A-249 Theranostic Algorithm: sensitivity and Specificity of a Composite Index Test for Guiding Treatment in Severe COVID-19

Abstract

Abstract Background Since the outbreak of the pandemic, severely affected COVID-19 patients receive standard of care at intensive care unit (ICU), but clinical trials are still ongoing for new expensive treatments. There is a need for guiding therapeutic decisions to restrain the public healthcare costs and to target the most efficient therapy. Based on the pathophysiology of SARS-CoV-2 on its ACE2-receptor, the angiotensin pathway is activated and circulating biomarkers of interest can be measured. This diagnostic test accuracy study describes how to classify COVID-19 patients in severe vs mild cases with the intention of guiding treatment decisions, to prevent long term sequels like pulmonary fibrosis. A theranostic approach is proposed, which combines in vitro diagnostics and personalized therapeutics. Methods N = 17 severely affected COVID-19 with acute respiratory distress syndrome (ARDS) at ICU were included and compared to N = 7 mild or asymptomatic COVID-19 cases and to N = 10 healthy controls. Follow-up samples have been collected during pulmonology consultation at three time points during one year post-ICU (N = 39). Non-parametrical statistics were used; data are expressed as [median (IQR)] and P < 0.05 is considered to be statistically significant. The biological markers angiotensin-(1-7) (Ang-(1-7)), transforming growth factor beta (TGF-β), ferritin, C-reactive protein (CRP from Roche) and Krebs von den Lungen 6 (KL-6 from Fujirebio) were measured in plasma samples. Those biomarkers represent a disequilibrium in lung protection, fibrogenesis, inflammation and hyperplasia of alveolar epithelial cells type II. Cut-off for a composite index was obtained through receiver operating curve (ROC). Results COVID-19 patients show higher Ang-(1-7) concentrations than healthy controls (P < 0.05). Critical patients at ICU show Ang-(1-7) deficit [124 pg/mL (81–154)], compared to asymptomatic COVID-19 cases [147 pg/mL (138–201)] (P < 0.05). COVID-19 patients with insufficient Ang-(1-7) are eligible for substitution therapy and novel expensive therapies beside standard of care, while the subpopulation of high Ang-(1-7) is analyzed with a composite index test including TGF-β, ferritin and CRP. A composite index of <34 differentiates asymptomatic COVID-19 from severe COVID-19 at ICU having acute respiratory distress syndrome (sensitivity = 100%, specificity = 80%, AUC = 0.933, LR = 5). The patients without risk for pulmonary sequels are not eligible for expensive treatment.After 1 year, KL-6 (<275 U/mL) and TGF- β (<34 968 pg/mL) normalise in only 25%, respectively 33% of severe COVID-19 patients. Within the first year of FU post-ICU, the mean KL-6 value in the patient population does not significantly change at different measuring moments (P = 0.63), while TGF-β decreases gradually at 3, 6 and 12 months post-ICU (P < 0.0001). Conclusion This theranostic algorithm predicts whether a patient infected by SARS-CoV-2 is at risk of developing clinical complications and supports clinical decision making concerning therapy. A supported therapeutic choice, based on objective biomarker measurements, could diminish long-term sequelae like pulmonary fibrosis. Clinical trials using companion drugs (e.g. Ang-(1-7) substitution therapy, MAS-receptor agonist), anti-fibrotics, interferon therapy, or monoclonal antibodies, could also rely on those biomarkers. The measurement of Ang-(1-7) for example, could categorize patients in the targeted subpopulation before administrating Ang-(1-7) substitution therapy, aiming at obtaining better treatment efficacy.