Cardiac adipose tissue volume assessed by computed tomography is a specific and independent predictor of early mortality and critical illness in COVID-19 in type 2-diabetic patients

Abstract

Background: Patients with type 2-diabetes mellitus (T2D), are characterized by visceral and ectopic adipose tissue expansion, leading to systemic chronic low-grade inflammation. As visceral adiposity is associated with severe COVID-19 irrespective of obesity, we aimed to evaluate and compare the predictive value for early intensive care or death of three fat depots (cardiac, visceral and subcutaneous) using computed tomography (CT) at admission for COVID-19 in consecutive patients with and without T2D.

Methods: Two hundred and two patients admitted for COVID-19 were retrospectively included between February and June 2020 and distributed in two groups: T2D or non-diabetic controls. Chest CT with cardiac (CATi), visceral (VATi) and subcutaneous adipose tissue (SATi) volume measurements were performed at admission. The primary endpoint was a composite outcome criteria including death or ICU admission at day 21 after admission. Threshold values of adipose tissue components predicting adverse outcome were determined.

Results: One hundred and eight controls [median age: 76(IQR:59-83), 61% male, median BMI: 24(22-27)] and ninety-four T2D patients [median age: 70(IQR:61-77), 70% male, median BMI: 27(24-31)], were enrolled in this study. At day 21 after admission, 42 patients (21%) had died from COVID-19, 48 (24%) required intensive care and 112 (55%) were admitted to a conventional care unit (CMU). In T2D, CATi was associated with early death or ICU independently from age, sex, BMI, dyslipidemia, CRP and coronary calcium (CAC). (p = 0.005). Concerning T2D patients, the cut-point for CATi was > 100 mL/m² with a sensitivity of 0.83 and a specificity of 0.50 (AUC = 0.67, p = 0.004) and an OR of 4.71 for early ICU admission or mortality (p = 0.002) in the fully adjusted model. Other adipose tissues SATi or VATi were not significantly associated with early adverse outcomes. In control patients, age and male sex (OR = 1.03, p = 0.04) were the only predictors of ICU or death.

Conclusions: Cardiac adipose tissue volume measured in CT at admission was independently predictive of early intensive care or death in T2D patients with COVID-19 but not in non-diabetics. Such automated CT measurement could be used in routine in diabetic patients presenting with moderate to severe COVID-19 illness to optimize individual management and prevent critical evolution.

Keywords: COVID-19; Cardiac adipose tissue; Computed tomography; Intensive care; Mortality; Type 2 diabetes (T2D).

Fig. 1

Illustration of CT imaging biomarkers assessed in two patients admitted for COVID-19 infection including chest CT for lung severity and cardiac, visceral and subcutaneous abdominal adipose tissue. Top row: A 44-year-old man from the control group presented extensive lung involvement with predominant ground-glass lesions and few consolidations with aeric bronchogram (A, arrows), with a low CATi score (B, red overlay), low VATi (C, red overlay) and low SATi (C, blue overlay). Bottom row: A 79-year-old man with type-2 diabetes shows moderately extensive lung involvement but with predominant consolidation lesions (D, arrows), with a high CATi score (E, red overlay), high VATi (C, red overlay) and high SATi (C, blue overlay). CATi cardiac adipose tissue indexed, VATi visceral adipose tissue indexed, SATi subcutaneous adipose tissue indexed (SATi)

Fig. 2

Determinants of severe short-term outcome in the overall study population

Fig. 3

Comparison of CATi (A), VATi (B), SATi (C) according to the short-term outcomes for diabetes and controls patients. Data are represented as median with interquartile range. CMU conventional medical care unit, ICU intensive care unit, CAT cardiac adipose tissue, VATI visceral adipose tissue, SAT subcutaneous adipose tissue: non significative, **p < 0.005

Fig. 4

Receiver operating curves (ROC) of the different fat depots (CATi, VATi and SATi) for the prediction of ICU or Death in T2D and control patients. For T2D patients, diagnostic performance of CATi (AUC = 0.67, p = 0.0011) and VATi (AUC = 0.62, p = 0.051) was higher compared to SATi (AUC = 0.54, p = 0.57). Adipose tissue imaging biomarkers were not statistically significant for the prediction of ICU or death in non-diabetic patients. AUC area under the curve CATi cardiac adipose tissue index. VATi visceral adipose tissue index. SATi subcutaneous adipose tissue index