Multiomics analysis across 12 studies identified hyperinflammatory and hypoinflammatory ARDS subphenotypes that exhibit significant differences in mortality and differential treatment responses.
Systematic Review
Do multiomics approaches identify distinct biological phenotypes that inform personalized therapy in adult ARDS patients?
Multiomics approaches can identify distinct biological subphenotypes of ARDS that have different prognostic trajectories and therapeutic responses, paving the way for precision medicine.
Abstract Introduction Acute Respiratory Distress Syndrome (ARDS) is a clinically heterogeneous syndrome with high mortality. The traditional single-definition approach overlooks critical underlying biological variation, limiting the efficacy of generalized therapeutic strategies. Multiomics—the integrated analysis of genomics, transcriptomics, proteomics, and metabolomics—offers a powerful lens to decode this complexity, potentially guiding personalized treatment. This systematic review synthesizes the current evidence on how multiomics approaches are defining novel biological ARDS phenotypes and informing therapy selection. Methods We systematically searched major databases (e.g., PubMed, Embase) for studies employing multiomics approaches (two or more omics platforms) to phenotype adult ARDS patients. Data extracted included the specific omics technologies used, the ARDS population studied, the key phenotypes and biomarkers identified, and the clinical implications, particularly regarding differential treatment response. Results Twelve key multiomics studies (2019-2025) and three comprehensive reviews were analyzed. The most consistently identified subphenotypes are the hyperinflammatory (or reactive) and hypoinflammatory (or uninflamed) groups, defined by distinct molecular signatures (e.g., high vs. low levels of IL-6, IL-8, TNFR-1, or specific gene/protein expression patterns). Multiomics successfully identified these groups across diverse etiologies, including COVID-19 and non-COVID ARDS, and various clinical trial cohorts (e.g., ROSE trial). Key molecular pathways implicated include innate immunity, sphingolipid signaling, interferon signaling, and mitochondrial dysfunction. Importantly, these omics-defined phenotypes exhibited significant differences in mortality and ventilator-free days.Furthermore, multiomics has demonstrated strong clinical relevance for guiding therapy. Patients with the hyperinflammatory phenotype showed improved outcomes with higher PEEP, glucocorticoids, and potentially simvastatin, but the same treatments were either neutral or potentially harmful in the hypoinflammatory group. Newer proteomic studies have also identified an “immune-suppressed/repair” phenotype that may be harmed by traditional immunosuppressive strategies. Multiomics data also enables the development of high-performing, protein- or metabolite-based predictive models that often outperform purely clinical models. Conclusion Multiomics is essential for dissecting the biological heterogeneity of ARDS, moving beyond clinical definitions to reveal distinct prognostic and therapeutic subphenotypes. The robust evidence supporting differential treatment responses (e.g., to PEEP and corticosteroids) based on molecular phenotype underscores the potential of multiomics to realize precision medicine in ARDS. Future research should focus on validating these omics-driven phenotypes in prospective clinical trials and translating complex molecular data into clinically actionable, point-of-care diagnostics. This abstract is funded by: None
Francis et al. (Fri,) conducted a systematic review in Acute Respiratory Distress Syndrome (ARDS). Multiomics approaches was evaluated on Biological ARDS phenotypes and differential treatment response. Multiomics analysis across 12 studies identified hyperinflammatory and hypoinflammatory ARDS subphenotypes that exhibit significant differences in mortality and differential treatment responses.