Long-term effects of superficial femoral vein ligation: Thirteen-year follow-up

Long-term effects of superficial femoral vein ligation: Thirteen-year follow-up Elna M. Masuda, MD, Robe r t L. Kistner , MD , and Eugene B. Ferris H I , MD , Honolulu, Hawaii

This study examines the late clinical, hemodynamic, and anatomic results of superficial femoral vein ligation performed in 35 extremities that were followed an average of 131/2 years (range, 5 to 22 years). Indications for interruption were to prevent recurrent embolization from distal deep venous thrombosis (14 cases), to prevent emboli in patients with contraindication to anticoagulants (eight cases), to prevent distal reflux in selected patients undergoing iliofemoral thrombectomy (11 cases), and to control reflux in failed venous reconstruction (two cases). Ligation was effective in the prevention of pulmonary emboli as indicated by no significant clinical events and 15 negative postligation ventilation-perfusion scans. Long-term clinical follow-up showed normal (class 0) or near-normal (class 1) extremities in 83%. Fourteen percent developed mild to moderate symptoms of pain or swelling but without ulceration (dass 2), and only one case (3%) had ulcerative sequelae (class 3). The only two findings that correlated with worse clinical outcome were the presence of an incompetent profunda femoris or an obstructed greater saphenous vein. Profunda femoris reflux was found in 60% (3/5) of patients with class 2 or 3 sequelae, which was significantly higher than the 14% (3/22) found in those patients with class 0 or I results (p < 0.05). Obstruction of the greater saphenous vein was found in 50% of those patients with class 2 or 3 results as opposed to 9% with class 0 or i results (p = 0.05). A large collateral vessel between the profunda femoris and the distal superficial femoral or popliteal vein was associated with poor long-term results. We could not identify a clear correlation between physiologic obstruction and the clinical state. These findings indicate that superficial femoral vein ligation is effective and can be used as an alternative treatment to prevent emboli arising from lower extremity deep venous thrombosis in cases of contraindication to or failure of anticoagulation. Superficial femoral vein obstruction appears to be well tolerated over the long-term when the profunda femoris is patent and competent or the greater saphenous vein is unobstructed. (J VAsc SLatG 1992;16:741-9.)

Superficial femoral vein ligation (SFVL) is a procedure that has been used through the years to prevent pulmonary emboli arising from the lower extremities. Previous investigators demonstrated its effectiveness in controlling embolil; however, the operation fell into disuse because of three problems. First, it was demonstrated that SFVL led to elevated venous foot pressures. 2 Second, despite studies showing minimal sequelae after ligation, others demonstrated alarmingly high rates o f ulceration

From the Department of Vasaalar Surgery, Straub Clinic and Hospital, Honolulu.

Supported by a grant from the Straub Pacific Health Foundation. Presented at the Fourth Annual Meeting of the American Venous

Forum, Coronado, Calif., Feb. 26-28, 1992. Reprint requests: Elna M. Masuda, MD, Straub Pacific Health

Foundation, 846 South Hotel St., Suite 303, Honolulu, HI 96813.

24/1/41202

and late symptoms, ss Finally, some investigators reported cases of recurrent emboli despite vein interruption, 6s and they became less enthusiastic in applying this procedure, although failures may have been attributed to inaccurate diagnostic testing before ligation. These controversial points have led many to believe that SFVL may be ineffective and may lead to clinically detrimental effects over the long-term.

Interest in the present study was stimulated by our overall observation that patients with ligation and division of the superficial femoral vein (SFV) did surprisingly well over the very long-term with minimal late sequelae. This presented a unique oppommity to investigate the late effects of SFV obstruction by the use o f various noninvasive tests including color-flow duplex scanning, phleborheography, air plethysmography, and arm/foot pressure differentials with reactive hyperemia.

741

742 Masuda, IGstner, and Ferris

Journal of VASCULAR

SURGERY

Table I. Extent and location of acute DVT before ligation

Iliofemoral + popliteal/tibial 11 Distal SFV, popliteal, tibial 12 Tibial veins only 10 TOTAL 33

PATIENTS AND METHODS

Between 1969 and 1986, 96 patients underwent ligation and division of the SFV at the Straub Clinic and Hospital. Because the primary intent of this study was to report the long-term effects of the SFVL procedure, only those patients with a minimum follow-up of 5 years were included. Twenty-six patients or 35 extremities with greater than 5-year follow-up form the basis of this report. Of the 70 patients not included in this study because of less than the minimum follow-up, nearly all had demonstrated good to excellent early clinical results. Of these, one half could not be evaluated because they had died, and the other half were presumed to be alive but had moved or could not be traced by telephone or mail.

Patient ages ranged from 36 to 81 years, with a mean of 55 years. There were 21 men and five women, with nine bilateral procedures.

Ligation was performed for the treatment of acute venous disease in 33 cases and for chronic venous disease in two cases. Specific indications for SFVL were to prevent recurrent pulmonary emboli in those with thromboemboli arising from the lower extremities (14 limbs), to prevent distal venous reflux in selected patients undergoing iliofemoral thrombectomy who had surgery for acute deep venous thrombosis (DVT) (11 limbs), to prevent emboli in those patients with contraindication to or failure of anticoagulation (eight limbs), or to control reflux in failed venous valve reconstruction (two limbs). In all cases diagnosis of acute DVT was confirmed by ascending venography, and chronic venous reflux was evaluated by descending venography. Patients with acute DVT were classified into one of three groups based on location and extent of thrombosis (Table I).

All subjects with acute DVT as the indication for surgery (24 patients, 33 limbs) underwent ventilation-perfusion (V/Q) scans before ligation. Postligation scans were obtained in 15 patients.

Postoperative evaluation of long-term results was assessed by clinical, anatomic, and physiologic testing. Clinical results were based on history and physical examination in all but one case; in this case information was obtained by telephone interview. Clinical outcome was graded on the basis of criteria

recommended by the joint councils of the Society for Vascular Surgery/International Society for Cardio- vascular Surgery in the Reporting Standards of Venous Diseaseg: class 0: Asymptomatic extremity; class 1: Mild symptoms of pain, swelling, and indurative skin changes; class 2: Moderate symptoms of pain, swelling, and indurative changes but no ulceration; and class 3: Any ulceration and/or severe symptoms of pain, swelling, and indurative skin changes.

Anatomic assessment was achieved by color-flow duplex scanning. Thirty extremities were evaluated by the QAD-1 angiodynograph (Quantum Medical Systems, Inc., Issaquah, Wash.) with a 5.0 MHz probe. Examination was performed in the 10- to 15-degree reverse Trendelenburg position with reflux elicited by the Valsalva maneuver at all levels. In 26 of 30 limbs, patency and competency were determined at six levels including the common femoral, SFV, popliteal, tibial, and greater saphenous veins (GSV) and the profunda femoris.

Physiologic assessment was achieved by one or more tests for hemodynamic obstruction in 32 extremities. Thirty-two limbs were examined by phleborheography, 15 by air plethysmography, and !1 by arm/foot pressure differentials with reactive hyperemia. Phleborheography was performed with the use of the technique described by Crardey.~° The patient was examined in the 10-degree reverse Trendelenburg position with the knees slightly flexed. Five cuffs were applied and placed around the thorax, thigh, proximal calf, midcalf, and ankle. The sixth cuffwas placed around the foot and was used to apply pressure to the foot at 100 mm Hg for 0.5 second at intervals of 0.5 second while the other cuffs recorded volume changes. Inflation of cuff 5 was used to examine venous emptying of the foot. The phleborheography was interpreted as abnormal and consistent with chronic obstruction if there was an increase in baseline and normal or visible respiratory waves.

The technique of air plethysmography has been presented in detail elsewhere. 11 Briefly, the patient was examined in the supine position, and the leg was elevated to 45 degrees to empty the veins. The patient was then placed in the standing position with weight on the opposite leg. Outflow fraction was used to determine the presence or absence of venous obstruction. Change in volume represented by the functional venous volume, ( W ) and venous filling time 90 {VFT 90) was used to estimate reflux. VFT is the time required to achieve 90% of filling. Venous filling index (VFI) was obtained from calculating VFI =

Volume 16 Number 5 November 1992 Superficial femoral vein ligation 743

% Limbs

100

80

60

40

20

0 GSV Patent

m Chss 0-1 Q Class 2-3

Fig. 1. Late clinical results with occlusion of both GSV and SFV.

GSV

90% VV/VFT 90. Ejection volume (EV) was calculated from the change in volume after one tiptoe movement. Residual volume (RV) was obtained after 10 tiptoe movements, and residual volume fraction (RVF) was calculated from RVF = RV/VV x 100. Ejection fraction (EF) was determined by EF = EV/VV x 100.

Arm/foot venous pressure differential with reactive hyperemic test was performed by means of the method described by Raju and Frederichs. 12 Venous pressures of the arm and foot were determined by separate venipunctures. Normal unobstructed limbs demonstrated pressure differentials of less than 4 mm Hg. Reactive hyperemic testing was used to elicit hemodynamic obstruction. After cuff occlusion at 300 mm Hg for 2 minutes, the cuff was released, and venous pressure changes were recorded. An increase in foot venous pressure after release of the cuffofless than 6 nun Hg indicated no significant hemodynamic obstruction. If the increase in foot pressure exceeded 6 mm Hg, this indicated significant obstruction under stress. Obstruction was graded as follows. Grade 1 represented a compensated state with normal differential at rest and normal reactive hyperemic pressure. Grade 2 demonstrated partial compensation with normal pressures at rest but more than 6 mm Hg rise in pressure with reactive hyperemia. Grade 3 showed partial compensation with more than a 4 mm Hg differential at rest and abnormal reactive hyperemia of more than 6 mm Hg. Grade 4 represented a fully decompensated state in which resting pressure was more than 4 mm Hg and

paradoxically, reactive hyperemic pressure was less than 6 mm Hg.

SFVL operative technique. The SFV was ligated and divided at the confluence of the profunda femoris and SFV. The division was performed flush with the common femoral vein to avoid creating a stump in which thrombus could develop and propagate prox- imally. At the completion of the operation, the two ends of the divided SFV were separated by 1 to 2 cm by the ligation and interruption process.

Statistical analysis. Fisher's exact test was used for 2 by 2 tables to determine whether the row variable was associated with the column variable.

RESULTS

Clinical results. During a mean follow-up of 131/2 years, normal or near-normal extremities were found in 83% (29 cases). This group consisted of 34% (12 cases) of class 0 or totally asymptomatic limbs and 49% (17 cases) of class 1 with mild symptoms of edema or discomfort. There were 14% or five cases of class 2 results consisting of moderate symptoms but no ulceration° Only one patient (3%) manifested late sequelae of ulceration. Most were not wearing elastic support at the conclusion of the study. In the class 0 or 1 groups, 83% (24/29) denied the use of any elastic support. This was in contrast to 50% (3/6) of the class 2 or 3 groups without elastic support.

Preligation V/Q scans were positive for pulmonary emboli in 15 of 24 patients. All patients with positive scans for pulmonary emboli before ligation

744 ~/lasuda, Kistner, and Ferris

lournal of VASCULAR

SURGERY

% Reflux

6C 60%

G CFV PFV* GSV TiMal

[] Class 0-I [ ] Class2-3 *p < 0.05

Fig. 2. Relationship of reflux at different levels and clinical outcome. CFV, Common femoral vein; PFV, profunda femoris.

underwent a second V/Q scan within 1 month after ligation. All 15 postligation scans showed no new events or resolution of previous emboli. No clinical cases of pulmonary emboli occurred after surgery.

Recurrent DVT after ligation occurred in 11.4% (four cases) during the follow-up period. There were three cases ofiliofemoral thrombosis and one case of thrombosis involving all major veins below the inguinal ligament. Superficial phlebitis of the GSV occurred in one patient.

The extent of DVT before ligation did not significantly influence clinical outcome. Although there was a trend toward worse results in those patients with massive DVT involvement of the iliofemoral veins as opposed to those with isolated calf vein DVT, this difference was not statistically significant.

Duplex imaging results. Duplex scanning was used to assess venous patency and valve competency and patterns of collateralization. The length of residual SFV obstruction and the degree of recanalization did not affect clinical results. Obstruction of both the GSV and SFV were found in six cases. Five patients had had ligations of the GSV before SFVL, and the remaining patient developed thrombosis of the GSV several years after femoral vein ligation. Clinical results and patency of GSV are shown in Fig, 1. A significantly higher proportion of better dinical results, class 0 or 1, was found in those patients with a patent GSV (91%) as opposed to those with GSV obstruction (50%). This difference was statistically significant (p = 0.05). The significance of both deep vein and SFV obstruction could not be sufficiently

analyzed because only two cases of profunda femoris obstruction occurred.

Segment-by-segment analysis comparing the level of reflux and the clinical state is shown in Fig. 2. The only segment that correlated with clinical outcome was reflux of the profunda femoris. Deep femoral vein incompetence was found in 60% (3/5) of those patients in classes 2 and 3 as opposed to 14% (3/22) of those patients in classes 0 or 1. This difference was statistically significant (p < 0.05). In all cases of profunda femoris reflux with class 2 or 3 sequelae, a large incompetent collateral vessel was identified between the profunda femoris and distal SFV or popliteal veins.

We identified six patterns of collateralization or recanalization of the SFV (Fig. 3)i Collateral vessels were identified in 70% (21/30) of cases. Of these, 57% were incompetent. The finding of a large incompetent profunda femoris and SFV or profunda femoris and popliteal collateral vessel correlated strongly with a worse clinical outcome. Of those with a profunda femoris collateral vessel, 75% had moderate to severe symptoms of class 2 or 3 clinical severity as opposed to only 4% of those with no profunda fern- otis collateral vessel (p = 0.004). No statistically significant correlation could be identified between number, size, or competency of collateral vessels and clinical outcome.

Hemodynamie assessment. Physiologic obstruction was indicated by a positive test in 81% (26/32) of phleborheographic studies, 67% (10/15) of air plethysmographic studies, and 36% (4/11) of arm/foot pressure differential studies. Obstruction was defined by criteria outlined in the Methods


A. B. C. N=8 N=9 N=4

Common femoral-~--" ~Great vein | \ saph. / \ve=n Protunda I ~ / femoral / I \ \ v e i n / / ~ \

O Obstr

Super. [~///] femoral -----~/~. vein v / ~

D. E. F. N=4 N=4 N=I

Fig. 3. Diagramatic representation of duplex scan findings demonstrating various patterns of collateral development or recanalization of chronic SFV obstruction. A, Total length of SFV obstructed with no large collateral vessels detected; B, proximal SFV obstruction with large collateral vessel bypassing obstruction; C, total obstruction SFV, with large venae comitantes; D, total obstruction SFV, with large deep femoral-SFV or deep femoral-popliteal vein collateral vessels; E, multiple small vessels or neovascularization bypassing level of interruption; and F, total occlusion of all major veins.

section. We could not identify a clear correlation between clinical severity and the absence or presence of physiologic obstruction indicated by these three tests. However, patients with worse clinical grades demonstrated abnormal parameters by air plethysmography, including elevated venous filling index and ejection fraction indicative of reflux and calf muscle pump dysfunction, respectively (Table II).

Hemodynamic obstruction correlated best with the extent o f DVT before ligation. Nearly all patients with iliofemoral or femoral-popliteal-tibial thrombosis showed obstruction by phleborheography or air plethysmography, whereas only approximately one

half of those with isolated tibia/vein DVT showed obstruction (Fig. 4). This relationship was not observed with the arm/fo0t pressure differential t e s t .

DISCUSSION

SFVL has been a topic of controversy since 1929, when Homans 13 first introduced the concept of ligation of the femoral vein to control emboli originating from the lower extremity. From their vast experience of more than 2000 cases, Donaldson et al.1 reported a 95% success rate in controlling emboli and indicated that the procedure was associated with

746 Masuda, Kistner, and Ferris

Journal of VASCULAR

SURGERY

Table II. Results of air plethysmography compared with clinical outcome

APG results

Limb VFI R VF EF No. Class (<2 ml/sec) (<35%) (>60%)

1 0 2.5* 40* 58* 2 0 1.16 22 59* 3 0 0.97 32 49* 4 0 1.75 47* 41- 5 1 1.86 21 73 6 1 1.81 13 85 7 1 1.68 21 63 8 1 6.1" 26 74 9 1 7.82* 40* 36*

10 1 5.6* 80* 21- 11 2 12.9" 54* 35* 12 3 8.1" 72* 28*

APG, Air plethysmography; VFI, venous filling index; RVF, residual volume fraction; EF, ejection fraction. *Abnormal.

very few long-term sequelae. In contrast, several investigators including Agrifoglio and E d w a r d s , 6

Mozes et al.7 and Adams and DeWeese 8 reported high rates of emboli of up to 26% despite venous interruption and observed late sequelae of venous insufficiency in as many as 54%. s

One major criticism of many of these early reports was the lack of adequate preoperative diagnostic testing. Before the 1970s, femoral vein ligations were not infrequently performed on the basis of clinical suspicion of pulmonary emboli or DVT. Because diagnosis was not always verified by objective testing, it was not surprising that emboli continued to occur despite vein ligation and division. A review of the literature (Table III) demonstrates some of the shortcomings of earlier reports- the lack of objective testing, differences in the incidence of late sequelae reported, and the lack of adequate description of late clinical effects.

This series presents the late effects of SFVL in patients who have undergone preoperative evaluation with ascending venography demonstrating DVT arising from the lower extremities. Results are in agreement with reports that have shown that SFV interruption is effective in controlling emboli arising from the lower extremities and causes minimum long-term sequelae. 4,14 In this series with the excep- tion of one patient, no cases of ulceration or severe stasis changes occurred at the end of the follow-up period that averaged 131/2 years. Eighty-three percent demonstrated either asymptomatic or near- asymptomatic extremities, and only 17% wore elastic stockings over the very long-term. In addition, no

clinically significant postiigation emboli occurred, and no new emboli were found in 15 postligation V/Q scans. Although results may have been skewed by the small proportion of patients with follow-up greater than 5 years represented by this study, the remaining patients with shorter follow-up likewise manifested few if any symptoms of chronic venous insufficiency, and this lends additional support to the benign course observed after ligation.

In 1948 Linton and Hard)" is warned surgeons of the importance of the profunda femoris in providing compensation when the SFV was ligated. Although we were unable to study the impact ofpatency of the profunda femoris on clinical results because of the lack of sufficient numbers, in dais series, patients with competent profunda femoris did significantly better and achieved a class 0 or 1 result in 90% of cases when compared with those with incompetent profianda femoris in which only 50% had a class 0 or 1 result. In addition, the presence of a large incompetent collateral vessel between the deep femoral vein and the distal SFV or proximal popliteal vein was strongly associated with poor clinical outcome, which was indicated by a statistically higher proportion of class 2 or 3 results. Erikkson and Almgren 16 pointed out the importance of deep femoral vein reflux when they analyzed results of venous reconstruction for chronic venous insufficiency and suggested that its functional state was of major importance to the venous hemodynamics of the limb.

The status of the GSV was also found to affect clinical outcome. Although the numbers are small, the data suggest that better results are achieved when the GSV is not obstructed.

Despite the fact that all patients were subjected to anatomic obstruction by virtue of the ligation procedure, not all patients showed evidence of physiologic obstruction. Physiologic obstruction was present in only 70% to 80% of patients with phleborheographs or air plethysmographs, and the lack of obstruction could not be attributed to the extent of collateral formation or the degree of recanalization. It is noteworthy that physiologic obstruction appeared to correlate best with the extent of disease or DVT before ligation. Nearly all patients with iliofemoral or femoral-popllteal-tibial thrombosis showed positive findings of physiologic obstruction, particularly by phleborheography, whereas physiologic obstruction was found in only one half of those with isolated DVT of the calf vein.

In this study we were unable to show a clear correlation between physiologic obstruction and seventy of symptoms. There does not appear to be


% Limb Obstruction

100

80

60

40

20

0

90%

(9/10) 80%

92%

(11/12)

Iliofemoral Distal SFV '5 popliteal/tibial popliteal/tibial

m PRG

N APG

D Arm/Foot Differential

60% 60%

(6/10) (3/5)

~ 3 3 %

Tibial Veins only

Fig. 4. Results of three tests for physiologic obstruction based on extent of DVT before ligation. PRG, Phleborheography; APG, air phlethysmography.

Table I lL Summary of literature-late sequelae after SFVL

% Late sequelae

Author Method of No. of Type not (yr of study) diagnosis patients Follow-up Swelling/pain Ulcers specified

Rob inson and Moser,~ (1954) Keisker and Bowers, 14 (1960) Szilagyi and Alsop, 19 (1949) Mozes et al., z (1964) Agrifoglio and Edwards, 6 (1961) Adams and DeWeese, 8 (1970) Young et al.,26 (1974) Louagie et al., ~ (1990)

Clinical 36 6.5 yr - - < 10 Clinical 86 2-10 yr 12 12 - Clinical 100 1-12 yr - - 27 Clinical 112 3 m o - 7 yr - 6.3 15.6 Clinical 21 5-10 yr 9.5 - - Venography 31 " U p to" 7.5 yr - - 54 Venography 20 2.9 yr (average) - - 45 Venography 73 3 yr (average) 12.5-25 - -

any single physiologic test of venous obstruction as yet that accurately reflects the clinical state. Killewich et al?7 demonstrated that maximum venous outflow did not correlate with the presence of postthrombotic symptoms; likewise, Raju and Fredericks 12 could not relate ulceration with physiologic grade of obstruction when they examined patients with arm/foot pressure differentials and reactive hyperemia. These reports suggested a closer relationship of ulceration with reflux rather than obstructioll. 12,17,1s This is consistent with our observation that a clear correlation cannot be found between physiologic obstruction and the clinical state and that reflux may be a more important determinant of outcome. All four air plethysmography parameters-venous filling index,

residual volume fraction, ejection fraction, and outflow fraction-were abnormal in moderate to se- verely symptomatic limbs; this suggests that other factors in addition to obstruction, such as venous competence and calf muscle pump function, are important to determine the clinical state in patients with obstruction of the femoral vein.

Several patterns of collateral vessels were identified by color-flow duplex scan. This included a large collateral vessel between the common femoral vein and SFV, multiple venae comitantes of various caliber, neovascularization or the development of numerous small vessels around the site of interruption, and a large incompetent communication between the profunda femoris and distal SFV or

748 _Masuda, IGstner, and Ferris

Journal of VASCULAR

SURGERY

proximal popliteal vein (Fig. 3). The importance of collateral vessels in SFV obstruction remains an enigma. Small collateral vessels can dilate to enor- mous caliber and be mistaken for the SFV or popliteal vein. 29 According to Mavor and Galloway, 2° small collateral vessels are normally valved, and when subject to obstruction and dilation, these valves may be rendered incompetent. In this series collateral vessels were detected in 70% of cases by color-flow duplex scanning. Of these, reflux was found in only 57%.

The only collateral vessel that showed any clinical significance was an incompetent collateral vessel between the profunda femoris and the SFV or popliteal vein. The finding of this collateral vessel correlated with the presence of ulceration or moderate to severe symptoms in 75% of cases. It is noteworthy that in all cases, this was associated with an incompetent profimda femoris vein. On the basis of an autopsy series of 61 normal extremities, Edwards and Robuck 21 noted that a large communication was normally found between the profunda femoris and distal SFV or popliteal vein in 10% of cases. In previous reports, 6'8 it was theorized that the profunda femoris collateral vessel provided a channel through which thromboemboli bypassed the inter- rupted SFV and led to pulmonary emboli despite ligation. In this series the SFVL has not been associated with postligation emboli and did not occur despite large profunda femoris collateral vessels.

Isolated popliteal and tibial reflux in patients with proximal vein ligation showed f ine relationship with the severity of venous insufficiency symptoms. Of those with popliteal reflux and proximal ligation, 47% had no symptoms or only minimal swelling, In contrast, Shull e t al. 22 fonnd that popliteal reflux was strongly associated with ulceration; these studies were performed in patients who were not subjected to ligation. The difference may be attributed to interruption of the SFV, which eliminates retrograde flow into a valveless, recanalized vein distally and may serve to protect the limb from the subsequent effect of chronic reflux,

Several investigators ls'2a have cited favorable clinical results after SFVL was used for the treatment of chronic venous insufficiency. Buxton and Coller 2a were the first to publish a series using SFVL for the treatment of ulceration in which six of 14 cases showed complete healing of ulcers at 1 year. Linton and Hardy is used a very aggressive approach and combined SFVL with removal of the saphenous system and reported excellent early results. However, the procedure was not widely used for the treatment of chronic venous insufficiency because, as men-

tioned previously, ligation of the SFV was associated with elevated ambulatory venous pressures. 2 Inter- pretation of these results are not clear because ambulatory venous pressures cannot be used to distinguish between reflux and obstruction. Further investigations will be needed to determine the role of vein ligation In the treatment of deep venous reflux.

Recently invesngators 24'2s have used the SFV as an arterial bypass conduit and have shown that when the SFV is defunctionalized, very minimal late sequelae ensue. Although similar to the present study of SFV interruption, the two are distinctly different. The series presented here focusses on venous obstruction limited to the SFV and examines the effect of obstruction on postthrombotic limbs. In contrast, use of the SFV for arterial bypass described by Schulman et al)4 and Schanzer et al.2s involve extremities without previous thrombosis: the SFV and frequently the popliteal vein are totally defunctionalized to the level of the tibial veins. Although these veins are without prior DVT and obstruction may be better tolerated, dissection and interruption of the popliteal or tibial veins may have different implications on the function of the extremity over the long-term when compared with results of this study.

For those interested in this operation, SFVL can be used as an alternative treatment to prevent recurrent pulmonary emboli from distal DVT or to prevent emboli in cases of contraindication to or failure of anticoagulation. When the source of the emboli is known to reside in the distal lower extremity and is confirmed by diagnostic studies, SFVL can be selected for treatment if the profunda femoris is patent and competent. The likelihood of emboli by way of a large collateral vessel from the proftmda femoris is possible but low. Other potential applications of SFVL include prophylactic or definitive treatment of chronic venous insufficiency; however, possible ben- efits await confirmation by future studies.

This serms shows that SFVL is a safe procedure with few serious long-term sequelae. On the one hand, ligation of the SFV protects those patients with massive DVT l i.e., iliofemoral-popliteal-tibial thrombosis) from the late effects of recanalization and reflux by way of the femoral vein; on the other hand, it does not impose a detrimental effect on those with isolated calf DVT over the long-term. It appears that the SFV is analogous to the superficial femoral artery in that obstruction is well tolerated and compensation may occur by way of a patent and competent profunda femoris or by an unobstructed GSV.


We express our grateful appreciation to Deborah Ann White for her assistance in manuscript preparation, Ann S. Harada for assistance with statistical analysis, and Darcy M. Kessler and Darren Y. Ogawa for their technical work.

REFERENCES 1. Donaldson GA, Linton RR, Rodkey GV. A twenty-year

survey of thromboembolism at the Massachusetts General Hospital, 1939-I959. N Engl ] Med 196i;265:208-14.

2. Decamp PT, Ward JA, Ochsner A. Ambulatory venous pressure studies in post-phlebitic and other disease states. Surgery I951;29:365-80.

3. Allen AW, Linton RR, Donaldson GA. Thrombosis and embolism. Ann Surg I943;i18:728-40. "

4. Robinson JR, Moyer CA. Comparison of late sequelae of common and superficial femoral vein ligations. Surgery 1954;35:690-7.

5. Louagie Y, Van Ruyssevek P, E1Hammouti F, et al. Ligarion of the superficial femoral vein in prevention of pulmonary embolism: an old fashion procedure? J CARDIOVASC 8URG 1990;31:416-23.

6. Agrifoglio G, Edwards EA. Venous stasis after ligation of femoral veins or inferior vena cava. JAM A 196I;178: 10I-7.

7. Mozes M, Adar R, Bogokowsky H, Agrnon M. Vein ligation in the treatment of pulmonary embolism. Surgery 1964;55: 621-9.

8. Adams JT, DeWeesc JA. Comparative evaluation of ligation and partial interruption of the femoral vein in the treatment of thromboebolic diseases. Ann Surg 1970;172:795-803.

9. Porter JM, Rutherford RB, Clagett GP, et al. Reporting standards in venous disease. J VASC SUV, G 1988;8: 172-81.

10. Cranley Jl. Phleborheography. In: Kempczinski RF, Yao JST, eds. Practical noninvasive vascular diagnosis. 2nd ed. Chicago: Year Book Medical Publishers, 1987:438-63.

11. Christopoulos DG, Nicolaides AN, Szendro G, Irvine AT, Bull M, Eastcott HHG. Air plethsmography and the effect of elastic compression on venous hemodynamics of the leg. J VASC SURG I987;5:i48-59.

12. Raju S, Fredericks R. Venous obstruction: an analysis of one hundred thirty-seven cases with hemodynamic, venographic, and clinical correlations. J VASC SURG 1991;14:305-13.

13. Homans J. Thrombosis of deep veins of lower legs, causing pulmonary embolism. N Engl J Meal 1934;21I:993-7.

14. Keisker HW, Bowers RF. Results obtained by superficial femoral vein ligation. Surgery 1960;47:224-9.

15. Linton RR, Hardy IB. Postthrombotic syndrome of the lower extremity, treatment by interruption of the superficial femoral vein and hgation and stripping of the long and short saphenous veins. Surgery 1948;24:452-68.

16. Eri -ksson I, Almgren B. Influence of the profunda femoris vein on venous hemodynamics of the limb. J VASC SURG 1986; 4:390-5. ,:,

17. KiUewich LA, Bedford GR, Beach KW, Strandness DE. Spontaneous lysis of deep venous thrombi: rate and outcome. J Vasc SURG 1989;9:89-97.

18. Killewich LA, Martin R, Cramer M, Beach KW, Strandness DE Jr. An objective assessment of the physiologic changes in the postthrombotic syndrome. Arch Surg 1985;120:424-6.

19. Szilagyi DE, Alsop JF. Early and late sequelae of therapeutic vein ligation for thrombosis of veins of lower limbs. Arch Surg 1949;59:633-66.

20. Mavor GE, Galloway JMD. Collaterals of the deep venous circulation of the lower limb. Surg Gynecol Obstet 1967; 125:561-71. :

21. Edwards EA, Robuck JD. Applied anatomy of the femoral vein and its tributaries. Surg Gynecol Obstet 1947;85:547- 57.

22. Shull KC, Nicolaides AN, Fernandes e Fernandes IF, et al. Significance of popliteal reflux in relation to ambulatory venous pressure and ulceration. Arch Surg 1979;i I4:i304-6.

23. Buxton RW, CoUer FA. Surgical treatment of long-standing deep phlebitis of the leg. Surgery 1945;18:663-9.

24. Schulman ML, Badhey MR, Yatco R. Superficial femoral- popliteal veins and reversed saphenous veins as primary femoropopliteal bypass grafts: a randomized comparative study. J VAse SURG 1987;6:1-i0.

25. Schaxzzer H, Chiang K, Mabrouk M, Pierce EC. Use of lower extremity deep veins as arterial substitutes: functional status of the donor leg. J VASC SURG 1991;14:624-7.

26. Young AE, Thomas ML, Browse NL. Comparison between sequelae of surgical and medical treatment of venous thromboembolism. Br Mcd I 1974;4:127-130.

Submitted April 14, 1992; accepted July 22, 1992.

Documents

Long-term effects of superficial femoral vein ligation: Thirteen-year follow-up