The Open Vein Hypothesis and Post-Thrombotic Syndrome: Not Dead Yet

Postthrombotic syndrome (PTS) is the most common long-term complication of acute lower extremity deep vein thrombosis (DVT), and prospective data suggest that 20–50% of individuals develop some manifestation of PTS following DVT.¹ Symptoms of PTS include limb pain, heaviness, fatigue, edema, itching, and venous claudication. Following acute DVT, up to 10% of individuals will develop venous ulcers, the most severe manifestation of PTS.² Not surprisingly, PTS is associated with a poor quality of life and dramatically increased healthcare costs.¹

The pathophysiology of PTS is incompletely understood. Acute DVT obstructs venous outflow while also causing vein dilation and disrupting normal venous valve function.³ These changes may persist beyond the acute phase and lead to chronic venous insufficiency, inflammation, and venous hypertension, all of which are key to the development of PTS. Patient factors associated with increased risk of PTS include obesity, advanced age, and recurrent ipsilateral DVT.^{4, 5} The extent of DVT, particularly in the iliofemoral venous segment, is also associated with a heightened risk of PTS.^{4, 5}

Although awareness, detection, and treatment of DVT have dramatically improved in recent decades, effective prevention and treatment of PTS is limited. Graduated compression stockings are often used following acute DVT to reduce soft tissue edema and venous hypertension, but, in a randomized controlled trial did not reduce the incidence of PTS compared to placebo stockings.⁶ A variety of venotonic drugs, such as micronized flavonoid fractions⁷ and horse-chestnut seed extract,⁸ have shown modest improvements in some PTS symptoms. These drugs have been tested only in small studies with unknown long-term benefits and unclear efficacy in preventing PTS.

Given the limits of conservative management for PTS, there is great interest in invasive therapies to prevent the development of PTS following acute DVT. Persistence of thrombus, particularly in proximal venous segments, worsens venous hypertension and PTS symptoms.⁹ Further, the presence and extent of residual vein obstruction directly associates with the development of PTS.¹⁰ The “open vein” hypothesis posits that removing this obstruction through pharmacomechanical means may prevent PTS. This hypothesis has been tested in two large randomized controlled trials to date.

In the CaVenT (Catheter-directed Venous Thrombolysis) study, 209 subjects with acute DVT located from the mid-thigh to the common iliac veins were randomized to standard care, including therapeutic anticoagulation and graduated compression stockings, or standard care plus catheter-directed thrombolysis (CDT).¹¹ At 24 months, subjects undergoing CDT were less likely to have developed PTS, and iliofemoral patency rates were higher with CDT. This was offset by a modest increase in risk of major and clinically relevant bleeding in the CDT arm.

CaVenT only utilized CDT, and it remained unknown if additional interventions, such as balloon venoplasty or venous stenting, would further reduce PTS incidence. More recently, the ATTRACT (Acute Venous Thrombosis: Thrombus Removal with Adjunctive Catheter-Directed Thrombolysis) trial randomized 692 patients with acute DVT to receive standard care, consisting of therapeutic anticoagulation and graduated compression stockings, or standard care plus pharmacomechanical catheter-directed thrombolysis (PCDT).¹² Patients were eligible if symptoms involved the femoral, common femoral, and/or iliac veins. The primary outcome was the development of PTS 6–24 months following intervention. PCDT did not reduce the overall risk of PTS, did not improve quality of life, and showed no reduction in PTS in the iliofemoral DVT subgroup.¹³ As in CaVenT, invasive therapy in ATTRACT also modestly increased the risk of major bleeding.¹²

The results of ATTRACT suggest that routinely performing PCDT for proximal DVT is without benefit, and, to many, marked the end of the intervention-based “open vein” era. Nonetheless, critical questions remained. Patients in both CaVenT and ATTRACT were eligible for enrollment 21 and 14 days, respectively, after the onset of symptoms, and perhaps earlier intervention could still be beneficial. There remains considerable heterogeneity in the affected venous segments treated in both studies, and perhaps limiting intervention to only individuals with proximal iliac disease is most appropriate.

Given this ongoing controversy, the current study by Li et al is a welcome addition to the literature that should help advance the “open vein” debate by specifically focusing on the time interval between DVT formation and restoration of blood flow (RBF).¹⁴ In this study, the authors developed a novel mouse model of proximal DVT and CDT. To induce DVT formation, the inferior vena cava (IVC) and its side branches were ligated, and these ligatures were left in place for at least 2 days to trigger coagulation cascade activation and thrombus formation. Absence or presence of blood flow within the IVC past the thrombus was confirmed using both intravital microscopy-based venography as well as Doppler ultrasound imaging.

This model allowed the investigators to restore blood flow through mechanical de-ligation at a series of timepoints. Compared to sham de-ligation, IVC de-ligation led to a greater likelihood of RBF at day 8 (~90% vs. ~50%; p<0.05). This occurred in a time dependent manner, and RBF was achieved more rapidly with de-ligation at day 2 compared to day 4. Early RBF, defined as occurring by day 4, was associated with a greater reduction in thrombus mass, width, and weight compared to late RBF (occurring after day 4), thus suggesting maximal RBF was not only dependent upon removal of proximal obstruction but also reduction in thrombus burden.

To recapitulate PCDT, the investigators also performed de-ligation at day 2 followed by recombinant tissue plasminogen active (rtPA) infusion at either day 4 or 6. In mice who had already achieve RBF by day 4, rtPA administration did not further improve thrombus burden at day 8. However, in mice without RBF at day 4, rtPA did increase the proportion with RBF at day 6, although this did not reach statistical significance (55.6% at day 4 vs. 76.9% at day 6; p>0.05). Thrombolysis did, however, reduce thrombus burden at day 8 (p<0.01). Late infusion of rtPA at day 6 had no impact on RBF rates or thrombus burden, again supporting a time-dependent link between thrombus formation and maximal impact following CDT.

Histological analysis showed that early RBF was associated with a reduction in vein wall collagen thickness as well as vein wall macrophage and fibroblast infiltrate. Similarly, early RBF led to a reduction in venous wall expression of numerous inflammatory genes, including IL-1β, MMP-2, and PAI-1. These findings were seen whether RBF was achieved with de-ligation alone or de-ligation plus rtPA and suggest early achievement of RBF may prevent some long-term pathological changes associated with DVT.

To help account for these animal data given the overall null findings from ATTRACT, the authors performed a post-hoc analysis from ATTRACT classifying participants based on the time interval from symptom onset to randomization. Overall, participants in the intermediate time frame (4–8 days from symptom onset to randomization) had the greatest benefit in terms of PTS symptoms (p<0.002). In contrast, subjects in the other groups (<4 days or >8 days between symptom onset and randomization) experienced no improvement in VEINES-QoL scores with PCDT (p>0.05).

Overall, these animal data suggest RBF in acute, proximal DVT, whether achieved by mechanical or pharmacomechanical means, is more likely to occur if the intervention occurs early, and early RBF by any means reduces long-term molecular and cellular consequences of DVT. In ATTRACT, subjects randomized early after symptom onset did not benefit from PCDT, and this may indicate that therapeutic anticoagulation alone is effective for fresh thrombus, highlighting the importance of rapidly achieved anticoagulation.¹⁵ In contrast, late intervention may occur beyond a reversible period of DVT sequelae. Clinicians may need to target individuals who have failed early medical therapy but before the long-term pathophysiological consequences of DVT are triggered.

There remain several questions from this study. Although the mouse model allows one to definitively determine the onset of thrombus, this is often not possible in clinical practice. Similarly, invasive imaging to fully characterize the absence or presence of blood flow past an acute DVT is unavailable for most patients. Clinicians need novel non-invasive imaging methods to help guide decision making in this patient population. Also, the mouse model did not incorporate therapeutic anticoagulation, and it is unclear how this could influence RBF as well as the changes seen in the vein wall.

Nonetheless, these findings provide important data suggesting a data-driven, personalized approach to iliofemoral DVT management may be possible in the future. In our own practices, PCDT remains an important therapy for a select group of patients with acute DVT. Given the findings of ATTRACT, it is unlikely we will have another large, randomized controlled trial of PCDT in DVT. Thus, we hope this study helps spur future translational and outcomes-based investigations of this topic.