVT. Anteroposterior view of a thigh contrast venogram in a paraplegic patient with lower extremity swelling and fever demonstrates a 2 em long nonocclusive thrombus (arrow). ities. An isolated posterior tibial thrombus occurred in one symptomatic and one asymptomatic extremity. A thrombus occurred in the peroneal veins in 23 symptomatic and in 14 asymptomatic extremities. A peroneal thrombus was contiguous or associated with a femoropopliteal thrombus in 18 symptomatic extremities but in only two asymptomatic extremities. A thrombus was isolated to the peroneal veins in three symptomatic extremities and in nine asymptomatic extremities. Figure 2 DVT isolated to one member of a duplicated popliteal vein. Anteroposterior view, knee contrast venogram in a patient with lower extremity pain, swelling, erythema, and warmth and a past history of multiple episodes of prior DVT. A 2.5 em near-occlusive acute thrombus (arrows) is present in the medial member of a duplicated popliteal vein. The finding of irregular wall margins is consistent with recanalization of prior episodes of DVT. Muscular Calf Veins The soleal veins (sinuses) were thrombosed in 23 symptomatic and in 15 asymptomatic extremities. A soleal thrombus was contiguous or associated with a femoropopliteal thrombus in 16 symptomatic and in only three asymptomatic extremities. An isolated soleal thrombus occurred in three symptomatic and in seven asymptomatic extremities (Fig. 3). The gastrocnemius veins were thrombosed in 12 symptomatic extremities, always in continuity with a popliteal thrombus. In asymptomatic patients, a gastrocnemius thrombus occurred in two extremities, one in continuity with a popliteal thrombus and one in association with a thrombus elsewhere in the calf. There were no cases of an isolated gastrocnemius vein thrombosis. DISCUSSION Regional Distributio11 of a Thrombus. The results of our study are in agreement with preceding work concerning the distribution of lower extremity acute DVT.I -7 First, as shown in Table 2, a thrombus was more common in the calf veins than in the femoropopliteal segment or the iliac veins. Second, a thrombus was often restricted to the calf (Table 2). Third, a thrombus was only rarely isolated to the iliac, femoral, and popliteal regions (5%, three of 59 extremities, symptomatic and asymptomatic) but was commonly isolated to the deep veins of the calf (51%, 30 of 59 extremities, symptomatic and asymptomatic). The findings of this study emphasize the clinical

6 248 ACUTE LOWER EXTREMITY DEEP VENOUS THROMBOSIS J Ultrasound Med 13: , 1994 nation when clinical circumstances require unequivocal information about the status of the calf veins. This approach may be particularly warranted in asymptomatic patients, considering the prevalence of isolated calf DVT noted in this study. Alternatively, utilize repeated ultrasonographic surveillance of the popliteal veins might be used to detect proximal extension of an undetected calf vein thrombus. Iliac Vein Thrombosis Figure 3 sojeal vein (sinus) DVT. Lateral view, calf contrast venogram in an asymptomatic patient 5 days after total knee replacement. A 10.5 em near ~ occlusive thrombus (arrows) is present in a pair of soleal veins. importance of calf vein thrombosis in patients with acute DVT and are particularly compelling with respect to asymptomatic patients. A thrombus was iso lated to the calf in 88% of asymptomatic patients as compared with 24% of symptomatic patients (Table 2). The prevalence of calf vein thrombus in a given patient population may influence the choice of d iagnostic studies for acute DVT. The value of duplex sonography for detecting an acute calf vein thrombus is subject to question. Although Polak and associates!! and Yucel and coworkers9 were successful in examining the calf veins with ultrasonography in patients with suspected acute DVT, other authors report considerable difficulty with sonographic examination of calf veins. o-tj If the calf veins cannot be examined consistently with ultrasonography, then it may be argued that venography should be the initial exami- AU the iliac thrombi (common or external) in this series occurred in continuity with common femoral thrombi in keeping with the accepted premise that isolated iliac system thrombosis is uncommon. In an autopsy and venographic series encompassing 486 patients, only four instances of isolated mac vein thrombosis were reported. 2-4 This represents a prevalence of just under 1% in patients with acute DVT, which is not an insignificant figure considering that a thrombus in this location would likely be large and may embolize to the pulmonary circulation. An isolated iliac thrombus generally can be detected with contrast venography, but technical diligence often is required to satisfactorily opacify the iliac veins. The detection of an isolated iliac thrombus with ultrasonography is largely dependent upon Doppler assessment of venous flow characteristics at the common femoral level, since direct sonographic visualization of the iliac veins frequently is obscured by overlying bowel gas. A nonocclusive thrombus may not produce Doppler flow abnormalities at the common femoral level. Deep Femoral Vein Thrombus Only one isolated thrombus occurred in the femoral system in this series, and this was locauzed to the deep femoral vein in an asymptomatic patient. The prevalence of isolated deep femoral vein thrombosis has not been studied extensively, but it appears to be uncommon. Havigl found no isolated deep femoral thrombi in his autopsy series of 261 cadavers. The superior portion of the deep femoral vein typically is well visualized sonographically; therefore, detection of an isolated thrombus in this location would be likely. Deep femoral vein thrombus may be overlooked with venography, however, if the contrast material does not reflux into this part of the venous system. Common and Proximal Superficial Femoral Vein Thrombus A thrombus within the common and superficial femoral veins usually is extensive in symptomatic patients, as shown in our series. In the femoropopliteal

7 J Ultrasound Med 13: , 1994 ROSE ET AL 249 segment between the insertion of the lesser saphenous vein into the popliteal vein and the deep femoral vein into the common femoral vein, tributaries are small. Thus, occlusive thrombosis in one portion leads to stasis and often subsequent antegrade and retrograde propagation involving the entire segment. All six common femoral and 12 of 14 superficial femoral vein thrombi were continuous with popliteal thrombi. Because thrombosis in the femoral system generally is extensive, accurate diagnosis usually is not difficult with either contrast venography or sonography. Small, isolated thrombi can occur in the femoral veins,i4.is however. One asymptomatic patient in this series harbored a 3 em superficial femoral vein thrombus associated (but not contiguous) with calf vein thrombus. It is possible for small nonocclusive thrombi in the femoral system to be overlooked with duplex sonography. 16 Care should be taken, therefore, to ensure that the femoral veins are visualized in their entirety. No data exist in the literature with respect to the prevalence of a thrombus isolated to the adductor segment of the superficial femoral vein-a segment that may be difficult to visualize sonographically. In our series, an adductor thrombus was contiguous with a sonographically easily detected proximal (superficial femoral) or distal (popliteal) thrombus in 12 symptomatic extremities. In one asymptomatic extremity, however, a 0.9 em thrombus was localized to the adductor segment. An associated (but not contiguous) thrombus in the calf was easily detected in this patient, but the adductor segment thrombus was overlooked with color duplex flow imaging. Duplication of the superficial femoral vein is common, as seen in our data (18% of extremities) and in May's series (38%).1 7 Duplication of any venous segment might be a cause of false-negative imaging results if the thrombus that is present in one channel is overlooked because the other channel is patent. 18 In no case in this series was a thrombus isolated to one member of a duplicated superficial femoral vein pair, and it appears, therefore, that such an occurrence is uncommon. No corroborative data exist in the literature with respect to this assumption. Popliteal Vein Thrombus In 21 of 23 cases in this series, a thrombus in the popliteal vein was extensive and was contiguous with a femoral or calf vein thrombus. Such a thrombus is easily detected with venography or ultrasonographic imaging. A thrombus may be isolated to the popliteal system, however, as occurred in two cases in this series. A symptomatic patient had a thrombus isolated to one member of a duplicated popliteal vein pair. Duplication of the popliteal vein is common (36% of extremities in our series and 44% in May's series). J7 Most popliteal vein duplications result from failure of the posterior tibial and peroneal veins to coalesce within the calf. Instead, these veins extend well into the popliteal space before they unite. The detection of a thrombus in a duplicated popliteal system generally does not pose a problem for contrast venography so long as both the peroneal and posterior tibial systems are well opacified. The risk of overlooking an isolated thrombus in a duplicated popliteal system probably is greater during sonographic examination, particularly if the calf veins cannot be examined in detail. Anterior Tibial Vein Thrombus Based on the belief that an isolated thrombosis of the anterior tibial veins is extremely uncommon, it has sometimes been argued that routine examination of the anterior tibial veins is unnecessary if the remainder of the lower extremity deep veins are normal. In our population, two of five anterior tibial thrombi in asymptomatic extremities were isolated to this segment. This finding suggests that, in fact, it may be important to visualize the anterior tibial veins. An isolated anterior tibial thrombus appears to be more common in asymptomatic patients than in symptomatic persons. All 12 anterior tibial thrombi in symptomatic patients were associated or contiguous with a thrombus elsewhere in the venous system. Posterior Tibial aud Peroneal Vein Thrombus A thrombus was infrequently isolated to the posterior tibial veins but commonly was isolated to the peroneal veins (Table 3). An isolated peroneal thrombosis was particularly common in asymptomatic patients (nine of 14 cases with peroneal thrombosis) in keeping with the hypothesis that lower extremity DVT commonly originates in either the peroneal veins or the soleal veins (discussed laterp.u. 7 The predisposition of thrombosis in the peroneal veins may relate to the stasis of flow behind the cusps of the multiple valves or to propagation arising from the soleal sinuses that drain into the peroneal veins. The higher frequency of isolated peroneal vein versus posterior tibial vein thrombosis in asymptomatic patients is unexplained, as both share drainage from the soleal sinuses and have a large number of valves. Nevertheless, the results of this study emphasize the importance of scrutiny of the peroneal veins, particu larly in asymptomatic persons.

8 250 ACUTE LOWER EXTREMITY DEEP VENOUS THROMBOSIS J Ultrasound Med 13 : 24~250, 1994 Soleal Vein Thrombosis An additional important point about calf DVT revealed by this study is the prevalence of soleal sinus thrombosis in both symptomatic and asymptomatic patients. Soleal thrombosis was venographically identified in 67% (23 of 34) of symptomatic extremities with DVT and in 60% (15 of 25) of asymptomatic extremities (a soleal thrombus may have been present but undetected in additional cases); a thrombus was isolated to the soleal veins in 9% (three of 34) of symptomatic extremities and in 28% (seven of 25) of asymptomatic extremities. The high frequency of soleal sinus thrombosis also was noted by Nicolaides and colleagues, 6 who found a 98% incidence (48 of 49) of soleal thrombus in symptomatic extremities with DVT and an 18% incidence of isolated soleal sinus thrombosis. The soleal veins (or sinuses) are, in essence, saccu lar veins embedded within the muscle that bears their name. The soleal veins are thought to be important in the pathogenesis of calf DVT owing to sluggish flow conditions within these vessels. The visualization of the soleal veins is difficult with both contrast venography and sonographic imaging. Our data suggest that in spite of such difficulty, these veins should be sought diligently in the course of venous imaging studies in both symptomatic and asymptomatic patients. Gastrocnemius Vein Thrombosis The final point that our study makes is that gastrocnemius thrombosis appears to be relatively unimportant from an imaging perspective. Although gastrocnemius thrombosis occurred in 35% (12 of 34) of symptomatic extremities and in 8% (two of 25) of asymptomatic extremities with DVT, such a thrombus invariably was contiguous with an easily detected popliteal vein thrombus. These data suggest that the gastrocnemius veins need not be examined with the same diligence that we have suggested for the soleal veins. We could find no corroborative data in the literature regarding the prevalence of isolated gastrocnemius thrombosis. The gastrocnemius veins drain into the popliteal vein. They are most prominent in the medial head of the gastrocnemius muscle and are multiple and small in caliber. Because of their small size and multiplicity, they may be difficult to identify sonographically. REFERENCES 1. Sevitt S, Gallagher N: Venous thrombosis and pulmonary embolism. A clinico-pathological study in injured and burned patients. Br J Surg 48:475, McLachlin j, Paterson jc: Some basic observations on venous thrombosis and pulmonary embolism. Surg Gynecol Obstet 93:1, Havig 0: Deep vein thrombosis and pulmonary embolism: An autopsy study with multiple regression anal ysis of possible risk factors. Acta Chir Scand 478(Suppl 1):1, Nylander G, Olivecrona H: The phlebographic pattern of acute leg thrombosis within a defined urban population. Acta Chir Scand 142:505, Cotton LT, Clark C: Anatomic localization of venous thrombosis. Ann R Call Surg Engl 36:214, Nicolaides AN, Kakkar VV, Field ES, et al: The origin of deep vein thrombosis: A venographic study. Br J Radial 44:653, Kakkar VV, Howe CT, Flanc C, et al: Natural history of postoperative deep--vein thrombosis. Lancet 2:230, Polak JF, Colter 55, O'Leary DH: Deep veins of the calf: Assessment with color duplex flow imaging. Radiology 171 :481, Yucel EK, Fisher JS, Egglin TK, et al: calf venous thrombosis: Diagnosis with compression US. Radiology 179:443, Lensing AWA, Prandoni P, Brandjes D, et al: Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med 320:342, Rose SC, Zwiebel WJ, Nelson BD, et al; Symptomatic lower extremity deep venous thrombosis: Accuracy, limitations, and role of color duplex flow imaging in diagnosis. Radiology 175:639, Elliott CG, Suchyta M, Rose SC, et al: Duplex ultrasonography for the detection of deep vein thrombi after total hip or knee arthroplasty. Angiology 44:26, Rose SC, Murdock LE, Zwiebel WJ, et a): Color Doppler flow imaging for detection of acute calf deep venous thrombosis in asymptomatic postoperative patients. JVIR 4:111, Ginsberg JS, Caco CC, Brill-Edwards PA, et al: Venous thrombosis in patients who have undergone major hip or knee surgery: Detection with compression US and impedance plethysmography. Radiology 181:651, 1991 IS. Stamatakis JD, Kakkar VV, Lawrence D, et al: The origin of thrombi of the deep veins of the lower limb: A venographic study. Br J Surg 65:449, Barnes RW, Nix ML, Barnes CL, et al: Perioperative asymptomatic venous thrombosis; Role of duplex scanning versus venography. j Vase Surg 9;251, May R; Surgery of the Veins of the Leg and Pelvis. Philadelphia, WB Saunders, 1979, p Quinn KL, Vandeman FN: Thrombosis of a duplicated femoral vein: Potential error in compression ultrasound diagnosis of lower extremity deep venous thrombosis. J Ultrasound Med 9:235, 1990