(Circulation. 1996;94:2228-2234.)
© 1996 American Heart Association, Inc.
Articles |
the Division of Cardiology and Department of Medicine, Cedars-Sinai Medical Center Research Institute and the University of California, Los Angeles, School of Medicine; and Comedicus Inc, Minneapolis, Minn (T.C.H.).
Correspondence to Sanjay Kaul, MD, Vascular Physiology Laboratory, Division of Cardiology, Cedars-Sinai Medical Center, 8700 Beverly Blvd, Los Angeles, CA 90048. E-mail kaul@csmc.edu.
| Abstract |
|---|
|
|
|---|
Methods and Results Effects of local infusion of DMHD/NO, intravenous aspirin, and heparin on ST were evaluated in an ex vivo porcine AV shunt model. Nitinol stents (2 mm in diameter, n=120) were placed in a tubular chamber and perfused with blood from pigs (n=13) at a shear rate of 2100 s-1 for 20 minutes. ST was quantified by measurement of dry thrombus weight (TW). Effects on platelet aggregation (PA), blood pressure, bleeding time, and activated clotting time (ACT) were also examined. There was a dose-dependent inhibition of ST and PA by DMHD/NO. TW was reduced by 95% (1±2 versus 16±4 mg control, mean±SD, P<.001), and PA was reduced by 75% (4±3 versus 14±9
/min control, P<.05) at the highest dose of 10 µmol/L. DMHD/NO had no effects on bleeding time, ACT, or blood pressure. In contrast, aspirin (10 mg/kg), despite inhibiting PA, had no effects on TW (12±5 versus 16±8 mg control, P=.3). Heparin (200 U/kg) reduced TW by 33% (14±4 versus 21±3 mg control, P<.05) and prolonged ACT.
Conclusions Local delivery of DMHD/NO produced a 15-fold inhibition of acute ST at high-shear flow without producing adverse systemic hemostatic or hemodynamic effects. Thus, treatment with DMHD/NO may be an effective strategy for prevention of stent thrombosis.
Key Words: stents platelets thrombosis endothelium-derived factors aspirin
| Introduction |
|---|
|
|
|---|
NO is an endogenous molecule that plays an important role in regulation of vasomotor tone,9 platelet interaction with the vessel wall,10 11 12 and smooth muscle cell replication13 after vascular injury. We recently showed that a short-lived (half-life of <1 minute), NO-releasing diazeniumdiolate type of NO donor, DMHD/NO, reduced acute platelet-dependent thrombus formation in injured arterial strips by ninefold in an extracorporeal rabbit perfusion model.14
There is little information about the effects of NO on platelet deposition on metallic bioprosthetic surfaces, and less is known about its effects on stent thrombosis. Our hypothesis is that local delivery of a rapidly acting NO donor, DMHD/NO, may effectively inhibit platelet-rich stent thrombosis. We therefore examined the effects of DMHD/NO on acute stent thrombosis, systemic hemodynamics, and hemostatic parameters in an ex vivo porcine AV shunt model under high-shear flow conditions. For comparison, the effects of conventional antiplatelet and antithrombotic agents, aspirin and heparin, were also evaluated.
| Methods |
|---|
|
|
|---|
The ex vivo extracorporeal perfusion system used in this study to examine acute stent thrombosis is shown in Fig 1
. The left common carotid artery and internal jugular vein were cannulated with 8F introducer sheaths to establish an extracorporeal circuit. A tubular perfusion chamber with an ID of 2.0 mm (kindly provided by Dr Juan J. Badimon, Mount Sinai Medical Center, New York City, NY)15 was placed between the arterial and venous circuits. The arterial cannula was connected to the inlet of the perfusion chamber, and the outflow was connected through a variable-speed peristaltic pump (model 7521-40, Masterflex, Cole-Palmer Instrument Co) to the venous cannula. The chamber and the tubing were immersed in a temperature-controlled water bath at 37°C. A transit-time Doppler flow probe (Transonic Systems Inc) was interposed in the circuit to continuously monitor the blood flow. A second peristaltic pump was used to infuse DMHD/NO upstream from the perfusion chamber (Fig 1
). Arterial blood gases were periodically monitored and normalized by adjustment of ventilation rate and volume. Invasive blood pressure, O2 saturation, ECG, and rectal temperature were continuously monitored. Body temperature was maintained at 37°C with a temperature-controlled blanket. Mean blood pressure was maintained at
90 mm Hg before drug administration by adjustment of the depth of anesthesia and volume expansion with intravenous 0.9% saline. At the conclusion of the experiment, the animals were euthanatized under anesthesia by a bolus dose of sodium pentobarbital (50 mg/kg IV).
|
Coronary Stents
The stents (n=120) tested were 7-mm-long slotted-tube ACT-One stents made from the nickel-titanium alloy nitinol, with a strut thickness of 0.007 in (Advanced Coronary Technology). Stents weighed an average of 16 mg and had a silicon carbide gritblasted surface finish, which creates a uniformly roughened surface that is highly thrombogenic in this model.16 Stents were preexpanded by a mandrel to an OD of 2.0 mm before being mounted in the perfusion chamber.
Drugs
Heparin sodium and sodium nitrite were purchased from Sigma Chemical Co. The NO donor DMHD/NO and the carrier DMHD were kindly provided by Comedicus Inc. Collagen was obtained from Chronolog Corp.
Stock solution of DMHD/NO was prepared by dissolving the solid material in ice-cold 0.1 mol/L NaOH to minimize drug decomposition. Before being administered, the drug was reconstituted in 0.1 mol/L PBS at pH 7.4. Sodium nitrite and DMHD were prepared by dissolution of these compounds in PBS just before administration. Lysine-buffered aspirin (Bayer AG) was dissolved in water just before use according to the instructions of the manufacturer.
Experimental Protocol
Blood was collected for baseline determination of platelet aggregation, leukocyte and platelet counts, hematocrit, BT, and ACT. All animals subsequently received low-dose anticoagulation with heparin (an initial bolus of 50 U/kg IV and repeated at 25 U/kg IV every hour to keep the ratio of treated to control activated partial thromboplastin time
1.5) to avoid clotting inside the tubing system.
After a 60-minute stabilization period, stents were mounted in the perfusion chamber and perfused with Krebs solution for 60 seconds at 37°C. Blood was then perfused through the perfusion chamber at 70 mL/min for 20 minutes. This flow rate was estimated to generate a local shear rate of about 1500 s-1 at the tube surface and about 2100 s-1 at the stent surface.17 At the end of the perfusion period, Krebs buffer was circulated through the chamber for 30 seconds at 40 mL/min to wash off unattached cells and blood from the stent and the perfusion system. The changes from buffer to blood and vice versa were achieved by a switch valve without the introduction of stasis in the chamber. The stents were removed from the chamber, dried, and weighed. The perfusion chamber and the catheters were then thoroughly cleaned with normal saline to clear any blood before another stent was mounted. Images of stents were obtained at x15 magnification with a videomicroscope (Sony CCD TR 51) and image analysis software (Bioscan, Optimas Corp).
Control stents exhibited significant amounts of thrombus formation, which was time dependent and reached a plateau at
15 to 20 minutes at the given flow conditions. On the basis of these data, a perfusion time of 20 minutes was chosen for subsequent perfusion studies. Perfusion experiments were performed at baseline and after treatment with drugs or vehicles in a randomized order. To avoid rapid inactivation of NO released by DMHD/NO by hemoglobin in blood, DMHD/NO (0.1, 1.0, and 10 µmol·L-1·min-1) was infused continuously at 1 mL/min for the entire perfusion period. Aspirin (5 and 10 mg/kg) and heparin (50, 100, and 200 U/kg) were administered systemically as intravenous bolus doses. Perfusion experiments were performed 30 minutes after intravenous bolus administration of aspirin to allow for peak antiplatelet effects of this drug. ACT was checked before perfusion studies with heparin to ensure optimum anticoagulant effect. Platelet aggregation studies were also done to confirm antiplatelet effects of aspirin before each perfusion. In a few selected animals (n=3), perfusion studies were repeated at frequent time intervals after the administration of drugs to determine whether the antithrombotic effects of these drugs were short- or long-lived. Based on the results of these experiments, a minimum 45- to 90-minute recovery period was allowed between administrations of drugs for subsequent perfusion studies. Time-control studies to ensure recovery of thrombus formation were performed before every perfusion study with each drug. The vehicle for the NO donor, the carrier nucleophile DMHD, and sodium nitrite (nitrite ion is the end product of NO oxidation in aqueous medium) were given to selected animals (n=4).
To determine whether DMHD/NO-mediated inhibition of thrombus formation was due to a dethrombotic or a lytic effect, the effects of DMHD/NO were examined on stents with preexisting thrombus in three experiments.
Blood samples were collected in selected animals at the beginning and at the end of each experiment to determine hematocrit and platelet and leukocyte counts.
Platelet Studies
Platelet Aggregation Assay
Blood (3 mL) was collected in siliconized test tubes containing 3.8% sodium citrate (Becton Dickinson Vacutainer System). Aggregation studies were performed within 20 minutes after collection of blood. A whole-blood impedance aggregometer (Chronolog Corp) was used to measure platelet aggregation induced by collagen (5 µg/mL). Aggregation was expressed as rate of increase in electrical impedance. To examine the effects of DMHD/NO on platelet aggregation, untreated blood samples were incubated with the drug in vitro for 1.0 minute before the addition of collagen. To examine the effects of aspirin and heparin on platelet aggregation, blood samples were collected 30 minutes after intravenous infusion of aspirin and heparin.
Platelet cGMP Assay
cGMP levels in platelets were measured in PRP. PRP was obtained by centrifugation of venous blood at 140g for 10 minutes. To examine the effects of DMHD/NO on platelet cGMP concentration, untreated PRP samples were incubated with the drug in vitro for 2 minutes. The platelet count in each plasma sample was determined, and 1.0-mL samples were then centrifuged at 3000g for 2 minutes to obtain a platelet pellet. One milliliter of 6% trichloroacetic acid was added to each platelet pellet and vortexed for 1 minute. Samples were then centrifuged at 10 000g for 15 minutes, and the aqueous phase was frozen in liquid nitrogen and stored at -70°C before assay. Each sample was thawed, trichloroacetic acid in the supernatant was extracted by four washes with water-saturated ether, and the sample was evaporated to dryness in a Speed-vac Concentrator (Favant). The sample was reconstituted in buffer and assayed by commercially available radioimmunoassay kits (Amersham). Results were expressed as pmol cGMP/109 platelets.
BT and ACT
BT was estimated by the following procedure. An incision was made on the posterior surface of the ear with a number 11 surgical knife. The ear was then placed in a beaker of isotonic saline at 37°C. The time between incision and cessation of bleeding was recorded as BT.18 ACT was measured with a Hemochron 400 machine (International Technidyne Corp) in standard fashion.19
Statistical Analysis
Data are presented as mean±SD. Multiple-group comparisons were made with one-way ANOVA based on the general linear model to account for differences in group sample sizes. If P<.05, then post hoc pairwise comparisons of the mean were performed with the Bonferroni inequality.
| Results |
|---|
|
|
|---|
|
The effects of treatment with DMHD/NO, aspirin, and heparin on stent thrombus weight are shown in Fig 3
. DMHD/NO produced a dose-dependent inhibition of thrombus weight. Thrombus weight was reduced by 15-fold compared with controls at the highest dose of 10 µmol/L (from 16±4 to 1±2 mg, P<.001). Neither DMHD, the vehicle for DMHD/NO, nor sodium nitrite, a product of NO metabolism, reduced stent thrombus weight. Aspirin (both 5 and 10 mg/kg IV doses) and conventional anticoagulant doses of heparin (50 to 100 U/kg IV) had no effects on thrombus formation on stents. At a higher dose (200 U/kg), however, heparin produced a small but statistically significant reduction of thrombus weight (from 21±3 to 14±4 mg, P<.05).
|
Treatment with DMHD/NO (10 µmol/L) did not reduce thrombus weight on stents with preexisting thrombi (16±5 versus 18±7 mg, P=.4). The antithrombotic effects of DMHD/NO and heparin were short-lived; thrombus formation on stents was restored to baseline within 30 to 45 minutes and 45 to 90 minutes after the administration of the NO donor and heparin, respectively.
Platelet Studies
Effects of DMHD/NO, aspirin, and heparin on platelet aggregation are shown in Fig 4
. There was a dose-dependent inhibition of platelet aggregation by DMHD/NO, with a 75% reduction compared with controls at the highest concentration of 10 µmol/L (from 14±9 to 4±3
/min, P<.05). There was no significant change in platelet aggregation with either DMHD or sodium nitrite. Although aspirin did not produce any significant effects on stent thrombosis, platelet aggregation was markedly inhibited at the lowest dose of 5 mg/kg IV. Heparin produced a small, statistically insignificant (P=.3) effect on platelet aggregation, with a 28% reduction at the highest dose.
|
Platelet cGMP levels increased from 2.6±0.1 pmol/109 platelets at baseline to 4.4±0.5 pmol/109 platelets in DMHD/NO-treated samples (P<.05, n=3).
Hematological Parameters
The effects of drugs on bleeding and ACT are summarized in the Table
. DMHD/NO had virtually no effect on either BT or ACT. Aspirin produced a slight but insignificant prolongation of the BT and had no effects on ACT. Heparin produced a slight but insignificant increase in BT but significantly prolonged ACT at higher doses of 100 and 200 U/kg. There were no episodes of overt bleeding in any of the animals studied.
|
A small, nonsignificant decline in platelet count (312±87x103 to 273±145x103/µL, P=.3) was observed at the end compared with the beginning of the experimental protocol. Leukocyte count (18±7x103 to 15±9x103/µL) and hematocrit (27±4% to 30±7%) (n=5 each) were essentially unchanged.
Hemodynamic Parameters
DMHD/NO did not produce any significant effects on arterial blood pressure and heart rate, even at doses that produced reduction in thrombus formation on stents. Mean arterial blood pressure and heart rate were 89±16 mm Hg and 118±9 bpm, respectively, before and 85±15 mm Hg and 120±7 bpm after infusion of DMHD/NO (10 µmol/L). Aspirin and heparin had no significant effects on arterial blood pressure (data not shown).
The control substances DMHD and sodium nitrite had no significant effects on either blood pressure, BT, or ACT (data not shown).
| Discussion |
|---|
|
|
|---|
We have taken an ex vivo system that was developed for studying platelet responses to vascular injury or to thrombogenic bioprosthetic surfaces15 17 and successfully adapted it to characterize the events resulting in acute stent thrombosis. This model closely simulates in vivo arterial thrombus formation and has been widely accepted for studies in experimental animals15 17 and in humans20 for the evaluation of thrombotic mechanisms and therapeutic strategies. Although the clinical relevance of stent thrombosis in this experimental model has not been defined, the reproducibility and simplicity of this ex vivo system make it attractive to study the interaction of blood elements with stent surface and to make inferences about the efficacy of therapeutic interventions.
Stent Thrombosis
Studies in patients21 and animal models22 23 show platelet-rich thrombi to be the primary component of acute stent occlusion. Platelet-dependent stent thrombosis is probably related to three factors: (1) the topographic characteristics, stent geometry, and chemical properties of the stent surface16 ; (2) shear stress generated at the site of stenting16 17 or other coexisting stenoses in the stented vessel; and (3) deep vessel wall injury.23
In the present model, thrombosis is mediated by the interaction of blood elements under high-shear flow with the stent surface. This model excludes the contribution of vessel wall injury. The robust thrombus formation seen in our experimental model may be due to the purposefully thrombogenic rough surface of the prototype nitinol stents and the high-shear flow conditions that were used in our studies. The shear rate of 2100 s-1 was chosen to simulate conditions of inadequate stent deployment or stent placement in small vessels. Small target reference vessel diameter (<3.0 mm) is a risk factor for coronary stent thrombosis.1 2 Fewer than 50% of coronary stenoses can currently benefit from stenting because a majority of obstructive lesions are confined to segments of coronary arteries <3.0 mm in diameter, in which the incidence of stent thrombosis is increased.
Effects of NO Donor
Several recent studies suggest that NO plays an important role in modulating plateletvessel wall interaction after vascular injury.9 10 11 12 This study may be the first to demonstrate the potential of NO to reduce the thrombogenicity of endovascular metal stents. Our previous experience with DMHD/NO revealed it to be a potent and short-lived (half-life of <1 minute) NO donor. We documented a marked reduction in acute platelet-dependent thrombus formation on injured arterial strips when DMHD/NO was delivered locally in a rabbit extracorporeal perfusion model.14
DMHD/NO is one of several new diazeniumdiolate NO donor compounds that are collectively called NONOates.24 25 26 These agents nonenzymatically release NO with predictable first-order kinetics, and their biological half-lives range from 2 seconds to 2 weeks.24 25 26 Several NONOate compounds have shown potent vasorelaxant,24 antiplatelet,14 25 and antimitotic26 effects in vitro and in animal models.
Local delivery of DMHD/NO may offer the advantage of dissociating the desired regional antiplatelet effects from systemic effects on hemodynamics and hemostasis. Local delivery circumvents the rapid inactivation of NO by hemoglobin before high concentration is achieved at the stented site. Rapid inactivation of NO may also explain the lack of hypotensive and hemostatic effects seen in our study.
Mechanism of Action of DMHD/NO
The marked impairment of platelet aggregation by DMHD/NO suggests one possible mechanism for reducing thrombus formation on stents. Although aspirin produced a similar inhibition of platelet aggregation, it was not effective in preventing stent thrombosis. This finding suggests that the inhibitory effects of DMHD/NO on stent thrombosis may be mediated by reduction in platelet adhesion rather than inhibition of platelet aggregation. An alternative explanation is that NO effectively inhibits platelet aggregation under low- as well as high-shear conditions, whereas aspirin may be effective only under static or low-shear conditions (as encountered in an aggregometer). Another potential mechanism for the inhibitory effects of DMHD/NO on stent thrombosis may be activation of plasma protein(s), which rapidly coat and "passivate" the stent surface, thereby rendering it thromboresistant. We do not know of any evidence to support this. A significant thrombolytic effect is unlikely because there was no reduction in preexisting platelet thrombi on the stent surfaces. Data with the vehicle carrier and sodium nitrite (a metabolite of NO oxidation) suggest that thrombus formation was specifically inhibited by the generation of NO. The molecular basis for these effects is the activation of platelet-soluble guanylyl cyclase, resulting in the increased formation of cGMP.12
The swine model limits the evaluation of precise mechanisms of DMHD/NO action. There are currently no monoclonal antibodies for measuring platelet P-selectin (an activation-dependent surface marker) expression in this species. We have recently shown, however, that DMHD/NO markedly inhibits P-selectin expression on stimulated human platelets.27 Thus, DMHD/NO most likely acts at the step of platelet activation.
Effects of Aspirin and Heparin
Although platelet aggregation was markedly reduced by conventional doses of aspirin, it was ineffective in inhibiting thrombus formation on stents. We recently observed that aspirin in doses of >20 mg/kg IV may reduce stent thrombosis in this model. These findings are consistent with observations of Michelson et al28 that high-dose aspirin is required to inhibit high shearinduced platelet adhesion despite marked inhibition of platelet aggregation at lower doses.
Systemic administration of heparin was not effective in reducing stent thrombosis. Potential explanations include (1) the release of heparin-neutralizing proteins such as platelet factor 4 and ß-thromboglobulin,29 30 (2) steric or ionic exclusion of heparin from the thrombin formed within the thrombus,31 (3) the inability of heparin to inactivate meizothrombin formed on the platelets,32 and (4) the fact that heparin contains fractions that activate platelets directly, thereby contributing to the development of arterial thrombosis. We recently showed that heparin in concentrations of <1.0 U/mL (within therapeutic range) augments platelet aggregation and P-selectin expression in humans.33 In contrast, very high systemic doses of heparin (>200 U/kg, as used in this study) have been shown to inhibit platelet-dependent thrombus formation.29 34
Study Limitations
There are several limitations of this ex vivo model that preclude conclusions applicable to clinical practice. The shear rates and thrombus burden generated in this model are much greater than expected in larger vessels (>3.0 mm), which are most often stented. Failure of aspirin and heparin to inhibit stent thrombosis may be related, in part, to this high shear rate. The experimental conditions were chosen to simulate conditions of inadequate stent deployment or stent placement in small vessels in which the higher shear rates encountered are likely to contribute to increased thrombosis. However, increased thrombosis in underdeployed stents is also probably related to other physical and physiological factors, including higher metal density, low lesion compliance, and possibly greater mural injury. The extracorporeal perfusion model used in this study precludes the evaluation of the effects of vessel wall injury on stent thrombosis and the evaluation of vasomotor influences of NO donors at sites of stent deployment, which would have a significant effect on local shear stress and blood flow and therefore impact platelet adhesion and thrombus formation.
These additional effects of NO donors may have an important effect in their overall clinical utility. The link between NO and its observed inhibitory effects on stent thrombosis is indirect, because we did not measure NO concentrations directly at the target area (stent site) in this study. Furthermore, this experimental model allows us to study events resulting in acute but not subacute stent thrombosis.
Study Implications
We have shown that local administration of DMHD/NO, a new type of NO donor, is highly effective in inhibiting acute platelet-dependent stent thrombosis under high-shear flow conditions in an ex vivo porcine AV shunt model compared with aspirin and heparin. Our findings suggest that the use of stents in smaller-diameter vessels might be rendered feasible if thrombosis under high-shear flow conditions could be effectively inhibited. Local delivery of NO, either from DMHD/NO-coated stents or from local infusion of the NO donor, may offer such an avenue and potentially affect the problem of stent thrombosis. The long-term safety of these compounds in humans, however, remains to be established.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
| Footnotes |
|---|
Received January 12, 1996; revision received May 1, 1996; accepted May 20, 1996.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J.-S. Wang, Y.-S. Li, J.-C. Chen, and Y.-W. Chen Effects of Exercise Training and Deconditioning on Platelet Aggregation Induced by Alternating Shear Stress in Men Arterioscler. Thromb. Vasc. Biol., February 1, 2005; 25(2): 454 - 460. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Rukshin, B. Azarbal, P. K. Shah, V. T. Tsang, M. Shechter, A. Finkelstein, B. Cercek, and S. Kaul Intravenous Magnesium in Experimental Stent Thrombosis in Swine Arterioscler. Thromb. Vasc. Biol., September 1, 2001; 21(9): 1544 - 1549. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. L. Bouchie, H. Hansen, and E. P. Feener Natriuretic Factors and Nitric Oxide Suppress Plasminogen Activator Inhibitor-1 Expression in Vascular Smooth Muscle Cells : Role of cGMP in the Regulation of the Plasminogen System Arterioscler. Thromb. Vasc. Biol., November 1, 1998; 18(11): 1771 - 1779. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Gilutz Extracorporeal Circulation Circulation, January 13, 1998; 97(1): 117 - 117. [Full Text] |
||||
![]() |
S. Kaul, T. Z. Naqvi, M. C. Fishbein, B. Cercek, J. J. Badimon, T. C. Hutsell, S. Thomas, M. Molloy, and P. K. Shah Local Delivery of an Ultra-short-acting Nitric Oxide- releasing Compound, DMHD/NO, Is Highly Effective in Inhibiting Acute Platelet-Thrombus Formation on Injured Arterial Strips Journal of Cardiovascular Pharmacology and Therapeutics, January 1, 1997; 2(3): 181 - 193. [Abstract] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1996 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |