BACKGROUND
SeptiCyte® RAPID is a host response-based assay to assess for the probability of sepsis in patients with systemic inflammation. Severe COVID-19 can be considered an example of viral sepsis. The purpose of this study was to evaluate whether SeptiCyte® RAPID could be used as a triage tool for COVID-19 patients at a hospital.
METHOD
COVID-19 positive (RT-qPCR) patients (n=59) were recruited at Hospital Foch, France. Blood samples were collected in EDTA vacutainers after patient admission to Emergency Department or Intensive Care Units (ICU). Samples were processed on SeptiCyte® RAPID cartridges, on a Biocartis Idylla™ instrument, per the SeptiCyte® RAPID CE-IVD package insert. COVID-19 disease severity was assessed by chest CT scan as percentage of lung damage, by these criteria: (1) peripheral, bilateral ground glass opacity (GGO) or (2) multifocal GGO of rounded shape with or without consolidation or visible intra-lobular lines (crazy paving). Patients were then clinically classified as “Mild”, “Moderate”, “Extensive”, “Severe”, or “Critical” based on <10%, 10-25%, 25-50%, 50-75%, >75% lung damage, respectively. CT scans were done within 24 hours of SeptiCyte® RAPID testing.
RESULTS
SeptiCyte® RAPID can distinguish COVID-19 patients showing higher lung damage as defined by lung CT scan (Critical and Severe) from those with a milder presentation (Mild and Moderate) with an AUC of 0.85 (Figure 1). The SeptiCyte® RAPID score was significantly elevated for patients in ICU vs. those not in ICU, distinguished with an AUC of 0.82 (Figure 2A). The SeptiScore® of Critical and Severe patients in the ICU were higher relative to the Moderate and Mild cases not in ICU (AUC = 0.87, Figure 2B). To evaluate the use of SeptiCyte® RAPID with other clinical parameters to triage Critical and Severe from Moderate and Mild cases, logistic regression models were generated in combination with other clinical parameters (Figure 3). Any logistic regression model containing SeptiScore® with clinical parameters (Left Panel) performs better than the singleton biomarker models or models that do not contain SeptiScore® (Right Panel).
CONCLUSION
SeptiCyte® RAPID can be processed using EDTA blood samples. The assay can distinguish COVID-19 cases with Severe or Critical presentation on CT from those with Mild or Moderate presentation.
FIGURES
Figure 1
SeptiCyte® RAPID scores stratified by radiologic severity. Panel A shows the score ranges in the various CT based COVID severity categories. Panel B shows the score ranges when the Critical and Severe groups are combined and compared relative to the Moderate and Mild cases.
Figure 2
SeptiCyte® RAPID scores stratified by patient severity and hospital location. Panel A. SeptiCyte® RAPID scores stratified by the need for ICU admission. Two patients’ samples were excluded from this analysis because their blood was drawn >1 day after ICU admission. Panel B. SeptiCyte® RAPID scores stratified by radiologic severity and hospital location. Three patients each with moderate (in ICU, black) or severe (not in ICU, black) lung damage are plotted separately, because their oxygen requirements differed significantly from expectation based on lung damage assessment.
Figure 3
Combining SeptiCyte® RAPID with other clinical variables: CRP, D-dimer, Lactate, Monocytes, CD16 Monocytes for triaging COVID-19 patients. Samples included in analysis were from the 1st time point only out of temporal samples collected for a few patients comparing Critical and Severe COVID-19 patients in ICU, (n=13) to Moderate and Mild cases, not in ICU, (n=24). Missing values were imputed using a multivariable imputation algorithm Amelia [1].
AUTHORS
Victor Gravrand1, François Mellot;2, Felix Ackerman3, Maric Christine Ballester4, Benjamin Zuber5, James Kirk6, Thomas Yager6, Krupa Navalkar6, Tiffany Pascreau6, Eric Farfour1, Marc Vasse7
1 Biology Department – Suresnes (France) / 2 Radiology Department – Suresnes (France) / 3 Internal Department – Suresnes (France) / 4 Emergency Department – Suresnes (France) / 5 Intensive Care Unit – Suresnes (France) / 6 Immunexpress – Seattle, WA (United States) / 7 Biology Department & UMRS 1176 – Suresnes (France)
[Abstract # 04729 Session: 12c]
REFERENCES
1. Honaker, J., King, G., & Blackwell, M. (2011). Amelia II: A Program for Missing Data. Journal of Statistical Software, 45(7), 1–47.
ACKNOWLEDGMENTS
This project has been supported in part by the Biomedical Advanced Research and Development Authority (BARDA), part of the HHS Office of the Assistant Secretary for Preparedness and Response under Contract No. 75A50120C00125.