Author + information
- Received June 21, 2012
- Revision received February 22, 2013
- Accepted March 12, 2013
- Published online October 15, 2013.
- Gabrielle Sarlon-Bartoli, MD, PhD∗,†,
- Youssef Bennis, MD†,
- Romaric Lacroix, MD, PhD†,‡,
- Marie Dominique Piercecchi-Marti, MD, PhD§,
- Michel A. Bartoli, MD, PhD∗,
- Laurent Arnaud‡,
- Julien Mancini, MD, PhD‖,
- Audrey Boudes, MD∗,
- Emmanuelle Sarlon, MD¶,
- Benjamin Thevenin, MD∗,
- Aurelie S. Leroyer, PhD†,
- Christian Squarcioni, MD#,
- Pierre Edouard Magnan, MD∗,
- Françoise Dignat-George, MD, PhD†,‡ and
- Florence Sabatier, MD, PhD†∗∗∗ ()
- ∗Service de Chirurgie Vasculaire, Faculté de Médecine de Marseille, Aix-Marseille Université, Assistance Publique Hôpitaux de Marseille-Hôpital de la Timone, Marseille, France
- †INSERM UMR-1076, Faculté de Pharmacie, Aix-Marseille Université, Marseille, France
- ‡Laboratoire d'Hématologie, Assistance Publique Hôpitaux de Marseille-Hôpital de la Conception, Marseille, France
- §Service de Médecine Légale Droit de la Santé, Faculté de Médecine de Marseille, Aix-Marseille Université, Assistance Publique Hôpitaux de Marseille-Hôpital de la Timone, Marseille, France
- ‖Service de Santé Publique, Assistance Publique Hôpitaux de Marseille-Hôpital de la Timone, Marseille, France
- ¶INSERM U669, Faculté de Médecine Paris Sud, Paris, France
- #Service de Neurologie Vasculaire, Assistance Publique Hôpitaux de Marseille-Hôpital de la Timone, Marseille, France
- ∗∗Laboratoire de Culture et Thérapie Cellulaire, INSERM CICBT 510, Assistance Publique Hôpitaux de Marseille-Hôpital de la Conception, Marseille, France
- ↵∗Reprint requests and correspondence:
Dr. Florence Sabatier, INSERM UMR 1076, Faculté de Pharmacie, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 05, France.
Objectives This study sought to analyze whether the plasmatic level of leukocyte-derived microparticles (LMP) is associated with unstable plaques in patients with high-grade carotid stenosis.
Background Preventive carotid surgery in asymptomatic patients is currently debated given the improvement of medical therapy. Therefore, noninvasive biomarkers that can predict plaque instability are needed. The LMPs, originating from activated or apoptotic leukocytes, are the major microparticle (MP) subset in human carotid plaque extracts.
Methods Forty-two patients with >70% carotid stenosis were enrolled. Using a new standardized high-sensitivity flow cytometry assay, LMPs were measured before thromboendarterectomy. The removed plaques were characterized as stable or unstable using histological analysis according to the American Heart Association criteria. The LMP levels were analyzed according to the plaque morphology.
Results The median LMP levels were significantly higher in patients with unstable plaque (n = 28; CD11bCD66b+ MP/μl 240 [25th to 75th percentile: 147 to 394], and CD15+ MP/μl 147 [60 to 335]) compared to patients with stable plaque (16 [0 to 234] and 55 [36 to 157]; p < 0.001 and p < 0.01, respectively). The increase in LMP levels was also significant when considering only the group of asymptomatic patients with unstable plaque (n = 10; CD11bCD66b+ MP/μl 199 [153 to 410] and CD15+ MP/μl 78 [56 to 258] compared with patients with stable plaque (n = 14; 20 [0 to 251] and 55 [34 to 102]; p < 0.05 and p < 0.05, respectively). After logistic regression, the neurologic symptoms (odds ratio: 48.7, 95% confidence interval: 3.0 to 788, p < 0.01) and the level of CD11bCD66b+ MPs (odds ratio: 24.4, 95% confidence interval: 2.4 to 245, p < 0.01) independently predicted plaque instability.
Conclusions LMP constitute a promising biomarker associated with plaque vulnerability in patients with high-grade carotid stenosis. These data provide clues for identifying asymptomatic subjects that are most at risk of neurologic events.
Surgery for symptomatic carotid stenosis is the gold standard for avoiding recurrence of ipsilateral stroke (1). Management of symptom-free patients is more debated because with systematic use of antiplatelets, statins and angiotensin-converting enzyme inhibitors, the stroke risk of a carotid stenosis is reduced and close to that of preventive surgery (2,3). Identification of biological and imaging markers of atherosclerotic plaque vulnerability is important to predict which patients are most at risk for neurologic events. Membrane microparticles (MP) released by apoptotic or activated cells (such as endothelial cells, leukocytes, and platelets) have been reported to be sensitive markers of atherothrombotic disease progression and potential effectors of vascular dysfunction (4–6). Interestingly, increased leukocyte-derived microparticles (LMP) levels are present in the plasma of patients with cardiovascular risk factors (7) or subclinical atherosclerosis as assessed by ultrasonography (8). The LMPs were found to be the major MP subpopulation in extracts from human plaques after carotid endarterectomy (9). Given the critical involvement of leukocyte activation in plaque formation and destabilization (10), our objective was to test whether plasmatic level of LMP in patients with high-grade carotid stenosis undergoing systematic surgery could be useful to identify patients with unstable plaques.
Forty-two patients with high-grade carotid stenosis who were undergoing carotid endarterectomy in the vascular surgery department of Timone Hospital were enrolled from June 2009 to June 2010 after informed consent was obtained and procedures approved by the local human ethics committee. During the inclusion period, 38 patients were not included because of unavailable biological or histological samples or patient's rejection or patients with patent clinical infection signs, chronic inflammatory diseases, or cancer. The NASCET (North American Symptomatic Carotid Endarterectomy Trial) clinical criteria (1) were used to classify patients as neurologically symptomatic or asymptomatic. High-grade carotid stenosis was defined on computed tomography as a ≥70% reduction in diameter using the NASCET trial technique (1). For the LMP and high-sensitivity C-reactive protein analyses, blood sampling was performed just before surgery after an overnight fast. Surgically removed carotid plaques were collected for histopathological analysis. A group of 20 age-matched healthy volunteers with no cardiovascular risk factors and no medications were recruited as controls.
Leukocyte-derived microparticle measurement using high-sensitivity flow cytometry
Venous blood was collected into 0.129 mol/l tri-sodium citrate Vacutainer tubes (BD Diagnostics, Franklin Lakes, New Jersey). Platelet-free plasma was prepared within 4 h using 2 serial centrifugations (15 min at 1,500g, 2 min at 13,000g) and stored at −80°C until use. For the MP labeling, 30 μl freshly thawed platelet-free plasma was incubated for 30 min at room temperature with 10 μl conjugated monoclonal antibodies (CD11b-FITC [clone Bear1'] and CD66b-FITC [clone UCHT1] or CD15-FITC [clone 80H5]) or isotype and concentration-matched control antibodies from Beckman Coulter (Marseille, France). The platelet-free plasma was then diluted in phosphate-buffered saline without calcium and magnesium. Then, 30 μl Cyto-Count Fluorospheres (Dako, Trappes, France) was added to determine the absolute MP numbers. The LMP determination was performed on a high-sensitivity flow cytometer (Gallios Instrument, Beckman Coulter). Protocol standardization was based on a blend of fluorescent size-calibrated beads (0.1, 0.3, 0.5, and 0.9 μm [Megamix-Plus prototype, Biocytex, Marseille, France]) (11). To obtain better sensitivity in the low size range, the Gallios flow cytometer was set up in the “W2” mode. The lower and upper limits of the MP gate were defined using 0.3-μm and 0.9-μm beads, respectively (12). To evidence the specific advantage of high-sensitivity flow cytometry, LMP determination was also performed on a previous-generation instrument FC500 (Beckman Coulter) with a lower limit of MP gate at 0.5 μm beads-equivalent. The LMP were defined as the CD11bCD66b or CD15 positive events included in the MP region, and the results were expressed as the number of MP/μl of plasma.
Histological assessment of the carotid plaques
The samples were fixed for 24 h in 4% neutral buffered formalin, dehydrated in graded alcohols, cleared in xylene, and embedded in paraffin. Hematoxylin-and-eosin staining were performed on serial sections (5 μm) of each sample, and morphological characteristics of the plaques were established according to the classification defined by the American Heart Association (AHA) (13). The stable lesions are AHA type V: fibrous connective tissue together with extracellular lipids and laminated acellular collagen without endothelial disruption (Va) or only fibrous conjunctive tissue (Vb). The unstable lesions are type VI: ulceration of the endothelial surface (VIa) or recent intraplaque hemorrhage (VIb) or intraplaque thrombosis (VIc). Six sections were analyzed from each patient.
First, the patient characteristics were compared according to the plaque morphology. The results were expressed as median (25th percentile to 75th percentile). Bivariate comparisons were conducted using the Mann-Whitney U test for continuous characteristics and the chi-square or Fisher exact tests for categorical characteristics. Wilcoxon matched pairs test was used to compare LMP levels from the 2 flow cytometry strategies. Spearman's rank correlation coefficients were calculated to determine the strength of the association between the continuous characteristics. A receiver-operating characteristic curve analysis was then conducted to identify the threshold level of LMPs that was capable of predicting an unstable plaque with the shortest distance from the upper left corner of the receiver-operating characteristic curve (minimizing [(1-sensitivity)² + (1-specificity)²]). Positive predictive value (PPV) and negative predictive value (NPV) were analyzed.
A multivariate analysis was carried out by creating a forward stepwise logistic regression model to estimate the probability of an unstable plaque. First, we verified that neurologic events and CD11b66b+ MP level were independently associated with the presence of unstable plaque. Then, age, sex, treatments (antiplatelets, statins, angiotensin-converting enzyme inhibitors), and factors associated with the outcome at the p < 0.20 level in the bivariate analyses (LMP, hypertension, diabetes mellitus, hypercholesterolemia, smoker, stenosis degree, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein) were considered eligible to enter into the logistic regression model. This forward approach was based on the log-likelihood ratio test (entry threshold: p < 0.05). Low-density lipoprotein cholesterol level was secondarily removed from the candidate variables because it was far from statistical significance and had 1 missing datum. The logistic regression model was applied to the whole population and to the population of asymptomatic patients. The only difference was that antiplatelets were not considered in the second analysis because all the asymptomatic patients had such a treatment. Statistical analyses were performed using SPSS version 17.0. (SPSS, Chicago, Illinois). Any p value <0.05 was considered statistically significant.
In the whole population, patients with stable and unstable plaques displayed similar characteristics except for neurologic symptoms, significantly higher in the patients with unstable plaque, and hypercholesterolemia, significantly higher in the patients with stable plaque. Overall, 23 patients were asymptomatic. No significant differences were observed between the asymptomatic patients classified according to plaque morphology (Table 1).
Age of healthy controls and of the whole population of patients was similar as attempted (healthy controls: 67 [66 to 69] years, and entire population of patients: 73 [63 to 78] years; p = 0.32).
Circulating leukocyte-derived microparticles are increased in patients with unstable plaque
In the whole population of patients with carotid stenosis, the median plasmatic LMP levels, as assessed by CD11b66b+ or CD15+ labeling and high-sensitivity flow cytometry analysis, were significantly higher compared to those of healthy controls (181 [70 to 324] MP/μl and 90 [51 to 233] MP/μl vs. 2 [0 to 5] MP/μl and 2 [0 to 16] MP/μl, respectively; p < 0.05). In the whole population of patients, the levels of both types of LMPs were significantly correlated (rho = 0.57; p < 0.001; n = 42).
Circulating leukocyte-derived microparticles independently predict plaque instability
A receiver-operating characteristic analysis indicated that the cutoff point for LMP that best predicted the presence of unstable plaque was 120 MP/μl for CD11b66b+ MPs with an 82% sensitivity, a 71% specificity, an 85% PPV, and a 67% NPV (Table 2). For CD15+ LMP, the cut-off was 58 MP/μl with a 79% sensitivity, a 71% specificity, a 85% PPV, and a 62% NPV (Fig. 3A).
In the asymptomatic patients, the cutoff points for LMP that best predicted the presence of unstable plaque were found to be the same (120 for CD11b66b+ MPs with a 90% sensitivity, a 69% specificity, a 69% PPV, and a 90% NPV; and 58 for CD15+ MPs with an 70% sensitivity, a 77% specificity, a 70% PPV, and a 77% NPV) (Fig. 3B).
After logistic regression, both the neurologic symptoms and the level of CD11b66b+ MP were identified as independent predictors of unstable plaques (neurologic symptoms odds ratio [OR]: 48.7, 95% confidence interval [CI]: 3.0 to 788, p < 0.01; CD11b66b+ MP >120 OR: 24.4, 95% CI: 2.4 to 245, p < 0.01). In the asymptomatic patients, only the level of CD11b66b+ MP was associated with unstable plaques (level >120 OR: 20.3, 95% CI: 1.9 to 219, p < 0.01).
The annual risk of ischemic stroke secondary to a ≥60% asymptomatic carotid stenosis has been reported to be 2% to 3% in randomized trials of endarterectomy versus best medical therapy (3). However, thanks to medical treatment improvement, this risk has been reduced to 0.4% (2), and surgery is likely to be performed for patients with a high neurologic risk. Thus, identification of markers of plaque instability is interesting to make this selection possible. However, none of the circulating markers previously tested for their ability to predict vulnerable carotid plaques allowed differentiation of stable and unstable plaques in asymptomatic patients (14). To our knowledge, this is the first study to demonstrate that high plasmatic level of LMP is associated with unstable plaque in asymptomatic patients with high-grade carotid stenosis. Thus, LMP behave as a new potential biomarker that could be used to identify asymptomatic subjects who are most at risk of neurologic events.
In a previous study (9), MP subsets were measured in plaque extracts after carotid endarterectomy in asymptomatic and symptomatic patients. Although a high proportion of plaque MP was found to originate from leukocytes, circulating level of LMP failed to differentiate the 2 groups of patients. Here, we used a new standardized high-sensitivity flow cytometry method to measure circulating LMP levels. This method, providing improved resolution and decreased background noise, allows the measurement of previously undetectable small MP subpopulation in a standardized manner (12). Consistently, LMP levels measured in patients with carotid stenosis were found to be higher than those determined using previous-generation flow cytometry that do not give reliable access to small-size events. In addition, we demonstrated that including these small-size MP fractions identifies an excessive leukocyte vesiculation occurring in patients with histologically defined unstable carotid plaque compared to patients with stable plaque, whereas the previous-generation method failed to do so. Elevation of LMP subsets remained significant when only asymptomatic patients were considered.
After adjustment for neurologic symptoms, LMP were independently associated with plaque instability. Importantly, medical therapy with angiotensin-converting enzyme inhibitors or statins is presumably able to stabilize atherosclerotic plaques and may influence circulating MP levels (15). However, we did not observe any difference in LMP levels between patients taking or not taking these drugs (data not shown). In addition, these treatments were included in the logic regression model attesting that the link between LMP and plaque vulnerability was independent of medical therapy. These data indicate that CD11b66b+ LMP levels are a promising marker of plaque instability. By contrast, an independent association between LMP expressing CD15 and plaque morphology could not be evidenced. This result suggests that, depending on the marker used, LMP subsets may have various pathological roles and clinical significances, and is consistent with the fact that these markers do not identify strictly similar subpopulations of leukocytes. In contrast to CD15, present mainly in granulocytes, CD11b is also expressed by leukocyte subpopulations that make significant contributions to atherosclerosis progression, such as monocytes, CD8+ T lymphocytes and NK cells (16). In addition, the high expression levels of CD11b and CD66b on parental leukocytes contribute to optimize the sensitivity for MP detection.
The pathophysiological mechanisms underlying the elevation of circulating LMP in patients with unstable carotid plaque remain to be elucidated. One hypothesis is that LMP are generated within the blood compartment or at the vessel wall interface as a result of inflammatory processes, which drives plaque instability. In our study, a role for systemic inflammation is unlikely because no significant elevation of high-sensitivity C-reactive protein was noted in patients with unstable plaque. In addition, the levels of phosphatidylserine-positive MP and endothelial-derived MP, classically increased during systemic inflammation, were similar in the 2 groups of patients (data not shown). Another hypothesis is that LMP originated from plaque rupture because high amounts of MP resulting from apoptotic leukocytes have been reported in vulnerable plaques (9). These MP could be a major determinant of plaque instability because they promote thrombogenicity (9), stimulate endothelial proliferation and intraplaque neovascularization through a CD40-to-CD40L interaction (17), and promote leukocyte recruitment (18) by the transfer of intercellular adhesion molecule-1 to endothelial cells.
The potential limitations of this study include the limited number of subjects and the absence of patient follow-up regarding neurologic events. Particularly, the limited number of subjects limits our inferences regarding the lack for confounding of the several cardiovascular risk factors measured, because of potential over-fitting of multiple logistic regression models. In the same way, the single dataset used to establish MP thresholds and to assess performance may have over-estimated sensitivity, specificity, and positive and negative predictive values. Future large-scale prospective studies are needed to establish whether LMP levels display prognostic value with respect to the occurrence of neurologic events, and whether LMP levels could be used as a noninvasive biomarker with decision-making value in an asymptomatic population of patients who are considered for preventive carotid endarterectomy. One prerequisite is also to extend data on the clinical stability of these markers.
Among the asymptomatic patients with high-grade carotid stenosis, increased circulating LMP level may be useful in identifying a subgroup of patients with histologically defined plaque instability. The LMP may constitute a promising biomarker for the diagnosis of carotid plaque vulnerability and provide encouraging clues for the identification of asymptomatic subjects at high neurologic risk who would benefit the most from carotid surgery.
The authors thank the paramedical teams of the vascular surgery department for their help in conducting this work. They also thank Patricia Stellmann and Patricia Berenger for their technical assistance with the MP analysis. They are grateful to Beckman Coulter for their technical support, and thank Biocytex for the kind gift of the Megamix-Plus standardization beads.
The authors have reported they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- leukocyte-derived microparticle
- negative predictive value
- positive predictive value
- Received June 21, 2012.
- Revision received February 22, 2013.
- Accepted March 12, 2013.
- American College of Cardiology Foundation
- Marquardt L.,
- Geraghty O.C.,
- Mehta Z.,
- Rothwell P.M.
- Morel O.,
- Toti F.,
- Hugel B.,
- et al.
- Mallat Z.,
- Hugel B.,
- Ohan J.,
- Lesèche G.,
- Freyssinet J.M.,
- Tedgui A.
- Rautou P.E.,
- Vion A.C.,
- Amabile N.,
- et al.
- Chironi G.,
- Simon A.,
- Hugel B.,
- et al.
- Leroyer A.S.,
- Isobe H.,
- Lesèche G.,
- et al.
- Angelillo-Scherrer A.
- Robert S.,
- Lacroix R.,
- Poncelet P.,
- et al.
- Stary H.C.,
- Chandler A.B.,
- Dinsmore R.E.,
- et al.
- Nomura S.,
- Shouzu A.,
- Omoto S.,
- Nishikawa M.,
- Iwasaka T.
- Leroyer A.S.,
- Rautou P.E.,
- Silvestre J.S.,
- et al.
- Rautou P.E.,
- Leroyer A.S.,
- Ramkhelawon B.,
- et al.