Introduction

Post-thrombotic syndrome (PTS) is one of the most common long-term complications of deep vein thrombosis (DVT). Anticoagulant therapy for DVT does not completely prevent venous valvular damage or persistent impairment of venous patency. Reflux and impaired venous outflow may result in chronic venous hypertension, with manifestations ranging from mild symptoms to advanced venous insufficiency and permanent functional limitation. Markel et al. [1] reported that approximately 20% of affected veins remained occluded two years after DVT, whereas complete recanalisation occurred in approximately 20–40% of cases. Residual post-thrombotic changes persisted in the remaining patients. In a study by Prandoni et al. [2] involving 180 patients after DVT, PTS was diagnosed in 48% of patients with isolated residual changes, 37.5% of those with reflux, and 54% of those with coexisting reflux and residual post-thrombotic changes. The reported incidence of PTS is 11–75%, largely because of differences in diagnostic criteria, definitions, and follow-up duration. Most cases develop within the first two years after DVT, although later presentation is also possible [3–6]. Prandoni et al. [3] estimated the cumulative incidence of PTS at 17% after one year, 23% after two years, 28% after five years, and 29% after eight years. Severe PTS was identified in 3% of patients after one year and 9% after five years.

Despite extensive research, the factors that are most important in the development of PTS in an individual patient remain uncertain, and reliable prediction is still subject to considerable error. Clinically relevant factors appear to include recurrent DVT in the same limb, proximal DVT, and incomplete venous recanalisation. Nevertheless, nearly asymptomatic chronic occlusion of the femoral or iliac vein is still encountered in clinical practice [7]. Although studies and current recommendations for the management of proximal DVT suggest potential benefits of early invasive restoration of venous patency, most patients continue to be treated with anticoagulation alone. This study assessed the long-term outcomes of conservative treatment for proximal DVT and examined the associations of the type and location of post-thrombotic changes with the occurrence and severity of PTS.

Material and methods

The study included 100 consecutive adult patients with a history of proximal lower-extremity DVT who were evaluated at the Vascular Diseases Outpatient Clinic of the Upper Silesian Medical Centre in Katowice-Ochojec in 2013–2014. Eligible patients had experienced the first episode of proximal lower-extremity DVT at least 12 months earlier. Exclusion criteria were recurrent DVT; inability to determine the timing of the first episode; previous fibrinolytic treatment; congenital deep venous obstruction; thrombosis of the inferior vena cava or a vascular malformation; injury to the venous system; previous injury or disease impairing the calf-muscle pump; contraindications to or intolerance of compression therapy, including an ankle-brachial index below 0.8; recurrent DVT or pulmonary embolism within 12 months before enrolment; advanced chronic venous disease before DVT (CEAP C4–C6); age below 18 years; and lack of consent to participate. Information was collected on the course of DVT, comorbidities, risk factors, anticoagulant treatment, and compression therapy. Clinical status was assessed using the CEAP classification and the Villalta scale, revised venous clinical severity score (rVCSS), venous disability score (VDS), and venous segmental disease score (VSDS). The assessment also included symptoms of chronic venous disease, limb circumference measurements, pulse examination, and measurement of the ankle-brachial index. All patients underwent duplex Doppler ultrasonography of the deep venous system to determine the presence, location, and extent of obstructive lesions and reflux. The study protocol was approved by the Bioethics Committee of the Silesian Chamber of Physicians.

Results

The median interval between DVT and study enrolment was 58 months (range, 12–342 months). The mean age was 54 years and the median age was 56 years (range, 20–92 years); the median age at the time of DVT was 45 years (range, 18–90 years). Comorbidities were present in 42% of patients at enrolment: hypertension in 26%, coronary artery disease in 22%, malignancy in 7%, diabetes mellitus in 3%, asthma in 2%, and rheumatoid arthritis in 2%. Obesity was present in 19%, and 13% had undergone major surgery. The causes of DVT are presented in Figure 1.

At the time of diagnosis, thrombotic changes involved both the iliac and femoropopliteal segments (iliofemoral DVT) in 26 patients (26%). Isolated iliac involvement was found in 3 patients (3%), whereas DVT was confined to the femoropopliteal segment in 71 patients (71%). Calf vein involvement was also present in 48 patients (48%). The location and degree of venous obstruction at the time of DVT diagnosis are presented in Figure 2.

All patients initially received low-molecular-weight or unfractionated heparin. After the acute phase, 62% received a vitamin K antagonist. None of the patients received a direct oral anticoagulant. The mean duration of anticoagulant treatment was 50 months, and 59% remained on continuous anticoagulation throughout the follow-up period. At the study assessment, 87% of patients used compression therapy: 71% regularly, 5% frequently, and 11% occasionally.

At the study assessment, CEAP clinical classes C3–C6 were present in 97% of patients. The predominant finding was oedema (C3) in 53%, followed by C4 changes in 36%, an active venous leg ulcer (C6) in 1%, and a healed ulcer (C5) in 4% [8].

The revised venous clinical severity score is 0–30 points. Scores in the study group ranged 0–18, with a mean of 7.0 ±3.67. The mean rVCSS score was 8.15 ±3.49 in patients with PTS according to the Villalta scale and 4.076 ±2.11 in those without PTS. Venous segmental disease scores were 0–8, with a mean of 3.0 ±1.89, against a maximum possible score of 20. Most patients (57%) rated their venous disability as 2 on using VDS score, corresponding to symptomatic venous disease permitting usual activities only with compression therapy and/or limb elevation. The mean VDS score was 1.5 ±0.75, and the maximum score was 4 PTS, defined as a Villalta score of at least 5, was diagnosed in 74% of patients; moderate PTS was present in 21% and severe PTS in 6% (Figure 3).

Pathological reflux within the venous system was identified in 75 patients. Reflux could not be assessed in 5 patients because of venous occlusion, which involved multiple levels in 3%. Deep venous reflux was most frequently detected in the popliteal and calf veins and was considerably less common in the femoral vein. Among patients with PTS, symptoms most commonly coexisted with pathological reflux (65%), whereas 16% had obstruction. In a separate categorisation of patients with PTS, pathological reflux was reported in 72%, obstruction in 1%, and coexisting obstruction and pathological reflux in 5%.

Subsequent analyses examined potential associations between abnormalities detected on duplex Doppler ultrasonography, including obstructive lesions, reflux, and their location, and the occurrence of PTS.

Association between the initial extent of proximal deep vein thrombosis and post-thrombotic syndrome

The extent of proximal DVT at diagnosis was not associated with significant differences in Villalta, rVCSS, or VDS scores. However, VSDS scores were significantly higher in patients with the involvement of three or four venous segments than in those with less extensive DVT (Kruskal-Wallis test, p < 0.05) (Figure 4).

Association between the extent of persistent obstruction at follow-up and post-thrombotic syndrome

A greater number of persistently occluded venous segments at follow-up was associated with significantly greater disease severity according to the Villalta, rVCSS, and VSDS scores. On the Villalta scale, persistent obstruction involving more than two segments was associated with significantly more severe PTS than complete recanalisation of the venous system (Mann-Whitney U test, p < 0.05) (Figure 5).

Association between the initial location of deep vein thrombosis and post-thrombotic syndrome

Neither the location nor the severity (full obstruction, partial thrombosis) of thrombotic changes in the iliac, femoral, popliteal, or calf veins at the time of DVT was associated with significant differences in Villalta, rVCSS, or VDS scores (p > 0.05). Changes in the distal femoral vein, popliteal vein, and calf veins were also not associated with significant differences in VSDS scores (Kruskal-Wallis test, p > 0.05). The only significant association concerned femoral vein changes at the groin. The venous segmental disease score was higher in patients with venous occlusion (median, 3.0; IQR, 2.0–5.0) than in those with changes involving less than 50% of the lumen (median, 2.0; IQR, 2.5–3.0) (Kruskal-Wallis test, p < 0.05). The severity of changes at this site was not associated with differences in Villalta, rVCSS, or VDS scores.

Association between the location of post-thrombotic changes at follow-up and post-thrombotic syndrome

At follow-up, associations between the location of post-thrombotic changes and the individual scale scores were more heterogeneous. Iliac vein changes were associated with significantly higher VSDS scores than the absence of post-thrombotic changes, whereas Villalta, rVCSS, and VDS scores did not differ. At the groin, VSDS scores were higher in patients with changes involving more than 50% of the femoral vein lumen (median, 5.0; IQR, 3.5–6.13) than in those with changes involving less than 50% (median, 3.0; IQR, 2.0–3.75) (Mann-Whitney U test, p < 0.05) with no difference regarding Villalta and rVCSS scores. Significant differences in Villalta and VSDS scores were observed for changes in the distal femoral vein; the highest scores occurred in patients with occlusion, compared with both those with changes involving less than 50% of the lumen and those with 50–90% luminal involvement. No differences were found in rVCSS or VDS scores. The severity of popliteal vein changes was not associated with Villalta, rVCSS, VDS, or VSDS scores (Mann-Whitney U test, p > 0.05). Calf vein analysis was not possible because all changes detected at follow-up involved less than 50% of the lumen (Figures 6, 7).

Association between persistent venous obstruction and post-thrombotic syndrome

Persistent iliac vein occlusion was not associated with a significant difference in the prevalence of PTS, defined as a Villalta score greater than 4 (χ² test of independence, p > 0.05). However, patients with persistent iliac vein occlusion had significantly higher VSDS scores (median, 6.00; IQR, 5.00–6.50 vs. median, 3.00; IQR, 2.00–3.00) and VDS scores (median, 7.50; IQR, 5.25–12.50 vs. median, 7.00; IQR, 5.00–9.00) (Mann-Whitney U test, p < 0.05), with no significant difference in rVCSS score.

At follow-up, the prevalence of PTS did not differ between patients with and without femoral vein occlusion (Villalta score > 4; χ² test of independence, p > 0.05). Similarly, persistent popliteal vein occlusion was not associated with significant differences in Villalta, rVCSS, VDS, or VSDS scores compared with patients without occlusion of this segment (Mann-Whitney U test, p > 0.05).

Association between venous reflux and post-thrombotic syndrome

The severity of PTS and chronic venous disease did not differ significantly in Villalta or VDS scores between patients with and without pathological femoral vein reflux (Mann-Whitney U test, p > 0.05). Patients with femoral vein reflux nevertheless had significantly higher rVCSS and VSDS scores and a higher CEAP clinical class than those without pathological reflux (Mann-Whitney U test, p < 0.05). Thus, femoral vein reflux was associated with greater severity of chronic venous disease.

Patients with popliteal vein reflux had higher VSDS scores (median, 3.00; IQR, 2.00–4.38) than those without reflux (median, 2.00; IQR, 0.00–3.00) (Mann-Whitney U test, p < 0.05). No corresponding differences were observed in Villalta, rVCSS, or VDS scores. The analysis of calf vein reflux yielded similar findings.

Association between the extent of reflux and post-thrombotic syndrome

The extent of reflux was associated with significant differences in VSDS scores (Kruskal-Wallis test, p < 0.05). Patients with pathological reflux in at least one assessed segment had more severe chronic venous disease than those without reflux. Villalta, VDS, and rVCSS scores did not differ significantly between the groups.

For the clinical component of the CEAP classification, the highest classes were observed in patients with reflux at all three assessed levels. Chronic venous disease was significantly more advanced in this group than in patients without reflux and those with reflux limited to two levels. These findings indicate that both the presence and greater extent of pathological reflux are associated with more severe chronic venous disease. The CEAP analysis also suggests that involvement of additional venous levels is associated with a less favourable clinical presentation.

Association of pathological reflux or venous obstruction and post-thrombotic syndrome

Pathological reflux or venous obstruction was associated with significantly higher VSDS scores than the absence of these abnormalities (p < 0.05). Villalta, VDS, and rVCSS scores did not differ significantly between the groups.

Discussion

Post-thrombotic syndrome is a common and clinically important long-term complication of lower-extremity DVT. Its clinical course is heterogeneous, and symptoms may become apparent within the first few months or many years after DVT [5, 9]. Despite advances in the diagnosis and treatment of venous thromboembolism, no single objective parameter can reliably predict PTS in an individual patient [10]. This reflects the complex pathophysiology of the syndrome, which includes persistent venous obstruction, valvular damage and reflux, venous hypertension, microcirculatory dysfunction, and variable capacity for compensatory venous drainage.

In the present cohort, PTS, defined as a Villalta score of at least 5, was diagnosed in 74% of patients. This proportion is high compared with most published estimates. In the meta-analysis by Valerio et al. [11], which included 49 studies and 14,171 patients, the pooled risk of PTS was 33.2%, whereas the risk of moderate or severe PTS was 11.8%. Risk varied substantially according to the most proximal extent of DVT: 51.1% after iliofemoral DVT, 30.7% after femoropopliteal DVT, and 22.8% after isolated distal DVT [11].

The high prevalence of PTS in our cohort may partly reflect the characteristics of the study population. Only patients with a history of proximal DVT who were evaluated at a specialist vascular clinic were enrolled. The median interval between DVT and assessment was 58 months, with a range of 12–342 months. This wide interval and recruitment at a specialist centre may have resulted in overrepresentation of patients with persistent symptoms. The high prevalence of advanced clinical chronic venous disease supports this interpretation: CEAP classes C3–C6 were present in 97% of patients. The prevalence observed in this cohort should therefore not be extrapolated directly
to the overall population after the first episode of DVT.

Interpretation of the findings requires consideration of the differences among the assessment tools. The Villalta scale, rVCSS, VDS, VSDS, and CEAP classification describe related but distinct dimensions of venous disease. The Villalta scale, recommended for the diagnosis and grading of PTS in clinical studies [12, 13], combines five patient-reported symptoms with six signs assessed by the examiner. The revised venous clinical severity quantifies the clinical severity of chronic venous disease, whereas the VDS describes the associated limitation of activity. The venous segmental disease score characterises the anatomical distribution of venous abnormalities, and the clinical component of CEAP classifies visible manifestations of chronic venous disease.

These tools are not equivalent and should not be used interchangeably. A higher VSDS score indicates a greater anatomical extent of reflux or obstruction but does not necessarily imply more severe symptomatic PTS. Similarly,
a higher CEAP clinical class indicates more advanced clinical manifestations of chronic venous disease but does not establish a post-thrombotic aetiology or fulfilment of the Villalta criteria.

The mean rVCSS score was higher in patients with PTS than in those without PTS (8.15 ±3.49 vs. 4.076 ±2.11). This confirms that PTS was accompanied by more severe clinical manifestations of chronic venous disease. Nevertheless, patients without PTS also obtained rVCSS points, and CEAP C3–C6 manifestations occurred in some patients who did not meet the Villalta criterion for PTS. The absence of PTS was therefore not equivalent to the absence of chronic venous disease.

This discrepancy may result from differences in the construction of the individual tools. The Villalta scale includes pain, cramps, heaviness, paraesthesia, and pruritus, none of which is specific to PTS. By contrast, the rVCSS and CEAP classification place greater emphasis on objective consequences of venous hypertension, including oedema, skin changes, and ulceration. Clinically relevant chronic venous disease may therefore occur despite a Villalta score below the diagnostic threshold for PTS, while nonspecific symptoms may increase the Villalta score.

The initial location of thrombotic changes was not associated with significant differences in subsequent Villalta, rVCSS, or VDS scores (however, the number of the iliac vein thrombosis cases in the study group was limited). Similarly, the extent of DVT during the acute phase did not differentiate the results of these scales. Involvement of three or four segments was, however, associated with a higher VSDS score than less extensive DVT. Greater
initial DVT extent may therefore have been associated with more extensive subsequent anatomical damage to the venous system, although no clear association with symptomatic PTS was demonstrated.

The importance of DVT location for PTS risk has not been established consistently in previous studies. Stain et al. [14] and Tick et al. [15] identified proximal location as a significant risk factor for PTS, whereas other studies did not confirm this association [3, 16]. Some authors have reported a lower risk of PTS after popliteal or calf vein thrombosis than after DVT involving more proximal segments [17, 18].

The recent meta-analysis by Valerio et al. [11] strengthens the evidence for the importance of proximal DVT extent. Compared with distal DVT, the odds ratio for PTS was 1.96 for proximal DVT and 3.54 for iliofemoral DVT. The corresponding odds ratios for moderate or severe PTS were 3.69 and 5.56, respectively. These findings indicate a graded increase in risk with a more proximal DVT location.

Impaired outflow due to persistent obstruction and venous hypertension caused by reflux are considered central to the pathogenesis of PTS. Some studies have reported a higher prevalence of PTS in patients with persistent obstructive changes [2], whereas others have emphasised valvular damage and reflux [19–23]. The available findings are not entirely consistent [24].

In our cohort, isolated occlusion of a single segment was generally not associated with a higher prevalence of PTS according to the Villalta scale. Persistent iliac vein occlusion did not differentiate the occurrence of PTS but was associated with higher VSDS and VDS scores. Similarly, femoral vein occlusion at the groin was primarily associated with a higher VSDS score. An isolated anatomical lesion may therefore be detectable on ultrasonography and reflected by an anatomical score without a corresponding increase in clinical scores.

The most consistent associations were observed in patients with extensive, persistent multilevel vein occlusion. Vein occlusion involving more than two segments was associated with higher Villalta, rVCSS, and VSDS scores than the absence of occlusion. Thus, the extent of a lesion and its effect on venous outflow may be more important clinically than the mere presence of a single abnormal segment.

An isolated obstruction may be partly compensated by collateral circulation, whereas multilevel changes involving major outflow vessels may exceed the compensatory capacity of the venous system. This interpretation is consistent with reports of a higher risk of PTS in the presence of persistent obstruction [2]. No single anatomical threshold can, however, determine the clinical significance of obstruction because collateral capacity, calf muscle pump function, and individual adaptive mechanisms also influence its consequences [5].

A second important mechanism is reflux caused by post-thrombotic valvular damage. Pathological reflux was reported in 72% of patients in the present cohort, and assessment was not possible in a further 5% because of extensive obstructive changes. This high prevalence confirms that valvular incompetence is a common long-term consequence of proximal DVT.

Femoral vein reflux was not associated with significant differences in Villalta or VDS scores. Patients with reflux nevertheless had higher rVCSS and VSDS scores and a higher CEAP clinical class. Reflux was therefore associated primarily with more severe objective manifestations of chronic venous disease and more extensive anatomical abnormalities, but it did not clearly differentiate symptomatic PTS.

For popliteal vein reflux, a significant difference was found only in VSDS scores, with no differences in Villalta, rVCSS, or VDS scores. Similar findings were observed for calf vein reflux. Detection of a haemodynamic abnormality therefore does not necessarily correspond to proportional clinical symptom severity. Similar anatomical abnormalities may produce different symptoms depending on collateral development, coexisting obstruction, calf muscle pump function, and other lower limb disorders.

Previous studies of the clinical relevance of reflux have yielded heterogeneous findings. Haenen et al. [25] reported correlations between PTS severity and reflux in the femoral and popliteal veins [20], and Asbeutah et al. [21] reported similar observations. By contrast, Kahn et al. [24] did not confirm a clear effect of coexisting post-thrombotic changes and reflux on PTS prevalence. This heterogeneity may reflect differences in the definition of PTS, diagnostic criteria for reflux, follow-up duration, and study populations.

A greater number of segments with reflux was associated with higher VSDS scores and a more advanced CEAP clinical class but did not influence Villalta, rVCSS, or VDS scores. It cannot therefore be concluded that involvement of each additional segment caused a proportional increase in PTS severity. The finding indicates only an association between more extensive reflux and broader anatomical damage to the venous system and selected clinical manifestations of the disease.

The presence of reflux, vein occlusion, or both in any segment was likewise associated with a higher VSDS score but did not influence the Villalta score. These observations confirm that the anatomical and haemodynamic consequences of DVT do not translate linearly into the Villalta score.

Early restoration of venous patency has been proposed as a means of reducing PTS risk. In the CaVenT trial, catheter-directed thrombolysis reduced the incidence of PTS after acute iliofemoral DVT [26]. The ATTRACT trial did not demonstrate a reduction in the overall incidence of PTS across the entire population with proximal DVT, although subgroup analyses suggested a possible reduction in moderate or severe PTS among appropriately selected patients with iliofemoral DVT [27]. Earlier studies of surgical thrombectomy also suggested that venous patency could be preserved in some patients [28, 29]. The present study cannot compare interventional and conservative treatment because patients who had received fibrinolysis were excluded and all participants received anticoagulation. Evidence concerning early thrombus removal therefore provides pathophysiological context but cannot support therapeutic conclusions based on our findings.

This study has important limitations. It was conducted at a single centre and included patients referred to a specialist vascular clinic, which limits generalisability. The assessment was cross-sectional, and the interval since DVT varied widely. Information on the initial location and extent of DVT was reconstructed from clinical records rather than collected using a uniform prospective imaging protocol. These limitations may have contributed to the absence of significant associations in some analyses.

In summary, the initial location and extent of proximal DVT in our group did not differentiate the subsequent diagnosis of PTS according to the Villalta scale in this cohort. This finding applies only to the analysed population with proximal DVT and should not be generalised to all DVT locations. Previous studies and the meta-analysis by Valerio et al. [11] indicate that PTS risk increases with the extent od more proximal located deep vein thrombosis in broader study group [14, 15]. In the present cohort, extensive persistent multilevel obstruction appeared to have greater clinical relevance. Reflux was associated primarily with more severe chronic venous disease according to the rVCSS, VSDS, and CEAP classification, whereas its association with PTS diagnosed using the Villalta scale was inconclusive. Differences among the scale results reflect distinct dimensions of venous disease and support their combined, but non-interchangeable, use.

Conclusions

Persistent multilevel obstructive lesions and pathological reflux are important factors associated with the severity of clinical symptoms in patients with a history of DVT.

Further studies are required to define the risk factors for post-thrombotic syndrome. Despite analysis of numerous variables, it is currently not possible to predict reliably which patients will develop PTS after an episode of proximal lower-extremity DVT.

Disclosures

1. Institutional review board statement: Not applicable.
2. Assistance with the article: None.
3. Financial support and sponsorship: None.
4. Conflicts of interest: None.

References