OUP user menu

Journal of Crohn's and Colitis: 10 (10)


Laurence J. Egan, Ireland

Associate Editors

Shomron Ben-Horin, IsraelSilvio Danese, ItalyPeter Lakatos, HungaryMiles Parkes, UKJesús Rivera-Nieves, USABritta Siegmund, GermanyGijs van den Brink, NLSéverine Vermeire, Belgium


Published on behalf of

Tissue factor exposing microparticles in inflammatory bowel disease

Julia Palkovits, Gottfried Novacek, Marietta Kollars, Gregor Hron, Wolf Osterode, Peter Quehenberger, Paul Alexander Kyrle, Harald Vogelsang, Walter Reinisch, Pavol Papay, Ansgar Weltermann
DOI: http://dx.doi.org/10.1016/j.crohns.2012.05.016 222-229 First published online: 1 April 2013


Background: Circulating procoagulant microparticles (MPs) are thought to be involved in the pathogenesis of venous thromboembolism in patients with inflammatory bowel disease (IBD). However, the exposure of tissue factor, the primary initiator of coagulation activation, on microparticles (TF+MPs) and its association with hemostasis activation has not yet been studied in IBD patients.

Methods: In this case–control study 49 IBD patients (28 Crohn's disease, 21 ulcerative colitis) and 49 sex- and age-matched, healthy controls were included. Clinical disease activity (Crohn's Disease Activity Index and Clinical Activity Index, respectively) was assessed and IBD-related data were determined by chart review. Numbers, cellular origin and procoagulant activity of TF+MPs in plasma were determined using flow cytometry and a chromogenic activity assay. D-dimer and high-sensitive C-reactive protein (CRP) served as markers for coagulation activation and inflammation, respectively. The primary endpoint was the number of TF+MPs in IBD patients compared to controls.

Results: Median number (interquartile range) of TF+MPs was higher in IBD patients than in controls (14.0 (11.9–22.8) × 103/mL vs. 11.9 (11.9–19.1) × 103/mL plasma, P = 0.029). This finding was due to generally higher plasma levels of MPs from platelets and leukocytes in IBD patients. However, the number of TF+MPs was neither correlated with their procoagulant activity and D-dimer nor with disease activity and CRP.

Conclusions: Increased numbers of circulating TF+MPs represent a new facet of hemostatic abnormalities in IBD. However, the lack of association with activation of the coagulation system and disease activity questions their pathogenetic role for venous thromboembolism in this patient group.

  • Microparticles
  • Tissue factor
  • Platelets
  • Venous thromboembolism
  • Inflammatory bowel disease

1 Introduction

Venous thromboembolism (VTE) is recognized as an extra-intestinal complication of patients with inflammatory bowel disease (IBD).19 The risk of VTE in IBD patients is two- to four-fold increased compared to non-IBD subjects3,5,7,9 and is considered to be a specific feature of IBD, since it is not shared with other chronic inflammatory diseases or chronic bowel diseases such as rheumatoid arthritis and celiac disease, respectively.5 The underlying pathogenesis of VTE in IBD seems to be multifactorial with acquired risk factors playing the most important role.10 At the time of diagnosis of VTE approximately two-thirds of patients reveal clinically active disease or IBD-specific complications such as fistula or abscess formation.2,5,10 The association between active disease and VTE has been confirmed very recently in a large cohort study.7 The prothrombotic tendency in IBD is reflected by high plasma levels of hemostasis activation markers.1012 According to the clinical finding of an association with disease activity the inflammatory process in IBD is thought to be the primary cause of hemostasis activation. For instance, secretion of biologically active molecules such as tumor necrosis factor-alpha and interleukin-1 can directly induce coagulation activation.1317

Emerging evidence suggests that microparticles (MPs) shed from various blood cells may contribute to the prothrombotic state in IBD.1820 MPs are membrane-coated vesicles which are shed from cells upon activation or apoptosis and harbor various antigens of the respective mother cell.21 In IBD, inflammatory processes are assumed to present the stimuli for cell activation and subsequent release of MPs. One of the most important properties revealed by MPs is their contribution to the initiation and propagation of coagulation.22,23 In vitro studies suggest that MPs contribute to thrombus formation by both a negatively charged phospholipid surface and exposure of procoagulant proteins such as tissue factor (TF), the primary initiator of coagulation activation.24 In pancreatic cancer, two prospective studies have recently shown a high risk of VTE among patients with high levels of MP-associated TF-activity.25,26 Also in diabetes mellitus MP characteristics have been shown to be related to a specific type of vascular complications.27 However, the role of TF exposure on MPs and its contribution to coagulation activation have not yet been investigated in IBD patients. Thus, our primary objectives in this prospective study were to 1) evaluate the number, cellular origin, and procoagulant activity of TF exposing microparticles (TF+MPs) in plasma of IBD patients compared to healthy controls and to 2) evaluate the impact of inflammation and coagulation system activation on the results of TF+MPs in IBD patients.

2 Patients and methods

The present investigation is a case–control study including IBD patients and control subjects. The study was approved by the ethics committee of the Medical University of Vienna and was performed in accordance with the Declaration of Helsinki and Good Clinical Practice. Prior to study enrollment, a written informed consent was obtained from each study subject.

2.1 Study population

IBD patients were recruited consecutively from the outpatient clinic of the Division of Gastroenterology and Hepatology, Department of Internal Medicine III of the Medical University of Vienna. Patients were eligible, when they had an established diagnosis of Crohn's disease (CD) or of ulcerative colitis (UC) (based on clinical, endoscopic, histological, and radiological criteria according to ECCO guidelines).28,29 Disease activities of CD and UC were assessed by the use of the Crohn's Disease Activity Index (CDAI)30 and the Clinical Activity Index (CAI),31 respectively. Quiescent disease was defined by a CDAI < 150 or a CAI < 4, whereas active disease was defined by a CDAI ≥ 150 or a CAI ≥ 4,32 respectively. Type of IBD, duration of disease, disease classification according to the Montreal classification,33 and medical therapy were determined by review of hospital records in all patients.

Control subjects were recruited from the outpatient clinic of the Division of Occupational Medicine, Department of Internal Medicine II of the Medical University of Vienna, and were eligible if there was no clinical evidence that they suffered from IBD.5 Subjects attended this department for their regular preventive medical check-up for exposure to solvents (xylene, toluene, etc.), heavy metals (lead, mercury, cadmium, etc.), radiation, or noise.5 This preventive medical check-up revealed no evidence of renal, liver, lung, and heart disease as well as abnormalities of the blood count. Both patients and controls were at least 19 years of age. Exclusion criteria for both patient groups included malignancy, diabetes mellitus, chronic renal failure, and the use of antiplatelet and anticoagulant drugs within 2 weeks before blood sampling.

A total of 49 patients (25 men and 24 women) with confirmed inflammatory bowel disease and 49 healthy control subjects matched for age and sex were enrolled. Demographic and clinical data of IBD patients and controls are presented in Table 1 .

View this table:
Table 1

Demographic and clinical data of control subjects and IBD patients.

Controls (n = 49)IBD total (n = 49)IBD quiescent n = 32)IBD active (n = 17)
Age (years)39 ± 1539 ± 1538 ± 1442 ± 16
Female, n (%)24 (49.0)24 (49.0)17 (53.1)7 (41.2)
CD/UC, n (%)28/21 (57.1/42.9)20/12 (62.5/37.5)8/9 (47.1/52.9)
Disease activity
CDAI for CD patients119 ± 13246 ± 33302 ± 103
CAI for UC patients5 ± 51 ± 110 ± 3
Duration of disease (years)11 ± 811 ± 810 ± 8
Disease location, n (%)
CD patients (n = 28)
L12 (7.1)1 (5.0)1 (12.5)
L25 (17.9)3 (15.0)2 (25.0)
L319 (67.9)15 (75.0)4 (50.0)
L4 (+ L1–L3)2 (7.1)1 (5.0)1 (12.5)
Perianal5 (17.9)3 (15.0)2 (25.0)
UC patients (n = 21)
Proctitis3 (14.3)2 (16.7)1 (11.1)
Left-sided colitis7 (33.3)4 (33.3)3 (33.3)
Extensive colitis11 (52.4)6 (50.0)5 (55.6)
Medical therapy
5-ASA, n (%)25 (51.0)17 (53.1)8 (47.1)
Corticosteroids, n (%)12 (24.5)5 (15.6)7 (41.2)
Azathioprine, n (%)14 (28.6)10 (31.3)4 (23.5)
Infliximab, n (%)1 (2.0)01( 5.9)
  • Values are numbers (percentage) or mean ± standard deviation, respectively.

  • CD, Crohn's disease; CAI, clinical activity index; CDAI, Crohn's disease activity index; IBD, inflammatory bowel disease; UC, ulcerative colitis; 5-ASA, 5-aminosalicylic acid.

  • Quiescent IBD was defined by a CDAI < 150 (CD) or a CAI < 4 (UC) and active disease was defined by a CDAI ≥ 150 or a CAI ≥ 4.

  • The location of Crohn's disease was defined according to the Montreal Classification: L1 ileal, L2 colonic, L3 ileocolonic, L4 isolated upper gastrointestinal tract (L4 is a modifier that can be added to L1–L3 when concomitant upper gastrointestinal disease is present).

  • Percentage of patients with CD or UC, respectively.

  • Some patients took more than one drug.

2.2 Sample collection and preparation

After puncture of an anticubital vein using a 21-gauge needle (Vacuette, Greiner Bio-One, Kremsmünster, Austria), the first 3 mL of blood was drawn into EDTA (ethylenediaminetetraacetic acid) Vacutainer tubes (Vacuette, Greiner Bio-One) for determination of the complete blood count (hematology automated analyzer: Sysmex XE-2100, Kobe, Japan). Without applying further venostasis, blood was drawn into 0.1 volume of 3.8% trisodium citrate (Vacuette, Greiner Bio-One) for determination of MPs, D-dimer, and high sensitive CRP. Within 1 h after blood sampling, platelet-free plasma was gained by one-step centrifugation (2600 g for 15 min at 4 °C). For analysis of MPs, 500 μL of supernatant was immediately placed on melting ice, whereas aliquots of the remaining plasma were stored at − 80 °C until determination of D-dimer (enzyme linked immunoassay, Asserachrom D-dimer, Boehringer Mannheim, Germany) and high sensitive CRP (Cardio Phase hsCRP Reagent, Nephelometer Analyzer II, Dade Behring, Marburg, Germany).

2.3 Flow cytometric analysis of microparticles

Flow cytometric analysis was performed as previously described by Hron et al34: 390 μL plasma was diluted in 845 μL annexin V binding buffer (10 mmol L− 1 Hepes/NaOH (pH 7.4), 140 mmol/L NaCl, 2.5 mmol/L CaCl2). To 95 μL of this mixture, 5 μL of anti-CD142-phycoerythrin (PE) (clone HTF-1, IgG1) and 5 μL of annexin V binding buffer were added to quantify total number of TF+MPs in plasma. Cellular origin of TF+MPs was evaluated by replacing 5 μL of annexin V binding buffer with 5 μL of one of the following cell-specific monoclonal antibodies: leukocytes (anti-CD45-Peridin–Chlorophyll–protein complex (PerCP), clone 2D1, IgG1), monocytes (anti-CD14-PerCP, clone MϕP9, IgG2b) or platelets (anti-CD41a-PerCP-Cy5.5, clone HIP8, IgG1). Monoclonal antibodies against Glycophorin A, available either fluorescein isothiocyanate (FITC) or PE, were used to determine MPs from red blood cells. All antibodies were purchased from Becton Dickinson except for the FITC- and PE-labeled monoclonal antibodies against Glycophorin A which were obtained from Immunotech (Marseille, France). After incubation in the dark for 15 min at 4 °C, 5 μL of annexin V-FITC was added, followed by a further incubation in the dark for 15 min at 4 °C.

After dilution in 1500 μL binding buffer, microparticles were analyzed on a FACScan cytometer using CELLQUEST-PRO software (BD Biosciences, San Jose, CA, USA; acquisition time 3 min, flow rate 60 mL min− 1) using forward and side scatters set at logarithmic gain. All annexin V positive events that were 1 μm (adjusted by fluorescent microbeads; Fluka, Buchs, Switzerland) or smaller in size were counted as microparticles. The thresholds to distinguish TF positive from TF negative events and cell marker positive and negative events were based on the findings with annexin binding positive microparticles. To ensure specific measurement of MPs, TF+MPs, and cell specific MPs, respective concentrations of isotype-matched control antibodies (all clone X40, IgG1) were analyzed in the same way to set threshold fluorescence levels. The detection limits for MPs-events were then determined by five independent experiments using annexin buffer instead of plasma. For each antibody combination, the highest value found in these experiments was defined as detection limit. The number of MPs in plasma was calculated as follows: Counts (K/mL plasma) = N [counted events] × (used volume [1610 μL] / used plasma volume [30 μL]) / acquisition volume [180 μL]. Intra-assay variabilities of our assay system (coefficient of variation) for healthy controls (with low levels of TF microparticles) were 6.4%, 19.7% and 42.0% for annexin positive binding MPs, CD41 positive MPs and TF+MPs, respectively. The amount of phosphatidylserine exposure on MPs was quantified by measurement of fluorescence signal intensity of annexin V binding on MPs (arbitrary units). In order to determine the fluorescence threshold value (background noise), a respective concentration of a FITC isotype-control antibody (clone X40, IgG1) was applied. For determination of the mean fluorescence signal of annexin V binding, only MPs with a fluorescence signal above the background noise (20 arbitrary units) were evaluated.

2.4 Procoagulant activity of TF+MPs

Procoagulant activity (PCA) of TF+MPs was quantified using a commercially available chromogenic assay (Actichrome TF, American Diagnostica Inc., Stamford, CT, USA) according to the manufacturers' instructions. To enable specific measurement of PCA of TF+MPs, soluble plasma TF was removed by centrifugation at 100,000 g. MP pellets were resuspended and stored in a TritonX-100 solution at − 80 °C until testing. For analysis, samples were thawed rapidly at 37 °C for immediate use. The absorbance of the reaction solutions was read by a micro-test plate reader (Dynatech MR 7000, Guernsey Channel Islands). The corresponding concentration of active TF was interpolated directly from a standard curve that was constructed using serial diluted lipidated TF standards.

2.5 Statistical methods

The primary endpoint was the number of TF+MPs in IBD patients compared to controls. Secondary endpoints were total number of MPs, number by their cellular origin, the procoagulant activity of TF+MPs and their association with clinical activity, high sensitive CRP, and D-dimer. The power calculation was based on the following considerations: the mean number of TF+MPs (primary endpoint) of healthy controls was defined as 100% (SD 25%). Assuming a 25% increase of TF+MPs (alpha 0.05, power 0.8) the estimated number of patients would be at least 16 for IBD subgroups (patients with quiescent and active disease, respectively). Data are presented as median with interquartile range (IQR), unless indicated otherwise. For MP values below the detection limit, levels were corrected to a value as high as the detection limit. Continuous and dichotomous data were analyzed by means of Mann–Whitney U-test and Wilcoxon signed rank test, respectively. Bivariate correlations were estimated by Pearson's correlation coefficient (r). Providing two-tailed significance levels, differences and correlations were considered statistically significant at P < 0.05. Statistical analysis was performed using SPSS version 17.0 (Chicago, IL, USA).

3 Results

3.1 Blood count

Compared to controls, platelet and leukocyte counts were significantly higher in IBD patients (Table 2). This difference was independent from the type of disease (CD/UC) (data not shown) and was more pronounced in active than in quiescent disease (Table 2).

View this table:
Table 2

Blood count and markers of coagulation activation and inflammation.

Controls (n = 49)IBD total (n = 49)PIBD quiescent (n = 32)PIBD active (n = 17)P
Hemoglobin (g/dL)13.7 (13.3–15.2)13.6 (12.1–14.7)0.06614.0 (12.6–14.8)0.54312.1(10.3–13.8)0.001
Leucocytes (G/L)6.0 (5.2–7.5)7.2 (6.4–9.2)0.0016.9 (6.0–8.0)0.0589.5 (6.5–12.0)0.001
Platelets (G/L)233 (183–276)293 (250–356)0.006282 (237–313)0.015378 (275–457)< 0.001
C-reactive protein (mg/dL)0.07 (0.03–0.15)0.39 (0.16–1.79)< 0.0010.26 (0.08–0.44)< 0.0013.02 (1.25–5.86)< 0.001
D-dimer (μg/mL)0.14 (0.08–0.26)0.44 (0.17–0.76)< 0.0010.34 (0.12–0.70)< 0.0010.56 (0.28–1.25)< 0.001
  • Values are median values (interquartile range).

  • P values are calculated from each patient group vs. the control group and determined by Mann–Whitney U test.

3.2 Number and cellular origin of microparticles

The total number of MPs did not differ between IBD patients and controls (313 [199–442] × 103/mL vs. 277 [218–327] × 103/mL, P = 0.34) as well as between patients with active or quiescent disease and controls (Table 3). With regard to the cellular origin of MPs (Table 3), patients with quiescent IBD but not active disease showed significantly higher numbers of platelet-derived MPs than control subjects (P = 0.039). Conversely, number of erythrocyte-derived MPs was significantly lower in patients with active disease compared with control subjects (P = 0.01), whereas no difference was found between patients with quiescent disease and controls. The number of leukocyte-derived MPs was slightly, but not significantly higher in patients with quiescent IBD compared with control subjects. This was in particular true for MPs originating from monocytes, where differences tightly failed to reach statistical significance (P = 0.066). Monocyte-derived MPs were found to constitute the main subpopulation of leukocyte-derived MPs accounting for at least 85% of total leukocyte-derived MPs.

View this table:
Table 3

Number, cellular origin, and phosphatidylserine exposure of microparticles in plasma of IBD patients and controls.

Controls (n = 49)IBD total (n = 49)PIBD quiescent (n = 32)PIBD active (n = 17)P
Microparticles (× 103/mL plasma)
Total MPs277 (218–327)313 (199–442)0.343329 (222–466)0.145272 (165–400)0.797
Platelet MPs (CD41a+)67 (33–96)103 (51–141)0.032109 (49–142)0.03992 (49–147)0.199
Leukocyte MPs (CD45+)5.1 (4.5–7.3)6.0 (4.5–9.1)0.1407.5 (4.5–9.2)0.1955.7 (4.5–9.1)0.273
Monocyte MPs (CD14+)5.4 (4.5–6.7)6.0 (4.5–8.9)0.1046.3 (4.5–9.2)0.0665.4 (4.5–7.3)0.567
Erythrocyte MPs (Glycophorin A+)128 (88–193)98 (54–150)0.02999 (72–169)0.20474 (45–136)0.010
Phosphatidylserine-exposure on microparticles (mean fluorescence intensity, arbitrary units)
Platelet MPs (CD41a+)183 (150–244)182 (147–227)0.699189 (145–227)0.832176 (155–226)0.634
Leukocyte MPs (CD45+)102 (73–134)91 (64–109)0.10887 (63–107)0.11891 (63–126)0.344
Monocyte MPs (CD14+)93 (72–125)81 (62–125)0.22779 (57–142)0.32281 (71–110)0.322
  • Values are median values (interquartile range).

  • P values are calculated from each patient group vs. the control group and determined by Mann–Whitney U test.

  • 4.5 K/mL represent the detection limit of leukocyte/monocyte MPs.

Comparing the active and the quiescent patient groups with each other, neither the total number of MPs, nor the number within any subgroup of cell-specific MPs showed statistically significant differences.

3.3 Tissue factor exposure of microparticles

In IBD patients the total number of TF+MPs was significantly higher than in controls (14.0 [11.9–22.8] × 103/mL vs. 11.9 [11.9–19.1] × 103/mL, P = 0.029) (Table 4). In patients with quiescent disease, TF+MP levels were about 40% higher compared to the control group (P = 0.018) (Table 4). The elevated number of TF+MPs was in part due to a significantly higher level of platelet-derived TF+MPs (P = 0.012). TF+MPs of leukocyte origin were also found to be higher in these patients but differences did not reach statistical significance. Nearly all leukocyte-derived TF+MPs co-expressed CD14 indicating monocyte origin. The total number of TF+MPs and numbers of TF+MPs from monocytes or platelets were not different between patients with active disease and controls (Table 4).

View this table:
Table 4

Number, cellular origin and procoagulant activity (PCA) of TF exposing microparticles (TF+MPs) in plasma of IBD patients and healthy controls.

Controls (n = 49)IBD total (n = 49)PIBD quiescent (n = 32)PIBD active (n = 17)P
Total TF+ MPs11.9 (11.9–19.1)14.0 (11.9–22.8)0.02916.3 (11.9–33.8)0.01812.9 (11.9–19.4)0.340
Platelet TF+ MPs (CD41a)4.8 (4.5–6.3)6.0 (4.5–8.7)0.0416.0 (4.5–11.6)0.0125.1 (4.5–6.3)0.676
Leukocyte TF+ MPs (CD45)3.9 (3.0–5.7)5.4 (3.0–7.8)0.2176.0 (3.0–8.1)0.1514.8 (3.0–6.6)0.713
Monocyte TF+ MPs (CD14)4.5 (3.0–6.0)5.1 (3.0–7.6)0.3635.5 (3.1–8.4)0.1374.5 (3.0–6.1)0.699
PCA of TF+ MPs (nmol/mL)2.6 (1.7–3.5)2.4 (1.4–3.2)0.3632.6 (1.4–3.2)0.4032.4 (1.5–3.0)0.543
  • Values are median (interquartile range) and represent number of MPs × K/mL plasma.

  • P values are calculated from each patient group vs. the control group and determined by Mann–Whitney U test.

  • Detection limits are defined as follows: 11.9 K/mL for total TF+ MPs; 4.5 K/mL for platelet TF+ MPs; 3.0 K/mL for leukocyte/monocyte TF+ MPs.

TF exposure strongly varied within the subpopulations of MPs. In control subjects 80% of leukocyte-derived and 85% of monocyte-derived MPs exposed TF. In contrast, TF was detectable in only 9% of platelet-derived MPs. These patterns of TF exposure were principally also found in patients with both quiescent and active IBD without significant differences (data not shown).

No TF exposure was found on erythrocyte-derived MPs as indirectly indicated by the fact that Glycophorin A-particles (size < 1 μm, regardless of annexin V-staining) did not co-stain with anti-CD-142-PE. We were unable to determine the cellular origin of 3.2 × 103/mL TF+MPs (26.9%) in controls, 4.3 × 103/mL TF+MPs (26.4%) in quiescent patients, and 3.0 × 103/mL TF+MPs (23.3%) in active patients. These TF+MPs might have been derived from endothelial cells.

Since platelet-derived TF+MPs seemed to play the key role for the elevation of total TF+MPs in patients with quiescent IBD, we tried to assess whether these results were attributable to the patient's elevated platelet count and/or to an enhanced release from activated platelets. Pooling data of patients and controls, a weak correlation was detectable between platelet-derived TF+MPs and platelet counts (r = 0.281, P = 0.018) Calculating the number of platelet-derived TF+MPs per 105 platelets, platelets of patients with quiescent IBD released comparable amounts of TF+MPs as platelets of controls (2.7 [2.0–4.3] × 103/105 platelets vs. 2.3 [1.9–3.4] × 103/105 platelets; P = 0.397). Remarkably, platelets of patients with active disease released significantly lower numbers of TF+MPs than platelets of controls (1.6 (1.2–3.0) × 103/105 platelets; P = 0.010).

Besides disease activity, the type of disease and medication might influence the shedding of microparticles. Interestingly, number of TF+MPs tended to be higher in patients with CD (17.7 (12.8–22.8) × 103/mL) than in patients with UC (11.9 (11.9–26.2) × 103/mL, P = 0.054). TF+MPs were comparable in patients with and without azathioprine medication (14.0 (13.1–19.1) × 103/mL vs. 14.2 (11.9–32.1) × 103/mL, P = 0.88). Moreover, levels of TF+MPs were not different between IBD patients with and without corticosteroid treatment (14.0 (11.9–19.4) × 103/mL vs. 17.9 (11.9–39.1) × 103/mL, P = 0.44), although numbers of TF+MPs from monocytes tended to be lower among patients on corticosteroids (3.0 vs. 5.1 × 103/mL; P > 0.05).

3.4 Phosphatidylserine

3.4.1 Exposure on microparticles

Besides TF exposure, the procoagulant activity of MPs is dependent on the presence of negatively charged phospholipids, mainly phosphatidylserine, on the outer membrane. In our study, the amount of phosphatidylserine on CD41a+MPs, CD45+MPs, and CD14+MPs did not significantly differ between patient groups and controls (Table 3). In all groups, phosphatidylserine exposure was almost two-fold higher on CD41a+MPs than on CD45+MPs and CD14+MPs, respectively (P ≤ 0.007 for each comparison).

3.4.2 Procoagulant activity of TF exposing microparticles (PCA of TF+MPs)

No difference was observed between PCA of TF+MPs of IBD patients and controls (2.4 (1.4–3.2) nmol/mL vs. 2.6 (1.7–3.5) nmol/mL, P = 0.36). IBD disease activity had no impact on PCA of TF+MPs (Table 3). Moreover, PCA of TF+MPs did not differ between patients with CD (2.4 (1.7–3.1) nmol/mL) and UC (2.4 (0.6–3.2) nmol/mL, P = 0.55). And finally, PCA of TF+MPs did not correlate with phosphatidylserine exposure on platelet and leukocyte derived MPs (r = − 0.054, P = 0.60 and r = 0.17, P = 0.10).

3.4.3 Markers of coagulation activation and inflammation

D-dimer levels of patients with both quiescent and active IBD were significantly increased compared with that of controls (Table 2). The difference between patients with quiescent and active IBD was not significant (P = 0.139). D-dimer levels did neither correlate with the total number of TF+MPs (r = − 0.056, P = 0.587) nor with the overall MP count (r = − 0.07, P = 0.495).

High sensitive CRP was determined to assess the level of inflammatory activity in patients with IBD and in control subjects. We found that patients with both quiescent and active disease showed significantly higher levels of high sensitive CRP than controls (Table 2). Furthermore, high sensitive CRP levels of patients with active IBD were significantly higher than that of patients with quiescent IBD (P < 0.001). High sensitive CRP did neither correlate with the total number of TF+MPs (r = − 0.052, P = 0.611) nor with TF+MPs of leukocytic origin (r = − 0.56, P = 0.586), but a significant correlation was found with D-dimer level (r = 0.44, P < 0.001). High sensitive CRP levels were more strongly correlated with disease activity in patients with CD (r = 0,60; P = 0.004) as compared to patients with UC (r = 0,37; P = 0.051). However, CRP neither correlated with TF + MPs in patients with CD (r = − 0.06, P = 0.6) nor with UC (r = 0.01, P = 0.97), respectively.

4 Discussion

Three studies have observed elevated levels of circulating MPs in the plasma of patients with IBD.1820 In this study we tried to complement this current evidence by evaluating the TF exposure of these MPs, since TF is known to play the key role in coagulation system activation. Our data revealed significantly higher levels of TF+MPs in patients with IBD as compared to age- and sex-matched controls. However, the TF dependent procoagulant activity of these MPs was not increased and levels of TF+MPs did not correlate with in vivo coagulation system activation.

In our study the total number of MPs, irrespective of TF exposure and cell origin, showed no differences between IBD patients and controls. However, with respect to the cell-specific subpopulations a markedly higher number of MPs from platelets and a trend of higher number of MPs from monocytes were found in patients with quiescent IBD as compared to controls. In fact, the 40% increased number of TF+MPs in quiescent IBD patients was mainly due to an elevated number of TF+MPs from platelets. It has been assumed that platelet activation occurs in the intestinal microcirculation after contact with intestinal microvascular endothelial cells.35,36 Since in vitro studies have demonstrated a significant release of MPs upon platelet activation,37 we guess that the enhanced shedding of TF+MPs from platelets of our patients is a consequence of platelet activation in the intestine. Another aspect is, that TF exposure on platelet-derived MPs is due to a specific TF transfer from monocytes to platelets via MPs.38 Thus, higher numbers of TF+MPs from platelets might reflect the influence of these cells on each other.3942

The procoagulant activity of TF on MPs was not different between patients and controls. Moreover, within the patient group no difference in MP-associated TF-activity was found with respect to disease activity and disease type, respectively. Although TF does not need to undergo proteolytic activation such as other coagulation factors, it is known that most of TF on membrane surfaces is physiologically encrypted and thus not active. Possibly, that TF on MPs is still encrypted in IBD and hence not active despite a high amount of phosphatidylserine on the outer membrane of MPs. This would explain the low and comparable TF-activity of MPs as well as the lack of association with inflammation and coagulation system activation, respectively. In our study, measurement of TF-activity of MPs was not performed at the very same day as the flow cytometric analysis and MPs were stored at − 80 °C until analysis. We cannot exclude that the freezing and thawing process has altered the membrane composition of MPs thereby significantly influencing the procoagulant activity of TF+MPs.

Elevated D-dimer levels found in our patients and the correlation with systemic inflammatory activity (represented by high sensitive CRP) are in line with the literature.43 Since MPs and especially TF+MPs correlated neither with inflammatory activity nor with D-dimer levels we question their role for activation of the coagulation system in IBD patients. Number of TF+MPs did not correlate with inflammation activity either. This seems paradoxical in so far as during inflammation activity, where alterations of the blood count were more pronounced than during quiescent disease, one would expect even higher numbers of TF+MPs than in patients with quiescent disease. Despite the significantly elevated platelet count in patients with active IBD, number of platelet-derived TF+MPs did not differ significantly from that of control subjects. This was reflected in a markedly reduced amount of platelet-derived TF+MPs per 105 platelets in active patients when compared to controls. There might be alternative interpretations of these findings: On the one hand, MPs might have been captured in intracapillary thrombi of the intestine, a phenomenon that was seen in pre-eclamptic women who showed an accumulation of platelet MPs in fibrin deposits in the placenta, whereas MP levels were decreased in the peripheral blood.44 On the other hand, a markedly reduced shedding of TF+MPs from platelets of these patients might have occurred due to inhibition of platelet activity by an unknown factor. In this connection, the question arises whether the IBD-specific therapy such as corticosteroids and immunosuppressives might have influenced the shedding of MPs.45,46 However, comparing the levels of TF+MPs among patients with and without corticosteroids and with and without azathioprine treatment, respectively, we could not find an impact of these medications on the results.

Finally, there are some limitations concerning study methods that have to be mentioned. It is known, that conventional FACS is limited by a low sensitivity for small-sized MP populations. As a consequence 11 patients with quiescent IBD (34%), 7 patients with active IBD (41%), and finally 29 control subjects (59%) revealed TF+MP levels below the detection limit and were thus rounded up to the detection limit. A more sensitive assay might have resulted in more significant differences of study results. Overall, evidence from the literature suggests that IBD is in any kind associated with elevated levels of MPs as compared to the healthy population. However, in detail the various results are quite heterogeneous due to different methodical approaches. For instance, Andoh et al. could most clearly demonstrate increased MP levels in both active CD and UC, correlating well with disease activity.19 In this study an ELISA system was used that only detected MPs of platelet origin. In contrast, the findings of Chamouard et al. were based on the use of an immunocapture assay measuring prothrombinase activity assay of MPs.18 They found increased levels of MPs in CD but not in UC and, interestingly, prothrombinase activity of MPs were much more pronounced in mild CD than in active disease as it was the case in our study. And finally, Pamuk et al. who investigated platelet activation and platelet-leukocyte complexes in UC patients did not find any elevation of platelet MPs at all.20 These results were based on flow cytometry assessments which were solely addressed to MPs of platelet origin. A further limitation might be the number of patients included in our study. We recruited the largest number of age- as well as sex-matched controls evaluating microparticles in IBD by now and, furthermore, the number of IBD patients was similar or even higher than in other studies on MPs in IBD. Nevertheless, we are aware of the potential limitation of the sample size for subgroup analyses, which, however, was not the primary objective of our study.

In conclusion, our data demonstrate increased levels of TF+MPs in the circulation of IBD patients. This observation was mainly attributable to a higher level of TF+MPs originating from platelets, obviously representing a new facet of the platelet abnormalities observed in IBD. Since the increased number of TF+MPs was not associated with an increased MP-associated TF-activity, activation of the coagulation system in vivo, and markers of inflammation, respectively, we have reason to doubt a crucial role of TF+MPs for the pathogenesis of venous thrombosis in IBD.

Conflict of interest



View Abstract