Clinical Outcome of Deep Vein Thrombosis Is Related to Thrombophilic Risk Factors

Abstract

Deep vein thrombosis (DVT) and the associated possible complication of pulmonary artery embolism (LAE) represent a recognized reason for significant perioperative morbidity and mortality. There is a risk of pulmonary artery embolism through embolization. The aim of the study was to investigate the influence of various risk factors on the clinical outcome of the therapy, particularly regarding whether maintenance therapy offers a benefit in terms of the frequency of bleeding and thrombotic events. 80 patients were included, some of them retrospectively from July 2018. The observational period was set to 12 months after the DVT event. In the present sample with n = 80, with 57.5% men and 42.5% women (after 12 months of observation: n = 78), a success rate of the therapies administered of 89.7% was recorded. Only 8.9% showed partial recanalization. 3.8% of the patients had a relapse (also beyond the localization of the leg and pelvic veins) and 8.8% had a residual thrombus during the first 12 months of observation. In this study, BARC (Bleeding Academic Research Consortium) and HAS-BLED (Hypertension, Abnormal renal and liver function, Stroke, Bleeding, Labile INR, Elderly, Drugs or alcohol) scores for identifying the risk of bleeding and Wells scores for assessing the risk of having a thrombosis were used. The Villalta score tested in this study showed significant correlations with residual thrombus (P < .001), recurrence within 12 months (P < .001), and the risk of bleeding (P < .001) and is capable to provide an assessment of the variables mentioned not only at the possible end of therapy but also at the start of anticoagulant therapy.

Introduction

Acute deep vein thrombosis (DVT) in the leg and pelvis is defined as partial or complete obstruction of vascular and/or muscle veins by blood clots. ¹

If part of the clot detaches, there is a risk that it will be carried along with the blood flow via the heart to the lungs. As a result, a pulmonary artery embolism (LAE) develops, the extent of which depends on the size of the clot. ¹

The average risk of developing a thrombosis in patients with age <60 years is 1 in 10 000 and age >60 years is 1 in 100. ³ DVT and LAE are major causes of morbidity and mortality during hospitalization. Around 1.6 million venous thromboembolism (VTE) events are recorded in the European Union each year. ²

The DVT of the leg and pelvic veins is localized in 4 levels with the indication of frequencies: iliac vein (10.0%), femoral vein (50.0%), popliteal vein (20.0%), and lower leg veins (20.0%). It is noted that 2/3 of leg DVT is in the left leg. 20.0% of all untreated lower leg vein thromboses can progress to the femoral veins and 20.0% of all femoral vein thromboses can then lead to an ascending pelvic vein thrombosis. Early diagnosis and therapy were essential. ¹

The diagnostic process starts by estimating the clinical probability of DVT using validated clinical prediction rules such as the Wells score. Further diagnostic procedures, such as compression ultrasound (CUS), should then be used for definitive clarification. ³ The Farbkodierte Doppler-Sonografie (FKDS=color-coded duplex sonography) enables functional statements to be made about the blood circulatory system in addition to the assessment of the vessel morphology.

Risk factors for DVT are divided into endogenous and exogenous factors. In old age, we found an imbalance between pro- and anticoagulation, which in turn can trigger, among other things, hypercoagulation and thus lead to hemodynamic difficulties. ⁴ Virchow's triad describes 3 main pathophysiological causes that lead to the development of thrombosis.^5,6 Many thrombosis patients fulfill most or even all of Virchow's triad^7–9: damage to the vascular endothelium, reduction in blood flow velocity, and changes in blood composition.

The most common causes of hereditary hypercoagulation are heterozygous factor V Leiden (population prevalence up to 7.0%) and prothrombin gene mutation (population prevalence up to 4.0%).^10–13

Other non-hereditary risk factors for the development of thrombosis are diverse^14–16: male gender, age >60 years, DVT or embolism in the past, and surgical interventions (especially of an orthopedic nature).

As soon as DVT has been diagnosed, therapeutic anticoagulation should be started with the aims of preventing a pulmonary artery embolism, preventing the thrombosis from spreading, re-canalizing the thrombosed vessel with preservation of the venous valves, and preventing a post-thrombotic syndrome (PTS).

Here, either the same drug (eg, new oral anticoagulants [NOAC]) as used in the later maintenance phase, only in higher doses or parenteral anticoagulants such as low molecular weight heparin (LMWH) or fondaparinux should be used. ¹

This initial therapy (medication immediately after diagnosis) should be followed by maintenance therapy (additional medication to prevent an early recurrence) for about 3 to 6 months. ¹ The duration of therapy differs between thromboses in the proximal and those in the distal leg veins, between provoked and unprovoked events and recurrences.

Of course, this takes place under a benefit-risk evaluation: longer-term continuation of anticoagulant therapy, considering an associated risk of bleeding. ¹

A risk of recurrence persists for many years after discontinuation of anticoagulant therapy. ¹⁷

The duration and form of therapy must therefore be individually optimized in the future. The main question is whether maintenance therapy could benefit beyond the 3 months of initial therapy regarding the occurrence of bleeding and thrombosis events under anticoagulant therapy.

Patients and Methods

Patients

This prospective, nonrandomized study was performed with approval by the institutional review board (ethical committee, University Hospital Frankfurt, Germany, approval number 246/18). A total of 80 patients with DVT who were older than 18 years were included. The study started on October 1, 2018 and ended on September 30, 2020.

In addition, diagnoses made from July 2018 on and patients with recurrence were considered. Patients who did not have a signed Informed Consent Form or whose reduced general condition 3 months after DVT (start of maintenance therapy) made maintenance appear inappropriate were not included.

The primary goal was to determine the clinical outcome with a possible benefit of maintenance therapy beyond 3 months of initial therapy (Figure 1).

Figure 1.

Schedule of assessments.

The data collected were evaluated, with a focus on the outcome of the therapy, especially when deciding on maintenance therapy. Possible correlations of outcome/course of therapy with age, gender, medication, risk factors, risk of bleeding, type of medication, and additional therapies (physiotherapy, mobilization) were examined. The type of anticoagulation that the patient received was not a focus in this trial. If the treatment with anticoagulants was stopped, the reason for this was noted by the treating physician.

Questionnaires and Scores

Two questionnaires were used for clinical data collection, one at the start of the study and one at the end of the study, and served as support for the treating physician in deciding for or against maintenance therapy by querying numerous risk factors.

The following scores were used: HAS-BLED, BARC, HERDOO2, Wells, and Villalta.

Additional Examinations

In addition to the ultrasound examination at the beginning, after 3 and 12 months, a thrombophilia screening was also carried out after 3 months. The following parameters were also determined^18,19: lupus anticoagulant (lupus inhibitors), cardiolipin antibodies (IgG + IgM), β2-glycoprotein 1 antibodies (IgG + IgM), protein C activity, factor VIII, prothrombin (factor II) gene mutation, factor V gene mutation (factor V Leiden), antithrombin, protein S antigen, free.

After 12 months, the patients also received a vein occlusion plethysmography.

The venous refill time is given in seconds, resulting in the following classification: normal finding: >25 seconds, first-degree venous dysfunction: 20 to 25 seconds, second-degree venous dysfunction: 10 to 19 seconds, third-degree venous dysfunction. Grades: <10 seconds. These parameters were then used together with other available scores, such as the Villalta score used in this study, to assess the PTS. ²⁰

Statistical Analysis

In the further statistical analysis, the clinical outcome was recorded as the primary target variable, and the relationship to the secondary target variables such as mode of therapy (therapy with/without maintenance therapy, co-therapies), numerous risk factors, age, and gender were examined. In addition to descriptive, exploratory statistical evaluations were planned. In the case of categorical variables, chi-square tests were carried out with a focus on testing the independence of 2 characteristics, and in the case of interval-scaled variables, product-moment correlations were carried out. Since the study contains numerous nominally scaled (categorical) variables, chi-square tests were primarily used to check the relationship between the variables regarding their stochastic independence. In addition, the mean values of interval-scaled variables between 2 random samples, which were formed based on the risk factors (eg, concomitant medication yes/no), were compared using t tests. The inferential statistical evaluations were carried out at a significance level of alpha = 5% and alpha = 10%. With Cramer's V, as a measure of the relationship between 2 nominally scaled variables, an effect size of the chi-square test can be specified. According to Cohen (1988), there was a weak association when Cramer's V has a value between 0.1 and 0.3. From a value ≥0.4, a medium relationship was assumed. A strong correlation between the 2 parameters was at Cramer's V values >0.574. The “IBM SPSS Statistics Version 24” software was used for the statistical evaluation of the data set.

Results

Analysis of the Patient Cohort

In the period from October 1, 2018 to September 30, 2020, 80 patients (n = 80) with deep vein thrombosis who met the inclusion criteria of the study were included in the angiology and hemostasis department of the University Hospital Frankfurt.

Overall, 57.5% of the patients were male and 42.5% female. The mean age was 51.8 years (±14.7, range 24-82 years).

In the extent of the thrombosis, a decision must be made as to whether it is long or short. 60.0% of the patients showed long-distance thrombosis, and 40.0% short-distance. Proximal thromboses occurred most frequently with 65.0%, and 32.5% were located proximally on the right. Of all patients with a proximal location, 30.6% had a pulmonary artery embolism.

Risk Factors

Thrombophilic risk factors were found in 43.8% of the patients. Figure 2 visualizes the frequency distribution of the risk factors.

Figure 2.

Frequency distribution of the risk factors (number of patients with the respective risk factor, multiple entries possible).

Several other thrombophilic conditions were excluded from this analysis as since the according laboratory parameters are not assessed as a standard of care at our hospital. The conditions that were not assessed include Sticky Platelet Syndrome and activated protein C resistance phenotypes stemming not from factor V Leiden mutation.

Thrombophilia risk factors were found in 43.8% of all patients (Figure 2). A significant connection between age, surgical intervention, and findings after 3 months could be determined. 84.4% of patients without surgery and 40.0% of patients with surgery showed complete recanalization (χ² = 6.4, df = 2, P < .05, V = .42).

Regarding the second graphic shown here, 67.5% of the patients had a Wells score ≥2, which indicates an increased risk of the presence of thrombosis (Figure 3). The Wells score is used to estimate the clinical likelihood of DVT. If this probability is high, then blood samples should not be taken to determine the D-dimers, but instead, a duplex sonography should be performed directly. The Wells score can be divided roughly into 4 categories: anamnesis, immobilization, clinical signs of the affected leg, and differential diagnoses. The risk factor chart already shows that 1 point is awarded for immobilization, 1 point for previous thrombosis, and 1 point for malignancy. Accordingly, more than 2 points can be achieved in this way, which corresponds to an increased clinical probability of the presence of thrombosis.

Figure 3.

Points achieved in the Wells score.

Concomitant Diseases

60.0% of the patients had comorbidities. The mean value of comorbidities occurring at the same time is M = 0.83. Delayed recanalization was observed in diabetics at both times compared to non-diabetics. In addition, the number of comorbidities occurring at the same time increased with increasing age. A significant connection between thrombus expansion and comorbidities in general and the number of comorbidities could be determined. Other significant correlations could be shown:

• –age, diabetes mellitus, and findings after 3 months/12 months (P < .05)
• –age and number of concurrent comorbidities (P < .05)
• –thrombus extension and concomitant diseases (P < .10)

Concomitant Medications and Scores

71.3% of patients had taken various medications regularly, for example against high blood pressure or pain. A higher score mean value in the HAS-BLED score could be determined in patients with medication, t(67.88) = −4.23, P < .001. Accordingly, they showed an increased risk of bleeding. 6.3% of the patients had a HAS-BLED score ≥3 and thus an increased risk of bleeding. 18.8% of the patients showed a HERDOO2 score ≥2 and thus an increased risk of recurrence.

Thrombophilia Screening

72.5% of patients showed a conspicuous result (rise/fall of 1 or more parameters) in the thrombophilia screening. 32.8% of the patients with a positive result in the thrombophilia screening had at least 1 thromboembolic event in the past (Figure 4). 11 patients with positive findings for thrombophilia showed normal findings on plethysmography, while most patients with positive findings had rank 2. 40 patients with positive results showed complete recanalization, 18 patients with positive results incomplete recanalization (Figure 5).

Figure 4.

Results of the thrombophilia screening (number of patients with the respective finding, multiple entries possible).

Figure 5.

Outcome regarding recanalization after 12 months based on the type of recanalization (number of patients with each finding).

A significant correlation was found between the Villalta score and the individual parameters of the thrombophilia screening (χ² = 46.51, df = 12 P < .001, V = .54). The most common causes of hereditary hypercoagulation: heterozygous factor V Leiden mutation and heterozygous prothrombin mutation were also found in this cohort with 16.25% and 5.0%.

There were no changes in the ultrasound findings in 78% of the patients comparing months 3 and 12. Significant correlations were found between the Villalta score and the findings after 3 months and between the Villalta score and the findings after 12 months (Figure 6).

Figure 6.

Findings of plethysmography divided into 4 ranks based on comorbidity/no comorbidity.

There were significant correlations between gender and findings after 12 months: more men than women showed complete recanalization at this time.

Initial Therapy

30.0% of all patients only received initial therapy.

These showed known and in part only temporary risk factors. These include immobilization, surgery, trauma, and hormonal contraception. 42.5% of patients were immobile prior to DVT, of which 35.3% were surgically immobilized.

Maintenance Therapy

58.8% of patients received maintenance therapy. The patients on maintenance therapy showed the lowest recurrence rate at 2.13%. The recurrence rate in patients with initial therapy was 4.17%, and in those with tumor-related maintenance therapy 12.50%. Based on the criteria listed in the current S2k guideline for the diagnosis and treatment of venous thrombosis and pulmonary embolism (valid until October 2020) for or against prolonged maintenance therapy with anticoagulants, 9 out of 12 concurrences of the risk factors in the decision-making process could be determined. There were increased risk factors of a permanent nature (compared to initial therapy), such as obesity, thrombosis events in the family, dyspnea, pulmonary artery embolism, and chronic venous insufficiency with grades 1 to 3 in plethysmography.

Venous Plethysmography and Villalta Score

The findings of the plethysmography can be divided into 4 ranks (rank 0-3). Here, patients with comorbidities presented a higher score. Both male and female patients could most frequently be classified as rank 2. On plethysmography, 18 rank zero patients also had a Villalta score of 0 points. 27 rank 2 patients also had a zero Villalta score. Overall, only 2 patients achieved a score of 5 to 9 points in the Villalta score, which corresponds to a mild PTS. No significant correlations could be demonstrated between plethysmography and smoking history, plethysmography and age, plethysmography and gender, plethysmography and thrombus extension, plethysmography and previous thrombosis events, and plethysmography and compliance when wearing compression stockings (Figure 7).

Figure 7.

Villalta score findings based on the 4 ranks of plethysmography.

Recurrence and Outcome

The success rate of the therapy was 89.7% (n = 78 with 2 dropouts). 8.97% of the patients showed partial recanalization, and 3.8% recurrence. Evaluated using the HAS-BLED score, 6.3% of the patients showed an increased risk of bleeding. 2.6% of the patients presented with a mild post-thrombotic syndrome (evaluated using the Villalta score). One patient showed no recanalization, instead a recurrence and a residual thrombus. Significant correlations between the Villalta score and bleeding risk and the Villalta score and residual thrombus could be observed.

Discussion

Regarding the patient population, it is noticeable that with an average age of 51.8 years, it is a relatively young patient population, with the increased risk only in patients over 60 years of age being 1:100/year. ²¹ Age as a risk factor for thromboembolic events is due to an associated increase in body weight, less physical activity, and longer and more frequent episodes of illness accompanied by immobilization. ¹

The average age of the patients in the study did not correspond to the risk group over 60 years of age. This can possibly be explained by the fact that this study was based on certain patient exclusion criteria. This included, among other things, a greatly reduced general condition, which makes further maintenance therapy no longer appear sensible and expedient.

A total of 58 patients showed a positive result in the thrombophilia screening as a possible trigger of a thrombotic event. Of these, 13 (68.4%) of the 19 patients were aged ≤40. The partly young age of the patients (the youngest patient was 24 years old) can be explained by this fact.

In the sonographic diagnosis of thrombosis, a distinction was made between long and short, proximal, and distal. Patients with chronic risk factors suffer more frequently from proximal thromboses, and those with transient risk factors more frequently from distal thromboses. ²² This explains the increased number of proximal thromboses in this study. 72.5% of the patients received a positive result in the thrombophilia screening, which is considered a trigger factor of a thrombotic event. This is called a chronic risk factor. Heterozygous factor V Leiden was detected in 16.3% and a heterozygous gene mutation of prothrombin in 5.0% of the patients.

Regarding factor VIII, a sustained elevation is required to be considered a thrombosis factor. The 48.75% of patients with an increased value must be viewed in a differentiated manner.

According to Bertina, 50% of the thrombotic events in a patient with hereditary thrombosis were also associated with acquired factors. ²³ This could be confirmed in the present study: 19 (23.8%) patients had 2, and 2 patients (2.5%) simultaneously had 3 positive results in the thrombophilia screening.

A significant connection was found between the positive result of the thrombophilia screening and the results of the plethysmography. ²⁴ A possible explanation can be found in the fact that of the 58 patients with a positive result in the thrombophilia screening, 32.8% of the patients already reported at least 1 previous thrombosis event in their anamnesis. In the context of a PTS, which should be assessed using plethysmography and the Villalta score, it is not uncommon for the venous valves to be destroyed by an inflammatory and connective tissue organization of the thrombus. A disturbed venous return ultimately results in the late complication of chronic venous insufficiency. The more thromboses were described in the history, the more stressed and stressed is the venous system. ²⁵

Patients with comorbidities also showed a higher degree of plethysmography. No finding deviates from the current literature, since plethysmography, as already described at the beginning of this work, evaluates the venous function and this can be impaired by numerous factors, including multimorbidity. ¹ The Villalta score, which is considered in conjunction with the findings of plethysmography to assess the PTS, did not provide clear results and made a differentiated classification difficult. Only 2.6% achieved a score of 5 to 9 points, presenting as mild PTS. Also, no significant correlations to other scores used or the wearing of compression stockings could be found. However, a crucial significant connection could be described, namely one between the Villalta score and a recurrence within 12 months and a residual thrombus. In the Villata score, parameters such as pain (in general and increased with calf compression), hardening, redness, overheating, edema, itching, and so on were queried. There were striking parallels with the Wells score, which is used at the beginning of the study to determine the clinical probability of thrombosis. Pain, edema, and hardening were also checked here. Accordingly, the Villalta score combines several components. It allows the differentiation of a post-thrombotic change by examining exactly those symptoms that occur with existing vein damage. In addition, these criteria already account for 4 of the 9 possible points of the Wells score. This finding is to be expected given that a recurrence, or even a residual thrombus, can present itself clinically like an initial thrombosis. In a synopsis of the findings from plethysmography and the Villalta score, it can therefore be stated that most of the present sample showed no, if any, mild PTS.

The existing significant connection between age, surgical intervention, and findings after 3 months in the age group 41 to 60 years can be explained by increased surgical interventions in the context of an operation or trauma in this patient sample. Possible considerations include, in addition to more frequent falls with age, also increased elective interventions, especially in orthopedics, for example, due to progressive arthrosis in old age with the need for a new joint.^26,27 When considering the respective concomitant diseases, it was noticeable that diabetics at both time points delayed recanalization compared to non-diabetics. These results of the present random sample agree with the current literature, where diabetes mellitus is ascribed to a 2-fold increase in the risk of thrombosis. The mechanisms of the exact development and the influence of other concomitant diseases have not yet been fully clarified. ²⁸ It is believed that diabetes increases the blood's tendency to clot, especially due to fluctuating blood sugar levels caused by meals or increased blood lipid levels. Increased derailments cause vascular damage and allow diabetic angiopathy to progress uncontrollably. ²⁹

The duration of the anticoagulation performed must also be considered.

As already mentioned in the current AWMF (Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften e. V/Association of the Scientific Medical Societies in Germany) guidelines, criteria for and against prolonged maintenance therapy with anticoagulants must now be defined. ¹ These risk factors influence the decision for or against an extension of therapy. From the present sample, 47 patients received maintenance therapy. Compared to those with initial therapy, patients with maintenance therapy showed an increased number of the following relevant risk factors: obesity, thrombosis events in the family, previous bleeding and thrombosis events, unclear genesis, male gender, smoking history, dyspnea, long-distance thrombus extent, proximal thrombus location, also bilaterally proximally, pulmonary artery embolism, partial recanalization after 12 months, rank 1 to 3 in plethysmography, residual thrombus, rank 1 and 3 in the HERDOO2 score and rank 2 in the Wells score. Based on the current AWMF guidelines, 9 of the 12 (75.0%) of the 12 potential risk factors described in patients who received maintenance therapy can be regarded as fulfilled. This is different for those patients who have been prescribed initial therapy. Here, known risk factors, some of which were only temporary, such as immobilization, surgical intervention, trauma, or even hormonal contraception should be mentioned. Current recommendations for the duration of anticoagulation, most of which have already been considered by the treating physicians in the sample, were discussed below^30–33: Most patients with a first-time thromboembolic event (proximal DVT or LAE) were anticoagulated for a limited period of 3-12 months. The decision to use indefinite anticoagulation should be individualized and based on the latest risk of recurrence and bleeding, especially considering the added value for the patient and patient preferences. Bleeding risk in this sample was determined using BARC and HAS-BLED scores, as recommended by the guidelines. Only a small percentage <10.0% of the sample showed such a bleeding risk. In addition, it was found that patients taking concomitant medication had a higher mean score in the HAS-BLED score and thus also an increased risk of bleeding. Since numerous medications influence blood coagulation, be it hypo- or hypercoagulation, this finding is not surprising. ³⁴

The recommendation for most patients with a first unprovoked proximal DVT, unprovoked symptomatic LAE, active cancer, or antiphospholipid syndrome with a low to moderate risk of bleeding is to prefer indefinite anticoagulation to a defined therapy of 3 to 12 months. The same is recommended for patients with a recurrent episode of unprovoked DVTs. In contrast, patients with provoked DVT and major short-term risk factors, such as surgical intervention or cessation of hormonal therapy, should not routinely receive indefinite anticoagulation. In addition, indefinite anticoagulation should be avoided in patients at increased risk of bleeding. For most patients with recurrent provoked DVT or a first episode of provoked DVT with irreversible, multiple, or minor risk factors, a first episode of unprovoked isolated distal DVT, or an unprovoked episode of incidentally discovered LAE, therapy must be individualized based on patient-specific risk factors for bleeding and thrombosis.

This sample made it clear that the course of recovery is often evident as early as 3 months after DVT. 56 patients and thus 70.9% (n = 79) already showed complete recanalization during the ultrasound control after 3 months. Even 9 months later, the deep and superficial leg veins of these patients were free and without any residuals or recurrences. It can thus be concluded that early therapy can be classified as sensible and effective, if necessary, also in combination with compression therapy, here especially regarding post-thrombotic changes.

A re-evaluation of the risk factors seems important in the clinical setting. Especially after 3 months, because then a potentially increased risk of bleeding under anticoagulation can also be estimated.

This study was planned as a pilot study in which we tried to include all patients presenting with DVT and LAE at the University Hospital Frankfurt during the trial period. We are aware that the relatively small sample size and the heterogeneity of the patient population represent a possible limitation of this study.

Limitations

An incomplete data set due to a lack of consistency in data collection is to be regarded as a possible limitation. There is also the risk of social desirability, which is caused by a lack of anonymity, especially regarding questions about alcohol consumption or the regular wearing of compression stockings. In addition, there may have been a change in the risk factors in the observation period of 12 months, which were not recorded again. The small number of patients in the study, which results from exclusion criteria such as comorbid patients, means that the results were not as representative as desired.