In this study, we investigated the ability of differences in EAT volume and radiodensity to predict mortality in patients with ESRD undergoing hemodialysis. Our findings indicated that higher EAT radiodensity was associated with a poorer prognosis after adjustment for confounding factors. However, EAT volume did not predict long-term mortality in patients undergoing prevalent hemodialysis. These findings support the notion that EAT has active biologic properties and suggest a potential role of EAT radiodensity as a biomarker for patients with ESRD.
EAT radiodensity is advantageous as a biomarker, as it is relatively simple to measure when compared with perivascular or pericoronary fat volume/density. Moreover, as EAT can be quantified from non-contrast CT 10, there is no risk associated with exposure to contrast medium, which is especially critical for patients with chronic kidney disease. These benefits make determining EAT radiodensity a useful strategy for identifying high-risk patients with ESRD or renal insufficiency. Further studies involving larger cohorts of patients with ESRD are required to validate the applicability of EAT radiodensity on non-contrast CT.
Analyzing CTA imaging sequences may help to elucidate the biological effects of vascular inflammation such as expression of adipogenic genes, average adipocyte size, and lipid accumulation in pericoronary adipose tissue 6. These findings suggest that high radiodensity in perivascular fat reflects prominent tissue inflammation. Considerable evidence suggests that EAT is closely related to coronary atherosclerosis 10,18−20; however, in our study, the severity of CAD was not correlated with EAT radiodensity. As reported in another study 21, it is possible that the degree of CAD was not accurately assessed due to the high calcium burden in patients with ESRD. Indeed, another study reported that calcium blooming artifacts are a major concern when assessing the degree of coronary artery stenosis 22.
However, EAT radiodensity may reflect myocardial states other than coronary artery atherosclerosis. Many studies have reported that EAT abnormalities are associated with atrial fibrillation 3,23,24, while another has noted that arrhythmogenic mechanisms may include adipocyte infiltration into the heart 25. Given that sudden cardiac death is the leading cause of mortality among patients with ESRD 15, additional studies are required to determine whether EAT radiodensity and fatal arrhythmias (e.g., ventricular tachycardia/fibrillation) are related.
It is well known that chronic inflammation is highly prevalent among patients with ESRD due to uremia-induced nutritional and catabolic alterations 26, and the presence of inflammation has been associated with increased mortality risk in these patients 27. Another study reported that inflammation, as determined based on higher CRP and interleukin (IL)-6 levels, was associated with a higher risk of sudden cardiac death among patients with ESRD 28. Our results are consistent with those of a previous study 29 in which the authors reported a significant association between CRP levels and mortality; however, EAT radiodensity was not associated with CRP levels. Therefore, we presumed that EAT radiodensity reflects a more heart-specific inflammatory state rather than systemic inflammation, and that this inflammatory response in the heart may contribute to mortality in patients with ESRD. As there is no fascia or other separating tissue, adipocytes in the epicardium may directly affect the myocardium, exacerbating myocardial lipotoxicity 30. Myocardial lipotoxicity and extra-cardiac adiposity result in increased heart weight and decreased systolic function 31, left ventricular hypertrophy, electrocardiographic abnormalities, and increased arrhythmogenicity 32. We believe that chronic myocardial inflammatory conditions also influence EAT and contribute to mortality in patients with ESRD.
Our findings indicate that, unlike EAT radiodensity, EAT volume is not associated with all-cause mortality in patients undergoing maintenance hemodialysis. A previous study reported that high EAT volume predicts mortality in patients undergoing incident hemodialysis 17. The difference in results is likely due to differences between the study populations. Previous studies have indicated that phosphate binders can increase EAT volume in patients undergoing hemodialysis 33 and that high-phosphate treatments can also affect EAT volume in these patients. Therefore, we presumed that qualitative assessments of EAT based on inflammatory cell deposits and fibrosis are superior to quantitative assessments of EAT such as volume in terms of predicting prognosis in patients with ESRD.
Our study is subject to limitations. First, our analysis was limited by a relatively small sample size, ultimately restricting its statistical power. Nevertheless, the longitudinal analysis of mortality was a strength; hence, conclusions regarding causality can be confidently made. Second, we were unable to assess CVD related outcomes in this study. However, several studies have reported the major cause of death in ESRD is CVD 15,34. Third, despite coronary calcium score being an important predictor of cardiac events, we were unable to assess the extent of coronary calcification due to the CTA protocol of our institution and retrospective study design. Last, we were only able to hypothesize about the pathophysiological relationship between CT-derived EAT radiodensity and all-cause mortality, which needs to be evaluated in dedicated large-scale studies.
In conclusion, Higher EAT radiodensity assessed by coronary CTA was associated with higher long-term all-cause mortality in patients with ESRD. Due to ease of measurement, EAT radiodensity may provide a useful method to noninvasively assess mortality risk in patients undergoing prevalent hemodialysis. Therefore, EAT radiodensity is a promising prognostic imaging biomarker in patients with ESRD.