Abstract
OBJECTIVE: To investigate the incremental value of pericoronary fat attenuation index (FAI) in routine coronary artery computed tomography angiography (CCTA) to identify culprit lesions in acute coronary syndrome (ACS). METHODS: We reviewed the CCTA data from 80 ACS patients and 40 individuals with stable coronary atherosclerosis. ACS patient plaques were categorized into culprit and nonculprit groups. The plaque-specific pericoronary FAI was assessed using the Perivascular Fat Analysis Tool. We applied a default prespecified window of -190 to -30 Hounsfield units (HU) and a broader prespecified window of -190 to 20 HU. FAI values within these prespecified windows and the types and severity of plaque stenosis were compared across the 3 groups. Additionally, we investigated high-risk characteristics of plaques in the ACS group and their correlation with FAI. The effectiveness and worthiness of FAI in identifying culprit lesions were analyzed based on the receiver operating characteristic curve. RESULTS: The FAI values under the 2 prespecified windows were higher in the culprit group than in the nonculprit and control groups (all P < 0.001). The culprit group showed the most mixed plaques and the most severe stenosis (all P < 0.001). In the ACS group, the FAI value was significantly lower around calcified lesions (-85.00 ± 9.97 HU) than around noncalcified (-78.00 ± 11.52 HU) and mixed plaques (-78.00 ± 9.24 HU) (both P < 0.001). The culprit group had more high-risk plaques, and high-risk plaques had higher FAI values than those without high-risk characteristics (-70.00 ± 7.67 HU vs -82.00 ± 10.16 HU, P < 0.001). The efficacy of FAI under the default prespecified window in identifying culprit lesions was higher compared than that under the broader prespecified window (area under the curve = 0.799 vs 0.761, P = 0.042), and the diagnostic cutoff values were -77 versus -58 HU. The FAI under the default prespecified window exhibited an incremental value for identifying culprit lesions, as compared with stenosis severity (area under the curve = 0.970 vs 0.939, P < 0.001). CONCLUSION: The culprit lesions have higher FAI than the nonculprit lesions and the controls. FAI is a worthy parameter for identifying culprit lesions in routine CCTA according to stenosis severity, and the default prespecified window is a better option.