Abstract
Background:
Epicardial adipose tissue represents a metabolically active visceral fat depot that is in direct contact with the left ventricular myocardium. Whilst it is associated with coronary artery disease, little is known regarding its role in aortic stenosis. We sought to investigate the association of epicardial adipose tissue with aortic stenosis severity and progression, myocardial remodelling and function, and mortality in asymptomatic patients with aortic stenosis.
Methods:
In a post-hoc analysis of 124 patients with asymptomatic mild to severe aortic stenosis participating in a prospective clinical trial, baseline epicardial adipose tissue was quantified on computed tomography (CT) angiography using fully automated deep learning-enabled software. Aortic stenosis disease severity was assessed at baseline and 1 year. The primary endpoint was all-cause mortality.
Results:
Neither epicardial adipose tissue volume nor attenuation correlated with aortic stenosis severity or subsequent disease progression as assessed by echocardiography or CT (p>0.05 for all). Epicardial adipose tissue volume correlated with plasma cardiac troponin concentration (r=0.23, p=0.009), left ventricular mass (r=0.46, p<0.001), ejection fraction (r=−0.28, p=0.002), global longitudinal strain (r=0.28, p=0.017), and left atrial volume (r=0.39, p<0.001). During the median follow-up of 48 [interquartile interval 26–73] months, a total of 23 (18%) died. In multivariable analysis, both epicardial adipose tissue volume (hazard ratio 1.82, 95% confidence interval 1.10–3.03; p=0.021) and plasma cardiac troponin concentration (hazard ratio 1.47, 95% confidence interval 1.13–1.90; p=0.004) were associated with all-cause mortality, after adjustment for age, body mass index and left ventricular ejection fraction. Patients with epicardial adipose tissue volume >90 mm3 had 3- to 4-times higher risk of death (adjusted hazard ratio 3.74, 95% confidence interval 1.08–12.96; p=0.037).
Conclusions:
Epicardial adipose tissue volume does not associate with aortic stenosis severity or its progression but does correlate with blood and imaging biomarkers of impaired myocardial health. The latter may explain the association of epicardial adipose tissue volume with an increased risk of all-cause mortality in patients with asymptomatic aortic stenosis.
Keywords: epicardial adipose tissue, aortic stenosis, computed tomography angiography, all-cause mortality
Graphical Abstract
INTRODUCTION
Aortic stenosis has become a major healthcare burden that continues to rise with improvements in life expectancy and aging population [1,2]. It is characterized by progressive valve narrowing, but also by left ventricular hypertrophic remodelling in response to increasing afterload. This hypertrophic response is initially adaptive in restoring wall stress and cardiac performance, but ultimately proves maladaptive resulting in left ventricular decompensation, symptom development and adverse cardiac events [3,4]. Unfortunately, we lack any form of medical therapy capable of favourably influencing events in either the valve or myocardium, although ensuring the timely referral of patients for aortic valve replacement is essential to prevent adverse clinical events [2]. However, decision on timing of valvular intervention remains difficult, as we lack reliable markers that predict disease progression, identify left ventricular decompensation or predict long-term mortality. There is therefore a pressing need to further explore the mechanisms driving aortic stenosis with respect to both the valve and myocardium so that we can improve patient risk stratification and the timing of valve replacement as well as identifying novel treatment targets.
Epicardial adipose tissue has recently attracted major interest and represents a metabolically active visceral fat depot that is enclosed by the pericardium and directly surrounds the myocardium [5]. Epicardial fat and the myocardium share the same blood supply and are anatomically and functionally contiguous. This close proximity, allows the epicardial fat to directly interact with the myocardium with several adipokines, including profibrotic and proinflammatory cytokines, capable of inducing myocardial remodelling and dysfunction [6]. Consequently, epicardial adipose tissue has been suggested as an important risk factor in various cardiovascular diseases such as coronary artery disease, heart failure, and atrial fibrillation [7,8]. In patients with severe symptomatic aortic stenosis undergoing aortic valve replacement, epicardial fat has been associated with short- and long-term outcomes [9,10], but little is known about its role in asymptomatic patients or those with a range of aortic stenosis disease severity, nor the mechanisms underlying any association.
The present study sought to investigate the association of epicardial adipose tissue with aortic stenosis disease severity, disease progression, left ventricular remodelling and function, as well as clinical outcomes in patients with asymptomatic mild to severe aortic stenosis.
METHODS
Study Design and Study Population
This is a post-hoc analysis of asymptomatic patients with calcific aortic stenosis who participated in the ‘Study Investigating the Effects of Drugs used to Treat Osteoporosis on the Progression of Aortic Stenosis’ (SALTIRE 2) trial. This was a single-centre parallel group double blind randomized controlled trial with Trial Steering Committee oversight and regional ethics committee (Scotland A Research Ethics Committee, 14/SS/0064) and Medicines and Healthcare products Regulatory Agency (EudraCT 2014–001112-19) approvals. The study was conducted in accordance with the Declaration of Helsinki and registered on clinicaltrials.gov (NCT02132026). All patients provided written informed consent.
Trial population and protocol has been published previously [11]. In brief, patients over 50 years of age with asymptomatic calcific aortic stenosis were recruited and underwent echocardiography, non-contrast CT calcium score, and contrast-enhanced cardiac CT angiography at baseline and 1 year. Patients received one of the three trial interventions: alendronic acid, denosumab, matched placebo. The trial intervention caused no demonstrable differences in disease progression [11].
Imaging Data
Echocardiography
Baseline and 1-year echocardiograms were performed by the same experienced sonographer (ACW), on the same echocardiography scanner (Affiniti, Philips Healthcare, Netherlands) within an accredited department using a standardized protocol according to the European guidelines [12]. Aortic stenosis severity and disease progression were assessed using the mean aortic valve gradient, peak aortic jet velocity and aortic valve area using the continuity equation. Mild aortic stenosis was defined by peak aortic jet velocity 2.6–2.9m/s, mean gradient <20mmHg, aortic valve area >1.5cm2; moderate aortic stenosis: peak aortic jet velocity 3.0–4.0m/s, mean gradient 20–40mmHg, aortic valve area 1.0–1.5cm2; and severe aortic stenosis: peak aortic jet velocity >4m/s, mean gradient >40mmHg, aortic valve area <1.0cm2. To assess the myocardial remodelling, the following was measured: left ventricular volume, left ventricular wall thickness, left ventricular mass, left ventricular ejection fraction estimation, left ventricular global longitudinal strain, indexed left atrial volume [13]. Left ventricular ejection fraction was visually estimated and categorized as normal (≥55%), mildly impaired (45%−54%), moderately impaired (36%−44%), or severely impaired (≤35%). Global longitudinal strain was measured from standard apical four-chamber, two-chamber, and three-chamber views according to current guidelines [14], and strain values from six basal, six mid and five apical segments of the left ventricle were averaged to obtain three regional longitudinal strain values. Left ventricular wall thickness and internal diameter in diastole was measured on parasternal long-axis views to derive left ventricular mass.
Computed Tomography Acquisition
Baseline CT imaging was performed on a 128-detector CT scanner (Biograph mCT, Siemens, Germany) using automated tube voltage modulation (CARE Dose 4D, 100–200 kV) with 0.75mm slice thickness for contrast-enhanced CT, tube voltage of 120 kV with 3mm slice thickness for non-contrast CT. Both were ECG-gated, performed in diastole and in held expiration. Beta-blockade was administered when resting heart rate was >65 beats per minute and no contraindications. Aortic valve CT calcium score was assessed by an experienced operator (TP) using Vitrea software (Vitrea Advanced, USA) as previously described [15].
Epicardial Adipose Tissue Analysis
Epicardial adipose tissue volume and attenuation were quantified on CT using fully automated deep learning-enabled software (Figure 1; appendix 1; Supplementary Figure 1). Briefly, the 3D U-Net [16] segmentation model from the nnU-Net framework [17] was used to train the deep learning model on heart masks annotated by expert readers on 200 CT scans from 4 different international cohorts (testing and internal validation, Supplementary Table 1). A total of 150 patients of the Scottish COmputed Tomography of the HEART (SCOT-HEART) trial were used as external validation to test its accuracy (Supplementary Figure 2). Epicardial adipose tissue volume and mean attenuation were subsequently generated from 3-dimensional fat voxels of the predicted heart mask between the HU limits of −190 and −30 HU. Epicardial adipose tissue volumes indexed for body surface area (BSA) were calculated to account for body size related variations.
Figure 1.
Deep learning based epicardial adipose tissue quantification
Example of fully automated deep learning enabled epicardial adipose tissue segmentation (red overlay) from computed tomography angiography (A-C), with 3-dimensional volume rendering image of epicardial adipose tissue shown in red (D).
Clinical Outcome
Follow-up was assessed in May 2023 by electronic chart review. The primary endpoint was all-cause mortality.
Statistical Analysis
Continuous variables are presented as mean ± standard deviation or median [interquartile interval]. Categorical variables are expressed as absolute values and percentages. Distributions of data were tested for normality with Shapiro-Wilk test and quantile-quantile plots. Continuous variables were compared using independent samples Student’s t-test or Mann-Whitney U-test; categorical variables were tested using chi-square (χ2) test. Comparisons between mild, moderate, and severe aortic stenosis categories were conducted as pairwise 2-group tests: mild versus moderate, moderate versus severe, mild versus severe. Correlations between continuous variables were assessed with linear regression analysis (Pearson r or Spearman ρ), with correlation coefficients <0.20 considered very weak, 0.20 to <0.40: weak, 0.40 to <0.60: moderate, 0.60 to <0.80: strong, 0.80 to 1.00: very strong. Rates of progression were calculated using the difference between baseline and 1-year time points for all imaging assessments. Kaplan–Meier curves and Cox proportional hazards regression analyses were used to determine the association of epicardial adipose tissue with all-cause mortality. The proportional hazards assumption was tested by plotting Schoenfeld residuals. High-sensitivity cardiac troponin I, N-terminal pro-B-type natriuretic peptide, epicardial adipose tissue volume and epicardial adipose tissue volume indexed to body surface area were expressed per 1 standard deviation increment. Collinearity of variables was assessed before inclusion in the multivariable model. Variables with p-value of <0.05 in univariable analysis were included in the multivariable model. Body mass index was also included in multivariable analysis because of its potential interaction with epicardial adipose tissue and its possible confounding effect on the association between epicardial adipose tissue and all-cause mortality. The optimal cut-off of epicardial adipose tissue volume for outcome prediction was derived from the intersection point of sensitivity and specificity in the receiver-operating characteristic curve (Youden index). A two-sided p value of <0.05 was considered statistically significant. Statistical analyses were performed using SPSS (v29.0.1.0) and R (v2023.12.0+369).
RESULTS
Patient Characteristics
Of the 150 patients included in the SALTIRE 2 clinical trial, 26 were excluded due to inadequate image quality or absence of follow-up data, leaving 124 patients included in the study. Patients had a median age of 72 [68 to 77] years, most were male (78%), the median epicardial adipose tissue volume was 100 [74 to 139] mm3 and 51 [40 to 66] mm3/m2 indexed to body surface area, and epicardial adipose tissue attenuation was −70 [−74 to −66] Hounsfield Units (HU) (Table 1). Higher epicardial adipose tissue volume (above the median value) was seen in male patients and in patients with higher body mass index (Supplementary Table 2), and those two clinical characteristics were also significant predictors of epicardial adipose tissue volume in multivariable linear regression analysis (Supplementary Table 3).
Table 1.
Patient characteristics of the total study population and of patients who survived versus those who died
| characteristic | Total Study population (n=124) | Patients who survived (n=101) | Patients who died (n=23) | p-value |
|---|---|---|---|---|
| Demographics | ||||
| Age, years | 72 [68 to 77] | 71 [66 to 77] | 74 [69 to 80] | 0.080 |
| Male gender, n (%) | 97 (78) | 75 (74) | 22 (96) | 0.025 |
| Body mass index, kg/m2 | 29 [27 to 33] | 29 [27 to 33] | 28 [26 to 34] | 0.765 |
| Hypertension, n (%) | 95 (77) | 76 (75) | 19 (83) | 0.452 |
| Hyperlipidaemia, n (%) | 73 (59) | 56 (55) | 17 (74) | 0.104 |
| Diabetes mellitus, n (%) | 26 (21) | 20 (20) | 6 (26) | 0.504 |
| Previous myocardial infarction, n (%) | 16 (13) | 13 (13) | 3 (13) | 0.982 |
| Previous coronary artery bypass grafting, n (%) | 15 (12) | 11 (11) | 4 (17) | 0.388 |
| Biochemistry | ||||
| High-sensitivity cardiac troponin I, ng/mL | 5.1 [2.9 to 8.9] | 5.0 [2.6 to 7.8] | 7.3 [4.1 to 19.1] | 0.035 |
| N-terminal pro-B-type natriuretic peptide, pg/mL | 202 [110 to 416] | 190 [102 to 411] | 216 [123 to 774] | 0.383 |
| Echocardiography | ||||
| Peak aortic jet velocity, m/s | 3.4 [3.0 to 3.9] | 3.4 [3.0 to 3.9] | 3.2 [2.9 to 3.9] | 0.380 |
| 1-year change in peak aortic jet velocity, m/s | 0.24 [0.08 to 0.46] | 0.30 [0.09 to 0.48] | 0.15 [0.00 to 0.23] | 0.090 |
| Mean gradient, mmHg | 23.8 [18.3 to 31.7] | 25.3 [18.3 to 31.7] | 20.7 [18.3 to 31.0] | 0.354 |
| 1-year change in mean gradient, mmHg | 2.84 [−0.58 to 7.34] | 3.60 [−0.33 to 7.55] | 1.67 [−1.00 to 4.33] | 0.164 |
| Aortic valve area, cm2 | 1.05 [0.86 to 1.26] | 1.05 [0.85 to 1.22] | 1.07 [0.88 to 1.37] | 0.346 |
| Left ventricular ejection fraction | 0.016 | |||
| 40 – 55% | 11 (9) | 6 (6) | 5 (22) | |
| > 55% | 113 (91) | 95 (94) | 18 (78) | |
| Left ventricular septal wall thickness, cm | 1.3 [1.2 to 1.4] | 1.3 [1.2 to 1.4] | 1.4 [1.3 to 1.4] | 0.098 |
| Left ventricular end-diastolic diameter, cm | 4.4 [4.1 to 4.8] | 4.3 [4.1 to 4.8] | 4.4 [4.1 to 5.1] | 0.455 |
| Left ventricular mass, g | 215 [180 to 265] | 208 [179 to 258] | 237 [188 to 290] | 0.436 |
| Left ventricular strain, % | −19.6 [−21.5 to −18.2] | −19.6 [−21.6 to −18.2] | −19.4 [−21.2 to −16.9] | 0.566 |
| Left atrial volume index, mL/m2 | 3.9 [3.6 to 4.3] | 3.9 [3.6 to 4.3] | 4.0 [3.8 to 4.6] | 0.084 |
| Computed Tomography | ||||
| Aortic valve calcium score, Agatston Units | 1126 [632 to 2153] | 1124 [622 to 2152] | 1163 [655 to 2358] | 0.822 |
| 1-year change in aortic valve calcium score, Agatston Units | 211 [77 to 396] | 246 [94 to 401] | 110 [37 to 286] | 0.055 |
| Epicardial adipose tissue volume, mm3 | 100 [74 to 139] | 94 [73 to 127] | 128 [92 to 199] | 0.014 |
| Indexed epicardial adipose tissue volume*, mm3/cm2 | 51 [40 to 66] | 48 [39 to 64] | 64 [48 to 96] | 0.007 |
| Epicardial adipose tissue attenuation, HU | −70 [−74 to −66] | −70 [−74 to −66] | −712 [−74 to −67] | 0.880 |
Significant values are highlighted in bold.
Indexed to body surface area
Aortic Stenosis Severity
Thirty-one patients had mild aortic stenosis, 72 moderate, and 21 severe aortic stenosis. There were no differences in baseline epicardial adipose tissue volume or mean attenuation according to aortic stenosis disease severity (Figure 2, Supplementary Table 4). No association was observed between echocardiographic measures of aortic stenosis severity and epicardial adipose tissue volume (mean gradient: r=−0.11, p=0.230) or mean attenuation (mean gradient: r=0.07, p=0.448) (Supplementary Figure 3A, Supplementary Table 5). Patients with epicardial adipose tissue volume above the median had no significant difference in peak aortic jet velocity, mean gradient, or aortic valve area compared to those with epicardial fat volume below the median (Supplementary Table 2). Equally, no associations between CT aortic valve calcium score and epicardial adipose tissue volume (r=0.04, p=0.66) or mean attenuation (r=0.11, p=0.24) were observed (Supplementary Figure 3B, Supplementary Table 5).
Figure 2.
Epicardial adipose tissue volume across aortic stenosis severity levels
There were no significant differences in epicardial adipose tissue volume between patients with mild, moderate and severe aortic stenosis (97 [80–130] mm 3 vs 99 [72–148] mm3 vs 110 [76–147] mm3, respectively; overall p=0.987).
Aortic Stenosis Disease Progression
Progression of aortic stenosis severity as assessed by 1-year change in peak velocity was 0.24 [0.08 to 0.46] m/s, by 1-year change in mean gradient was 2.84 [−0.58 to 7.34] mmHg, and 1-year change in calcium score was 211 [77 to 396] Agatston Units, and was not different between patients who survived and who died during follow-up (p>0.05 for all, Table 1). No correlation was observed between baseline epicardial adipose tissue volume and echocardiographic assessments of hemodynamic disease progression (1-year change in mean gradient: r=−0.06, p=0.542) or 1-year change in calcium score (r=−0.09, p=0.41) (Supplementary Figure 3C & 3D, Supplementary Table 5). Similarly, there were no correlations with baseline epicardial adipose tissue attenuation.
Myocardial Remodelling & Function
Patients with high epicardial fat volume had higher left ventricular septal wall thickness, end-diastolic diameter and mass, and left atrial volume compared to those with epicardial fat volume below the median (p<0.05 for all, Supplementary Table 2). Epicardial adipose tissue volume was positively correlated with left ventricular septal wall thickness, left ventricular end-diastolic diameter, left ventricular mass, left ventricular global longitudinal strain, left atrial volume; and negatively correlated with left ventricular ejection fraction as assessed on baseline echocardiography (p<0.05 for all, Table 2, Figure 3). Moreover, epicardial adipose tissue volume correlated with baseline plasma cardiac troponin concentration. Evaluation of interaction effect between epicardial adipose tissue volume and measures of aortic stenosis severity (aortic valve peak velocity, mean gradient, and aortic valve area) on myocardial remodelling parameters showed no significant interaction (all p>0.005, Supplementary Table 6). Epicardial adipose tissue attenuation correlated with left ventricular end-diastolic diameter (r=0.31, p<0.001) and left ventricular mass (0.22, p=0.018), but not with other echocardiographic or biochemical markers of myocardial remodelling (Table 2).
Table 2.
Correlation between epicardial adipose tissue and markers of myocardial remodelling in aortic stenosis
| Epicardial adipose tissue volume, mm3 | Indexed epicardial adipose tissue volume*, mm3/cm2 | Epicardial adipose tissue attenuation, HU | |
|---|---|---|---|
| Left ventricular septal wall thickness, cm | rho=0.30 p<0.001 |
rho=0.26 p=0.003 |
rho=0.08 p=0.405 |
| Left ventricular end-diastolic diameter, cm | rho=0.40 p<0.001 |
rho=0.30 p=0.001 |
rho=0.31 p<0.001 |
| Left ventricular mass, g | rho=0.46 p<0.001 |
rho=0.37 p<0.001 |
rho=0.22 p=0.018 |
| Left ventricular ejection fraction, % | rho=−0.29 p=0.001 |
rho=−0.32 p<0.001 |
rho=0.16 p=0.083 |
| Left ventricular global longitudinal strain, % | rho=0.28 p=0.017 |
rho=0.244 p=0.035 |
rho=−0.07 p=0.537 |
| Left atrial volume index, mL/m2 | rho=0.39 p<0.001 |
rho=0.33 p<0.001 |
rho=0.13 p=0.161 |
| High-sensitivity cardiac troponin I, ng/mL | rho=0.23 p=0.009 |
rho=0.19 p=0.032 |
rho=0.12 p=0.184 |
| N-terminal pro-B-type natriuretic peptide, pg/mL | rho=0.18 p=0.052 |
rho=0.18 p=0.051 |
rho=−0.04 p=0.674 |
Significant values are highlighted in bold.
Indexed to body surface area
Figure 3.
Relation between epicardial adipose tissue volume and markers of myocardial remodelling Epicardial adipose tissue volume correlated with left ventricular mass (rho=0.46, p<0.001) (A). Patients with mildly reduced ejection fraction had higher epicardial adipose tissue volume compared to patients with preserved ejection fraction (164 [97–202] mm3 vs 96 [74–128] mm3; p=0.009) (B).
Outcome Prediction
During the median follow-up of 48 [26–73] months, a total of 23 (18%) patients died. Patients who died were more likely to be male (96% versus 74%, p=0.025), to have an ejection fraction below 55% (22% versus 6%, p=0.016), a higher baseline plasma cardiac troponin concentration (7.3 versus 5.0 ng/mL, p=0.035), and higher epicardial adipose tissue volume (128 versus 94 mL, p=0.014) (Table 1, Figure 4).
Figure 4.
Study characteristics of patients who died during follow-up
Patients who died were more likely to be male (96% versus 74%, p=0.025) (A), to have an ejection fraction below 55% (22% versus 6%, p=0.016) (B), a higher baseline cardiac troponin concentration (7.3 versus 5.0 ng/mL, p=0.035) (C), and higher epicardial adipose tissue volume (128 versus 94 mL, p=0.014) (D).
On univariable Cox regression analysis, epicardial adipose tissue volume (hazard ratio 1.61, 95% confidence interval 1.11–2.33; p=0.012), plasma cardiac troponin I concentration (hazard ratio 1.43, 95% confidence interval 1.16–1.75; p<0.001) and left ventricular ejection fraction (hazard ratio 3.34, 95% confidence interval 1.12–9.13; p=0.019) were associated with all-cause mortality. Both epicardial adipose tissue volume (hazard ratio 1.82, 95% confidence interval 1.10–3.03; p=0.021) and cardiac troponin concentration (hazard ratio 1.47, 95% confidence interval 1.13–1.90; p=0.004) remained independently associated with all-cause mortality on multivariable analysis (Supplementary Table 7).
Similar results were obtained when epicardial adipose tissue volume was considered as a categorical variable (high versus low) with optimal cut-off of 90 mm3 derived from the receiver-operator-characteristic curve (area under the curve 0.67, sensitivity 83%, specificity 47%). Again, epicardial adipose tissue volume was the strongest independent predictor of all-cause mortality on multivariable analysis and patients with high epicardial adipose tissue volume had a 3 to 4-fold increased risk to experience death during follow-up (hazard ratio 3.74, 95% confidence interval 1.08–12.96; p=0.037, Figure 5, Supplementary Table 7).
Figure 5.
The epicardial adipose tissue volume threshold of 90 mm3 predicted all-cause mortality during follow-up
(A) Forrest plot for multivariable Cox regression analysis. Epicardial adipose tissue threshold of 90 mm3 was an independent predictor of all-cause mortality, and patients with high epicardial adipose tissue volume had 3- to 4-fold increase in risk of death. (B) Kaplan-Meier Curves demonstrating event-free survival using the 90 mm3 epicardial adipose tissue volume threshold. High epicardial adipose tissue volume was associated with adverse prognosis (logrank p=0.026). Epicardial adipose tissue volume quantified from CT angiography in (C) a patient in their 70s with moderate aortic stenosis and epicardial fat volume of 73 mm3 who did not experience death during follow-up, and in (D) a patient in their 70s with moderate aortic stenosis and epicardial adipose tissue volume of 192 mm3 who died during follow-up. * Per 1 standard deviation increment. HR = hazard ratio, CI = confidence interval.
DISCUSSION
In this prospective trial of asymptomatic patients with mild to severe aortic stenosis, we have demonstrated that epicardial adipose tissue volume was independently associated with all-cause mortality. Although there was no association with aortic stenosis disease severity or progression, epicardial fat volume was associated with markers of myocardial remodelling. Epicardial adipose tissue can therefore be considered as an early marker of impaired myocardial health with the potential to improve risk stratification and outcome prediction in those patients. Importantly, our data also suggest that epicardial adipose tissue may represent a treatment target for this important condition.
The myocardial remodelling response to aortic stenosis varies between individuals, is not predicted by stenosis severity alone, and may occur early in the disease process [18]. This suggests that unidentified local and systemic factors might contribute to the myocardial remodelling. Although beneficial in its initial phase by restoring wall stress and cardiac performance, the left ventricular hypertrophic response ultimately decompensates, driven by progressive myocardial fibrosis and cell death, leading to impaired cardiac function, symptoms and poor prognosis. Identifying factors that contribute to adverse myocardial remodelling in aortic stenosis is therefore an important goal to improve risk prediction, timing of intervention and ultimately patient outcome.
In our study, we found that epicardial adipose tissue volume may be one of those factors contributing to early adverse cardiac remodelling and worse patient outcomes. Indeed, we demonstrated an association between epicardial fat volume and increased left ventricular mass, reduced ejection fraction, and poorer rates of survival that was independent of aortic stenosis severity itself. Our interaction analysis further confirmed that the associations between epicardial adipose tissue volume and markers of myocardial remodelling were independent of aortic stenosis severity, highlighting its potential role as a distinct contributor to myocardial health. Moreover, the observed effect of epicardial fat volume was independent of body size as our results were similar when using volumes indexed to body surface area. Importantly, the body mass index of our cohort corresponded to overweight or mild obesity, suggesting that systemic influences of increased body fat on myocardial health may also play a role in this population. However, the unique pathophysiological effects of epicardial fat, distinct from other visceral fat depots, may exert additional strain on the left ventricle through localized metabolic and inflammatory signalling. Even after adjusting for aortic valve replacement, epicardial adipose tissue volume remained independently associated with all-cause mortality, highlighting its prognostic value irrespective of interventions aimed at relieving valve obstruction.
Given its anatomical location within the pericardium and in direct contact with the myocardium, epicardial adipose tissue may contribute to structural damage and adverse remodelling of the adjacent myocardium through its metabolic and pro-inflammatory activities [21]. At the same time, as a visceral fat depot, epicardial adipose tissue reflects broader systemic metabolic disturbances that promote myocardial remodelling. For example, increased visceral adiposity has been associated with elevated levels of pro-inflammatory adipokines such as leptin and resistin, which may drive maladaptive left ventricular hypertrophy, fibrosis and ultimately decompensation [22]. Thus, epicardial adipose tissue likely represents both a local and systemic driver of myocardial stress in the context of pressure overload, potentially worsening aortic stenosis related myocardial damage [19,20].
Epicardial adipose tissue has already been linked to several cardiovascular diseases [7,8], and the amount of epicardial adipose tissue has been related to impairment of systolic [23] and diastolic function [24]. However, only few studies have investigated the role of epicardial adipose tissue in patients with aortic stenosis so far. Studies focusing on patients undergoing aortic valve replacement have demonstrated an association between higher epicardial fat volume and worse short- [9], and long-term outcomes [10]. Coisne et al. showed that epicardial fat thickness measured by echocardiography was associated with left ventricular remodelling in aortic stenosis [25]. Similar, Arangalage et al. measured epicardial adipose tissue on non-contrast CT in 143 patients with asymptomatic aortic stenosis and demonstrated an association with left ventricular mass, independent of aortic stenosis severity [26]. We have confirmed and extended these results by obtaining epicardial fat volume from CT angiography and correlated our results with both extensive markers of myocardial remodelling (left ventricular mass, ejection fraction, global longitudinal strain, and left atrial dilatation) and long-term outcome of patients with all degrees of aortic stenosis severity. We have also proposed an optimal cut-off of epicardial adipose tissue volume of 90mm3 that identified patients with 3- to 4-fold increased risk of death. This now requires further external validation in other cohorts.
With all this in mind, epicardial fat could be proposed as a potential treatment target that might be favourably modified by drugs that reduce visceral fat, like glucagon-like peptide 1 receptor agonists. Indeed, favourable effects on the epicardial fat may underlie the beneficial effects of these drugs already observed in patients with heart failure with preserved ejection fraction [27] which shares many similarities with the myocardium of patients with aortic stenosis.
The attenuation of local regions of adipose tissue directly surrounding the coronary arteries may have prognostic value in patients with coronary artery disease [28]. However, when considering epicardial fat as a whole, its volume rather than attenuation is of clinical importance [29]. Indeed, in this study, whilst epicardial adipose tissue attenuation correlated with myocardial mass, no association was observed with other biochemical or imaging markers of myocardial remodelling or function and it was unable to predict mortality.
The most accurate imaging technique to assess epicardial fat burden is volumetric quantification, but measuring epicardial fat from CT has represented a cumbersome process with limited clinical applicability. Recently, fully automated quantification from non-contrast CT has been validated [30], although we are the first to our knowledge to implement a method for fully automated quantification from CT angiography. This is very relevant as CT angiography is increasingly incorporated into clinical workflows for evaluation of coronary artery disease and before transcatheter aortic valve replacement. Our method provides a clinically useful and readily available tool to obtain epicardial adipose tissue volume as an extra prognostic marker from standard CT angiography in an easy and quick way.
This study has some limitations which should be acknowledged. This is a post-hoc analysis of previously collected data, no power calculation was performed. The low event rate in this cohort may limit the statistical power to predict outcomes robustly. Continuous data for left ventricular ejection fraction were not available, and this variable was treated as categorical variable, which may reduce granularity in assessing its relationship with outcomes. The SALTIRE 2 trial, a single-centre study in Scotland, primarily included elderly Caucasian men, limiting broader generalizability. The applicability of the epicardial adipose tissue volume threshold of 90mm3 should be confirmed in larger external prospective studies. The causal mechanisms explaining the association between pro-inflammatory activity of epicardial adipose tissue and myocardial injury and patient outcome in aortic stenosis need to be further investigated.
To conclude, in this prospective trial of patients with asymptomatic mild to severe aortic stenosis, epicardial adipose tissue volume automatically derived from CT angiography was associated with blood and imaging biomarkers of myocardial remodelling as well as all-cause mortality. Epicardial adipose tissue volume may therefore represent a novel treatment target and a marker of impaired myocardial health that has the potential to improve risk stratification and outcome prediction in patients with aortic stenosis.
Supplementary Material
KEY MESSAGES.
What is already known on this topic:
Epicardial adipose tissue is a metabolically active visceral fat depot that has been linked to adverse cardiovascular outcomes
Little is known about its role in patients with aortic stenosis
What this study adds:
Epicardial adipose tissue volume can easily be obtained from contrast computed tomography using fully automated software
In patients with aortic stenosis, increased epicardial adipose tissue volume is associated with markers of adverse myocardial remodelling and all-cause mortality
How this study might affect research, practice or policy:
Automated quantification of epicardial adipose tissue from computed tomography angiography provides a clinically feasible approach for integrating epicardial fat assessment into routine cardiovascular imaging workflows
Epicardial adipose tissue could serve as an early marker of impaired myocardial health that has the potential to improve risk stratification and outcome prediction in patients with aortic stenosis
FUNDING
This study was supported by the National Heart, Lung, and Blood Institute, USA (grants 1R01HL148787, R01HL151266 and 1R01HL175875), the Miriam & Sheldon G Adelson Medical Research Foundation, and the Winnick Family Foundation. KG is supported by the Foundation for Polish Science, and Polish National Agency for Academic Exchange. MRD is supported by the British Heart Foundation (FS/SCRF/21/32010) and recipient of the Sir Jules Thorn Award for Biomedical research 2015 (15/JTA). DEN is supported by the British Heart Foundation (CH/09/002, RG/F/22/110093, RE/24/130012).
Footnotes
DISCLOSURE OF INTEREST
The authors have nothing to disclose
DATA AVAILABILITY STATEMENT
The data underlying this article will be shared on reasonable request to the corresponding author.
REFERENCES
- 1.Nkomo VT, Gardin JM, Skelton TN, et al. Burden of valvular heart diseases: a population-based study. Lancet 2006;368:1005–11. doi: 10.1016/S0140-6736(06)69208-8 [DOI] [PubMed] [Google Scholar]
- 2.Vahanian A, Beyersdorf F, Praz F, et al. 2021 ESC/EACTS Guidelines for the management of valvular heart disease. Eur Heart J 2022;43:561–632. doi: 10.1093/eurheartj/ehab395 [DOI] [PubMed] [Google Scholar]
- 3.Braunwald E. On the natural history of severe aortic stenosis. J. Am. Coll. Cardiol. 1990;15:1018–20. doi: 10.1016/0735-1097(90)90235-H [DOI] [PubMed] [Google Scholar]
- 4.Hein S, Arnon E, Kostin S, et al. Progression from compensated hypertrophy to failure in the pressure-overloaded human: Heart structural deterioration and compensatory mechanisms. Circulation 2003;107:984–91. doi: 10.1161/01.CIR.0000051865.66123.B7 [DOI] [PubMed] [Google Scholar]
- 5.Iacobellis G, Corradi D, Sharma AM. Epicardial adipose tissue: Anatomic, biomolecular and clinical relationships with the heart. Nat. Clin. Pract. Cardiovasc. Med. 2005;2:536–43. doi: 10.1038/ncpcardio0319 [DOI] [PubMed] [Google Scholar]
- 6.Iacobellis G, Barbaro G. Epicardial adipose tissue feeding and overfeeding the heart. Nutrition. 2019;59:1–6. doi: 10.1016/j.nut.2018.07.002 [DOI] [PubMed] [Google Scholar]
- 7.Raggi P, Alakija P. Epicardial adipose tissue: A long-overlooked marker of risk of cardiovascular disease. Atherosclerosis. 2013;229:32–3. doi: 10.1016/j.atherosclerosis.2013.02.030 [DOI] [PubMed] [Google Scholar]
- 8.Rosito GA, Massaro JM, Hoffmann U, et al. Pericardial fat, visceral abdominal fat, cardiovascular disease risk factors, and vascular calcification in a community-based sample the framingham heart study. Circulation 2008;117:605–13. doi: 10.1161/CIRCULATIONAHA.107.743062 [DOI] [PubMed] [Google Scholar]
- 9.Eberhard M, Stocker D, Meyer M, et al. Epicardial adipose tissue volume is associated with adverse outcomes after transcatheter aortic valve replacement. Int J Cardiol 2019;286:29–35. doi: 10.1016/j.ijcard.2019.01.068 [DOI] [PubMed] [Google Scholar]
- 10.Schulz A, Beuthner BE, Böttiger ZM, et al. Epicardial adipose tissue as an independent predictor of long-term outcome in patients with severe aortic stenosis undergoing transcatheter aortic valve replacement. Clin Res Cardiol 2024;:1–12. doi: 10.1007/s00392-024-02387-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pawade TA, Doris MK, Bing R, et al. Effect of Denosumab or Alendronic Acid on the Progression of Aortic Stenosis: A Double-Blind Randomized Controlled Trial. Circulation 2021;143:2418–27. doi: 10.1161/CIRCULATIONAHA.121.053708 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harkness A, Ring L, Augustine DX, et al. Normal reference intervals for cardiac dimensions and function for use in echocardiographic practice: a guideline from the British Society of Echocardiography. Echo Res Pract 2020;7:X1. doi: 10.1530/ERP-19-0050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Dweck MR, Loganath K, Bing R, et al. Multi-modality imaging in aortic stenosis: an EACVI clinical consensus document. Eur Heart J Cardiovasc Imaging 2023;24:1430–43. doi: 10.1093/ehjci/jead153 [DOI] [PubMed] [Google Scholar]
- 14.Voigt JU, Pedrizzetti G, Lysyansky P, et al. Definitions for a common standard for 2D speckle tracking echocardiography: consensus document of the EACVI/ASE/Industry Task Force to standardize deformation imaging. Eur Heart J Cardiovasc Imaging 2015;16:1–11. doi: 10.1093/ehjci/jeu184 [DOI] [PubMed] [Google Scholar]
- 15.Pawade T, Sheth T, Guzzetti E, et al. Why and How to Measure Aortic Valve Calcification in Patients With Aortic Stenosis. JACC Cardiovasc. Imaging 2019;12:1835–48. [DOI] [PubMed] [Google Scholar]
- 16.Çiçek Ö, Abdulkadir A, Lienkamp SS, et al. 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation. Lect Notes Comput Sci (including Subser Lect Notes Artif Intell Lect Notes Bioinformatics) 2016;9901 LNCS:424–32.http://arxiv.org/abs/1606.06650 (accessed 3 Apr 2024). [Google Scholar]
- 17.Isensee F, Jaeger PF, Kohl SAA, et al. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods 2021;18:203–11. doi: 10.1038/s41592-020-01008-z [DOI] [PubMed] [Google Scholar]
- 18.Dweck MR, Joshi S, Murigu T, et al. Left ventricular remodeling and hypertrophy in patients with aortic stenosis: Insights from cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2012;14. doi: 10.1186/1532-429X-14-50 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Alexopoulos N, McLean DS, Janik M, et al. Epicardial adipose tissue and coronary artery plaque characteristics. Atherosclerosis 2010;210:150–4. doi: 10.1016/j.atherosclerosis.2009.11.020 [DOI] [PubMed] [Google Scholar]
- 20.Paulus WJ, Tschöpe C. A novel paradigm for heart failure with preserved ejection fraction: Comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation. J. Am. Coll. Cardiol 2013;62:263–71. doi: 10.1016/j.jacc.2013.02.092 [DOI] [PubMed] [Google Scholar]
- 21.Hao S, Sui X, Wang J, et al. Secretory products from epicardial adipose tissue induce adverse myocardial remodeling after myocardial infarction by promoting reactive oxygen species accumulation. Cell Death Dis 2021;12:1–11. doi: 10.1038/s41419-021-04111-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Grymyr LMD, Nadirpour S, Gerdts E, et al. Left ventricular myocardial oxygen demand and subclinical dysfunction in patients with severe obesity referred for bariatric surgery. Nutr Metab Cardiovasc Dis 2021;31:666–74. doi: 10.1016/j.numecd.2020.10.009 [DOI] [PubMed] [Google Scholar]
- 23.Watanabe K, Kishino T, Sano J, et al. Relationship between epicardial adipose tissue thickness and early impairment of left ventricular systolic function in patients with preserved ejection fraction. Heart Vessels 2016;31:1010–5. doi: 10.1007/s00380-015-0650-8 [DOI] [PubMed] [Google Scholar]
- 24.Lin HH, Lee JK, Yang CY, et al. Accumulation of epicardial fat rather than visceral fat is an independent risk factor for left ventricular diastolic dysfunction in patients undergoing peritoneal dialysis. Cardiovasc Diabetol 2013;12:127. doi: 10.1186/1475-2840-12-127 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Coisne A, Ninni S, Ortmans S, et al. Epicardial fat amount is associated with the magnitude of left ventricular remodeling in aortic stenosis. Int J Cardiovasc Imaging 2019;35:267–73. doi: 10.1007/s10554-018-1477-z [DOI] [PubMed] [Google Scholar]
- 26.Arangalage D, Mathieu T, Nguyen V, et al. Epicardial adipose tissue volume is associated with left ventricular remodelling in calcific aortic valve stenosis. Arch Cardiovasc Dis 2019;112:594–603. doi: 10.1016/j.acvd.2019.06.005 [DOI] [PubMed] [Google Scholar]
- 27.Kosiborod MN, Abildstrøm SZ, Borlaug BA, et al. Semaglutide in Patients with Heart Failure with Preserved Ejection Fraction and Obesity. N Engl J Med 2023;389:1069–84. doi: 10.1056/nejmoa2306963 [DOI] [PubMed] [Google Scholar]
- 28.Tzolos E, Williams MC, McElhinney P, et al. Pericoronary Adipose Tissue Attenuation, Low-Attenuation Plaque Burden, and 5-Year Risk of Myocardial Infarction. JACC Cardiovasc Imaging 2022;15:1078–88. doi: 10.1016/j.jcmg.2022.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hell MM, Ding X, Rubeaux M, et al. Epicardial adipose tissue volume but not density is an independent predictor for myocardial ischemia. J Cardiovasc Comput Tomogr 2016;10:141–9. doi: 10.1016/j.jcct.2016.01.009 [DOI] [PubMed] [Google Scholar]
- 30.Commandeur F, Goeller M, Betancur J, et al. Deep Learning for Quantification of Epicardial and Thoracic Adipose Tissue from Non-Contrast CT. IEEE Trans Med Imaging 2018;37:1835–46. doi: 10.1109/TMI.2018.2804799 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article will be shared on reasonable request to the corresponding author.