Role of Surgery in Iliofemoral Venous Thrombosis

Role of Surgery in Thrombosis·

~liofemoral

Venous

Robert B. Rutherford, M.D. t

There are 3 basic options in the management of iliofemoral venous thrombosis: anticoagulation, thrombolysis, and thrombectomy. While some authors strongly defend a single favored approach, selective application of each modality makes the most sense. This presentation will attempt to define and justify appropriate indications for thrombectomy. HISTORICAL BACKGROUND

Early experiences with thrombectomy for iliofemoral venous thrombosis' were followed by a wave of enthusiasm in the 1960s with reports of good initial patency and of absence in the majority of painful swelling and evidence of venous stasis. Typical was the report by Haller and Abrams" of 85% patency in those operated on within 10 days of onset and "normal" legs with no edema in 81% of survivors. Then reports emerged of high rates of rethrombosis, especially in those with phlegmasia cerulea dolens," and failure to prevent stasis sequelae because, in spite of patency, valvular competency had been destroyed. Typical of the reappraisals that swung the pendulum away from thrombectomy in the late 1960s and "70s was a 5-year follow-up analysis by Lansing and Davis" of the same patients originally reported by Haller and Abrams. They showed that 94% of those followed up had sufficient edema and stasis changes to require elastic stockings and leg elevation, and, in addition, all those with venograms had insufficient valves. Further, 20f3 postoperative deaths (in 34 patients) were from pulmonary embolism, and there was a 30% wound complication rate, an average transfusion of 1,000 ml, and a mean hospital stay of12 days. This report suffered from a potential selection bias, since follow-up was only 50%, and venographic documentation even less, and because of the likelihood that patients with the worst results were heavily represented. Nevertheless, such reports plus the emergence of a new therapeutic option, thrombolysis, relegated thrombectomy to something of historic interest only in most clinics. However, several subsequent developments now *From the University of Colorado Health Sciences Center, Denver. tProfessor of Surgery, Division of Vascular Surgery. Reprint requests: Dr. Rutherford, University of Colorado Health Science Center, Vascular Surgery, Box C312, 4200 East Ninth Avenue, Denver 80262

force us to reappraise the role of thrombectomy: minimizing the rethrombosis rate by extending postoperative anticoagulant therapy and employing a temporary AV fistula, intraoperative venography, and other technical refinements, the development of noninvasive tests for more objective comparison of results, and the much more selective application of thrombectomy. A more balanced view that thrombectomy now is a better operation that can achieve specific goals in carefully selected patients will be presented below. PROBLEMS IN EVALUATING PUBUSHED REPORTS

There have been a few prospective randomized trials of heparin vs thrombolysfs'" and anticoagulant therapy with and without thrombectomy," but all three options have not been so compared. For the most part, one is still forced to compare reports that differ significantly in patient selection, time between onset and treatment, technical differences (eg, technique of thrombectomy, use of adjunctive AV fistula) extensiveness of thrombosis (eg, proportion of phlegmasia alba dolens vs phlegmasia cerulea dolens), type and duration of postoperative anticoagulant therapy, duration of follow-up, and outcome criteria (particularly clinical vs physiologic testing vs venography). All of these are important variables, as illustrated in the following discussion of the natural history of iliofemoral venous thrombosis and its modification by thrombectomy. NATURAL HISTORY

Iliofemoral thrombosis is unique when compared with more distally located deep venous thrombosis, in that three quarters of cases occur on the left, possibly because of compression of the left iliac vein by the right iliac artery and by a higher rate of failure to recanalize than femoropopliteal or tibial-soleal venous thrombosis.v" When patency is restored by the recanalization process but damaged valves remain incompetent, the patient is at risk for developing the typical "stasis" changes of brawny edema, pigmentation, subcutaneous fibrosis, and cutaneous atrophy, which can ultimately lead to ulceration. These changes are caused by ambulatory venous hypertension in the lower leg or "gaiter" zone. 10 With time, these occur in the majority of conservatively treated patients." When most physiSurgery in iliofemoral venous Ttvombosis (Robert B. Rutherford)

Table I-Iliofemoral Venous Thromb0Bi8: Outcome

Criteria

Early morbidity Discomfort, swelling . Hospitalization, cost Compartment syndrome, gangrene Late postphlebitic sequelae Obstruction (venous claudication) Valve reflux (stasis dermatitis, ulcers) Pulmonary embolism Clinically signiJicant pulmonary embolism Recurrent thrombosis Rethrombosis rate Distal thrombosis

cians refer to the "postphlebitic" syndrome, they consider only these sequelae. If re-canalization fails and patency is not achieved, another set of very different consequences prevails. Sufficient collaterals usually develop to handle venous outflow at rest but not the increased How demands produced by exercise. These patients suffer from "venous claudication" and typically have swollen thighs and prominent collaterals, with varicosities, in the proximal limb. As Nicolaides and Yao12 documented with physiologic testing, some patients de" velop 1, some both, and some neither of these 2 clinically distinguishable consequences of ileofemoral venous thrombosis, depending on (1) restoration of proximal patency and (2) preservation of distal (popliteal) valvular competence. Clearly, therapeutic outcome must be judged on the basis of both types of complications and by objective criteria, not just by the need for elastic stockings and leg elevation or by the frequency of stasis ulcers. THERAPEUTIC GOALS AND OuTcOME CRITERIA

Ideally one would like to avoid pulmonary embolism, eliminate both the early morbidity and late sequelae of iliofemoral venous thrombosis, and minimize the risk of recurrent or progressive thrombosis in the involved limb. These overall goals can be further subdivided into many specific outcome measures (Table 1), the degree or duration of which can serve as valid criteria for comparing the effectiveness of alternative therapeutic approaches. Before considering these comparisons between different treatments, one additional perspective is offered, namely, the importance of the distinction between phlegmasia alba dolens and phlegmasia cerulea dolens. While this distinction may occasionally be difficult in intermediate cases, it is quite obvious at the two extremes and, more important, significantly affects outcome and surgical indications. Much has been made of the physiologic basis for the striking differences between a bland "milk leg" on the one hand, and a tensely swollen, painful cyanotic limb

on the other, It is unlikely that associated arterial "spasm" is important in the pathogenesis of these clinical syndromes or that true difference simply reflects the extensiveness of the venous thrombosis." Propagation proximally and distally and out into venous tributaries progressively obstructs the venous outflow and intensifies swelling. Ultimately, venous outflow obstruction leads to impeded arterial inflow Experimentally, femoral vein ligation produces a 25-40% decrease in arterial inflow: 14 Compartmental swelling further impedes venous outflow; steadily accelerating its rate of development and ultimately reducing arterial inflow to ischemic levels. Venous gangrene, though relatively rare, can occur in phlegmasia cerulea dolens without any demonstrable arterial occlusive disease or spasm. However, the more extensive the thrombosis, the less likely thrombectomy will achieve either patency or preserve valve function, so reduction in acute morbidity is the only practical therapeutic goal. On the other hand, with lesser degrees of thrombosis, where thrombectomy gives the best long-term results, the acute morbidity is mild, and there is no threat of venous gangrene. Here, amelioration of postphlebitic sequelae must be the primary goal of thrombectomy or of any competitive therapy. EFFECIlVENESS OF ILIOFEMORAL VENOUS THROMBECTOMY IN ACHIEVING SPECIFIC THERAPEUTIC GOALS

Ultimately the indications for interventional therapy will be determined by its effectiveness in different clinical settings. Ordinarily when the procedure carries a significant mortality, this dominates the risk:benefit analysis. Because thrombectomy can be performed using local anesthesia through a groin incision without opening the abdomen, the risk should be minimal and depend primarily on the seriousness of the underlying disease and the ability to avoid pulmonary embolism or loss of blood faster than replacement. While the mortality in Haller's series was 9%, with 2 of3 fatalities related to pulmonary embolism, the technical progress witnessed in the ensuing 2 decades and the more careful selection of patients have markedly reduced the mortality risk. Thus, the application of thrombectomy can be based primarily on its effectiveness, relative to competitive forms of therapy, in reducing either the early morbidity or late sequelae of iliofemoral venous thrombosis. EARLY MORBIDITY

Reduction in pain and swelling and length ofhospital stay were not significantly different between the thrombectomy and the conservatively treated groups in the Scandinavian prospective randomized trial reported by Plate et al. 7 In general, the additional cost of CHEST I 89 I 5 I MAY, 1988 I SUpplement

435S

operation, incisional discomfort, and the risk of wound hematoma in anticoagulated patients will more than offset any immediate benefit of thrombectomy except in 1 group of patients-those with phlegmasia cerulea dolens with tense painful swelling who are not promptly relieved by heparin and elevation or any patient threatened by compartmental compression or venous gangrene.

LATE

SEQUELAE

The long-term benefits of thrombectomy relate to its ability to achieve proximal patency and/or distal valve competence and are influenced in turn by initial technical success and avoidance of rethrombosis. Initial success in achieving patency is in turn influenced by treatment delay and attention to technical detail, especially intraoperative venographic control. The rethrombosis rate is improved by continuing anticoagulant therapy for a longer period postoperatively and temporarily using an AV fistula. Because of the variable effect of these factors in older reported series, late patency rates vary from 21 to 71% and valve competence from 0 to 29% with postphlebitic sequelae developing in 18-81%.7 It is, therefore, not surprising that confusion exists regarding the efficacy of thrombectomy. The effect of treatment delay is well estabIished,2.8.15.16 so that few thrombectomies are now performed beyond 7 days, and it is usually done within 2 or 3 days of onset. The rethrombosis rate was high in the 1960s and '70S.3•17 Postoperative heparin administration has been shown to improve early patency from 72 to 93% and late patency from 49 to 67%.8.18 Since endothelial repair after venous thrombectomy takes at least 4 weeks, 19 even longer protection than afforded by in-hospital heparin administration is required. This is the rationale for this temporary (4-6 weeks) use of an AV fistula. With this maneuver, early rethrombosis rates have been reduced below 20%, being 13% in the Scandanavian prospective trial. Late patencies of 60-83% with postphlebitic sequelae as low as 0-17% have been achieved." In the Scandanavian trial, complete iliofemoral patency without significant filling defects was documented phlebographically in 76% of thrombectomy patients at 6 months, compared with 35% in the conservatively treated group. Distally in the femoropopliteal segment, twice as many thrombectomy cases had patent valves (52 vs 26%), and valve reflux was demonstrated in only one fourth of cases conservatively treated (9 vs 37%). Distal occlusion was similar (37 vs 39%) in both groups, but in those with an open femoropopliteal segment, competent valves were demonstrated in 86% of operated patients vs 41% of those conservatively treated. In terms of postphlebitic sequelae, 42 vs 7% had "no complaints." In addition, cases of leg swelling, new varicose veins, and venous claudication were all fewer in the thrombectomy

group. In thrombectomy patients, the rise in venous pressure with exercise was lower than in those conservatively treated, and dilated pelvic collaterals were seen more often in the latter group (65 vs 24% of operated patients). The Scandanavian trial is heavily quoted in this review because it is recent, prospectively randomized, used important adjunctive measures, and objectively compared the thrombectomy and conservative treatment groups by both phlebography and physiologic testing. It establishes important advantages for current thrombectomy techniques when selectively applied. Further, simpler methods of closing the AV fistula have removed the major objection to its use. iD Thrombectomy has not been prospectively compared with thrombolytic therapy, although an earlier experience in the same Scandanavian hospitals reporting the prospective analysis of thrombectomy vs heparin indicated that thrombectomy gave better resuits. 11 Thrombolysis has also shown improved patency relative to heparin therapy'" but, like early thrombectomy series, has not shown better valve competency. A recent prospective study by Kakkar et alit showed no difference between the streptokinase and heparin ,treated cases when evaluated later by physiologic testing. Finally, systemic streptokinase carries a small but significant risk of serious, even fatal, complications. 5 •6. 23 SUMMARY

Thrombectomy has a limited but definite role in the clinical management of patients with iliofemoral venous thrombosis. It is best applied selectively to achieve specific goals in two different groups of patients at either end of the disease spectrum. In relatively active healthy young patients with phlegmasia alba dolens operated on relatively soon after thrombosis, better protection against the late postthrombotic sequelae can be achieved. Patients with malignancy, underlying coagulopathy, or those who are inactive or have a limited life expectancy due to age or concurrent disease should not be operated on for bland thrombosis. At the other extreme, when phlegmasia cerulea dolens causes painful tense swelling, increases compartmental or decreases ankle pressures, and threatens limb viability in spite of heparinization and leg elevation, thrombectomy should be performed. REFERENCES

1 Mahomer H, Castleberry ~ Coleman WOe Attempts to restore function in major veins which are the site of massive thrombosis. Ann Surg 1957; 146:510 2 Haller]A)r, Abrams BL. Use of thrombectomy in the treatment of acute iliofemoral venous thrombosis in forty-five patients. Ann Surg 1963; 158:561 3 Karp RB, Wylie EJ. Recurrent thrombosis after iliofemoral venous thrombectomy. Surg Forum 1966; 17:147 4 Lansing AM, Davis WM. Five-year follow-up study of iliofemoral Surgery In Iliofemoral YanousThrombosis (Robert B. Rutherford)

venous thrombectomy. Ann Surg 1968; 168:620 5 Elliot MS, Immelman EJ, Jeffery ~ et ale A comparative randomized trial of heparin versus streptokinase in the treatment of acute proximal venous thrombosis: an interim report of a prospective trial. Br J Surg 1979; 66:838 6 Arnesen H, Heilo A, Jakobsen E, et ale A prospective study of streptokinase and heparin in the treatment of deep vein thrombosis. Acta Med Scand 1978;203:457 7 Plate G, Einarsson E, Ohlin ~ et ale Thrombectomy with temporary arteriovenous fistula: the treatment of choice in acute iliofemoral venous thrombosis. J Vase Surg 1984;1:867 8 Mavor GE, Galloway JMD. Iliofemoral venous thrombosis: pathological considerations and surgical management. Br J Surg

Yao JS,£ eds. Venous Problems. Chicago: Year Book Medical Publishers, 1978

15 Kistner RL, Sparkuhl MD. Surgery in acute and chronic venous

9 Cockett FB, Thomas M. The iliac vein compression syndrome.

disease. Surgery 1979; 85:31 16 Bertelsen S. Surgical management of iliofemoral thrombophlebitis. Acta Coo Scand 1969; 135:149 17 Johansson E, Nordlander S, Zetterquist S. Venous thrombectomy in the lower extremity: clinical, phlebographic and plethysmographic evaluation of early and late results. Acta Coo Scand 1973; 139:511 18 Lindhagen J, Haglund M, Haglund U, et al. Iliofemoral venous thrombectomy. J Cardiovasc Surg 1978;19:319 19 Einarsson E, Eklof B, Kuenzig M, et ale Scanning electronmicroscopy and Evans blue staining for assessing endothelial trauma and reconstruction after venous treatment. Scan Electron

10 Rutherford RB, Johnson G Jr. Nonoperative management of chronic venous insufficiency. In: Rutherford, RB, ed. Vascular surgery. Philadelphia: WB Saunders Co, 1984 11 Bauer G. A roentgenological and clinical study of the sequels of thrombosis. Acta Coo Scand 1942; 86:1 12 Nicolaides AN, Yao JS1: Investigation of vascular disorders. London: Churchill Livingstone Publishers, 1981 13 Brockman SK, Vasko JS. Observations on the pathophysiology and treatment of phlegmasia cerulea dolens with special reference to thrombectomy. Am J Surg 1965; 109:485 14 Wright CB, Hobson RW II, Swan KG, Rich NM. The pathophysiology of extremity venous occlusion. In: Bergan JJ,

20 Edwards WS. In discussion of Kroener JM, Bernstein EE Valve competence following experimental venous valve autotransplantation. Arch Surg 1981; 116:1467 21 Albrechtsson U, Andersson J, Einarsson E, et ale Streptokinase treatment of deep vein thrombosis and the post-thrombotic syndrome: follow-up evaluation of venous function in the postthrombotic legs. Arch Surg 1981; 116:33 22 Kakkar vv. Read before Symposium on Venous Thromboembolism, Scottsdale, Ariz, October 1984 23 Johansson E, Ericson K, Zetterquist S. Streptokinase treatment of deep venous thrombosis of the lower extremity. Acta Med Scand 1976; 199:89

1969; 56:45

Br J Surg 1965; 52:816

Microsc 1984; 1:273

CHEST I 89 I 5 I MAY, 1988 I Supplement

4378