Featured Article

Introduction
Post-thrombotic syndrome (PTS) is a common complication of acute deep vein thrombosis (DVT).¹ PTS commonly causes chronic, lifestyle-limiting symptoms such as limb heaviness/fatigue, swelling, and pain, with a minority of patients developing significant venous claudication, stasis dermatitis, and skin ulceration.² Catheter-directed thrombolysis (CDT) refers to a family of image-guided endovascular techniques in which a thrombolytic drug is delivered directly into the venous thrombus through a catheter embedded in the thrombosed vein.3 In modern practice, percutaneous mechanical thrombectomy (PMT) devices are used in conjunction with CDT to accelerate thrombolysis.⁴ By providing rapid thrombus elimination, these techniques have been hypothesized to enable more effective PTS prevention than anticoagulant therapy alone, and safer PTS prevention than systemic thrombolysis and surgical thrombectomy.

However, CDT is inconsistently utilized for initial acute DVT therapy for three main reasons: (1) absence of a multicenter randomized trial showing that CDT prevents PTS; (2) concern that CDT may increase bleeding complications; and (3) the additional healthcare resource utilization associated with CDT. Hence, the American College of Chest Physicians recommends that CDT be restricted to DVT patients with limb-threatening circulatory compromise.⁵ In contrast, the Society of Interventional Radiology states that the use of first-line adjunctive CDT (along with standard anticoagulant therapy) is reasonable to consider for ambulatory patients with extensive acute DVT, low bleeding risk, and long life-expectancy.⁶ This controversy is now particularly acute, since newer pharmacomechanical catheter-directed thrombolysis (PCDT) techniques appear to have overcome many limitations of earlier CDT methods.

The time is right for a major interdisciplinary collaboration between the endovascular specialists who perform CDT (who are more often “CDT proponents”) and the front-line physicians who determine initial DVT therapy (who are more often “CDT opponents”). However, the development of this much-needed effort has been hindered by rhetoric on both sides, which has propagated a number of myths, misconceptions, and inaccurate interpretations of the DVT literature.

Myths and Misconceptions
CDT proponents have often relied upon outdated studies of suboptimal methodologic quality in arguing in favor of CDT. Specifically, the following myths, misconceptions, and anachronisms regarding the DVT literature have been propagated:

The Ubiquity of PTS. Historical surgical series have estimated PTS rates to be 60–70% in proximal DVT patients and > 90% in patients with iliofemoral DVT (defined as DVT involving the iliac vein and/or common femoral vein).^7–9 However, although important in their time, these studies had major limitations, including the lack of available validated measures with which to evaluate PTS, the bias of surgical populations towards more severely affected DVT patients, and the inclusion of patients with recurrent DVT who are now known to be at increased risk for PTS. In fact, a number of modern prospective studies have placed the actual rate of PTS in anticoagulated patients with first-episode proximal DVT to be around 50% at 2 years.^10–12 With daily use of elastic compression stockings, this rate can probably be further reduced significantly.

The Frequency of Venous Ulceration.
CDT proponents have often relied upon anecdotal depictions of the most dramatic PTS cases to justify assumption of the risks of CDT. In particular, graphic depictions of venous ulcers have been a hallmark feature of many CDT presentations over the years. However, it is important to understand that only a small minority of the 400,000–500,000 venous ulcers in the United States are related to PTS, and that by far, the most common cause of venous ulcer is primary valvular insufficiency.¹³ The PTS-related venous ulcer is certainly a morbid complication, with major patient and societal consequences when it occurs, but only 4–7% of first-episode proximal DVT patients will develop an ulcer within 2 years of diagnosis.^10,11 This is important because the physicians who treat DVT on a daily basis are not likely to see many PTS cases that manifest as ulcers. Moreover, given the frequency with which DVT patients are lost to long-term follow up, physicians are unlikely to relate venous ulcers to the incident DVT episode. For this reason, overemphasizing ulcers as a dominant PTS manifestation may not affect physicians, since it may not be consistent with their own clinical observations. On the other hand, it would be far more worthwhile to encourage physicians to actively query their patients about the more common PTS symptoms (i.e., pain, swelling with activity) that are experienced by most PTS patients; if physicians actually did this routinely, the high frequency of PTS would readily become apparent.

The Failure of Anticoagulant Therapy. Whereas CDT proponents have often asserted that anticoagulant therapy does not prevent PTS, published studies have clearly identified the presence of recurrent ipsilateral DVT as an important factor in PTS development. In these studies, the relative risk of PTS is increased 3- to 6-fold when ipsilateral recurrence is present.^1,11 Although conflicting results have been observed in different studies, one recent study found that patients in whom long-term warfarin therapy was non-therapeutic > 50% of the time had a 2.7 times greater incidence of PTS.¹⁴ Therefore, pending further evidence, adequate anticoagulant therapy should be considered an important element of PTS prevention by CDT proponents and opponents alike.

The Importance of Anatomy.
It is important to recognize that the primary care physicians who see presenting DVT patients often view the basic venous thromboembolism (VTE) disease process differently from endovascular subspecialists. Most importantly, internists tend to view DVT as a disorder of hematology/biology, whereas vascular surgeons and interventional radiologists more often view it from an anatomical perspective. This has several important implications that have negatively affected effective communication and collaboration between physicians of differing backgrounds. First, the entity of “iliofemoral DVT” is commonly recognized as presenting fundamentally different physiological considerations and more severe disease manifestations within the endovascular communities, but is rarely distinguished from other forms of proximal DVT by other physicians.¹⁵ This disconnect is partly related to the fact that modern prospective studies have not clearly demonstrated differences in PTS rates between patients with iliofemoral DVT and proximal DVT of lesser anatomic extent, despite the common observation (by this author and others) that patients with iliac vein occlusion fare particularly poorly.^1,7,8 On the other hand, endovascular physicians tend to over-rely on anatomy as an explanation for pathology. For example, CDT proponents have argued that one potential advantage of CDT is the ability to treat underlying venous stenoses using stents (e.g., May-Thurner Syndrome). However, this argument has been taken to an extreme by some who assert that patients treated in this fashion do not require subsequent anticoagulant therapy since the “culprit” lesion has been treated. This practice should be strongly discouraged, since many patients have occult thrombophilias and remain at risk for fatal pulmonary embolism. Indeed, it is by no means clear whether a left iliac vein stenosis actually increases thrombosis risk, or merely affects the site of thrombosis in patients destined to clot for other reasons. Furthermore, in patients who develop recurrent DVT after CDT, endovascular physicians often assume that untreated anatomic lesions (e.g., stenosis or residual thrombus) are to blame. However, it is likely that many of these patients had ongoing risk factors for thrombosis (i.e., occult inherited risk factors) or were not optimally anticoagulated (e.g., the common finding of recurrent DVT in cancer patients treated with coumadin rather than low-molecular weight heparin).^5,16 In addition, there is increasing evidence that nonanatomical factors such as the genetics and biology of inflammation, lipid metabolism, thrombus amplification, and vascular injury may play a significant role in the development of PTS.^17,18

The Safety of Intrathrombus Drug Infusions.
Although CDT proponents commonly claim that intrathrombus infusion of thrombolytic drugs is safer than systemic infusions, there are no comparative studies to confirm this. In fact, the observed major bleeding rate (11.4%) in a large, prospective multicenter CDT registry was quite comparable to studies of systemic thrombolysis using streptokinase (14% in one commonly-cited meta-analysis).^19,20 Although it seems likely that contemporary PCDT techniques may enable substantial reductions in fibrinolytic drug dose, infusion time, and safety to be achieved, these relationships remain unproven at present.⁴

On the other hand, CDT opponents have often failed to recognize the impact of PTS on patients and society, the strength of the proof-of-concept evidence supporting early thrombus removal for DVT treatment, and the strong concordance of the indirect studies favoring CDT in the subset of patients with acute iliofemoral DVT.
The Impact of PTS On Patients and Society. In recent years, the impact of PTS on DVT patients has been clearly demonstrated in a number of studies. Three modern randomized trials have shown that despite anticoagulant therapy and use of elastic compression stockings, 25–46% of patients with proximal DVT will develop PTS within 2 years.^10–12 In addition, DVT patients with PTS experience significantly poorer quality of life (QOL) than those without PTS, and the degree of QOL impairment parallels the severity of PTS.^21,22 Moreover, although the economic impact of PTS has not been precisely quantified, PTS has been estimated to cause > 12% of the new U.S. cases of chronic venous disease (n = 150,474, direct cost $261 million) and venous ulcer (n = 20,556, direct cost $153 million) that occur yearly.13 In a recent study of long-term DVT complications following hip replacement surgery, the cost of severe PTS was $3,817 in the first year and $1,677 in subsequent years.23 Because established PTS is a lifelong condition for which consistently effective treatments are not available, its prevention is important and should represent an integral element of modern DVT care.

The Pathophysiology of PTS. There is good reason to believe that early thrombus removal may prevent PTS in DVT patients. The continued presence of thrombus within the deep venous system during the initial weeks and months after a DVT episode is believed to lead to PTS by two pathways: residual thrombus physically blocks venous blood flow (obstruction), and thrombosis leads to valvular reflux in (a) the thrombosed deep veins (via an inflammatory reaction which damages the venous valves) and (b) in uninvolved deep veins and superficial collaterals (due to compensatory dilatation, which separates the valve leaflets).^24,25 The presence of obstruction and/or valvular reflux produces ambulatory venous hypertension and leads to edema, tissue hypoxia and injury, progressive calf pump dysfunction, subcutaneous fibrosis, and skin ulceration. For this reason, methods to rapidly eliminate venous thrombus and restore unobstructed deep venous flow have been hypothesized to prevent valvular reflux, obstruction, and PTS. In fact, the presence of residual thrombus on ultrasound following DVT has been shown to predict recurrent VTE, a known risk factor for PTS.^26,27

Proof-of-Concept: A Randomized Trial of Contemporary Surgical Thrombectomy.
In a randomized trial of 58 acute iliofemoral DVT patients, Plate et al found that modern surgical venous thrombectomy plus anticoagulation provided more frequent symptom resolution (42% versus 7%, p < 0.005), iliac vein patency (76% versus 35%, p < 0.025), and valvular competence (52% versus 26%, p < 0.05) than anticoagulation alone at 6- month follow up.²⁸ At 5 years, symptom resolution (37% versus 18%, p = NS) and venous patency (71% versus 30%, p < 0.05) were more frequent and mean ambulatory venous pressure was lower (43 mmHg versus 60 mmHg, p < 0.05) in the surgical patients.²⁹ Although this was a single-center study with a small sample size, and while surgical thrombectomy is probably too invasive for routine use in DVT patients, this study does support the basic paradigm of early thrombus removal as a means of PTS prevention.

Proof-of-Concept: Randomized Trials of Systemic Thrombolysis.
Systemic thrombolysis refers to venous thrombus dissolution using a fibrinolytic drug administered via an intravenous line distant from the affected limb. The use of systemic thrombolysis using streptokinase, a first-generation fibrinolytic drug, was evaluated in a number of randomized trials. Overall, > 50% clot lysis (by quantitative analysis of venograms) was significantly more frequent in streptokinase recipients than heparin-alone recipients (62% versus 17%, p < 0.0001).²⁰ In 1979, Elliot et al found that PTS developed less frequently in acute DVT patients receiving streptokinase compared with heparin alone at a mean of 19 months (35% versus 92%, n = 51).30 In 1982, Arnesen et al found that venographic obstruction (56% versus 100%) and PTS (24% versus 67%) was less frequent in streptokinase recipients at a mean 6.5-year follow-up (p < 0.01, n = 42).³¹ In a 1990 randomized trial, Turpie et al observed a trend towards reduced PTS rates in patients with > 50% clot removal with systemic thrombolysis (25% versus 56%, p = 0.07).³² Although excessive bleeding complications have precluded its routine use in modern practice, the experience with systemic DVT thrombolysis provided important proof-of-concept support for the use of early thrombolytic therapy to prevent PTS.

Comparative CDT Studies. Three comparative studies have been concordant in finding CDT to provide better outcomes than anticoagulant therapy alone: (1) In a 2000 study, Comerota et al found that 68 acute iliofemoral DVT patients who underwent successful CDT had significantly fewer PTS symptoms, better physical functioning, less stigma, and less health distress at a mean 16-month follow up than 30 similar patients who received anticoagulation alone³³; (2) In 2001, AbuRahma et al described a prospective study in which 51 acute iliofemoral DVT patients were permitted to choose adjunctive CDT plus anticoagulation or anticoagulation alone. The CDT-treated patients had significantly greater frequent 6-month venous patency (83% versus 24%, p < 0.0001) and 5-year symptom resolution (defined as CEAP classes 0–2) (78% versus 30%, p = 0.0015)³⁴; (3) A single-center randomized trial showed that the use of adjunctive CDT resulted in better preservation of valvular competence (reflux in 11% versus 41%, p = 0.04) and venous function (72% versus 12%, p < 0.001) at 6-month follow up.³⁵ Although each of these studies had important methodological limitations, their concordant findings (along with the data for systemic thrombolysis and surgical thrombectomy) certainly support consideration of CDT for carefully selected patients with acute iliofemoral DVT.

Conclusion
Given the importance of the consequences of both the best-case (safe prevention of PTS) and worst-case (severe bleeding) scenarios, a multicenter randomized trial is certainly needed to better quantify the risk-benefit tradeoffs that are inherent in the selection of CDT as initial DVT therapy. The proper design of this study will critically depend upon an effective multidisciplinary collaboration. It is hoped that this article will assist in overcoming the ideological barriers to developing this type of partnership.