Associations between the serum triglyceride-glucose index and pericoronary adipose tissue attenuation and plaque features using dual-layer spectral detector computed tomography: a cross-sectional study.

Abstract

The triglyceride-glucose (TyG) index is a reliable alternative marker for insulin resistance (IR). Pericoronary adipose tissue (PCAT) can indirectly reflect coronary inflammation. IR and coronary inflammation play a key role in the development and progression of coronary atherosclerosis. Therefore, this study investigated the relationships between the TyG index, PCAT and atherosclerotic plaque characteristics to explore whether IR might lead to coronary artery atherosclerosis progression by inducing coronary inflammation. We retrospectively collected data on patients with chest pain who underwent coronary computed tomography angiography using spectral detector computed tomography at our institution from June to December 2021. The patients were grouped based on their TyG index levels: T1 (low), T2 (medium), and T3 (high). Each patient was assessed for total plaque volume, plaque load, maximum stenosis, the plaque component volume proportion, high-risk plaques(HRPs), and plaque characteristics (including low attenuation plaques, positive remodeling, a napkin ring sign, and spot calcification). PCAT quantification was performed on the proximal right coronary artery using the fat attenuation index (FAI) measured from a conventional multicolor computed tomography image (FAI120kVp), a spectral virtual single-energy image (FAI40keV), and the slope of the spectral HU curve (λHU). We enrolled 201 patients. The proportion of patients with maximum plaque stenosis, positive remodeling, low-density plaques, and HRPs increased as the TyG index level increased. Moreover, the FAI40keV and λHU significantly differed among the three groups, and we identified good positive correlations between FAI40keV and λHU and the TyG index (r = 0.319, P <0.01 and r = 0.325, P <0.01, respectively). FAI120kVp did not significantly differ among the groups. FAI40keV had the highest area under the curve, with an optimal cutoff value of -130.5 HU for predicting a TyG index value of ≥9.13. The multivariate linear regression analysis demonstrated that FAI40keV and λHU were independently positively related to a high TyG index level (standardized regression coefficients: 0.117 [P <0.001] and 0.134 [P <0.001], respectively). Patients with chest pain and a higher TyG index level were more likely to have severe stenosis and HRPs. Moreover, FAI40keV and λHU had good correlations with the serum TyG index, which may noninvasively reflect PCAT inflammation under insulin resistance. These results could help explain the mechanism of plaque progression and instability in patients with insulin resistance might be related to IR-induced coronary inflammation.