Epicardial Adipose Tissue in Cardiovascular Disease: Integration of Imaging, Biomarkers, and Clinical Outcomes

Abstract

This thesis investigates the clinical significance of epicardial adipose tissue as an imaging biomarker in cardiovascular disease through three complementary studies: a systematic review and meta-analysis, a pooled analysis of randomised controlled trials, and a 10-year outcome study. EAT, a metabolically active visceral fat depot, has emerged as a potentially modifiable marker of cardiometabolic risk. The first study presents a systematic review and meta-analysis of four clinical studies (N=189), evaluating the effect of sodium-glucose co-transporter 2 inhibitors on EAT in patients with cardiovascular disease. Across studies, SGLT2 inhibitors significantly reduced EAT volume, as measured by CT and MRI. This reduction was associated with decreases in TNF-α and body weight, supporting both anti-inflammatory and metabolic mechanisms. These findings highlight EAT as a responsive imaging biomarker that may reflect therapeutic benefit. The second study involves a pooled cardiac MRI analysis of two randomised controlled trials (DAPA-LVH and REFORM) including 96 patients with heart failure or left ventricular hypertrophy. Dapagliflozin treatment over 12 months significantly reduced EAT thickness, BMI, and LV mass compared to placebo. However, these reductions did not correlate with changes in inflammatory markers or cardiac function, suggesting that the structural cardiac benefits may arise through mechanisms not directly mediated by systemic inflammation. This supports the hypothesis that EAT reduction maybe one contributor to the overall cardiometabolic benefits of SGLT2 inhibitors. The third study examines long-term outcomes over 10 years in a sub cohort of 145 patients of SUMMIT study. Higher baseline EAT was associated with hospital admission and mortality in univariate models. Multivariable analysis identified Aβ40 and MMP-12 as independent predictors of mortality, and ESM-1 as a protective factor for hospital admission. EAT retained borderline or univariate significance, suggesting its prognostic value may depend on the clinical context and integration with circulating biomarkers. Collectively, these studies support the role of EAT as a non-invasive imaging biomarker that offers prognostic insight and potential utility for therapeutic monitoring in high-risk cardiometabolic populations.