Circulating endothelial signatures correlate with worse outcomes in COVID-19, respiratory failure and ARDS

Ana C Costa Monteiro1, Harry Pickering2, Aartik Sarma3

  • 1University of California School of Medicine- Los Angeles, Los Angeles, CA, USA. Acostamonteiro@mednet.ucla.edu.

PubMed

Insights

Elevated endothelial cell signatures (ECS) in circulation correlate with worse outcomes in respiratory failure patients. This transcriptomic approach offers a non-invasive method to evaluate endothelial damage in ARDS and COVID-19.

Area of Science:

  • Pulmonary Medicine
  • Genomics
  • Critical Care Medicine

Background:

  • Elevated circulating endothelial cells (CECs) are linked to poor outcomes in ARDS and COVID-19.
  • A consensus proteomic phenotype for CECs is lacking.
  • This study explores a transcriptomic approach to identify endothelial cells in circulation.

Purpose of the Study:

  • To determine if elevated endothelial cell signatures (ECS) in circulation correlate with worse respiratory outcomes.
  • To validate a transcriptomic deconvolution method for quantifying ECS.
  • To assess the association between ECS and mortality/respiratory failure severity.

Main Methods:

  • Unsupervised bulk-transcriptome deconvolution was used to quantify ECS percentage.
  • Two cohorts were analyzed: pediatric invasive mechanical ventilation (CAF-PINT) and adult COVID-19 (IMPACC).
  • Primary outcome was 28-day mortality; secondary outcomes included respiratory trajectories.

Main Results:

  • Higher day 0 ECS% was observed in non-survivors versus survivors in both pediatric and adult COVID-19 cohorts.
  • Each 1% increase in baseline ECS% significantly associated with mortality (aOR 1.36).
  • Increased baseline ECS% correlated with worse respiratory trajectories, including fatal outcomes.

Conclusions:

  • Quantifying ECS via deconvolution supports a transcriptomic approach for non-invasive evaluation of endothelial damage.
  • This method aids in understanding the link between endothelial damage and ARDS.
  • Utilizes novel deconvolution of circulating transcriptomic data for mechanistic insights.
Abstract