Radiol Cardiothorac Imaging. 2026 Jun;8(3):e250404. doi: 10.1148/ryct.250404.
ABSTRACT
Purpose To investigate whether normalizing pericoronary adipose tissue (PCAT) attenuation to epicardial fat attenuation (EAT) reduces susceptibility to CT acquisition settings, particularly tube potential, and improves diagnostic performance for impaired myocardial flow reserve (MFR) and its prognostic value compared with PCAT attenuation in real-world heterogeneous coronary CT angiography (CCTA) protocols. Materials and Methods This retrospective study included patients who underwent CCTA and nitrogen 13 ammonia PET from September 2019 to December 2024. CCTA was performed on three CT systems with varying tube potentials (70-120 kV). PCAT attenuation was semiautomatically measured around the proximal right coronary artery, and EAT attenuation was automatically assessed. A PET-derived MFR of less than 2 was used as the reference standard for impaired MFR. Factors related to PCAT metrics, diagnostic odds ratios (ORs) for impaired MFR using multivariable logistic regression, and exploratory time-to-event analysis of major adverse cardiac events were evaluated. Results The analysis included 74 patients (median age, 72 years [IQR, 60-79 years]; 41 male). PCAT attenuation was affected by the tube potential (P < .001) and electrocardiography gating (P = .01). In contrast, the PCAT/EAT ratio was associated with impaired MFR (P = .01) and coronary stenosis (P = .009). It helped predict an MFR less than 2 (OR per 0.05-decrease, 1.82 [95% CI: 1.16, 3.01]) and showed higher diagnostic performance than did uncorrected PCAT (area under the receiver operating characteristic curve, 0.71 vs 0.56; P = .006). A PCAT/EAT ratio less than 1 showed a hypothesis-generating association with lower event-free survival in patients without severe stenosis (log-rank P = .045). Conclusion The PCAT/EAT ratio reduced scan-related variability and outperformed PCAT attenuation in diagnosing impaired MFR. Keywords: CT Angiography, Coronary Artery Disease, Epicardial Adipose Tissue, Inflammation, Atherosclerosis, Image Processing, Computer-Assisted Supplemental material is available for this article. © RSNA, 2026 See also commentary by Murphy in this issue.
PMID:42165745 | DOI:10.1148/ryct.250404