Atrial fibrillation (af) is the most frequent arrhythmia in the adult ages and related with the one third of the hospitalizations for rhythm disorders [1]. Morbidity and mortality of the patients with atrial fibrillation increase significantly according to the patients without af [2]. Stroke is the most important cause of morbidity at these patients. Non-valvular af (nvaf) increases the risk of stroke 5-folds and valvular af (vaf) increases the risk 17-folds [3-5]. In order to decrease the morbidity and mortality due to the atrial fibrillation, medical and electrical treatments are heavily used [6-8].

Hemodynamic changes seen in the structural heart diseases, neurohumoral activation, cellular interactions and other factors cause atrial tension and interstitial fibrosis [9]. These changes affect the electrical and structural features of the atrial tissue that plays an important role in the development of af.

As a result of atrial fibrillation, endothelial impairment and prothrombic endothelial changes develop at atriums [10]. Many structural changes that are defined in the atrial tissue of patients with atrial fibrillation are related with the changes which develop in the signal transducing pathways at the cell level [11] like st2/il-33 (suppression of tumorigenicity 2/interleukin 33) pathway’s role on myocardium remodeling [12].

A group of cellular surface molecule that maintains the adhesive interactions between the endothelial cells and leucocytes are defined. It is shown that these molecules that are called adhesion molecules are involved in the regulation of events such as histogenesis, embryogenesis, cell growth, cell differentiation and inflammation [13]. Some adhesion molecules that increase in the blood stream can contribute to intravascular thrombus development by increasing the cell collection and adhesion [11]. Intercellular adhesion molecule-1 (icam-1) which is an adhesion molecule from the immunoglobulin super family is expressed particularly from endothelial cells, lymphocytes, smooth muscle cells, monocytes and macrophages [14]. Another adhesion molecule in this group is called vascular cell adhesion molecule-1 (vcam-1) and it is expressed especially depending on endothelial cells, antigen presenting cells, stromal cells of bone marrow, embryonic tissue and synovial tissue [15]. Endocardial adhesion molecules (icam-1 and vcam-1) can play a role in inflammation and prothrombic mechanisms or relationship between these two roles. In many studies, vcam-1 and icam-1 have been claimed to be responsible for the atrial thrombus, valvular and myocardial damage [10,16-20].

In patients with valvular atrial fibrillation (vaf), the risk of developing of atrial thrombus and thromboembolic event is much higher than in the subjects with nvaf. We consider that this situation can be related with the adhesion molecules. In this study, we aimed to compare the levels of icam-1 and vcam-1 in patients with vaf, nvaf and normal sinus rhythm.

Patient group

We used 45 subjects with normal sinus rhythm as the control group and 44 af patients whose coronary arteries were found to be normal between January 2008 and May 2009 as the experiment group. 27 of this group were with nvaf and 17 of them were with vaf.

Demographic features, major cardiovascular risk factors (hypertension, hyperlipidemia and diabetes) and laboratory findings were all recorded.

As a result of the echocardiographic tests, patients with stenosis or failure symptoms due to the rheumatic valvular disease or atrial fibrillation and patients who were administered surgical valvular replacement/repair due to these causes were defined as vaf. The remaining cases of atrial fibrillation outside of them were considered as nvaf. In order to detect the svcam-1 and sicam-1 levels, venous blood samples were obtained and all samples, at the same time, were stored at -80C. Serum vcam-1 and icam-1 levels were measured by using the elisa technique with ray bio human sicam-1 elisa kit for icam and ray bio human svcam-1 elisa kit for vcam.

After all patients had been informed about the study, approval forms were signed and taken together with the approval of the ethical committee.

Continuous variables were defined as mean ± standard deviation and quantitative variables were defined as number and percentage. Appropriateness of the variables for normal distribution was evaluated with “shapiro-wilk” test. At binary comparisons, appropriateness of continuous variables for normal distribution was compared with “t test” and “mann-whitney u test”. Continuous variables in the multiple groups (>2) were compared with one way variance analysis (anova). Values which were statistically and significantly different between the groups in the one way variance analysis were compared with multiple comparison methods such as multiple comparison procedures and tukey hsd (tukey’s honestly significant difference test). Qualitative variables in all of the groups (gender, diabetes, hypertension, smoking etc.) were compared with chi-square test. For all of the statistical analyses, “spss for windows version 12” software was used. Calculated p value<0.05 was accepted as statistically significant.

Total 89 patients of which 44 with af and 45 with normal sinus rhythm at the electrocardiography as control patients who do not have coronary artery disease were included in the study. Demographic features of the groups were shown in Table 1. While both groups were similiar in terms of gender (male in the control group n=16 (35.6%), male in the atrial fibrillation group=15 (%34.1), p=0.19), there was a statistically significant difference in terms of the average ages (51 ± 8 vs. 62 ± 10, p<0.001) between af and the control group though the average ages were similar in vaf and the control group. There wasn’t any statistically significant difference between the groups in terms of diabetes, hypertension, hyperlipidemia, and family history. Biochemical parameters bun, creatinine, total cholesterol and ldl levels were higher in the af group, while hdl and triglyceride levels were higher in the control group. Any statistically significant difference between the groups was not detected in terms of crp levels (Table 1). When the control group and vaf group were compared, any statistically significant difference wasn’t detected in terms of age, bun, total cholesterol and ldl levels (p>0.5) (Table 2). Vcam-1 and icam-1 distribution within the atrial fibrillation and control groups are shown at Figure 1. There wasn’t any difference detected in scope of icam-1 between the groups. Vcam-1 level was higher in the atrial fibrillation group when compared with the control group (1169 ± 246 mg/dl, 1072 ± 229 mg/dl, p=0.056) (Figure 1).

Table 1: Continuous variables are shown as mean ± standard deviation; quantitative variables are shown as numbers and percantages. NVAF: nonvalvular atrial fibrillation, VAF: valvular atrial fibrillation, BUN: blood urea nitrogen, LDL: Low-density lipoprotein, HDL: High-density lipoprotein, LVEF: left ventricular ejection fraction, VCAM: vascular cell adhesion molecule, ICAM: intercellular adhesion molecule.

Table 2: VAF: valvular atrial fibrillation, BUN: blood urea nitrogen, LDL: Low density lipoprotein, VCAM: vascular cell adhesion molecule, ICAM: intercellular adhesion molecule.

Figure 1: Distribution of VCAM-1 levels (1072 ± 229 mg /dl vs 1169 ± 246 mg / dl, p=0.056) and ICAM-1 levels at both groups (671 ± 132 mg / dl vs 706 ± 180 mg / dl, p=0.313). AF: Atrial fibrillation, ICAM: intercellular adhesion molecule, VCAM: vascular cell adhesion molecule.

In the atrial fibrillation group, there were 17 (39%) vaf and 27 (61%) nvaf patients. In the valvular atrial fibrillation group, there was history of valvular surgery at 8 (47%) patients of which 6 for mitral valvulus and 2 for mitral valve replacement (mvr)+aortic valve replacement (avr). A statistically significant difference was detected between the valvular af, non-valvular af and control groups in terms of age and left ventricular ef values. Any difference between the valvular af, nonvalvular af and control groups in terms of icam-1 value (748 ± 203 mg/ dl vs 680 ± 163 mg/dl vs 671 ± 132, p=0.216), however, there was a statistically significant difference between the groups in term of vcam- 1 levels (1260 ± 291 mg/dl vs 1113 ± 198 mg/dl vs 1072 ± 229 mg/dl, p=0.021) (Figure 2). Vcam-1 level, which is higher (1260 ± 291 mg/dl vs. 1072 ± 229 mg/dl, p=0.016) in the vaf group than the control group is found to be the primary reason for the difference mentioned between the af and control groups in the analyses conducted. In multivariate cox regression analysis, high vcam-1 levels and high age correlation was significant (p=0.01) (Table 3).

Table 3: Multivariate cox regresyon analysis, VCAM-1 cut off value >1200 mg/dl.

Figure 2: Distribution of VCAM-1 levels (1260 ± 291 vs 1113 ± 198 vs 1072 ± 229, p=0.021) and ICAM-1 levels at study groups (671 ± 132 vs 680 ± 163 vs 748 ± 203, p=0.216). NVAF: non-valvular atrial fibrillation, VAF: valvular atrial fibrillation, VCAM: vascular cell adhesion molecule.

In this study, we aimed to compare the icam-1 and vcam-1 levels at vaf and nvaf patients with healthy control group. Vcam-1 levels were found to be statistically significantly higher especially in patients with vaf according to the nvaf patients and control group.

Af causes significant increase in morbidity, mortality and health costs due to hemodynamic disorders, stroke and other thromboembolic complications. Overall in af, decrease in blood flow, increase in the activation of coagulation system and endothelial changes play a role in the development of thrombus. Adhesion molecules help cells to adhere both to each other and to the extracellular matrix. Some adhesion molecules that increase in the blood flow can contribute to development of intravascular thrombus by increasing the collection of cells and adhesion [11].

Pathophysiological importance of adhesion molecules in atrial fibrillation was supported by the study conducted by kamiyama. In this trial; after administering 8 hours of rapid atrial pace in the left atrium of rabbit at the endocardium increase in adhesion molecules with leucocyte adhesion was shown [17]. These data support the claim that adhesion molecules can trigger development of thrombus by achieving the adhesion of trombocytes, lecocytes and monocyte to the surface of endocardium [18]. Goette et al. Showed that vcam-1 expression increases both in paroxysmal and in persistant af [10]. Hammwöhner et al. suggested that the levels of plasma vcam-1 and icam-1 levels increase in patients with af and vcam-1 level is an independent risk factor atrial thrombus [21]. On the other hand, chen et al. Reported that there aren’t any differences between the patients with lone af and control group in terms of vcam-1 and icam-1 levels [22]. Canbaz et al. in a study in which they evaluated the postoperative af development, didn’t detect any relationship between the af development and vcam-1 levels [23]. Compliant with these data, in our study vcam-1 levels of patients with nvaf weren’t different from those of the control subjects.

Levels of vcam-1, icam-1 and von willebrand factor were reported to be significantly higher in the patients with rheumatic heart disease (rhd) when compared with the patients without rhd [20]. Particularly in rhd cases with af, a significant increase at icam-1 levels was observed when compared with the rhd patients with normal sinus rhythm [20]. In a study conducted by yetkin et al. plasma vcam-1, icam-1 and e-selectin levels were shown to increase significantly in patients with rheumatic mitral stenosis when compared with the control group [24]. Müller et al. showed that vcam expression increases not only in inflamed cuspis but alsoin degenerative cuspis in which there is not any finding of inflammation [25]. Similarly, shahi et al. found in a study that vcam-1, icam-1 and e-selectin levels are increased in nonrheumatic aort stenosis and they suggested that non-rheumatic aort stenosis can have an inflammatory or immunological etiology [26]. Zhang et al. showed that plasma vcam-1 and icam-1 levels are higher in the patients with rheumatic heart disease when compared with the control group and for the reason of this issue they suggested that this can be one of the pathological mechanisms of myocardial damage, endothelial dysfunction and cuspis damage in patients with rheumatic heart disease [20].

In our study, we could not find any difference between the groups in terms of icam-1 level. However, we observed that vcam-1 levels are significantly higher in the vaf patients when compared with the control group subjects that don’t have rhd with sinus rhythm. Although there was not any atrial thrombus in the patient group of this study, high vcam-1 levels can be related with increased risk of atrial thrombus and stroke.

Prothrombotic endothelial changes (endocardial remodeling) develop at the fibrillated atriums. Cai et al. in animal studies, reported that, during af expression of no, synthetase decreases leading to decrease in no production contributing to the thrombus development by affecting the release of adhesion molecules [27]. There are studies that suggest that, during af oxidative, stress via nADPH oxidase increases and stimulates the production of adhesion molecules [28,29]. Vcam-1 levels which are one of the adhesion molecules that were evaluated in this study were found to be high. Multivariate regression analysis showed that high age and high vcam-1 level correlation is significant. Af is an age related disorder so that this result is not a surprise.

One of the most important limitations of this study is the lack of prospective follow-up for the thromboembolic complications of af. Another limitation is, no doubt, the low number of patients included in the study. Larger studies with a number of patients will be able to better explain the role of these molecules in atrial fibrillation. Levels of adhesion molecules are affected by multiple factors such as smoking, hypertension, age, hyperlipidemia, low level of hdl and medications [30-34]. Thus, it is very difficult to achieve a full homogenization in scope of these factors. Also, in our study the potential effects of these factors on the adhesion molecules cannot be ruled out.

While icam-1 levels are alike in patients with valvular atrial fibrillation and patients with sinus rhythm, vcam-1 levels are higher at patients with valvular atrial fibrillation. We suggest that high vcam-1 levels can be a risk for the development of thromboembolic events in this group.