Reduced levels of circulating EPCs have been shown to be independent predictors of atherosclerotic disease progression 37. Indeed, endothelial integrity is a fine balance between endothelial damage and repair 37. Since atherosclerotic risk factors are associated with reduced numbers and function of circulating EPCs, it is possible that progression of atherosclerosis is 'driven' by an impairment of EPC repairing capacity 9.

In addition, disease processes that 'damage' the endothelium per se lead to endothelial cell detachment, resulting in increased levels of circulating endothelial cells (CECs) in the blood 38. CECs are thought to be mature cells that have detached from the intimal monolayer in response to endothelial injury and are a different cell population to EPCs 38. There is increasing evidence to support a relation between endothelial damage/dysfunction and CECs 3940. An inverse relationship has been found between EPCs and CECs, as elevated numbers of CECs have been seen in patients with CAD whereas a reduction in the number of circulating EPCs has been associated with CAD 394041.

An inverse relationship has been found between circulating EPCs and CAD severity, independently of traditional risk factors 42 (Table 2). Indeed, patients with multivessel CAD, had significantly lower EPC counts as compared to those without - for every increase in 10 clusters of CFU-ECs, a patient's likelihood for multivessel CAD declined by 20% 42. In contrast, Guven et al 43 showed that the number of EPCs was increased among patients with significant CAD especially in those requiring coronary intervention, and EPC numbers correlated with the maximum angiographic stenosis severity. The discrepancy between these two studies might be attributable to the degree of ischemia experienced by patients included as more severe ischemia in patients required intervention could be responcible for EPC mobilization.

Measurement of EPCs is of predictive value for cardiovascular outcomes in stable CAD patients. For example, Werner et al measured EPCs positive for the CD34 and KDR in 519 patients with CAD confirmed on coronary angiography, and evaluated the association between baseline levels of EPCs and major adverse cardiac event rates at 12 months 34. After adjustment for age, sex, vascular risk factors and other relevant variables, increased levels of EPCs were associated with a reduced risk of death from cardiovascular causes, a first major cardiovascular event, revascularization and hospitalization. This association was independent of severity of CAD, cardiovascular risk factors and drug therapies known to influence cardiovascular outcomes.

Of interest, EPC number and function is closely associated with coronary endothelial function. In 90 patients with CAD coronary endothelial function was assessed using an intracoronary acetylcholine with quantitative coronary angiography 44. The number of circulating CD133⁺ or CD34⁺/KDR⁺ EPCs and EPC function estimated by proliferation of CFU-ECs inversely correspondent with the degree of endothelial function impairment 44. Multivariate analysis showed that the number of EPCs predicted severe endothelial dysfunction independently of classical cardiovascular risk factors.

Acute coronary syndromes (ACS) are associated with increased levels of inflammatory and haematopoietic cytokines, which in turn can mobilise progenitor cells from the bone marrow 454647. Of the many studies in this patient group, Shintani et al were the first to describe a rapid and significant increase in the CD34⁺ EPC numbers, which reached a maximum at 7 days after the onset of ischaemia in acute myocardial infarction (AMI) 48. Similarly, Massa et al described spontaneous mobilisation and a 5.8-fold increase of CD34⁺ progenitor cells, which peaked about 3 hours after the onset of symptoms, significantly decreased after 7 days, and reached levels comparable with those of healthy subjects within 2 months 46. Other cell subpopulations, such as CD34⁺/CD117⁺, CD34⁺/CXCR4⁺, CD34⁺/CD38⁺ and CD34⁺/CD45⁺ broadly follow the same pattern as that of EPCs 4546. Sufficient EPC numbers, as well as the capacity to differentiate into mature endothelial cells are considered to be essential for myocardial functional recovery and infarct size reduction after AMI 49.

Of note, the type of revascularization (ie, thrombolytic therapy or primary angioplasty) does not seem to affect EPC mobilization 50. However, ischemia per se may be a primary factor for EPC mobilization, given a significant rise of EPC numbers in patients with unstable angina 50. An inflammatory state, as shown by high C-reactive protein levels, is also involved in the modulation of adhesive properties of EPCs in ACS 51.

Biphasic changes of the number of CD34+ cells and EPCs have been observed in patients with heart failure. EPCs are significantly up-regulated in mild heart failure (NYHA class I-II) but their mobilization is severely depressed in patients with advanced heart failure (NYHA III-IV) irrespectively of the origin of the disease 5253. Depletion in EPCs has been correlated with high levels of TNF-α indicating that endothelial precursors arem additional 'victim' of excessive inflammation seen in heart failure 52. These data are in accordance with observations in animal model that statin-induced prevention of left ventricular dysfunction was strongly associated the ability of statins to mobilize EPCs 54.

The proportion of CD34+ cells in hospitalised patients with congestive heart failure increases with the improvement of clinical status and is correlated with BNP levels. Of note, Michowitz et al have recently demonstrated that the CFU-EC numbers, together with age and presence of diabetes were independent predictor of all-cause mortality in 107 patients with congestive heart failure 55.

Restenosis after coronary artery stenting remains a significant problem in the practice of interventional cardiology 56. Prevention of restenosis may be promoted by endothelial regeneration through the administration of growth factors, endothelial cell seeding, vessel reconstruction with autologous endothelial cell/fibrin matrix, and the use of stents designed to attract and capture EPCs 56.

Traditionally, the regeneration of injured endothelium has been believed to be due to the migration and proliferation of neighboring endothelial cells. This hypothesis has been challenged by animal studies, which reveal that transfusion of EPCs led to reduced intimal thickening of the injured arterial wall, as a result of accelerated reendothelialization 57. Stents coated with an integrin-binding cyclic Arg-Gly-Asp peptide appear to limit coronary neointima formation and accelerates endothelialization by attracting EPCs, at least in an animal model 58.

What are the implications for therapy? The MAGIC cell trial demonstrated that G-CSF therapy with intracoronary infusion of peripheral blood stem cells improved cardiac function, and promoted angiogenesis in patients with acute myocardial infarction(AMI) 59. At the same time, the concern over aggravation of restenosis was raised. However, the MAGIC Cell-3-DES trial showed that G-CSF - based stem cell therapy was both feasible and safe with DES implantation, eliminating the risk of restenosis 60.

Nonetheless, G-CSF administration itself has been shown to be associated with enhanced neointimal hyperplasia, possibly by the stimulation of excessive proliferation and the migration of smooth muscle cells, thus promoting re-stenosis of stented arteries 61. Moreover, the possibility remains that EPCs may be independently responsible for stent restenosis, as a strong correlation has been shown between circulating CD34⁺ cells and late luminal loss following coronary angiography 62. In a multivariate regression model, a change in CD34+ cells independently predicts late lumen loss 63. However, CD34 cells may be the ones that aggravated in stent restenosis per se. CD34 cells are not authentic EPCs but myeloid stem cells/progenitors that may differentiate into smooth muscle cells depending on the situation or in the presence of platelet derived growth factor 63.

The role of EPCs in process of post-arterial injury recovery/impairment is pretty complex. The ability of EPCs to home to arterial neointima was initially shown in a experimental minipig model 64. Inoue et al subssequently confirmed that an increase in EPC numbers following bare-metal stent implantation in patients with CAD and the extent of EPC elevation correlated with the risk of stent restenosis 65. In contrast, DESs (eg, sirolimus-eluting stents) are capable of preventing procedure-related EPC mobilization 65.

Rapamycin has been also shown to inhibit proliferation, migration, and differentiation of human EPCs in vitro, suggesting that the targeting of EPCs may be an additional mechanism of their successful prevention of re-stenosis by the employment of DESs 66. Indeed, a reduction of EPC numbers may be one mechanism through which drug-eluting stents actually inhibit restenosis. Of note, cell-cycle inhibitors (sirolimus and paclitaxel) reduce neointimal formation by impeding smooth muscle cells proliferation and migration, as well as cause impairment of the normal process of healing of the injured arterial wall, leading to delayed re-endothelialization 67.

Are EPCs really a negative factor in interventional procedures? In contrast to the data where increased numbers of EPCs lead to excessive neointima hyperplasia, Matsuo et al have demonstrated that circulating EPCs in patients with in-stent restenosis grow less colonies of CFU-ECs and these cells have signs of increased senescence compared to patients with successful arterial healing 68. Increased number of senescent EPCs were an independent predictor of stent restenosis, but these data also suggest that functional characteristics of EPCs - rather than numbers per se - may be critical for efficient re-endothelisation. However, one has to keep in mind that circulating CD34+ cells are not exclusively made up of endothelial precursors but also of hematopoietic stem cells and attraction of CD34+ cells may also involve homing of certain populations of inflammatory leukocytes to sites of vascular damage.

The numbers of circulating EPCs may be significantly affected by cardiac surgery, including coronary artery bypass grafting and valve surgery. For example, Smadja et al demonstrated selective mobilisation of endothelial progenitors after different cardiac operations 69, probably in response to tissue damage and application of cardiovascular bypass. Indeed, the mobilization of EPCs from the bone marrow may be promptly stimulated by cardiopulmonary bypass but the extent of such mobilisation is significantly affected by pre-existing risk factors and the European System for Cardiac Operative Risk Evaluation (EuroSCORE) score. Further research is needed to establish the dynamics of EPC during or following cardiac surgery, and how this affects prognosis 70.

Statin therapy is associated with an increase in the number of circulating EPCs in patients with CAD 20 (Table 3). The increase in EPCs and improvement of their migratory capacity was significant as early as at 1 week after the initiation of treatment with atorvastatin, with a 3-fold increase at 3 to 4 weeks of therapy 20. The mobilization of circulating EPCs, along with their enhanced functional activity may contribute to the beneficial effects of statins in patients with CAD 20. The migration and incorporation of EPCs to the sites of re-endothelialization was found significantly increased after statin administration 2071.

While short-term treatment with statins has been seen to increase both EPC number and function, more long-term treatment with statins may have effects to the contrary. Indeed, there is a biphasic effect of statins on EPC counts. In one study of 144 patients with angiographically documented CAD, the administration of statins for more then 4 weeks resulted in a significant decrease of EPC numbers, as determined by flow cytometry and in vitro culture 72. However, only CFU-ECs were evaluated in this study, and as previously discussed, this special subpopulation of EPCs is probably monocytic in origin, with limited capacity to be incorporated into the microvasculature, but may still promote angiogenesis by the release of proangiogenic factors 73. These observations were further substantiated by Deschaseaux et al 74, who demonstrated that statin-induced depression of EPCs was only attributed to CFU-ECs, whereas the ECFC populations, were preserved or even enhanced in the circulation by long-term statin treatment.

In summary, the effects of statins on EPCs are complex. They help to mobilize early EPCs, which is especially important in acute ischaemic conditions such as AMI, but strong long-term positive effects of statins may be partly explained by augmentation of ECFC, which are believed to be 'true EPCs' responsible for vasculogenesis.

Angiotensin II subtype 1-receptor blockade increases the number of EPCs, an effect which seems to be common to all angiotensin II receptor antagonists 75. VEGF appears to be involved in angiotensin receptor blocker-mediated EPC stimulation 32.

Min et al showed that treatment with ramipril 5 mg daily for 4 weeks in patients with CAD was associated with an approximately 1.5-fold increase in the number of circulating EPCs within 1 week of initiating treatment 33. This trend persisted with increased levels to approximately 2.5-fold throughout the 4-week study period. In addition, increases in the functional activity of EPCs - as assessed by their proliferation, migration, adhesion and in vitro vasculogenesis capacity - was also seen. Broadly similar effects were also seen with enalapril, as shown in a study by Wang et al, whereby patients on enalapril displayed a significant increase in circulating EPCs in response to ischemic stress 76. Enalapril also caused a 6-fold increase in the contribution of bone marrow-derived EPCs to the ischemia-induced neovascularization.

Administration of the PPAR-gamma agonists has been associated with elevation of EPC numbers and improvement of their function 77. For example, rosiglitazone was shown to attenuate the negative effects of C-reactive protein on EPCs and enhance NO-production of EPCs 78. Other available data indicate that positive effects of PPAR-gamma agonists on EPCs may be one of mechanisms on the cardiovascular system 7879. In addition to the effects of PPAR-gamma agonists on EPCs, another PPAR subtype -the PPAR agonists- has recently attracted attention. For example, Han et al showed PPAR agonists modulate CFU-EC promoting vasculogenesis 80. There are also provasculogenic effects of PPAR agonists on ECFCs 81.

Oestrogens augment production and survival of EPCs, thereby increasing the circulating levels of these cells 82. Moreover, oestrogens not only increase EPC count, but also their ability to effectively home to the sites of vascular lesions 82. Oestradiol may enhances EPC mobilization though NO-mediated pathways 83.

Treatment options for CAD include the mechanical dilatation of one or more areas of narrowing within the coronary artery using percutaneous coronary intervention (PCI). The original 'balloon-only' technique was improved by the use of bare-metal stents (BMS). However, BMSs are still prone to restenosis in up to 30% of patients 84. To reduce stent restenosis, the use of DESs which can elute antimitotic medications has become widespread. However, DES use has been associated with late stent thrombosis 8586. Indeed, the risk of stent thrombosis with DES is greater than that seen with BMS due to impaired endothelial healing and delayed endothelialisation 87.

In contrast, endothelialisation with a third type of stent, the CD34 antibody coated stents based on EPC capture technology, usually occurs within 7 days in animal studies 88. Rapid endothelialisation and restoration of endothelial function may have several advantages, including reduced restenosis and reduced stent thrombosis rate. In addition, early cessation of dual antiplatelet therapy may be possible, reducing the inherent risk of this.

Effectiveness of CD34 covered stents has been tested in a series of studies performed under the HEALING (Healthy Endothelial Accelerated Lining Inhibits Neointimal Growth) program. The first of these studies was the Healthy Endothelial Accelerated Lining Inhibits Neointimal Growth-First In Man (HEALING-FIM) Registry which demonstrated feasibility and safety of EPC capture by CD34 covered stents in stable CAD 89. Subsequently, the HEALING II and the HEALING III studies, and ultimately the real-world e-Healing multi-center registry of the Genous™ pro-healing EPC capturing stent (OrbusNeich, Netherlands) have been performed.

The e-Healing registry established that the rates of cardiac death, MI and target lesion revascularization with the CD34 covered stents were low and comparable to the early Taxus registry studies 90. These stents were also tested for use in primary PCI in patients with AMI. Implantation of these stents was associated with relatively low rate of major adverse cardiac events: 1.6% in-hospital, 4.2% at 30 days, 5.8% at 6 months, and 9.2% at 1 year; there was 1 patient each with acute and subacute stent thrombosis but no incidence of late stent thrombosis 91. When used in 'high risk' patients (33% with diabetes, 73% with ACS, 8% with left ventricular dysfunction, 9% with multivessel intervention, 56% type B2/C lesions) 14 month follow-up revealed an event rate of 13% for noncardiac death and AMI, 13% for repeated percutaneous revascularization and 4% for de novo lesions 92. No bypass surgery was preformed. The study showed that the CD34-covered stents were safe and effective, with satisfactory immediate results and mid-term outcome, without evidence of stent thrombosis

In a recent report from the ongoing multi-centre, randomized the Tri-stent adjudication (TRIAS) study CD34-covered stents were as effective as paclitaxel-eluting stents for the prevention of re-stenosis during 12-month follow-up; however, 4 cases of out-of-hospital thrombotic events were reported for paclitaxel-covered stents, whilst no such events were reported in patients with implanted CD34-covered stents 93.