Author + information
- Received September 14, 2009
- Revision received December 23, 2009
- Accepted January 6, 2010
- Published online May 11, 2010.
- Smriti Saraf, MB BS, MPhil†,
- Christos Christopoulos, LRCP&SI, MB, BCh, BAO†,
- Imen Ben Salha, MB BS†,
- David J. Stott, DPhil, MSc‡ and
- Diana A. Gorog, MD, PhD*,†,* ()
- ↵*Reprint requests and correspondence:
Dr. Diana A. Gorog, Imperial College, South Kensington Campus, SW7 2AZ London, United Kingdom
Objectives Our objective was to assess endogenous thrombolytic activity in acute coronary syndrome (ACS) patients and relate this to their likelihood of future adverse cardiovascular events.
Background Spontaneous lysis of platelet-rich thrombi is an important defense mechanism against lasting occlusion. Despite convincing evidence for the role of endogenous fibrinolysis in ACS, the prognostic value of plasma fibrinolytic markers in assessing risk is limited. We employed a novel global test which, in addition to platelet reactivity, allows assessment of endogenous thrombolytic activity to identify ACS patients who remain at risk of cardiovascular events.
Methods We used the global thrombosis test (GTT) to assess thrombotic and thrombolytic status in 300 ACS patients receiving dual-antiplatelet therapy. The test assesses the time required to form an occlusive thrombus, the occlusion time (OT), and the time to lyse this, the lysis time (LT). The end point of the study at 12 months' follow-up was the composite of death, nonfatal myocardial infarction, or stroke.
Results The OT and LT were both prolonged in ACS patients compared with normal volunteers (p < 0.001). LT ≥3,000 s occurred in 23% of ACS patients versus none of the normal subjects and was a significant and independent predictor of cardiovascular death and nonfatal myocardial infarction in a multivariate model adjusted for cardiovascular risk factors. LT ≥3,000 s was the optimal cutoff value for predicting 12-month major adverse cardiovascular events (hazard ratio [HR]: 2.52, 95% confidence interval: 1.34 to 4.71, p = 0.004) and cardiovascular death (HR: 4.2, 95% confidence interval: 1.13 to 15.62, p = 0.033). HR increased further as LT increased. No association was found between OT and the risk of major adverse cardiovascular events.
Conclusions Assessment of endogenous thrombolytic status based on the lysis of platelet-rich thrombi from native blood using the point-of-care GTT can identify ACS patients at risk of future cardiac events.
The clinical manifestation of an acute thrombotic event is determined by the balance between the propensity for thrombus formation and the efficacy of the endogenous thrombolytic processes.
Thrombotic events continue to occur despite treatment of high-risk patients with dual-antiplatelet medication. Nonresponsiveness to antiplatelet drugs remains a major limitation in the prevention of future thrombotic episodes in patients who experience acute coronary syndrome (ACS). Previous studies estimated that 5.5% to 56.8% of the population are aspirin resistant (1), and the prevalence of clopidogrel nonresponsiveness was 21% (2). Although fewer published data are available for dual (aspirin and clopidogrel) nonresponsiveness, the incidence in a recent study involving 746 patients was only 6% (3). Despite several prospective clinical trials demonstrating a statistically significant relationship between nonresponsiveness to antiplatelet medication and subsequent thrombotic events, still greater differences exist between the predictive power of the different platelet function tests used. All platelet function tests currently in clinical use measure the response of platelets to only 1 specific agonist (4), despite the involvement of many other physiologically important agonists such as high shear stress and thrombin in thrombogenesis. However, the major limitation of platelet function tests is that they are performed on citrate-anticoagulated blood, which prevents the assessment of thrombin generation by activated platelets, a major determinant of occlusive platelet-rich thrombus formation (4). Further, only stable, fibrin-stabilized platelet aggregates (thrombi) can cause lasting occlusion of arterial flow.
To assess the overall thrombotic status, in addition to platelet reactivity, the endogenous thrombolytic activity should also be measured. However, no global test of endogenous thrombolytic activity has been available for routine clinical use. The ability of blood to induce lysis of a platelet-rich thrombus is an important defense mechanism against lasting occlusion. Acute myocardial infarction (MI) has been regarded a result of the failure of timely spontaneous thrombolysis (5). Although many individual components of the fibrinolytic system can be measured by various laboratory techniques, the overall fibrinolytic status is difficult to ascertain from the plasma level of 1 or even several fibrinolysis activity markers. Thus, the conventional point-of-care assays measuring only the response of platelets to selected agonists, without taking into consideration the activation of the coagulation system (thrombin generation) and the endogenous fibrinolytic/thrombolytic activity, cannot give a realistic assessment of an individual patient's thrombotic status.
The global thrombosis test (GTT) (Montrose Diagnostics, London, United Kingdom), is a novel comprehensive test of platelet reactivity, coagulation (thrombin generation), and spontaneous (endogenous) thrombolytic activity. As this test is performed from native (nonanticoagulated) blood, it is genuinely different, and free from many of the shortcomings of conventional platelet tests, which employ citrate-anticoagulated blood (4,6,7). The aim of the present study was to investigate whether this novel test of overall thrombotic status could identify patients with ACS who, despite dual-antiplatelet medication, remain at risk of future cardiovascular events.
Three hundred patients hospitalized with ACS were sampled during their index admission. The ACS was defined by the presence of at least 2 of the following: ischemic chest pain, elevation of cardiac enzymes, and dynamic electrocardiographic changes compatible with ischemia at rest. All subjects had been established on dual-antiplatelet therapy, having received a loading dose of aspirin 300 mg and clopidogrel 300 mg on index admission, followed by a daily dose of 75 mg aspirin and 75 mg clopidogrel. The exclusion criteria listed in Table 1were applied. The clinical characteristics of our study population are shown in Tables 2 and 3.⇓⇓Patients who had received unfractionated or low molecular weight heparin (LMWH) were sampled a minimum of 48 h after discontinuation of these. Sampling was performed 5 ± 3 days after admission. Fasting was not required. All patients received dual-antiplatelet medication for a whole year for follow-up.
A control group of 100 healthy normal volunteers (age 38 ± 11 years) not taking medication was also tested. Informed written consent was obtained from all subjects, and the study was approved by the local research ethics committee.
Assessment of thrombotic and thrombolytic status
Blood samples were taken from an antecubital vein using an 18-G butterfly cannula using a 2-syringe technique. The first 3 ml blood was used for routine blood tests and the next 2 3-ml samples were used for thrombotic status assessment.
Thrombotic status and endogenous thrombolytic activity was assessed using the GTT. The GTT is a novel, point-of-care assay that employs native (nonanticoagulated) blood. The instrument assesses the time taken to create a shear-induced thrombus under physiological conditions and in the second phase of the test, measures the time to achieve endogenous thrombolysis of the thrombus created during the first phase of the test. The principle of the GTT has previously been described in detail (8). The instrument measures the time (d) between 2 consecutive blood drops. This time interval increases gradually as flow slows down and at an arbitrary point (d ≥15 s, before reaching complete occlusion), the end point of the measurement is displayed (occlusion time [OT], in seconds). Restart of blood flow after occlusion is due to spontaneous thrombolysis (lysis time [LT], in seconds). If lysis does not occur until 6,000 s after OT (LT cutoff time), “no lysis” is displayed and recorded. Coefficient of variation was assessed by testing 1 healthy volunteer 10 times over a period of 4 weeks, under similar conditions, and by testing 10 volunteers twice, at 48-h intervals. The coefficient of variation was also assessed by testing 10 patients with ACS twice, at 24-h intervals.
In a subgroup of 100 patients, thrombotic status was also assessed using the VerifyNow assay (Accumetrics, San Diego, California) and compared with thrombotic status measured using the GTT.
Venous blood samples were taken and anticoagulated with sodium citrate 0.109 mol/l (ratio 9:1). The VerifyNow system is a turbidometry-based optical detection device that measures platelet-induced aggregation in a system containing fibrinogen-coated beads (9). The instrument measures changes in light transmission and thus the rate of aggregation in whole blood. In the cartridge of the VerifyNow P2Y12 assay, there is a channel in which inhibition of the adenosine diphosphate (ADP) P2Y12 receptor is measured. This channel contains ADP as platelet agonist and prostaglandin E1 as a suppressor of intracellular free calcium levels, to reduce the nonspecific contribution of ADP binding to P2Y1 receptors. Results are expressed as P2Y12 reaction units (PRU). Recently, in patients with ACS, a cutoff level of PRU ≥240 was shown to be predictive of major cardiovascular events (10). The VerifyNow system does not assess thrombolytic status.
Data collection and follow-up
Patients were recruited into the study during their index admission, and baseline demographics obtained. Follow-up was done at 3-month intervals, and source documents of major adverse cardiac event (MACE) obtained. The primary end point of the study was a composite of occurrence of cardiovascular death, nonfatal MI, or stroke (MACE) at 1 year.
The end points of the study were as follows: 1) cardiovascular death, defined as death in the presence of ACS, significant cardiac arrhythmia, or refractory congestive heart failure, or death attributed to cardiovascular cause at postmortem; 2) nonfatal MI (a rise in serum troponin I or an increase in creatine kinase-myocardial band isoenzyme at least twice the upper normal limits with at least 1 of the following: acute onset of prolonged [≥20 min] typical ischemic chest pain; ST-segment elevation of at least 1 mm in 2 or more contiguous electrocardiographic leads, or ST-segment depression ≥0.5 mm in ≥2 contiguous leads; or T-wave inversion >1 mm in leads with predominant R waves; or 3) stroke, defined as the presence of a new focal neurologic deficit thought to be vascular in origin, with signs or symptoms lasting more than 24 h, preferably supported by an imaging procedure such as a computed tomography or magnetic resonance imaging.
Unpaired ttests were used for comparison of normally distributed variables, and Mann-Whitney Utest used for nonnormally distributed variables. Dichotomous variables were compared by chi-square test or Fisher's exact test, as appropriate. Where necessary, log transformations were employed. Correlations were analyzed using Spearman's coefficient of rank correlation.
The OT and LT were separately related to MACE. To establish whether OT outside the normal range may be predictive of MACE, a normal range was established from mean ± 2 SDs of normal volunteers.
Ability of the test to discriminate between patients with and without MACE was evaluated by receiver-operating characteristic curve analysis. The optimal cutoff value was calculated by determining the LT value providing the greatest sum of sensitivity and specificity. In addition, to investigate the relationship between increasing LT and MACE, Cox regression was performed on LT divided into bands of 1,000 s. Kaplan-Meier survival methods with log-rank tests were used to compare survival curves. Univariate and multivariate hazard regression models of Cox were used, respectively, to identify risk factors for clinical end points and to adjust for potential confounders that were associated with clinical end points on univariate analysis. Net reclassification improvement (NRI) was used to assess the added predictive ability of LT >3,000 for MACE.
A significance level was defined as p < 0.05. All analysis was performed with SPSS version 16.0 (SPSS Inc., Chicago, Illinois).
Completed follow-up was available in 297 patients. The coefficient of variation was 6.2% for OT and 20.9% for LT in normal volunteers, and 7% for OT and 19% for LT in ACS patients. These values are very similar to the CV obtained in earlier studies using this technique (8,11,12).
The distribution of OT and LT of ACS patients and of normal volunteers is shown in Figure 1.The OT in both the normal and the ACS groups was normally distributed. In ACS patients, OT was significantly prolonged compared with that of normal volunteers (428 ± 155 s vs. 378 ± 96 s, p < 0.001). However, survival analysis did not demonstrate a relationship between OT and MACE.
The distribution of LT was positively skewed in both normal subjects and ACS patients (Figs. 1B and 1D). The LT was significantly prolonged in ACS patients compared with healthy volunteers (median 1,053 [978 to 1,125) s vs. 1,362 [1,240 to 1,514) s, p < 0.001). Some 23% of ACS patients had LT ≥3,000 s, compared with none of the normal subjects, with a significant number of patients demonstrating markedly impaired thrombolytic status with LT ≥5,000 s.
Thrombolytic status was strongly predictive of MACE. The event-free survival curves for cardiovascular death, nonfatal MI, and stroke according to LT are shown in Figure 2.Receiver-operating characteristic curve analysis showed that LT level significantly discriminated between patients with and without MACE, with an area under the curve of 0.63 (95% confidence interval [CI]: 0.51 to 0.69; p < 0.05). A LT ≥3,000 s was identified as the optimal cut point to predict MACE outcome, with sensitivity of 60% and specificity of 80% (Fig. 3).Above this LT level, the hazard ratio (HR) increased as the LT increased (Fig. 4).Breakdown of MACE according to LT is shown in Figure 5and Table 4.
Using univariate Cox regression analysis, LT (cutoff 3,000 s) was associated with a significantly higher risk of both cardiovascular death (HR: 4.2, 95% CI: 1.13 to 15.62, p = 0.03) and nonfatal MI (HR: 2.09, 95% CI: 1.02 to 4.27, p = 0.04) whereas no association was detected for stroke, probably because only 1 stroke occurred in the entire group.
The following variables were interrogated for effects on OT and LT: all patient characteristics in Table 2; the echocardiographic, angiographic, and interventional characteristics in Table 3; recent LMWH (>2 days earlier) and recent thrombolysis (>5 days earlier). Only statin was found to have a direct effect on OT, with patients on statin therapy exhibiting a longer (less thrombotic) OT than patients not on statin therapy (439 ± 157 s vs. 359 ± 120 s, p = 0.02). Of all the variables, only age was related to LT, with increasing age relating to increased LT (p = 0.008), although there was also a trend toward a relationship between LT and troponin (p = 0.055). There was a weak negative association between OT and LT (r = −0.2, p = 0.07).
Univariate analysis suggested that of all the earlier variables, only the following were related to MACE: age (p = 0.005), diabetes mellitus (p = 0.03), low hemoglobin (p = 0.005), low hematocrit (p = 0.03), peripheral vascular disease (p = 0.002), angiotensin-converting enzyme inhibitor treatment (p = 0.016), and oral nitrate therapy (p = 0.001). Only a very small number of patients had peripheral vascular disease.
The following variables were then entered into the final model of multivariable logistic regression analysis: age, diabetes, angiotensin-converting enzyme inhibitor, and nitrate. Multivariate analysis showed that LT remained a predictor of MACE after adjustment for these risk factors (MACE HR: 2.63, 95% CI: 1.4 to 4.95, p = 0.003).
Net reclassification improvement showed that the inclusion of LT 3,000 s in a model containing 3 significant predictors (age, diabetes, and ACE inhibitor) significantly added to the model effectiveness (net reclassification improvement 0.259, p < 0.001).
Among ACS patients, 15% were deemed prothrombotic based on results of the VerifyNow assay (PRU <240), compared with 5% based on GTT OT status (OT <200 s). There was a negative relationship between GTT OT and VerifyNow PRU values (r = −0.2, 95% CI: −0.3 to −0.01, p = 0.01), so that as OT increased (less thrombotic), so PRU decreased (less thrombotic). The results of the VerifyNow assay were also assessed for relationship to MACE. Both as a continuous variable, and using a cutoff of PRU ≥240, we found no relationship between VerifyNow results and MACE in our population.
Our results confirm that despite effective dual-antiplatelet medication, some patients with ACS remain at risk of recurrent thrombotic events, and that is at least partly related to impaired endogenous thrombolytic activity.
Dual-antiplatelet therapy (aspirin + clopidogrel) is the current standard of care in patients with ACS. In this study, we tested 100 cardiac patients for clopidogrel resistance, using the most credited point-of-care VerifyNow P2Y12 assay. Our finding of 15% resistance (depending on the criteria of resistance) was confirmed recently in a much larger number of patients (13). Likewise, the low incidence (5%) of nonresponsiveness as measured by the GTT assay supports recently reported data of 6% dual (aspirin and clopidogrel) nonresponsiveness (3). Thus, the much lower number of patients exhibiting dual nonresponsiveness assessed by GTT compared with the VerifyNow P2Y12 assay supports the view that the VerifyNow P2Y12 assay may fail to accurately quantify platelet inhibition by clopidogrel (14), and, importantly, platelet contribution to coagulation by thrombin generation. As such, the possibility of inhibition of thrombin generation by aspirin (15–17) and clopidogrel (18) is not taken into account in the assessment of thrombotic status by tests that use anticoagulated blood. The LMWH that patients received on admission could not interfere with the GTT assay, as this was performed >48 h after the last dose of LMWH, by which time heparin had completely cleared from the circulation (19).
Although the thrombotic status of ACS patients on dual-antiplatelet medication was no less favorable than that of healthy normal subjects not taking medication, the thrombolytic status was markedly impaired in many ACS patients. Emerging evidence supports the assumption that acute MI is a failure of timely, spontaneous thrombolysis (5). In 585 patients with ST-segment elevation MI, electrocardiographic or angiographic spontaneous reperfusion was observed in 14.9% and 14.7%, respectively. Those with spontaneous reperfusion had lower mortality, lower composite of death/shock/congestive heart failure, and significant reduction in death or reinfarction (20). In a recent study of 710 patients with ST-segment elevation MI undergoing percutaneous coronary revascularization, spontaneous reperfusion was observed in 22% of patients, and at 30 days, was associated with significantly lower incidence of death, congestive heart failure, and recurrent ACS (21). It has also been suggested that impaired fibrinolysis determines the outcome of coronary angioplasty (22). Large prospective studies provided evidence for a statistically significant association between inhibited fibrinolysis as indicated by increased levels of fibrinolysis inhibition markers such as tissue-type plasminogen activator, plasminogen activator inhibitor (PAI)-1, thrombin activatable fibrinolysis inhibitor (TAFI), plasmin-antiplasmin complex, lipoprotein(a), and an increased risk of recurrent MI or sudden cardiac death (23–28). However, multivariate analysis of the measured plasma concentrations by these fibrinolysis markers showed weak or no prognostic value of these tests.
From the several global clot lysis assays described in the past few years, we have chosen the GTT technique, which has the advantage of measuring thrombolysis as opposed to clot lysis. Although worthy of further prospective confirmation, our results suggest a strong association between impaired spontaneous thrombolysis and MACE. A very recent study involving 335 young survivors of a first arterial thrombosis reported that low plasma fibrinolytic potential, measured by clot LT, and found in 10% of the population, was associated with a 2-fold increase in relative risk of arterial thrombosis (29). Our findings show that in patients with ACS, the increase in risk associated with impaired lysis can even be higher.
The main limitation of our study is that patients were only sampled once during their index admission, generally when they had stabilized on medical therapy. Firstly, evidence has been provided that the antiplatelet effects of aspirin (in particular, the inhibition of thrombin generation) (15,30) and clopidogrel (31,32) are dose dependent. When we measure platelet reactivity in patients under the effect of the loading dose of 300 mg aspirin and 300 mg clopidogrel, such measurements do not necessarily reflect steady-state platelet function during the maintenance doses of 75 mg/day.
Secondly, heparin administered immediately after admission may have interfered with the GTT measurement. The GTT OT is determined not just by the rate of platelet aggregation but also by the formation and the effect of thrombin, generated by activated platelets. Based on published data regarding the plasma half-life of heparin, at the time of sampling, we anticipated no interference from heparin with the test. However, there is some evidence that heparin binds to platelets (33), and such membrane-bound heparin (whose effect is not detectable by the activated partial thromboplastin time test) may affect platelet behavior for much longer periods. With its strong antithrombin effect (34), such platelet-bound heparin may have interfered with our GTT measurements. Had sampling been performed about 10 days after admission, when patients were established on maintenance doses of antiplatelet drugs and also interference by heparin could have been excluded, it is possible that the platelet reactivity may have been predictive of MACE, and impaired thrombolysis may have been even more predictive of future events. Furthermore, it remains unknown whether the impaired thrombolysis was part of an acute phase response or whether it reflects a chronic impairment of thrombolysis that influences late outcome.
Possible diurnal variation in LT was not investigated. Further studies are required to investigate the effect of medication on thrombolytic state, both in the acute and the chronic state. The study was underpowered to specifically investigate thrombolytic status in specific subgroups of ACS patients such as those with ST-segment elevation MI, diabetes, smoking, or the effects of ACS medications on LT. Furthermore, the effect of percutaneous coronary revascularization and stenting on thrombolytic status remains to be investigated.
To date, chronic modulation of fibrinolytic state as a pharmacologic target has largely been ignored, mainly due to difficulties in measuring a clinically meaningful marker. Major inhibitors of fibrinolysis such as PAI-1 and TAFI can be targeted directly. Platelets contain approximately 90% of the total amount of PAI-1 present in blood, and this PAI-1 together with TAFI are released only from activated platelets (35). Of all platelet agonists, at physiological [Ca2+], only thrombin can induce release of PAI-1 and TAFI from storage granules of platelets. For this reason, in addition to direct inhibitors of PAI-1 and TAFI, inhibition of thrombin would achieve both objectives (36). Development of orally acting, safe thrombin inhibitors would prevent the release of PAI-1 and TAFI at sites of platelet activation. Of the 2 standard antiplatelet drugs, aspirin and clopidogrel, clopidogrel does not affect the fibrinolytic system (37), whereas at high doses, aspirin is claimed to exert fibrinolytic effects (15,38,39). To boost spontaneous thrombolytic activity by endogenous profibrinolytic agents, interferon-alpha synthesized by leukocytes could be targeted. Recently, high-dose aspirin was shown to be capable of boosting interferon-alpha production by leukocytes (40). However, at the commonly used dose of 75 or 80 mg a day, aspirin is unlikely to exert a profibrinolytic effect. It is also claimed that dihydropyridine calcium antagonists increase fibrinolytic activity independently of their antihypertensive action (41). Future studies will be required to assess whether pharmacological modulation of endogenous thrombolytic status will lead to a reduction in thrombotic events over and above that which can be achieved with dual-antiplatelet therapy.
The present study suggests for the first time that impaired endogenous thrombolysis in ACS patients is an independent risk factor for future thrombotic events. The results not only potentially dichotomously identify patients at high risk of future events, but also highlight the potential importance of modulation of spontaneous thrombolysis as a novel therapeutic target in such patients.
- Abbreviations and Acronyms
- acute coronary syndrome
- adenosine diphosphate
- confidence interval
- global thrombosis test
- hazard ratio
- low molecular weight heparin
- lysis time
- major adverse cardiovascular event
- myocardial infarction
- occlusion time
- plasminogen activator inhibitor
- P2Y12 reaction units
- thrombin activatable fibrinolysis inhibitor
- Received September 14, 2009.
- Revision received December 23, 2009.
- Accepted January 6, 2010.
- American College of Cardiology Foundation
- Gori A.M.,
- Marcucci R.,
- Migliorini A.,
- et al.
- Swan H.J.
- Marcucci R.,
- Gori A.M.,
- Paniccia R.,
- et al.
- Ratnatunga C.P.,
- Edmondson S.F.,
- Rees G.M.,
- Kovacs I.B.
- van der Meijden P.E.,
- Feijge M.A.,
- Giesen P.L.,
- Huijberts M.,
- van Raak L.P.,
- Heemskerk J.W.
- Cushman M.,
- Lemaitre R.N.,
- Kuller L.H.,
- et al.
- Morange P.E.,
- Saut N.,
- Alessi M.C.,
- et al.
- Undas A.,
- Brummel-Ziedins K.E.,
- Mann K.G.
- Dangas G.,
- Mehran R.,
- Guagliumi G.,
- et al.
- Aleil B.,
- Jacquemin L.,
- De Poli F.,
- et al.
- Horne M.K. III.,
- Chao E.S.
- De Candia E.,
- De Cristofaro R.,
- Landolfi R.
- Bhattacharyya M.,
- Karmohapatra S.K.,
- Bhattacharya G.,
- Bhattacharya R.,
- Sinha A.K.
- Vergouwen M.D.,
- Vermeulen M.,
- de Haan R.J.,
- Levi M.,
- Roos Y.B.