Thrombosis Research 133 (2014) 736–742

Contents lists available at ScienceDirect

Thrombosis Research

j ourna l homepage: www.e lsev ie r .com/ locate / th romres

Regular Article

Increased adhesive properties of neutrophils and inflammatory markersin venous thromboembolism patients with residual vein occlusion andhigh D-dimer levels

Kiara C.S. Zapponi a,⁎, Bruna M. Mazetto a, Luis F. Bittar a, Aline Barnabé a, Fernanda D. Santiago-Bassora a,Erich V. De Paula b, Fernanda A. Orsi a, Carla F. Franco-Penteado a,Nicola Conran a, Joyce M. Annichino-Bizzacchi a

a Hematology and Hemotherapy Center, University of Campinas - UNICAMP, São Paulo, Brazilb Department of Clinical Pathology, University of Campinas - UNICAMP, São Paulo, Brazil

Abbreviations: VTE, Venous thromboembolism; RVO,⁎ Corresponding author at: Hematology andHemotherap

Campinas, SP 13083–878, Brazil. Tel.: +55 19 35218755.E-mail address: kiarazapponi@gmail.com (K.C.S. Zapp

http://dx.doi.org/10.1016/j.thromres.2014.01.0350049-3848/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 4 September 2013Received in revised form 12 December 2013Accepted 28 January 2014Available online 6 February 2014

Keywords:InflammationVenous thromboembolismCell adhesionPlateletsNeutrophilsErythrocytes

Background: Venous thromboembolism (VTE) develops via a multicellular process on the endothelial surface.Although widely recognized, the relationship between inflammation and thrombosis, this relationship hasbeen mostly explored in clinical studies by measuring circulating levels of inflammatory cytokines. However,the role of inflammatory cells, such as neutrophils, in the pathogenesis of VTE is not clear in humans.Aims: To evaluate the adhesive properties of neutrophils, erythrocytes andplatelets in VTEpatients and to correlatefindings with inflammatory and hypercoagulability marker levels.Methods: Study group consisted of twenty-nine VTE patients and controls matched according to age, gender andethnic background. Adhesive properties of neutrophils, erythrocytes and platelets were determined using a staticadhesion assay. Neutrophil adhesionmolecules expressions were evaluated by flow cytometry. Inflammatory andhypercoagulability marker levels were evaluated by standard methods. Residual vein occlusion (RVO) was evalu-ated by Doppler ultrasound.Results:No significant difference could be observed in platelet and erythrocyte adhesion between VTE patients and

controls. Interestingly, VTE patients with high levels of D-dimer and RVO, demonstrated a significant increase inneutrophil adhesion, compared to controls and remaining patients. Inflammatory markers (IL-6, IL-8, TNF-α)were also significantly elevated in this subgroup, compared to other VTE patients. Adhesive properties of neutro-phils correlatedwith IL-6 andD-dimer levels. Neutrophils adhesionmolecules (CD11a, CD11b and CD18)were notaltered in any of the groups.Conclusion: These findings not only support the hypothesis of an association between inflammation and hyper-coagulability, but more importantly, highlight the role of neutrophils in this process.

© 2014 Elsevier Ltd. All rights reserved.

Introduction

Venous thromboembolism (VTE) comprises deep venous thrombosis(DVT) and pulmonary embolism (PE). VTE is a common multifactorialdisease with an annual incidence of 1–3:1000 in the general population[1,2]. Its pathogenesis depends on the interaction of genetic and acquiredfactors [3], and despite several already-identified risk factors, 30-40% ofcases are considered idiopathic [4,5]. Many VTE patients experienceresidual vein occlusion (RVO) [6,7]; in addition, several authors havedemonstrated that RVOandhypercoagulability, characterized by elevatedD-dimer levels, may be associated with the risk of recurrence [8–12].

Residual vein occlusion.y Center, Rua Carlos Chagas 480,

oni).

The relationship between inflammation and coagulation is bidirec-tional. Inflammation leads to the activation of coagulation, and coagula-tion also considerably affects inflammatory activity [13]. Accordingly, ithas been demonstrated that the interaction of the thrombus with thevenous wall originates a locoregional inflammatory response and cyto-kine release [14].

Despite the fact that the link between inflammation and thrombosishas been widely recognized, this relationship has beenmostly exploredin studies about the interactions between the inflammatory cytokinesand the coagulation cascade [15–19]. Of note, inflammatory cells suchas circulating neutrophils have been mostly studied in animal modelsof VTE, however, new insights about these cells are necessary to bettercomprehend its role in physiopathology of VTE in humans.

Thrombus formation is a multicellular process and develops on thesurface of the endothelium, presenting a laminar structure consisting ofplatelets, leukocytes, erythrocytes and fibrin [20–23]. A recent study,

737K.C.S. Zapponi et al. / Thrombosis Research 133 (2014) 736–742

using intravital microscopy, demonstrated that reduction of blood flowinduces a proinflammatory endothelial phenotype that initiates neutro-phil recruitment. Recruited neutrophils start fibrin formation via bloodcell-derived tissue factor (TF),which is the decisive trigger to themassivefibrin deposition that is characteristic of DVT. Furthermore, platelets arecritical for DVT propagation, as they support neutrophil accumulationand promote neutrophil extracellular trap formation, which also sup-ports thrombus propagation [21].

Therefore, the aim of our studywas to evaluate the adhesive proper-ties of circulating neutrophils, erythrocytes and platelets in a cohort ofpatients with VTE, and to correlate our findings with inflammatoryand hypercoagulability markers.

Materials and Methods

Patients and Controls

Twenty-nine patients from the Hemostasis Outpatient Clinic of theUniversity of Campinas, with history of VTE, were consecutive enrolledbetween June of 2010 and July of 2012. The exclusion criteria wereadopted in order to provide a more homogenous group of patientsand to exclude pre-existing inflammatory or immune conditions thatcould confound the results. They were: age below 18 and above70 years-old at the time of VTE, thrombosis at an unusual site, cancer,severe liver or kidney diseases, autoimmune diseases includingantiphospholipid antibody syndrome, pregnancy, use of any anticoagu-lant and/or antiplateletmedicationwithin 10 days prior to sample collec-tion. The healthy individuals, included as controls, were matched bygender, age and ethnicity and were selected from the same geographicregion. Exclusion criteria for controls were the same as for patients.This studywas performed in accordancewith the Declaration of Helsinkiandwas approved by the ethics committee of the University of Campinas(CAAE: 0281.0.146.000-10).

Materials

Fibronectin (FN) from human plasma, human thrombin (TB),Histopaque®-1119, Histopaque®-1077, Human fibrinogen (FB) andDrabkin’s solution, were bought from Sigma-Aldrich (St Louis, MO,USA). Alexa Fluor® 488 mouse anti-human CD11b (Mac-1), FITCmouse anti-human CD11a (LFA-1) and PE mouse anti-human CD18were purchased from BD Biosciences Pharmingen (San Jose, CA, USA).The following ELISA kitswere also purchased; IL-8 (BDOptEIATM, BDBio-sciences Pharmingen, San Diego, CA, USA), IL-6 and TNF-α (Quantikine,R&DSystems, Minneapolis, MN, USA).

Sample Collection

Bloodwas obtained from participants at the time of their enrollment.Cell-adhesion and flow cytometry assays were performed on the sameday, as detailed below. For the evaluation of inflammatory markers andD-dimer levels, serum and plasma samples were obtained and storedat −80 °C until analysis.

Red Cell Separation and Static Adhesion Assay

Red cell of peripheral blood was isolated (50 μL, 2x108 cells/mL, inHBSS) and seeded in triplicate onto 96-well plates previously coatedwith 60 μL FN (20 μg/mL FN in PBS) and cells were allowed to adherefor 30 min, 37 °C, 5% CO2. Following incubation, non adherent cellswere discarded and plates were washed thrice with PBS, before adding50 μL HBSS to each well. Varying concentrations of the original cellsuspension were added to empty wells to form a standard curve. Fiftymicroliters of Drabkin’s solution was added to each experimental welland the wells of the standard curves; absorbance was then measuredat 540 nm (Versamax; Molecular Devices, Sunnyvale, CA, USA),

following incubation for 10 min at 37 °C. Adherence was calculated bycomparing absorbance of unknowns with those of the standard curve[24].

Preparation of Washed Platelets and Static Adhesion Assay

The preparation of washed platelets and static platelet adhesionassay were based on a previously reported method [25]. Briefly, plate-lets (50 μL of 1x108 platelets/mL in Krebs solution) were seeded onto96-well plates previously coated with 50 μg/mL FB, in the presence orabsence of TB (5 μL); cells were allowed to adhere for 15 minutes at37 °C. Following incubation, non-adherent platelets were removed(gentle washing with Krebs solution). Wells containing adherent plate-letswere incubatedwith acid phosphatase substrate solution (0.1 mol/Lcitrate buffer, pH 5.4, containing 5 mmol/l p-nitrophenyl phosphate and0.1% Triton X-100) for 1 h at room temperature (RT). The reaction wasstopped by the addition of 2 mol/L NaOH and p-nitrophenol productionwas determinedwith amicroplate reader at 405 nm (Versamax;Molec-ular Devices, Sunnyvale, CA, USA). The percentage of adherent cells wascalculated by comparing acid phosphatase activity to that of a standardcurve with known numbers of platelets from the same individual.Platelet acid phosphatase activity is not affected by the functional stateof the cell or by agonist stimulation [26]. As conditions that restrictplatelet aggregation we utilized absence of flow/stirring and presenceof Mg2+ [27]. Thus, platelets quantified were deemed to be adheredplatelets and not aggregated platelets.

Isolation and Neutrophil Adhesion Assays from Peripheral Blood

Neutrophils were isolated as described previously [28], and the sus-pensions of these cells with purities of greater than 92% were used im-mediately in assays; contaminating cells were mainly lymphocytes andeosinophils. Briefly, neutrophils (50 μL of 2x106 cells/mL in RPMI) wereseeded onto 96-well plates previously coatedwith 20 μg/ml fibronectin(FN); cells were allowed to adhere for 30min (37 ºC, 5% CO2). Followingincubation, nonadhered cells were discarded and wells washed twicewith PBS. RPMI (50 μl) was added to each well and varying concentra-tions of the original cell suspension were added to empty wells toform a standard curve. Percentage cell adhesionwas calculated bymea-suring themyeloperoxidase content (MPO) of eachwell and comparingto a standard curve of cells in suspension, determinedwith amicroplatereader at 492 nm (Versamax; Molecular Devices, Sunnyvale, CA, USA).

Ultrasonographic Assessment of RVO

Doopler ultrasound assessment of RVO was performed only on par-ticipants who had had DVT. Vascular Doppler imaging was assessed bythe same physician using a Siemens Sonoline G40 (Siemens Healthcare,Mountain View, CA, USA), with a convex transducer of 3–5 MHz andlinear 5–10 MHz. The incompressible venous segment without colorflow was considered as presenting RVO [29].

Plasma D-dimer Levels

Plasma D-dimer levels was measured by an immunoturbidimetricassay (BCS-XP, Siemens Healthcare, Marburg, Germany). Levels below0.55 mg/L were considered as normal.

Serum Levels of Inflammatory Biomarkers

Serum leveIs of IL-8, IL-6 andTNF-αweremeasuredusing commerciaIenzyme-linked immunosorbent assay (ELlSA). Serum high sensitive CRP(hs-CRP) levels were measured with a nephelometric method, usinga BN ProSpec System (Plasma Protein Analysis- Siemens HealthcareDiagnostics, Marburg, Alemanha). IL-6, IL-8, TNF-α and CRP were evalu-ated only in the VTE patients.

738 K.C.S. Zapponi et al. / Thrombosis Research 133 (2014) 736–742

Flow Cytometry

Isolated neutrophils (1 × 106 cells/mL in PBS) were incubated withAlexa Fluor® 488 mouse anti-human -CD11b (Mac-1), FITC mouseanti-human CD11a (LFA-1) and PE mouse anti-human CD18, for30 min, at 4 °C, in the dark. After washing in PBS, cells were fixed in1% paraformaldehyde/PBS and then analyzed (10,000) at 488 nm on aBectonDickinson FACScalibur (BD Biosciences, EUA) and CellQuest Soft-ware was used for acquisition. Data are expressed as mean fluorescenceintensities (MFI) compared to a negative isotype control.

Statistical Analysis

All assays were performed in triplicate for individual samples anddata are presented as median and range for each group (n refers tothe number of individuals in each group). Datawere compared betweengroups using the Mann–Whitney and Kruskal-Wallis nonparametrictests. The Wilcoxon matched pairs test was used to compare groupsbefore and after treatment with specific agents. Categorical data wereanalyzed using the Chi-square test. Correlation was analyzed using theSpearman’s test. Statistical significance was established as p ≤ 0.05.Commercial statistical software (R Development Core Team version2.0, 2011, Vienna, Austria) was used.

Results

Demographic and Clinical Data

During the enrollment period, 385 DVT patients were attended, 356could not be enrolled for the study because of the exclusion criteria orbecause they denied participating, the remaining 29 (7.53%) were in-cluded. Additionally, we included 29 healthy individuals as controls atthe same time the patients were included. Demographic parameterssuch as age, gender and ethnicity were similar between patients andcontrols, as indicated in Table 1.

VTE was provoked in 11 patients (37.93%) and the risk factors wereexposure to estrogens (pregnancy, hormonal contraceptive, puerperi-um and hormone replacement therapy; n = 6), surgery (n = 3) andlong-distance travel (n = 2). Hereditary thrombophilias were presentin three patients (10.34%), 2with the Factor V Leiden heterozygousmu-tation and 1 with the factor II 20210A heterozygous mutation. Patients'clinical and laboratory data are summarized in Table 1.

Table 1Demographic and clinical data.

Demographic data Controls (N = 29) Patients (N = 29) P

Female/male 17/12 17/12 1Caucasian/non-Caucasian 16/13 16/13 1Age (years) 43.45 (21–72) 44.52 (20–72) 0.76Time after VTE (months) – 34.59 (12–72) –

Clinical and laboratory features of patientsProximal DVTa – 24 (82%) –

Spontaneous VTE episode – 18 (62%) –

Provoked VTE episode – 11 (37.93%) –

Hereditary thrombophilia 3 (10%)

RVOb + higher D-dimer levels – 15 (51%) –

Comorbidities 5 8 0.058Hypertension 4 7Dyslipidemia 1 1Diabetes – 1Hypothyroidism – 3

Gender and ethnic are expressed as number of individuals. Age and *time after VTE areexpressed as median (range). Clinical and laboratory data are expressed as cases number(percentage). Comorbidities are expressed as number of individuals. The P values werecalculated by chi-square and Mann–Whitney tests.

a DVT (deep vein thrombosis).b RVO (residual vein occlusion).

From the 29 patients evaluated, 15 presented RVO and 14 didpresent RVO. The median time from the VTE event was 27 monthsand 29 months in RVO positive vs. RVO negative patients respectively(P = 0.678). None patient with RVO presented diabetes or any inflam-matory disease. However, 1 presented hypothyroidism, 3 presented arte-rial hypertension and 1 presented both diseases. Three patients withoutRVO presented arterial hypertension, from those 1 had also diabetes, 1presented hypothyroidism and 1 dyslipidemia. In the control group, 4subjects presented hypertension and 1 presented dyslipidemia. As bothgroups of study presented similar clinical conditions, we consider thatprobably these diseases did not interfere on our results.

Evaluation of Adhesion of Erythrocytes and Platelets from VTE Patients toFibronectin and Fibrinogen in vitro

As shown in Fig. 1 and Table 2, the in vitro static adhesion propertiesof erythrocytes to FN fromVTE patients were similar to those of controls(Table 2 and Fig. 1.A). In addition, platelet adhesive properties to FB,both basal and after thrombin stimulation, were also similar betweenpatients and controls (Table 2 and Fig. 1.B). When evaluated in a sub-group of patients that presented high D-dimer levels and RVO (n =15; 51%), there was not also a difference in the adhesive properties oferythrocytes and platelets compared to controls and remaining patients(Table 3).

As expected, platelet adhesion to FBwas significantly increased afterthe thrombin stimulus (50 mU/ml) both in patients and controls(Fig. 1.C).

The Adhesive Properties of Neutrophils are Increased in VTE Patients withRVO and High D-dimer Levels

Adhesive properties of neutrophils to FN were also similar betweenVTE patients and controls (Table 2 and Fig. 2.A). However, when evalu-ated in a subgroup of patients that presented high D-dimer levels andRVO (n = 15; 51%), a significant increase in the adhesive properties ofneutrophils could be observed, when compared to controls and remain-ing patients, respectively (24.68%, 19.29%, 18.23%, P= 0.0179,) (Table 3and Fig. 2.B). Of note, the absolute neutrophil count was not differentbetween patients and controls (3.57 × 103/μL vs 3.51 × 103/μL, P =0.912 respectively).

Inflammatory Markers are Significantly Elevated in Patients with IncreasedNeutrophil Adhesion Properties. Neutrophils Adhesion Properties Correlatewith IL-6 and D-dimer Levels

In the fifteen patients with increased neutrophil adhesion proper-ties, levels of IL-6, IL-8 and TNF-α were significantly increased whencompared with the remaining 14 patients (Table 4). In addition, therewas a correlation between neutrophil adhesive properties to FN andlevels of IL-6 (r = 0.3815; P = 0.0375), and D-dimer (r = 0.3831;P = 0.0367) in VTE patients (Fig. 3), but there was not correlationwith IL-8, TNF-α and CRP (not shown).

Expression of Thee Adhesion Molecules, LFA-1 (CD11a/CD18) and MAC-1(CD11b/CD18), on the Surface of Neutrophils

In order to assess the mechanism of the increased adhesive proper-ties of neutrophils to FN, the expressions of LFA-1 and MAC-1 integrinswere determined in neutrophils from patients and controls by flowcytometry (Table 5). As demonstrated in Table 5, the LFA-1 and MAC-1integrins were not differently expressed on the neutrophil surface ofpatients that presented high D-dimer levels and RVO, when comparedto controls and remaining patients, both in terms of mean fluorescenceintensities (MFI) (Table 5) or percentage of positive cells (not shown).

Fig. 1. (A) Basal static adhesion of erythrocytes to fibronectin (FN) in VTE patients compared to controls. (B) Platelet adhesion to fibrinogen (FB) in VTE patients compared to controls,before and after TB stimulation. (C) Increased platelet adhesion to FB after TB-stimulation in controls andVTE patients. Results are expressed as % 0 cells adhered to FN or FB. Bars indicatedmedian values. TB = thrombin. Figure (A) (B) the P values were calculated by Mann Whitney U test, figure (C) was calculated by Wilcoxon test.

739K.C.S. Zapponi et al. / Thrombosis Research 133 (2014) 736–742

Discussion

Although several acquired or genetic risk factors contributing to thepathogenesis of VTE have been described, in about 30-40% of cases,none of these factors can be identified [4,5]. In recent years, the inter-play between coagulation and inflammation has been acknowledgedas an important pathogenic factor for VTE [13,14]. However, althoughwidely recognized, the relationship between inflammation and throm-bosis has beenmostly explored in clinical studies measuring circulatinglevels of inflammatory cytokines [13,15–18]. At the same time, recentstudies in mouse models of VTE have provided important insights intoa novel role of erythrocytes, platelets and neutrophils in the pathogen-esis of this disease [21,30–32]. Therefore, we evaluated the adhesiveproperties of erythrocytes, platelets and neutrophils in VTE patientsand controls.

Our study design did not include patients in the acute phase of VTE,when the transient risk factors are possibly still present, and changes in

Table 2Comparative analysis of adhesive properties of platelets, erythrocytes and neutrophils ofVTE patients and controls.

Controls Patients P

Erythrocytes(n = 29)

7.52 (0.97–16.30) 7.45 (0.13–18.54) 0.888

Platelets(non-stimulated)(n = 25)

15.57 (7.60–55.94) 16.49 (8.50–46.31) 0.322

Platelets (TB-stimulus)(n = 25)

25.21 (11.68–68.55) 34.66 (16.47–77.98) 0.183

Neutrophils(n = 29)

19.29 (7.98–32.66) 22.13 (7.63–58.55) 0.099

Data are expressed as median % adhesion (range). The P values were calculated by theMann–Whitney test.

cell adhesion could also play an important role. In this study, we wereinterested in evaluating if an inflammatory condition, mediated byadhesion of blood cells, and an associated hypercoagulability state, werepresent in patients even after long time since the DVT event (between1–6 years after the thrombotic event).

Using a static adhesion assay, wewere able to demonstrate increasedadhesive properties of neutrophils in the subgroup of VTE patients asso-ciated with a higher risk of recurrent disease.

The static adhesion assay has been recently employed in studies,especially with sickle cell disease patients [24,25,28,33,34], supportingthe description of the increased adhesive properties of platelets andneutrophils in the pathogenesis of this condition. Since VTE is a diseaseoccurring under conditions of low shear stress, we believe that the staticadhesion assay is an adequate tool to investigate the role of circulatingblood cells in the pathogenesis of this condition.

We did not observe alteration in erythrocyte adhesion in our VTEpatients when compared with controls, and also between the VTE sub-groups. It is possible that relatively non adherent erythrocytes canbecome entrapped in fibrin during thrombus formation [22], in a pro-cess that could be independent of the intrinsic adhesive properties ofthese cells. However, an active role of these cells in the pathogenesisof VTE cannot be excluded by our results, as recently suggested by theobservation that erythrocytes may contribute to thrombin generationthrough the exposure of cell surface phosphatidylserine [35]. In addi-tion, hematocrit and related hematologic variables such as hemoglobinand red blood cells count can be considerated as risk factors for VTE in ageneral population [36].

We next evaluated platelet adhesion in patients. During inflammatoryconditions, platelets roll along and adhere to the activated endothelium[37,38]. Accordingly, we observed an increase of platelet adherence tofibrinogen in response to thrombin in both patients and controls. Never-theless, we could not demonstrate alterations in the adhesive properties

Table 3Comparative analysis of adhesive properties of platelets, erythrocytes and neutrophils, among groups: controls, patients with normal D-dimer levels and without RVO and patients withhigh D-dimer levels and RVO.

Controls Patients with normal DDa

and without RVObPatients with higher

DD and RVOp⁎

Platelets(non-stimulated)

15.57(7.60–5.94)n = 25

17.94(11.78–46.31)

n = 12

15.47(8.50–41.61)

n = 130.429

Platelets(TB–stimulus)

25.21(11.69–68.55)

n = 25

40.30(19.83–77.98)

n = 12

30.97(16.47–58.18)

n = 130.149

Erythrocytes 7.52(0.97–16.30)

n = 29

7.10(2.34–12.79)

n = 14

7.53(0.13–18.54)

n = 150.976

Neutrophil 19.29(7.98–32.66)

n = 29

18.23(7.63–58.55)

n = 14

24.68(14.11–42.24)

n = 150.018

Data are expressed as median % adhesion (range).a DD (D-dimer).b RVD (residual vein occlusion).⁎ P Value was calculated by Kruskal-Wallis test.

740 K.C.S. Zapponi et al. / Thrombosis Research 133 (2014) 736–742

of platelets tofibrinogen inVTEpatients comparedwith controls, and alsobetween the VTE subgroups, with or without thrombin stimulation.Whether a different result would be obtained in the acute phase of VTEremains to be determined.

Finally, although not statistically significant, we did observe a trendtowards an increase in the adhesive properties of neutrophils in VTEpatients. When evaluated in more detail, patients mainly drove thisnon-significant increase with known markers of hypercoagulability(D-dimer and RVO), classically associated with a higher risk of VTErecurrence [8–12]. In fact, when evaluated as a separated group, the pop-ulation of VTE patients with high D-dimer and RVO presented a signifi-cant increase in the adhesive properties of neutrophils.

During an immune response, circulating neutrophils are recruited tosites of inflammation,where they contribute to tissue repair. The activa-tion and recruitment of these cells are mediated by interaction of cyto-kines, chemokines and the binding of leucocyte integrins to endothelialadhesion molecules [39]. Thus, in order to investigate the mechanismsof increased neutrophil adhesion in our patients, we evaluated the sur-face expression of adhesionmolecules, LFA-1 (CD11a/CD18) andMAC-1(CD11b/CD18) on neutrophils, as well as the circulating levels of otherinflammatory markers.

The surface expressions of the CD11a, CD11b and CD18 moleculeswere not significantly altered in the subgroups of VTE patients and con-trols. However, changes in cell adhesion can be mediated both by

Fig. 2. (A) Neutrophil adhesion to FN in VTE patients compared to controls. (B) Comparative anD-dimer (DD) levels and without residual vein occlusion (RVO) and patients with high DD leveFN. Bars indicate median values. FN = fibro nectin. The P values were calculated by MannWh

mobilization of integrins to the cell surface, or by changes in their affin-ity and avidity (mediated largely by integrin clustering) [40–44]. The af-finity of MAC-1 (CD11b/CD18) for its ligands, for example, isdynamically regulated by inside-out signaling, producing a form of theintegrin that binds to its substrate with a much greater affinity [41]. Re-cent structural and biophysical data predict that LFA-1 andMAC-1 existin at least three conformational states [44]. It is, therefore, possible thatthe increased adhesive properties of neutrophils observed may be me-diated by alterations in integrin affinity rather than expression.

In regard to circulating levels of inflammatorymarkers, IL-6, IL-8 andTNF-α were significantly elevated in patients with higher neutrophiladhesive properties, when compared to the remaining patients. In addi-tion, the adhesive properties of neutrophils were positively correlatedwith IL-6 and D-dimer levels, suggesting a possible relationship be-tween these factors. Proinflammatory cytokines and other mediatorsare able to activate the coagulation system and down-regulate naturalanticoagulant pathways [13]. Moreover, thrombin and factor Xa, bybinding to protease-activated receptors in endothelial andmononuclearcells, are able to stimulate the production of several cytokines, thusdemonstrating a bidirectional relationship between inflammation andcoagulation [14]. IL-6 augments the expression of adhesion moleculessuch as VCAM-1 (vascular cell adhesion molecule-1) and ICAM-1(main ligand of integrin LFA-1 and MAC-1 on neutrophils) in inflamedsites andon endothelial cells, and induces theproductionof chemokines

alysis of neutrophil adhesion to FN among groups: Healthy controls, patients with normalls and RVO. Neutrophil adhesion is expressed as the percentage of neutrophils adhered toitney and Kruskal-Wallis tests.

Table 4Increased serum levels of IL-6, IL-8, TNF-α in VTE patientswith RVO and highD-dimer levels.

Patients with normalDDa and without RVO

(n = 14)

Patients with higherDD and RVO(n = 15)

p⁎

IL-6 pg/mL 0.84 (0.18–2.68) 2.08 (0.71–6.37) 0.011IL-8 pg/mL 15.74 (5.88–40.96) 28.72 (12.60–191.1) 0.027TNF-α pg/mL 2.02 (1.20–6.64) 4.50 (2.03–13.76) 0.024CRP pg/mL 0.14 (0.001–1.77) 0.35 (0.02–1.05) 0.093

Data expressed as median (range).a DD (D-dimer); RVO (residual vein occlusion).⁎ P Value was calculated by the Mann–Whitney U test.

741K.C.S. Zapponi et al. / Thrombosis Research 133 (2014) 736–742

such as IL-8, IL-6 itself, and acute phase proteins such as C-reactive pro-tein (CRP) [39,45,46]. The IL-6 receptor, present in neutrophils, pro-motes neutrophil recruitment and activation, which could change theexpression and activity of adhesionmolecules [39]. It is therefore possi-ble that higher levels of IL-8, TNF-α and particularly IL-6, could lead tothe increased adhesive properties of neutrophils, triggering a viciouscircle involving inflammation, increased neutrophil adhesion and acti-vation of coagulation.

Our study has several limitations. First, we did not evaluate celladhesion properties under high shear stress. However, as discussedabove, we believe that static adhesion is a more adequate model forVTE, given the low shear stress in the venous circulation, where VTEoccurs. Secondly, we only evaluated the quantitative expression ofadhesion molecules, as such, we cannot rule out the possibility thatchanges in their affinity could explain our results. Third, other pathwaysof neutrophil adhesion, such as the α4-integrin/VCAM-1 [47], were notevaluated and could also play a role on thrombosis. It would be interest-ing to evaluate this pathway in VTE patients in further studies. Forth,due to a strict inclusion criteria a great number of patientswere excluded.The exclusion of patients with chronic diseases, as cancer, inflammatoryand autoimmune diseases, has contributed for the small population sizeof the study and possibly for the fact that some results failed on reachingstatistical significance. However, we considered crucial the elimination of

Fig. 3. (A) Positive correlation of neutrophil adhesive propertieswith levels of IL-6 pg/mL in VTElevels in VTE patients (n = 29). DD (D-dimer). P = and r = values refer to Spearman’s correla

Table 5Comparative analysis of expressions of CD11a, CD11b and CD18 on the surface among groups: conlevels and RVO.

Controls(n = 29)

Patients wDDa and w

(n =

CD 11a (MFI) 884 (106–1591) 827 (379CD11b (MFI) 3676 (22–9274) 3791 (144CD18 (MFI) 2616 (1123–5296) 2592 (843

Data are expressed as median fluorescence intensities (MFI) on the surface of 10,000 neutropha DD (D-dimer).b RVO (residual vein occlusion).⁎ P value was calculated by Kruskal–Wallis test.

these potential confounders, avoiding drawingwrong conclusion and im-proving the reliability of our results. Hence, larger clinical and prospectivestudies are needed to validate these results.

In conclusion, we demonstrated increased adhesive properties ofneutrophils in VTE patients with high D-dimer levels and RVO, andthe association of these findings with high levels of IL-6, IL-8 and TNF-α.These findings not only support the hypothesis of an association betweeninflammation and hypercoagulability, as suggested by previousstudies, but more importantly, highlight the role of neutrophils inthis process.

Conflict of Interest Statement

The authors state that they have no conflict of interest.

Acknowledgements

The study receivedfinancial support from the Fundação deAmparo àPesquisa do Estado de São Paulo (FAPESP, no 2010/12352-0) and CAPES.The Hematology andHaemotherapy Centre, UNICAMP forms part of theNational Institute of Blood (INCT de sangue – CNPq/MCT).

References

[1] Rosendaal FR. Risk factors for venous thrombotic disease. Thromb Haemost Aug1999;82(2):610–9.

[2] Heit JA, O'Fallon WM, Petterson TM, Lohse CM, Silverstein MD, Mohr DN, et al.Relative impact of risk factors for deep vein thrombosis and pulmonary embolism:a population-based study. Arch Intern Med Jun 10 2002;162(11):1245–8.

[3] Rosendaal FR. Venous thrombosis: amulticausal disease. Lancet Apr 3 1999;353(9159):1167–73.

[4] Fowkes FJ, Price JF, Fowkes FG. Incidence of diagnosed deep vein thrombosis in thegeneral population: systematic review. Eur J Vasc Endovasc Surg Jan 2003;25(1):1–5.

[5] Morange PE, Tregouet DA. Deciphering the molecular basis of venous thromboembo-lism: where are we and where should we go? Br J Haematol Feb 2010;148(4):495–506.

[6] Prandoni P, Noventa F, Ghirarduzzi A, Pengo V, Bernardi E, Pesavento R, et al. Therisk of recurrent venous thromboembolism after discontinuing anticoagulation in

patients (n=29). (B) Positive correlation of neutrophil adhesive properties with DDmg/Ition.

trols, patientswith normal D-dimer levels andwithout RVO and patientswith highD-dimer

ith normalithout RVOb

14)

Patents with higherDD and RVO(n = 15)

p⁎

–1158) 613 (338–937) 0.153–9712) 2705 (238–7962) 0.740–4304) 1742 (356–5399) 0.427

ils, as measured by flow cytometry. Data expressed as median (range).

742 K.C.S. Zapponi et al. / Thrombosis Research 133 (2014) 736–742

patients with acute proximal deep vein thrombosis or pulmonary embolism. A pro-spective cohort study in 1,626 patients. Haematologica Feb 2007;92(2):199–205.

[7] Prandoni P, Lensing AW, Bernardi E, Villalta S, Bagatella P, Girolami A. The diagnosticvalue of compression ultrasonography in patients with suspected recurrent deepvein thrombosis. Thromb Haemost Sep 2002;88(3):402–6.

[8] Verhovsek M, Douketis JD, Yi Q, Shrivastava S, Tait RC, Baglin T, et al. Systematicreview: D-dimer to predict recurrent disease after stopping anticoagulanttherapy for unprovoked venous thromboembolism. Ann Intern Med Oct 72008;149(7):481–90 [W94].

[9] Tosetto A, Iorio A, Marcucci M, Baglin T, Cushman M, Eichinger S, et al. Predictingdisease recurrence in patients with previous unprovoked venous thromboembolism:a proposed prediction score (DASH). J Thromb Haemost Jun 2012;10(6):1019–25.

[10] CarrierM, RodgerMA,Wells PS, RighiniM, G LEG. Residual vein obstruction to predictthe risk of recurrent venous thromboembolism in patientswith deep vein thrombosis:a systematic review and meta-analysis. J Thromb Haemost Jun 2011;9(6):1119–25.

[11] Cosmi B, Legnani C, Cini M, Guazzaloca G, Palareti G. D-dimer levels in combinationwith residual venous obstruction and the risk of recurrence after anticoagulationwithdrawal for a first idiopathic deep vein thrombosis. Thromb Haemost Nov2005;94(5):969–74.

[12] Cosmi B, Legnani C, Iorio A, Pengo V, Ghirarduzzi A, Testa S, et al. Residual venousobstruction, alone and in combination with D-dimer, as a risk factor for recurrenceafter anticoagulation withdrawal following a first idiopathic deep vein thrombosisin the prolong study. Eur J Vasc Endovasc Surg Mar 2010;39(3):356–65.

[13] Levi M, van der Poll T. Inflammation and coagulation. Crit Care Med Feb2010;38(2 Suppl.):S26–34.

[14] Levi M, van der Poll T, Buller HR. Bidirectional relation between inflammation andcoagulation. Circulation Jun 8 2004;109(22):2698–704.

[15] van Aken BE, den Heijer M, Bos GM, van Deventer SJ, Reitsma PH. Recurrent venousthrombosis and markers of inflammation. Thromb Haemost Apr 2000;83(4):536–9.

[16] Reitsma PH, Rosendaal FR. Activation of innate immunity in patients with venousthrombosis: the Leiden Thrombophilia Study. J Thromb Haemost Apr 2004;2(4):619–22.

[17] Christiansen SC, Naess IA, Cannegieter SC, Hammerstrom J, Rosendaal FR, Reitsma PH.Inflammatory cytokines as risk factors for a first venous thrombosis: a prospectivepopulation-based study. PLoS Med Aug 2006;3(8):e334.

[18] Esmon CT. The impact of the inflammatory response on coagulation. Thromb Res2004;114(5–6):321–7.

[19] Matos MF, Lourenco DM, Orikaza CM, Bajerl JA, Noguti MA, Morelli VM. The role ofIL-6, IL-8 and MCP-1 and their promoter polymorphisms IL-6 -174GC, IL-8 -251ATand MCP-1 -2518AG in the risk of venous thromboembolism: a case–controlstudy. Thromb Res Sep 2011;128(3):216–20.

[20] Furie B, Furie BC. Mechanisms of thrombus formation. N Engl J Med Aug 282008;359(9):938–49.

[21] von Bruhl ML, Stark K, Steinhart A, Chandraratne S, Konrad I, Lorenz M, et al. Mono-cytes, neutrophils, and platelets cooperate to initiate and propagate venous throm-bosis in mice in vivo. J Exp Med Apr 9 2012;209(4):819–35.

[22] Saha P, Humphries J, Modarai B, Mattock K, Waltham M, Evans CE, et al. Leukocytesand the natural history of deep vein thrombosis: current concepts and future direc-tions. Arterioscler Thromb Vasc Biol Mar 2011;31(3):506–12.

[23] Manfredi AA, Rovere-Querini P, Maugeri N. Dangerous connections: neutrophils andthe phagocytic clearance of activated platelets. Curr OpinHematol Jan 2010;17(1):3–8.

[24] Gambero S, Canalli AA, Traina F, Albuquerque DM, Saad ST, Costa FF, et al. Therapywith hydroxyurea is associated with reduced adhesion molecule gene and proteinexpression in sickle red cells with a concomitant reduction in adhesive properties.Eur J Haematol Feb 2007;78(2):144–51.

[25] Proenca-Ferreira R, Franco-Penteado CF, Traina F, Saad ST, Costa FF, Conran N. In-creased adhesive properties of platelets in sickle cell disease: roles for alphaIIbbeta3-mediated ligand binding, diminished cAMP signalling and increased phos-phodiesterase 3A activity. Br J Haematol Apr 2010;149(2):280–8.

[26] Bellavite P, Andrioli G, Guzzo P, Arigliano P, Chirumbolo S, Manzato F, et al. A color-imetric method for the measurement of platelet adhesion in microtiter plates. AnalBiochem Feb 1 1994;216(2):444–50.

[27] Eriksson AC, Whiss PA. Characterization of static adhesion of human platelets inplasma to protein surfaces in microplates. Blood Coagul Fibrinolysis Apr 2009;20(3):197–206.

[28] Assis A, Conran N, Canalli AA, Lorand-Metze I, Saad ST, Costa FF. Effect of cytokinesand chemokines on sickle neutrophil adhesion to fibronectin. Acta Haematol2005;113(2):130–6.

[29] Prandoni P, Prins MH, Lensing AW, Ghirarduzzi A, AgenoW, Imberti D, et al. Residualthrombosis on ultrasonography to guide the duration of anticoagulation in patientswith deep venous thrombosis: a randomized trial. Ann Intern Med May 5 2009;150(9):577–85.

[30] Brill A, Fuchs TA, Chauhan AK, Yang JJ, De Meyer SF, Kollnberger M, et al. vonWillebrand factor-mediated platelet adhesion is critical for deep vein thrombosisin mouse models. Blood Jan 27 2011;117(4):1400–7.

[31] Fuchs TA, Brill A, Duerschmied D, Schatzberg D, Monestier M, Myers Jr DD, et al.Extracellular DNA traps promote thrombosis. Proc Natl Acad Sci U S A Sep 72010;107(36):15880–5.

[32] Fuchs TA, Brill A, Wagner DD. Neutrophil extracellular trap (NET) impact on deepvein thrombosis. Arterioscler Thromb Vasc Biol Aug 2012;32(8):1777–83.

[33] Silveira AA, Dominical VM, Lazarini M, Costa FF, Conran N. Simvastatin abrogates in-flamed neutrophil adhesive properties, in association with the inhibition of Mac-1integrin expression and modulation of Rho kinase activity. Inflamm Res Feb2013;62(2):127–32.

[34] Canalli AA, Franco-Penteado CF, Saad ST, Conran N, Costa FF. Increased adhesiveproperties of neutrophils in sickle cell disease may be reversed by pharmacologicalnitric oxide donation. Haematologica Apr 2008;93(4):605–9.

[35] Whelihan MF, Zachary V, Orfeo T, Mann KG. Prothrombin activation in bloodcoagulation: the erythrocyte contribution to thrombin generation. Blood Nov 12012;120(18):3837–45.

[36] Braekkan SK, Mathiesen EB, Njolstad I, Wilsgaard T, Hansen JB. Hematocrit andrisk of venous thromboembolism in a general population. The Tromso study.Haematologica Feb 2010;95(2):270–5.

[37] Tabuchi A, Kuebler WM. Endothelium-platelet interactions in inflammatory lungdisease. Vascul Pharmacol Oct-Dec 2008;49(4–6):141–50.

[38] Wagner DD, Frenette PS. The vessel wall and its interactions. Blood Jun 1 2008;111(11):5271–81.

[39] Mihara M, Hashizume M, Yoshida H, Suzuki M, Shiina M. IL-6/IL-6 receptor systemand its role in physiological and pathological conditions. Clin Sci (Lond) Feb2012;122(4):143–59.

[40] Hogg N, Leitinger B. Shape and shift changes related to the function of leukocyteintegrins LFA-1 and Mac-1. J Leukoc Biol Jun 2001;69(6):893–8.

[41] Shimaoka M, Takagi J, Springer TA. Conformational regulation of integrin structureand function. Annu Rev Biophys Biomol Struct 2002;31:485–516.

[42] Arnaout MA, Mahalingam B, Xiong JP. Integrin structure, allostery, and bidirectionalsignaling. Annu Rev Cell Dev Biol 2005;21:381–410.

[43] Hu P, Luo BH. Integrin bi-directional signaling across the plasma membrane. J CellPhysiol Feb 2012;228(2):306–12.

[44] Montresor A, Toffali L, Constantin G, Laudanna C. Chemokines and the signalingmodules regulating integrin affinity. Front Immunol 2012;3:127.

[45] RomanoM, SironiM, Toniatti C, Polentarutti N, Fruscella P, Ghezzi P, et al. Role of IL-6and its soluble receptor in induction of chemokines and leukocyte recruitment. Im-munity Mar 1997;6(3):315–25.

[46] Rincon M. Interleukin-6: from an inflammatory marker to a target for inflammatorydiseases. Trends Immunol Nov 2012;33(11):571–7.

[47] Ibbotson GC, Doig C, Kaur J, Gill V, Ostrovsky L, Fairhead T, et al. Functional alpha4-integrin: a newly identified pathway of neutrophil recruitment in critically ill septicpatients. Nat Med Apr 2001;7(4):465–70.