Classification of patients with COVID-19 by blood RNA endotype: A prospective cohort study (preprint)

ABSTRACT

Background: Although the development of vaccines has considerably reduced the severity of COVID-19, its incidence is still high. Hence, a targeted approach based on RNA endotypes of a population should be developed to help design biomarker-based therapies for COVID-19. Objectives: We evaluated the major RNAs transcribed in blood cells during COVID-19 using PCR to further elucidate its pathogenesis and determine predictive phenotypes in COVID-19 patients. Study design: In a discovery cohort of 40 patients with COVID-19, 26,354 RNAs were measured on day 1 and day 7. Five RNAs associated with disease severity and prognosis were derived. In a validation cohort of 153 patients with COVID-19 treated in the intensive care unit, we focused on prolactin (PRL), and toll-like receptor 3 (TLR3) among RNAs, which have a strong association with prognosis, and evaluated the accuracy for predicting survival of PRL-to-TL3 ratios (PRL/TLR3) with the areas under the ROC curves (AUC). The validation cohort was divided into two groups based on the cut-off value in the ROC curve with the maximum AUC. The two groups were defined by high PRL/TLR3 (n=47) and low PRL/TLR3 groups (n=106) and the clinical outcomes were compared. Results: In the validation cohort, the AUC for PRL/TLR3 was 0.79, showing superior prognostic ability compared to severity scores such as APACHE II and SOFA. The high PRL/TLR3 group had a significantly higher 28-day mortality than the low PRL/TLR3 group (17.0% vs 0.9%, P<0.01). Conclusions: A new RNA endotype classified using high PRL/TLR3 was associated with mortality in COVID-19 patients.