Accelerated fibrin clot degradation is associated with arterial thromboembolism in patients following venous thrombosis: a cohort study

Several lines of evidence have suggested that patients following venous thromboembolism (VTE) are at higher risk of arterial thromboembolism (ATE). Prothrombotic fibrin clot characteristics were reported in individuals with cardiovascular risk factors. We investigated whether specific fibrin clot properties measured after 3–4 months of anticoagulation characterize VTE patients with subsequent ATE. We enrolled 320 patients following VTE aged below 70 years (median age, 46). Ten patients were lost to follow-up. ATE occurred in 21 individuals after a median 54 (31–68) months during a follow-up of 87.5 months (incidence 0.94%; 95% confidence interval [CI], 0.59–1.4 per patient-year). Patients with ATE had faster fibrin clot degradation, reflected by maximum rate of D-dimer increase during plasma clot lysis induced by tissue-type plasminogen activator (D-Drate) at baseline. Clot permeability, turbidimetric variables, clot lysis time, and thrombin generation were unrelated to ATE. Univariable Cox proportional hazards analysis showed that age, diabetes, and D–Drate were risk factors for subsequent ATE. Increased D–Drate (by 0.001 mg/L/min; hazard ratio, 1.08; 95% CI 1.02–1.14) was an independent predictor of ATE after adjustment for potential confounders. Faster fibrin clot degradation at 3 months since VTE may increase the risk of ATE among VTE patients during follow-up.

ATE during follow-up. Ten patients (3.1%) were lost to follow-up. The median follow-up period was 87.5 (77-95) months. ATE was diagnosed in 21 (6.8%, 0.9% per year) patients (9 men, 12 women) after a median follow-up of 54  months. Fourteen patients (4.5% of the whole group) had MI (including 7 ST-elevated MI and 7 non-ST elevated MI), 6 (1.9%) had ischemic stroke and 1 patient was diagnosed with peripheral arterial disease complicated with acute popliteal artery occlusion. The incidence rate of ATE was 0.94% (95% CI, 0.59-1.4%) per patient-year. The corresponding incidence rates of MI and ischemic stroke were 0.62% (95% CI, 0.36-1.02%) per patient-year and 0.27% (95% CI, 0.11-0.56%) per patient-year (Table 1). Five subjects of the 21 ATE patients (23.8%) were also diagnosed with recurrent VTE. In all of them ATE was observed after recurrent VTE when patients were on anticoagulation.
Patients who experienced ATE during follow-up were older by about 6 years and more frequently suffered from diabetes. There were no other intergroup differences in demographics or comorbidities (Table 1, Supplemental Table 2). Regarding routine laboratory investigations patients with ATE during follow-up had 23% lower high-density lipoprotein cholesterol compared to the remainder (1.14 [1.01-1.55] mmol/L versus 1.48 [1.21-1.73] mmol/L; p = 0.009). There were no differences in fibrinogen, D-dimer, and thrombin generation between patients with ATE and the remainder.
When we divided patients into three groups, i.e. patients with recurrent VTE, patients with ATE, and those free of both ATE and recurrent VTE (Table 1), we observed that patients with ATE were older than those from the latter group. Patients with ATE had 26 Table 2).
Fibrin clots and ATE. Patients with diagnosed ATE compared with the remainder had baseline lower K s which reflects a smaller average pore size in fibrin networks, as well as a higher rate of increase in D-dimer levels (D-D rate ) ( Table 2, Fig. 1). The two groups did not differ in lag phase, ΔAbs, maximum D-dimer concentrations (D-D max ), and CLT (  1) compared to the remainder (n = 289), without any differences in other fibrin variables, including K s . We compared patients free of both recurrent VTE and ATE during follow-up with those with recurrent VTE and those with ATE. We found that the latter group had 11.6% higher D-D rate and 25% shorter CLT than those with recurrent VTE (Table 2, Fig. 2).
There were no differences in fibrin clot properties between patients with MI and ischemic stroke (data not shown).
Fibrin clot and recurrent VTE. Individuals with recurrent VTE (n = 83) had 12% lower K s , 9.1% lower lag phase and 23.5% longer CLT compared to subjects without both recurrent VTE and ATE during followup (Table 2, Fig. 2). Patients with recurrent VTE compared with those free of recurrent VTE, regardless ATE differed significantly in fibrin clot properties (Supplemental Table 4). The results correspond to the findings obtained after a median 44 months of follow-up 20 . Previously, seventy-seven (25%) subjects diagnosed previously with recurrent VTE were characterized by 12% lower K s, 9% shorter lag phase in the fibrin formation assay and 25% longer CLT 20 .

Discussion
The present study is the first to assess a comprehensive set of plasma fibrin clot properties in a cohort of VTE patients as potential risk factors for ATE in order to address the hypothesis of the fibrin-related mechanisms linking VTE and ATE. Contrary to our expectations, we demonstrated that in a largely middle-aged cohort of VTE patients an increased rate of plasma fibrin clot degradation expressed as higher D-D rate characterized patients with subsequent ATE 27 . The study showed that other fibrin clot properties, including clot permeability and CLT extensively explored in VTE and cardiovascular disease, did not identify young and middle-aged VTE patients at an increased risk of ATE in the future. Our findings suggest that faster enzymatic fibrin clot degradation could reflect the instability of fibrin networks prone to fragmentation in vivo and could increase the risk of ATE. This study appears to indicate that the prothrombotic clot phenotype observed in young and middle-aged subjects after 3 months following VTE might be associated with reduced risk of ATE in the following years. The current observations suggest the complex role of fibrin clot structure and function in the pathophysiology of thromboembolism. Table 2. The comparison of fibrin clot proprieties. K s, fibrin clot permeability; ΔAbs, maximum absorbance at the plateau phase; D-D max, maximum D-dimer concentrations; D-D rate , rate of increase in D-dimer levels; CLT, clot lysis time; for other abbreviations see Table 1. Values are given as median (interquartile range). In terms of ATE two patients excluded from the previous analysis were included in the current follow-up study. *Five patients with both recurrent VTE and ATE were included into recurrent VTE group as this event occurred first. † Indicates statistical difference between patients without both VTE and ATE, and patients with recurrent VTE. ‡ Indicates statistical difference between patients with recurrent VTE and patients with ATE. The presented differences between groups remained statistically significant after adjustment for age, sex, diabetes and fibrinogen.
Of paramount importance is the plasma-based assay in which we were able to show faster clot lysis in patients who developed ATE during follow-up. In the assay introduced by Collet et al. 28 in 1999 to analyze fibrin properties in nephrotic patients and healthy controls, the previously formed plasma clot was perfused with a buffer containing high tissue-type plasminogen activator (tPA) concentration, similar to that observed in patients treated with tPA-based thrombolysis 29 . Our modification of the original approach was used for the first time to assess the impact of statins and other drugs on fibrin clot properties in 2006 30 . This assay was also applied by our group to evaluate efficiency of fibrinolysis in patients with ATE. We found higher D-D rate and D-D max in patients with cryptogenic ischemic stroke as compared to the control group 23 . In contrast, lower D-D rate was observed in MI patients who survived in-stent thrombosis 31 and patients with a history of limb ischemia of unknown cause as compared to the control group 32 . In line with the present study, this measure of clot lysis did not predict recurrent VTE 20 . Our present observation suggests that various lysis assays should be used to highlight specific fibrin clot abnormalities in a particular disease with lysis abnormalities, without any single test appropriate to all conditions in which fibrinolysis is disturbed as shown in previous studies [32][33][34] . In our study the modified assay by Collet et al. 28 enabled to show subtle changes in fibrin clot lysability that could be of importance in ATE prediction, while other commonly used assays did not differentiate subjects with VTE at risk of ATE. Importantly we did not observe any differences in CLT between patients who developed the subsequent ATE and the remainder.
Mechanisms behind the observed link between ATE and faster clot degradation remain unclear. Atherogenic effects of fibrin degradation products have been demonstrated in several studies. Corban et al. 35 have reported that fibrin degradation products were associated with larger atherosclerotic plaques and necrotic core areas. They suggested that higher fibrin degradation products might be a marker of subclinical rupture or erosion of the  www.nature.com/scientificreports/ plaque 35 . Moreover, fibrin was found to stimulate the production of proinflammatory molecules, interleukin-1, interleukin-8 and intracellular adhesion molecule-1 [36][37][38] Fibrin degradation products stimulate the migration of monocytes and local fibrinolysis 39 .
On the other hand, the role of fibrin in atherothrombosis is complex. It has been suggested that fibrinogen and fibrin are involved in the development of early atherosclerotic lesions and their progression [39][40][41] . In hyperlipidemic and hypercoagulable mouse models (mice carrying the factor V Leiden and mice being thrombomodulin mutants) Seehaus et al. 27 observed that hypercoagulabity leaded to larger atherosclerotic plaques and plaque stability with less necrotic cores, and that anticoagulant treatment reduced plaque stability 27 . Borissoff et al. 42 reported enhanced procoagulant state and higher endogenous thrombin potential in the early atherosclerotic lesions as compared to stable advanced atherosclerotic lesions. They concluded that blood coagulation and the resultant increased formation of fibrin contribute to a more stable atherosclerotic plaques 42 . In animal studies it has been shown that in mice less prothrombotic phenotype was associated with reduced atherosclerosis or less early-stage atherosclerotic lesions 43 . Based on the current findings, it might be speculated that faster lysis with enhanced fragmentation of fibrin meshworks can predispose to embolic events. Moreover, faster clot degradation could be associated with more vulnerable plaques with more fragile fibrin deposits on the surface of atherosclerotic lesions, and therefore it could be a risk factor of ATE. It remains to be established whether similar observation can be made in all subjects at risk of ATE beyond those with a history of VTE episode.
In our cohort there were no differences in the risk of ATE between patients with unprovoked VTE versus provoked VTE. This finding, though based on a relatively small number of patients, was contrary to our expectations, because several studies, including its meta-analysis, have shown that patients after unprovoked VTE were at higher risk of ATE compared to those with provoked VTE, however some reports failed to observe such differences related to the type of VTE 4,6,44 . Of note, we observed an increased fibrinogen concentration at 3 months since the event in patients with provoked VTE. Elevated fibrinogen levels are observed commonly in patients with cardiovascular disease and show associated with cardiovascular risk factors 45,46 . This parameter is also the key determinant of fibrin clot properties 18 .
Several study limitations should be acknowledged. The study population was limited, however this cohort was well described with a large set of hemostatic parameters and followed for, on average, more than 7 years. All laboratory measurements were done in a single point time, at 3 months and this time point is of key importance in clinical decision making following VTE. Changes over time in all the variables measured cannot be ruled out, however in our opinion the impact of the results obtained specifically after 3 months of anticoagulation has been shown suggesting persistent abnormalities affecting clinical outcomes during follow-up 19,23 . The influence of drugs including ASA on fibrin clot properties is possible 18 . We did not perform microscopic assessment of plasma clots, but in our previous studies D-D rate showed no association with fiber diameter or pore size in clots obtained using scanning electron microscopy. The current findings cannot be easily extrapolated to elderly VTE patients in whom ATE and VTE occur more commonly, since we excluded such patients from this study. The same holds true for anticoagulated VTE patients because the majority of the current cohort stopped anticoagulation after a few months of treatment except those who developed recurrent VTE episodes.
In conclusion, we demonstrated that patients with ATE which is experienced a few years since VTE are not characterized by the prothrombotic fibrin clot phenotype in a cohort of patients aged 70 years or less, which is in contrast to the association between such phenotype and recurrent VTE. The finding suggesting that VTE patients with subsequent ATE during follow-up had a higher rate of clot degradation at high concentrations of rtPA, measured in vitro after 3 months of anticoagulation supports growing evidence for fibrin involvement in a wide spectrum of thromboembolic episodes. It needs further investigations whether D-D rate determined at 3 months since the index VTE may help identify VTE patients who need close surveillance for ATE and assessment of cardiovascular risk. Further studies should be performed to corroborate our results in larger cohorts and elucidate pathophysiological mechanisms behind the contribution of fibrin degradation and the clinical outcomes.

Patients and methods
A total of 368 patients with a history of VTE were screened for meeting the eligibility criteria for the study between October 2008 and June 2010. The cohort was described in detail in our previous paper 20 . Briefly, the eligible patients following the first-ever isolated DVT or combined with pulmonary embolism were recruited among those referred to our center for diagnostic work-up. The diagnosis of DVT and pulmonary embolism was established as reported 20 . We excluded patients with known cancer, severe thrombophilia, recent acute ATE, acute infection or severe kidney or liver failure. Patients with VTE received standard anticoagulant treatment with vitamin K antagonists for 3 to 12 months according to the guidelines 47 .
At enrolment data on cardiovascular risk factors were collected. Hypertension was regarded as increased blood pressure (systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg) or antihypertensive treatment. Hypercholesterolemia was defined as a total cholesterol level above 5 mmol/l, low-density lipoprotein cholesterol above 3 mmol/l or statin treatment; diabetes as fasting glucose of ≥ 7.0 mmol/l, non-fasting glucose ≥ 11.1 mmol/l, antidiabetic treatment or previously diagnosed diabetes. Obesity was defined as body mass index equal to or greater than 30 kg/m 2 . Heart failure was recognized based on typical symptoms and reduced left ventricular ejection fraction (< 40%).
All experimental protocols were approved by the Bioethics Committee of the Jagiellonian University. From all study participants the informed written consent in accordance with the Declaration of Helsinki was obtained. All methods were carried out in accordance with relevant guidelines and regulations.

Follow-up.
Patients were followed-up on a 6-month basis since enrolment. Clinical data was collected every six months via a visit in the outpatient clinic or phone calls. www.nature.com/scientificreports/ The primary composite endpoint was ATE defined as ischemic stroke, MI or peripheral arterial thromboembolic event. Stroke was diagnosed based on persistence of typical symptoms for more than 24 h confirmed on magnetic resonance imaging or computed tomography 48 . The diagnosis of MI was established based on typical symptoms, changes on the electrocardiogram and increased myocardial necrosis biomarkers 49 . Peripheral arterial thromboembolic event was defined as acute peripheral artery occlusion 50 . ATE episodes related to invasive procedures were excluded. The secondary endpoint of the study was symptomatic VTE diagnosed based on positive findings of color duplex sonography. Recurrent DVT in the same leg as the index event was diagnosed when new no-compressibility of venous segment was observed or when there was an increase of at least 4 mm in the residual diameter. Anticoagulation treatment was prescribed again in patients with recurrent VTE. Laboratory investigations. After 3 months (12 to 15 weeks) of anticoagulant treatment since DVT, antecubital blood samples for laboratory investigation were taken from fasting patients in the morning hours (8 to 10 AM). Patients who were treated with vitamin K antagonists were first temporarily switched to a low-molecularweight heparin for 10-14 days. Blood samples were taken after 16-24 h since the last injection. We assessed blood cell count, lipid profiles, glucose, creatinine, and international normalized ratio using routine laboratory techniques. Firstly, blood samples (vol/vol, 9:1 of 3.2% trisodium citrate) were centrifuged at 2000 × g for 10 min within 30 min of the draw, then we removed supernatant, aliquoted and stored it at − 80 °C until analysis. We determined fibrinogen using the Clauss method. High-sensitivity C-reactive protein was assessed by nephelometry (Siemens). We measured plasma concentrations of D-dimer, tPA and plasminogen activator inhibitor-1 antigens using immunoenzymatic assays (American Diagnostica). Thrombophilia screening was conducted as described 51 .
Fibrin clot permeability. Fibrin clot permeation was assessed as described 30 . Briefly, we mixed 60 µl of plasma with 60 µl of the coagulation trigger containing 1 IU/ml human thrombin and 20 mM CaCl 2 . Immediately, 100 µl of prepared assay was transferred to a plastic cylinder made from a serological pipette (Sarstedt, Nümbrecht, Germany). Subsequently, after 2 h of incubation at a room temperature, we connected tubes containing the clots via plastic tubing to a reservoir of a buffer (0.05 M tris-HCl, 0.15 NaCl, pH 7.5). The volume of a buffer flowing through the gels was measured within 60 min. Bromophenol blue was used after experiments to find potential leaks, thanks to this procedure we were able to detect and discard the defective clots. We calculated the permeability coefficient, an indirect measure of the average pore size in the fiber network, using the equation: K s (× 10 -9 cm 2 ) = Q × L × η/t × A × ∆P, where Q (cm 3 ) is the flow rate at time t (s), L (cm) is the length of the fibrin gel, η (dyne × s/cm 2 ) is the viscosity of the liquid, A (cm 2 ) is the cross-sectional area and ∆p (dyne/cm 2 ) is differential pressure. The intraassay coefficient of variation was 6.8%.

Turbidity measurements.
To initiate polymerization plasma citrated samples from each patient were mixed 2:1 with a Tris buffer containing 0.6 IU/mL human thrombin (Sigma) and 50 mM CaCl 2 . We used spectrophotometer to read absorbance (ΔAbs) at 405 nm. A lag phase as time to the start of fibrin polymerization, slope of the polymerization curve, along with maximum absorbance at plateau were assessed 19 .
Plasma clot lysis assays. Fibrinolysis efficiency was examined using two assays at 2 various concentrations of recombinant tPA (rtPA) 30,52 . In the first assay, CLT was assessed. As previously described 20 , 100 µl citrated plasma was mixed with 15 mmol/L CaCl 2 , 0.6 pM human tissue factor (Innovin, Siemens), 12 µmol/L phospholipid vesicles and 60 ng mL -1 tPA (rtPA, Boehringer Ingelheim, Ingelheim, Germany). The mixture was transferred to a microtiter plate. Absorbance was measured at 405 nm at 37 °C. CLT was determined as the time from the midpoint of the clear-to-maximum-turbid transition (clot formation), to the midpoint of the maximum-turbid-to-clear transition (the lysis of the clot).
In the second assay, fibrin clot lysis was assessed using a dynamic lysis assay according to Collet et al. 28 with some modifications 30 . Fibrin clots obtained in the same manner as for fibrin clot permeation were washed with Tris buffer and perfused with the same buffer containing 0.2 µmol/L rtPA (Boehringer Ingelheim). The lysis rate was determined by fibrin degradation reflected D-dimer levels in the effluent using ELISA (American Diagnostica). The D-dimer level was assessing at 20 min intervals until when the fibrin gel collapsed under the pressure, usually after 80-120 min. We measured the maximum rate of increase in D-dimer levels (D-D rate ) and maximum D-dimer concentrations (D-D max ) as previously described 20 . Calibrated automated thrombogram (CAT). The commercial reagents (Thrombinoscope, BV, Maastricht, Netherlands) were used to conduct the CAT assay 19 . Shortly, 20 µl of a starting reagent containing 5 pM recombinant relipidated TF, 4 mM phospholipids, 100 mM CaCl 2 and 2.5 mM fluorogenic substrate was added to 80 µl of plasma sample. We determined the 3 following variables: the peak thrombin, the endogenous thrombin potential, and the time to thrombin using the Fluoroskan Ascent microplate fluorometer (Thermo Fisher Scientific, Vantaa, Finland).
Statistical analysis. Continuous variables were presented as mean ± standard deviation or median (interquartile range). The Shapiro-Wilk test was used to verify the assumption of the normal distribution of continuous variables. The Student's or the Welch's t-test based on the equality of variances for normally distributed variables were used to compare two groups. The Mann-Whitney U-test was performed to compare two groups for non-normally distributed continuous variables. The categorical (qualitative) variables were presented as the number (percentages) and the Chi-squared test (or Fisher exact test) was used to compare them between groups www.nature.com/scientificreports/ Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.