Background: Increasing evidence supports microRNAs (miRNAs) as potential biomarkers for the management of viral respiratory infections and critical patients.

Aim: First, to validate candidate miRNAs as biomarkers for clinical decision-making in critical COVID-19. Second, to construct a miRNA panel for early prediction of fatal outcomes in the intensive care unit (ICU).

Methods: Multicenter cohort study including 503 critically ill COVID-19 patients from 19 Spanish ICUs (CIBERESUCICOVID). A 16-miRNA qPCR panel, based on previously published data from our group, was analyzed in plasma samples collected within 48 hours after ICU admission.

Results: Nine candidate miRNAs were validated as biomarkers of all-cause ICU mortality (FC<0.83, FDR<0.05). Cox regression analysis revealed that low expression levels of eight candidates were associated with a higher risk of death. LASSO regression method constructed a miRNA panel (miR-16-5p, miR-192-5p, miR-323a-3p and miR-451a) which predicts the risk of in-ICU mortality (HR 2.53). This finding was supported by the Kaplan?Meier analysis. The miRNA panel improved the prognostic capacity for ICU mortality of APACHE-II (C-index 0.71, DeLong test p-value 0.055) and SOFA (C-index 0.67, DeLong test p-value 0.001) scores, and a clinical model (C-index 0.74, DeLong test-p-value 0.035). This improvement was also observed for 28 and 90-day mortality. Gene set enrichment analysis identified mechanistic pathways involved in SARS-CoV infection and inflammation.

Conclusion: The miRNA panel provides additional value for early prediction of fatal outcomes in critically ill COVID-19 patients.

Funding: ISCIII(PI20/00577,COV20/00110,CP20/00041,FI21/00187)_UNESPA.