Decreased pre-surgical CD34⁺/CD144⁺ cell number in patients undergoing coronary artery bypass grafting compared to coronary artery disease-free valvular patients

Atherothrombosis is the leading global cause of morbimortality [1]. A growing burden of experimental and clinical evidence highlights the importance of endothelial progenitor cells (EPCs) in the pathogenesis of this disease [2, 3]. EPCs originate in the bone marrow and migrate in response to ischemic stimuli, especially stromal-derived factor (SDF-1) [4]; they differentiate towards the endothelial lineage or support local angiogenesis by resident endothelial cells. In their differentiation process, EPCs possess the early markers CD34 and KDR and acquire further markers of endothelial commitment (such as CD144, also known as VE-cadherin) [2].

Several studies describe a lower number of CD34⁺/KDR⁺ cells in patients with coronary artery disease (CAD); these low numbers seem to be inversely correlated with cardiovascular risk factors [4]; moreover, this EPC decrease has been directly linked to CAD [5] and to precede future cardiovascular events [6]. Lower CD34⁺/CD144⁺ numbers have been described in atherosclerotic patients, as well as the role of statins to differentiate CD34⁺/KDR⁺ to CD34⁺/CD144⁺ [7].

Other reports, however, suggest a higher number of EPCs in atherosclerosis [8]. A population-based study shows a direct correlation between CD34⁺/KDR⁺ and Framingham risk factors [9]. In addition, long-term statin treatment has been related to a reduced number of CD34⁺/KDR⁺ in patients with CAD [10]. This discrepancy may be explained by the possible oscillation of cell number depending on the vascular disease state, which depends on the intensity of the ischemia [11]. Higher apoptosis and lower viability have been observed in cultured EPCs from atherosclerotic patients, especially diabetics [12, 13]. High glucose concentrations impair adhesive properties of EPCs under flow conditions [14]. The recent clinical evidence suggests a role of bone marrow-origin EPC exhaustion in advanced atheroscelrotic cardiovascular disease [15]. However, EPC exhaustion has equally been described in degenerative aortic stenosis and it may mediate endothelial damage in this disorder [16]. Both CABG and valvular replacement surgery dramatically increase the number of EPCs, however these EPCs seem to be functionally impaired after both interventions [17]. The question arises whether CABG with advanced atherosclerotic disease and valvular replacement patients with aortic stenosis and angiographic abscence of coronary lessions possess different EPC levels. Thus, we aimed to assess the number of CD34⁺/CD144⁺ and CD34⁺/KDR⁺ cells in the peripheral blood of patients undergoing coronary artery bypass grafting (CABG) with angiographic demonstration of CAD, compared to valvular surgery patients with angiographic demonstration of absence of CAD. The ultimate goal was to elucidate which vascular disorder involves a deeper alteration of EPC numbers.

Our results show a lower basal number of CD34⁺/CD144⁺ cells in the pre-surgical blood of patients who underwent CABG surgery compared to aortic stenosis valvular replacement patients (Figure 3, panel B). Of note, being CABG or valvular was related to CD34⁺/CD144⁺ cells even after the P value was adjusted to the number of comparisons (Table 4). This suggests that this parameter could be related to angiographic CAD in severe and advanced cardiovascular disease.

The clinical characteristics of our patients merit some comments (Tables 1 and 2); although in these non-primary endpoints the P values were not adjusted to the number of comparisons and thus should be interpreted in an exploratory manner. CABG had a larger amount of atherosclerotic risk factors and related medications (Table 1) and higher numbers of monocytes, fibrinogen and platelets (Table 2), which reproduces what was found in other series [22]. This reinforces the broadly accepted idea that atherosclerosis acts as an inflammatory stimulus, a concept supported by the finding of augmented CRP levels in a subgroup of our CABG patients (3.4 ± 1.67 mg/l), In the intragroup analysis we found that only dyslipidemia had a significant effect on CD34⁺/CD144⁺ numbers in CABG patients (Table 3). Of note, the plasmatic concentration of several cytokines (TGF-β1, SDF-1, IL-6 and TNF-α) in CABG and valvular patients remained unchanged (Table 2). Therefore, these parameters were not specific enough to discriminate between patients with angiographic presence or absence of coronary disease in our study.

The numbers of CD34⁺/CD144⁺ cells in the peripheral blood of cardiac surgery patients (both CABG and valvular) is higher than reported elsewhere for CD34⁺/KDR⁺ [5, 6], and than the one which was obtained by us in a cohort of healthy volunteers [18]. A population-based study shows that CD34⁺/KDR⁺ cells and SDF-1 levels are directly associated with Framingham risk factors, whereas this correlation becomes inverse when CD34⁺/KDR⁺ are compared to vessel injury parameters such as intima-media thickness [9]. Our results show that CD34⁺/KDR⁺ and CD34⁺/CD144⁺ from healthy subjects without atherosclerotic risk factors are clearly decreased compared to ischemic non-atherosclerotic patients (Figure 3). Thus, our data support the idea that EPCs are released in response to ischemia and they reach different levels in CABG and valvular patients. These findings agree with the emerging concept that EPCs are released in response to ischemia, being further depleted in advanced cardiovascular disease [9]. This may help to explain why several studies show that these cells are reduced in cardiovascular disease [4–6], whereas they are physiologically mobilized from the bone marrow in response to ischemia [2, 3]. The idea of EPC exhaustion in atherosclerosis is supported by recent clinical evidence [15]. However, in the present study this decrease is found in a significant manner only for CD34⁺/CD144⁺, not for CD34⁺/KDR⁺ (Figure 3). Given that CD34⁺/CD144⁺ is considered as a more differentiated EPC phenotype [7], it could be postulated that this advanced differentiation may be disrupted in CABG patients due to a longer history of atherosclerotic vascular damage. Lower CD34⁺/KDR⁺ numbers for atherosclerotic populations described elsewhere [6] were not assessed in patients where CAD was angiographically demonstrated and were candidates for surgical CABG. Our data also show a lower number of CD34⁺/CD144⁺ cells despite a higher rate of cardiovascular drug consumption (Table 1). In the case of statins, the particular history of CABG patients (long-term, severe and advanced angiographic CAD) may also explain the lack of CD34⁺/KDR⁺ decrease [10] and subsequent differentiation-related CD34⁺/CD144⁺ increase [7] which has been related to statins.

In addition to CAD, aortic valve stenosis is associated to endothelial damage and decreased EPC number and function has been described in these patients [16]. A proper EPC function may be essential for the favorable outcome of surgical grafts [23]. Both in CABG and in valve replacement surgery, EPCs are dramatically released in response to surgical stress, although they seem to be functionally impaired [17]. Restoration of a proper EPC function in both surgical conditions may prove a promising therapeutic strategy. Nevertheless, no data were available about the comparative basal levels of EPCs in the pre-surgical blood of CABG and valvular replacement patients. In the present study we show that the number of CD34⁺/CD144⁺ cells is decreased in the pre-surgical blood of CABG patients, compared to valvular patients with angiographic absence of CAD. Both cardiac surgery patients had a higher abundance of CD34⁺/CD144⁺ cells compared to healthy controls. This may possess a putative diagnostic and prognostic value.

This is the first study which demonstrates a relative EPC depletion in the pre-surgical blood of patients undergoing CABG compared to patients undergoing valve replacement, where CAD was angiographically demonstrated or ruled out, respectively. Given that EPCs seem to play an important role in cardiac surgery grafting, this finding may help to improve our understanding for the simulation of this reparatory mechanism. The study of the underlying mechanisms of this EPC depletion in CAD will be the subject of further research efforts.