Liquid Biopsy of Extracellular Microvesicles Predicts Future Major Ischemic Events in Genetically Characterized Familial Hypercholesterolemia Patients Rosa Suades, Teresa Padró, Javier Crespo, Alessandro Sionis, Rodrigo Alonso, Pedro Mata, Lina Badimon Objective—Circulating microvesicles (cMVs) exert regulatory roles in atherothrombosis. Patients with familial hypercholesterolemia (FH) that are at high risk for premature cardiovascular events (CVEs) have previously shown high levels of cMVs related to disease severity. However, much remains unknown about their value as markers of CVE. We sought to investigate the prognostic cMV signature for future major CVE presentation in patients with FH. Approach and Results—Liquid biopsies from genetically characterized patients with FH from the SAFEHEART (Spanish Familial Hypercholesterolemia Cohort Study)-cohort without clinical manifestation of disease at entry that were going to suffer a CVE within a mean period of 3.3±2.6 years postsampling (CVE, N=92) and from age/cardiovascular risk factor/treatment-matched patients with FH that did not suffer an event within the same time-period (non-CVE, N=48) were investigated. cMVs were phenotyped by flow cytometry to identify activated parental cells. Patients with CVE had higher number of overall procoagulant annexin V+-cMVs than non-CVE (P<0.05). Pan-leukocyte-derived and neutrophil- derived cMVs, as well as activated platelet-derived cMVs, were significantly higher in patients with CVE. Baseline number of cMVs derived from lymphocytes, neutrophils, and activated platelets were positively associated with mortality at follow-up (P<0.05). Patient-risk calculated by classical cardiovascular risk-factor scores did not correlate with cMVs. Inclusion of the cMV signature into the SAFEHEART risk model for patients with FH for the prediction of ischemic events increased the area under the curve from 0.603±0.050 to 0.768±0.042 (P<0.005). Conclusions—Patients with FH who are going to suffer a CVE within a mean period of 3.3 years, despite being treated according to guidelines, have ongoing innate immune cell and platelet activation. The proposed cMV signature is a prognostic marker for accelerated atherosclerosis and clinical event presentation in patients with FH. Downloaded from http://ahajournals.org by on November 6, 2020 Visual Overview—An online visual overview is available for this article.   (Arterioscler Thromb Vasc Biol. 2019;39:1172- 1181. DOI: 10.1161/ATVBAHA.119.312420.) Key Words: biomarkers ◼ leukocytes ◼ mortality ◼ platelets ◼ prognosis. A therosclerosis is a multifactorial disease that is the lead- ing cause of morbidity and mortality worldwide.1 One of the main challenges of cardiology is the identification of disease that, in addition to the classical risk factors, may have clinical utility to detect those patients who can suffer an unpre- dictable CVE.7 These findings in a CVE prone patient popula- individuals at risk of developing a major ischemic cardiovas- tion, as FH, may be later on extended to all patients who need cular event (CVE).2 Specifically, heterozygous familial hy- new predictive biomarkers to improve their management.8,9 percholesterolemia (FH) patients, characterized by high LDL Although the atherosclerotic risk is 3- to 13-fold higher in (low-density lipoprotein)-cholesterol plasma levels, show pre- patients with FH than in those without this genetic disease, mature development of atherothrombotic events.3 Indeed, there the impact of acute ischemic events is highly variable among is an increased risk of acute CVE in patients with FH.4,5 Recent patients with FH,10 implying the need for better biomarkers for efforts have aimed to investigate novel biomarkers to identify this heterogeneous high-risk population. Besides, FH remains those patients with the highest risk of developing a major CVE widely underdiagnosed and undertreated worldwide.11 Thus, (MACE). Presently, in the context of dyslipidemia, current FH offers a good clinical model to study new biomarkers de- laboratory assessment of cardiovascular risk mainly relies on rived from cells directly involved in the disease process that lipoproteins.6 Liquid biopsies try to identify cell markers of may accurately report on disease progression. Received on: December 17, 2018; final version accepted on: March 31, 2019. From the Cardiovascular-Program ICCC, Research Institute Hospital Santa Creu i Sant Pau, IIB-Sant Pau, Barcelona, Spain (R.S., T.P., J.C., L.B.); Cardiology Unit, Department of Medicine Solna, Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden (R.S.); CIBERCV Instituto de Salud Carlos III, Madrid, Spain (T.P., J.C., A.S., L.B.); Acute and Intensive Cardiac Care Unit, Cardiology Department, Hospital Santa Creu i Sant Pau, Barcelona, Spain (A.S.); Fundación Hipercolesterolemia Familiar, Madrid, Spain (R.A., P.M.); Department of Nutrition, Clínica Las Condes Santiago, Chile (R.A.); and Cardiovascular Research Chair, UAB, Barcelona, Spain (L.B.). The online-only Data Supplement is available with this article at https://www.ahajournals.org/doi/suppl/10.1161/ATVBAHA.119.312420. Correspondence to Lina Badimon, B Pharm, MSc, PhD, FESC, FAHA, Cardiovascular-Program ICCC, Research Institute Hospital Santa Creu i Sant Pau, Sant Antoni Maria Claret 167, 08025 Barcelona, Spain. Email lbadimon@santpau.cat © 2019 American Heart Association, Inc. Arterioscler Thromb Vasc Biol is available at https://www.ahajournals.org/journal/atvb DOI: 10.1161/ATVBAHA.119.312420 1172 Suades et al   Circulating MVs in Cardiovascular Risk Assessment   1173 with FH presented a MACE such as unstable angina, acute myocar- Nonstandard Abbreviations and Acronyms dial infarction, sudden death, or ischemic stroke (CVE, N=92) within a mean period of 3.31±2.6 years postsampling as compared to FH a-pMVs activated platelet-derived microvesicles patients without a CVE within the same follow-up period of time AUC area under the curve (8 years; non-CVE [nCVE], N=48; Table I in the online-only Data AV annexin V Supplement). The Coordinating Center was responsible for managing cMVs circulating microvesicles the patient follow-up, and patients were contacted annually by tel- ephone by trained staff using a standardized phone call protocol.23 CVE cardiovascular event The CVE and nCVE groups were matched by age, sex, and dem- FH familial hypercholesterolemia ographics as well as LLT and cardiovascular comorbidities (Table I FRS Framingham risk score in the online-only Data Supplement, Graphic Abstract [the Graphic hs-CRP high-sensitivity C-reactive protein Abstract shows the schematic overview of the study design]). Clinical data (sociodemographic, lifestyle, medical, and therapeutic) were LDL low-density lipoprotein obtained from all subjects using a standardized report form at the LLT lipid-lowering therapy inclusion (Table I in the online-only Data Supplement). Data related ℓMVs lymphocyte-derived microvesicles to LLT included statin type, dose, time, and compliance. All selected LMVs leukocyte-derived microvesicles subjects had been prescribed an LLT according to current guide- lines.4,24–26 Maximum statin dose (simvastatin 40 mg/d, fluvastatin 80 mMVs monocyte-derived microvesicles mg/d, atorvastatin 80 mg/d, and rosuvastatin 20–40 mg/d), as well nCVE non-CVE as maximum combined therapy, and maximum LLT were defined as nMVs neutrophil-derived microvesicles previously reported.27,28 Adherence to LLT was assessed by an indi- PFP platelet-free plasma rect method with a single question.29 No patient presented pregnancy or had past history of cancer because these conditions are known to pMVs platelet-derived microvesicles independently impair cMV number. ROC receiver operating characteristic The study was approved by the Local Ethics Committee of the SAFEHEART Spanish Familial Hypercholesterolemia Cohort Study Investigación Clínica Fundación Jimenez Diaz (protocol’s number: TSP-1 thrombospondin-1 01/09) was conducted according to the Declaration of Helsinki, and a written informed consent was obtained from all participants before the study. The results of the study are presented in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology Circulating microvesicles (cMVs) are small phospholipid guidelines30 as well as Transparent Reporting of a Multivariable extracellular vesicles shed by stressed, activated, senescence, Prediction Model for Individual Prognosis or Diagnosis statement.31 and apoptotic cells. cMVs play a pivotal role on atherosclerosis, inflammation, and thrombosis by means of cell signaling and Blood Sampling, Laboratory Measurements, intercellular crosstalk.12–16 Increased levels of cMVs have been Downloaded from http://ahajournals.org by on November 6, 2020 Genetic Analysis, and cMV Isolation found in a wide variety of atherothrombotic disorders in associ- Venous blood samples were collected after a 10 to 14 hours of fasting ation with the disease severity.17–20 Patients with FH had previ- from the cubital vein without tourniquet using a 20-gauge needle into ously shown to have high levels of phenotypically characterized serum, EDTA, and 3.8% sodium citrate tubes, for biochemical, geno- cMVs even when their LDL levels were controlled by lipid- typing, and microvesicle analysis, respectively.32 All samples were lowering therapy (LLT).21,22 However, their prognostic value as processed identically and within the first 2 hours after extraction. For lipid profile, total cholesterol, triglycerides, and HDL (high-density markers of major cardiovascular disease remains poorly defined. lipoprotein)-cholesterol were measured by standardized enzymatic Differential expression levels of cell-specific cMVs may methods; serum LDL-cholesterol concentration was calculated using lead to improve our understanding of pathophysiological mech- the Friedewäld formula33; and lipoprotein (a) levels were measured anisms related to clinical outcomes and help us to predict its using a turbidimetric method. Genetic diagnosis of FH was made using a DNA-microarray (LIPOCHIP).34 Mutations were classified as occurrence. Profiling of cMVs shed by injured cardiovascular receptor negative or receptor defective, depending on their functional cells may provide information on ongoing disease and, hence, class. Platelet-free plasma (PFP) for analysis of the cMV fraction was be markers of future adverse ischemic CVE. To the best of our prepared by 2-step high-speed centrifugation (1258×g, 20 minutes at knowledge, no previous studies have addressed whether there is room temperature) to ensure the complete removal of cells, as pre- a specific cMV signature in patients with FH at high cardiovas- viously described.22 PFP samples were tested with a cell counter for the absence of residual cells. PFP aliquots were snap-frozen in liquid cular risk of developing a MACE. This study aims to investi- nitrogen and immediately stored, for an identical time interval, at gate whether MV level and phenotype in asymptomatic patients −80°C until flow cytometric studies.35 with FH are able to predict an adverse cardiovascular outcome. cMV Isolation Materials and Methods Frozen PFP aliquots (250 µL) were thawed on melting ice for 1 The data that support the findings of this study are available from the hour and centrifuged at 20000×g 30 minutes to pellet cMVs. The corresponding author upon reasonable request. cMV-enriched pellets were washed once with citrate-PBS solution (citrate-PBS; 1.4 mmol/L phosphate, 154 NaCl, 10.9 mmol/L triso- dium citrate, and pH 7.4) before a second equal centrifugation step Study Design and Population was made. Finally, the remaining cMV pellets were resuspended in The present study comprised patients with clinical and genetic diag- citrate-PBS to a final volume of 100 µL. nosis of heterozygous FH from the SAFEHEART (Spanish Familial Hypercholesterolemia Cohort Study),21,23 a multicenter, nationwide, long-term, prospective cohort study of a molecularly defined patients Flow Cytometry Analysis with heterozygous FH. None of the patients with FH participating Triple-label flow cytometric analysis was performed as previously in the present study (N=140) had suffered a clinical event before described.19,21,22,36 Briefly, washed cMV suspensions were incubated entering the SAFEHEART cohort. Sixty-six percent of the patients with combinations of AV (annexin V) with 2 cell surface-specific 1174   Arterioscler Thromb Vasc Biol   June 2019 monoclonal antibodies, or the isotype-matched control antibodies (Table the moment of inclusion,4,24–26 and there was no significant dif- I in the online-only Data Supplement). Samples were diluted before ference on either the type or the duration of the treatment be- being immediately analyzed on a FACSCantoII Becton Dickinson flow cytometer. Acquisition was performed for 1 minute per sample at low- tween CVE and nCVE groups. There were also no differences in flow rate. cMVs were identified and quantified based on their FSC/SSC the usage of other cardiovascular drugs. Patients with CVE had (forward scatter/side scatter) characteristics (obtained with gain settings a nonsignificant trend to higher ApoB levels and clinical pre- in the logarithmic scale) according to their size, binding to AV, and reac- sentation of xanthomas compared with patients without CVE. tivity to cell-specific monoclonal antibody. Gate limits were established Patients with CVE predominantly presented an AMI (N=41, as previously described using a Flow-Check Size Range Calibration Kit (Polysciences) and an in vitro platelet-derived microvesicle population 45%) while unstable angina occurred in 21 patients (23%), sud- as positive control (Figure I in the online-only Data Supplement). The den death in 16 (17%), and ischemic stroke in 14 (15%) during lower detection limit was placed as a threshold above the electronic follow-up. Only the first event during follow-up was considered. background noise of the flow cytometer and a threshold was set at SSC parameter. Data were analyzed with FACSDiva software (version 6.1.3, Becton Dickinson). The concentration (number of cMVs per μL of Liquid Biopsies: cMV Profile in plasma) was determined according to Nieuwland procedure,19,21,22 based Patients With and Without CVE on sample’s volume, flow cytometer’s flow rate, and the number of flu- Cell origin and cell activation markers were analyzed to identify orescence-positive events. Intraassay cardiovascular of cMV counts was the cMV signature of patients with CVE at follow-up. Figure 1 3.1%, while interassay cardiovascular was 5.4%. illustrates the pattern of cMVs according to their parental cell marker. The total number of AV-positive cMVs (AV+-cMVs) Statistical Analysis was significantly higher in patients with CVE than in patients Results are reported as median±interquartile range except for contin- without CVE (P<0.05, Figure 1A). While levels of total cir- uous variables of patient characteristics that are given as mean (±SEM). Normal sample distribution was determined via Kolmogorov-Smirnov culating endothelial-derived (CD62E+-eMVs; Figure 1C), red test. Statistical analysis was performed by the nonparametric Mann- blood cell-derived (CD235a+-erythrocyte-derived microves- Whitney. P values were adjusted by the false discovery rate using the icles; Figure 1D), and platelet-derived microvesicles (CD61+- graphically sharpened method described by Benjamin and Hochberg.37 and CD41+-pMVs; Figure 1E) as well as bearing tissue factor The new calculated adjusted Q values (false discovery rate adjusted) (CD142+; Figure 1F) did not significantly change, there were indicate the probability of false positives for variables considered to be significant. In the present study, it refers to Q values <0.05, which significant increases in white cell type-derived (Figure 1B) and represents >95% truly positives for differentially expressed cMV activated platelet-derived cMVs (Figure 1E) between patients subtypes. Strength of the association between continuous variables with and without CVE. Indeed, pan-leukocyte-derived cMVs was calculated by Spearman correlation. Risk estimations based (CD45+-LMVs) were significantly higher in patients with CVE on the Framingham Risk Score (FRS),24 Systematic Coronary Risk as compared to patients without CVE. About leukocyte-specific Downloaded from http://ahajournals.org by on November 6, 2020 Evaluation,38 Registre Gironí del Cor,39 and SAFEHEART-CVEs risk estimator8 equations were performed for each individual. Binary lo- subsets of the innate immunity cell-derived cMVs, patients with gistic regression models were performed to estimate predicted prob- CVE presented higher numbers of neutrophil-derived cMVs abilities for CVE with combination of cMVs, as well as cardiovascular (CD15+-nMVs), while no alteration in lymphocyte-derived risk factor models. To avoid overfitting in the multivariable model, we (CD3+-ℓMVs) and monocyte-derived (CD14+-mMVs) cMVs combined cMV biomarkers by using binary logistic regression as a single score. Cox proportional hazards regression (time-to-event) anal- (Figure 1B). These derangements of white blood cell-derived ysis was performed using the predicted probabilities for combination of cMVs suggest a proinflammatory profile in patients prone to de- cMVs and individual values for the CVD risk scores. The performance velop a MACE. Numbers of activated monocytes (CD14+/11b+- of the models was evaluated with respect to their discrimination and a-mMVs) cMVs were significantly reduced in patients with calibration ability on the basis of the Hosmer-Lemeshow type χ2 good- CVE as compared to patients without CVE. Furthermore, cMVs ness of fit statistic test. An associated receiver operating characteristic (ROC) curve analysis was calculated to evaluate cMVs as independent carrying platelet activation epitopes (activated αIIBβ3 [PAC1+], prognostic biomarkers. Then, the areas under the dependent ROC thrombospondin 1 [TSP1+] and P-selectin [CD62P+]-a-pMVs) curves (AUC) from each model were compared using the DeLong were significantly increased in patients with CVE suggesting method.40 StatView 5.0.1 (Abacus Concepts) and SPSS statistics 22.0.0 a higher level of platelet reactivity despite of similar levels of (SPSS) were used for all statistical tests. A P<0.05 was considered sta- total pMVs between both groups (Figure 1E). When we stratify tistically significant and all reported P values were 2-tailed. our population by sex, there is a similar cMV profile associated to CVE (Table III in the online-only Data Supplement). Among Results patients with CVE, a correlation was found between type of mu- Characteristics of the Patient Population tation (null/defective) and pMVs bearing TSP-1 (thrombospon- Baseline clinical characteristics of the studied population are din-1; Figure II in the online-only Data Supplement), showing given in Table I in the online-only Data Supplement. Between that those patients with a lifetime null LDL-receptor function CVE and nCVE groups, there were no significant differences in have higher platelet activation. the clinical, sociodemographic, and laboratory characteristics, as well as in cardiovascular risks factors. The mean age of patients cMVs and Mortality in Patients With High-Risk FH without CVE was 60, whereas the mean age was 58 in patients Specific populations of cMVs were related to mortality at fol- with CVE. There were more men than women in both CVE low-up in the high-risk FH patient population. Similar to the and nCVE groups, specifically, 70% and 58%, respectively. All CVE-driven cMV signature, cMVs derived from white blood patients were at high cardiovascular risk because they presented cells (Figure 2A) and activated platelets (Figure 2B) were asso- genetically identified and clinically diagnosed heterozygous FH. ciated with mortality at follow-up. Concerning cMVs derived All patients were under LLT according to current guidelines at from leukocyte lineage cells, baseline numbers of cMVs derived Suades et al   Circulating MVs in Cardiovascular Risk Assessment   1175 Figure 1. Changes in total, white blood cell, red blood cell, endothelial cell, and platelet-derived circulating microvesicles (cMVs) between cardiovascular event (CVE) and non-CVE (nCVE) high-risk familial hypercholesterolemic (FH) patients. Box and whisker plots showing levels (per µL of platelet-free plasma [PFP]) of (A) total annexin V-positive (phosphatidylserine-rich [PS]+), (B) specific white blood cell-derived cMVs from leukocytes (LMVs), neutrophils (nMVs), lympho- cytes (ℓMVs), monocytes (mMVs), and activated monocytes (a-mMVs), (C) endothelial cell-derived microvesicles (eMVs), (D) erythrocyte-derived microvesicles (ErMVs), (E) platelet-derived microvesicles (pMVs) bearing parental cell markers (PECAM-1 [CD31+], glycoprotein IIIa [CD61+], and glycoprotein IIb/IIIa [CD41a+]) and platelet activation (a-pMVs) epitopes (activated αIIBβ3 [PAC1+], P-Selectin [CD62P+] and thrombospondin-1 [TSP-1+]), and (F) cMVs bearing tissue factor [CD142+] (TF+-cMVs) in patients with CVE (N=92) compared with patients with nCVE (N=48). Data are expressed as median (interquartile range). *P<0.05 and **P<0.01 for difference using Mann-Whitney U nonparametric test. §Q value <0.05 and §§Q value <0.01 after false discovery rate (FDR)-adjustment. from lymphocytes (CD3+-ℓMVs) and neutrophils (CD15+- these data indicate a higher proinflammatory and prothrombotic Downloaded from http://ahajournals.org by on November 6, 2020 nMVs) were positively associated with mortality at follow-up state in those patients prone to die within 3 years. (Figure 2A; P<0.05), while low levels of mMVs (CD14+-mMVs) were significantly found in patients who died at follow-up. cMVs and CRP in Patients With High-Risk FH Besides, numbers of survivors and nonsurvivors at follow-up To further investigate whether systemic inflammation of patients also directly correlated with activated platelet-derived cMVs with high-risk FH was related to cMV phenotype, cMV data bearing TSP-1 epitope (TSP1+-a-pMVs, P<0.05; Figure 2B). were analyzed according to the level of hs-CRP (high-sensitiv- Interestingly, white blood cell- and activated platelet-derived ity C-reactive protein; below and above 1.5 mg/L). Although cMVs were also associated with mortality at follow-up in both levels of hs-CRP in patients with CVE (2.68 mg/L) and patients sexes (Table III in the online-only Data Supplement). Altogether without CVE (1.90 mg/L) were nonsignificantly (Table I in the Figure 2. Specific circulating microvesicles (cMVs) associate with exitus. Box and whisker plots showing levels (per µL of platelet-free plasma [PFP]) of (A) white blood cell-derived cMVs from lymphocytes (ℓMVs), monocytes (mMVs), and neutrophils (nMVs), and (B) microvesicles derived from activated platelets (a-pMVs) carrying thrombospondin-1 [TSP-1+] marker in non-exitus (N=102) and exitus (N=38) patients. Data are expressed as median (interquartile range). AV indicates Annexin V. *P<0.05 and **P<0.01 for difference using Mann-Whitney U nonparametric test. §Q value <0.05 and §§Q value <0.01 after false dis- covery rate (FDR)-adjustment. 1176   Arterioscler Thromb Vasc Biol   June 2019 probabilities of LMVs (CD45+), nMVs (CD15+), pMVs (CD31+ and CD41a+/61+), and a-pMVs (CD62P+) showed the high- est AUC value: 0.745±0.043 (P<0.001; 95% CI, 0.661–0.830; Figure 5A), indicating that clustering these cMVs provides a sig- nificant prognostic value for future CVE in patients with high- risk FH. We also tested the added predictive ability of cMVs by examining the improvement in global AUC obtained when adding the combined probability for the leukocyte- and activated platelet-derived cMV (L/p-cMV) signature to the best cardio- vascular risk factor models for our specific high-risk FH pop- ulation: FRS for coronary heart disease (% of risk at 10 years) and SAFEHEART-CVE 10-year risk estimator. To avoid over- fitting in the multivariable model, the cMV biomarkers were combined in a single score equation by using logistic regression Figure 3. Circulating microvesicles (cMVs) and hs-CRP (high-sensitivity C-reactive protein). Box and whisker plots showing levels (per µL of platelet- for binary outcome. Specifically, the Hosmer-Lemeshow type χ2 free plasma [PFP]) of (A) neutrophil-derived microvesicles (nMVs) and (B) goodness of fit test was of χ2(8)=6.28 (probability >χ2=0.615) microvesicles derived from activated platelets (a-pMVs) carrying thrombo- for cMV model, χ2(8)=4.49 (probability >χ2=0.810) for FRS- spondin-1 [TSP-1+] marker in patients classified by low (N=73) and high cMV model, of χ2(8)=10.81 (probability >χ2=0.213) for cMVs (N=62) hs-CRP levels. Data are expressed as median (interquartile range). AV indicates Annexin V. *P<0.05 for difference using Mann-Whitney U nonpara- and SAFEHEART 5-year risk model and of χ2(8)=7.28 (prob- metric test. §Q value <0.05 after false discovery rate (FDR)-adjustment. ability >χ2=0.507) for cMVs and SAFEHEART 10-year risk model. Interestingly, the inclusion of this cMV signature into the FRS-CVFR-model for the prediction of major ischemic online-only Data Supplement), in those patients with higher events increased the AUC from 0.574±0.051 (P=0.150, [95% levels of hs-CRP there were positive and significant correlations CI, 0.475–0.674]) with only classical risk factors-FRS model with the degree of cell activation and cMV shedding (Figure 3) to 0.760±0.042 (P<0.001, [95% CI, 0.677–0.843]; P<0.005, likely reflecting the proinflammatory state of high-risk FH pa- Figure 5B). Similarly, the L/p-cMV signature enhanced sub- tient population. In line with previous data, cMVs derived from stantially the discrimination power of the SAFEHEART 10-year neutrophils (CD15+-nMVs; Figure 3A) and cMVs carrying the risk model for the prediction of CVE with an improvement in platelet activation marker TSP-1 (TSP1+-a-pMVs; Figure 3B) AUC from 0.603±0.050 (P=0.045, [95% CI, 0.505–0.701]) were higher in patients with higher levels of hs-CRP (P<0.05). Downloaded from http://ahajournals.org by on November 6, 2020 to 0.771±0.042 (P<0.001, [95% CI, 0.688–0.853]; P<0.005; This trend was also found when sex was taken into account Figure 5C). Results about the prognosis of patients based on (Table III in the online-only Data Supplement). cMVs and FRS model as well as cMVs and SAFEHEART 5- and 10-year CVE risk estimator models are shown in Table IV Cardiovascular Risk Scores Are Poor Predictors in the online-only Data Supplement. Multiple linear regression of Future CVE in Patients With High-Risk FH analysis was performed to rule out multicolinearity between Using ROC curve analysis distinct models of cardiovascular risk cMV variables (Table V in the online-only Data Supplement). score estimators were investigated for their accuracy in identify- Similarly, the prognostic value of cMVs for future MACE was ing those patients prone to develop a MACE. FRS for coronary also investigated in males and females. C statistics analyses for heart disease (% of risk at 10 years) was calculated using the high- the L/p-cMV signature showed significant and high AUC values risk Framingham Heart Study equations of the NCEP (National in both sexes (Figure III in the online-only Data Supplement). Cholesterol Education Program).24 Patients were stratified accord- Adding this cMV signature into the cardiovascular risk factor ing to their FRS and there was no significant difference between models (FRS and SAFEHEART 10-year risk) for the prediction patients with and without CVE. Total numbers of AV+-cMVs did of major ischemic events increased the AUCs in all cases. Of not show a significant correlation with FRS (Figure 4A). In a note, the L/p-cMV signature enhanced substantially the discrim- similar fashion, primary prevention cardiovascular risk estima- ination power of the cardiovascular risk factor models in the fe- tors Systematic Coronary Risk Evaluation (SCORE; Figure 4B) male subgroup (Figure IIIB in the online-only Data Supplement). and Registre Gironí del Cor (Regicor; Figure 4C) showed also poor discriminatory power for future CVE presentation. As ex- Discussion pected, both SAFEHEART-CVE risk estimator at 5- and 10-year FH is a monogenic disorder associated with lifelong ex- showed a predictive value for future MACE in our population posure to very-high LDL-cholesterol plasma levels and slightly below the significance threshold (Figure 4D), suggesting increased risk (up to 20 times) of early atherothrombotic that other contributing factors such as cMVs might help to im- disease.3,4 Coronary artery disease is the main cause of prove the prediction of CVE in these patients. morbidity and mortality in individuals with FH.5 We have previously reported that patients with FH at high risk for Prognostic Value of Liquid Biopsy premature ischemic CVE have high levels of circulating cMVs in Patients With FH MVs related to atherosclerotic plaque burden (measured by ROC curve analysis was also used to assess the accuracy of al- CMR) even when their LDL-cholesterol levels were con- tered cMV signature in predicting future MACE in the high-risk trolled by LLT.21,22 This is the first study to demonstrate FH patient cohort. C statistics analysis for combined predicted that cMVs are significant prognostic indicators of major Suades et al   Circulating MVs in Cardiovascular Risk Assessment   1177 Downloaded from http://ahajournals.org by on November 6, 2020 Figure 4. Discriminatory performance of cardiovascular risk scores in the prediction of future major cardiovascular events (CVEs). A, Correlation between total AV+- circulating microvesicles (cMVs) with individual Framingham Risk Score for each individual (N=140) by Spearman correlation nonparametric test. A–D, Receiver operating characteristic (ROC) curve analyses (C statistics) of distinct cardiovascular risk factor models: Framingham Risk Score [FRS] (A), Systematic Coronary Risk Evaluation (SCORE; B) and Registre Gironí del Cor (Regicor; C); SAFEHEART (Spanish Familial Hypercholesterolemia Cohort Study) CVEs risk estimator (D), for the prediction of major CVE presentation in high cardiovascular risk familial hypercholesterolemic (FH) patients (N=140). AV+-cMVs, annexin V-positive cMVs. AUC indicates area under the curve; and AV, Annexin V. ischemic events and the combination of these cMV-based Nozaki et al44 reported that increased levels of CD144+-cMVs biomarkers have a high predictive value in high-risk asymp- were independent predictors of acute coronary syndromes and tomatic patients with FH. cardiovascular death in a population with coronary artery di- Interactions between blood and vascular cells are key sease. Besides, Sinning et al45 found elevated levels of CD31+/ features in atherothrombosis progression. A growing body AV+-cMVs also in patients with stable coronary artery disease, of evidence suggests that blood cell-derived MVs might play assessed by coronary angiography, which later develop a major an important role in vascular function by means of intercel- adverse CVE. Albeit risk stratification in patients with coronary lular communication,41 predominantly on inflammatory,42 and artery disease is highly relevant, the latter association was highly thrombotic processes,43 which are crucial in the development influenced by diabetes mellitus, a condition known to present of CVE. Accordingly, our findings strongly indicate that cMV high circulating levels of CD31+/AV+-MVs.46 Interestingly, those shedding directly relates to the progression of atherosclerotic subjects from PREDIMED trial (Prevention With Mediterranean lesion leading to adverse cardiovascular outcomes in both men Diet) who despite following a Mediterranean diet supplemented and women. In particular, we described increased levels of with extra virgin olive oil still suffered a CVE during follow-up, cMVs originated from pan-leukocytes, neutrophils, and acti- presented increased levels of ℓMV,47 extending our previous find- vated platelets in a molecularly defined population of patients ings in which CD3+/CD45+-ℓMVs were able to accurately dif- with FH that are going to suffer a MACE within the next few ferentiate between lipid-rich and fibrous atherosclerotic plaques years. And, consistently, white blood cell- and platelet-derived in asymptomatic patients with FH.21 Indeed, the prevalence cMV phenotype was also directly related to mortality at fol- of subclinical atherosclerotic burden of patients with FH has low-up. Few studies have addressed the potential link between been shown by aortic magnetic resonance imaging48 and coro- cMVs and CVE presentation in primary prevention and none nary computed tomographic angiography.5 In secondary pre- of them in asymptomatic high-risk patients with FH without vention, contradictory results exist. In this regard, high levels previous cardiovascular disease. of CD14+/11b+-mMVs were shown to be predictors of future 1178   Arterioscler Thromb Vasc Biol   June 2019 Downloaded from http://ahajournals.org by on November 6, 2020 Figure 5. Circulating microvesicles (cMVs) as prognostic markers of future major cardiovascular events (CVEs). Discriminatory power of (A) cMV signature in comparison with (B) Framingham Risk Score (FRS) and (C) SAFEHEART (Spanish Familial Hypercholesterolemia Cohort Study) CVEs risk estimator for the prediction of major CVE presentation in high cardiovascular risk familial hypercholesterolemic (FH) patients (N=140). Prognostic value (*P<0.05) was deter- mined by receiver operating characteristic (ROC) curve analyses (C statistics) and the comparison of the ROC-derived area under the curves (AUCs) by the DeLong method. L/p-cMVs indicates leukocyte- and activated platelet-derived cMVs. MACE in patients with ST-segment–elevation myocardial infarc- occurs in the coronary circulation of acute coronary syndrome tion.49 However, low levels of CD11b+-LMVs were associated patients after percutaneous coronary intervention.53 Recently, with both high-risk atherosclerotic lesions and early recurrence CD15+-MVs derived from neutrophils increased inflammatory of CVE in patients with acute coronary syndromes after coro- response and mortality in a mouse model of sepsis.54 Indeed, nary stenting.50 In agreement, in the present study, mMVs were Lewis X Antigen (CD15) is known to act as an adhesive sub- significantly decreased in patients with CVE and negatively as- strate for endothelium, a monocyte activator as well as a signal- sociated to cardiovascular mortality. Analysis of CD14+-mMVs ing initiator of proinflammatory effects in monocytes.55 nMVs might reflect the heterogeneity of blood monocytes and in-depth were previously shown to induce myeloperoxidase-mediated analyses of the differential subset-specific mMV contribution to damage of vascular endothelium.56 Similarly, nMVs showed atherothrombosis disease deserves further examination. inflammatory and adhesive properties (specifically with mono- Our data reflects ongoing activation of proinflammatory cytes and ECs).57 In addition, proteolytically active transmem- cells in the evolution of atherogenesis even in the absence of brane proteases carried by nMVs promote accelerated risk of overt systemic inflammation as measured by hs-CRP. High human abdominal aortic aneurysm in smokers.58 numbers of cMVs originated from the white blood cell lin- Here, in our study, nMVs were raised in those patients eage, specifically pan-leukocytes (CD45+-LMVs) and neu- with high hs-CRP levels, as a surrogate marker of systemic trophils (CD15+-nMVs), were associated with enhanced inflammation in patients with FH. Consistently, it has been cardiovascular complications in high-risk patients with FH. found that low hs-CRP is associated with reduced risk of in- Accordingly, LMVs were found to be the main pool of MVs cident stroke, coronary heart disease, and death in high cardi- within atherosclerotic plaques51 and patients with high-grade ovascular risk patients.59 Thus, altogether our results support carotid stenosis17 and asymptomatic subjects with subclinical the relevance of the inflammatory process in the pathogenesis atherosclerotic burden52 presented also with high levels of of CVD. Although statins downregulate inflammatory media- LMVs, reinforcing the idea of LMV contribution to atheroscle- tors, as shown in the hallmark JUPITER trial (Justification rotic plaque progression. Of note, an acute release of nMVs for the Use of Statins in Prevention: an Intervention Trial Suades et al   Circulating MVs in Cardiovascular Risk Assessment   1179 Evaluating Rosuvastatin),60 patients with FH constitute a spe- outcomes improves risk stratification of the before mentioned cial population with a lifetime exposure of the vasculature classical risk factor models by a substantial increase in the to high LDL-cholesterol levels and, consequently, they are AUC of ROC curve analysis, as indicative of the significant characterized by chronic vascular cell activation and ensuing and sensitive predictive ability of cMVs in this context. inflammation, in line with the emerging concept of residual in- There are well-described differences between men and flammatory risk as shown in the CANTOS trial (Canakinumab women in the clinical presentation, pathophysiology, and re- Antiinflammatory Thrombosis Outcome Study).61 sponse to treatment of CVD that might be explained by both Not only proinflammatory but also prothrombotic MVs biological and environmental processes. This is also true for are found as bioeffectors of CVE in patients with FH. cMVs patients with cardiovascular risk factors such as dyslipidemia. from activated platelets (TSP1+- and CD62P+-a-pMVs) cor- Because women show a lower perception of cardiovascular related with CVE presentation and mortality at follow-up as risk and are less keen to seek medical care when experiencing well as with high levels of hs-CRP, suggesting that MV shed- symptoms, their prognosis is worse than in men. In addition, ding from activated platelets reflect a deleterious environment sex differences in the field of cardiovascular risk prediction re- in the vascular compartment of patients with FH predisposing main to be further elucidated. Thus, here, we have investigated to adverse cardiovascular outcomes. Despite it is well known cMVs as prognostic factors of future MACE by stratifying our that pMVs exert key roles in all pathophysiological stages data by sex and showing that sex condition does not influence of atherothrombotic disease worsening and exacerbating its the cardiovascular risk based on cMV profile. progression,13 no previous study has pointed out circulating The current study has some caveats and limitations that pMVs as predictive markers of MACE. warrant discussion. The main caveats are that measurements It is well known that lipid infiltration in the vessel wall of circulating MVs in daily clinical practice are still limited induces tissue factor release from vascular resident cells.62 because of inherent complex technology needs and that there We and others have previously reported increased numbers of is a lack of standardization. A second aspect is the relatively TF-rich cMVs in patients with FH.22,63 In this study, all patients small sample size that does not represent a more general pop- are FH and the numbers of TF+-cMVs in those patients with FH ulation of patients with FH. However, the study evaluates a prone to develop a MACE are not significantly different from highly detailed group of patients with molecularly confirmed those not suffering an event. Intriguingly, FH patients with a FH diagnosis and documented cardiovascular ischemic event similar degree of cardiovascular risk and atherosclerotic plaque presentation, a rather unique characteristic that confers a re- burden can evolve distinctly in terms of CVE presentation. Our markable strength to a study in the field of circulating MVs, results seem to unveil a fine-tuning regulation of all cells present usually centered on a case-control approach. Further larger Downloaded from http://ahajournals.org by on November 6, 2020 in the vascular compartment in the atherosclerotic process. studies are warranted to assess the tentative extension of these Prevention strategies with state-of-the-art LLT and car- findings to patients without FH but also to wider asympto- diovascular risk factor control have shown to be effective in matic populations and contexts. reducing cardiovascular outcomes. However, a high number Taken together, our data shed light into a cMV-mediated of patients still suffer unexpectedly from CVE. Because of a modulation of pathophysiological mechanisms in athero- lack of genetic testing, FH remains underdiagnosed11 posting thrombosis progression indicating that leukocytes and plate- asymptomatic FH population at the highest risk of major ad- lets are activated in some patients with FH, although treated verse CVE. Current risk scores for primary prevention are not as per guidelines. A specific cMV signature directly predicts a appropriate for high-risk patients with FH and long-term ex- detrimental vascular milieu that will trigger CVE and cardio- posure to LDL-cholesterol plasma levels likely because cardi- vascular death. Early prediction of vascular injury may allow ovascular risk might be underestimated. Indeed, no risk score better risk stratification and improved prognosis. achieved significance in the prediction of future CVE in our patient group. Although a recent study has suggested that pool Acknowledgments cohort equations substantially overestimate atherosclerotic We thank the Spanish Familial Hypercholesterolemia Foundation cardiovascular risk in a middle-aged adult population,64 bet- (FHF) and physicians that contribute to the FHF-patient cohort ter existing risk prediction tools to accurately reclassify and SAFEHEART (Spanish Familial Hypercholesterolemia Cohort stratify patients in the clinical management of FH are urgently Study) recruitment. We also thank the Fundació d’Investigació needed. This is particularly important because of the new po- Cardiovascular (FIC)-Fundación Jesús Serra, Barcelona, Spain, for their continuous support. We are indebted to the patients with familial tent therapeutic options now available to reduce LDL as the hypercholesterolemia (FH) and relatives for their valuable contribu- PCSK9 (proprotein convertase subtilisin/kexin 9) inhibitors. tion and willingness to participate. Hence, we investigated the added value of detrimental cMV signature on top of main cardiovascular risk scores for primary Sources of Funding prevention namely Systematic Coronary Risk Evaluation and This study was supported by grants from Spanish Ministry of Economy Registre Gironí del Cor and for high-risk patients including and Competitiveness of Science (SAF2016-76819-R to L. Badimon); FRS for hard coronary events and the SAFEHEART risk esti- Institute of Health Carlos III, ISCIII (TERCEL—RD16/0011/0018; mator. As indicated by the C statistics, the COX Proportional and FIS PI16/01915 to T. Padró); We thank the continuous support Hazards Regression Model combining cMVs profiles pro- of the Generalitat of Catalunya (Secretaria d’Universitats i Recerca del Departament d’Economia i Coneixement de la Generalitat, 2017 vides high discriminatory power between patients with and SGR 1480); and FEDER “Una Manera de Hacer Europa”. R. Suades without CVE. An important finding of the present study is that is the recipient of a European Society of Cardiology (ESC) Research L/p-cMV signature beyond being associated to cardiovascular Grant R-2017–070. 1180   Arterioscler Thromb Vasc Biol   June 2019 Disclosures 19. Suades R, Padró T, Crespo J, Ramaiola I, Martin-Yuste V, Sabaté M, Sans-Roselló J, Sionis A, Badimon L. Circulating microparticle sig- None. nature in coronary and peripheral blood of ST elevation myocardial in- farction patients in relation to pain-to-PCI elapsed time. Int J Cardiol. References 2016;202:378–387. doi: 10.1016/j.ijcard.2015.09.011 20. Suades R, Padró T, Vilahur G, Martin-Yuste V, Sabaté M, Sans-Roselló J, 1. Sanz J, Moreno PR, Fuster V. The year in atherothrombosis. J Am Coll Sionis A, Badimon L. Growing thrombi release increased levels of Cardiol. 2013;62:1131–1143. doi: 10.1016/j.jacc.2013.06.045 CD235a(+) microparticles and decreased levels of activated platelet- 2. Fuster V, Kovacic JC. Acute coronary syndromes: pathology, diagnosis, derived microparticles. Validation in ST-elevation myocardial infarction genetics, prevention, and treatment. Circ Res. 2014;114:1847–1851. doi: patients. J Thromb Haemost. 2015;13:1776–1786. doi: 10.1111/jth.13065 10.1161/CIRCRESAHA.114.302806 21. Suades R, Padró T, Alonso R, López-Miranda J, Mata P, Badimon L. 3. Berberich AJ, Hegele RA. The complex molecular genetics of familial Circulating CD45+/CD3+ lymphocyte-derived microparticles map lipid- hypercholesterolaemia. Nat Rev Cardiol. 2019;16:9–20. doi: 10.1038/ rich atherosclerotic plaques in familial hypercholesterolaemia patients. s41569-018-0052-6 Thromb Haemost. 2014;111:111–121. doi: 10.1160/TH13-07-0612 4. Nordestgaard BG, Chapman MJ, Humphries SE, et al; European 22. Suades R, Padró T, Alonso R, Mata P, Badimon L. High levels of TSP1+/ Atherosclerosis Society Consensus Panel. Familial hypercholesterolaemia CD142+ platelet-derived microparticles characterise young patients with is underdiagnosed and undertreated in the general population: guidance high cardiovascular risk and subclinical atherosclerosis. Thromb Haemost. for clinicians to prevent coronary heart disease: consensus statement of the 2015;114:1310–1321. doi: 10.1160/TH15-04-0325 European Atherosclerosis Society. Eur Heart J. 2013;34:3478a–3490a. 23. Mata N, Alonso R, Badimón L, et al. Clinical characteristics and doi: 10.1093/eurheartj/eht273 evaluation of LDL-cholesterol treatment of the Spanish Familial 5. Perez de Isla L, Alonso R, Mata N, et al. Coronary heart disease, periph- Hypercholesterolemia Longitudinal Cohort Study (SAFEHEART). Lipids eral arterial disease, and stroke in familial hypercholesterolaemia: insights Health Dis. 2011;10:94. doi: 10.1186/1476-511X-10-94 from the safeheart registry (spanish familial hypercholesterolaemia cohort 24. National Cholesterol Education Program (NCEP) Expert Panel on study). Arterioscler Thromb Vasc Biol. 2016;36:2004–2010. Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults 6. Hoefer IE, Steffens S, Ala-Korpela M, et al; ESC Working Group (Adult Treatment Panel III). Third report of the National Cholesterol Atherosclerosis and Vascular Biology. Novel methodologies for bio- Education Program (NCEP) Expert Panel on Detection, Evaluation, and marker discovery in atherosclerosis. Eur Heart J. 2015;36:2635–2642. Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel iii) doi: 10.1093/eurheartj/ehv236 final report. Circulation. 2002;106:3143–3421. 7. Khot UN, Khot MB, Bajzer CT, Sapp SK, Ohman EM, Brener SJ, 25. Wierzbicki AS, Humphries SE, Minhas R; Guideline Development Group. Ellis SG, Lincoff AM, Topol EJ. Prevalence of conventional risk factors Familial hypercholesterolaemia: summary of NICE guidance. BMJ. in patients with coronary heart disease. JAMA. 2003;290:898–904. doi: 2008;337:a1095. doi: 10.1136/bmj.a1095 10.1001/jama.290.7.898 26. Reiner Z, Catapano AL, De Backer G, et al. Esc/eas guidelines for the 8. Perez de Isla L, Alonso R, Mata N, et al. Predicting cardiovascular events management of dyslipidaemias: the task force for the management of dys- in familial hypercholesterolemia: the safeheart registry (spanish familial lipidaemias of the european society of cardiology (esc) and the european hypercholesterolemia cohort study). Circulation. 2017;135:2133–2144. atherosclerosis society (eas). Eur Heart J. 2011;32:1769–1818. 9. Villa G, Wong B, Kutikova L, Ray KK, Mata P, Bruckert E. Prediction 27. Perez de Isla L, Alonso R, Watts GF, et al. Attainment of ldl-cholesterol of cardiovascular risk in patients with familial hypercholesterolae- treatment goals in patients with familial hypercholesterolemia: 5-year Downloaded from http://ahajournals.org by on November 6, 2020 mia. Eur Heart J Qual Care Clin Outcomes. 2017;3:274–280. doi: safeheart registry follow-up. J Am Coll Cardiol. 2016;67:1278–1285. 10.1093/ehjqcco/qcx011 28. Pijlman AH, Huijgen R, Verhagen SN, Imholz BP, Liem AH, Kastelein JJ, 10. Khera AV, Won HH, Peloso GM, et al. Diagnostic yield and clinical utility Abbink EJ, Stalenhoef AF, Visseren FL. Evaluation of cholesterol low- of sequencing familial hypercholesterolemia genes in patients with se- ering treatment of patients with familial hypercholesterolemia: a large vere hypercholesterolemia. J Am Coll Cardiol. 2016;67:2578–2589. doi: cross-sectional study in The Netherlands. Atherosclerosis. 2010;209:189– 10.1016/j.jacc.2016.03.520 194. doi: 10.1016/j.atherosclerosis.2009.09.014 11. Sturm AC, Knowles JW, Gidding SS, et al; Convened by the Familial 29. Gehi AK, Ali S, Na B, Whooley MA. Self-reported medication adherence Hypercholesterolemia Foundation. Clinical genetic testing for familial and cardiovascular events in patients with stable coronary heart disease: hypercholesterolemia: JACC scientific expert panel. J Am Coll Cardiol. the heart and soul study. Arch Intern Med. 2007;167:1798–1803. doi: 2018;72:662–680. doi: 10.1016/j.jacc.2018.05.044 10.1001/archinte.167.16.1798 12. Badimon L, Suades R, Arderiu G, Peña E, Chiva-Blanch G, Padró T. 30. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Microvesicles in atherosclerosis and angiogenesis: from bench to bed- Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting side and reverse. Front Cardiovasc Med. 2017;4:77. doi: 10.3389/fcvm. of Observational Studies in Epidemiology (STROBE) statement: guide- 2017.00077 lines for reporting observational studies. PLoS Med. 2007;4:e296. doi: 13. Badimon L, Suades R, Fuentes E, Palomo I, Padró T. Role of platelet- 10.1371/journal.pmed.0040296 derived Microvesicles as crosstalk mediators in atherothrombosis and future 31. Collins GS, Reitsma JB, Altman DG, Moons KG; TRIPOD Group. pharmacology targets: a link between inflammation, atherosclerosis, and Transparent reporting of a multivariable prediction model for individual thrombosis. Front Pharmacol. 2016;7:293. doi: 10.3389/fphar.2016.00293 prognosis or diagnosis (TRIPOD): the TRIPOD statement. The TRIPOD 14. Ridger VC, Boulanger CM, Angelillo-Scherrer A, et al. Microvesicles in Group. Circulation. 2015;131:211–219. doi: 10.1161/CIRCULATIONAHA. vascular homeostasis and diseases. Thromb Haemost. 2017;117:1296– 114.014508 1316. doi: 10.1160/TH16-12-0943 32. Kong F, Zhang L, Wang H, Yuan G, Guo A, Li Q, Chen Z. Impact of 15. Suades R, Padró T, Badimon L. The role of blood-borne microparticles in collection, isolation and storage methodology of circulating microvesicles inflammation and hemostasis. Semin Thromb Hemost. 2015;41:590–606. on flow cytometric analysis. Exp Ther Med. 2015;10:2093–2101. doi: doi: 10.1055/s-0035-1556591 10.3892/etm.2015.2780 16. Boulanger CM, Loyer X, Rautou PE, Amabile N. Extracellular vesicles 33. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration in coronary artery disease. Nat Rev Cardiol. 2017;14:259–272. doi: of low-density lipoprotein cholesterol in plasma, without use of the pre- 10.1038/nrcardio.2017.7 parative ultracentrifuge. Clin Chem. 1972;18:499–502. 17. Sarlon-Bartoli G, Bennis Y, Lacroix R, Piercecchi-Marti MD, Bartoli MA, 34. Alonso R, Defesche JC, Tejedor D, Castillo S, Stef M, Mata N, Arnaud L, Mancini J, Boudes A, Sarlon E, Thevenin B, Leroyer AS, Gomez-Enterria P, Martinez-Faedo C, Forga L, Mata P. Genetic diagnosis Squarcioni C, Magnan PE, Dignat-George F, Sabatier F. Plasmatic level of familial hypercholesterolemia using a DNA-array based platform. Clin of leukocyte-derived microparticles is associated with unstable plaque Biochem. 2009;42:899–903. doi: 10.1016/j.clinbiochem.2009.01.017 in asymptomatic patients with high-grade carotid stenosis. J Am Coll 35. Yuana Y, Bertina RM, Osanto S. Pre-analytical and analytical issues in the Cardiol. 2013;62:1436–1441. doi: 10.1016/j.jacc.2013.03.078 analysis of blood microparticles. Thromb Haemost. 2011;105:396–408. 18. Sionis A, Suades R, Sans-Roselló J, Sánchez-Martínez M, Crespo J, Padró T, doi: 10.1160/TH10-09-0595 Cubedo J, Ferrero-Gregori A, Vila-Perales M, Duran-Cambra A, Badimon L. 36. Suades R, Padró T, Alonso R, Mata P, Badimon L. Lipid-lowering therapy Circulating microparticles are associated with clinical severity of persistent with statins reduces microparticle shedding from endothelium, plate- ST-segment elevation myocardial infarction complicated with cardiogenic lets and inflammatory cells. Thromb Haemost. 2013;110:366–377. doi: shock. Int J Cardiol. 2018;258:249–256. doi: 10.1016/j.ijcard.2017.10.044 10.1160/TH13-03-0238 Suades et al   Circulating MVs in Cardiovascular Risk Assessment   1181 37. Benjamini Y, Hochberg Y. On the adaptive control of the false discovery coronary syndromes. J Thromb Haemost. 2011;9:1870–1873. doi: fate in multiple testing with independent statistics. J Educ Behav Stat. 10.1111/j.1538-7836.2011.04418.x 2000;25:60–83 51. Leroyer AS, Isobe H, Lesèche G, Castier Y, Wassef M, Mallat Z, 38. Perk J, De Backer G, Gohlke H, et al; European Association for Binder BR, Tedgui A, Boulanger CM. Cellular origins and thrombogenic Cardiovascular Prevention & Rehabilitation (EACPR); ESC Committee activity of microparticles isolated from human atherosclerotic plaques. for Practice Guidelines (CPG). European Guidelines on cardiovascular J Am Coll Cardiol. 2007;49:772–777. doi: 10.1016/j.jacc.2006.10.053 disease prevention in clinical practice (version 2012). The Fifth Joint 52. Chironi G, Simon A, Hugel B, Del Pino M, Gariepy J, Freyssinet JM, Task Force of the European Society of Cardiology and Other Societies Tedgui A. Circulating leukocyte-derived microparticles predict subclinical on Cardiovascular Disease Prevention in Clinical Practice (constituted atherosclerosis burden in asymptomatic subjects. Arterioscler Thromb Vasc by representatives of nine societies and by invited experts). Eur Heart J. Biol. 2006;26:2775–2780. doi: 10.1161/01.ATV.0000249639.36915.04 2012;33:1635–1701. doi: 10.1093/eurheartj/ehs092 53. Martinez GJ, Barraclough JY, Nakhla S, Kienzle V, Robertson S, Mallat Z, 39. Marrugat J, Solanas P, D’Agostino R, Sullivan L, Ordovas J, Cordón F, Celermajer DS, Patel S. Neutrophil-derived microparticles are released Ramos R, Sala J, Masià R, Rohlfs I, Elosua R, Kannel WB. [Coronary into the coronary circulation following percutaneous coronary intervention risk estimation in spain using a calibrated framingham function]. Rev Esp in acute coronary syndrome patients. Biosci Rep. 2017;37:BSR20160430. Cardiol. 2003;56:253–261. 54. Johnson BL 3rd, Midura EF, Prakash PS, Rice TC, Kunz N, Kalies K, 40. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas Caldwell CC. Neutrophil derived microparticles increase mortality and under two or more correlated receiver operating characteristic curves: a the counter-inflammatory response in a murine model of sepsis. Biochim nonparametric approach. Biometrics. 1988;44:837–845. Biophys Acta Mol Basis Dis. 2017;1863(10 pt B):2554–2563. doi: 41. Mause SF, Weber C. Microparticles: protagonists of a novel commu- 10.1016/j.bbadis.2017.01.012 nication network for intercellular information exchange. Circ Res. 55. Lo SK, Golenbock DT, Sass PM, Maskati A, Xu H, Silverstein RL. 2010;107:1047–1057. doi: 10.1161/CIRCRESAHA.110.226456 Engagement of the Lewis X antigen (CD15) results in monocyte activa- 42. Hoyer FF, Giesen MK, Nunes França C, Lütjohann D, Nickenig G, tion. Blood. 1997;89:307–314. Werner N. Monocytic microparticles promote atherogenesis by modulat- 56. Pitanga TN, de Aragão França L, Rocha VC, Meirelles T, Borges VM, ing inflammatory cells in mice. J Cell Mol Med. 2012;16:2777–2788. doi: Gonçalves MS, Pontes-de-Carvalho LC, Noronha-Dutra AA, dos-Santos WL. 10.1111/j.1582-4934.2012.01595.x Neutrophil-derived microparticles induce myeloperoxidase-mediated 43. Suades R, Padró T, Vilahur G, Badimon L. Circulating and platelet- damage of vascular endothelial cells. BMC Cell Biol. 2014;15:21. doi: derived microparticles in human blood enhance thrombosis on ath- 10.1186/1471-2121-15-21 erosclerotic plaques. Thromb Haemost. 2012;108:1208–1219. doi: 57. Gasser O, Hess C, Miot S, Deon C, Sanchez JC, Schifferli JA. 10.1160/TH12-07-0486 Characterisation and properties of ectosomes released by human polymor- 44. Nozaki T, Sugiyama S, Koga H, Sugamura K, Ohba K, Matsuzawa Y, phonuclear neutrophils. Exp Cell Res. 2003;285:243–257. Sumida H, Matsui K, Jinnouchi H, Ogawa H. Significance of a mul- 58. Folkesson M, Li C, Frebelius S, Swedenborg J, Wågsäter D, Williams KJ, tiple biomarkers strategy including endothelial dysfunction to improve Eriksson P, Roy J, Liu ML. Proteolytically active ADAM10 and ADAM17 risk stratification for cardiovascular events in patients at high risk for carried on membrane microvesicles in human abdominal aortic aneurysms. coronary heart disease. J Am Coll Cardiol. 2009;54:601–608. doi: Thromb Haemost. 2015;114:1165–1174. doi: 10.1160/TH14-10-0899 10.1016/j.jacc.2009.05.022 59. Penson PE, Long DL, Howard G, Toth PP, Muntner P, Howard VJ, 45. Sinning JM, Losch J, Walenta K, Böhm M, Nickenig G, Werner N. Circulating Safford MM, Jones SR, Martin SS, Mazidi M, Catapano AL, Banach M. CD31+/Annexin V+ microparticles correlate with cardiovascular out- Associations between very low concentrations of low density lipopro- Downloaded from http://ahajournals.org by on November 6, 2020 comes. Eur Heart J. 2011;32:2034–2041. doi: 10.1093/eurheartj/ehq478 tein cholesterol, high sensitivity C-reactive protein, and health outcomes 46. Berezin AE, Kremzer AA, Berezina TA, Martovitskaya YV. The pat- in the Reasons for Geographical and Racial Differences in Stroke tern of circulating microparticles in patients with diabetes mellitus with (REGARDS) study. Eur Heart J. 2018;39:3641–3653. doi: 10.1093/ asymptomatic atherosclerosis. Acta Clin Belg. 2016;71:38–45. doi: eurheartj/ehy533 10.1080/17843286.2015.1110894 60. Ridker PM, Danielson E, Fonseca FA, Genest J, Gotto AM Jr, Kastelein JJ, 47. Chiva-Blanch G, Suades R, Crespo J, Vilahur G, Arderiu G, Padró T, Koenig W, Libby P, Lorenzatti AJ, MacFadyen JG, Nordestgaard BG, Corella D, Salas-Salvadó J, Arós F, Martínez-González MA, Ros E, Fitó M, Shepherd J, Willerson JT, Glynn RJ; JUPITER Study Group. Rosuvastatin Estruch R, Badimon L. CD3(+)/CD45(+) and SMA-α(+) circulating mic- to prevent vascular events in men and women with elevated C-reactive pro- roparticles are increased in individuals at high cardiovascular risk who tein. N Engl J Med. 2008;359:2195–2207. doi: 10.1056/NEJMoa0807646 will develop a major cardiovascular event. Int J Cardiol. 2016;208:147– 61. Ridker PM, Everett BM, Thuren T, et al; CANTOS Trial Group. 149. doi: 10.1016/j.ijcard.2016.01.211 Antiinflammatory therapy with canakinumab for atherosclerotic disease. 48. Caballero P, Alonso R, Rosado P, Mata N, Fernández-Friera L, N Engl J Med. 2017;377:1119–1131. doi: 10.1056/NEJMoa1707914 Jiménez-Borreguero LJ, Badimon L, Mata P. Detection of subclin- 62. Llorente-Cortés V, Otero-Viñas M, Camino-López S, Llampayas O, ical atherosclerosis in familial hypercholesterolemia using non- Badimon L. Aggregated low-density lipoprotein uptake induces mem- invasive imaging modalities. Atherosclerosis. 2012;222:468–472. doi: brane tissue factor procoagulant activity and microparticle release in 10.1016/j.atherosclerosis.2012.02.043 human vascular smooth muscle cells. Circulation. 2004;110:452–459. 49. Chiva-Blanch G, Bratseth V, Ritschel V, Andersen GØ, Halvorsen S, doi: 10.1161/01.CIR.0000136032.40666.3D Eritsland J, Arnesen H, Badimon L, Seljeflot I. Monocyte-derived cir- 63. Nielsen MH, Irvine H, Vedel S, Raungaard B, Beck-Nielsen H, Handberg culating microparticles (CD14+, CD14+/CD11b+ and CD14+/CD142+) A. The Impact of lipoprotein-associated oxidative stress on cell-specific are related to long-term prognosis for cardiovascular mortality in microvesicle release in patients with familial hypercholesterolemia. Oxid STEMI patients. Int J Cardiol. 2017;227:876–881. doi: 10.1016/j.ijcard. Med Cell Longev. 2016;2016:2492858. doi: 10.1155/2016/2492858 2016.11.302 64. Emdin CA, Khera AV, Natarajan P, Klarin D, Baber U, Mehran R, Rader DJ, 50. Faille D, Frere C, Cuisset T, Quilici J, Moro PJ, Morange PE, Bonnet JL, Fuster V, Kathiresan S. Evaluation of the pooled cohort equations for Alessi MC. CD11b+ leukocyte microparticles are associated with high- prediction of cardiovascular risk in a contemporary prospective cohort. risk angiographic lesions and recurrent cardiovascular events in acute Am J Cardiol. 2017;119:881–885. doi: 10.1016/j.amjcard.2016.11.042 Highlights • Overall levels of circulating microvesicles that exert key regulatory roles in atherothrombosis are found to be significantly elevated in patients with high-risk familial hypercholesterolemic 3 years before cardiovascular event onset. • Pan-leukocyte-derived and activated neutrophil-derived circulating microvesicles as well as circulating microvesicles bearing markers of platelet activation (P-selectin and thrombospondin-1) are significantly increased in patients about to suffer an ischemic event. • A leucocyte and platelet activation-based circulating microvesicle signature emerge as an accurate prognostic marker that is able to improve the prediction of ischemic events in patients with asymptomatic high cardiovascular risk familial hypercholesterolemia.