Secondary Logo

Journal Logo

Clinical Aspects

Characterization of Microvesicles in Septic Shock Using High-Sensitivity Flow Cytometry

Lehner, Georg Franz; Harler, Ulrich; Haller, Viktoria Maria; Feistritzer, Clemens; Hasslacher, Julia; Dunzendorfer, Stefan; Bellmann, Romuald; Joannidis, Michael

Author Information
doi: 10.1097/SHK.0000000000000657



The inflammatory reaction found in septic shock is orchestrated by cytokines that are able to activate the endothelium and to induce the release of extracellular vesicles, called microvesicles (MV) (1). MV, frequently referred to as microparticles, are sized approximately between 100 nm and 1 μm (2, 3). They can be released by cells either constitutively, upon stimulation or apoptosis. They are able to mediate pleiotropic inflammatory signals during sepsis (4), and presumably play a role in the propagation as well as in the initiation of blood coagulation (1). Current evidence suggests that MV can be cause and effect of cellular injury (1, 4–6). Thus, modulating MV generation or composition might be an elegant tool to prevent tissue injury in future. Endothelial pathology, activation, and apoptosis are considered key features of sepsis. Although there is evidence supporting the hypothesis of endothelial damage in sepsis and septic shock (7), the examination of the state of the endothelium in vivo may be difficult (8), since isolation of endothelial cells might induce alterations such as apoptosis itself. Endothelial-derived microvesicles (EMV) are considered a promising diagnostic tool to assess the state of the endothelium in diseases associated with endothelial activation and damage, such as ANCA-associated small vessel vasculitis (9). Recent evidence suggests that EMV are associated with the occurrence of disseminated intravascular coagulation (DIC) in septic shock (5). Since there is no uniform definition of EMV, different approaches were used in former studies, impeding direct comparison (5, 10–17). MV carry epitopes from their parent cell on their surface. Thus, subtypes of MV originating from distinct cell types can be differentiated by flow cytometry. However, common flow cytometers have a detection limit of approximately 0.5 μm bead-equivalents and are thus missing smaller sized MV. There is evidence that the use of high-sensitivity flow cytometry (hsFC), which is able to resolve MV down to a size of approximately 0.3 μm bead-equivalents, provides access to new biological information by increasing the sensitivity for smaller and previously undetectable MV (18).

This study was aimed at testing the hypothesis that increased levels of EMV reflect endothelial pathology which is considered to be a key element in the pathophysiology of sepsis. Therefore, we analyzed the signature of MV circulating in blood during the early phase of septic shock using hsFC as well as associations between counts of different MV subtypes and severity of disease.


Thirty consecutive eligible patients from a tertiary medical intensive care unit (ICU) with septic shock and 18 healthy volunteers were enrolled in the study between January 2012 and September 2013 (Supplemental Figure 1, Supplemental Digital Content 1, at The study was approved by the Ethics Committee from the Medical University Innsbruck (protocol UN 2705a 244/4.20). Subjects provided written informed consent either prior to enrollment or post-hoc. Septic shock patients were eligible for the study if they fulfilled at least two systemic inflammatory response criteria (19), in the presence of a suspected or proven infection, were dependent on vasopressors (norepinephrine), and if study enrollment was feasible within 48 h after onset of septic shock or ICU admission. Subjects were excluded if they were younger than 18 years, moribund, pregnant, breast feeding, or had a preceding episode of sepsis during the same admission.

Vital parameters were obtained, routine laboratory values measured as well as acute physiology and chronic health evaluation (APACHE II) score, simplified acute physiology score (SAPS II), and sequential organ failure assessment score (SOFA) computed from patients with septic shock. Additionally, International Society on Thrombosis and Haemostasis (ISTH)-DIC score was calculated and patients with an ISTH-DIC score of at least five were classified as having an overt DIC.

Sample collection and preparation

Blood was drawn in 3 mL S-Monovette tubes (Sarstedt, AG & Co, Nümbrecht, Germany) containing 3.2% citrate after discarding the first 3 mL. Sampling sites were an arterial line in septic shock patients and a cubital vein in healthy volunteers. A 21 G needle (BD Valu-Set, Becton Dickinson, Schwechat, Austria) was used for drawing blood from healthy volunteers by applying a mild tourniquet. Blood was immediately centrifuged at 20°C in a Rotanta 46 RC (Hettich, Tuttlingen, Germany) with 1,550 g for 15 min with acceleration and deceleration set at 8. The supernatant plasma was then centrifuged at 20°C with 13,000 g for 2 min in a Micro R22 (Hettich) centrifuge set at maximum acceleration and deceleration. Platelet-free plasma (PFP) was obtained, flash frozen in liquid nitrogen, and stored at −80 °C. Upon analysis, aliquots of PFP were thawed at 37°C (20) in a waterbath for 2 min and then kept on wet ice.

Specimen preparation and flow cytometric analysis of microvesicles

Three different antibody panels were used for characterization of MV (Supplemental Table 1, Supplemental Digital Content 1, at

  1. Panel one contained AnnexinV-FITC (1.5 μg/mL [final concentration]) to label phosphatidylserine on MV, CD31-PE (PECAM; platelet endothelial cell adhesion molecule; 0.39 μg/mL) which is present on MV from endothelium, platelets as well as leukocytes and CD42b-APC (GPIbα; glycoprotein Ib alpha; 0.78 μg/mL; all from BD Pharmingen, San Jose, CA) as well as CD41-PC7 (GPIIb; glycoprotein IIb; 0.78 μg/mL; Beckman-Coulter, Miami, FL) to label platelet-derived MV.
  2. Panel two consisted of CD144-FITC (VE-cadherin; vascular endothelial-cadherin; 1.5 μg/mL) to label all endothelial MV and CD62E-PE (ELAM-1; endothelial-leukocyte adhesion molecule 1; 0.78 μg/mL) as well as CD106-APC (VCAM-1; vascular cell adhesion protein 1; 0.70 μg/mL; all from BD Pharmingen) to label MV released during endothelial activation.
  3. To determine the percentage of CD31+/CD41− MV that stain positive for common leukocyte markers, rethawed PFP was analyzed with a third panel (leukocyte panel) as well as corresponding isotype controls, analogously to the procedure used for panel one. Panel three consisted of AnnexinV-FITC (0.5 μg/mL; BD Pharmingen), CD31-PE (PECAM; 0.391 μg/mL; BD Pharmingen), CD41-PC7 (GPIIb; 0.78 μg/mL; Beckman-Coulter) as well as CD45-Pacific Blue (leukocyte-common antigen; 4.545 μg/mL; Beckman-Coulter) to label leukocyte-derived MV, CD14-PerCP5.5 (myeloid cell-specific leucine-rich glycoprotein; 0.962 μg/mL; eBioscience, Vienna, Austria) to label monocyte-derived MV, CD66b-APC (carcinoembryonic antigen-related cell adhesion molecule 8; 0.566 μg/mL; eBioscience) as marker for granulocyte-derived MV, CD20-APC-Alexa Fluor 700 (B-lymphocyte surface antigen; 0.968 μg/mL; eBioscience) to label B-cell-derived MV and CD3-Krome Orange (T3, associated with T-cell receptor; 2.381 μg/mL; Beckman-Coulter) as marker for T-cell-derived MV.

For the setup of panel one platelet poor plasma (i.e. plasma obtained after centrifugation of citrated blood at 1550 g for 15 min) was used. To establish labeling procedures and to set up compensations as well as gates of endothelial-specific markers (i.e. panel two) PFP from healthy volunteers was enriched with EMV generated in vitro by stimulating renal microvascular endothelial cells with tumor necrosis factor alpha (Supplemental Figure 2, Supplemental Digital Content 1, at, analogously to Jimenez et al. (21). For the setup of panel three PFP from healthy volunteers was enriched with MV generated in vitro by stimulating leukocytes from healthy volunteers with 100 ng/mL tumor necrosis factor (TNF) alpha (Sigma-Aldrich, Saint Louis, MO) for 3 h and from leukocytes stimulated with 100 μg/mL lipopolysaccharides from Escherichia coli (LPS; Sigma-Aldrich) for 24 h.

Following incubation of PFP with the antibody panels for 30 min at room temperature, CytoCount beads (DakoCytomation, Glostrup, Denmark) and 500 μL of AnnexinV binding buffer (BD Pharmingen) in case of panel one and panel three or phosphate-buffered saline (PBS from PAA Laboratories, Pasching, Austria) in case of panel two were added. Isotype controls prepared analogously were run in parallel to assess background signals. These controls were labeled with corresponding panels of AnnexinV-FITC as well as matched irrelevant antibodies of the same isotype used at the same concentration as specific antibodies and diluted in calcium-free PBS. AnnexinV and antibodies were filtered with a 0.1 μm filter (Millipore, Darmstadt, Germany), PBS and AnnexinV binding buffer with a 0.2 μm syringe filter (Sarstedt, Nümbrecht, Germany).

Specimens were measured with a Gallios™ flow cytometer with the threshold set at forward scatter by using Gallios Cytometry List Mode Data Acquisition and Analysis Software 1.2™ (both Beckman Coulter, Bra, CA). Measurement was performed for 100 s at medium flow rate, which we found in preliminary experiments to be long enough to analyze a reliable number of low abundant MV, such as endothelial MV, and short enough to prevent clotting of samples and altering MV counts. Flow cytometric data were analyzed with Kaluza® version 1.2 (Beckman Coulter). The gating strategy included MV in a size range between 0.3 and 1.0 μm polystyrene bead-equivalents (LB3 and 89904 from Sigma-Aldrich) that showed a positivity concerning the aforementioned markers (Supplemental Figure 3, Supplemental Digital Content 1, at In the following, MV binding to AnnexinV due to their phosphatidylserine-rich surface are termed AnnexinV+ MV. The number of MV per microliter PFP was determined by referring to CytoCount beads for panels one and two. The percentage of CD31+/CD41− that stain positive for CD45, CD14, CD66b, CD20, or CD3 was calculated for panel three.

Quality control of flow cytometry data

Despite the above-mentioned thorough antibody and specimen handling false-positive fluorescent signals were evident in several samples in preliminary experiments, particularly altering counts of endothelial markers and AnnexinV-positive events. Thus, isotype controls were measured in parallel to check specimens for background signals and artifacts. MV counts measured with specific antibodies were compared to the number of positive events (i.e. background signals or artifacts) measured in corresponding control samples. The number of measured MV was only used for statistical analysis if its artifact count was less than 20% of the specific signal in panel one and panel three or less than 5 positive events per microliter in panel two.

In vitro stimulation of monocytes and analysis of monocyte-derived MV

To investigate whether subtypes of leukocytes are capable of generating CD31+/CD41− MV that cannot be labeled with one of the above-mentioned common leukocyte markers (panel three), the generation of MV from THP-1 monocytes was induced in vitro. THP-1 monocytes (ATCC TIB-202, Middlesex, UK) were cultured at 37°C in a humidified atmosphere with 5% CO2 in RPMI 1,640 medium supplemented with 10% fetal calf serum, 1% penicillin/streptomycin, and 1% L-glutamine (all from Biochrom, Berlin, Germany). For experiments 2 × 106 THP-1 cells per milliliter were stimulated with 100 μg/mL LPS (Sigma-Aldrich) in 0.2 μm filtered RPMI 1640 without supplements for 20 h. Following stimulation, cell suspension was harvested and centrifuged at 1,000 g for 10 min at 26°C in a 320R centrifuge (Hettich). MV containing supernatant was analyzed with the leukocyte panel (i.e. panel three) in flow cytometry as mentioned above. The percentage of CD31+/CD41− THP-1 derived MV showing positivity for CD45, CD14, CD66b, CD20, or CD3 was calculated (n = 4). CD45 expression on THP-1 cells was verified by using the above-mentioned leukocyte panel and a FcR Blocking Reagent (Miltenyi Biotec, Bergisch Gladbach, Germany) by flow cytometry (n = 4).

Cytokine measurements

For analysis of inflammatory mediators (including interleukin (IL) 6, IL 10, TNF alpha, and soluble E-selectin (sE-selectin)) in plasma samples, a Human Inflammation Panel kit (EPX200-12185-901, affymetrix eBioscience, Vienna, Austria) was used. Plasma samples were thawed at room temperature, diluted 1:2 with Universal Assay Buffer (provided by the manufacturer), and assayed according to the manufacturer's instructions. Measurements were done with a Bio-Plex 200 System, acquisitions and analyses were performed by using Bio-Plex Manager 6.0 (both from Bio-Rad Laboratories Inc, Hercules, CA).

Statistical analysis

Statistical analyses were performed with GraphPad Prism version 5 (GraphPad Software Inc, La Jolla, CA) and SPSS versions 21 and 23 (IBM, Armonk, NY). Quantitative data are presented as median and 25th to 75th quartiles, if not indicated otherwise. Data were tested for normality by using the Shapiro–Wilk test. For comparison of non-parametric values between groups Mann–Whitney U test was used. Correlations between non-parametric variables were analyzed by Spearman rank-order test. A P value below 0.05 was considered statistically significant.


Patients characteristics

Seven female and 23 male patients with septic shock as well as 10 female and 8 male healthy volunteers were enrolled in the study. Median (25th to 75th quartiles) time from initiation of vasopressors to blood draw was 12.6 (3.1–21) h. Median age in the septic shock group was 62.5 (52.0–72.5) compared with 27.5 (26–35) years in the control group. SAPS II score was 65.5 (46.8–78.5), APACHE II score 27 (20–32.3), and a SOFA score 13 (9.8–15) in the septic shock group. ICU mortality was 66.7%, hospital mortality 70%. In 66.7% of patients blood culture revealed a positive result. Primary sepsis focus was the lung in 40%, the urinary tract in 13.3%, skin or soft tissue in 6.7%, and abdomen in 3.3%. In 36.7% no distinct focus could be determined in addition to a positive blood culture. Patients with septic shock had significantly elevated levels of IL-6, IL 10, TNF-alpha, and sE-selectin (Supplemental Table 2, Supplemental Digital Content 1, at Additional patient characteristics are listed in Table 1.

Table 1:
Characteristics of patients with septic shock in total and stratified according to hospital survival

Microvesicle analysis

Both patients with septic shock as well as healthy volunteers did have low levels of MV carrying endothelial-specific epitopes such as CD144, CD62E, or CD106 (Fig. 1). Although statistical analysis showed significantly elevated CD144+ (P = 0.007) as well as CD62E+ (P = 0.009) and significantly lower CD106+ (P = 0.031) MV in patients with septic shock compared with healthy volunteers, those measured MV counts were mostly in the low range of the accuracy of flow cytometric MV detection. The characteristics of the five patients showing higher levels of EMV (CD62E+) are shown in Supplemental Table 3, Supplemental Digital Content 1, at All MV counts are summarized in Table 2.

Fig. 1:
Subtypes of MV per microliter PFP in patients with septic shock compared with healthy volunteers.MV indicates microvesicles, ns, not significant, PFP, platelet-free plasma, * = P <0.05, ** = P <0.01, *** = P <0.001.
Table 2:
Absolute microvesicle counts (MV/μL) in patients with septic shock and healthy volunteers analyzed in 0.3–1 μm (hsFC) and in 0.5–1 μm gates (corresponding to sdFC)

Patients with septic shock showed a 3-fold higher count of CD31+/CD41− MV (58.5 (26.4–101.2) vs. 19.5 (12.8–25.4) MV/μL; P <0.001) compared with healthy volunteers. There was a significant correlation between the number of CD31+/CD41− MV and leukocyte count in patients with septic shock (rs = 0.64; P <0.001). The percentage of CD31+/CD41− MV that stain positive for at least one of the analyzed leukocyte markers (CD45, CD14, CD66b, CD20, or CD3) was 8.9 (3.6–18.6)% in healthy controls and 16.1 (8.7–27.7)% in patients with septic shock (P of absolute values <0.05).

No significant difference could be detected in AnnexinV+, CD31+, CD41+, or CD42b+ MV (Fig. 1). However, we found a highly significant correlation between the numbers of CD41+ MV as well as CD42b+ MV and platelet count in patients with septic shock (rs = 0.733; P <0.001 and rs = 0.576; P = 0.001, respectively).

No effect on MV counts could be detected for aspirine (n = 9), heparin (n = 14), vasopressin (n = 3), paracetamol (n = 4), or erythrocyte concentrates (n = 4).

To quantify the improved detection by hsFC, we performed secondary analyses by setting the gate between 0.5 μm and 1 μm (by referring to Megamix beads (Biocytec, Marseille, France)), which corresponds to MV counts measured with standard flow cytometry (Table 2). Overall, we found a 2.1-fold increase in median. This gain in sensitivity resulted in increased detection of several MV subtypes, such as CD62E+, CD144+, or CD31+/CD41− MV. Additionally, we calculated the amounts of MV sized between 0.3 μm and 0.5 μm (Supplemental Table 4, Supplemental Digital Content 1, at Although the main findings remained comparable in this size range it is remarkable that the count of AnnexinV+ MV was higher in patients with septic shock compared with healthy controls (317.5 (218.2–499.6) vs. 183.1 (135.4–343.9) MV/μL; P = 0.039).

Microvesicles in patients with septic shock who died within 48 h

Six patients (20%) with septic shock died within 48 h after blood samples were drawn. The time from initiation of vasopressors to death ranged from 6.2 to 47.5 h (median 18.3 h) in these patients. They showed significantly higher counts of CD41+ (639.1 (321.3–969.7) vs. 221.5 (119.5–456.9) MV/μL; P = 0.037) compared with survivors (n = 24). Furthermore, they showed significantly elevated levels of CD31+/CD41−/AnnexinV− (51.9 (24.9–259.8) vs. 18.9 (9.7–31) MV/μL; P = 0.028) MV (Fig. 2A), whereas CD31+/CD41−/AnnexinV+ MV were not significantly different (P = 0.445). There were no significant differences in CD144+, CD62E+, or CD106+ MV, platelet or leukocyte count between these two groups. Detailed data are presented in Supplemental Table 5, Supplemental Digital Content 1, at

Fig. 2:
Subgroup analyses of endothelium- (CD144+, CD62E+) MV, platelet- (CD41+), and presumably leukocyte-derived MV (CD31+/CD41−/AnnexinV−).A, Patients with septic shock who did not survive 48 h and patients who did survive more than 48 h after blood draw. B, Patients with septic shock stratified according to SOFA scores at the day of blood draw. C, Patients with septic shock stratified according to ISTH-DIC score. The Bars represent medians and interquartile ranges. AV indicates AnnexinV; ISTH, International Society of Thrombosis and Haemostasis; MV, microvesicles; ns, not significant; SOFA, sequential organ failure assessment score, * = P <0.05, ** = P <0.01.

Microvesicles in patients with septic shock and high SOFA score

Patients with septic shock were stratified according to the SOFA score on day of blood draw (Fig. 2B). The median SOFA score of the study population was used as cutoff. Patients with a SOFA score above 13 (n = 12) had significantly higher counts of CD31+/CD41−/AnnexinV− (45.4 (24.1–87.3) vs. 14.3 (8.6–25.3) MV/μL; P = 0.003) compared to those with a SOFA score equal to or below 13 (n = 18). No differences in the numbers of CD144+, CD62E+, CD106+, or CD41+ MV were observed.

Microvesicles in patients with septic shock and disseminated intravascular coagulation

Analysis of patients with overt DIC as classified by a ISTH-DIC score of at least five (n = 14; 47%) revealed lower counts of CD41+ MV (135.3 (96.5–321.3) vs. 385.5 (219.6–747.5) MV/μL; P = 0.040) as well as a lower platelet count (61.0 (29.3–79) vs. 131.5 (104–204.8) G/L; P <0.001) as compared with patients with an ISTH-DIC score below five (n = 16; 53%). Although there was no significant difference in leukocyte count (9.7 (3.7–20.9) vs. 13.8 (7.9–21.1) G/L; P = 0.473), there was a tendency to higher counts of CD31+/CD41−/AnnexinV− MV (25.6 (18.9–65) vs. 13.7 (7.1–33.5); P = 0.072) in patients with overt DIC compared with patients without overt DIC, respectively (Fig. 2C). There was no significant difference concerning the counts of CD144+, CD62E+, or CD106+ MV.

Characterization of microvesicles released from stimulated monocytes in vitro

After stimulation, THP-1 cells released CD31+/CD41− MV (Supplemental Figures 4 and 5, Supplemental Digital Content 1, at, of which only 26.8% (3.2) [mean (standard deviation)] expressed at least one of the analyzed leukocyte markers (CD45, CD14, CD66b, CD20, or CD3), although parent cells clearly expressed CD45 (Supplemental Figure 6, Supplemental Digital Content 1, at


This is the first study analyzing MV down to a size of 0.3 μm bead-equivalents in patients with septic shock. Although our patients with septic shock had increased counts of several MV subtypes, the numbers of endothelial-specific MV, considered as a manifestation of endothelial pathology, were low. Furthermore, we found a significant association between early mortality and counts of CD41+ MV, reflecting excessive platelet activation, as well as CD31+/CD41−/AnnexinV− MV.

Three-fold higher counts of CD31+/CD41− MV were measured in our patients with septic shock compared with healthy volunteers. Similar findings were reported by Soriano et al. who found elevated levels of CD31+/CD42− MV in patients with severe sepsis. CD31+ (PECAM-1) and simultaneously platelet marker negative (i.e. CD41− or CD42−) MV are frequently interpreted as EMV (16, 22). However, CD31 (PECAM-1) is also present on subtypes of leukocytes and MV originating from them. Based on the numeric discrepancy between CD31+/CD41− MV and MV expressing endothelium-specific markers (i.e. CD144, CD62E, or CD106) in our study, we hypothesize that CD31+/CD41− MV are mostly not endothelium derived. Furthermore, this subtype did significantly correlate with leukocyte count in patients with septic shock. Alternatively, this correlation might be explained by a release of CD31+/CD41− MV from endothelial cells as consequence of leukocyte activation. Interestingly, one patient who was aplastic showing a leukocyte count below 0.1 G/L and a platelet count of 8 G/L exhibited extremely low levels of CD31+/CD41− MV, also suggesting that this subtype may originate from leukocytes. To substantiate this assumption, we additionally reanalyzed samples with a leukocyte panel, including CD45, CD14, CD66b, CD20, and CD3, to determine the percentage of CD31+/CD41− that stain positive for common leukocyte markers. Although CD31+/CD41− MV that carry leukocyte markers were significantly elevated in patients with septic shock, we found that the percentage of CD31+/CD41− MV that stained positive with at least one leukocyte marker was only around 16%. Although this percentage is higher than the 7% as reported by Soriano et al. (16), who used only a single leukocyte marker in a less severe ill cohort of patients, it is not entirely explaining the origin of increased CD31+/CD41− MV in our cohort. Therefore, we investigated in vitro whether subtypes of leukocytes may generate CD31+/CD41− MV that cannot be stained with one of the above-mentioned commonly used leukocyte markers. We found that upon stimulation THP-1 monocytes release CD31+/CD41− MV of which only 26.8% stain positive for one of those leukocyte-specific markers. These findings suggest that at least a relevant part of CD31+/CD41− MV might originate from leukocytes.

Our second major finding was that despite small albeit statistically significant differences of CD144+ (VE-Cadherin), CD62E+ (ELAM-1), and CD106+ (VCAM-1) MV between patients with septic shock and healthy volunteers, counts of MV labeled with these endothelium specific markers were mostly in the low range of the accuracy of flow cytometric MV detection. Only occasional patients with septic shock had markedly increased and measureable levels of MV carrying endothelium specific markers (i.e. CD144+, CD62E+, and CD106+ MV) (Fig. 1). These patients (Supplemental Table 3, Supplemental Digital Content 1, at were characterized by presence of liver injury, coagulation abnormalities and a highly proinflammatory state as well as increased levels of soluble markers of endothelial activation (i.e. soluble E-selectin). Two former studies, in which endothelium-specific markers were used, revealed findings similar to our study: Nieuwland et al. (14) detected increased counts of CD62E+ (ELAM-1) MV in single patients with DIC and menigococcal sepsis, although statistics did not reveal a significant difference between septic patients and healthy volunteers in their study. In the second study low levels of endothelium-specific MV such as CD144+ (VE-Cadherin) and CD62E+ MV (ELAM-1) were reported in septic patients (11).

Furthermore, we analyzed subgroups of patients, which we expected to exhibit a significant endothelial pathology usually contributing to a reduction of antithrombotic properties and to an increased risk of a fatal outcome (23), i.e., patients with overt DIC, patients who died within 48 h and patients with a more severe organ dysfunction (i.e. high SOFA score). The latter two subgroups representing patients with the most pronounced disease had increased counts of CD31+/CD41−/AnnexinV− MV. Since this subtype is negative for the apoptosis-marker AnnexinV we assume that these MV are mostly released by cellular activation rather than by apoptosis, as shown for endothelial cells (21). However, endothelial-specific MV counts were not elevated. Thus, it may be assumed that fatal septic shock is accompanied by leukocyte activation and subsequent release of CD31+/CD41−/AnnexinV− MV. We also consider excessive platelet activation a feature of advanced and lethal septic shock, because patients who died within 48 h did have a significantly higher count of CD41+ (GPIb) MV.

Patients with overt DIC exhibited significantly lower counts of CD41+ MV, probably resulting from a low platelet count in these patients, a parameter that effects the DIC score itself. The finding that CD41+ MV significantly correlated with platelet count is supporting this assumption. Recently, Delabranche et al. (5) reported an association between EMV defined as CD31+ MV as well as CD105+ (i.e. endoglin) MV and early DIC in septic shock. However, we could not detect a significant difference in specific EMV subtypes (i.e. CD144+, CD62E+, or CD106+ MV) between patients with and patients without overt DIC in our study. Possible explanations for these discrepant findings might be that Delabranche et al. used solid phase capturing assays, which might capture even smaller MV than hsFC and applied other markers to analyze EMV.

Strengths and limitations

A major strength of this study is the meticulous exclusion of artifacts inherent to MV analysis in flow cytometry, including filtering of liquids, antibodies, corresponding isotype-control antibodies, and AnnexinV just before usage. However, in critically ill patients with multiple organ failures, sources of artifacts might be multifactorial including high bilirubin, plasma lipids, or different medications among others (24, 25). Some of those factors are not controllable and some might even not have been identified yet. Therefore, a novel approach was applied in our study to increase specificity: if an unspecific signal in a control sample surpassed a defined threshold, its corresponding value measured in the sample labeled with specific antibodies was excluded from subsequent statistical analyses. Although this method might reduce the sensitivity it increases specificity, which is of great importance, especially for low abundant MV subtypes such as EMV.

Second, our data show that hsFC increases sensitivity and discrimination properties of the method as compared with standard flow cytometry. About two times higher median MV counts could be detected by hsFC compared with standard flow cytometry. By increasing the detection rate low abundant MV subtypes such as MV of endothelial origin can be more reliably quantified. Although the overall signal pattern did not change dramatically when comparing the different size ranges (Table 2, Supplemental Table 4, Supplemental Digital Content 1, at slightly changed signals could be detected for distinct MV subtypes, e.g., AnnexinV+ MV that were significantly higher in septic patients in the size range 0.3 of 0.5 μm. Although these findings are compatible with the study from Robert et al. (18), who reported different proportions of small and large MV, depending on the MV subset as well as the clinical status, the biological and clinical relevance of the observed differences remain to be elucidated.

Limitations of this study might comprise the relatively low number of patients included and the difference in age between patients with septic shock and healthy volunteers. Still, stratification according to age did not reveal significant differences in MV counts, neither in septic patients nor in healthy subjects (data not shown). One might argue that healthy volunteers are not an ideal control population (26). However, we decided to use healthy volunteers rather than ICU patients as control group to decrease heterogeneity, which is an approach chosen by several other studies investigating MV in septic shock (11, 14). Another limitation might be the fact that only patients during the early phase of septic shock were assessed. We focused on the first 48 h of septic shock to include severely ill patients with the worst prognosis during a highly inflammatory phase. This is reflected by high levels of cytokines (27), with a predominant increase of proinflammatory mediators (Table 1), as characteristic for early sepsis. Although hsFC is able to resolve MV down to a size of approximately 0.3 μm, this approach does not provide information about MV sized below this detection limit. Since MV circulating in plasma which are sized below approximately 1 μm were analyzed, no conclusions can be drawn concerning aggregated MV, as described by Heloire et al. (28). Although several epitopes known to be characteristic for EMV were analyzed, it cannot be excluded that EMV loose specific epitopes that usually are present on the parent cells. Finally, it should be noted that blood sampling sites were not identical in septic shock patients and healthy controls which might influence the findings.

In conclusion, this study shows that, despite improved sensitivity by hsFC, only occasional patients with septic shock exhibit increased counts of endothelial-specific MV. The sickest patients show features of excessive platelet and presumably leukocyte activation. Although we could demonstrate that CD31+/CD41− MV, which were significantly elevated in septic shock, can be released by leukocytes, the exact origins of this MV subtype remain to be elucidated.


1. Meziani F, Delabranche X, Asfar P, Toti F. Bench-to-bedside review: circulating microparticles—a new player in sepsis? Crit Care 2010; 14:236.
2. Thery C, Ostrowski M, Segura E. Membrane vesicles as conveyors of immune responses. Nat Rev Immunol 2009; 9:581–593.
3. Gyorgy B, Szabo TG, Pasztoi M, Pal Z, Misjak P, Aradi B, Laszlo V, Pallinger E, Pap E, Kittel A, et al. Membrane vesicles, current state-of-the-art: emerging role of extracellular vesicles. Cell Mol Life Sci 2011; 68:2667–2688.
4. Mastronardi ML, Mostefai HA, Meziani F, Martinez MC, Asfar P, Andriantsitohaina R. Circulating microparticles from septic shock patients exert differential tissue expression of enzymes related to inflammation and oxidative stress. Crit Care Med 2011; 39:1739–1748.
5. Delabranche X, Boisrame-Helms J, Asfar P, Berger A, Mootien Y, Lavigne T, Grunebaum L, Lanza F, Gachet C, Freyssinet JM, et al. Microparticles are new biomarkers of septic shock-induced disseminated intravascular coagulopathy. Intensive Care Med 2013; 39:1695–1703.
6. Meziani F, Tesse A, Andriantsitohaina R. Microparticles are vectors of paradoxical information in vascular cells including the endothelium: role in health and diseases. Pharmacol Rep 2008; 60:75–84.
7. Mutunga M, Fulton B, Bullock R, Batchelor A, Gascoigne A, Gillespie JI, Baudouin SV. Circulating endothelial cells in patients with septic shock. Am J Respir Crit Care Med 2001; 163:195–200.
8. Hotchkiss RS, Tinsley KW, Swanson PE, Karl IE. Endothelial cell apoptosis in sepsis. Crit Care Med 2002; 30 (5 Suppl):S225–228.
9. Erdbruegger U, Grossheim M, Hertel B, Wyss K, Kirsch T, Woywodt A, Haller H, Haubitz M. Diagnostic role of endothelial microparticles in vasculitis. Rheumatology 2008; 47:1820–1825.
10. Aras O, Shet A, Bach RR, Hysjulien JL, Slungaard A, Hebbel RP, Escolar G, Jilma B, Key NS. Induction of microparticle- and cell-associated intravascular tissue factor in human endotoxemia. Blood 2004; 103:4545–4553.
11. Joop K, Berckmans RJ, Nieuwland R, Berkhout J, Romijn FP, Hack CE, Sturk A. Microparticles from patients with multiple organ dysfunction syndrome and sepsis support coagulation through multiple mechanisms. Thromb Haemost 2001; 85:810–820.
12. Mortaza S, Martinez MC, Baron-Menguy C, Burban M, de la Bourdonnaye M, Fizanne L, Pierrot M, Cales P, Henrion D, Andriantsitohaina R, et al. Detrimental hemodynamic and inflammatory effects of microparticles originating from septic rats. Crit Care Med 2009; 37:2045–2050.
13. Mostefai HA, Meziani F, Mastronardi ML, Agouni A, Heymes C, Sargentini C, Asfar P, Martinez MC, Andriantsitohaina R. Circulating microparticles from patients with septic shock exert protective role in vascular function. Am J Respir Crit Care Med 2008; 178:1148–1155.
14. Nieuwland R, Berckmans RJ, McGregor S, Boing AN, Romijn FP, Westendorp RG, Hack CE, Sturk A. Cellular origin and procoagulant properties of microparticles in meningococcal sepsis. Blood 2000; 95:930–935.
15. Ogura H, Kawasaki T, Tanaka H, Koh T, Tanaka R, Ozeki Y, Hosotsubo H, Kuwagata Y, Shimazu T, Sugimoto H. Activated platelets enhance microparticle formation and platelet-leukocyte interaction in severe trauma and sepsis. J Trauma 2001; 50:801–809.
16. Soriano AO, Jy W, Chirinos JA, Valdivia MA, Velasquez HS, Jimenez JJ, Horstman LL, Kett DH, Schein RM, Ahn YS. Levels of endothelial and platelet microparticles and their interactions with leukocytes negatively correlate with organ dysfunction and predict mortality in severe sepsis. Crit Care Med 2005; 33:2540–2546.
17. Dignat-George F, Boulanger CM. The many faces of endothelial microparticles. Arterioscler Thromb Vasc Biol 2011; 31:27–33.
18. Robert S, Lacroix R, Poncelet P, Harhouri K, Bouriche T, Judicone C, Wischhusen J, Arnaud L, Dignat-George F. High-sensitivity flow cytometry provides access to standardized measurement of small-size microparticles—brief report. Arterioscler Thromb Vasc Biol 2012; 32:1054–1058.
19. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, Schein RM, Sibbald WJ. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992; 101:1644–1655.
20. Trummer A, De Rop C, Tiede A, Ganser A, Eisert R. Recovery and composition of microparticles after snap-freezing depends on thawing temperature. Blood Coagul Fibrinolysis 2009; 20:52–56.
21. Jimenez JJ, Jy W, Mauro LM, Soderland C, Horstman LL, Ahn YS. Endothelial cells release phenotypically and quantitatively distinct microparticles in activation and apoptosis. Thromb Res 2003; 109:175–180.
22. Lacroix R, Robert S, Poncelet P, Dignat-George F. Overcoming limitations of microparticle measurement by flow cytometry. Semin Thromb Hemost 2010; 36:807–818.
23. Levi M, van der Poll T. Endothelial injury in sepsis. Intensive Care Med 2013; 39:1839–1842.
24. Larson MC, Luthi MR, Hogg N, Hillery CA. Calcium-phosphate microprecipitates mimic microparticles when examined with flow cytometry. Cytometry A 2013; 83:242–250.
25. van Ierssel SH, Hoymans VY, Van Craenenbroeck EM, Van Tendeloo VF, Vrints CJ, Jorens PG, Conraads VM. Endothelial microparticles (EMP) for the assessment of endothelial function: an in vitro and in vivo study on possible interference of plasma lipids. PLoS One 2012; 7:e31496.
26. Cain DJ, del Arroyo AG, Ackland GL. Uncontrolled sepsis: a systematic review of translational immunology studies in intensive care medicine. Intensive Care Med Exp 2014; 2:6.
27. Gogos CA, Drosou E, Bassaris HP, Skoutelis A. Pro- versus anti-inflammatory cytokine profile in patients with severe sepsis: a marker for prognosis and future therapeutic options. J Infect Dis 2000; 181:176–180.
28. Heloire F, Weill B, Weber S, Batteux F. Aggregates of endothelial microparticles and platelets circulate in peripheral blood. Variations during stable coronary disease and acute myocardial infarction. Thromb Res 2003; 110:173–180.

Disseminated intravascular coagulation; endothelium; extracellular vesicles; microparticles; microvesicles; septic shock; sepsis; APACHE II; acute physiology and chronic health evaluation score; ELAM-1; endothelial-leukocyte adhesion molecule 1; EMV; endothelial-derived microvesicles; FS; forward scatter; GPIbα; glycoprotein Ib alpha; GPIIb; glycoprotein IIb; hsFS; high-sensitivity flow cytometry; ICU; intensive care unit; IQR; interquartile range; ISTH-DIC; International Society on Thrombosis and Haemostasis score; MV; microvesicles; PECAM; platelet endothelial cell adhesion molecule; PFP; platelet-free plasma; SAPS II; simplified acute physiology score; SIRS; systemic inflammatory response; SOFA; sequential organ failure assessment score; VCAM-1; vascular cell adhesion protein 1; VE-cadherin; vascular endothelial-cadherin

Supplemental Digital Content

© 2016 by the Shock Society