Altered fibrin clot structure/function in patients with idiopathic venous thromboembolism and in their relatives

Venous thromboembolism (VTE) involving deep-vein thrombosis (DVT) and pulmonary embolism (PE) affects 1 to 3 per 1000 persons per year.¹  The pathogenesis of VTE is multifactorial. It is well known that common risk factors, such as surgery, trauma, prolonged immobilization, malignancy, pregnancy, or oral contraceptives, are consistently associated with VTE.^2,3  The frequency of hereditary thrombophilia in unselected patients with unprovoked VTE is approximately 25%.^3,4  The most common thrombophilic factors are prothrombotic polymorphisms, such as factor V Leiden and prothrombin G20210A.^3,4  However, thrombophilia screening fails to identify predisposing factors in 30% to 50% of idiopathic VTE patients.²  Moreover, a family history of VTE has been reported to be associated with VTE occurrence irrespective of major congenital and acquired risk factors.⁵ 

The thrombin-mediated conversion of plasma fibrinogen into fibrin and the formation of clots relatively resistant to lysis are the final step in the blood coagulation. A balance between clot formation and fibrinolysis largely determines clot stability. The structure of a fibrin clot is influenced by environmental and genetic factors, especially concentrations and functions of fibrinogen.⁶  It has been shown that formation of clots composed of compact thin fibrin fiber networks, being resistant to fibrinolysis, predisposes to arterial thrombotic events. Reduced clot permeability and impaired fibrinolysis have been reported in patients with acute myocardial infarction (MI),⁷  a history of MI,^8,9  and after cryptogenic stroke.¹⁰  Data on links between clot properties and VTE are sparse. Curnow et al¹¹  showed that hypercoagulable patients with arterial thrombosis or VTE, pregnancy complications, or autoimmune diseases have increased fibrin generation and reduced fibrinolysis. It has been demonstrated that increased D-dimer levels indicating enhanced fibrin formation and degradation represent an independent risk factor for future VTE.¹²  Hypofibrinolysis in a clot lysis assay has been shown in subjects after the first DVT episode.¹³  We hypothesized that abnormal fibrin clot features other than hypofibrinolysis can be detected in VTE patients and clot structure/function differ in relatives of VTE patients compared with controls with no history of thrombotic events, as in relatives of MI patients.¹⁴ 

The aim of this study was to investigate whether fibrin clot properties are altered in VTE patients and their relatives compared with controls.

We investigated white subjects 18 to 65 years of age in 3 categories: (1) 100 patients with a history of first-ever idiopathic VTE; (2) 100 first-degree relatives of the patients in the first category with no history of VTE, MI, or stroke; and (3) 100 age- and sex-matched asymptomatic subjects with no personal or family history of VTE, MI, or stroke. The diagnosis of VTE was established by a positive finding of color duplex sonography (visualization of an intraluminal thrombus in calf, popliteal, femoral, or iliac veins). The diagnosis of PE was based on the presence of typical symptoms and positive results of high-resolution spiral computed tomography. Patients were analyzed in subgroups according to the initial presentation of VTE (DVT alone, PE alone, or DVT combined with PE). Proximal DVT was defined as thrombosis involving the popliteal or more proximal deep vein segments. Distal (calf) DVT was diagnosed if thrombosis involved any vein segment distal to the knee joint. The VTE patients were recruited at least 7 months after a thrombotic event if anticoagulation with vitamin K antagonists was discontinued at least 4 weeks earlier. To exclude active DVT in relatives and controls at blood sampling, compression test was performed and yielded negative results.

The base population of subjects younger than 65 years with a first-ever previous idiopathic VTE episode, defined at screening as having no history of cancer, surgery, major trauma, plaster cast, or hospitalization in the past month, or pregnancy or delivery in the past 3 months, was composed of 324 patients. We excluded 108 (32.3%) patients because of deficiency of antithrombin, protein C, protein S, factor V Leiden, prothrombin 20210A mutation, lupus anticoagulant, or significant levels of anticardiolipin or anti–β₂-glycoprotein I antibodies. Another 116 patients were excluded because of the presence of other exclusion criteria, such as nephrotic syndrome (n = 1), inflammatory bowel disease (n = 2), overt malignancy (despite negative evaluation before referral; n = 5), any inflammatory states (C-reactive protein [CRP] > 10 mg/L, n = 9); international normalized ratio (> 1.2, n = 1), use of oral contraceptives or hormone replacement therapy (n = 20), atrial fibrillation (n = 2), heart failure New York Heart Association (NYHA) II-IV class (n = 5), diabetes mellitus (glucose > 7mM or use of hypoglycemic agents; n = 11), serum creatinine more than 120μM (n = 4), thrombosis other than DVT (n = 3), massive PE (n = 3), elevated plasma D-dimer levels (> 500 μg/L, n = 19), and residual thrombosis in DVT patients (n = 31); coexistence of 2 or more exclusion criteria was found in 98 patients. No cases of paroxysmal nocturnal hemoglobinuria, polycythemia vera, essential thrombocytemia, or use of thienopyridines were detected in the patients screened. Finally, 100 patients were confirmed eligible and included in the study. Relatives (n = 100) of 169 screened persons were recruited (using the same screening protocol), mainly brothers and sisters of the 72 patients, including 2 first-degree relatives of each of 28 patients and one relative of each of 44 patients. Controls (n = 100) recruited from 189 screened persons using the same screening protocol included members of the hospital staff, investigators' friends or collaborators, and their family members. Relatives and controls were matched for age, sex, smoking, and medications.

The Jagiellonian University School of Medicine Ethical Committee approved the study, and participants provided informed consent in accordance with the Declaration of Helsinki.

Fasting blood samples were drawn from an antecubital vein with minimal stasis at 8 to 10 AM. Lipid profiles, blood morphology, glucose, creatinine, and international normalized ratio were assayed by routine laboratory techniques. Blood samples (vol/vol, 9:1 of 3.2% trisodium citrate) were centrifuged at 2560g for 20 minutes, and supernatant was aliquoted and stored at −80°C. Fibrinogen was determined using the Clauss method. High sensitivity CRP was measured by nephelometry (Dade Behring). Fasting total plasma homocysteine (Hcy) levels were measured using an immunoassay (IMX System, Abbott Diagnostics). Immunoenzymatic assays were used to determine in citrated plasma prothrombin fragments 1.2 (F1.2), a marker of thrombin formation (Dade Behring); D-dimer, a marker of fibrin turnover (American Diagnostica); tissue-type plasminogen activator (tPA); and plasminogen activator inhibitor-1 (PAI-1; American Diagnostica). All measurements were performed by technicians blinded to the origin of the samples. Intra- and interassay coefficients of variation were less than 7%.

Fibrin clot permeation was determined using a pressure-driven system.^14,15  Briefly, 20mM calcium chloride and 1 U/mL human thrombin (Sigma-Aldrich) were added to 120 μL of citrated plasma. After 2 hours of incubation in a wet chamber, tubes containing the clots were connected via plastic tubing to a reservoir of a buffer (0.01M Tris, 0.1M NaCl, 1mM ethylenediaminetetraacetic acid, pH 7.5), and its volume flowing through the gels was measured within 60 minutes. A permeation coefficient (K_s), which indicates the pore size, was calculated from the equation: K_s = QxLxη/txAxΔp, where Q is the flow rate in time t; L, the length of a fibrin gel; η, the viscosity of liquid (in poise); A, the cross-sectional area (in square centimeters), and Δp, a differential pressure (in dynes per square centimeter). The interassay and intraassay coefficients of variation were 8.7% (n = 48) and 6.9% (n = 90), respectively.

Citrated plasma was mixed (3:2) with 0.7 IU/mL thrombin, 0.1% Tween 80, and 20mM calcium chloride in Tris-buffered saline, and then clots were formed in tubes prepared as in the permeation experiments.¹⁵  After centrifugation at 6000g for 60 seconds, the volume of the supernatant evacuated from the tubes was assessed on the basis of the difference in weight of the tube. Compaction was expressed as this volume divided by the initial plasma volume used to form the fibrin clot. The interassay and intraassay coefficients of variation were 7.5% (n = 60) and 7.1% (n = 82), respectively.

Plasma citrated samples were mixed 2:1 with a Tris buffer containing 0.6 U/mL human thrombin (Sigma-Aldrich) and 50mM calcium chloride to plasma initiated polymerization. Absorbance was read at 405 nm with a Perkin-Elmer Lambda 4B spectrophotometer (Molecular Devices). The lag phase of the turbidity curve, which reflects the time required for initial protofibril formation and maximum absorbance at the plateau phase (ΔAbs), indicating the number of protofibrils per fiber, was recorded.^15,16  The interassay and intraassay coefficients of variation were 7.9% (n = 96) and 6.7% (n = 150), respectively, for the lag phase, and 8.1% (n = 96) and 6.9% (n = 150), respectively, for ΔAbs.

To assess efficiency of clot lysis, 2 different methods with slight modifications were used. In the first assay,¹⁷  100 μL of citrated plasma was diluted with 100 μL of a Tris buffer, containing 20mM calcium chloride, 1 U/mL human thrombin (Sigma-Aldrich), and 14μM recombinant tissue-type plasminogen activator (rtPA; Boehringer Ingelheim). Assembly kinetics were monitored by spectrophotometry at 405 nm in a microplate reader (Molecular Devices). The time required for a 50% decrease in clot turbidity (t_50%) was chosen as a marker of susceptibility to fibrinolysis. The interassay and intraassay coefficients of variation were 8.7% (n = 90) and 7.3% (n = 120), respectively. In the second assay,¹⁸  fibrin clots, formed as for permeation evaluation, were perfused with a Tris buffer containing 0.2μM rtPA (Boehringer Ingelheim). D-dimer levels were measured every 30 minutes in the effluent. The experiment was stopped, usually after 90 to 120 minutes, whereas the fibrin gel collapsed under the pressure. Maximum rate of increase in D-dimer levels and maximum D-dimer concentrations were analyzed. The interassay and intraassay coefficients of variation were 9.7% (n = 48) and 8.1% (n = 48), respectively, for the former, and 9.4% (n = 48) and 8.3% (n = 48), respectively, for the latter variable.

The FXIII Val34Leu and α fibrinogen Thr312Ala polymorphisms were determined by the polymerase chain reaction followed by restriction fragment length polymorphism analysis as described previously.^19,20 

Data are expressed as mean plus or minus SD or as median (interquartile range), unless otherwise stated. The Kolmogorov-Smirnov test was used to assess conformity with a normal distribution. Mean values of continuous variables between the 3 groups were compared by 2-way analysis of variance followed by Tukey post-hoc test. Categorical values were analyzed using the χ² test or Fisher exact test as appropriate. For nonparametric analysis, differences between groups were assessed by the Kruskal-Wallis test. Correlations between the individual parameters were calculated using the Spearman rank correlation method. To identify independent factors, we used linear regression analysis in which a P value of .05 or less in a simple regression analysis was used as the criterion for entry into the model. A P value less than .05 was considered statistically significant.

Of the 100 VTE patients, 66 subjects had DVT without symptomatic PE (proximal DVT, n = 36), and 34 subjects had symptomatic PE (DVT combined with PE, n = 22; and PE alone, n = 12). Median time from the event was 13 months (range, 7-23 months). The mean duration of anticoagulation was 11 months (range, 4-20 months). No recurrences were observed between anticoagulation withdrawal and blood collection. A positive family history was noted in 38 VTE patients, including 11 patients with 2 or more relatives with VTE. Among the group of VTE relatives, only 18 subjects reported 2 or more first-degree family members with VTE.

All 3 groups did not differ with respect to age, sex, smoking status, arterial hypertension, lipid profile, glucose, creatinine, tHcy, CRP, and F1.2 (Table 1). VTE patients had slightly higher fibrinogen, D-dimer, tPA, and PAI-1 compared with healthy controls and relatives (Table 1). None of these variables correlated with the time from the VTE occurrence (P > .2).

The distributions of the FXIII Val34Leu genotypes were similar in all groups, for VTE patients (Val34Val, 58%; Val34Leu, 27%; Leu34Leu, 15%), for relatives (Val34Val, 55%; Val34Leu, 29%; and Leu34Leu, 16%), and for controls (Val34Val, 60%; Val34Leu, 28%; and Leu34Leu, 12%) (P > .05 for all comparisons). Similarly, no intergroup Thr312Ala genotype differences were noted with the following frequencies in VTE patients, relatives, and controls: Thr312Thr, 54%, 55%, and 56%; Thr312Ala, 35%, 33%, and 35%; Ala312Ala, 11%, 12%, and 9%, respectively (P > .05 for all comparisons). No differences in distributions of both polymorphisms were observed in subgroups of PE and DVT patients (data not shown).

Plasma fibrin clot variables significantly differed between the groups (Table 2). VTE patients had 29.5% lower clot permeability compared with controls (P < .001). A difference (by 18.2%) in this variable between VTE relatives and controls was also significant (P = .007). Clot compaction differed among groups in a similar manner, although relative intergroup differences were smaller (both P < .05). Lysis time (t_50%) was 31.9% longer in VTE patients (P < .001) and 19.2% longer in relatives (P = .012) compared with controls. Duration of the lag phase in a turbidimetric assay was similar in the 3 groups. VTE patients displayed 23.7% higher absorbance than that found in the control group (P = .004), with no significant difference between relatives and controls. Time courses of the D-dimer release from fibrin clots demonstrated that maximum rates of fibrinolysis in this assay did not differ among groups; however, maximum D-dimer levels measured in the effluent were elevated in VTE patients compared with controls with no difference between controls and relatives (Table 2).

The presence of FXIII Val34Leu and alpha fibrinogen Thr312Ala variants showed no associations with clot properties in VTE patients (data not shown). In controls and relatives, only K_s was lower in FXIII Leu34 carriers compared with Val34Val homozygous subjects (8.59 ± 1.18 vs 9.43 ± 1.09 10⁻⁹ cm²; P = .034; and 7.08 ± 1.07 vs 8.02 ± 0.89 10⁻⁹ cm²; P = .012, respectively).

The DVT and PE patients were similar with regard to demographic variables, cardiovascular risk factors, and laboratory results, including fibrinogen, CRP, tHcy, F1.2, and fibrinolytic markers (data not shown). The proportions of patients taking aspirin or statins in both groups were also similar (13 [19.7%] vs 5 [14.7%], and 14 [21.2%] vs 6 [17.6%], P > .1, respectively). The same was true for other classes of drugs (data not shown). Compared with patients with DVT alone, PE patients (with or without concomitant DVT) had 13.1% greater permeability, 15.2% lower compaction, 10.2% shorter lysis time, 10.9% lower maximum ΔAbs, and 10.7% higher maximum D-dimer levels released from clots (Table 3). Moreover, all clots made from plasma of PE patients collapsed at 80 minutes of percolation during the lysis assay, whereas in the DVT alone group, the corresponding median time was 100 minutes (range, 80-140 minutes; P = .01). Similarly, distributions of FXIII Val34Leu and α fibrinogen Thr312Ala allelic variants in DVT and PE patients as well as their relatives were similar (data not shown).

The comparison of the PE + DVT (n = 22) and PE alone (n = 12) subgroups showed no significant differences between the 2 subgroups with regard to plasma fibrin clot properties (Table 3).

Correlations of fibrin clot variables in all 3 groups are shown in Table 4. As expected, all the variables correlated with fibrinogen. None of the variables showed association with plasma F1.2 concentrations. No differences associated with sex, smoking, or any medication used were observed in clot variables. There were no associations among clot variables and the time that elapsed from VTE event and the duration of anticoagulation (r < 0.2, P > .1).

Multiple linear regression model incorporating relatives and control subjects with K_s and t_50% as dependent variables (Table 5) showed that subject status (relative vs control), fibrinogen (per 1 g/L), and CRP (per 1 mg/L) remained independent common predictors of both fibrin variables. Multiple linear regression model incorporating only VTE patients (Table 6) revealed that of age, fibrinogen, CRP, D-dimer, PAI-1, and tPA, only fibrinogen (per 1 g/L), CRP (per 1 mg/L), and D-dimer (per 100 μg/L) remained independent predictors of both K_s and t_50%. No independent predictors for all the remaining fibrin clot parameters except fibrinogen were identified. Analysis incorporating all 300 persons studied showed that being a VTE patient, fibrinogen, and CRP were identified as independent predictors of both K_s and t_50% (Table 7). Moreover, D-dimer predicted K_s, whereas tPA and PAI-1 predicted t_50%.

This study is the first to demonstrate that altered clot properties and reduced susceptibility to lysis occur not only in patients with previous VTE, but also in their first-degree relatives compared with well-matched controls with no personal or familial history of arterial or venous thrombotic events. In VTE patients with no established acquired or genetically determined thrombotic risk factors, including thrombophilic defects, we found that plasma fibrin clots contain smaller pores and are composed of thicker fibers; however, fibrils are formed after similar lag phases, compared with those made from plasma obtained from healthy persons and patients' relatives. Importantly, we provided evidence that VTE patients and their relatives share some common clot features, such as lower permeability, reduced compaction, greater maximum absorbance, and longer lysis time, compared with controls. This suggests a genetic background of idiopathic VTE. However, all these features, except maximum absorbance, were more pronounced in VTE patients, indicating that additional factors or previous thrombosis per se enhance alterations in fibrin clot variables.

Susceptibility of fibrin clots to lysis in VTE patients is probably a crucial fibrin feature from a pathophysiologic viewpoint. It is known that impaired transport of proteins through networks results from compact clot structure, and a fibrin clot with loosely packed fibers has a faster fibrinolysis rate.⁶  Using a different methodology (ie, clot lysis assay with addition of exogenous thrombin), we confirmed previous findings that impaired plasma fibrin clot lysis characterizes patients after the first episode of DVT.¹³  Another approach based on the clot permeation system appears less sensitive to VTE-associated changes in fibrin clot structure and function. A discrepancy in results of different lysis assays used in the current study has been observed previously in other clinical settings.^7,10,15  We extended available data by showing that, after exclusion of several known thrombotic risk factors, a relative decrease in fibrinolytic activity in a plasma-based fibrin assay¹⁷  is even more pronounced compared with controls. Given recent findings showing no differences in lysis time for patients after idiopathic DVT versus controls from affected families,²¹  it is probable that a choice of a clot lysis assay to determine efficiency of fibrinolysis is of importance while considering the utility of various lysis assays in clinical settings. In the current study, we failed to show differences between VTE patients and controls in the velocity of D-dimer release from fibrin clots subjected to rtPA in a perfusion assay and used successfully in other patients' groups.^8,17  Differences in concentrations of reagents and assay design appear to determine sensitivity of the assay to differentiate between thrombotic patients and controls. Further studies on larger populations are needed to establish which methodologic approach is optimal to test clot lysis.

Of note, our findings indicate a marked similarity in fibrin clot characteristics observed in the CAD patients and those with previous VTE. Given the growing evidence for the links between cardiovascular disease and VTE,²²  our data support this concept by showing that certain clot features may contribute to increased risk of VTE among patients at high cardiovascular risk. Because very recently the Tromso study has demonstrated that family history of MI is a risk factor not only for MI, but also for VTE, both total and unprovoked,²³  it is tempting to speculate that alterations in fibrin clot structure/function observed in relatives of MI¹⁴  and VTE patients might represent an additional potential explanation for a link between arterial and venous thrombosis.

An intriguing finding is a difference in fibrin properties between PE and DVT patients. Less compact fibrin structure associated with faster clot lysis seems to characterize subjects who have experienced PE, regardless of the presence of concomitant DVT or not. It might be speculated that a constellation of higher K_s, and shorter lysis time, found in PE patients, contributes to clot fragmentation followed by pulmonary embolization. This finding might help identify subjects at risk of PE. However, given the limited number of patients with PE in the current study, these novel observations should be interpreted with caution.

Mechanisms behind altered clot structure/function in VTE patients and its relatives remain unclear. The Leeds Family Study showed that genetic factors contribute modestly to variance in fibrin clot measures, including lysis time (range, 10%-40%), whereas the contribution of environmental factors, being poorly characterized, is much larger.²⁴  Increased levels of fibrinogen, which is a major predictor of fibrin clot properties, results in faster activation rate, a more dense and tight fibrin network, as well as the formation of thicker fibers.^6,16  VTE patients displayed associations between some clot features and CRP levels, which, if elevated, can reduce clot permeability and lysis as shown in patients at risk of atherosclerotic vascular disease and those with acute MI.^8,17  Similar correlations were observed in all 3 groups in the current study. Although increased Hcy levels have been reported to decrease clot permeability and susceptibility to lysis in coronary patients and healthy controls,²⁵  in VTE patients those tHcy levels did not differ from those measured in relatives and controls, no such effects have been observed indicating that in VTE patients some unknown confounders operate. Although thrombin concentrations alter fibrin structure in in vitro experiments on purified fibrinogen,²⁶  we did not observe any associations between plasma F1.2 levels and any fibrin variables in any group. Endogenous thrombin potential, but not plasma F1.2, has been shown to be associated with VTE.²⁷  It remains to be established whether endogenous thrombin potential correlates with fibrin clot properties determined in the current study. It should be highlighted that alterations in fibrin function/structure, found in VTE patients, cannot be attributed to the activation of coagulation observed in the acute phase of thrombosis because thrombin markers were similar in all 3 groups and there were no associations of the time from the VTE episode and clot features. The effect of 2 polymorphisms reported to alter fibrin clot structure/function, ie, FXIII Val34Leu²⁸  and α fibrinogen Thr312Ala,²⁹  has been ruled out in VTE patients, although other studies also yielded results suggesting such associations.⁶  However, similarities in fibrin features observed in VTE patients and relatives suggested an influence of other as yet unidentified genetic factors.

There is evidence that aspirin, statins, and angiotensin-converting enzyme inhibitors, used in the prevention and treatment of cardiovascular disease, may increase the size of clot pores and enhance clot lysis.^6,7,17,30  We failed to show differences in fibrin clot properties associated with the administration of any of these drugs in this study. Lack of drug-induced differences might be related to the patients' characteristics and a low percentage of subjects treated. Regarding statins, 2 other factors could be of importance. In the current study, statin-treated patients, who received mostly simvastatin or atorvastatin at a dose of 10 or 20 mg/day, had relatively low cholesterol levels, which have been shown to be associated with only 4.4% increase in clot permeability and 11% reduction in lysis time after 40 mg/day simvastatin.³¹ 

Our study has several limitations. The number of study participants was limited. However, all 3 groups were meticulously selected to minimize the impact of known potential confounders. We determined each variable at a single time point. However, there were no differences in clot variables depending on the time from the VTE occurrence. Exclusion of anticoagulated patients (resulting from warfarin interference with fibrin tests) and those with 2 or more VTE episodes resulted in omission of subjects who could have more pronounced fibrin alterations. Our experimental approach did not allow analysis of the effect of blood cells and platelets on fibrin clot structure/function, which can alter, for example, fibrinolysis.³²  Scanning electron microscopy of fibrin clots has not been performed. We think that functional plasma-based assays provide more valuable insights into a role of fibrin clots in thrombosis and other diseases. Plasma levels of FVIII, FIX, and FXI that have been identified as risk factors for VTE were not determined. However, to our knowledge, there have been no reports showing the impact of their increased levels on clot variables except clot lysis time determined using a tissue factor-induced and rtPA-induced plasma-based lysis assay, showing weak associations with most coagulation factors.¹³  Moreover, because lipoprotein(a) levels have been reported to adversely affect clot properties,³³  we cannot exclude that this variable, not measured in the present study, may contribute to VTE-related differences in clot properties.

In conclusion, we have shown that patients with VTE and their relatives have characteristic features of fibrin clot structure/function, such as rapid formation of dense clots resistant to lysis, and fibrin clot properties differ between PE patients and those with DVT alone. Our study provides new insights into the pathophysiology of idiopathic VTE that might have practical implications. Prospective studies are needed to assess the prognostic value of fibrin clot properties in persons before thrombotic events.

This work was supported by Jagiellonian University School of Medicine grant K/ZDS/000565 (A.U.).

Contribution: A.U. conceived and designed the study, conducted laboratory investigations, and drafted the manuscript; K.Z. designed the study, recruited patients, and interpreted data; M.C.-D. recruited patients, performed clinical workup, and interpreted data; A.L.-K. performed clinical workup and interpreted data; A.S. performed clinical workup and conducted laboratory investigations; K.C. recruited patients, conducted laboratory investigations, and interpreted data; and W.T. designed the study, revised the manuscript, and gave final approval.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Anetta Undas, Institute of Cardiology, Jagiellonian University School of Medicine, 80 Pradnicka St, 31-202 Krakow, Poland; e-mail: mmundas@cyf-kr.edu.pl.