Mediastinal adipose tissue as an active player in cardiovascular disease: a multimodality imaging narrative review Authors: Lorenza Cananzi, Federico Greco, Andrea Buoso, Caterina Bernetti, Luca Pugliese, Gianfranco Di Gennaro, Bruno Beomonte Zobel, Carlo Augusto Mallio Published: 2026 ABSTRACT AbstractBackground and ObjectiveMediastinal adipose tissue comprises several distinct fat depots, including epicardial, pericoronary, pericardial, and paracardial adipose tissue, which surround the heart and great vessels and actively contribute to cardiovascular pathophysiology. Under physiological conditions, these adipose compartments exert protective metabolic and mechanical functions; however, in pathological states, their expansion and remodeling promote inflammation, endothelial dysfunction, fibrosis, and atherosclerotic progression. This narrative review aims to summarize the current evidence regarding the anatomical distribution, imaging assessment, and clinical implications of mediastinal adipose tissue in cardiovascular disease.MethodsA narrative review was conducted using PubMed, Scopus, and Web of Science databases, covering publications up to January 2026. Only articles published in English were considered. Original research articles and relevant reviews focusing on imaging, pathophysiology, and clinical implications of mediastinal adipose tissue were included. Reference lists of selected studies were additionally screened to identify additional relevant publications.Key Content and FindingsComputed tomography (CT) enables highly reproducible volumetric and qualitative assessment of cardiac adipose tissue, including inflammatory markers such as the fat attenuation index (FAI). Magnetic resonance imaging (MRI) provides excellent tissue characterization without ionizing radiation, whereas echocardiography represents a simple and widely available screening tool. Increased epicardial, pericoronary, and pericardial adipose tissue (PAT) burden has been consistently associated with coronary artery disease (CAD), atrial fibrillation (AF), and heart failure (HF), particularly HF with preserved ejection fraction (HFpEF). Emerging evidence also supports the potential role of artificial intelligence (AI) and radiomics for automated quantification and cardiovascular risk stratification.ConclusionsMediastinal adipose tissue is increasingly recognized as an active biomarker and potential therapeutic target in cardiovascular disease. Multimodality imaging plays a central role in its evaluation by providing complementary anatomical, functional, and inflammatory information. Further standardization of imaging methodologies and validation of treatment-related changes are needed to support future clinical applications and personalized cardiovascular risk assessment.Keywords: Atrial fibrillation (AF), cardiac adipose tissue, coronary artery disease (CAD), cardiovascular imaging, heart failure (HF) Introduction Anatomy & physiology The mediastinum contains several adipose depots that surround the heart and great vessels. These include: (I) epicardial adipose tissue (EAT); (II) pericoronary adipose tissue (PCAT); (III) pericardial adipose tissue (PAT); and (IV) paracardial adipose tissue. EAT lies between the myocardium and the visceral pericardium (Figure 1), whereas PCAT represents a subset of EAT that specifically surrounds the coronary arteries. PAT is located between the visceral and parietal layers of the pericardium, while paracardial fat is found outside the parietal pericardium, contributing to the overall complexity of cardiac adipose distribution (Figure 2) (1). These adipose compartments are now recognized as metabolically active tissues, rather than inert energy stores. They secrete a wide range of adipokines, cytokines, and inflammatory mediators. Because no fascial barrier separates EAT from the underlying myocardium, EAT is in direct contact with the myocardium. This anatomical continuity allows EAT to exert paracrine and endocrine effects on the myocardium and coronary arteries (2). Under physiological conditions, EAT provides several protective functions, including mechanical cushioning, thermoregulation, immune support, and metabolic buffering through the storage and release of free fatty acids. In pathological states, however, these same tissues may promote oxidative stress, endothelial dysfunction, fibrosis, and atherosclerosis progression (3). In addition to cardiac adipose compartments, the mediastinum may also contain other fat-related entities that should be considered in the broader differential diagnosis. These include mediastinal lipomatosis, typically characterized by diffuse, non-encapsulated fat accumulation often linked with obesity, and steroid therapy, and thymic fatty involution, a physiological or age-related process in which the thymus is progressively replaced by adipose tissue (4,5). Although these conditions are usually benign and distinct from cardiac adipose depots, their recognition is important in comprehensive mediastinal imaging assessment. Axial non-contrast-enhanced chest CT image at the level of the heart. The yellow ROI delineates EAT, defined as the fat located between the outer surface of the myocardium and the visceral layer of the pericardium. The ROI follows the cardiac contours and highlights tissue with attenuation values consistent with adipose tissue on non-contrast CT, allowing visualization and assessment of epicardial fat distribution. CT, computed tomography; EAT, epicardial adipose tissue; ROI, region of interest. Axial non-contrast chest CT image at the level of the heart. The yellow ROI delineates PAT, which is located outside the EAT. PAT is bounded internally by the visceral pericardial layer, separating it from the EAT, and externally by the parietal pericardial layer, which separates it from the paracardial adipose tissue. The ROI follows the pericardial contours and highlights tissue with low attenuation values characteristic of adipose tissue on non-contrast CT, enabling visualization and assessment of pericardial fat distribution. CT, computed tomography; EAT, epicardial adipose tissue; PAT, pericardial adipose tissue; ROI, region of interest. Clinical implications Clinically, increased volumes of EAT and PAT have been consistently associated with coronary artery disease (CAD), heart failure (HF), and atrial fibrillation (AF) (6,7). EAT has also emerged as a potential therapeutic target in cardiovascular prevention and treatment, given its close anatomical and functional interaction with the myocardium and coronary vessels. In this context, EAT represents an active cardiovascular risk marker reflecting underlying metabolic and inflammatory processes, and its assessment may enable earlier and more refined cardiovascular risk stratification, supporting a more personalized approach to patient management (8). Given its growing clinical relevance, accurate identification and quantification of mediastinal fat through imaging modalities such as computed tomography (CT), magnetic resonance imaging (MRI), and echocardiography have become increasingly important. This review aims to summarize the current evidence on the anatomical distribution, imaging appearance, quantification methods, and clinical implications of mediastinal adipose tissue. We present this article in accordance with the Narrative Review reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2845/rc). Methods A narrative literature review was conducted to identify relevant studies on mediastinal adipose tissue and cardiovascular disease. A comprehensive search of electronic databases, including PubMed, Scopus, and Web of Science, was performed covering publications up to January 2026. Only articles published in English were considered. Original research studies, focusing on imaging, pathophysiology, and clinical implications of cardiac adipose tissue were included. Reference lists of selected articles were also screened to identify additional relevant studies. The search strategy summary is presented in Table 1. Table 1The search strategy summaryItemsSpecificationDate of searchInitial search: December 2025; final update: January 2026Databases searchedPubMed, Scopus, and Web of ScienceSearch terms usedTerms related to epicardial adipose tissue, pericoronary adipose tissue, pericardial adipose tissue, cardiac imaging, coronary artery disease, atrial fibrillation, and heart failure were used. Both MeSH terms and free-text keywords were applied. Detailed search strategies for one database are provided in the supplementary materialTimeframeArticles published up to January 2026, without time-span limitationsInclusion and exclusion criteriaOnly articles published in English were considered. Original research articles and relevant reviews focusing on imaging, pathophysiology, and clinical implications of cardiac adipose tissue were included. Case reports, editorials, conference abstracts, and non-English articles were excludedSelection processThe selection process was conducted independently by two radiologists. Any discrepancies were resolved through discussion and consensus between the reviewersAny additional considerationsReference lists of selected articles were additionally screened to identify further relevant studiesMeSH, medical subject headings. The search strategy summary MeSH, medical subject headings. Imaging appearance and quantification of mediastinal fat CT Appearance on CT EAT, PAT, and PCAT appear hypodense on CT, with typical attenuation ranging from −190 to −30 Hounsfield units (HU) (9-11). CT provides high spatial resolution and three-dimensional visualization of the heart and surrounding fat depots, allowing volumetric reconstruction for total and regional fat volumes (12,13). Beyond volumetric assessment, tissue attenuation provides insight into adipose tissue composition and metabolic activity. The fat attenuation index (FAI), quantifies local variation in fat density, serving as a marker of inflammation and remodeling. Increased PCAT attenuation has been associated with coronary inflammation whereas PAT attenuation tends to rise during pro-fibrotic remodeling (11). Quantification EAT volumes can be quantified using semi-automated or fully automated algorithms (quantifying both EAT volume and attenuation). Volumetric quantification of EAT by cardiac CT is highly reproducible. Several studies have evaluated how contrast enhancement, tube current, and the use of non-electrocardiogram (ECG)-gated chest CT affect EAT quantification. The reference standard is generally considered to be 120 kV non-contrast, ECG-gated cardiac CT. Contrast-enhanced scans tend to underestimate EAT volume compared with non-contrast studies (mean difference, ~31 mL; 95% limits of agreement: 27 to −89 mL) (14). However, despite this systematic bias, studies have shown good agreement between EAT volumes measured on contrast-enhanced and non-contrast CT using the standard HU threshold of −190 to −30 HU, suggesting that contrast CT may still provide reliable relative estimates when acquisition parameters are consistent. Measurement errors can also occur due to partial volume effects, where small structures within a voxel can alter the calculated attenuation value. Consequently, voxel size and image resolution must be considered when comparing EAT volume and radiodensity across studies (15). Differences in contrast use, cardiac phase, and HU threshold can affect measurements (12). The CT attenuation of PCAT reflects the lipid-to-aqueous balance within adipocytes, correlating with cell size and differentiation, and shows a gradual decrease from pericoronary to epicardial fat (16). Higher EAT radiodensity has been associated with increased coronary atherosclerotic burden, major adverse cardiovascular events, and conditions such as myocardial infarction with non-obstructive coronary arteries and Takotsubo syndrome. EAT volume measured by cardiac CT is highly reproducible but standardized reference values are still lacking or not universally accepted. Spearman et al. (17) reported the threshold of cardiac CT measured EAT volume >125 mL as a predictor of cardiac pathology. Increased radiodensity of PCAT may reflect local inflammation and atherosclerosis (11,14). PAT volume correlates with cardiometabolic risk and metabolic syndrome (18). EAT Manual tracing of the visceral pericardium and myocardial border; voxels between these borders within −190 to −30 HU are counted (11,19). 3D reconstruction allows total and regional volume calculation. Acquisition may be contrast-enhanced or non-contrast, and during systole or diastole, affecting measurements slightly (12). EAT volume correlates with sex, age, and CAD severity; mean EAT volume in Coronary Artery Disease-Reporting and Data System (CAD-RADS) 5 was 149 cm3, higher in men (20). PCAT Measured around main coronary arteries right coronary artery (RCA), left anterior descending artery (LAD), and left circumflex artery (LCX), within a radial distance equal to the vessel diameter (9). PCAT attenuation reflects coronary inflammation, predicting CAD progression and response to therapy (11). PAT Quantified using single-slice or volumetric CT, including scans performed for other indications. Men have higher PAT volume than women and it increases with body mass index (BMI), particularly in men. PAT volume correlates with cardiometabolic risk and metabolic syndrome (18). Differences in contrast use, cardiac phase, and HU threshold can affect measurements (12). Clinical correlations EAT volume correlates with: CAD severity, number of diseased vessels, luminal stenosis; CAD-RADS score and coronary calcium score; male sex and older age. PCAT attenuation correlates with local coronary inflammation and early atherosclerosis (11). PAT volume serves as a marker of cardiometabolic risk and metabolic syndrome (18). CT measurements of EAT and PAT can be used for risk stratification and opportunistic screening (13,18). MRI MRI appearance and general advantages MRI allows non-invasive and accurate assessment of EAT and PAT thickness and volume. It is operator-independent, not limited by acoustic windows, and offers excellent soft-tissue contrast for detailed fat assessment (13). Compared with CT, MRI avoids radiation exposure and does not require contrast agent administration for fat quantification. However, MRI examinations are longer, more costly, less available, and may be affected by lower spatial resolution and thicker slices than CT, which can reduce segmentation accuracy. MRI techniques for fat assessment Cine steady-state free precession imaging In steady-state free precession imaging, signal intensity is primarily determined by the tissue T1/T2 ratio. Because both blood and fat exhibit similarly high T1/T2 ratios, they appear with very bright signal intensity, whereas myocardium—characterized by a lower T1/T2 ratio—shows comparatively lower signal. This intrinsic contrast allows steady-state free precession sequences to delineate adipose tissue from the atrial and ventricular myocardial walls with superior clarity compared with other noninvasive imaging modalities. Consequently, cardiac magnetic resonance (CMR) represents an optimal technique for assessing the presence, distribution, and extent of EAT (21). Cine steady-state free precession sequences allow dynamic visualization of cardiac motion and provide bright signal for fat and blood due to their high T2/T1 ratio. The pericardium appears as a thin curvilinear structure of intermediate-to-low signal intensity, surrounded by high-signal epicardial and paracardial fat, improving anatomical delineation. Time-resolved cine steady-state free precession may help discriminate epicardial from paracardial fat by exploiting their different deformation patterns with cardiac motion (22). Fat-water separation (Dixon) techniques Fat-water separation methods—including 2-point, 3-point, and multi-echo Dixon—are widely used for fat quantification due to their robustness and reproducibility Fat-only images show bright fat signal, whereas water-only images highlight the pericardium as a thin bright line, facilitating compartment discrimination (23). Three-dimensional multi-echo Dixon techniques (e.g., cinedixon) enable high-resolution volumetric assessment of EAT. Time-resolved cine Dixon improves border delineation and reduces inter-observer variability compared with single-phase Dixon, leveraging cardiac motion to differentiate epicardial and paracardial compartments (22). Fat suppression and additional sequences Short tau inversion recovery and other fat-suppressed sequences may enhance pericardial visualization but are less commonly used for quantitative assessment (23). Quantification MRI allows reliable volumetric quantification of EAT with high reproducibility (24). EAT volume can be measured using 3D Dixon sequences or time-resolved cine Dixon approaches (22). Manual delineation of EAT on cine steady-state free precession long-axis and 4-chamber images is feasible and correlates well with CT-derived volume (25). Measuring pericardial fat area from a single long-axis cine view also correlates strongly with 3D CT-derived pericardial fat volumes (26). MRI-derived fat-signal fraction correlates strongly with CT-based EAT volume (r=0.92) and inversely with CT fat attenuation (r=−0.73), indicating that MRI captures both volume and tissue quality. Differences between MRI and CT volumes arise from spatial resolution, slice thickness, and acquisition parameters. In adolescents, 3D phase-sensitive inversion recovery based techniques enable precise and reproducible quantification of total EAT volume, minimizing the influence of inter-individual fat distribution variability (27). Clinical implications MRI-derived EAT parameters have demonstrated prognostic value. Increased EAT thickness measured by CMR was independently associated with higher risk of myocardial infarction, stroke, HF, and cardiac death over long-term follow-up (24). Adding EAT volume index to traditional CMR markers such as LVEF, perfusion defects, and late gadolinium enhancement improves prediction of major adverse cardiovascular events (28). EAT quality measured by CMR T1-mapping may outperform volume as a predictor of adverse cardiovascular outcomes, suggesting that lipid composition and inflammation play relevant roles (29). Fat-water separation combined with late gadolinium enhancement can distinguish benign myocardial fat from pathological fibrofatty infiltration in arrhythmogenic cardiomyopathy or ischemic disease (23). Ultrasound Normal echocardiographic appearance EAT can be visualized on transthoracic echocardiography (TTE) as an echo-free layer located between the outer myocardial wall and the visceral pericardium (14). It is most prominent along the right ventricular free wall, where the thickness is typically greatest (30). In some cases—particularly when the adipose layer is abundant (>15 mm) or inflamed—EAT may appear more hyperechoic. Thickness of epicardial fat can be measured at the echo and it is variable: it can range from 1 mm up to 23 mm (in obese patients) (31). PAT appears as a hypoechoic space anterior to the EAT and the parietal pericardium. Unlike epicardial fat, PAT does not substantially deform during the cardiac cycle, a feature sometimes used to distinguish the two compartments (13). A known limitation is that epicardial fat may mimic pericardial effusion, requiring careful interpretation to avoid misclassification (14,30). Echocardiographic techniques for EAT measurement TTE represents a simple, noninvasive, and inexpensive technique for the study of cardiac fat. EAT thickness is typically measured perpendicularly to the myocardial wall in parasternal long-axis, parasternal short-axis, and apical long-axis views. Measurements are performed at end-systole, when the adipose layer is least compressed, ensuring the most accurate and reproducible values (31). The preferred site for measurement is the right ventricular free wall because it consistently exhibits the greatest epicardial fat thickness and allows optimal beam alignment (30). Echocardiographic EAT may appear echo-free or hyperechoic, depending on the amount of fat present or inflammatory changes (32). Average thickness is typically derived from three cardiac cycles, and a thickness >5 mm has been proposed as a threshold identifying increased cardiometabolic risk (14). Clinical correlations and applications Echocardiographically measured EAT thickness has demonstrated strong associations with several cardiometabolic and cardiovascular conditions. Increased EAT thickness correlates with metabolic syndrome, insulin resistance, type 2 diabetes, CAD, and subclinical atherosclerosis (31). It also reflects visceral adiposity more reliably than anthropometric measurements and may serve as an accessible marker for cardiometabolic risk stratification. Although TTE is inferior to CT and MRI for detailed tissue characterization or precise volumetric quantification (13), it remains valuable for screening and large-scale clinical studies due to its low cost, wide availability, and absence of radiation exposure (32). Comparison In summary, each imaging modality provides complementary information in the evaluation of cardiac adipose tissue. Echocardiography is widely available and useful for screening but limited in quantitative assessment. CT offers highly reproducible volumetric and attenuation-based measurements, at the cost of radiation exposure. MRI provides excellent tissue characterization without radiation, although it is less available and more time-consuming. A structured comparison of these modalities is presented in Table 2. Table 2A comparative summary of the main advantages and limitations of TTE, CT, and MRI in the assessment of cardiac adipose tissueParameterTTECTMRICostLowModerateHighRadiation exposureNoneYes (ionizing radiation)NoneSpatial resolutionLimitedExcellentGood (inferior to CT)Tissue characterizationLimitedGood (attenuation-based; indirect assessment of inflammation)Excellent (fat-water separation and tissue composition)QuantificationThickness measurement (semi-quantitative)Accurate volumetric and attenuation quantificationAccurate volumetric and qualitative assessmentAvailabilityWidely availableWidely availableVariable (center-dependent)Operator dependenceHighLowLowAcquisition timeShortShortLongerMain advantagesLow cost; bedside availability; useful for screeningHigh spatial resolution; pericoronary fat assessmentNo ionizing radiation; superior tissue characterizationMain limitationsLimited reproducibility; lack of volumetric assessmentRadiation exposure; use of contrast agents in CCTAHigh cost; longer acquisition time; limited availabilityCCTA, coronary computed tomography angiography; CT, computed tomography; MRI, magnetic resonance imaging; TTE, transthoracic echocardiography. A comparative summary of the main advantages and limitations of TTE, CT, and MRI in the assessment of cardiac adipose tissue CCTA, coronary computed tomography angiography; CT, computed tomography; MRI, magnetic resonance imaging; TTE, transthoracic echocardiography. Clinical correlations of mediastinal fat CAD In recent years, a consistent body of evidence has shown that both the quantity and quality of epicardial and pericoronary fat are associated with coronary atherosclerosis and its clinical manifestations. Several studies have shown that increased epicardial or pericardial fat volume mirrors cardiometabolic derangements and correlates with the presence and severity of CAD. Dey et al. reported that pericardial fat volume quantified on non-contrast CT was strongly associated with coronary calcium burden and metabolic syndrome, representing a reproducible marker of visceral adiposity and metabolic risk (33). Similar associations were observed by Aydın et al., Gorter et al., Kazemi et al., and Sunil Kumar et al., who consistently demonstrated that higher epicardial or pericoronary fat thickness or volume was linked to greater coronary plaque burden, stenosis severity, and the presence of multiple cardiovascular risk factors (34-37). Overall, these studies indicate that greater epicardial and pericardial fat burden is consistently associated with more severe coronary atherosclerotic disease. In parallel, Soliman et al. identified a pericardial fat volume threshold predictive of CAD, while Khidr et al. showed that increased epicardial fat measured by cardiac MRI was associated with more complex coronary disease, particularly among diabetic patients (24,38). Beyond volumetric burden, a second major line of evidence highlights the role of PCAT attenuation as a noninvasive marker of local coronary inflammation. Konishi et al. first demonstrated that pericoronary fat density reflects inflammatory changes around the vessel wall, a concept reinforced by subsequent CT studies using advanced techniques such as dual-layer spectral imaging, which identified higher PCAT attenuation around high-risk plaques (39,40). Goeller et al. linked elevated RCA PCAT attenuation to pro-inflammatory cytokines and showed that it independently predicted major adverse cardiovascular events during long-term follow-up (41). The seminal work by Antonopoulos et al. provided mechanistic validation, showing that inflamed coronary vessels modulate adipocyte lipid content in adjacent perivascular adipose tissue, enabling the development of the FAI as a sensitive marker of vascular inflammation (3). Multiple clinical studies, including those by Jing et al., Yang et al., and Ma et al., confirmed that elevated PCAT attenuation or lesion-specific FAI correlates with plaque vulnerability, instability, and the presence of culprit lesions in non-ST-segment elevation acute coronary syndromes (ACSs), outperforming traditional stenosis-based metrics (42-44). Importantly, both epicardial fat volume and attenuation have demonstrated significant prognostic relevance. Mahabadi et al. showed that increased EAT volume and attenuation are independently associated with type-I myocardial infarction, while Du et al. reported that lesion-specific EAT volume predicts myocardial ischemia beyond the degree of luminal stenosis (45,46). Extending these findings to non-obstructive CAD, Zheng et al. demonstrated that PCAT attenuation independently predicts major adverse cardiovascular events and improves risk stratification beyond conventional clinical and plaque features (47). Similar associations were observed in patients with myocardial infarction with non-obstructive coronary arteries and Takotsubo syndrome, where Gaibazzi et al. found elevated pericoronary fat attenuation as a marker of coronary inflammation, and in longitudinal analyses by Zhang et al., who reported that higher baseline pericoronary FAI predicts future plaque formation and progression toward necrotic core-rich morphology (48,49). Complementing CT-based studies, Morales-Portano et al. showed that echocardiographic assessment of EAT thickness is a strong predictor of adverse outcomes in CAD, highlighting the prognostic utility of both volumetric and thickness-based fat measurements across imaging modalities (50). Collectively, these studies converge on a unified concept: cardiac adipose tissue is not merely a passive fat depot but an active biomarker of coronary atherosclerosis, vascular inflammation, and adverse cardiovascular risk. Quantitative measures of epicardial and pericoronary fat provide complementary information to stenosis severity and plaque morphology, and emerging CT-based indices, particularly PCAT attenuation and the FAI, hold promise for refined risk stratification, early detection of high-risk coronary lesions, and improved understanding of the inflammatory mechanisms underpinning CAD. Beyond its role in stable CAD, EAT is increasingly recognized as a contributor to ACSs, where it may enhance local inflammatory responses and influence post-infarction remodeling processes (51). The main imaging studies investigating the relationship between mediastinal adipose tissue and CAD are summarized in Table 3. Table 3Summary of key imaging studies investigating the role of epicardial, pericoronary, and PAT in CAD, including study design, patient population, imaging modality, and main clinical and prognostic findingsStudy, yearStudy designPatient populationImaging modalityMain findingsDey et al., 2010 (33)Observational201 patients; mean age 56±11 years; 45% femaleNon-contrast CTPericardial fat volume was significantly associated with coronary calcium score and metabolic syndrome, and correlated with visceral adiposity and metabolic risk factorsAydın et al., 2015 (34)Cross-sectionalPatients with and without CADCoronary CT angiographyPericoronary EAT thickness was higher in CAD and associated with cardiometabolic risk factors and disease severityGorter et al., 2008 (35)Observational60 patients with suspected CADCardiac CTQuantification of epicardial and pericoronary fat showed good reproducibility and was associated with obesity and metabolic syndromeMahabadi et al., 2010 (10)Observational (segment-based)Coronary artery segmentsCardiac CTPericoronary fat volume was associated with local atherosclerotic plaque burden in the corresponding coronary segmentsKonishi et al., 2010 (39)ObservationalPatients with ACS or suspected CADCTPericardial fat inflammation was associated with the presence of CADChen et al., 2021 (40)Observational104 patients with CADDual-layer spectral CTPCAT attenuation was higher around high-risk plaques; low-keV reconstructions improved plaque characterizationGoeller et al., 2021 (41)Observational/prognosticPatients undergoing coronary CTACoronary CT angiographyPCAT attenuation was associated with inflammatory biomarkers, coronary calcification, and adverse cardiovascular eventsJing et al., 2024 (43)Observational428 patientsCoronary CT angiographyIncreased pericoronary fat attenuation was associated with greater CAD severityAntonopoulos et al., 2017 (3)Translational/prospectiveClinical and validation cohortsCoronary CTA and ex vivo analysisCT-derived perivascular fat attenuation reflected coronary inflammation and enabled detection of high-risk coronary diseaseKhidr et al., 2025 (24)ObservationalPatients with CAD and diabetesCardiac MRIEpicardial fat volume quantified by MRI was associated with greater CAD complexityMorales-Portano et al., 2018 (50)ProspectivePatients with CADEchocardiographyEAT thickness was associated with CAD severity and predicted adverse cardiovascular outcomesACS, acute coronary syndrome; CAD, coronary artery disease; CTA, computed tomography angiography; CT, computed tomography; EAT, epicardial adipose tissue; MRI, magnetic resonance imaging; PCAT, pericoronary adipose tissue; PAT, pericardial adipose tissue. Summary of key imaging studies investigating the role of epicardial, pericoronary, and PAT in CAD, including study design, patient population, imaging modality, and main clinical and prognostic findings ACS, acute coronary syndrome; CAD, coronary artery disease; CTA, computed tomography angiography; CT, computed tomography; EAT, epicardial adipose tissue; MRI, magnetic resonance imaging; PCAT, pericoronary adipose tissue; PAT, pericardial adipose tissue. AF Different studies have highlighted the pivotal role of dysfunctional EAT in atrial remodeling and the development of AF. EAT secretes pro-inflammatory cytokines such as interleukin (IL)-1β, IL-6, and tumor necrosis factor-α (TNF-α), which reach myocardial tissue via paracrine or vasocrine pathways, promoting atrial structural and electrical remodeling and creating a substrate for AF (52). Umar et al. found that EAT volume is significantly greater in AF patients even in the absence of left atrial (LA) fibrosis, supporting a direct role of EAT in AF pathogenesis (53). Several studies have shown a direct association between increased pericardial volume and the presence of AF. In a meta-analysis, Al-Makhamreh et al. demonstrated that patients with AF consistently exhibit higher pericardial fat volumes than healthy controls, and that pericardial fat is independently associated with AF (54). Evidence also indicates that increased atrial fat contributes to AF progression. van Rosendael et al. reported that LA-EAT is elevated in paroxysmal AF compared with sinus rhythm, but does not further increase in persistent AF, whereas LA volume progressively enlarges with disease severity, suggesting that LA-EAT accumulation may be an early marker of AF (55). Teixeira et al. further showed that higher EAT volume predicts recurrence of AF after catheter ablation, highlighting its role in disease persistence (56). Recent systematic reviews and meta-analyses indicate a potential association between increased peri-LA-EAT burden, as well as higher LA-EAT attenuation on CT, and AF recurrence after catheter ablation. Nevertheless, findings on EAT attenuation remain heterogeneous and should be interpreted with caution (57-59). Beyond volume, CT-derived EAT characteristics such as density or attenuation provide additional prognostic information. Gaibazzi et al. found that peri-atrial EAT density independently predicted AF, outperforming volume alone, while Huber et al. demonstrated that lower EAT attenuation was associated with atrial remodeling and recurrence after ablation, reflecting metabolically active, pro-arrhythmic tissue (48,60). Cohen-Dor et al. extended these findings using radiomic analysis, showing that LA-EAT texture features can non-invasively identify AF (61). Cardiac MRI studies provide complementary insights into the atrial substrate. Chahine et al. reported that LA fibrosis and EAT independently impair LA function, with EAT negatively correlating with conduit strain, while Nakamori et al. confirmed that LA-epicardial fat is higher in AF patients and that combining LA fat measurements with LA volume improves detection of AF beyond structural remodeling alone (62,63). Skoda et al. highlighted that the spatial overlap in the atrial wall between LA-EAT and fibrosis is limited and does not correlate with AF stage, suggesting distinct mechanisms of interaction between these two substrate (64). Sex-specific differences have also been reported. Zhu et al. found that postmenopausal women have lower total EAT but higher periatrial-to-total (P/T) EAT ratios compared to men leading to the following conclusion: female sex and P/T EAT ratio were identified as independent predictors of AF recurrence, while age and periatrial EAT volume independently predicted major adverse cardiovascular events (similar in both sexes) (65). TTE also provides clinically useful measurements. Chao et al. demonstrated that EAT thickness predicts AF recurrence after ablation, with distinct cutoffs for paroxysmal and non-paroxysmal AF, offering a low-cost tool for patient risk assessment (66). Collectively, these findings support a multifaceted role of cardiac fat in AF development, progression, and recurrence, with both volumetric and qualitative imaging markers serving as potential tools for risk stratification and guiding therapeutic interventions. Key imaging studies evaluating the association between mediastinal adipose tissue and AF are summarized in Table 4. Table 4Overview of imaging studies assessing the association between cardiac adipose tissue and AF, focusing on volume, distribution, and qualitative characteristics of epicardial and periatrial fat and their relationship with AF onset, progression, and recurrenceStudy, yearType of studyPatient populationImaging techniqueKey findings/resultsvan Rosendael et al., 2022 (55)Prospective cohort300 patients: 100 sinus rhythm, 100 paroxysmal AF, 100 persistent/permanent AFCardiac CTLA-EAT mass was higher in paroxysmal AF compared with sinus rhythm but did not further increase in persistent/permanent AF. LA volume increased progressively with AF severity, suggesting LA-EAT as an early marker and LA volume as a marker of advanced diseaseGaibazzi et al., 2019 (48)Case-control160 patients: 80 AF, 80 controlsCoronary CTALA fat volume was higher in AF (5.42±2.94 vs. 4.16±2.55 mL, P=0.007), but not after adjustment for LA size. LA fat attenuation was significantly higher in AF (−69.15 vs. −76.82 HU, P<0.0001) and independently predicted AF (AUC =0.829 vs. 0.775)Huber et al., 2024 (60)ObservationalPatients undergoing pulmonary vein isolationCardiac CTLower EAT attenuation was associated with LA structural and electrical remodeling and AF recurrence after ablation, reflecting adipocyte hypertrophy and pro-arrhythmic EATCohen-Dor et al., 2025 (61)Retrospective matched case-control280 patients: 120 AF, 160 controlsNon-contrast ECG-gated cardiac CTRadiomic features of LA-EAT accurately predicted AF using machine learning, supporting LA-EAT radiomics as a non-invasive biomarker for early detection and risk stratificationUmar et al., 2025 (53)ProspectiveAF patients without LA fibrosis (low-voltage zones) and matched controlsCardiac CTEAT volume was greater in AF patients even in the absence of LA fibrosis, supporting a direct role of EAT in AF pathogenesis independent of structural remodelingTeixeira et al., 2024 (56)Prospective cohort305 AF patients; mean age 57.5 years; 24-month follow-upCardiac CTEAT volume predicted AF recurrence after ablation; patients with EAT >92 cm3 had nearly double recurrence risk (HR =1.95, P=0.008)Zhu et al., 2022 (65)ProspectivePostmenopausal women and matched men undergoing first ablationCardiac CTWomen had lower total EAT but higher P/T EAT ratio. Female sex and P/T EAT ratio independently predicted AF recurrence; periatrial EAT predicted major adverse cardiovascular eventsChahine et al., 2024 (62)Prospective101 AF patientsCMR (Dixon fat-water separation + LGE)LA fibrosis was negatively associated with global longitudinal strain, while LA-EAT negatively correlated with conduit strain. Both correlated with LA volume, but only fibrosis was associated with reduced LA emptying fraction. EAT contributes independently to LA functional impairmentSkoda et al., 2024 (64)Observational42 AF patients, 9 controls3D Dixon LGE-CMRSpatial overlap between LA-EAT and fibrosis was limited and not associated with AF stage, suggesting distinct or indirect mechanisms of interactionNakamori et al., 2018 (63)Prospective105 patients: 53 AF, 52 controls3D multi-echo Dixon CMRAF patients had higher LA-EAT (28.9 vs. 14.2 mL) and larger LA volume (110.8 vs. 89.7 mL). LA-EAT was independently associated with AF (OR =1.42/mL), and its combination with LA volume improved detection (c-statistic 0.88 vs. 0.74)Chao et al., 2013 (66)Prospective283 AF patients: 227 paroxysmal, 56 non-paroxysmalTTENon-paroxysmal AF patients had greater EAT thickness (7.0 vs. 5.9 mm). AF recurrence was 33.6% over 16 months. EAT thickness independently predicted recurrence; cutoffs of 6 mm (paroxysmal AF) and 6.9 mm (non-paroxysmal AF) were useful for risk stratificationAF, atrial fibrillation; AUC, area under the curve; CMR, cardiovascular magnetic resonance; CT, computed tomography; CTA, computed tomography angiography; EAT, epicardial adipose tissue; ECG, electrocardiogram; HR, hazard ratio; HU, Hounsfield units; LA, left atrium; LGE, late gadolinium enhancement; OR, odds ratio; P/T, periatrial-to-total; TTE, transthoracic echocardiography. Overview of imaging studies assessing the association between cardiac adipose tissue and AF, focusing on volume, distribution, and qualitative characteristics of epicardial and periatrial fat and their relationship with AF onset, progression, and recurrence AF, atrial fibrillation; AUC, area under the curve; CMR, cardiovascular magnetic resonance; CT, computed tomography; CTA, computed tomography angiography; EAT, epicardial adipose tissue; ECG, electrocardiogram; HR, hazard ratio; HU, Hounsfield units; LA, left atrium; LGE, late gadolinium enhancement; OR, odds ratio; P/T, periatrial-to-total; TTE, transthoracic echocardiography. HF Epicardial, pericardial, and pericoronary adiposity appears to play an important role in the development, progression, and prognosis of HF, particularly HF with preserved ejection fraction (HFpEF). Large cohort studies have shown that higher pericardial fat volume independently predicts incident HF. In the Multi-Ethnic Study of Atherosclerosis, elevated pericardial fat volume was associated with a 44% higher HF risk in women and 13% in men, with thresholds ≥70 cm3 in women and ≥120 cm3 in men marking significantly increased risk, particularly for HFpEF (67). Higher PCAT attenuation and pericoronary FAI are independently associated with HFpEF and future HF hospitalizations, suggesting that inflamed cardiac fat directly promotes myocardial dysfunction (68,69). CMR studies demonstrate that HFpEF patients exhibit increased EAT volumes, which correlate with impaired biventricular function, adverse remodeling, and metabolic comorbidities (70-72). Regional EAT accumulation is independently linked to functional impairment of adjacent chambers and elevated N-terminal pro-B-type natriuretic peptide (NT-proBNP), suggesting a direct local effect on diastolic dysfunction in HFpEF (73). In contrast, in HF with reduced ejection fraction (HFrEF), lower EAT volume and higher density are associated with HF with improved ejection fraction, indicating that both the amount and quality of cardiac fat influence myocardial performance (44). Echocardiography and CT-derived metrics further highlight the prognostic relevance of EAT. Increased EAT thickness predicts HFpEF progression and impaired diastolic function, independent of BMI or chamber size, while reductions in EAT, for instance following metabolic or bariatric interventions, correlate with functional improvement (73-75). Overall, these findings underscore a multifaceted role of cardiac adiposity in HF, where volumetric, qualitative, and inflammatory properties of epicardial and pericardial fat serve as key determinants of disease risk, phenotype, and potential therapeutic targets. The principal imaging studies assessing the role of cardiac adipose tissue in HF are summarized in Table 5. Table 5Main evidence on the role of epicardial, pericardial, and PCAT in HF, with particular emphasis on HFpEF, reporting imaging findings, pathophysiological associations, and prognostic implicationsStudy, yearType of studyPatient populationImaging techniqueKey findings/resultsKenchaiah et al., 2021 (67)Prospective cohort6,785 participants from MESA without pre-existing CVDCardiac CTHigher pericardial fat volume was associated with increased HF risk over 15.7 years. For each 1-SD increase, HF risk increased by 44% in women and 13% in men. Elevated pericardial fat volume (≥70 cm3 in women, ≥120 cm3 in men) was associated with approximately doubled HF risk in women and a 53% increase in men. Pericardial fat volume was a stronger predictor of HFpEF than HFrEF and remained independently associated with outcomes after adjustment for body fat distribution and inflammatory biomarkersNishihara et al., 2023 (68)Observational607 outpatients without obstructive CADCardiac CTHigher PCAT attenuation was associated with HFpEF across all coronary arteries, supporting a link between pericoronary inflammation and HFpEF pathophysiologyNakashima et al., 2025 (69)Retrospective cohort1,196 patients undergoing coronary CTACardiac CTHigher FAI values in the LAD (≥63.4 HU; HR =4.8) and LCX (≥61.6 HU; HR =4.5) independently predicted HF hospitalizations in patients with HFpEF over a median follow-up of 4.3 years, supporting pericoronary inflammation as a prognostic markerSchulz et al., 2024 (73)Observational68 symptomatic patients with diastolic dysfunctionCardiac MRIHigher EAT volume was associated with elevated HFpEF diagnostic scores, higher NT-proBNP levels, and a greater prevalence of HFpEF (64% vs. 37%). Regional EAT accumulation was linked to functional impairment of adjacent chambers, suggesting a local effect on diastolic dysfunctionSong et al., 2025 (70)Observational104 HFrEF, 226 HFpEF, and 172 control subjectsCardiac MRIPatients with HFpEF had higher EAT volume (51±21 mL) than those with HFrEF (32±14 mL) and controls (33±19 mL). In HFpEF, increased EAT was associated with worse biventricular function, whereas in HFrEF, higher EAT correlated with better biventricular function, suggesting phenotype-specific effectsMenghoum et al., 2025 (71)Cohort104 HFpEF, 16 pre-HF, and 26 controlsCardiac MRIEAT volume was significantly higher in HFpEF and was associated with metabolic comorbidities, impaired biventricular function, and adverse remodelingWang et al., 2024 (72)Multicenter cohort692 patients with HF with mid-range EF or HFpEFCardiac MRIHigher EAT volume was independently associated with worse outcomes (HR =1.62, 95% CI: 1.42–1.86; P<0.001) and improved the predictive performance of multivariable models including comorbidities, NYHA class, extracellular volume, LVEF, BMI, age, and NT-proBNPDhore-Patil et al., 2024 (74)ReviewPatients with obesity-related HFpEFEchocardiographyEAT thickness increases as HFpEF progresses. TTE-based assessment may help evaluate the contribution of EAT to coronary disease progression and impaired LV diastolic function via reduced subendocardial perfusionSchulz et al., 2024 (73)Observational68 patients with diastolic dysfunction and preserved EFEchocardiography + right heart catheterization + cardiac MRIHigher EAT volumes were associated with a greater prevalence of HFpEF (64% vs. 37%) and with impaired atrial and ventricular strain, particularly under stress. Regional EAT was independently linked to functional impairment of adjacent chambersTanahashi et al., 2025 (75)Prospective75 severely obese patients undergoing sleeve gastrectomyEchocardiographyBaseline EAT was independently associated with diastolic dysfunction, and reductions in EAT after surgery correlated with functional improvement, supporting a potential causal role of EAT in cardiac dysfunction1-SD, one standard deviation; BMI, body mass index; CAD, coronary artery disease; CI, confidence interval; CT, computed tomography; CTA, computed tomography angiography; CVD, cardiovascular disease; EAT, epicardial adipose tissue; EF, ejection fraction; FAI, fat attenuation index; HF, heart failure; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; HR, hazard ratio; HU, Hounsfield units; LAD, left anterior descending artery; LCX, left circumflex artery; LV, left ventricular; LVEF, left ventricular ejection fraction; MESA, Multi-Ethnic Study of Atherosclerosis; MRI, magnetic resonance imaging; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; PCAT, pericoronary adipose tissue; TTE, transthoracic echocardiography. Main evidence on the role of epicardial, pericardial, and PCAT in HF, with particular emphasis on HFpEF, reporting imaging findings, pathophysiological associations, and prognostic implications 1-SD, one standard deviation; BMI, body mass index; CAD, coronary artery disease; CI, confidence interval; CT, computed tomography; CTA, computed tomography angiography; CVD, cardiovascular disease; EAT, epicardial adipose tissue; EF, ejection fraction; FAI, fat attenuation index; HF, heart failure; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; HR, hazard ratio; HU, Hounsfield units; LAD, left anterior descending artery; LCX, left circumflex artery; LV, left ventricular; LVEF, left ventricular ejection fraction; MESA, Multi-Ethnic Study of Atherosclerosis; MRI, magnetic resonance imaging; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; PCAT, pericoronary adipose tissue; TTE, transthoracic echocardiography. Lipomatosis of the interatrial septum Lipomatous hypertrophy of the interatrial septum (LHIS) is a benign fatty infiltration of the interatrial septum characterized by a dumbbell- or hourglass-shaped morphology due to sparing of the fossa ovalis, typically diagnosed when septal thickness exceeds 2 cm (Figure 3) (76,77). Histologically, LHIS consists of a non-encapsulated proliferation of adipose tissue with entrapped and hypertrophied myocytes, often including fetal fat cells, distinguishing it from true lipomas, which are encapsulated and composed of mature adipocytes. The lesion commonly extends from the coronary sinus to the aortic root (78). Lacaita et al. suggested that LHIS may arise from EAT, yet represents a distinct fat compartment with higher CT density than EAT, paracardial, and PCAT; [18F] fluorodeoxyglucose positron emission tomography (FDG-PET) showed tracer uptake in 83.3% of patients, consistent with brown fat characteristics, suggesting LHIS may behave differently from conventional EAT (79). Kılıçkap et al. also highlighted that LHIS may form part of a broader metabolic adipose phenotype, as it frequently coexists with increased epicardial and visceral fat in obese individuals (76). Axial non-contrast chest CT image at the level of the atria showing LHIS, a benign fatty infiltration of the interatrial septum. The yellow ROI highlights adipose tissue within the interatrial septum, characterized by low attenuation values on non-contrast CT. This fatty infiltration produces a characteristic thickened, dumbbell- or hourglass-shaped morphology due to sparing of the fossa ovalis, consistent with interatrial septal lipomatosis. CT, computed tomography; LHIS, lipomatous hypertrophy of the interatrial septum; ROI, region of interest. Imaging plays a central role in diagnosing LHIS. Echocardiography (TTE and transesophageal echocardiography) typically reveals bilobar septal thickening with sparing of the fossa ovalis, producing the characteristic dumbbell shape (78,80). CT shows a non-enhancing, smoothly marginated, homogeneous fat-attenuation mass, often measuring 6–9 cm craniocaudally, 3.6–6.2 cm along the septum, and 1.5–4.8 cm perpendicular to it. In 10 out 12 patients, also epicardial fat is frequently increased (81). Cardiac MRI provides the most detailed tissue characterization, identifying fatty septal masses and any entrapped or hypertrophied myocytes, with fat-suppressed sequences confirming the adipose content (78). In some cases, atypical MRI signal intensities may be observed due to variations in myocyte hypertrophy, reflecting the spectrum of LHIS appearances rather than pathology. Although most patients with LHIS remain asymptomatic, the condition can occasionally be associated with atrial arrhythmias, including AF, supraventricular tachyarrhythmias, and junctional rhythms. Proposed mechanisms include disruption of atrial myocyte architecture and conduction pathways due to infiltration of the interatrial septum and right atrial wall, bleeding into the lesion, or coexistent CAD in typically elderly, obese patients. Septal thickness and lesion location, particularly near the crista terminalis and superior vena cava (SVC)-right atrium (RA) junction, may increase arrhythmic risk, and in selected cases, surgical resection has resolved arrhythmias. Rarely, LHIS may also cause complications such as SVC obstruction, HF, or pericardial effusion (80,82). Overall, LHIS represents a distinct benign adipose phenotype within the heart, often incidental but occasionally clinically significant, with imaging modalities providing accurate characterization and differentiation from other cardiac masses. From a diagnostic perspective, it is important to distinguish cardiac adipose tissue from other fat-containing cardiac conditions. Cardiac lipomas are rare benign tumors that most commonly present with cardiomegaly and appear on imaging as well-defined, nonmobile, homogeneous fatty masses, whose relationship with the coronary arteries is crucial for surgical planning (83). In addition, higher amounts of right and left ventricular epicardial fat are observed in arrhythmogenic right ventricular dysplasia/cardiomyopathy, particularly adjacent to the left ventricle, correlating with disease severity and aiding in differentiation from healthy individuals (84). Therapeutic implications The recognition of EAT as an active endocrine and inflammatory depot has important therapeutic implications. Beyond its value as an imaging biomarker, EAT is increasingly regarded as a modifiable target, because reductions in its volume and inflammatory activity may reflect or contribute to improved cardiometabolic risk profiles. Therapeutic strategies aimed at EAT include pharmacological treatment and lifestyle-based interventions (85). Among pharmacological approaches, SGLT2 inhibitors have shown consistent associations with reductions in EAT thickness or volume, together with improvements in metabolic and inflammatory parameters. Recent reviews suggest that these effects may be partly mediated by direct activity on EAT, where SGLT2 expression has been reported, as well as by indirect mechanisms including weight loss, improved insulin sensitivity, natriuresis, and attenuation of adipose tissue inflammation (86). GLP-1 receptor agonists also represent a promising strategy. Human EAT expresses incretin-related receptors, including GLP-1 receptor agonists, providing a biological rationale for a direct response to incretin-based therapies. Transcriptomic data suggest that GLP-1R activation may promote a metabolically favorable phenotype (e.g., reduced adipogenesis and enhanced fatty acid oxidation) (87). Lifestyle interventions remain a cornerstone of EAT modulation. Caloric restriction, weight loss, and structured exercise programs have been associated with reductions in EAT burden, although the magnitude of change may be variable. Nevertheless, lifestyle measures are clinically important because they address the broader metabolic milieu that promotes EAT expansion and dysfunction, and they should be considered complementary to pharmacological therapy rather than an alternative. Overall, current evidence supports the concept that EAT is not only a marker of cardiovascular risk but also a potentially modifiable therapeutic target. However, further studies are needed to clarify whether treatment-induced changes in EAT translate into incremental reductions in cardiovascular events beyond those achieved through overall weight loss and risk-factor control (88). Future directions Future directions include the integration of artificial intelligence (AI) and radiomics for automated and reproducible EAT quantification. These approaches enable extraction of quantitative features, including texture and potentially spatial characteristics, providing additional information beyond conventional volumetric and attenuation-based metrics. AI-driven analysis may improve risk stratification and support large-scale studies and personalized cardiovascular risk assessment, although further validation and standardization are required before widespread clinical implementation (89-91). Conclusions Mediastinal adipose tissue is an active and clinically relevant component of cardiovascular disease. Epicardial and pericoronary fat are consistently associated with coronary atherosclerosis, AF, and HF, particularly HFpEF. Multimodality imaging enables their reliable assessment, with CT providing quantitative and inflammatory markers, MRI offering advanced tissue characterization, and echocardiography supporting accessible clinical evaluation. Cardiac adipose tissue therefore emerges as both a biomarker of cardiovascular risk and a potential therapeutic target. Standardization of imaging approaches and validation of treatment-related changes remain key priorities. Supplementary The article’s supplementary files as Acknowledgments None. Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Footnotes Reporting Checklist: The authors have completed the Narrative Review reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2845/rc Funding: None. Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1-2845/coif). C.A.M. serves as an unpaid editorial board member of Quantitative Imaging in Medicine and Surgery. The other authors have no conflicts of interest to declare. References Associated Data Supplementary Materials The article’s supplementary files as