Published ahead of print on February 28, 2008, doi:10.1164/rccm.200710-1458OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200710-1458OC
Circulating Endothelial Microparticle Levels Predict Hemodynamic Severity of Pulmonary Hypertension1 Division of Cardiology, Department of Medicine, University of California, San Francisco, San Francisco, California Correspondence and requests for reprints should be addressed to Yerem Yeghiazarians, M.D., F.A.C.C., F.A.H.A., F.S.C.A.I., Division of Cardiology, University of California, San Francisco, 505 Parnassus Avenue, San Francisco, CA 94143-0103. E-mail: yeghiaza{at}medicine.ucsf.edu
Rationale: Circulating microparticles (MPs) are submicron membrane fragments shed from damaged or activated vascular cells. Endothelial MPs are a biological marker of dysfunctional endothelium. Vascular remodeling and endothelial dysfunction are involved in pulmonary hypertension (PH). Objectives: We tested the hypothesis that circulating MPs are increased in patients with PH and that identifiable subgroups of MPs predict the hemodynamic severity of this condition progression. Methods: Patients (n = 24; age, 54 ± 4 yr) undergoing right heart catheterization for precapillary PH without any endothelium-active vasodilator therapy participated in the study. Age- and sex-matched healthy control subjects (n = 20) were included. Endothelial (PECAM+ [CD31+]/ CD41–, VE-cadherin+ [CD144+], and E-selectin+ [CD62e+]), platelet (CD41+), leukocyte-derived (CD45+), and annexin V+ MPs were measured by flow cytometry in platelet-free plasma from venous blood. Measurements and Main Results: Levels of circulating endothelial PECAM+, VE-cadherin+, E-selectin+, and leukocyte-derived MPs, but not platelet and annexin V+ MPs, were increased in subjects with PH compared with control subjects (P < 0.01 each). PECAM+ and VE-cadherin+ MP levels significantly correlated with mean pulmonary artery pressure (r = 0.92 and r = 0.87, respectively), pulmonary vascular resistance (r = 0.78 and r = 0.73), and mean right atrial pressure (r = 0.43, and r = 0.46) and correlated inversely with cardiac index (r = –0.59 and r = –0.52). These relationships were not observed for other MP subgroups, and persisted in multivariate analysis after adjustment for confounding factors. Conclusions: In subjects with precapillary PH, levels of circulating endothelial and leukocyte MPs were increased compared with control subjects. In addition, levels of PECAM+ and VE-cadherin+, but not E-selectin+, endothelial MPs predicted hemodynamic severity of the disease.
Key Words: pulmonary hypertension • endothelium • microparticles • hemodynamics
Pulmonary hypertension (PH) is a complex, multifactorial disease defined by a mean pulmonary arterial pressure (  ) greater than 25 mm Hg at rest (1). Severe obstructive remodeling of the pulmonary vessel wall, multifocal thrombosis, and enhanced vasoconstriction contribute to the increase in pulmonary vascular resistance (2, 3). Several aspects of the pathophysiology of the disease remain largely unknown, including changes in the biology of pulmonary vascular endothelium. Endothelial cell morphology is changed during PH with disorganized proliferation and expression of an "activated" cell phenotype, characterized by modified expression of molecules involved in angiogenesis and local cell growth suppression (4). Although its initial cause has not been clearly elucidated, this endothelial activation results in increased secretion of cellular growth factors, impaired balance between vasorelaxant and vasoconstrictive factors, enhanced expression of adhesion molecules, loss of endothelial barrier function, and increased secretion of procoagulant factors. All these changes appear to act synergistically in the progression of the disease (4). Microparticles (MPs) are shed membrane vesicles released during apoptosis and/or activation of various cell types (5, 6). An increase in the number of circulating MPs of various origins has been reported in association with several cardiovascular diseases (5). Moreover, circulating endothelial microparticles (EMPs) have been reported as a marker of endothelial injury and systemic vascular remodeling (7). The number of EMPs appears to be correlated with the degree of endothelial dysfunction in vivo in patients with chronic renal failure (7) or coronary artery disease (8). Furthermore, in addition to being a marker of endothelial injury, EMPs could be playing an active role in exacerbating endothelial dysfunction. Both in vivo– and in vitro–generated isolated EMPs can induce endothelial dysfunction when incubated with rat aortic rings in vitro, by decreasing the production of nitric oxide (7, 9). To date, there are no available data assessing the presence of increased circulating microparticles, including the endothelial subpopulation, in the setting of PH. To test the hypothesis that MPs might be predictive of PH hemodynamic severity, we investigated the levels and cellular origin of circulating MPs in patients with this condition and correlated these findings with indices of disease severity.
Inclusion Criteria Patients who had already been scheduled for right heart catheterization as part of planned work-up for PH were asked to participate in this study. Subjects at least 18 years old with precapillary PH and not receiving endothelium-active vasodilator therapy (current treatment with endothelin-1 receptor antagonists, prostacyclin analogs, or type 5 phosphodiesterase inhibitors) were eligible for inclusion. Exclusion criteria included the following: left ventricular ejection fraction less than 50%, left ventricular end-diastolic pressure greater than 15 mm Hg; recent history (<3 mo) of pulmonary embolism; aortic or mitral regurgitation or stenosis. Moreover, we excluded subjects with conditions known to be associated with an increase in circulating microparticle number, such as chronic renal failure (creatinine clearance < 50 ml/min/m2), acute coronary syndromes, hemolytic anemia syndromes (sickle cell disease and β-thalassemia), and uncontrolled systemic hypertension (defined as diastolic blood pressure > 120 mm Hg and no hypertensive drugs for > 48 h) (5). Plasma hemoglobin, high-sensitivity C-reactive protein (hsCRP), brain natriuretic peptide, and low-density lipoprotein cholesterol were measured in every patient. For subsequent analysis, we divided the patients with PH into two subgroups, depending on the absence or the presence of at least one cardiovascular risk factor (systemic hypertension, diabetes, dyslipidemia, active or former smoking status, body mass index > 28 kg/m2). The control group consisted of healthy age- and sex-matched subjects without cardiovascular risk factors or other medical conditions who did not undergo right heart catheterization. The study protocol was approved by the Committee on Human Research (University of California, San Francisco, San Francisco, CA) and all subjects gave written, informed consent.
Right Heart Catheterization
Circulating MP Measurement Various types of circulating microparticles were analyzed, as reported previously (7, 11): CD31+ (PECAM+)/CD41+ (GPIIb receptor+) platelet MPs, CD62e+ (E-selectin+) endothelial MPs, CD144+ (VE-cadherin+) endothelial MPs, CD31+ (PECAM+)/CD41– endothelial MPs, CD45+ (leukocyte common antigen+) leukocyte-derived MPs (LMPs), and whole apoptotic annexin V+ MPs. Circulating MP levels were expressed as events per µL (ev/µL).
Statistical Analysis
Baseline Characteristics of Study Population Demographics and clinical characteristics of patients with PH and control subjects are shown in Table 1. The control subjects were age- and sex-matched healthy subjects with no past medical history and they were not taking any medications. The etiologies, functional classification, and hemodynamic characteristics of the PH subjects are depicted in Table 2. Twenty-four subjects were enrolled (67% women; mean age, 54 ± 3 yr) with 63% of the patients having a diagnosis of pulmonary arterial hypertension (PAH). No significant difference was observed for baseline characteristics between the PH and control groups, except for body mass index (Table 1). The  was 41 ± 2.4 mm Hg, and was significantly higher in patients with PAH compared with subjects with chronic pulmonary disease–related PH (46.1 ± 2.9 vs. 30.6 ± 1.8 mm Hg; P < 0.01 by Student t test).
Increased Levels of Circulating MPs in Patients with PH Results of the MP measurements are presented in Figure 1. We observed a significantly higher number of circulating endothelial MPs in the PH group compared with the healthy control group, as assessed by the levels of CD62e+ (E-selectin+) (303 ± 59 vs. 75 ± 12 ev/µl; P = 0.001), CD144+ (VE-cadherin+) (932 ± 158 vs. 199 ± 33 ev/µl; P < 0.001), and CD31+ (PECAM+)/CD41– EMPs (2,357 ± 328 vs. 781 ± 132 ev/µl; P < 0.001). Moreover, the number of circulating leukocyte-derived CD45+ LMPs was also increased in the PH group (306 ± 49 vs. 139 ± 22 ev/µl; P = 0.05), but no significant difference was observed for platelet MPs (3,763 ± 1,013 vs. 3,163 ± 555 ev/µl; P = 0.59) and annexin V+ MPs (3,674 ± 909 vs. 2,991 ± 587 ev/µl; P = 0.98). We observed similar results when comparing the patients with PH with patients (n = 8) who underwent right heart catheterization but did not fulfill PH criteria (data not shown).
Among the patients with PH, we observed significantly higher values of PECAM+ and VE-cadherin+ EMPs in patients with PAH versus those with chronic pulmonary disease–related PH (respectively, 3,003 ± 432 vs. 1,302 ± 254 and 1,211 ± 221 vs. 458 ± 90 ev/µl; P < 0.01 for both), whereas no difference was observed for other MP subgroups (Figure 2).
We investigated the influence of coronary artery disease risk factors on our results: the patients with PH were classified in two subgroups, depending on the presence or absence of cardiovascular risk factors. Group 1 (n = 12; age, 55.6 ± 2.7 yr) included patients with PH ( = 38.1 ± 2.1 mm Hg) and diabetes and/or a history of dyslipidemia, systemic hypertension, or smoking, whereas group 2 (n = 12; age, 51.3 ± 5.5 yr) included patients with PH ( = 43.2 ± 4.2 mm Hg) without any of the other conditions. We did not observe any significant differences between the values of E-selectin+ (350 ± 86 vs. 256 ± 82 ev/µl; P = 0.435), PECAM+ (2,016 ± 362 vs. 2,698 ± 547 ev/µl; P = 0.31), and VE-cadherin+ (744 ± 143 vs. 1,120 ± 280 ev/µl; P = 0.25) EMPs; CD45+ MPs (318 ± 71 vs. 297 ± 70 ev/µl; P = 0.84); PMPs (3,964 ± 1,412 vs. 3,561 ± 1,514 ev/µl; P = 0.81); and annexin V+ MPs (4,014 ± 1,370 vs. 3,334 ± 1,239 ev/µl; P = 0.72) between these two groups. However, the levels of PECAM+, VE-cadherin+, and E-selectin+ EMPs (all P < 0.01) and CD45+ MPs (P < 0.05) were significantly higher in each of these two groups compared with the healthy control subjects. Similarly, we did not find any significant differences in the levels of EMPs in patients being treated with steroids or statins compared with subjects not taking these medications (data not shown). There was a strong correlation between the number of VE-cadherin+ and PECAM+ EMPs among all subjects with PH (r = 0.92, P < 0.0001). Moreover, E-selectin+ EMPs were positively correlated with the number of CD45+ MPs (r = 0.68, P = 0.009) but not with other EMPs, either PECAM+ EMPs (r = 0.10, P = 0.63) or VE-cadherin+ EMPs (r = 0.04, P = 0.98). E-selectin+ EMPs were also the only MP subgroup that positively correlated with values of hsCRP (r = 0.51, P = 0.035); this relation was not observed for the healthy subjects (r = –0.19, P = 0.433). No correlation was found in the PH group between any of the endothelial MP subgroups and age, creatinine, low-density lipoprotein cholesterol, or body mass index. Interestingly, levels of both VE-cadherin+ and PECAM+ EMPs correlated positively with the hemoglobin level in patients with PH (r = 0.53, P = 0.008 and r = 0.58, P = 0.003, respectively).
Variations of MP Values in Systemic versus Pulmonary Circulation
Circulating Endothelial MPs Are Independent Predictors of the Hemodynamic Severity of Untreated PH Among the subjects with PH, correlated significantly with the baseline values of hemoglobin (r = 0.55, P = 0.005) and brain natriuretic peptide (r = 0.692, P = 0.001). We found strong and significant correlations between VE-cadherin+ and PECAM+ venous circulating EMPs and several parameters used to evaluate severity of pulmonary hypertension: , pulmonary vascular resistance, cardiac index, and right atrial pressure (Table 4 and Figure 3). Notably, there was no significant relation between these hemodynamic parameters and E-selectin+ EMP levels (Table 4). Furthermore, no correlation was found with the other circulating venous MP subpopulations and hemodynamic severity of PH. Finally, PECAM+ EMP (r = 0.63, P = 0.005) and VE-cadherin+ EMP (r = 0.61, P = 0.007) levels were also correlated with brain natriuretic peptide values (Figure 4), whereas no relation was observed with the other MP subpopulations (data not shown). The correlations between and PECAM+ or VE-cadherin+ EMPs were also observed within the subgroup of patients with PH who were classified as having PAH (n = 15) as per the World Health Organization (WHO) classification and in the patients with chronic pulmonary disease–related PH (n = 8; Figure 5). The same correlations were observed within patients with (group 1) and without (group 2) coronary artery disease risk factors (data not shown). Because the hemoglobin level has been previously reported to be correlated with EMP values and to influence EMP release through mediation of shear stress, we have included this parameter in our analysis (12). Multivariate regression analysis demonstrated persistence of significant correlations between or pulmonary vascular resistance and PECAM+ or VE-cadherin+ EMPs after adjustment for age and hemoglobin levels in all the patients (Table 5). No association existed between circulating MPs and clinical parameters of functional capacity in patients with PH: specifically, no correlation was found between 6-minute walking distance and levels of endothelial, leukocyte-derived, platelet-derived, or annexin V+ MPs. Moreover, levels of circulating MPs (any subgroup) did not differ significantly among patients in WHO class 2, 3, or 4 (data not shown).
We demonstrate that in patients with PH there is an increase in the number of several populations of circulating EMPs and LMPs, compared with a control population. Moreover, our results show that two specific subgroups of EMPs, specifically CD144+ (VE-cadherin+) and CD31+ (PECAM+)/CD41– EMPs, were independent predictors of the hemodynamic severity of pulmonary hypertension, whereas CD62e+ (E-selectin+) EMPs were associated with hsCRP values among the patients with PH, suggesting different origins and properties of the different types of EMPs. Endothelial dysfunction is a pivotal element in the development and progression of PH (3, 4). However, the mechanisms involved in the development of this endothelial injury and its involvement in pulmonary arterial remodeling are complex and remain poorly understood (2, 3). Despite several reports of impaired endothelial function in various types of PH, such as changes in nitric oxide synthase expression (13), endothelin-1 clearance, or local prostacyclin synthesis (14), the assessment of endothelial structural damage in humans in vivo remains challenging. Bull and coworkers (15) have shown evidence of endothelial lesions in patients with various etiologies of PH by reporting that there is an increase in the number of circulating endothelial cells with an activated phenotype, a marker of vascular injury (16). Nevertheless, the precise origin of circulating endothelial cells remains ambiguous as some of these cells may originate from the diseased endothelium whereas others may originate from the bone marrow (15). Circulating EMPs have been shown to result from apoptosis or activation of endothelial cells (5, 6). The levels of CD144+ and CD31+/CD41– EMPs have been reported as an index of endothelial injury in patients with end-stage renal failure (7) or coronary artery disease (8), and correlated closely with the degree of systemic endothelial dysfunction in these patients. Thus, the presence of an increased number of EMPs may relate to structural damage of the endothelium in pulmonary hypertension.
Our work provides evidence of a strong relationship between endothelial destruction and hemodynamic severity in patients with PH in vivo. Using another marker (circulating endothelial cell numbers), Bull and coworkers (15) previously reported an association between endothelial damage and pulmonary artery pressure assessed by echocardiography. However, echocardiography has several limitations for The exact mechanism underlying the association of EMPs and PH that we have observed in this study is unknown. The role of endothelial dysfunction in the causation and disease progression of PH has yet to be identified. EMPs have been reported as both markers of vascular injury and also, on their own, can impair vascular function (5–7). Many factors ("sentinel events") can injure the endothelium and potentially contribute to the increased release of EMPs (inflammation, modification of lumenal pressure regimen and blood flow, direct drug toxicity [methamphetamine], HIV infection, release of proliferative cytokines, or autoimmunity) (4, 5). However, EMPs can also interact with the endothelium and inhibit NO synthesis (7, 9), acting as a paracrine factor to potentiate endothelial dysfunction (5). The generation of EMPs could thus impair vasodilatory properties of the vessels and may mediate progression of the disease, leading to the remodeling of pulmonary arteries.
Surprisingly, we did not find a correlation between CD62e+ EMPs and any of the hemodynamic parameters of pulmonary hypertension. We did show, however, that this specific EMP population was related to values of hsCRP and leukocyte-derived MPs, suggesting potential differences in origins and pathophysiological mechanisms underlying EMP release. Jimenez and coworkers (11) reported that various types of stimuli led to the release of EMPs with various phenotypes by endothelial cells in vitro. CD62e+ EMPs were more abundant after induction of activation by tumor necrosis factor-
Our study has several limitations. Because PH is an uncommon condition, the number of patients in the PH group was limited, and the patients had heterogeneous etiologies. Yet, the increase in MP number was concordant and correlations to hemodynamic severity were robust. Furthermore, we observed significant correlations between CD31+/CD41– and CD144+ EMPs and In summary, the present study demonstrates an increase in the numbers of EMPs and LMPs in patients with precapillary PH. We also show that some of the EMP subgroups (CD31+/CD41– and CD144+ EMPs) predict the hemodynamic severity of pulmonary hypertension, whereas other subgroups (CD62e+ EMPs) were associated with inflammation. We have shown a relationship between markers of endothelial injury and PH hemodynamic severity. Further studies are needed to focus on defined subgroups, disease stages, and clinical outcomes. Because some of the hemodynamic parameters have been related to outcomes in patients with some categories of PH, our results raise the possibility that EMP measurement could be used to assess prognosis or might also be used to monitor the benefit of endothelium-protective therapy (6).
Supported by an educational grant from the Wayne and Gladys Valley Foundation, the Hartford Foundation, the UCSF Cardiology Council, and the Castle Foundation (Y.Y. and W.G.). This article has an online supplement, which is available from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200710-1458OC on February 28, 2008 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form October 1, 2007; accepted in final form February 25, 2008
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
2 test for categorical variables. The Student t test, one-way analysis of variance, or the Mann-Whitney U test was used to compare quantitative variables between groups. Quantitative variables with nonnormal distribution were log transformed. Univariate correlations were assessed by Pearson's r test, depending on distribution normality. To estimate the predictive value of circulating MP levels in predicting hemodynamic severity of PH a multivariate linear regression analysis was performed. Differences were considered significant at P < 0.05. Statistical analysis was performed with SPSS 14.0 software for Windows (SPSS, Inc., Chicago, IL). 
, whereas CD31+ EMPs were produced predominantly during endothelial destruction (
