Endothelial progenitor cells: Implications for cardiovascular disease



Endothelial progenitor cells (EPCs) reside in the bone marrow and are mobilized into the circulation by specific stimuli such as certain drugs, ischemia, and exercise training. Once in the circulation EPCs are thought to participate in the maintenance of the endothelial cell layer. Recently it was clearly demonstrated that the amount and function of EPCs is significantly impaired in different cardiovascular diseases. Furthermore, the level of circulating EPCs predicts the occurrence of cardiovascular events and death from cardiovascular causes and may help to identify patients at increased cardiovascular risk. After demonstrating the beneficial effect of applied EPCs in several animal experiments, these cells were also used to treat humans with different cardiovascular diseases. This review will focus on the characterization and liberation of EPCs from the bone marrow, as well as on the most important clinical cardiovascular diseases for which EPCs were used therapeutically. © 2008 International Society for Advancement of Cytometry

Since the first description of endothelial progenitor cells (EPCs) by Asahara et al. in 1997 (1), a huge amount of literature has been published on EPCs in cardiovascular diseases. Adult bone marrow is a rich reservoir of tissue-specific stem and progenitor cells. Among these, a scarce population of cells described as EPCs can be mobilized by various stimuli into the circulation to contribute to the neoangiogenic process or to contribute to the repair of the damaged endothelial cell layer. A point heavily discussed in the current literature is the proper identification of circulating EPCs using either flow cytometry or specified cell culture conditions and what are specific triggers for the liberation of the cells from the bone marrow.

Despite all the uncertainty of proper identification of the cells, EPCs were already used in clinical studies. After demonstrating a beneficial effect of EPCs applied to animals with hind-limb ischemia (2) and myocardial infarction (3), the group of Prof. Strauer in Düsseldorf published the first human intervention study using EPCs in patients after myocardial infarction (4, 5). This review will focus in the first part on the characterization and liberation of EPCs from the bone marrow, and in the second part we will discuss the role of EPCs in different cardiovascular diseases.


The most uncertain point in the discussions around EPCs is their proper characterization and quantification. A summary of the current used methods is shown in Figure 1. Morphological criteria are definitively not enough to clearly identify EPCs. More specific methods such as flow cytometry with proper antibodies or enrichment of EPCs by cell culture methods followed by specific staining has to be applied. In the following sections we will discuss the methods in more detail.

Figure 1.

Schematic drawing of the isolation and characterization of endothelial progenitor cells starting with circulating blood. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Flow Cytometry Analysis

At the moment there is no best marker described to characterize and to quantify circulating EPCs. This point makes it really difficult to standardize and compare the quantification of EPCs among different published studies. In most cases a combination of a hematopoietic stem cell marker, such as CD34, CD117 (cKit), and CD133, with a marker for endothelial cells, such as KDR (vascular endothelial growth factor (VEGF) receptor 2) and Ve-cadherin, is used to identify EPCs [summarized in (6, 7)]. In general, the best approach to characterize and quantify EPCs would be to include as many markers as possible. Nevertheless, this is limited by the low abundance of EPCs in the peripheral blood, which represent between 0.01 and 0.0001% of mononuclear cells (MNCs) depending on several factors such as antibody affinity or state of health of the individual. A solution for all these difficulties is the combination of CD34/KDR, most often used in the current literature. A typical analysis of EPCs from peripheral blood is shown in Figure 2. MNCs are isolated by density gradient centrifugation, followed by the incubation with specific antibodies. Subsequently, the washed and fixed cells are analyzed by flow cytometry. As first step the CD34pos cells are identified by proper gating. In a second step only the CD34pos cells are analyzed for the expression of KDR. In this example, the EPC frequency is around 0.005% from total MNCs. What are proper negative controls to set the gate for cell quantification? Several different negative controls are currently used in the different studies, ranging from isotype specific antibodies labeled with the same colors as the specific antibodies (8–10), addition of no antibodies at all to see the autofluorescence of the cells, to the fluorescence minus one (FMO) strategy [reviewed in (11)]. Another factor complicating the characterization of EPCs is the fact that the expression of the often used stem cell marker CD34 is not only restricted to stem cells, but is also expressed on mature microvascular endothelial cells (12, 13). Therefore, CD133, a more immature hematopoietic stem cell marker that is absent on mature endothelial cells and monocytic cells (14) was introduced as a better marker for EPCs. It was postulated that CD133pos/KDRpos cells more likely reflect immature progenitor cells, whereas CD34pos/KDRpos may also represent shedded cells of the vessel wall. In addition, it is suggested by several authors that more immature EPCs in the bone marrow are CD133pos/CD34pos/KDRpos/Ve-cadherinneg, whereas circulating EPCs lose the early marker CD133 and start to express more endothelial cell marker [reviewed in (6)]. Therefore they are defined as CD133neg/CD34pos/KDRpos/Ve-cadherinpos.

Figure 2.

A representative example for quantification of EPCs is depicted. Isolated MNCs are incubated with isotype control antibodies (negative control) or with specific antibodies (FITC-CD34 and PE-KDR). In a first evaluation step the amount of CD34pos cells is determined (R1). In a second step, KDR-expressing cells are determined in R1 (histogram).

In conclusion, the proper identification of EPCs is still uncertain, but for practical reasons the definition CD34pos/KDRpos is most often used. Nevertheless, more basic research is necessary to clearly define EPCs and distinguish them from other adult stem cells.

Cell Culture

Since the first report of Asahara et al. in 1997, classical isolation methods include adherence culture of total peripheral MNCs (1) and the use of magnetic beads coated with anti-CD34 (1), anti-CD133 (15), or anti-CD14 (16–18) antibodies. After the isolation, the cells are plated on cell culture dishes coated with either fibronectin (19–21) or gelatin (8, 9, 22). Already these variations in the coating strategy may have an influence on the results obtained by cell culture with respect to cell number, cell phenotype, and cell function. Another important factor regulating the fate of the plated cells is the cell culture media ranging from EBM-2 (23–25), medium-199 (22, 26, 27), to x-vivo-20 (28) as basal media, and 20% fetal calf serum (1), brain extract (1, 26), to commercial available single quots of VEGF, insulin-like growth factor (IGF), basic fibroblast growth factor (23, 25) as supplements. An identification criteria often used for EPCs in the culture of MNCs for 4 or 7 days is the double-positive staining of EPCs for endothelial-specific lectin (e.g., Ulex europaeus agglutini-1) and Dil-labeled acetylated low-density lipoproteins (Ac-LDL) (26, 29, 30). An example of such a staining is shown in Figure 3. At least two types of EPCs have been described using these classical culture methods. “Early” EPCs appear within 4–7 days of culture, are spindle-shaped, and express both endothelial (von Willebrand factor) and monocytic (CD14) markers (31). “Late” EPCs develop after 2–3 weeks of culture and have the characteristics of endothelial lineage-like cobblestone pattern (1, 8) and the expression of endothelial nitric oxide synthase (8). However, the origin of the double-positive EPCs is still under debate. It could be demonstrated by Gulati et al. that the vast majority of EPCs arose from a CD14pos subpopulation of peripheral blood cells, whereas late outgrowth endothelial cells develop exclusively from the CD14neg fraction (32). Nevertheless, with respect to the characteristics and functional properties Urbich et al. reported that adherent cells derived from CD14pos or CD14neg MNCs show equal expression of endothelial marker proteins and capacity for clonal expansion (17). Another approach to culture EPCs is a preplating procedure. This procedure claims to avoid contamination for early-adherent cells such as differentiated monocytic or possible mature endothelial cells (33). After replating nonattached cells onto fibronectin in endothelial specific medium, typical clusters of round cells centrally with spindle-shaped cells sprouting at the periphery appear. Some but not all of the cells inside the colony-forming unit are indeed double positive for Ac-LDL and lectin staining (27). Is there a possibility to tune EPCs in culture, so that their angiogenic potential is increased? In a recent study Hu et al. investigated the effect of low-dose aspirin on the proliferation and the functional capacity of EPCs in cell culture (34). These experiments clearly demonstrated that aspirin up to a concentration of 10–100 μmol/L increased the migratory capacity of the cells without influencing the proliferation. Beside the activation of EPC migration, specific activation of the main thrombin receptor PAR-1 by the receptor-activating peptide SFLLRN led to concentration-dependent increase in late-EPC proliferation (35).

Figure 3.

A representative image of EPCs after 7 days in culture and stained with DilacLDL (red) and FITC-lectin (green). Cells only positive for FITC-lectin are marked with 1, cells only positive for the uptake of Dil-acLDL are marked with 2, and EPCs positive for both markers are marked with 3. Nuclei are counterstained with Hoechst (blue). For cell culture, cells were isolated from peripheral blood using Ficoll density centrifugation, plated on gelatin-coated cell culture dishes, and cultured in EGM-2 media for 7 days. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Functional Analysis

Beside the concentration of circulating EPCs, their functional capacity is also of great importance. Currently, there are at least three assays described in the literature: (1) migration assay evaluating the capacity of EPCs to migrate along a chemokine gradient, (2) matrigel tube formation assay evaluating the capacity of EPCs to incorporate into a network of mature endothelial cells, and (3) hindlimb ischemia (HLI) model investigating the capacity of EPCs to restore blood flow in vivo.

Migration assay

To perform a migration assay, EPCs are isolated in most cases by cell culture techniques as described earlier. The Boyden chamber assay, originally introduced by Boyden for the analysis of leukocyte chemotaxis, is based on a chamber of two medium-filled compartments separated by a microporous membrane. In general, cells are placed in the upper compartment and are allowed to migrate through the pores of the membrane into the lower compartment, in which chemotactic agents are present. After an appropriate incubation time, the membrane between the two compartments is fixed and stained by either Hoechst or Di-LDL, and the number of cells that have migrated to the lower side of the membrane is determined. For the functional analysis of EPCs chemotactic agents such as VEGF (21, 24), SDF-1 (36–38) or adiponectin (39) are often used. Using this migration assay it could clearly be demonstrated that statins increase the migratory capacity of circulating EPCs (21) and that this function shows an inverse relation with cardiovascular risk factors (24).

Matrigel tube formation assay

In line with the idea that EPCs participate in angiogenesis, it is reasonable to test this ability in vitro. One of the most specific test for angiogenesis is the measurement of endothelial cells or there progenitors to form three-dimensional structures (tube formation) [reviewed in (40, 41)]. With the discovery that Matrigel [a matrix-rich product prepared from Engelbreth–Holm–Swarm (EHS) tumor cells whose primary component is laminin (42, 43)] can evoke endothelial cell tube formation within 24 h, tube formation assays have achieved a prominent place in the array of angiogenesis measurements (44, 45). This assay is currently also often used to analyze the capacity of EPCs to participate in angiogenesis. To perform such an analysis, EPCs are isolated from MNCs by cell culture techniques, labeled with DiLDL or FITC-lectin, and coplated with human coronary artery endothelial cells. For evaluation, the proportion of EPCs in tubules is determined (20, 38, 44) (see Fig. 4). One word of caution, however, is that cultured cells of nonendothelial origin, such as fibroblasts, may also exhibit a response to matrigel (46).

Figure 4.

Assessment of functional capacity by Matrigel assay. FITC-lectin-labeled EPCs (green, marked with arrows) were cocultured with Dil-acLDL human coronary artery endothelial cells (red) for 24 h on a matrigel matrix. Cell nuclei are stained with Hoechst (blue). For quantitative evaluation the amount of EPCs (green) incorporated into the tubes are counted. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

HLI model

The murine HLI model is frequently used for studying the angiogenic potential of injected cells. Therefore, after operative resection of one of the femoral artery in athymic mice, in which angiogenesis is characteristically impaired (47), labeled cells, for example, EPCs, are injected. To follow the angiogenetic potential of the injected cells over time, laser Doppler perfusion imaging is used to measure blood flow. The beauty of this system is that the in vivo potential of the injected cells can be investigated [reviewed in (48, 49)]. Using this system, in the year 2000, Kalka et al. were able to demonstrate by histological assessment that injected EPCs incorporate into sites of neovascularization (50). Other than in ischemic tissue and very rarely in the spleen, EPCs were detected neither in other organs nor in the contralateral nonischemic hindlimb. The peak incorporation was achieved within 3–7 days postadministration and the administration of EPCs led to a reduced limb loss and increased limb salvage. These investigations using the HLI model were of great importance, since they demonstrated for the first time that EPCs also participate in vivo in neovascularization.


It is believed that the majority of EPCs originate from the bone marrow. Stem cells within the bone marrow usually exist in a quiescent state and specific signals stimulate the stem cells to differentiate and to be mobilized into the systemic circulation. Beside humoral factors such as cytokines (51–53), hormones (51, 53, 54), chemokines (37, 55), and drugs (56–58), exercise training (ET) has been shown to increase the number of circulating EPCs (8, 9, 59, 60).

Humoral Factors

Enforced mobilization of hematopoietic stem cell from the bone marrow has been observed in all species investigated so far, suggesting that similar mechanisms are involved. It appears that during mobilization the bone marrow becomes a playground of a complex interplay between cytokines/chemokines, potent proteases, and adhesion molecules. Although the understanding of the molecular pathways leading to a mobilization of EPCs from the bone marrow is not fully clear, several studies in the current literature demonstrate that VEGF is one of the most potent molecules triggering an EPC release [reviewed in (61)]. VEGF expression is dramatically upregulated by hypoxia, which represents a critical force that drives adult vasculogenesis. Mechanistically it is thought that VEGF activates matrix metalloproteinase-9 (MMP-9) which cleaves membrane bound kit ligand to release soluble kit ligand (also known as stem cell factor) (62). This then stimulates cKitpos stem cells to migrate from a quiescent bone marrow niche to the vascular zone, thereby translocating the cells into a proliferative state (63). The fact that the activation of MMP-9 is a necessary intermediate in the mobilization process is supported by the observation that cytokine-induced progenitor cell mobilization was markedly impaired in MMP-9−/−mice (62). The central role of MMP-9 in the mobilization of EPCs is also supported by studies in eNOS−/− mice (64). Aicher et al. showed that mice deficient in endothelial nitric oxide synthase (eNOS−/−) exhibit reduced VEGF-induced mobilization of EPCs and increased mortality after myelosuppression. Mechanistically, MMP-9 was significantly reduced in the bone marrow of these mice, thereby reducing the mobilization of EPCs into the circulation. Besides VEGF, other humoral factors have been described to mobilize EPCs from the bone marrow. For example Li et al. showed in a murine bone marrow transplantation model that placental growth factor increased tumor vasculogenesis by enhancing EPC recruitment (65). Erythropoietin (Epo) is another factor influencing EPC mobilization. In an animal study, Heeschen et al. demonstrated that the subcutaneous injection of human recombinant Epo resulted in an increased number of stem and progenitor cells in the bone marrow and elevated numbers of circulating EPCs (66). In addition, in a multivariate regression model, serum levels of Epo were significantly associated with the number of circulating EPCs (66). A third factor implicated as a chemokine for EPCs is the stromal cell-derived factor-1α (SDF-1α). SDF-1α has the potency to induce the expression of VEGF (67), and the local delivery enhances neovascularization of an ischemic hindlimb (68). Mechanistically, it is proposed that SDF-1α stimulated the mobilization of EPCs via an enhancement of protein kinase B (Akt) and endothelial nitric oxide synthase activity (68).

Physical Exercise

In addition to pharmacologic therapies, ET as a nonpharmacologic intervention may have the potency to increase the number and function of circulating EPCs. This concept is based on the findings that exercise intervention programs are able to improve endothelial function in patients with coronary artery disease and chronic heart failure (69, 70), and that EPCs are thought to participate in the repair of damaged endothelial cell layers (7, 71). In a first study our group investigated the impact of a single exercise bout on the mobilization of EPCs from the bone marrow (8). Using flow cytometry and cell culture methods to quantify the number of circulating EPCs, a significant time-dependent increase was noted in patients with symptomatic coronary artery disease after a single episode of exercise-induced ischemia. This was confirmed by Rehman et al. as well as by van Craenenbroeck et al. (72). Both studies showed that a symptom-limited treadmill or bicycle exercise test could acutely increase the amount of EPCs. Possible mechanisms for this fast increase in EPCs (10 min to 2 h) may be an exercise-induced increase of VEGF (8, 72) or exercise-induced changes in shear stress, resulting in mobilization of the cells (73). Since all these studies investigated the effect of a single exercise bout on the mobilization of EPCs, one has to ask the question if ET over a longer time period also can influence the number and function of EPCs. The first study addressing this question was a study by Laufs et al. (59). In their study they investigated in an animal model as well as in humans with stable coronary artery disease the impact of training on EPCs. They clearly showed that in the animals as well as in the human subjects an ET intervention over a time period of 3 or 4 weeks significantly increased the number of circulating EPCs. As possible mechanisms they identified either increased VEGF levels or nitric oxide (NO). The importance of NO for the exercise-induced release of EPCs was supported by the analysis of eNOS knockout mice. These results could be confirmed by several other groups investigating patients with coronary artery disease (74), peripheral vascular occlusive disease (9), chronic heart failure (75), hemodialysis (76), or even health controls (77, 78). Even in school children, a recent study from our group confirmed the impact of exercise on EPCs (79). In summary, all published studies demonstrated so far that exercise and an ET program in diseased or healthy individuals has the capacity to induce various stimuli to increase the release of EPCs from the bone marrow. Can we do too much exercise, so that we decrease the amount of EPCs? One form of a strenuous exercise is performing a marathon race. Analyzing healthy individuals directly after the race, and even some days later, revealed that circulating progenitor cells (CPCs), defined as CD34pos or CD133pos, significantly decreased (80, 81), whereas the amount of EPCs was unchanged (81).


For over 10 years, bone marrow-derived EPCs have been studied as a novel biomarker to assess the severity of cardiovascular diseases, and as a potential new strategy in regenerative medicine. Cell-based therapy to stimulate postnatal vasculogenesis or to repair vascular integrity is being evaluated for cardiovascular diseases with excess morbidity and mortality. In the following sections, several diseases for which EPCs are already used as therapeutic approach are discussed.

Acute Myocardial Infarction

The intrinsic regenerative capacity of the heart is limited. The damaged and diseased cardiomyocytes are removed and replaced by scare tissue. Therefore, myocardial infarction can lead to chronic heart failure, a condition with high 5-year mortality. This occurs despite substantial improvement in interventional and pharmacological therapy of acute myocardial infarctions. The introduction of new medical therapies such as ACE-inhibitors let to a significant reduction in postmyocardial infarction mortality (risk reduction 23% vs. Placebo) (82). Also the implementation of an interventional revascularization procedure to standard therapy let to an additional reduction in mortality in those patients (83). The new discoveries on the regenerative potential of stem and progenitor cells resulted in a variety of preclinical and clinical studies which examined the use of these cells for regeneration of the heart and restoration of cardiac function.

Preclinical studies

In a mouse model of acute myocardial infarction the application of c-kit positive bone marrow cells into the infarct border zone led to a 68% myocardial regeneration in this area (84). However, the transdifferentiation of these bone marrow cells into cardiomyocytes is discussed controversially. Whereas Orlic et al. reported a cell differentiation with loss of c-kit-surface receptors, expression of connexin 43, expression of transcription factor MEF 2 as markers for maturing myocytes and marker for electrical coupling (84), others did not find markers for transdifferentiation into cardiac myocytes (85). They showed that bone marrow cells could fuse with host cardiomyocytes, therefore expressing both sets of markers (85, 86).

A functional improvement of left ventricular (LV) ejection fraction was reported by Kocher et al. (87) after the intravenous administration of bone marrow-derived CD34pos cells into mice 48 h after LAD ligation. Furthermore, a reduction of myocardial remodeling/scar due to enhanced neoangiogenesis was reported (87). The same results, only with lower doses of CD34pos EPCs (5–10%), were seen after local transplantation of the cells into infarcted rats (3). To explain the beneficial effects on myocardial function, several theories are discussed. Neovascularization of the peri-infarct-zone through direct or paracrine means are evident (3, 87). Also a reduction in ischemia and an improvement of perfusion, leading to a better regional contractility were shown (87, 88). Another mechanism to explain the effects on LV function is the regeneration of lost/dead myocardial tissue by the transplanted cells, but as discussed earlier the transdifferentiation of cells into myocytes is questionable. Nevertheless, several studies showed that this regeneration led to increased contractility and improvement of cardiac function (84, 89–92).

Another important point for the improvement of cardiac function is the paracrine secretion (e.g., VEGF, SDF 1α, IGF 1) from the transplanted cells. The secreted factors are beneficial for myocardial survival by preservation of ischemia, enhanced neovascularization (87), reduction of apoptosis (87, 88), and modulating the immune response to the infarct injury. Many preclinical studies also used several factors to increase EPC mobilization as an alternative strategy for cell therapy avoiding the need for cell isolation/preparation. VEGF (51), G-CSF (93), stem cell factor, and erythropoietin (94) have been shown to increase circulating EPCs in preclinical models leading to myocardial neovascularization and improved myocardial function in artificial myocardial infarction.

Another cell type showing promising preclinical results are very small embryonic-like stem cells, which are discussed in more detail by Zuba-Surma et al. in this issue of the journal (95). The results of the preclinical trials initiated a variety of clinical studies for restoring myocardial function in patients suffering from acute myocardial infarction.

Clinical studies

Until now 14 randomized trials (Table 1) with a total of 811 patients have been reported. It is very difficult to compare these trials directly, since they used various cell types, different manufacturing and therapy-protocols as well as different endpoints. All patients received PCI with stent implantation for reperfusion of the infracted area and the time of cell administration was for all patients within 7 days of symptom onset. Sample sizes of the trial were relatively small ranging from 20 to 204 participants. In 12 trials direct bone marrow aspiration was used, whereas in 2 trials G-CSF administration with subsequent cell-isolation from the blood was done. Follow-up data are available for up to 12 months, meaning that long-term follow-up data are lacking. The different trials had a wide heterogeneity of endpoints ranging from mortality endpoint (five studies) to morbidity with reinfarction rate (seven studies), restenosis (seven studies), revascularization (five studies), and readmission, arrhythmias, adverse events, quality of life, and reoperation. A recent metaanalysis done by Martin-Rendon showed favorable effects in the BMC-treated groups regarding LV systolic volume reduction, LV end diastolic volume reduction, reduction in myocardial infarction area, and an moderate improvement of the LV ejection fraction (109).

Table 1. Summary of studies using EPCs for treatment of acute myocardial infarction
Ruan et al. (96)9/1153.4/53.5BMC aspirationDiluted serumWithin 7 days6 monthsMyocardial contractility
Karpov et al. (97)10/10n.a.BMC aspiration?>7 days6 monthsSafety
Ge et al. (98)10/1053.8/58.2BMC aspirationn.a.Within 7 days6 monthsLV function
Huang et al. (99)20/2044.5/43.7BMC aspirationSaline heparinizedWithin 7 days6 monthsLV function
Kang et al. (100)25/2552.0/53.2BMC after G-CSF injectionNone>7 days6 monthsLV function
Lunde et al., ASTAMI (101)47/5054.8/53.6BMC aspiration (0.68 × 108 MNC)Plasma heparinizedWithin 7 days6 monthsLV function
Meyer et al., BOOST (102)30/3050.0/51.3BMC aspiration (2.46 × 108 MNC)Plasma heparinizedWithin 7 days18 monthsLV function
Janssens et al. (103)33/3448.5/46.9BMC aspirationSaline + 5% serumWithin 7 days4 monthsLV ejection fraction
Schächinger et al., repair-AMI (104)101/10348.3/46.9BMC aspiration (2.36 × 108 MNC)Medium + 20% serumWithin 5 days4 monthsLV function
Penicka et al. (105)14/1039.0/39.0BMC aspiration (26.4 × 108 MNC)n.a.Within 7 days4 monthsLV function
Meluzin et al. (106)20/20//2041.0/42.0BMC aspiration (1.0 × 107 MNC or 1.0 × 108 MNC)Cell suspensions mediaWithin 7 days12 monthsLV function
Suare de Lezo et al. (107)10/1037.0/39.0BMC aspiration (9 × 108 MNC)Saline heparinizedWithin 7 days3 monthsLV function
Li et al. (108)35/2350.0/51.0BMC after G-CSF injectionn.a.>7 days6 monthsLV function, adverse events

Although the observed improvement in LV function is moderate with a mean of 2.99% improvement compared with the controls, it may be possible that this moderate improvement in LV function leads to an improved long-term follow-up. Another important point still lacking is if the application of cells leads to an improvement in clinical endpoints and quality of life. Another way for stem cell mobilization from the bone marrow is the administration of granulocyte-colony-stimulating factor (G-CSF). Few studies with a limited number of patients suffering from myocardial infarction have evaluated the use of G-CSF as stimulus for myocardial regeneration. In an early safety study this therapy was associated with an increased risk for coronary restenosis in patients after myocardial infarction. The effects of this therapy on LV function and target vessel revascularization are summarized in a recently published metaanalysis by Zohlnhöfer et al. They found no effect of an additional G-CSF therapy after myocardial infarction with regard to LV-ejection fraction. Also restenosis or target vessel revascularization was not influenced by this therapy (110).

Chronic Myocardial Infarction/Coronary Total Occlusion

Despite the promising results in patient with acute myocardial infarction, some authors focused on EPC therapy in patients with chronic myocardial infarction or chronic total occlusion of a the coronary artery. These patients often suffered from chronic heart failure based on a reduced LV function. So far, only a limited number of small trials were performed for the aforementioned indication. In experimental models of myocardial infarction (LAD ligation model) the transplantation of stem cells of different origins resulted in an improvement of ventricular function, reduction of scar size, myocardial apoptosis rates, and an increased neovascularization (111). Liu et al. showed an increase in capillary density associated with an improvement of ventricular function and a reduction in ventricular size (112). Functional data were collected by serial myocardial perfusion imaging with SPECT. Three months after stem cell transplantation into the underperfused myocardial segments, the authors found an enhanced perfusion in the patients receiving cells, whereas the control group worsened (113). These effects could be enhanced by preincubation of the stem cells with cardiomyogenic growth factors leading to a cardiomyogenic specification. Using these cells in an infarction model an improved functional recovery was evident when compared with the transplantation of unmodified stem cells (114). Also for patients with old (chronic) myocardial infarctions this cell therapy seemed to be a ray of hope. The first feasibility study was reported in six patients undergoing cardiac surgery. Stem cells were injected into the infarct border zone (115). Strauer et al. reported a study in chronic myocardial infarction patients with transplantation of stem cells 5 months to 8.5 years after the index infarction into the infracted coronary vessel. As in animal models, they found a reduction in infarct size, an increase in LV function, and wall motion velocity in the treatment group. No change was obvious in the nonrandomized control group (116). Only few other studies with a small number of patients are further reported for stem cell therapy in this indication. Schots et al. demonstrated using labeled stem cells that these cells are capable of homing (117), and a small randomized study showed that this therapy is effective in improving the movement of the former infracted region. A larger randomized study by Erbs et al. (118) dealt with chronic total occlusions (>3 months) after myocardial infarction. This study showed after successful recanalization of the occluded vessel and afterwards randomization into a stem cell therapy group or a control group, a reduction in hibernating myocardial segments by 31% for the stem cell group whereas no change was found in the control group. Furthermore, global LV-ejection fraction was significantly increased by 7.2% at the stem cell group after 3 months compared with no change in the control group. A possible pathophysiological explanation for the beneficial effect of the stem cell treatment may be the resulted increase in coronary flow reserve by 43% after 3 months compared with the placebo group (118).

EPC capturing stents

Circulating EPCs have been identified as a key factor for endothelialization (1). After vascular injury like percutane transluminal coronary angioplasty (PTCA) and stent implantation, the development of an early functional endothelial cell layer has been identified as a factor for the prevention of neointimal proliferation. Therefore, EPC-capture stents, covered with CD34 antibodies, were developed for coronary therapy. In a first human study the application of this coronary stent system was safe and feasible for de novo coronary artery disease (119). Recently, a study of intermediate and long-term follow-up results in 129 patients with acute myocardial infarction receiving an EPC-capturing stent was published. The authors found a low cumulative major cardiac event rate (4.2% at 30 days, 5.8% at 6 months). So this might be another promising therapeutic option for intrinsic stem cell therapy.

Dilated Cardiomyopathy

First promising results for stem cell therapy in nonischemic dilated cardiomyopathy (DCM) were reported by Nagaya et al. (120). The aim of the study was to investigate whether transplanted mesenchymal stem cells (MSCs) induce myogenesis and angiogenesis in a rat model of DCM. And indeed, MSC transplantation was associated with significantly increased capillary density and decreased collagen volume fraction in the myocardium, resulting in an improved cardiac function. The first-in-man study was performed by Seith et al. (121). Forty-four patients with an ejection fraction ≤35%, normal coronary arteries, and New York Heart Association (NYHA) functional class II or more were randomly assigned to a stem cell therapy arm (24 patients) or a control arm (20 patients). Six months after autologous transplantation of bone marrow-derived MNCs (BM-MNCs), LV ejection fraction showed a small but significant improvement of 5.4%. This was associated with an augmentation in NYHA functional class in the treatment arm (at least one functional class in 16 patients, compared with only two patients improving in control arm). Endomyocardial biopsies, performed at the beginning and 3 months after transplantation, showed an insignificant increase in the ratio of capillaries to myocytes. Similar data were shown in the Düsseldorf autologous bone marrow cells in DCM trial (122). In this nonrandomized trial the intracoronary transplantation of BM-MNCs led to an increase in LV ejection fraction from 17% ± 1% up to 26% ± 3%. Moreover, this was associated by improved exercise capacity and measured by maximal workload and maximal oxygen uptake.

Peripheral Arterial Occlusive Disease

Peripheral arterial occlusive disease (PAOD) of the lower extremities is a common syndrome that becomes more prevalent worldwide (123, 124). Despite improvements in revascularization by either endovascular approaches or open surgery, approximately one-fourth remain unsuitable for or fail endovascular or operative therapy. However, pharmacological treatment alone has shown poor or no efficacy until now (125). Therefore, the field of cell-based therapy has expanded also in patients with PAOD. Kalka et al. (50) administered ex vivo expanded human EPCs to athymic nude mice with HLI. Histological evidence of the cell incorporation into sites of neovascularization was associated with an increase in capillary density and an enhancement of limb blood flow recovery in mice receiving human EPCs. Moreover, the rate of limb loss was significantly reduced. Murohara et al. reported similar effects after transplantation of human cord blood-derived EPCs (26). In another HLI model, the intramuscular injection of BM-MNCs after 24 h of culture under hypoxia increased microvessel density and blood flow rate in the ischemic muscle compared with controls (126). These promising results from experimental studies promoted the initiation of clinical trials. One of the first was the Therapeutic Angiogenesis by Cell Transplantation (TACT) study by Tateishi-Yuyama et al. (127), including a pilot study followed by a controlled and randomized study. In the pilot trial, 25 patients with unilateral ischemia of the leg were injected with BM-MNCs into the gastrocnemius of the ischemic limb and with saline into the less ischemic limb. Ankle-brachial pressure index (ABI) and transcutaneous oxygen pressure (TcO2) in legs implanted with BM-MNCs were significantly improved after 4 and 24 weeks. Moreover, pain-free walking time increased by about 3 min. The following randomized study was performed in 22 patients with bilateral leg ischemia, who were randomly injected with BM-MNCs in one leg and peripheral blood-MNCs in the other as a control. A significant improvement in ABI (0.09 [0.06 to 0.11]; P < 0.0001), TcO2 (13 mm Hg [9 to 17]; P < 0.0001), rest pain (−0.85 [−1.6 to −0.12]; P = 0.025), and pain-free walking time (1.2 min [0.7 to 1.7]; P < 0.0001) was seen at 4 weeks in legs treated with BM-MNCs. These results remained stable until the 24-week follow-up. Two other studies, using a similar approach in seven and eight patients with CLI, confirmed the safety and feasibility of intramuscular BM-MNC transplantation in PAOD (128, 129). Moreover, Higashi et al. assessed an improvement in endothelium-dependent vasodilatation after BM-MNC implantation in these patients with limb ischemia. Since endothelial dysfunction is the initial step in pathogenesis of atherosclerosis (130), restoration of endothelial function may prevent or at least delay the progression of atherosclerosis. We investigated the safety and potentially beneficial effects of an intraarterial application of autologous CPCs in patients with infrapopliteal PAOD and CLI. Therefore, seven patients with critical PAOD were treated with an intraarterial infusion of autologous CPCs isolated from peripheral blood after G-CSF stimulation. Twelve weeks after CPC administration, the pain-free walking distance increased from 6 ± 13 to 195 ± 196 m (P = 0.016). A significant increase in ABI, TcO2, flow-dependent vasodilation, flow reserve in response to adenosine, and endothelium-dependent vasodilation was observed. We concluded that in patients with CLI without surgical or interventional options, CPC application is safe and feasible and may improve both functional and clinical indices (10). Beside cell transplantation ET has been shown to improve regional perfusion in ischemic syndromes. Thus, we were interested in the effects of ET and ischemia on mobilization and functional activation of blood-derived progenitor cells in patients with PAOD. And indeed, ET led to a significant and time-dependent 5.2-fold increase in the fraction of CD34pos/KDRpos CPCs. Moreover, the function of CPCs analyzed by Matrigel assay was significantly enhanced after 4 weeks. These aforementioned changes in number and function of CPCs were associated with an improvement in symptom-free and maximal walking distances (9).


The exact characterization of EPCs is still a matter of considerable debate, and so far, most studies used a combination of CD34 and KDR to characterize this cell population. Maybe a clearer definition like CD133neg/CD34pos/KDRpos/VE-Cadherinpos or others would help to standardize future studies. A lot of positive effects in conjunction to EPC liberation in different conditions was described. And there is no doubt that liberation of these cells from the bone-marrow is caused by different substances and conditions like cytokines, hormones, drugs like statins and ET. Beside an influence on the amount of the cells, an impact of certain substances on the functional activity of the cells was reported. At this time a lot of small/ middle clinical trials dealing with different clinical conditions were performed to investigate the effect of EPC liberation/ injections on different clinical conditions/diseases. The greatest population treated with autologous stem cells were patients after acute myocardial infarctions. In this population a small but positive effect on myocardial regeneration compared with standard therapy was described. It seems to be that patients with the worst LV ejection fraction have the highest profit from this intervention. Based on the fact that this is a young field of investigation there are no long term data available and regarding the relatively small study populations there are no data on the improvement of clinical endpoints. But not only patients with acute myocardial infarctions, also those with old (chronic) myocardial infarctions and chronic total occlusions seem to have a benefit from EPC injection after revascularization, probably because of an improvement of endothelial function. A summary of diseases which may be suitable for the treatment with EPCs or other stem cells is depicted in Figure 5.

Figure 5.

Schematic drawing of conditions influencing the liberation of EPCs from the bone marrow, and its therapeutical application in various cardiovascular diseases. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

A promising field of investigation is the application of EPC capturing stents for coronary lesions. Here we have a device where endothelialization is more rapid and, therefore, the risk of subacute/late stent thrombosis may be much smaller. Nevertheless, for a wider clinical use of EPCs, larger studies are needed to define the patient population having the best benefit from treatment and to document an impact on mortality.