Prognostic value of non-alcoholic fatty liver disease, pericoronary fat attenuation index and computed tomography-derived fractional flow reserve in diabetic patients with suspected coronary artery disease.

BACKGROUND: This study aimed to investigate the predictive value of non-alcoholic fatty liver disease (NAFLD), pericoronary fat attenuation index (FAI), and computed tomography-derived fractional flow reserve (CT-FFR) for major adverse cardiovascular events (MACE) in diabetic patients with suspected coronary artery disease.

METHODS: This study retrospectively included 466 diabetic patients who underwent coronary computed tomography angiography (CCTA) and non-contrast chest CT from January 2017 to December 2018. The clinical data and imaging parameters of patients were collected. MACE was defined as rehospitalization for unstable angina, non-fatal myocardial infarction, late coronary revascularization, cardiac death, and all-cause mortality.

RESULTS: During a median follow-up of 70 months, 76 patients experienced MACE. NAFLD (hazard ratio [HR] = 2.248; 95 % confidence interval [CI]: 1.362-3.710; P = 0.002), pericoronary FAI > -71.42 HU (HR = 4.063; 95 % CI: 2.316-7.131; P < 0.001), and CT-FFR ≤ 0.80 (HR = 6.023; 95 % CI: 2.567-14.132; P < 0.001) associated with MACE independently. We developed six MACE prediction models: model 1: traditional clinical risk factors; model 2: model 1 + NAFLD; model 3: model 2 + CCTA characteristics; model 4: model 3 + CT-FFR ≤ 0.80; model 5: model 3 + pericoronary FAI > -71.42 HU; model 6: model 5 + CT-FFR ≤ 0.80. The areas under curve of above six models were 0.698, 0.741, 0.861, 0.890, 0.917 and 0.936 respectively.

CONCLUSIONS: NAFLD, pericoronary FAI, and CT-FFR were independent predictors of MACE in diabetic patients. The multiparametric model 6 demonstrated the optimal predictive performance for MACE.