Konstantinos Stellos, Medizinische Klinik III, Kardiologie und Kreislauferkrankungen, Eberhard-Karls Universität Tübingen, Otfried-Müller Str. 10, 72076 Tübingen, Germany. Tel.: +49 17696543101; fax: +49 7071295749. E-mail: firstname.lastname@example.org
Summary. Aims: Blood cell infiltration and inflammation are involved in atrial remodelling during atrial fibrillation (AF) although the exact mechanisms of inflammatory cell recruitment remain poorly understood. Platelet-bound stromal cell-derived factor-1 (SDF-1) is increased in cases of ischemic myocardium and regulates recruitment of CXCR4+ cells on the vascular wall. Whether platelet-bound SDF-1 expression is differentially influenced by non-valvular paroxysmal or permanent atrial fibrillation (AF) in patients with stable angina pectoris (SAP) or acute coronary syndrome (ACS) has not been reported so far. Methods and results: A total of 1291 consecutive patients with coronary artery disease (CAD) undergoing coronary angiography were recruited. Among the patients with SAP, platelet-bound-SDF-1 is increased in patients with paroxysmal AF compared with SR or to persistent/permanent AF (P < 0.05 for both). Platelet-bound SDF-1 correlated with plasma SDF-1 (r = 0.488, P = 0.013) in patients with AF and ACS, which was more pronounced among patients with persistent AF (r = 0.842, P = 0.009). Plasma SDF-1 was increased in persistent/permanent AF compared with SR. Patients with ACS presented with enhanced platelet-bound-SDF-1 compared with SAP. Interestingly, among patients with ACS, patients with paroxysmal or persistent/permanent AF presented with an impaired platelet-bound SDF-1 expression compared with patients with SR. Conclusions: Differential expression of platelet-bound and plasma SDF-1 was observed in patients with AF compared with SR which may be involved in progenitor cell mobilization and inflammatory cell recruitment in patients with AF and ischemic heart disease. Further in vivo studies are required to elucidate the role of SDF-1 in atrial remodeling and the atrial fibrillation course.
The natural history of atrial fibrillation (AF), the most common arrhythmogenic disorder, is characterized by a gradual worsening with time involving three stages: (i) electrical remodeling which develops within the first days of AF; (ii) a reduction of atrial contractility during AF which enhances atrial dilatation and hence may contribute to the persistence of AF; and (iii) tachycardia-induced structural remodeling which also occurs as a result of heart failure and other underlying cardiovascular diseases . Atrial remodeling is a multifactorial inflammatory process involving myocyte hypertrophy, apoptosis, necrosis, alterations in the composition of the extracellular matrix, changes in the expression of ionic channels and atrial hormones and inflammatory cell infiltration [2,3]. Inflammatory markers such as C-reactive protein (CRP), tumor necrosis factor, interleukins and cytokines have been shown to be elevated in AF [4,5]. Human AF causes local cardiac platelet activation within minutes of onset, as shown by increased P-selectin expression .
The chemokine stromal cell-derived factor-1 (SDF-1; CXCL12) regulates many vital functions of mature blood cells and progenitor cells including chemotaxis, migration, transmigration, adhesion, activation and differentiation. We have recently shown that activated platelets express on their surface and subsequently secrete the chemokine SDF-1, thereby supporting further primary adhesion on the vascular wall and differentiation of progenitor cells into endothelial progenitor cells supporting angiogenesis and vascular regeneration . Moreover, our group has previously reported that platelet-bound SDF-1 is increased in cases of acute coronary syndrome (ACS) compared with patients with stable coronary artery disease (CAD) and positively correlates with the number of circulating CD34+ progenitor cells . Whether platelet-bound SDF-1 expression is differentially regulated by non-valvular paroxysmal or persistent/permanent atrial fibrillation (AF) in patients with stable CAD or ACS has not been reported so far.
The aim of the present study was to evaluate the expression of platelet-bound SDF-1 in patients with non-valvular paroxysmal and persistent/permanent AF in a large cohort of patients with coronary artery disease.
Patients and methods
A total of 1291 consecutive patients with CAD undergoing coronary intervention were recruited in a pre-planned time period. In all, 714 patients with suspected or known coronary artery disease with typical symptoms for stable angina pectoris (SAP) were referred to our hospital for coronary angiography according to the ACC/AHA guidelines for coronary angiography . Patients with SAP had either typical angina on exertion and/or a pathological exercise test and were negative for markers of myocardial ischemia (troponin I and creatine kinase). In all, 577 patients presented in our emergency room with acute coronary syndrome, as previously defined [10,11], and immediately underwent percutaneous coronary intervention (PCI). In the means of a subgroup analysis, patients were further classified as patients with paroxysmal non-valvular AF (parox. AF, defined as self-terminating, usually within 48 h) or persistent/permanent non-valvular AF (pers. AF, defined as an episode lasting longer than 7 days) or with sinus rhythm (SR). Patients with parox. AF presented with AF at the time of hospital admission, when blood sampling was performed. Arterial blood was drawn from the sheath that was introduced into the femoral artery at the beginning of coronary intervention and after administration of 2500 U of unfractionated heparin. Arterial blood was filled into 5-mL vials containing citrate phosphate dextrose adenine (CPDA) for flow cytometry and EDTA for ELISA. CPDA probes were analyzed by flow cytometry according to standard methods. Patients with an acute myocardial infarction were preloaded with aspirin and clopidogrel before cardiac catheterization. In case glycoprotein (GP) IIb/IIIa antagonist therapy was needed, administration was performed after blood sampling and therefore could not influence our experimental results. The study was approved by the institutional ethical committee and all subjects gave their written informed consent.
Whole-blood flow cytometry
Platelets obtained from 1291 consecutive patients were studied for surface expression of SDF-1 by flow cytometric analysis as previously described . Conjugated monoclonal antibodies were used to measure platelet SDF-1 surface expression (clone 79014, fluorescein isothiocynate-FITC; R&D Systems, Minneapolis, MN, USA), with a two-color flow cytometry in patients’ whole blood as previously described. In brief, 10 μL CPDA blood was resuspended 50 : 1 with phosphate buffer saline (PBS; Invitrogen Corporation, Paisley, Scotland, UK) and was incubated for 30 min with the relevant conjugated antibodies in the dark at room temperature. After staining, the cells were fixed with 0.5% paraformaldehyde and stored at 4 °C until fluorescence activated-cell sorting (FACS) was performed with a FACS-Calibur flow cytometer (Becton-Dickinson, Heidelberg, Germany). CD42b-PE served as a control antibody to identify the platelet population in the whole blood. Specific monoclonal antibody binding was expressed as mean fluorescence intensity (MFI) and was used as a quantitative measurement of platelet proteins’ surface expression.
Plasma levels of SDF-1 were determined in a subgroup of 585 consecutive patients with symptomatic coronary artery disease at the time of platelet determination using a commercially available ELISA kit according to the manufacturer’s instructions (R&D Systems). EDTA plasma vials were centrifuged for 15 min at 10 000 g within 30 min of collection. Probes were aliquotted and stored at −20 °C before analysis. The lower detection limit of this assay is 18 pg mL−1. The mean centered coefficient of variation for soluble SDF-1 was 3.2%, thus allowing a relatively good reproducibility of our measurements.
Data presentation and statistical analysis
Data are presented as mean ± standard deviation (SD). Continuous variables were tested for normal distribution with the Kolmogorov–Smirnov test. Student’s t-test and anova analysis followed by Scheffé post hoc analysis were used to assess differences between two or three groups, respectively. Comparison of categorical variables was generated using Pearson’s chi-square test. Correlations were assessed with Pearson’s correlation coefficient test. A univariate analysis of variance was applied to determine the influence of possible cofactors in SDF-1 expression in different patient groups. All tests were two-tailed and statistical significance was considered for P-values < 0.05. All statistical analyzes were performed using SPSS version 19 for windows (SPSS Inc., Chicago, IL, USA).
The patients’ characteristics are shown on Tables 1 and 2. Among the patients with SAP (n = 714), compared with SR (n = 631), significantly elevated platelet-bound SDF-1 expression was detected in patients with paroxysmal AF (n = 43; P < 0.001) and persistent/permanent AF (n = 40; P = 0.047) (SR vs. parox. AF vs. pers. AF: MFI ± SD: 21.15 ± 16.49 vs. 34.72 ± 35.35 vs. 24.66 ± 21.90; P < 0.001; Fig. 1A). Performing a univariate analysis of variance for platelet-bound SDF-1 in paroxysmal AF vs. SR, we observed that specific medication (AT1-receptor blockers, beta-blockers and aspirin) and gender, but not cardiovascular risk factors, age or left ventricular function, influenced the increased platelet-bound SDF-1 expression in paroxysmal AF (Table 3). A subsequent univariate analysis of variance evaluating the interaction of the above significant factors with the patients’ group showed that the platelet-bound SDF-1 increase in paroxysmal AF was associated with gender and beta blocker therapy (Table 3). On the other hand, plasma levels of SDF-1 were increased in persistent (n = 22; P = 0.026), but not paroxysmal, non-valvular AF (n = 18; P = 0.795) compared with SR (n = 285) (SR vs. paroxysmal AF vs. persistent AF: mean ± SD: 2018.02 ± 433.83 vs. 1945.66 ± 359.89 vs. 2282.87 ± 561.64 pg mL−1; P = 0.018; Fig. 1B). Performing a univariate analysis of variance for plasma SDF-1 in persistent AF vs. SR, we observed that the increase of plasma SDF-1 in persistent AF compared with SR was independent of medication use, cardiovascular risk factors, gender, left ventricular function and age (Table 4).
Table 1. Patients’ baseline characteristics (SAP)
Total (n = 714)
SR (n = 631)
AF (n = 83)
Parox. AF (n = 43)
Pers. AF (n = 40)
P-value, SR vs. paroxysmal AF vs. persistent. SR, sinus rhythm; AF, atrial fibrillation; SAP, stable angina pectoris; CAD, for coronary artery disease; LVEF, left ventricular ejection fraction; ACE, angiotensin-converting enzyme.
Age, years (mean ± SD)
68.7 ± 9.9
68.1 ± 10
73.4 ± 7.9
71.5 ± 8.3
75.3 ± 7
Gender, female, n (%)
Cardiovascular risk factors, n (%)
Family history of CAD
LVEF, n (%)
Normal (> 55%)
Slightly reduced (45–55%)
Low (< 35%)
Premedication, n (%)
Vitamin K antagonist
Table 2. Patients’ baseline characteristics (ACS)
Total (n = 577)
SR (n = 514)
AF (n = 63)
Parox. AF (n = 42)
Pers. AF (n = 21)
P-value, SR vs. paroxysmal AF vs. persistent. SR, sinus rhythm; AF, atrial fibrillation; ACS, acute coronary syndrome; CAD, coronary artery disease; LVEF, left ventricular ejection fraction; ACE, angiotensin-converting enzyme.
Age, years (mean ± SD)
68.4 ± 12.7
67.7 ± 12.9
73.5 ± 9.7
72.9 ± 10.5
74.7 ± 7.8
Gender, female, n (%)
Cardiovascular risk factors, n (%)
Family history of CAD
LVEF, n (%)
Normal (> 55%)
Slightly reduced (45–55%)
Low (< 35%)
Premedication, n (%)
Vitamin K antagonist
Table 3. Univariate analysis of variance for platelet SDF-1 in paroxysmal AF vs. SR in patients with SAP
Among patients with ACS (n = 577), there was no significant difference regarding platelet-bound SDF-1 among SR, parox. AF and pers. AF (P = 0.325; Fig. 1C). On the other hand, increased plasma SDF-1 was observed in patients with persistent AF compared with patients with SR (P = 0.026) and compared with patients with non-valvular paroxysmal AF (P = 0.036; SR vs. paroxysmal AF vs. persistent AF: mean ± SD: 2028.90 ± 440.68 vs. 1967.01 ± 524.39 vs. 2475.08 ± 718.20 pg mL−1; Fig. 1D).
Among patients with AF and SAP a significant correlation was observed between CRP and the expression of platelet-bound SDF-1 (r = 0.399, P = 0.002).
Examining the possible associations between these two factors in patients with AF, we observed that platelet-bound SDF-1 correlated well with plasma SDF-1 among patients with AF and ACS (r = 0.488, P = 0.013), but not with SAP. The correlation coefficient was even higher among patients with persistent AF (r = 0.842, P = 0.009).
Last but not least, we examined whether AF influences SDF-1 expression in patients with ACS. Indeed, although platelet-bound SDF-1 expression was increased in ACS among patients with SR (P = 0.008), patients with paroxysmal AF presented with significantly decreased SDF-1 expression in ACS compared with SAP (P = 0.017), whereas patients with persistent AF did not show any difference between SAP and ACS (P = 0.562; Fig. 1E).
The major findings of the present study are: (i) paroxysmal AF, but not persistent/permanent AF is associated with an increased platelet-bound SDF-1 expression in patients with SAP; (ii) plasma SDF-1 is only increased in patients with persistent/permanent AF compared with SR, independent of clinical presentation of CAD; (iii) among patients with AF and SAP, CRP is positively associated with platelet-bound SDF-1; (iv) platelet-bound SDF-1 correlated positively with plasma SDF-1 in patients with AF, especially with persistent AF and ACS; and (v) patients with AF present with a deficient SDF-1 enhancement after a myocardial infarction which may influence myocardial regeneration.
Sustained AF in goats leads to predominantly structural changes in the atrial myocytes similar to those seen in ventricular myocytes from chronic hibernating myocardium, which explain the depressed contractile function of atrial myocardium after cardioversion and provide a link between these two diseases . Thus, both AF itself and the underlying heart disease may be associated with the development of the arrhythmogenic substrate of atrial remodeling . Whether the most potent chemokine of progenitor cells and leukocytes, SDF-1, is associated with atrial fibrillation is poorly studied so far. In the present study, we recruited 1291 consecutive patients with symptomatic CAD undergoing PCI and analyzed the association of non-valvular atrial fibrillation on SDF-1 expression. In the present study we have shown that in patients with stable CAD, platelet-bound SDF-1 is associated with an increased expression in patients with paroxysmal AF, but not in persistent/permanent AF, compared with sinus rhythm. This finding is in accordance with previous studies reporting that paroxysmal AF activates platelets and enhances platelet aggregation [14,15]. On the other hand, in patients with ACS, paroxysmal AF did not cause any further SDF-1 expression on the surface of platelets compared with SR. However, it should be taken into account as a limitation to the present study that our findings are based on a large number of subgroup analyzes, with different numbers of subjects present in each group.
Although platelet-bound SDF-1 is increased in patients with ACS compared with SAP , in the present study we show that this effect of ACS was lost in patients with paroxysmal or persistent/permanent AF as a result of possible temporal increased expression and subsequent secretion of platelet-derived SDF-1 resulting in an increase of the plasmatic form of SDF-1. The lack of an increase of platelet-bound SDF-1 surface expression in ACS may be explained by the fact of atrial fibrillation-induced chronic platelet activation [14,16,17] and subsequent secretion of platelet-bound SDF-1 in plasma, as previously reported in vitro and in vivo [7,18]. This deficient platelet response in patients with AF and ACS may influence mobilization and further vascular and myocardial regeneration, as it has been recently shown that platelet-bound SDF-1 expression correlates with the number of circulating progenitor cells in patients with ACS and mediates adhesion and differentiation of progenitor cells to endothelial progenitor cells [8,19]. Regarding plasma SDF-1, the present results are in accordance with a previous study reporting that plasma SDF-1 is only increased in patients with persistent/permanent AF .
To answer the question whether or not SDF-1 is actively involved in atrial remodeling/regeneration, further experimental in vivo studies are required to investigate the exact role of platelet-bound and plasma SDF-1 in atrial electrical and structural remodeling. However, reviewing the literature, we could identify functions of SDF-1 that if not explain, at least support our hypothesis: (i) an increase in spontaneous Ca2+ release in patients with AF probably as a result of an upregulation of the sarcoplasmic reticulum Ca2+ release channel activity, contributes to the development of AF . SDF-1 is reported to be involved in cardiomyocyte calcium homeostasis regulation in rat hearts . (ii) In ischemic heart disease, where platelets play a crucial role on its genesis and progression, atrial remodeling is accelerated , indicating a possible link between platelets and paroxysmal AF. To support this hypothesis, another recent study has shown that pre-operative platelet activation, as assessed by sCD40L levels, is a novel predictor of postoperative AF, independent of systemic endothelial function, vascular redox state and systemic inflammation . Platelet-bound SDF-1 expression is increased upon platelet activation  and therefore it may be involved in this process. However, it should be taken into account that the present study population comprised patients with concomitant CAD, and therefore our findings should be interpreted carefully in patients with AF in the absence of evidence of ischemic heart disease (systolic heart failure, arterial hypertension and lone AF). Further studies are needed to evaluate the effect of AF in patients with AF in the absence of evidence of ischemic heart disease. (iii) Patients with persistent AF present with increased plasma SDF-1 levels and CD34+ progenitor cell number compared with SR, which both correlate with atrial natriuretic peptide . In the present study, we show a correlation between platelet-bound and plasma SDF-1 in patients with AF, and especially persistent AF and ACS; (iv) Considering the effect of SDF-1 on hematopoietic progenitor cell mobilization and differentiation, it is remarkable that CD34+ hematopoietic progenitor cells are increased in the blood of patients with persistent atrial fibrillation . (v) CRP correlates with the left atrium volume, suggesting a relationship between inflammation and atrial remodeling . In the present study, we observed a significant association between CRP and platelet-bound SDF-1 in patients with AF and ACS. (vi) SDF-1 is crucially involved in mechanisms responsible for myocardial hypertrophy [27,28], which is a risk factor for atrial fibrillation incidence. Thus, further studies are needed to address the role of myocardial hypertrophy in patients with atrial fibrillation.
New strategies for the prevention and termination of AF should be built on our knowledge of the pathophysiological mechanisms of atrial remodeling . Further, in vivo studies are required to elucidate the role of SDF-1 in atrial remodeling and the atrial fibrillation course.
All authors have significantly contributed to the manuscript: K. Stellos, B. Bigalke and H.J. Weig conceived the hypothesis and study design, K. Stellos, A. Rahmann and K. Sopova conducted the statistical analyzes and wrote the first draft of the manuscript. A. Rahmann, M. Ruf, A. Kilias and R. Jorbenadze were responsible for the recruitment of patients S. Weretka, K. Stamatelopoulos, T. Geisler and M. Gawaz provided critical input at all stages. All authors critically reviewed and contributed to the manuscript.
The study was supported by the Deutsche Forschungsgemeinschaft (DFG) Sonderforschungsbereich/Transregio19 ‘Inflammatory Cardiomyopathy– molecular pathogenesis and therapy’ to K.S. and M.G., by the Fortune Project of the Eberhard-Karls- University of Tübingen, the Hirnliga e.V. and the ‘Nachwuchswissenschaftler 2011′ programme of the Goethe-University Frankfurt to K.S.
Disclosure of Conflict of Interest
The authors declare that none of them pertain a relationship with pharmaceutical companies, biomedical device manufacturers or other corporations whose products or services are related to the subject matter of the article. The authors state that they have no conflict of interest.