The ATTRACT trial may seem more attractive than it first looks for the management of acute deep vein thrombosis!

Post-thrombotic syndrome (PTS) affects almost half of patients who develop deep venous thrombosis ¹ (DVT) and represents a significant impediment to both patient functionality ¹ and healthcare costing. ² The syndrome manifests as a result of dysfunctional venous outflow, with the ensuing tissue oedema resulting in an acute inflammatory response within the interstitium. This in turn causes the spectrum of debilitating symptoms known as PTS, which may eventually lead to tissue loss. ³

With persistent outflow obstruction, a predictor of future PTS, ⁴ the aim of the pharmacomechanical venous intervention is to reduce thrombus burden and hence maintain an ‘open vein’ with the hypothesis that this reduces chronic obstructive symptomatology. However, there remains a paucity of level-one data addressing its use.

The recent publication of the Acute Venous Thrombosis: Thrombus Removal with Catheter-Directed Thrombolysis (ATTRACT) trial ⁵ has provided venous interventionists with more robust evidence to guide the management of patients with acute ilio-femoral DVT. This powered multi-centre trial represents the largest dataset to address this challenging patient cohort with participants randomised to pharmacomechanical thrombolysis plus standard treatment or standard treatment alone. Interventions involved initial catheter-directed thrombolysis (CDT) and either interval or immediate thrombectomy with stenting reserved for those with a residual 50% venous stenosis following thrombus extraction. The primary outcome was the development of PTS at 6–24 months as defined by the validated Villalta score, ⁶ limb ulceration or need for further deep venous intervention. Further outcomes included severity of PTS, quality of life (QOL) and importantly major bleeding events. The findings of this long-term and ambitious randomised controlled trial (RCT) undoubtedly raise more questions than it answers, and there is a need for additional prospective data to determine the best management strategies for specific patient cohorts. In contrast with the existing data,^7–9 Vedantham et al. ⁵ identified no difference between groups with regard to the reported PTS (47% vs. 48%, p = 0.56) at 24 months. This may have been down to the design of the trial, which included patients with acute DVT not only within the femoro-popliteal segments but also those with more proximal iliac thrombosis. If enrolment had been limited to ilio-femoral cases, in which there is a higher risk of developing late PTS, the study may have had a higher probability of meeting the primary outcome measure. Thus, while the ATTRACT study recruited significantly more proximal DVT patients than previous RCT’s,^7,8,10 it may well be underpowered to offer definitive PTS outcome data. Of note, subgroup analysis did suggest that the severity of PTS was significantly lower in the study cohort at all timepoints compared to the control. However, intervention offered no improvement in QOL compared to standard therapy with the added detriment of higher peri-procedural major bleed rates (1.7% vs. 0.3%, p = 0.049).

Indeed, the ATTRACT primary outcome contradicts, to some degree, previous well-regarded evidence examining the use of endovenous intervention. The CaVent study group published two- and five-year outcomes^7,8 examining 209 patients, with confirmed ilio-femoral DVT’s. These patients were randomised to CDT plus standard therapy and standard therapy alone. In contrast, the authors noted significantly fewer presentations of PTS at two years in those treated with CDT (41.1% vs. 55.6%, p = 0.047) amounting to an absolute risk reduction of 14.4% and a number needed to treat of seven. Importantly, 20 bleeding events occurred in the treatment group three of which were deemed major. At five years, ⁸ the effects of CDT were more pronounced with 43% of those in the intervention group reporting PTS compared to 71% in the control (p = 0.0001) corresponding to a number needed to treat of only four. However, this benefit failed to translate into an improvement in QOL with no difference identified between groups. Naturally, the results of this smaller study from four centres may be somewhat less generalizable with a suggestion ¹¹ that better compliance (non-significant) with compression therapy and anticoagulation may have skewed data in favour of the CDT cohort.

While the ATTRACT trial provides interesting data, there are some notable limitations. The authors report significant losses to the follow-up assessment (11.5%) within a moderate-sized trial. In addition, multiple modes of intervention and anticoagulation protocols were utilized resulting in significant heterogeneity across participating sites. Ganatra et al. ¹² note, in their correspondence with the authors, the anatomical variability of included DVT’s with 42% of presentations involving the femoral segment alone. Furthermore, it comments on the higher proportion of patients requiring venous stenting compared to those in the CaVent study^7,8 (28% vs. 5.7%) as a potential cause of higher post-intervention PTS through stent occlusion.

Further, randomised evidence from the TORPEDO ¹⁰ trial identified a significantly lower rate of PTS at six months (3.4% vs. 27.2%, p=<0.001) among 183 patients with proximal DVT’s randomised to endovenous intervention plus standard therapy than in those who had standard therapy alone.

These data^7,8,10 are further consolidated by meta-analyses from both Watson et al. ⁹ and Lu et al. ¹³ The Cochrane data ⁹ report the outcomes of 17 small RCT’s including 1103 patients. Included trials utilized variations of systemic, loco-regional or CDT in conjunction with standard therapy with those receiving any form of thrombolysis having significantly lower rates of PTS than their counterparts (RR = 0.66, 95% CI: 0.53 to 0.81; p < 0.0001). Again, intervention was associated with significantly higher bleed rates. However, it is important to note that only two trials specifically addressed CDT with overall study data deemed of moderate quality. Most recently, Lu et al. ¹³ reported the outcomes of 10 trials examining the use of CDT with significantly lower rates of PTS identified in the study group. Interestingly, subgroup analysis of RCT’s alone did not suggest a PTS benefit from CDT. Furthermore, while CDT resulted in a significant reduction in the risk of developing severe PTS, it had no role to play in preventing mild PTS.

As a result, the use of endovenous intervention for acute proximal DVT is tentatively advocated in the guidelines of both the Society for Vascular Surgery (2012) and the American Heart Association (2014)^14,15; however, these recommendations are based on low-quality evidence. Further, major European guidelines^16,17 published pre-ATTRACT advocate iliac stenting in chronic iliac occlusive disease, secondary to thrombosis, to limit progression to post-thrombotic symptoms with meta-analysis data from Seager et al. ¹⁸ suggesting their use to be both ‘promising and safe’. It is important to note that widespread consensus is yet to be established with the American College of Chest Physicians ¹⁹ recommending oral anticoagulation alone over CDT in patients presenting with proximal DVT in their 2016 guideline.

In this context, the ATTRACT data have generated significant debate regarding the use of acute interventional thrombolysis and subsequent thrombectomy. The trial is a well-designed study and is by some way the largest body of work in this patient cohort. While data regarding PTS in the interventional group are somewhat surprising, it is important to note the significant reduction in the severity of PTS in all comers. In addition, the longer term follow-up reported by the CaVent authors ⁸ identified the effects of intervention appeared to improve with time suggesting that the ATTRACT primary outcome may yet change with longer term analysis.

Thus, when the risk of bleeding associated with thrombolysis is taken into account and balanced with the merits of catheter-based intervention, it would appear that, based on the current data, considered patient selection offers the most favourable risk to benefit ratio. As such, based on the current evidence, interventionists should adopt a more measured approach to acute presentations of ilio-femoral DVT. Endovascular thrombolysis/thrombectomy should be reserved for highly functional relatively younger individuals with acceptable life expectancies and with severe symptomatic presentations within 14 days to limit the detrimental effects of uncontrolled bleeding. The ATTRACT should not be viewed as a negative outcome RCT, but one that points to the fact that there certain patient cohorts who may well benefit from early intervention rather than anticoagulation alone.