| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Thrombosis |
From the Carolina Cardiovascular Biology Center (A.C.K., J.C.Y., S.S.S.) and the School of Public Health (M.J.H.), The University of North Carolina, Chapel Hill; the Department of Medicine, Division of Cardiology (E.M.O.), Duke University Medical School, Durham, North Carolina; and the Division of Cardiovascular Medicine (G.L., Z.P., S.R.S., S.S.S.), The Gill Heart Institute, The University of Kentucky, Lexington.
Correspondence to Susan Smyth, MD, PhD, Division of Cardiovascular Medicine, The Gill Heart Institute, 900 S. Limestone Street, 326 Charles T Wethington Building, Lexington, KY 40536-0200. E-mail SusanSmyth{at}uky.edu
| Abstract |
|---|
|
|
|---|
Methods and Results Patients undergoing nonurgent PCI of a native coronary artery were randomized to receive adjunctive therapy with bivalirudin or UFH+eptifibatide. Interleukin (IL)-6 and C-reactive protein (CRP) transiently increased in both groups after PCI. In the UFH+eptifibatide, but not the bivalirudin group, myeloperoxidase (MPO) levels were elevated 2.3-fold above baseline (P=0.004) immediately after PCI. In an in vitro assay, heparin and to a lesser extent enoxaparin, but not bivalirudin or eptifibatide, stimulated MPO release from and binding to neutrophils and neutrophil activation. A mouse model of endoluminal femoral artery denudation was used to investigate further the importance of MPO in the context of arterial injury.
Conclusions Adjuvant therapy during PCI may have undesired effects on neutrophil activation, MPO release, and systemic inflammation.
Unfractionated heparin+eptifibatide was associated with higher myeloperoxidase levels after percutaneous coronary intervention. Heparin, but not bivalirudin, fondaparinux, or eptifibatide, increased MPO release from and binding to isolated neutrophils. These results indicate that adjuvant therapy during PCI may have undesired effects on neutrophil activation, MPO release, and systemic inflammation.
Key Words: platelets neutrophils myeloperoxidase percutaneous coronary intervention adjunctive therapy
| Introduction |
|---|
|
|
|---|
Given their distinct modes of action, it is not readily apparent why bivalirudin would provide similar efficacy as compared with heparin plus GPIIb/IIIa antagonists. GPIIb/IIIa antagonists are potent inhibitors of platelet aggregation.5 However, GPIIb/IIIa antagonists do not block platelet activation and may theoretically act as partial platelet agonists.6 In addition, heparin can also activate platelets7,8 and may have deleterious effects on other vascular cells. Bivalirudin, in contrast, does not directly prevent platelet aggregation, but through direct thrombin inhibition,9 may reduce platelet and vascular cell activation and thereby lower local and systemic inflammatory responses.
To gain insight into the efficacy of bivalirudin in comparison to GPIIb/IIIa antagonists plus UFH, we performed serial analysis of markers of platelet activation and systemic inflammation in patients undergoing PCI for stable angina randomized to receive either bivalirudin or UFH+ eptifibatide. The results indicated that PCI is associated with an acute inflammatory reaction regardless of adjunct therapy and that immediately after the procedure levels of myeloperoxidase (MPO), a hemoprotein released by activated neutrophils and to a lesser extent monocytes, were elevated in the UFH+eptifibatide group. Because elevated levels of MPO have been associated with poor outcomes in patients presenting with chest pain and acute coronary syndromes, we investigated the effects of adjuvant therapy on MPO release by isolated neutrophils in the absence and presence of activated platelets. In addition, we examined the role of MPO in the response to arterial injury in a mouse model. Our results suggest that heparin may have undesired effects during PCI by enhancing neutrophil function and MPO release.
| Methods |
|---|
|
|
|---|
Study Protocol
A total of 24 patients undergoing nonurgent PCI were randomized to receive bivalirudin alone or UFH+the GPIIb/IIIa antagonist eptifibatide. All patients received oral clopidogrel (300 mg) and were on aspirin before the intervention. Bivalirudin was administered as a bolus (0.75 mg/kg i.v.) before the procedure and then infused at 1.75 mg/kg/h for the duration of the procedure. Patients randomized to UFH+eptifibatide received a bolus of heparin (50 to 70 U/kg) before the procedure and 2 boluses of eptifibatide (180 µg/kg i.v.) 10 minutes apart before the procedure followed by an infusion of 3 µg/kg/min for 18 hours. Blood was collected before the intervention, and then immediately, 18 hours, 72 to 96 hours, and 7 days after PCI.
In Vitro Assays
Methodological details of the in vitro assays can be found in an online supplemental materials (available online at http://atvb. ahajournals.org)
Statistics
For the clinical study, inflammatory and platelet markers were compared over time to baseline levels by Wilcoxon rank sum test with continuity correction using a 2-sided test. Differences between the 2 groups at a given time point were analyzed by t test. Spearman analysis was used to assess the correlation of parameters over time. SAS software version 8 (SAS Institute) was used for analysis. Results from in vitro assays are presented as mean±SEM and are derived from at least 3 separate experiments run in duplicate or triplicate and were analyzed by t test or ANOVA, where appropriate. Significance was defined as P=0.05.
| Results |
|---|
|
|
|---|
|
|
Markers of Platelet Activation
There were no differences in either P-selectin exposure or binding of the GPIIb/IIIa activation-dependent antibody PAC-1, as measured by mean fluorescent intensity and percentage of positive platelets, in unstimulated blood before or after PCI in either treatment group (supplemental Figure IA and IB). In the 7 days after PCI, there was a steady reduction in P-selectin exposure in TRAP-treated blood in both treatment groups, which may reflect the fact that all patients were started on daily clopidogrel at the time of their intervention (supplemental Figure IC). As has been previously reported,10 immediately after PCI there were lower sCD40L levels in patients randomized to UFH+eptifibatide (P=0.024; supplemental Figure ID). A correlation between sCD40L levels and binding of PAC-1, a monoclonal antibody which recognizes the activated form of GPIIb/IIIa (integrin
IIbß3), was observed immediately after PCI (r=0.597; P=0.007). By 18 hours, sCD40L returned to baseline in both groups. No difference in plasma RANTES levels, a chemokine released by activated platelets, was observed between the treatment groups (data not shown). Likewise, no differences in circulating platelet-leukocyte aggregates in whole blood were observed after PCI in the 2 groups (data not shown).
Inflammatory Markers
PCI was associated with an acute inflammatory response, as judged by approximately 6-fold and 2-fold elevations in IL-6 and CRP levels, respectively, at 18 hours after the procedure in both treatment groups (P=0.05 compared with baseline; Figure 1A and 1B). Circulating levels of tissue factor and soluble intercellular adhesion molecule-1 (ICAM-1), a cell adhesive protein shed from activated endothelium and leukocytes, were similar in both treatment groups (data not shown). Surprisingly, immediately after the procedure, myeloperoxidase (MPO) levels were elevated 2.3-fold above baseline in the UFH+eptifibatide (7.5±0.7 versus 3.2±0.6 ng/mL; P=0.002; Figure 1C), but not in the bivalirudin group (2.9±0.6 versus 2.6±0.8 ng/mL). There was an inverse correlation between MPO levels and sCD40L immediately after PCI (r=0.452; P=0.0047).
|
Effect of Adjunctive Therapy on MPO Release From Neutrophils
To understand the potential effects of adjunctive therapy on white blood cell function, we sought to determine whether these drugs influenced MPO release from isolated neutrophils. When coincubated with isolated neutrophils, neither bivalirdin nor eptifibatide altered baseline MPO release in cell supernatant as determine by ELISA (Figure 2A). However, UFH dose-dependently increased MPO in the neutrophil supernatant, with a maximum effect observed at 2.5 U/mL (Figure 2A). Approximately 50% of endogenous MPO added to plasma was recovered in the assay (49.6±0.03; n=3). The detection of exogenously added MPO was slightly but consistently reduced in the presence of UFH (data not shown), suggesting that we may be underestimating the effects of heparin on MPO levels in both the in vitro assays and in the clinical samples. The addition of eptifibatide to heparin did not significantly alter MPO levels (Figure 2A). Enoxaparin, a low molecular weight heparin, stimulated release of MPO from isolated neutrophils (Figure 2B), although the maximal levels were lower than those observed with UFH. In contrast, fondaparinux, a pentasaccharide that selectively inhibits factor Xa, did not increase MPO levels when incubated with neutrophils at therapeutic plasma concentrations (Figure 2B).
|
A recent report indicates that MPO may stimulate an autocrine feedback loop in neutrophils via binding to integrin
Mß2 (CD11b/CD18).11 Therefore, we investigated the ability of heparin to activate this autocrine feedback loop, and observed that in the presence of heparin, neutrophils acquired the ability to bind MPO. Less than <10% of the neutrophils bound MPO in the presence of vehicle, but 45.74±12% of the cells bound MPO in the presence of heparin (mean±SEM; n=3). In comparison, phorbol ester (PMA) treatment stimulated MPO binding to 97±1.4% of the cells. (Figure 2C).
The effects of MPO on neutrophil function were investigated by measuring markers of degranulation and respiratory burst. Incubation of neutrophils with either heparin (10U/mL) or MPO (1 µg/mL) increased matrix metalloproteinase (MMP)-9 release to 126±9% (P=0.014) and 130±22% (P=0.084) of vehicle-treated, but neither increased elastase release. In the same assay, PMA increased neutrophil MMP-9 release to 282±46% and elastase to 134% of control. Respiratory burst was monitored by labeling neutrophils with DHR123, a fluorescent probe that increases fluorescence on oxidation by hydrogen peroxide (H2O2) generated as a result of the respiratory burst. DHR123 fluorescence doubled in the presence of MPO (MFI 644±235 versus 321±70; P<0.001), indicating that MPO binding to neutrophils elicited respiratory burst and generation of H2O2 (Figure 2D).
To ensure that we were not missing a platelet-mediated effect of eptifibatide on neutrophil MPO release, neutrophils and platelets were coincubated in the presence of shear. Interestingly, the addition of activated, but not resting, platelets to isolated neutrophils reduced the levels of MPO in the supernatant by 20% (P=0.031; supplemental Figure IIA). Pretreatment of platelets with either eptifibatide or antibodies to CD40L restored levels of MPO in the supernatant (supplemental Figure IIB). Bivalirudin had no effect on MPO release from neutrophils in the presence of platelets.
MPO and Arterial Injury in Mice
Elevated levels of MPO have been associated with adverse outcomes in patients presenting with chest pain,12,13 but have not been extensively studied in the context of PCI. To study the heparinMPO nexus in the context of arterial injury, we used a well-characterized model in which an angioplasty guide wire is used to denude the endothelium and trigger endoluminal arterial injury.1416 We have previously reported that within hours neutrophils are recruited to the site of injury by adherent platelets in this model. Immunohistochemical staining of the vessels at 6 hours after injury confirmed the presence of MPO (Figure 3A). Chlorotyrosine protein adduct formation, produced by reaction with MPO-generated hypochlorous acid, was also detected along injured vessels (Figure 3A). Pretreatment of mice with heparin (200 U, i.p.) 30 minutes before surgery increased MPO levels in injured vessels (Figure 3B). Fondaparinux (200 µg, i.p.) did not affect MPO levels, consistent with observations in the in vitro assay. Finally, to determine whether MPO contributes to the development of intimal hyperplasia in this model, injury was performed in wild-type mice and mice deficient in MPO (MPO/). At 28 days after femoral wire injury, neointimal area was 37% smaller in MPO/ mice as compared with wild-type control group (8600±2,200 µm2 versus 5400±1,500 µm2; n=10 arteries per genotype; Figure 3C), but the difference was not statistically significant (P=0.245). Likewise, no statistically significant differences were observed in lumen or media areas or intima/media ratios (lumen 20 500±6500 µm2 versus 27 900±8500, P=0.498; media 12 400±2000 µm2 versus 13 200±3000, P=0.827).
|
| Discussion |
|---|
|
|
|---|
Our studies with isolated neutrophils indicate that heparin and enoxaparin, but not bivalirudin, fondaparinux, or eptifibate, stimulate MPO release. Thus, the use of heparin as adjunct therapy during PCI may account for higher levels of circulating MPO. Additionally, we observed that heparin promotes the binding of endogenously-released MPO to neutrophils. Recently, MPO binding to integrin
Mß2 (CD11b/CD18) was proposed to function in an autocrine feedback loop that activates neutrophils independently of catalytic activity of the enzyme.11 We observed that both heparin and MPO stimulated MMP-9 release and respiratory burst in neutrophils, suggesting that that heparin may modulate the activation state of neutrophils via the cytokine-like effects of MPO.
The fact that the addition of activated platelets decreased MPO levels suggests that MPO may bind to platelet-stimulated neutrophils, platelet-neutrophil aggregates, or activated but not resting platelets. Pretreatment with either antibodies to CD40L or eptifibatide blunted this effect, suggesting that the effect may be mediated by surface expression of CD40L on activated platelets and/or the activated conformation of GPIIb/IIIa.
The significance of transient elevations of plasma MPO levels after PCI is unclear, but there is reason to believe that MPO may be a mediator of cardiovascular disease. For example, MPO has recently been shown to contribute to adverse left ventricular remodeling after myocardial infarction in an animal model.20 A major unfavorable outcome of stent implantation is the development of intimal hyperplasia and clinical restenosis. In a recent report, brief treatment of rat carotid arteries with MPO and H202 elicited intimal hyperplasia through the production of hypochlorous acid.21 Our results indicate that both MPO and chlorotyrosine-adduct formation accumulate along arteries after endoluminal injury in a mouse model. Heparin increased levels of vessel-associated MPO. However, we failed to observe a statistically significant difference in intimal area in mice deficient in MPO. Although these results may reflect the lack of a role for MPO in the injury response process, they need to be interpreted cautiously, as our studies may have been underpowered to detect a small difference in vessel areas in normal and MPO-deficient animals. Additionally, murine neutrophils contain only a fraction of the amount of MPO found in their human counterparts.22 In a mouse model of atherosclerosis, lack of MPO did not attenuate lesion development23; yet, transgenic mice expressing human MPO in macrophages have exaggerated atherosclerosis.24 Thus, differences in vascular cells and the injury response in mice may limit the extrapolation of our observations to humans.
We did not observe differences in platelet activation, as measured by platelet P-selectin expression or monoclonal antibody PAC-1 binding, in samples of peripherally-collected whole blood up to 1 week after PCI in patients who received UFH+eptifibatide as compared bivalirudin. These results are similar to those reported by Keating et al,18 although in their study, UFH+eptifibatide therapy was associated with elevated platelet surface expression of P-selectin and platelet-leukocyte aggregates in blood sampled from the coronary ostium before PCI. It is possible that the use of clopidogrel at the time of intervention blunts any effect of adjunctive therapy on markers of platelet activation, and it is likely that the gradual decline in agonist-induced activation of platelets that we observed over the 7 days after PCI was related to the administration of clopidogrel. In patients receiving UFH+eptifibatide, levels of soluble CD40L were lower immediately after the procedure, which would be consistent with the observation that eptifibatide blocks CD40L liberation from platelets.10,25
As has been shown by others,26,27 our results indicate that PCI is associated with an acute inflammatory response in patients with stable coronary artery disease, as manifest by a rise in IL-6 and C-reactive protein (CRP) levels in the first 24 hours after the procedure. In other studies, use of the GPIIb/IIIa antagonists abciximab and eptifibatide has been associated with a postprocedural reduction in IL-6 and CRP values as compared with placebo,28,29 and a direct comparison of the two agents indicated an equivalent effect on postprocedure inflammatory markers.30 Our results in a small population of stable patients suggests that there may not be a substantial difference in the change in IL-6 and CRP levels postprocedure in patients receiving bivalirudin or UFH+eptifibatide as adjunctive antithrombotic therapy.
In conclusion, our results suggest that the use of bivalirudin or UFH+eptifibatide in patients with stable coronary artery disease undergoing PCI is associated with similar effects on general inflammatory and platelet markers. Use of UFH+eptifibatide may be associated with transient elevations in MPO levels, perhaps as a direct consequence of actions of heparin on white blood cells. Larger clinical studies are required to understand the prognostic significance of transient elevations in MPO levels in this setting.
| Acknowledgments |
|---|
Sources of Funding
This work was supported by a research grant from the Medicines Company (S.R.S.), by NIH grants K08HL70304, R01HL07421, and P01HL080166 (S.S.S.), by NIH grants to the University of North Carolina P30DK34987 (G.I. histology core facility) and RR00046 (General Clinical Research Center), and by a Scientist Development Grant from the American Heart Association (G.L.).
Disclosures
E.M.O. has stockownership in Medtronic, Response Biomedical, and Savacor and is a consultant for Invoise, Liposcience, Response Biomedical, and Savacor. S.R.S. has received honoraria for serving as an advisor or consultant to the Medicines Company, Sanofi Aventis, AstraZeneca, Lilly, and Daiichi Sankyo. S.S.S. is the recipient of an Atorvastatin Research Award from Pfizer.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J. Shih, S. A. Datwyler, S. C. Hsu, M. S. Matias, D. P. Pacenti, C. Lueders, C. Mueller, O. Danne, and M. Mockel Effect of Collection Tube Type and Preanalytical Handling on Myeloperoxidase Concentrations Clin. Chem., June 1, 2008; 54(6): 1076 - 1079. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Li, J. M. Sanders, M. H. Bevard, Z. Sun, J. W. Chumley, E. V. Galkina, K. Ley, and I. J. Sarembock CD40 Ligand Promotes Mac-1 Expression, Leukocyte Recruitment, and Neointima Formation after Vascular Injury Am. J. Pathol., April 1, 2008; 172(4): 1141 - 1152. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |