Going deeper, reaching broader: proteomic ARDS phenotyping in East Asian critical care

<title>Extract</title> The phenotyping of acute respiratory distress syndrome (ARDS) has reached a pivotal moment, transitioning from discovery and validation towards real-time clinical implementation. The seminal 2014 latent class analysis (LCA) defined two reproducible inflammatory phenotypes (hyperinflammatory and hypoinflammatory) that reframed ARDS as a syndrome comprising distinct biological endotypes – patterns identifiable across aetiologies that could be leveraged for predictive enrichment in clinical trials [1]. What followed was an extraordinary arc: external validation across cohorts [2, 3], expansion of ARDS phenotyping in acute respiratory failure beyond ARDS [4, 5], predictive modelling with parsimonious models for bedside use [6], real-time testing of technical feasibility for biomarker-based phenotyping with the PHIND study (Clinical Evaluation of a Point of Care assay to identify PHenotypes IN acute respiratory Distress syndrome), towards large scale clinical trials for prospective testing of clinical utility of subphenotyping as with the PANTHER trial (Precision medicine Adaptive Network platform Trial in Hypoxemic acutE respiratory failuRe) [7] (figure 1). This evolution, from discovery to validation, model refinement, feasibility testing and early implementation, has brought precision critical care within reach.