Prognostic value of epicardial adipose tissue volume in combination with coronary plaque and flow assessment for the prediction of major adverse cardiac events. Issue 148 (March 2022)

• Record Type: Journal Article
• Authors: Brandt, Verena; Bekeredjian, Raffi; Schoepf, U. Joseph; Varga-Szemes, Akos; Emrich, Tilman; Aquino, Gilberto J.; Decker, Josua; Bayer, Richard R.; Ellis, Lauren; Ebersberger, Ullrich; Tesche, Christian
• Is Part Of: European journal of radiology. Issue 148(2022)
• Publication Date: 2022-03
• Subjects: Spiral computed tomography -- Coronary artery disease -- Epicardial adipose tissue; AI Artificial intelligence -- AUC Area under the curve -- CAC Coronary artery calcium -- CAD Coronary artery disease -- CCTA Coronary computed tomography angiography -- CT-FFR CT-derived fractional flow reserve -- EAT Epicardial adipose tissue -- HR Hazard Ratio -- ICA Invasive coronary angiography -- MACE Major adverse cardiac events -- MDCT Multidetector computed tomography -- ML Machine Learning -- NPV Negative predictive value -- PCAT Pericoronary adipose tissue -- PPV Positive predictive value -- ROC Receiver-operating characteristics; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2022.110157
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: CT-derived EAT assessment demonstrates high discriminatory power to predict MACE. EAT shows superior diagnostic performance over Morise score, CCTA-derived plaque measures and CT-FFR. A combined model of these markers demonstrated incremental MACE prediction beyond clinical risk score. Abstract: Purpose: The purpose of this study was to determine whether EAT volume in combination with coronary CT angiography (CCTA)-derived plaque quantification and CT-derived fractional flow reserve (CT-FFR) has prognostic implication with major adverse cardiac events (MACE). Methods: Patients ( n = 117, 58 ± 10 years, 61% male) who had previously undergone invasive coronary angiography (ICA) and CCTA were retrospectively analyzed. Follow-up was performed to record MACE. EAT volume and plaque measures were derived from non-contrast and contrast-enhanced CT images using a semi-automatic software approach, while CT-FFR was calculated using a machine-learning algorithm. The diagnostic performance to identify MACE was evaluated using univariable and multivariable Cox proportional hazards analysis and concordance (C)-indices. Results: During a median follow-up period of 40.4 months, 19 events were registered. EAT volume, CCTA ≥ 50% stenosis, and CT-FFR were significantly different in patients developing MACE (all p < 0.05). The following parameters were predictors of MACE in adjusted multivariable Cox regression analysis (hazard ratio [HR]): EAT volume (HR 2.21, p = 0.023), indexed EAT volume (HR 2.03, p = 0.035), and CCTA ≥ 50% (HR 1.05, p = 0.048). A model including Morise score, CCTA ≥ 50% stenosis, and EAT volume showed significantly improved C-index to Morise score alone (AUC 0.83 vs. 0.66, p = 0.004). Conclusions: Facing limitations in conventional cardiovascular risk scoring models, this observational study demonstrates that the prediction performance of our proposed method achieves a significant improvement in prognostic ability, especially when compared to models such as Morise score alone or its combination with CCTA and CT-FFR.