Epicardial adipose tissue thickness in COVID-19 hospitalized patients: a tool for risk stratification

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Coronavirus Disease 2019 (COVID-19) impacted public health systems,overwhelming the intensive care units(ICU)(1).Epicardial adipose tissue (EAT) thickness is a potential novel parameter, which can be assessed using standard computer tomography(CT) for the prediction of worse prognosis in COVID-19(2,3). Purpose We aimed to investigate the association of right ventricle (RV)-EAT thickness with the need for invasive mechanical ventilation,vasopressor support or ICU admission and in-hospital mortality in COVID-19. Methods We analyzed 310 consecutive hospitalized patients with confirmed COVID–19 by RT-PCR, between March and April 2020.EAT thickness was assessed during the acute setting of the disease using low dose non-contrast chest CT.Maximal EAT thickness was determined on axial image series at the level of the RV free wall perpendicular to the surface of the heart(Figure 1).Intra- and interobserver reproducibility for the RV-EAT thickness measurement was assessed in 20 random patients after two weeks,by the primary and a second investigator.Data included demographics,clinical evaluation,comorbidities,treatment and complications.Peak lactate dehydrogenase (LDH), neutrophil-lymphocyte ratio (NLR) and C-reactive protein (CRP) were defined as the highest level during hospitalization.The primary combined endpoint was ICU admission,invasive mechanical ventilation and vasopressor therapy.The secondary outcome was in-hospital mortality. Results Median age was 64 years (interquartile range:53 to 79 years,58.1% males).106(34.2%) patients reached the primary endpoint.In-hospital mortality rate was 19.5% (59 patients).Among patients with combined endpoint,the mortality rate was 35.8% (38 patients).RV-EAT thickness was higher in patients with combined endpoint (5.0 ±2.6 mm versus 4.3 ± 2.2 mm, p = 0.021).Additionally, patients with the composite endpoint had more diabetes mellitus (p = 0.028) and history of coronary artery disease (p = 0.020).Multivariable analysis showed that RV-EAT thickness predicted the primary endpoint,irrespective of risk factors and disease severity (p = 0.014,OR 1.157,95%CI 1.030-1.300; p = 0.031,OR 1.146,95%CI 1.013-1.298,respectively)(Figure 2).Moreover, peak CRP and peak LDH were associated with both endpoints(Figure 2).However, RV-EAT thickness was not predictive for mortality (p = 0.561, OR 1.039, 95%CI 0.913-1.183). Intraobserver and interobserver reproducibility were good 0.88 (95%CI 0.66-0.95) and 0.86 (95%CI 0.65-0.94). Conclusion RV-EAT thickness,easily and rapidly assessed by standard low dose non-contrast chest CT was associated with higher incidence of ICU admission, need for mechanical ventilation and vasopressor support in hospitalized COVID-19 patients.Although no independent association between RV-EAT and in-hospital mortality was found, RV-EAT thickness may serve as surrogate marker of severity, before the rise of inflammatory biomarkers and may reflect inflammation changes within the myocardium in COVID-19. Abstract Figure 1. Examples of right ventricle ep Abstract Figure 2. Predictors of invasive mechani