Re-modelling of venous thrombosis

Background

Deep vein thrombosis (DVT) in the lower extremity veins resolves over time in about two-thirds of the patients. The beginning lysis occurs within minutes after the development of venous thrombosis. Multiple cellular signalling pathways, effectors and growth factors are involved in an intense inflammatory response to thrombus that causes cellular migration and local degranulation of mediators such as cytokines (i.e. tumor necrosis factor-α, interleukin-6, macrophage inflammatory protein-1α, monocyte chemoattractant protein-1). The role of cell mediators in conjunction with the fibrinolysis system in venous thrombosis is under investigation.

Prompt DVT treatment may halt the thrombotic process preventing chronic venous obstruction, preserving valve function, and precluding the development of post-thrombotic syndrome (PTS). The significance of early diagnosis and timely fashioned treatment of VTE is critical. This paper analyses the mechanisms of vein wall damage secondary to venous thrombosis and the rationale for early treatment.

Basic concepts on thrombus resolution and vein wall damage

The role of inflammatory cellular migration and mediators in venous thrombosis has been investigated. In some experimental models, the concept of inflammation involves neutrophil migration followed by monocyte degranulation and aggregation of platelets that have been associated with thrombus formation. ¹ Subsequently, thrombus lysis occurs through enzymes produced by the same neutrophils and monocytes that intensify the inflammatory process leading to thrombus resolution. ²

Addressing the question about temporal relationship of DVT resolution, a mice model was used to gauge metalloproteinase (MMP) especially MMP-2 and MMP-9 activity and intrathrombus macrophage migration. ³ Findings of increased expression of MMPs and local macrophage activity were demonstrated. ³ More recently it was shown that doxycycline halts the production of MMP-9 in a mice model. It was concluded that MMP-9 is part of the process but not critical for thrombus resolution. It remains to be defined if inhibition of other MMPs or local inflammatory markers using different drugs or even monoclonal antibodies will have any impact on DVT resolution.

In addition to the inflammatory markers, the fibrinolytic system also plays its role in DVT resolution. The urokinase plasminogen activator and plasmin activator inhibitor-1 were found to be involved in vein wall re-modelling after an episode of DVT. ⁴ In cases where activation of the fibrinolytic system is deficient or not sufficient, alternative treatment options other than only heparin such as catheter-directed thrombolysis (CDT) or pharmacomechanical thrombectomy may be of assistance in reversing vein wall changes and to preserve valve function.

Signalling pathways and effectors are also part of the process and change in a short period of time. The toll-like receptors (TLR) mainly TLR-9 have been accountable for a potentiation of inflammatory response. ⁵ Microparticles (MP) are vesicles derived from activated platelets, leucocytes and endothelial cells that are associated with thrombosis and inflammation.^6,7 In addition to MP, E- and P-selectins are also involved in thrombogenesis. ⁸ Inhibition of selectin decreases thrombus formation and therefore has a potential therapeutic role. ⁹ Many other factors also play a role in thrombus re-modelling such as chemokines, pro-angiogenic factors, endothelial progenitor cells, hameoxygenase and CD40.

Distribution of venous thrombosis and its correlation with vein re-modelling

Duplex ultrasound (DU) has been utilized to study vein wall changes after an episode of deep vein thrombosis (DVT). Vein wall re-modelling and the timespan from initial events started after a thrombotic event to resolution has been recorded using DU. The location of venous thrombosis and the thrombus load is related to recanalization rates.

The thrombus load is much greater in proximal veins such as in the iliofemoral segment compared with calf veins. Gradual recanalization was observed in a prospective longitudinal study that demonstrated an exponential thrombus regression rate in the femoral and popliteal veins over time. ¹⁰ In this study calf vein thrombus recanalization rate was even higher than in the femoropopliteal segment due to smaller thrombus burden in those veins. ¹⁰ Two potential factors associated with higher recanalization rate in calf veins are calf muscle pump function and reduced thrombus load.

Chronological changes of vein re-modelling and haemodynamics have been demonstrated. In a prospective observational study, 20 patients with acute iliofemoral vein thrombosis receiving conventional anticoagulation were followed periodically for five years. ¹¹ After five years, 70% of the patients had obstructive lesions of the iliac veins with only minor changes occurring during that time. Venous outflow continuously improves with conventional anticoagulation following iliofemoral thrombosis; however, valvular competence and muscle pump function were constantly pathological. ¹¹ In a prospective study of 73 limbs with iliac, femoral and/or popliteal DVT, DU imaging at five years showed complete obstruction in five (6.9%), partial obstruction in 32 (43.8%) and no evidence of obstruction in 36 (36%). ¹² The chance that reflux would develop was significantly higher in limbs with partial compared with those with complete recanalization. ¹² A prospective study of 70 limbs with femoropopliteal and calf vein DVT demonstrated complete thrombus recanalization in 76% of the segments, partial recanalization in 20% and occlusion in 5%. ¹³ Predominantly, vein occlusion was found in the femoral vein (21%) at one year. ¹³ They also found that faster recanalization occurred in the calf veins than in proximal veins, no calf veins with reflux, and that multisegment DVT had a significantly higher incidence of deep vein insufficiency than single segment DVTs at one year. ¹³ Another smaller prospective study of 20 limbs with acute iliofemoral DVT followed for 24 months also showed that valve incompetence was present in at least a quarter of patients despite fast haemodynamic improvement during the first three months after DVT. ¹⁴

Outcomes of patients with DVT also depends to a great degree on single versus multiple thrombosed segments. The likelihood of recanalization decreases as the location of the thrombosed segment moves proximally. ¹⁵ Less than 20% of iliac vein thromboses recanalize, rendering the vein completely patent with no luminal obstructions. ¹⁶ Some studies found that nearly 95% of popliteal or calf DVT recanalize spontaneously and completely compared with about 50% of femoral DVT. ¹⁷

The clinical impact of thrombus location, affected number of segments and its recanalization rates have been reported.^15,18,19 The incidence of recurrence is higher in proximal DVT than in distal DVT. They result in more severe and long-lasting patient morbidities than those with popliteal or below knee DVT. ²⁰ Post-thrombotic syndrome is more frequently seen in patients with iliofemoral DVT than those who have either popliteal or calf DVT.^15,18,19 In a study of 120 limbs with a first episode of DVT found a higher incidence of PTS in patients with thrombosis in multiple segments. Venous claudication was present only when iliac veins were involved (7/37 iliofemoral DVT versus 0/40 femoropopliteal DVT, P = 0.004). ¹⁵ No patient with isolated common femoral, iliac or IVC thrombosis developed venous claudication. ¹⁵ A prospective study of 52 limbs with calf DVT was conducted to evaluate re-modelling rates of the thrombus, its echo texture and vein diameter. Of those 52 limbs, 23 (44%) had either thrombus propagation or lysis. ²¹ Proximal propagation into the popliteal vein occurred in 7 (13%) limbs and one patient had a pulmonary embolism confirmed by computed tomography of the chest seven days after the diagnosis of calf DVT. ²¹ Another prospective study of 1412 patients with an initial DVT episode revealed a 1.3-fold higher risk of developing PTS when a patient sustains an iliofemoral DVT compared with a popliteal vein DVT (relative ratio 1.3, 95% confidence interval 1.1–1.6). ¹⁸ A multicenter Canadian study enrolling 387 patients showed that patients with more severe PTS (as measured by the Villalta score) were found to have a common femoral or iliac vein DVT rather than a calf DVT (2.23 increase in score for iliofemoral versus distal DVT; P < 0.001). ¹⁹ Proximal thrombi have an important impact on social and economic aspects of our society, and therefore understanding how vein re-modelling occurs is of critical importance. ²²

Interventions to halt vein wall damage: the timeslot for reversible changes

The use of oral anticoagulants such as warfarin in the long-term to treat DVT has been implicated in increased risk of PTS when compared with low-molecular-weight heparin (LMWH). In a systematic review comparing the efficacy of LMWH versus oral anticoagulation demonstrated an increased risk of developing PTS in patients treated with oral anticoagulants for ≥3 months corroborating the findings of a previous study that compared tinzaparin to warfarin. ²³ In addition, the likelihood of complete recanalization of an affected segment was higher in patients treated with LMWH. It remains to be investigated whether or not the costs and the risks of using LMWH versus warfarin as a first option to treat DVT are justified.

Catheter-direct thrombolysis for DVT has gained popularity due to its lower risk of bleeding complications compared with massive doses of thrombolytics that were previously used when thrombolytics were administered systematically.^24,25 Timing for treatment is important when indicating endovenous procedures for DVT. The rationale of early thrombus removal is to maintain valve function, decrease vein wall damage and halt vein re-modelling and ultimately prevent thrombotic complications such as DVT recurrence and PTS. More recently, pharmacomechanical thrombectomy (PMT) was added to the armamentarium to treat iliofemoral DVT.^26,27 It consisted of a device that disrupts the thrombus either by using a rotational catheter or continuous saline sprays with subsequent aspiration of thrombus particles. The advantages of PMT versus CDT rely on a faster process to lyse the thrombus, fewer venograms needed to confirm thrombus resolution and lesser amount of lytics given. ²⁸

Conclusion

The time sequence for thrombus re-modelling in humans has not been established in detail. It is not known when the venous thrombi may resolve and what are the factors that promote partial, complete or no lumen recanalization. In the absence of detailed studies with rigorous follow-up that will remain not well understood. Evidence from the existing studies indicates that calf thrombi resolve faster than those in the iliac veins at a slower rate. New information from the inflammatory pathway may lead to development of more efficient drugs for prevention and treatment of thrombosis.