Abstract
Background. Many complications associated with congestive heart failure (CHF) have a thrombosis-related aetiology. Platelets play an important role in thrombogenesis, but it is not clear whether circulating platelets actively participate in thrombosis-related complications associated with CHF.
Objective. To determine whether soluble P-selectin, platelet surface P-selectin, and total platelet P-selectin as indices of platelet activation in CHF patients—compared to ‘disease controls’ and ‘healthy controls’—and to assess their prognostic value in CHF.
Methods. We measured soluble P-selectin (sP-sel, by enzyme-linked immunosorbent assay, ELISA), total platelet P-selectin (pP-sel, by a novel ‘platelet lysate’ assay), platelet surface P-selectin (CD62P%G) and platelet surface CD63 (CD63%G) expression by flow cytometry—in 108 patients with stable congestive heart failure (all with left ventricular ejection fraction (LVEF) <50%). Levels were compared with 50 healthy controls and 70 ‘disease controls’ (patients with coronary artery disease with normal left ventricular systolic function).
Results. CHF patients and disease controls had higher sP-sel, CD62P%G and CD63%G than healthy controls. There were no significant correlations between sP-sel, pP-sel, CD62P%G and CD63%G with ejection fraction (all P>0.05). There were no differences in these markers when ischaemic and non-ischaemic aetiologies of CHF were compared. After a median follow-up of 490 days (range 340–535), there were 7 deaths, 15 hospitalizations for worsening heart failure, 1 for cardiac resynchronization therapy, 4 for revascularizations, 4 for myocardial infarctions, and 1 stroke. None of the platelet markers were predictive of the composite end-point at follow-up.
Conclusions. Patients with stable CHF exhibit evidence of abnormal platelet activation, despite usage of antiplatelet agents. These abnormalities did not determine prognosis and were broadly similar to those seen in ‘disease controls’ indicating that platelet abnormalities in CHF may simply be related to associated comorbidities.
Introduction
Patients with congestive heart failure (CHF) are at increased risk of venous thromboembolism, stroke, and sudden death. The incidence of stroke is approximately 4%–5% in severe heart failure compared to 0.5% of the general population in the same age group Citation1. Sudden cardiac death may also have a similar thrombotic origin (i.e. intracardiac or intracoronary). These thrombotic complications in patients with CHF have been attributed to a prothrombotic state, by fulfilment of Virchow's triad for thrombogenesis, that is, abnormal blood constituents, blood flow, and vessel wall Citation2.
Platelet abnormalities are well recognized in CHF patients Citation3–5. P-selectin is of great interest because of its role in modulating interactions between platelets, leucocytes, and endothelium Citation6. Abnormal surface P-selectin expression Citation7, Citation8 and soluble P-selectin (sP-sel) levels Citation9 have been reported in decompensated heart failure. High levels of the latter may promote coagulation in animal models Citation10. However, it is unclear whether these circulating platelets actively participate in the coagulopathy of CHF. In an experimental model, Michelson et al. Citation11 have shown that circulating degranulated platelets rapidly lose surface P-selectin (that is, CD62P), and that levels in the plasma pool rise, and, more importantly, those P-selectin-negative platelets continue to circulate and function. However, it is unclear whether or not this phenomenon occurs in humans, or the timescale of the shedding of P-selectin.
Despite the common use of these approaches in assessing platelet activation, few studies have compared the two methods in the same groups of patients and controls. Indeed, it remains unclear whether sP-sel measurement or CD62P expression on platelets by flow cytometric analysis (CD62P%G) is a more reliable marker of in-vivo platelet activation Citation12. Another alternative to quantifying membrane and sP-sel is to simply measure the total mass of P-selectin (pP-sel) in a detergent lysate of a given number of platelets (e.g. 108), thus providing the index of the mass of P-selectin per platelet Citation13. However, abnormalities of pP-sel and sP-sel in CHF patients have not been previously explored. In addition, it is unclear whether or not the amount of P-selectin in each platelet varies in CHF, and whether or not the amount of P-selectin in each platelet is altered by antithrombotic therapy. Certainly, platelet alpha-granule secretion in response to adenosine diphosphate (ADP) is independent of the lipoxygenase- and cyclooxygenase-dependent metabolites of arachidonic acid and aspirin, and so may not decrease ADP-induced P-selectin expression on platelets or sP-sel Citation14, Citation15. From a clinical perspective, the Antithrombotic Trialists’ Collaboration Citation16 meta-analysis suggests that antiplatelet therapy produced a highly significant 22% reduction in vascular deaths. However, recent trials in CHF (WASH Citation17 and WATCH Citation18) have suggested that aspirin usage was associated with increased hospitalization for heart failure.
The aim of the present study is to explore the interrelationships between CD62P, sP-sel, and pP-sel in CHF. We hypothesized the following: 1) platelets from CHF have increased pP-sel levels, and are related to platelet CD62P expression (CD62P%G) and sP-sel levels; and 2) these platelet indices are of clinical significance in predicting adverse cardiovascular events in CHF patients.
Key messages
Many complications associated with congestive heart failure (CHF) have a thrombosis-related aetiology. Platelets play an important role in thrombogenesis, but it is not clear whether circulating platelets actively participate in thrombosis-related complications associated with CHF.
In this study we found that patients with stable CHF exhibit evidence of abnormal platelet activation, despite usage of antiplatelet agents.
These abnormalities did not determine prognosis and were broadly similar to those seen in ‘disease controls’ indicating that platelet abnormalities in CHF may simply be related to associated comorbidities.
Methods
We prospectively recruited stable CHF out-patients with documented left ventricular systolic dysfunction, defined as those with an ejection fraction <50%, either by echocardiography, radionuclide imaging, or left ventriculography within the preceding 3 months. Subjects with CHF were classified according to the New York Heart Association (NYHA) criteria, with class I to II being no or mild symptoms and class III to IV being moderate to severe symptoms. Patients were excluded from the study for the following criteria: recent (<3 months) acute coronary syndrome, CHF decompensation, cerebrovascular accident, or infections. Subjects with acute or chronic liver or renal disease, serological evidence of hepatitis B virus, known Human Immunodeficiency Virus (HIV) infection, malignancy, connective tissue disease, and treatment with antibiotics, hormone replacement therapy, immunosuppressive, and cytotoxic drugs were also excluded. We screened 111 patients and 3 were excluded: 1 with respiratory tract infection, 1 with connective tissue disease and 1 with malignancy. These CHF patients were compared with ‘disease controls’ (defined as patients with angiographic evidence of coronary disease (i.e. significant (>50% stenosis, or occlusive) coronary artery disease) but with normal systolic left ventricular function) and ‘healthy controls’ (recruited from among healthy hospital staff and from subjects attending the hospital for hernia repairs, varicose vein procedures, or other relatively minor operations). All healthy control subjects had no clinical evidence of vascular, metabolic, neoplastic, diabetic, or inflammatory disease on careful history, examination, and routine laboratory tests. Clinical characteristics of patients and controls are summarized in . The study was conducted in accordance with the Declaration of Helsinki and with the approval of the local research ethics committee. Written informed consent was obtained from all participants.
Table I. Base-line characteristics of patients and controls.
Laboratory
For the ‘platelet lysate’ assay, venous blood was drawn from an antecubital vein with minimal stasis, into Vacutainer tubes (Becton Dickinson, UK) containing 0.5 mL 3.8% sodium citrate, and was centrifuged at 170 g (1000 rpm) for 10 minutes, and the supernatant, platelet-rich plasma (PRP), aspirated and placed into a fresh tube. The platelet count was performed using a haemocytometer (ADVIA 120, Bayer, Newbury, Berks, UK). The PRP was then spun at 1500 g (3000 rpm) for 20 minutes to produce a platelet pellet that was resuspended in sterile saline solution to give a concentration of 2×108/mL. This was lysed with an equal amount of 0.1% phosphate-buffered saline (PBS)-tween and then stored for batch analysis for P-selectin by ELISA to give pP-sel. Intra-assay coefficients of variation for all enzyme-linked immunosorbent assays (ELISAs) were <5%; interassay variances were <10%.
Blood for platelet flow cytometric analysis was collected according to the method described in the European Working Group on Clinical Cell Analysis consensus protocol Citation19. For this analysis venous blood was taken from an antecubital vein with minimal stasis, into Vacutainer tubes (Becton Dickinson, UK) containing 4.5 mL CTAD (citrate, theophylline, adenosine, and dipyridamole). All blood samples were processed promptly on a flow cytometer (FACScan, BD Biosciences, Oxford, UK) within 10 minutes of venepuncture. CaliBRITE three-colour beads and FACSComp software were used to set photomultiplier tube voltages, fluorescence compensation, spectral overlap, and sensitivity. Forward (size-dependent) (FSC) and 90° sideways scatter (SSC) were set at logarithmic gain. Mouse IgG2a and IgG1 antikeyhole limpet haemocyanin monoclonal antibodies conjugated to phyco-erythrin were used as isotype-negative controls to define non-specific platelet binding. All monoclonal antibodies were obtained from BD Biosciences. Data acquisition and analysis was performed with CELLQuest software version 3.1 (Becton Dickinson, UK).
For platelet flow cytometry, 100 µL fresh blood was diluted with 800 µL of phosphate-buffered saline (PBS) (Sigma, UK). For determination of the percentage P-selectin positivity, 15 µL of the diluted blood was incubated with 10 µL each of anti-CD42b-FITC (fluorescein isothiocyanate, marking glycoprotein 1b) and anti-CD62P-PE (binds to platelet membrane bound p-selectin). Similarly, a 15 µL aliquot of diluted blood was added to 10 µL each of anti-CD42b-FITC and anti-CD63-PE (birds to platelet lysosomal membrane glycoprotein) for determination of%CD63, which was measured as another platelet activation marker Citation20, to allow a comparison with P-selectin. This was then diluted by 800 µL of PBS immediately before flow cytometry. Platelet events were gated for size and fluorescence in the FSC versus CD42b dot plot based on the characteristic signal. A gate region (R1) was set up to acquire all data except those, which were, fluorescence-negative and with a high side SSC profile (i.e. leukocytes). In the list mode 10,000 events were acquired at a flow rate of <500 particles per second and subsequently analysed for CD62P or CD63 expression. Results are expressed as per cent events CD62P- or CD63-positive in terms of the total number of CD42b-positive events in the appropriate (R1) gate. The inter- and intra-assay coefficients of variation were <5% and <8%, respectively.
Clinical follow-up
All patients were followed up by regular out-patient visits. Clinical information regarding major adverse cardiac events (death, myocardial infarction, stroke, hospitalization for worsening heart failure, revascularization, and cardiac resynchronization therapy) during a median follow-up of 490 days (range 340–535) was collected.
Power calculations
CHF carries a significant 10%–50% annual adverse event rate, depending on severity. We conservatively estimate an annual event rate of 33%. If we hypothesized that an increase of sP-sel of one half of a standard deviation will indicate an increased risk of adverse outcome (P < 0.05, 1-β = 0.8), we needed 25 patients to have had an end-point and 25 to be free of an end-point. As this end-point rate (i.e. 50%) is unrealistic, we would recruit instead 100 patients to be confident of getting at least 25 end-points (i.e. an end-point rate of 25%). As we intended to perform other analyses (e.g. pP-sel), we recruited 108 patients to provide the additional confidence required of multiple regression analyses and subanalyses, such as that for disease severity.
Statistical analysis
Following a test of statistical normality, the data were expressed as mean with standard deviation (SD) or as median with interquartile range (IQR) for the normally distributed data and non-parametrically distributed data, respectively. Comparisons between patients, disease and healthy controls were analysed by analysis of variance (ANOVA) or Kruskal-Wallis test, as appropriate. Correlations were performed by Spearman's correlation method. Stepwise multiple regression analyses were used to determine the predictors of abnormal base-line plasma markers. Survival probabilities were calculated using the univariate log-rank test and Kaplan-Meier method; for multivariate survival analysis, a Cox proportional hazards analysis was used. Statistical significance was accepted at the 0.05 level (two-sided). Statistical analyses were undertaken using SPSS software v. 12 (SPSS Inc., Chicago, Illinois). All other statistical calculations were performed on a microcomputer using a commercially available statistical package (Minitab release 13, Minitab 13, Minitab Inc., State College, PA, USA). A P-value of <0.05 was considered as statistically significant.
Results
We studied 108 patients (86 men; mean age 64±11 years) with CHF (mean LVEF 34±8%), who were compared with 70 disease controls (51 men; mean age 64±10 years, mean LVEF 64±8%) and 50 healthy controls (31 men; mean age 61±11 years, mean LVEF 72±12%) (). Systolic blood pressure was significantly lower in CHF patients compared to disease controls and healthy controls. Lower cholesterol levels were noted in the patient groups, reflecting the use of statins.
Patients with CHF and disease controls had higher sP-sel (P = 0.005), CD62P%G (P = 0.001), and CD63%G (P = 0.013) compared to healthy controls (a). There was no difference in pP-sel between the three study groups. Patients with NYHA III–IV had higher CD62P%G compared to those in NYHA I–II (P = 0.036) (b). There were no significant differences in these platelet markers between ischaemic and non-ischaemic heart failure (c). There were also no significant differences in these platelet markers in relation to the degree of left ventricular systolic dysfunction (d). There were ethnic differences whereby South Asians had higher median sP-sel levels than Afro-Caribbeans [82.5 (42.3–110) vs 47.5 (34.1–58.8) ng/ml; p = 0.013 by Mann-Whitney test].
Table II. Platelet abnormalities and activation in systolic heart failure.
(b) Comparison of platelet abnormalities according to NYHA classification.
(c) Platelet abnormalities and activation in ischaemic versus non-ischaemic heart failure.
(d) Platelet abnormalities according to severity of left ventricular ejection fraction.
Correlations
SP-sel and pP-sel were positively correlated (Spearman correlation, r=0.374, P < 0.0001), whilst CD62P%G was negatively correlated with sP-sel (r= − 0.257, P = 0.012) and pP-sel (r= − 0.205, P = 0.049) ().
Table III. Correlations between various platelet indices/markers.
Prognosis
In the cohort of median follow-up of 490 days (range 340–535), there were 7 deaths, 15 hospitalizations for worsening heart failure, 1 for cardiac resynchronization therapy, 4 for revascularizations, 4 for myocardial infarctions, and 1 stroke. Using a multivariate approach, in a multivariate Cox regression analysis model (that included base-line independent cardiovascular risk factors, medical history, and medications) none of the platelet research indices were statistically significant predictors for outcome ().
Table IV. Multivariate analysis of the predictive value of the platelet indices in heart failure patients.
Discussion
Our study demonstrates platelet abnormalities in CHF, although most may simply relate to associated comorbidities, given the non-significant trend towards higher sP-sel levels in CHF compared to disease controls. Previous studies have shown increased platelet surface P-selectin in decompensated heart failure Citation7, Citation8, and we show that this situation also exists in stable CHF patients, despite the high usage of antiplatelet agents. We found no significant difference in these platelet markers between ischaemic and non-ischaemic heart failure.
The elevated CD62P%G, CD63%G, and sP-sel in CHF are consistent with other studies that have reported platelet abnormalities in this condition. We found no relationship between platelet activation and LVEF, although there was a non-significant trend toward higher sP-sel levels in NYHA class III–IV. We also observed higher median sP-sel levels in South Asians than in other ethnic groups, in keeping with previous studies Citation21.
Although the origin of platelet activation in CHF remains to be established, increased platelet activity might be related to elevation of cytosolic free calcium concentrations Citation22, triggered by elevated levels of tumour necrosis factor Citation23, or may be affected by enhanced sympathoadrenal activation Citation24 and catecholamine release in CHF Citation25. Since platelet abnormalities were no different between CHF patients and the disease control group, it is possible that platelet abnormality may be largely due to underlying ischaemic heart disease or aetiological factors for CHF. Indeed, diabetes mellitus, hypercholesterolaemia, and hypertension—but not vascular disease per se—are associated with persistent platelet activation in vivo Citation26. Our subgroup analysis suggests no significant difference in platelet abnormalities between ischaemic heart failure and idiopathic cardiomyopathy.
Gurbel et al. Citation27 compared platelet surface and sP-sel in patients presented with chest pain and reported that the two platelet indices were not significantly correlated. In our cohort, these indices were negatively correlated, which may indicate the shedding of P-selectin into the circulation, causing a rise in sP-sel levels, in keeping with previous studies Citation11, Citation12. Our data also show a positive correlation between pP-sel and sP-sel, confirming the source of P-selectin. We did not find any difference in pP-sel between the disease groups and healthy controls, which may be due to antiplatelet therapy use. During platelet activation, it is possible that the differential expression and shedding of P-selectin—rather than the actual pool of P-selectin in the platelets—is more important. This does not, however, exclude the possible source of P-selectin shedding from the endothelium in CHF. Indeed, P-selectin may provide an important link between inflammation and coagulation. Certainly, activated platelets in circulation have been shown to initiate an acute inflammatory response by inducing P-selectin expression on the endothelium and consequently stimulating leukocyte rolling on the vessel wall Citation25. This was followed by release of von Willebrand factor from Weibel-Palade bodies, which could potentially increase the thrombotic tendency.
The prognostic value of P-selectin is so far unclear, and various studies in healthy subjects and CHF patients have been inconclusive Citation28–31. A strength of our study is our comparison of the various platelet activation markers, their interrelationships, and the implications for prognosis. Of note, we found that none of the platelet markers were prognostically important in CHF, suggesting that platelet activation—whilst present—may not have a major role in the pathophysiology of adverse outcomes in CHF. The major limitation of this study is that it is a small study with median follow-up of 490 days. Other limitations include its cross-sectional design and some imbalance in comorbidities. For example, the CHF patients had significantly lower blood pressure than both controls groups. Finally, it remains uncertain whether the abnormal platelet activation in CHF, despite usage of antiplatelet agents, reflects true ‘aspirin resistance’ per se—or even non-compliance—with antiplatelet therapy Citation32.
In conclusion, patients with stable CHF exhibit evidence of abnormal platelet activation, despite usage of antiplatelet agents. These abnormalities did not determine prognosis and were broadly similar to those seen in ‘disease controls’ indicating that platelet abnormalities in CHF may simply be related to associated comorbidities.
Acknowledgements
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References
- Nguyen KN, Ivar A, Kjekshus J. Interaction between enalapril and aspirin on mortality after acute myocardial infarction: subgroup analysis of the Cooperative New Scandinavian Enalapril Survival Study II (CONSENSUS II). Am J Cardiol. 1997; 79: 115–9
- Lip GYH, Gibbs CR. Does heart failure confer a hypercoagulable state?. J Am Coll Cardiol. 1999; 33: 1424–6
- Chung I, Lip GYH. Platelets and heart failure. Eur Heart J. 2006; 27: 2623–31
- Jafri SM, Ozawz T, Mammen E, Levine TB, Johnson C, Goldstein S. Platelet function, thrombin and fibrinolytic activity in patients with heart failure. Eur Heart J. 1993; 14: 205–12
- Erne P, Wardle J, Sanders K, Lewis SM, Maseri A. Mean platelet volume and size distribution and their sensitivity to agonists in patients with coronary artery disease and congestive heart failure. Thromb Haemost. 1988; 59: 259–63
- Blann AD, Nadar SK, Lip GYH. The adhesion molecule and P-selectin and cardiovascular disease. Eur Heart J. 2003; 24: 2166–79
- O'Connor C, Gurbel PA, Serebruany VL. Usefulness of soluble and surface-bound P-selectin in detecting heightened platelet activity in patients with congestive heart failure. Am J Cardiol. 1999; 83: 1345–9
- Stumpf C, Lehner C, Eskafi S, Raaz D, Yilmaz A, Ropers S, et al. Enhanced levels of CD154 (CD40 ligand) on platelets in patients with chronic heart failure. Eur J Heart Fail. 2003; 5: 629–37
- Gibbs CR, Blann AD, Watson RD, Lip GY. Abnormalities of hemorheological, endothelial, and platelet function in patients with chronic heart failure in sinus rhythm. Circulation. 2001; 103: 1746–51
- Andre P, Hartwell D, Hrachovinova I, Saffaripour S, Wagner DD. Pro-coagulant state resulting from high levels of soluble P-selectin in blood. Proc Natl Acad Sci U S A. 2000; 97: 13835–40
- Michelson AD, Barnard MR, Hechtman HB, MacGregor H, Connolly RJ, Loscalzo J, et al. In vivo tracking of platelets: circulating degranulated platelets rapidly lose surface P selectin but continue to circulate and function. Proc Natl Acad Sci U S A. 1996; 93: 11877–82
- Ferroni P, Pulcinelli FM, Lenti L, Gazzaniga PP. Is soluble P-selectin determination a more reliable marker of in vivo platelet activation than CD62P flow cytometric analysis?. Thromb Haemost. 1999; 81: 472–3
- Kamath S, Blann AD, Caine GJ, Gurney D, Chin BS, Lip GY. Platelet P-selectin levels in relation to plasma soluble P-selectin and beta-thromboglobulin levels in atrial fibrillation. Stroke. 2002; 33: 1237–42
- Rinder CS, Student LA, Bonan JL, Rinder HM, Smith BR. Aspirin does not inhibit adenosine diphosphate-induced platelet α-granule release. Blood. 1993; 82: 505–12
- Storey RF, Judge HM, Wilcox RG, Heptinstall S. Inhibition of ADP-induced P-selectin expression and platelet-leukocyte conjugate formation by clopidogrel and the P2Y12 receptor antagonist AR-C69931MX but not aspirin. Thromb Haemost. 2002; 88: 488–94
- Antithrombotic Trialists’ Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ. 2002;324:71–86.
- Cleland JG, Findlay I, Jafri S, Sutton G, Falk R, Bulpitt C, et al. The Warfarin/Aspirin Study in Heart failure (WASH): a randomized trial comparing antithrombotic strategies for patients with heart failure. Am Heart J. 2004; 148: 157–64
- Massie B, Krol W, Ammon S, Armstrong P, Cleland J, Collins J. Warfarin and Antiplatelet therapy in Chronic Heart Failure (WATCH), 9 March 2004. Available at: http://www.acc04online.acc.org/
- Schmitz G, Rothe G, Ruf A, Barlage S, Tschope D, Clemetson KJ, et al. European Working Group on Clinical Cell Analysis: consensus protocol for the flow cytometric characterization of platelet function. Thromb Haemost. 1998; 79: 885–96
- Israels SJ, McMillan-Ward EM, Easton J, Robertson C, McNicol A. CD63 associates with the alphaIIb beta3 integrin-CD9 complex on the surface of activated platelets. Thromb Haemost. 2001; 85: 134–41
- Markovitz JH, Kulkarni K, Goldschmidt-Clermont P, Kiefe CI, Rustagi P, Sekar P, et al. Increased platelet activation and fibrinogen in Asian Indians. Eur Heart J. 1998; 19: 720–6
- Negrescu EV, Sazonova LN, Baldenkov GN, Mukharliamov NM, Mazaev AV, Tkachuk VA. Relationship between the inhibition of receptor-induced increase in cytosolic free calcium concentration and the vasodilator effects of nitrates in patients with congestive heart failure. Int J Cardiol. 1990; 26: 175–84
- Pignatelli P, de Biase L, Lenti L, Tocci G, Brunelli A, Cangemi R, et al. Tumor necrosis factor-alpha as trigger of platelet activation in patients with heart failure. Blood. 2005; 106: 1992–4
- Anfossi G, Trovati M. Role of catecholamines in platelet function: pathophysiological and clinical significance. Eur J Clin Invest. 1996; 26: 353–70
- Dole VS, Bergeier W, Mitchell HA, Eichenberger SC, Wagner DD. Activated platelets induce Weibel-Palade-body secretion and leukocyte rolling in vivo: role of P-selectin. Blood. 2005; 106: 2334–9
- Davi G, Gresele P, Violi F, Basili S, Catalano M, Giammarresi C, et al. Diabetes mellitus, hypercholesterolaemia, and hypertension but not vascular disease per se are associated with persistent platelet activation in vivo. Circulation. 1997; 96: 69–75
- Gurbel PA, Kereiakes DJ, Serebruany VL. Soluble P-selectin is not a surrogate marker for platelet P-selectin: evidence from a multicenter chest pain study group. J Thromb Thrombolysis. 2001; 11: 155–62
- Ridker PM, Buring JE, Rifai N. Soluble P-selectin and the risk of future cardiovascular events. Circulation. 2001; 103: 491–5
- Gurbel PA, Gattis WA, Fuzaylov SF, Gaulden L, Hasselblad V, Serebruany VL, et al. Evaluation of platelets in heart failure: Is platelet activity related to etiology, functional class, or clinical outcomes?. Am Heart J. 2002; 143: 1068–75
- Chin BS, Blann AD, Gibbs CR, Chung NAY, Conway DS, Lip GY. Prognostic value of interleukin-6, plasma viscosity, fibrinogen, von Willebrand factor, tissue factor and vascular endothelial growth factor levels in congestive heart failure. Eur J Clin Invest. 2003; 33: 941–8
- Yin WH, Chen JW, Jen HL, Chiang MC, Huang WP, Feng AN, et al. The prognostic value of circulating soluble cell adhesion molecules in patients with chronic congestive heart failure. Eur J Heart Fail. 2003; 5: 507–16
- Gasparyan AY, Watson T, Lip GYH. The role of aspirin in cardiovascular prevention: implications of aspirin resistance. J Am Coll Cardiol. 2008; 51: 1829–43