Our study compared the ability of EAT thickness to predict MACE’s in patients with CAD. We found that EAT thickness significantly associated with MACE’s development; with a higher risk than the other markers tested.
EAT constitutes a type of VAT with intimate relation with the heart and coronary arteries, which owns unique characteristics regarding adipocyte's hypertrophy, biochemical and metabolic activities [22]. EAT is taught to exert pro-atherogenic effects acting as a source of pro-inflammatory mediators in a close proximity to coronary arteries and myocardium. Indeed, the relationship between EAT with coronary atherogenesis, plaque volume and stability has been described, as well as its predicting ability for clinical outcome and MACE’s [11, 23,24,25]. Consistently, in the present study higher values of EAT were observed in the group that developed MACE’s during the follow up period; whereas factors like lower E/A measures, diabetes mellitus and the type of CAD may also be involved. We observed that EAT maintained as independent risk factor for MACE’s after weighing potential interactions by logistic regression analysis.
In addition, EAT significantly related with cardiometabolic risk markers like BMI, LVEF and the degree of complexity of coronary arteries, reflected by the Syntax score; being consistent with previous reports also describing MACE’s predicting ability of such markers [11, 14, 25, 26]. However, the pathophysiological pathways leading cardiovascular risk underlying to each marker may be different. For example, BMI may be linked to low-grade inflammation and the “obesity paradox” risk phenomenon [17, 18]; while LVEF may be associated to myocardial injury, ventricular arrhytmia, pump failure with activation of renin-angiotensin-aldosterone and adrenergic systems leading to negative myocardial remodeling and deterioration in contractile function [15] and coronary severity may influence due to the degree of vascular damage, although the relation between coronary stenosis with plaque stability or future MACE’s is still debatable [16].
Therefore, we specifically compared EAT’s predicting ability of MACE's versus BMI, LVEF and complexity of coronary arteries in a study population with CAD, owning demographic characteristics and cardiovascular risk similar to previous reports [23]. In comparison with the other markers, only EAT thickness resulted significantly higher in patients who developed MACE’s; with an estimated risk for MACE's of almost four. Consistently, higher baseline values of EAT thickness have been predictive of MACE's, and independent association with similar risk values have been obtained in several studies [10, 11, 14, 23, 25]. Moreover, we found that EAT provided acceptable diagnostic performance for MACE’s, which was similar to LVEF and with higher sensitivity than Syntax score and BMI. The AUC in our study is lower than other reports [10]. To our knowledge, this is the first study comparing EAT thickness, BMI, LVEF and complexity of coronary arteries regarding their quantification, MACE’s association and diagnostic abilities.
In the present study, we performed echocardiographic measure of EAT thickness and used the median value to estimate the risk for developing MACE’s. However, there is a high variability in the way of measure EAT and the value considered of risk for MACE’s between studies. Some have used quartiles or 90th percentile of EAT basal measures, or even the EAT remodeling observed between basal measure and a 4-year subsequent measure; obtained through cardiac-CT with measures performed either as parallel, equidistant axial planes or full vertical length images of the heart [10, 25, 27]. Others have used the mean EAT, as echocardiographically determined, or receiver operating characteristics-obtained cutoff value best discriminating MACE’s; from measures of the thickest point at the echo-free space between the pericardium and the outer wall of the myocardium, performed usually at the free wall of the right ventricle from the parasternal long-axis views [11, 14]. Regardless the measurement method, EAT's predicting ability for MACE’s remains highly consistent and risk association values reported are very close to our result.
Moreover, we found that EAT showed a variable discrimination range for MACE's and different prediction ability, related to the location where EAT was measured. This finding has also been previously observed [27]. Although EAT measures on a single slice at specific level of the heart highly correlate with measures of total EAT burden [24], it seems that there is a location-dependent risk, at least observed in the prediction of coronary atherogenesis. According to studies and meta-analysis, EAT thickness measured at the left AV groove is associated with obstructive CAD risk [28, 29]; whereas scanty information exists regarding EAT's location-dependent risk for MACE’s. Currently, there is no guidelines or consensus regarding its clinical use of EAT measure. Some recommendations are based on its relation with MRI measures or with coronary angiography findings; moreover, echocardiographic cutoff values for increased EAT measures has been suggested [5]. However, echocardiographic method of EAT measure that best reflects its MACE’s prediction ability is to be defined.
The strength of our study are both, the performance analysis for MACE’s discrimination when echocardiographic EAT measurements are collected at different anatomical references; and the comparison of MACE’s prediction risk regarding other markers. However, major limitations of the present study includes: first, potential selection bias because patients were recruited from those referred for echocardiographic examination; second, no survival analysis at different time points for MACE’s was performed; and third, the relatively small number of patients and cardiac endpoints.
In conclusion, the ability of EAT thickness in predicting MACE’s was better than the other markers tested in a population with CAD. The influence of selective location of EAT measure over MACE’s discrimination performance deserves further study due to potential usefulness for cardiovascular outcome prediction.