The dual role of Iliac vein compression in the occurrence of pulmonary embolism

Abstract

AbstractSection Objective

See More

This study explores the risk factors for pulmonary embolism in patients with acute left lower extremity deep vein thrombosis and investigates the dual role of left iliac vein compression in the occurrence of pulmonary embolism.

AbstractSection Methods

In a single-center retrospective study, patients diagnosed with left iliac vein compression combined with acute left lower extremity deep vein thrombosis who completed pulmonary artery CT imaging examinations were consecutively enrolled from 2013 to 2023. The demographic characteristics, risk factors, comorbidities, onset time, degree of left iliac vein compression, and the relationship with the occurrence of pulmonary embolism were collected and analyzed. Logistic regression analysis was used to evaluate the odds ratios and 95% confidence intervals of pulmonary embolism across different levels of factors, and Restricted Cubic Spline were used to assess the association between pulmonary embolism and the degree of left iliac vein compression.

AbstractSection Results

A total of 120 patients were included, of which 48 were diagnosed with pulmonary embolism. Regression analysis found that left iliac vein compression was a protective factor against pulmonary embolism (B=-1.736, p < 0.001). Restricted Cubic Spline revealed a dual role of iliac vein compression in the occurrence of pulmonary embolism (p = 0.002): within the range of 0-46.7%, there was an increasing trend in the risk of pulmonary embolism as the degree of iliac vein compression increased; in the range of 46.7%-100%, as the degree of left iliac vein compression increased, the incidence of pulmonary embolism showed a decreasing trend; for patients with left iliac vein compression exceeding 62%, the incidence of pulmonary embolism was significantly reduced (OR < 1).

AbstractSection Conclusions

In patients with acute left lower extremity deep vein thrombosis, the degree of left iliac vein compression may initially promote and subsequently inhibit the occurrence of pulmonary embolism. We should be vigilant about the risk of pulmonary embolism in patients with moderate degrees of left iliac vein compression.

Introduction

Pulmonary embolism (PE) is the third most common acute cardiovascular event worldwide, following myocardial infarction and stroke[1], and accounts for 0.3% of emergency department visits [2]. PE is also a common cause of death, with more than 15% of patients dying within three months of being diagnosed with PE[3,4,5]. 25% of patients with PE present with sudden death as the initial manifestation [6]. Recent studies have shown that the mortality rate of PE, especially that of high-risk PE, is even on an upward trend [7, 8]. Up to 50% of patients with deep vein thrombosis (DVT) in the lower extremities may develop PE[9]. In patients with proximal DVT of the lower extremities, such as those with left iliofemoral vein thrombosis, the incidence of PE may be even higher [10]. Left iliac vein compression (LIVC) promotes DVT, especially iliofemoral vein DVT [11]. However, whether severe LIVC, while promoting DVT, also prevents thrombus detachment to the pulmonary arteries and thereby reduces PE has been rarely studied. This study compared the risk factors for PE among patients with LIVC combined with DVT in the left lower limb, exploring the different impacts of left iliac vein compression on the occurrence of PE.

Methods

Patients and procedures

Between July 2013 and July 2023, patients admitted to our hospital with LIVC combined with acute DVT of the left lower extremity were consecutively collected in this study. Exclusion criteria included a previous history of DVT or PE, or having received an inferior vena cava filter implantation. In our study center, combined computed tomography pulmonary angiography (CTPA) and CT venography (CTV) were routinely performed for patients with suspected PE or proximal DVT. All included patients were diagnosed with acute left lower extremity DVT using color Doppler ultrasound and had undergone combined CTPA and CTV. The presence or absence of PE was confirmed by CTPA, and the degree of iliac vein compression was evaluated using CTV. To calculate the degree of LIVC, the ratio of the diameter at the most severely compressed site of the left iliac vein to the diameter of the non-compressed distal part of the iliac vein was determined. The degree of LIVC was then expressed as the difference between 1 and this ratio [12, 13]. The measurement of the iliac vein and assessment of stenosis were independently performed by two board-certified radiologists blinded to patient information. When there was disagreement between the two reviewers, a third senior radiologist determined the final degree of compression. Meanwhile, to exclude the influence of mechanical thrombectomy or inferior vena cava filter placement on the occurrence of PE, we excluded patients who only completed the CTPA examination after undergoing mechanical thrombectomy or inferior vena cava filter placement. The collected data included demographic characteristics (gender, age, BMI), risk factors, comorbidities, onset time, and whether pulmonary embolism was present. A patients flow diagram inclusion and management is shown in Fig. 1.

Fig. 1

Patients flow diagram

Statistical analysis

The Kolmogorov-Smirnov test was used to assess the normality of the samples. For quantitative data that followed a normal distribution, they were expressed as mean ± standard deviation, and group comparisons were made using the t-test. If the data didn’t follow a normal distribution, they were presented as median (interquartile range), and group comparisons were conducted using the Mann-Whitney U test. Qualitative data were expressed as frequencies (percentages) and compared between groups using the χ2 test. Logistic regression analysis was employed to evaluate the odds ratios (OR) and 95% confidence intervals (CI) for PE across different levels of factors. Restricted cubic splines (RCS) were used to assess the association between PE and risk factors.

Results

Patient characteristics

A total of 352 patients with LIVC combined with acute DVT of the left lower limb were identified. Patients with a history of DVT or PE, having already accepted an inferior vena cava filter implantation, or failed to complete a CTPA examination were excluded. Ultimately, 120 patients were included in the study, among whom 48 were diagnosed with PE. The initial anticoagulation treatment for patients was heparin-based, with low molecular weight heparin preferred according to the patient’s renal function; otherwise, unfractionated heparin was used.

Among the 120 patients 39.2% were male, with an average age of 60.5 ± 13.7 years, and a median BMI of 26 kg/m² (IQR 24, 28 kg/m²). The median degree of LIVC was 62.5% (IQR 38.25%, 81%). The median onset time (time from the onset of DVT or PE symptoms to the formal diagnosis of PE) was 5 days (IQR 3.0, 7.0 days). Sedentary lifestyle (37.5%), major surgery or trauma (26.7%), and thrombophilia (hereditary thrombophilia, including 5 cases of antithrombin III deficiency, 1 case of protein C deficiency, and 4 cases of protein S deficiency) or cancer (22.5%) acted as the main risk factors for venous thrombosis. Hypertension (51.7%) and diabetes (19.2%) were the primary comorbidities. Patient baseline characteristics are shown in Table 1.

Table 1 The demographic characteristics, risk factors, comorbid conditions, onset time, and the degree of left iliac vein compression in 120 patients

Comparisons between groups showed that the degree of LIVC in patients with PE was significantly lower than in patients without PE (51.5% vs. 69%, p = 0.021). A violin plot illustrated the distribution of the degree of LIVC in patients with and without PE (Fig. 2 A), indicating that patients with a LIVC degree around 50% had a higher incidence of PE, whereas those with a LIVC degree close to either extreme had a lower incidence of PE.

Fig. 2

Violin plots illustrating the distribution of the degree of LIVC in patients with and without PE (A) and in patients with and without thrombus in the left iliac vein (B)

A total of 96 patients (80%) had thrombi extending proximally to the left iliac vein, with a median degree of LIVC of 69%. In contrast, 24 patients (20%) had thrombi extending proximally only to the left femoral or popliteal vein, with a median degree of LIVC of 40%. A violin plot displayed the distribution of the degree of LIVC in patients with and without thrombus in the left iliac vein (Fig. 2B), showing that patients with proximal DVT had a higher degree of LIVC (p = 0.001).

Risk factors for PE

Univariate regression analysis showed that the degree of LIVC was a protective factor for PE (OR 0.176; 95% CI 0.070–0.442; p < 0.001), and age was a risk factor for PE (OR 3.804; 95% CI 1.389–10.420; p = 0.009). Gender, BMI, risk factors, comorbidities, and onset time were not significantly associated with the occurrence of PE (p > 0.05). Regression analysis and forest plot evaluating the OR and 95% CI for PE across different levels of factors are shown in Fig. 3 (For some factors, the lower bounds of the OR’s 95% CI were too low or the upper bounds too high to be fully displayed on the forest plot. The complete specific OR values and 95% CIs were provided in the table next to the forest plot.). Multivariate regression analysis confirmed that the degree of LIVC was an independent protective factor for PE (OR 0.192; 95% CI 0.083–0.444; p < 0.001), and age was an independent risk factor for PE (OR 2.808; 95% CI 1.221–6.461; p = 0.015).

Fig. 3

Regression analysis and forest plot

The RCS results confirmed the manner in which the influencing factors affected the occurrence of PE (Fig. 4). The degree of LIVC had a dual effect on the occurrence of PE (P for nonlinear = 0.002) (Fig. 4 A). Within the range of 0-46.7%, there was an increasing trend in the incidence of PE (OR of PE) as the degree of LIVC increases. In contrast, within the range of 46.7%−100%, there was a decreasing trend in the incidence of PE (OR of PE) as the degree of LIVC increases. For patients with a LIVC degree exceeding 62%, the incidence of PE was significantly reduced (OR < 1). Age had a unidirectional promoting effect, rather than a dual effect on PE (P for non-linearity = 0.424) (Fig. 4B). The risk of PE showed an increasing trend with advancing age.

Fig. 4

RCS results confirming that the degree of LIVC had a dual effect on the occurrence of PE (A) and that age had a unidirectional promoting effect on PE (B)

Discussion

PE has become the third most common acute cardiovascular event and a major cause of death [1, 3]. Most pulmonary emboli originate from deep vein system of the lower limbs [14, 15]. LIVC is an important cause of stasis and promotes the development of venous thrombosis according to Virchow’s triad [11, 16, 17]. This study, based on measurements of the degree of LIVC during the acute phase of thrombosis, found that patients with DVT in the left iliac vein had significantly greater LIVC compared to those with thrombosis distal to the left femoral vein. Although the accuracy of measuring LIVC during the acute phase might be affected by the presence of acute thrombus, this study preliminarily suggested that LIVC may be associated with proximal thrombus formation. Proximal DVT in the lower limbs is related to an increased incidence of PE[9, 18], and may lead to severe, potentially fatal conditions [19]. Therefore, LIVC not only promotes the formation of proximal DVT in the left lower limb but also increases the potential risk of PE.

Recently, a few studies have identified a protective effect of severe LIVC against the occurrence of PE. In a retrospective study involving 3,101 patients, Zhang et al. found that although the prevalence of DVT in the left lower limb was higher than in the right lower limb, the prevalence of PE in patients with left lower limb DVT was lower [20]. In another retrospective study involving 1,476 patients, Chen et al. found that patients with left-sided DVT had a lower incidence of symptomatic PE compared to those with right-sided DVT. Specifically, the rate of symptomatic PE in patients with left iliac vein thrombosis was 5.4%, which was lower than that in patients with left non-iliac vein DVT, right iliac vein DVT, and right non-iliac vein DVT (12.8%, 10.1%, and 16.6%, respectively; p< 0.001)[21]. In a retrospective study involving 226 patients with lower extremity DVT, Shi et al. found that the risk of PE in patients with left lower limb DVT was significantly lower than in those with right lower limb DVT. Additionally, when the degree of LIVC exceeded 42.9%, this increased compression became a protective factor against PE[22]. In a retrospective study, Paik et al. examined 114 patients with DVT, with or without PE. They found that the shortest length of the left iliac vein in patients with DVT combined with PE was 11.3 mm, which was significantly longer than the 6.7 mm observed in patients with DVT not complicated by PE. The ROC curve analysis revealed that when the compressed length of the left iliac vein was less than 7.6 mm, the risk of PE significantly increased [23].

Additionally, some researchers have reported that in patients with DVT associated with iliac vein compression undergoing thrombectomy, no symptomatic PE occurred even without the preemptive placement of an inferior vena cava filter [24].

We hypothesize that moderate degrees of LIVC may promote the formation of venous thrombi in the left lower limb but fail to prevent these thrombi from detaching and traveling to the pulmonary arteries, thus increasing the risk of PE. In contrast, severe LIVC may act as a natural physical barrier that prevents thrombi from the left lower limb from reaching the pulmonary arteries, thereby reducing the risk of PE. As the degree of compression increases, the role of left iliac vein compression shifts from being a risk factor to a protective factor for PE.

Unfortunately, previous studies have not thoroughly explored this dual effect. The present study, through RCS analysis, found that a 46.7% degree of LIVC appeared to be a critical threshold for its dual effect on PE. Although the promoting effect of LIVC on PE was not statistically significant when the compression degree was less than 46.7%, we believe that further studies with larger sample sizes are needed to confirm this observation. This study confirmed the protective effect of severe LIVC (compression degree greater than 62%) against PE, and suggested that this protective effect became stronger as the degree of compression increased further. It should be noted that in this study, measurements of LIVC were all performed within the acute phase of DVT. Since the majority of patients had more than 50% LIVC and thus received iliac vein stenting during treatment, we were unable to repeat measurements of the original degree of LIVC during the subsequent non-acute phase, nor could we compare or verify the accuracy of the compression measurements taken during the acute phase of DVT. Nevertheless, CTV remains one of the preferred methods for assessing the degree of left iliac vein compression within the acute phase of DVT to date.

The Antithrombotic Therapy for VTE Disease guideline (CHEST Guideline and Expert Panel Report) states that placement of inferior vena cava filters is not recommended for patients with VTE who are receiving anticoagulant therapy. However, some researchers remain concerned about the risk of PE in patients with proximal DVT [25]. Our study further clarifies the risk factors for PE. It suggests that there may be no need for excessive concern regarding PE in patients with severe LIVC, whereas PE should still be carefully monitored in those with moderate LIVC. Therefore, we believe that for patients with left proximal DVT combined with moderate LIVC, inferior vena cava filter placement should be more actively considered when there are contraindications to anticoagulation or thrombus progression despite anticoagulation. However, in patients with moderate LIVC who are eligible for anticoagulation and in whom thrombus does not progress during anticoagulation, whether a more proactive inferior vena cava filter placement strategy is necessary requires further investigation.

Through multivariate regression analysis, this study found that age was an independent risk factor for PE in patients with LIVC (OR 2.808; 95% CI 1.221–6.461; p = 0.015). Further analysis using RCS curves didn’t reveal any dual effect of age on PE (P for non-linearity = 0.424), indicating that age had a unidirectional promoting effect on PE. In patients with LIVC combined with acute left DVT, the risk of PE increases with advancing age.

In contrast to this study, other research has indicated that gender may be a risk factor for PE[22]. The impact of gender or other factors on PE requires further exploration.

Limitations

There are several important limitations to the study. Firstly, this was a single-center retrospective study, and the findings may be affected by an insufficient sample size. Secondly, the study assumed that all PE emboli originated from thrombi in the left lower limb. However, it cannot completely rule out the possibility of a small number of in situ pulmonary artery thrombi formation. Despite these limitations, this study represents an exploration into the factors influencing PE as well as the dual role mechanisms of these influencing factors. The research contributes valuable insights into how different levels of left iliac vein compression can have varying impacts on PE risk. Future large-scale prospective studies are needed to validate these findings.

Conclusions

In patients with acute left DVT complicated by LIVC, age was a risk factor for PE. The degree of LIVC had a dual effect on PE. Within the range of 0-46.7%, as the degree of LIVC increases, there was a trend towards an increased incidence of PE. However, within the range of 46.7%−100%, as the degree of LIVC increases, there was a trend towards a decreased incidence of PE. Patients with a LIVC degree exceeding 62% had a significantly lower incidence of PE. In patients with left lower limb DVT, particular attention should be paid to the risk of PE in those with moderate LIVC.