Elastic compression stockings for prevention of the post-thrombotic syndrome in patients with and without residual vein thrombosis and/or popliteal valve reflux

The efficacy of elastic compression stockings (ECS) for the prevention of post-thrombotic syndrome (PTS) arising after a proximal deep-vein thrombosis (DVT) is controversial. ^1,2 Although most randomized studies showed a substantial reduction of PTS with the use of ECS,^3-5 the recent, large SOX clinical trial that used sham stockings as a comparator failed to confirm these findings.⁶ Accordingly, most international guidelines no longer recommend routine use of ECS for the prevention of PTS.^1,7 Nevertheless, ECS are still commonly prescribed in clinical practice.⁸

A recent systematic review showed that patients with DVT who at 6 weeks or later had either residual vein thrombosis (RVT) or popliteal valve reflux (PVR) had a higher risk of subsequent PTS than those without these findings.⁹ As these ultrasound features are associated with longstanding venous hypertension that could be modified by compression therapy,¹⁰ the early identification of RVT and PVR has the potential to identify a subgroup of patients who may still benefit from the use of ECS.

In a prospective cohort study of 869 patients with a proximal DVT that was either unprovoked or associated with transient risk factors, we observed an increased risk of PTS in those with RVT.¹¹ Briefly, all patients received initial treatment with unfractionated or low-molecularweight heparin followed by vitamin K antagonists according to international guidelines, with individual treatment duration (ranging between 3 and 24 months) based on each patient’s preferences and risk profile. Patients with recent (<2 years) ipsilateral DVT and those requiring indefinite anticoagulation were not eligible.¹¹ Patients were advised to wear ECS (30-40 mmHg at the ankle) for at least 2 years, and were followed-up for 3 years. They were instructed to report in a booklet how long they wore the stockings, the use of not permitted stockings, and the occurrence of any adverse effects impairing their use. At 3 months, an ultrasound assessment was done to document the presence of RVT (vein diameter under maximum compressibility >4 mm)¹¹ and PVR (retrograde flow through the popliteal valve after manual compression of the mid-thigh >0.5 seconds, which persisted after repeating the maneuver with a tourniquet).¹² The Villalta scale was used to assess the development of PTS every 6 months. A score of 5 to 14 in two, even non-consecutive, assessments indicated non-severe PTS, while a score ≥15 or the presence of a skin ulcer in a single assessment indicated severe PTS.^1,11,13

Here we report the risk of PTS in relation to therapeutic adherence to ECS and the presence of RVT, PVR or both in the 861 patients who survived at least 6 months. Two trained physicians who were unaware of the patients’ other details or study outcomes assessed the adherence to ECS. Patients who used the ECS for ≥70% of daytime for the first year were considered adherent (‘stockings’ group). Patients who did not accept the advised ECS, discontinued ECS use during the first year, or used the ECS <70% of daytime were considered nonadherent (‘non-stockings’ group).

The main demographic and clinical characteristics of the two groups were compared using standard methods. The hazard ratio (HR) for the effect of ECS on PTS development in the whole population, as well as in patients with and without RVT and/or PVR was estimated using the proportional hazard Cox regression model. Interaction terms were defined between RVT and ECS, and between PVR and ECS. A minimal significant model was achieved by a likelihood ratio-guided forward stepwise variable selection method. In each of the four subgroups of patients with or without RVT and/or PVR in the ‘stockings’ or ‘non stockings’ group, the cumulative incidence of PTS was estimated by the Kaplan-Meier method, tested by the log-rank test, and graphically represented by product-limit survival estimates by the final minimal significant model.

Figure 1.Cumulative incidence of patients free of post-thrombotic syndrome in each of the four subgroups according to the presence of residual vein thrombosis and/or popliteal valve reflux and the use of elastic compression stockings. PTS: post-thrombotic syndrome; ECS: elastic compression stockings; RVT: residual vein thrombosis; PVR: popliteal valve reflux.

Table 1.Demographic and clinical characteristics of the study patients, separately in the ‘stockings’ and ‘non-stockings’ groups.

Of the 861 patients, 511 (59.3%) belonged to the ‘stockings’ group, and the remaining 350 (40.7%) to the ‘non-stockings’ group. The two groups had substantially comparable demographic and clinical characteristics (Table 1). RVT and/or PVR was detected in 539 patients (62.6%). Of these, RVT alone was found in 299 (55.5%), PVR alone in 115 (21.3%) patients, and the combination of RVT with PVR in 125 (23.2%).

PTS developed in 249 of the 539 patients (46.2%; severe in 35, 6.5%) with RVT and/or PVR, and in 90 of the 322 (28.0%; severe in 11, 3.4%) without RVT and/or PVR (HR=2.18; 95% confidence interval [95% CI], 1.73-2.74). PTS developed in 162 of the 511 patients (31.7%; severe in 19, 3.7%) in the ‘stockings’ group, and in 177 of the 350 (50.6%; severe in 25, 7.1%) in the ‘non-stockings’ group (HR=0.64; 95% CI: 0.51-0.79; P<0.001).

In patients with RVT and/or PVR, PTS developed in 114 of the 328 (34.8%) in the ‘stockings’ group (severe in 14, 4.3%), and in 135 of the 211 (64.0%) in the ‘nonstockings’ group (severe in 19, 9.0%), for hazard ratios of all and severe PTS of 0.52 (95% CI: 0.41-0.66; P<0.001) and 0.41 (95% CI: 0.21-0.83; P=0.013), respectively. In patients without RVT and/or PVR, PTS developed in 48 of the 183 (26.2%) patients in the ‘stockings’ group (severe in 5, 2.7%), and 42 of the 139 (30.2%) patients in the ‘non-stockings’ group (severe in 6, 4.3%), for hazard ratios of all and severe PTS of 0.95 (95% CI: 0.62-1.44; P=0.80) and 0.59 (95% CI: 0.18-1.95; P=0.59), respectively (Table 2). In patients with RVT and/or PVR, the 36- month PTS-free survival figures were 35.2% (95% CI: 28.7-41.7) in the ‘non-stockings’ group, and 64.0% (95% CI: 58.7-69.3) in the ‘stockings’ group (P<0.001). In patients without RVT and/or PVR, the respective figures were 69.3% (95% CI: 61.5-77.1) and 73.5% (95% CI: 67.0-78.0), respectively (P=0.43) (Figure 1).

Of utmost importance, while the interaction term for the use of ECS and presence of RVT was highly significant (P=0.037), the term for the use of ECS and presence of PVR was not (P=0.46).

Our study strongly suggests that in patients with proximal DVT, adequate use of ECS provides a clinically important reduction in any and severe PTS in patients with ultrasound evidence of RVT (with or without PVR) at 3 months, whereas in patients without RVT such an effect is absent. In a clinical context dominated by persistent uncertainty on the necessity for compression therapy to prevent PTS, our study has the potential to revive a stalled discussion.

Table 2.Incidence of post-thrombotic syndrome in patients according to the use of compression elastic stockings and the presence of residual vein thrombosis and poplitieal valve reflux.

Recently, a hypothesis-generating meta-analysis showed a more than two-fold higher incidence of PTS in patients with ultrasound evidence of RVT at least 6 weeks after the index DVT.⁹ In patients with PVR, the incidence of PTS was also increased but only by onethird. In hindsight, this small increase could be easily explained by the confounding effect of RVT, which can also occur in combination with PVR. Indeed, in our study 48% of patients with PVR also had RVT, but in the multivariate minimal-significant Cox model, there was no independent effect of PVR on the incidence of PTS, or on the relative efficacy of adequate use of ECS.

Our observation is pathophysiologically plausible. Indeed, in the absence of longstanding vascular damage, venous hypertension and subsequently PTS are unlikely to occur.^9,14 This is also consistent with the demonstration that PTS is unlikely to develop in individuals with a limited thrombotic burden and in those with isolated calf DVT.¹

Our results are robust, as they are based on a prospective observation of patients with proximal DVT who were followed up for as long as 3 years.¹¹ Furthermore, the assessment of the adequacy of use of ECS was done by physicians who were unaware of clinical outcomes or potential confounders. To minimize the effect of potential confounders, patients with recent ipsilateral DVT were excluded, as were patients with a short life expectancy and those requiring indefinite anticoagulation. The main limitation of our study was the lack of random allocation to ECS or no ECS. However, the two study groups were virtually comparable for demographic and clinical characteristics.

The clinical implication of our observations in combination with results of contemporary studies is that in patients without RVT at 3 months, ECS can be safely withheld as long as leg complaints have disappeared, ideally after completing the first 6 months following the thrombotic episode.¹⁵ However, there is still uncertainty regarding patients with substantial damage to their venous system, as demonstrated by the presence of RVT at 3 months. Hence, we believe that trials should be initiated to assess the effect of ECS in patients with proximal DVT and RVT. As our patients were instructed to use ECS from the beginning, and the likelihood of RVT at 3 months was substantial (>50%), we think that both in clinical practice and in further confirmatory studies ECS should be given as soon as possible after the acute DVT to all patients while awaiting the ultrasound test, which has the potential to help decisions on the subsequent approach.

In conclusion, our results show that the ultrasound assessment of RVT in patients with proximal DVT has the potential to identify those who are likely to benefit from the long-term use of ECS. While awaiting confirmation from properly randomized clinical trials, they are, in our opinion, plausible enough to inform the long-term management of patients with proximal DVT.

Footnotes

• Received July 21, 2021
• Accepted September 1, 2021

Correspondence

PAOLO PRANDONI - prandonip@gmail.com

Disclosures: no conflicts of interest to disclose.

Contributions: PP, AWAL and MHP designed the study; SV, RP and DT contributed patients, and analyzed and interpreted data; FN and MHP performed the statistical analysis; GP provided the administrative support; PP, AWAL and MHP drafted the manuscript; all authors critically wrote or revised the intellectual content of the manuscript, reviewed and/or commented on each draft, and approved the final version for submission.