Introduction: COVID-19 ARDS has raised significant concerns as the underlying pathophysiology of the disease remains unclear. To better understand this, we conduct serum proteomics, transcriptomics, and metabolomics analysis on serum samples of COVID-19 ARDS patients. We hypothesize that the multi-omics profiles shift through the progression of the disease and differ based on patient oxygenation response and dependence on mechanical ventilation. Methods: This is a prospective cohort study on COVID-19 ARDS patients. After diagnosis, serum was collected at baseline (day 0), day 3, and day 6. Protein and metabolite levels were quantified using LC-MS, and RNA-seq was used in transcriptomic analysis. Then, differential analysis was conducted between time points and subgroups, categorized by ventilator-free-days (VFD > 7 vs. VFD ≤ 7) and oxygenation improvement (>20% rise in PaO2/FiO2 vs. ≤20% rise in PaO2/FiO2), using Welch’s t-test, adjusted with Benjamini-Hochberg FDR. Functional enrichment analysis was used to characterize the biological activities. Results: Serum samples from 17 patients were collected at baseline, day 3, and day 6. Longitudinally, 53 RNA transcripts, mostly associated with antiviral response, were downregulated, and 1 RNA transcript was upregulated. No significant temporal changes were observed in protein abundance or metabolite levels. By ventilation duration, the long-term ventilation group demonstrated a marked increase in metabolite levels. Finally, proteomics and transcriptomics differed most dramatically between the oxygenation improvement group and the non-oxygenation improvement group at baseline (1705 different RNA transcripts; 66 different proteins). Gene-ontology enrichment indicated relative upregulation of T cell-mediated immunity and ribosomal pathways in non-improvers, whereas improvers were enriched for granule-membrane components and metal-ion binding. Conclusions: Overall, a decrease in antiviral transcripts was observed throughout the progression of COVID-19 ARDS. In patients with prolonged ventilation, a continuous increase in metabolite levels was present. In patients with improved oxygenation, an early change in protein and transcript levels was observed. Here, the dramatic differences in multi-omic patterns might have the potential to identify high-risk patients.
Li et al. (Sun,) studied this question.