Abstract Introduction ARDS is a heterogeneous syndrome defined by alveolar injury, and current approaches for distinguishing patients with maladaptive from reparative responses to injury are limited. This poses critical barriers to developing lung-targeted therapies and delivering them to patients likely to benefit. Here we aimed to 1) identify alveolar host-response biomarkers associated with ARDS diagnosis and severity; and 2) determine their relationship with mortality, accounting for the presence and severity of ARDS. Methods We analyzed a cohort of critically ill patients at the University of Washington (2014-2025) who i) received invasive mechanical ventilation; ii) had at ≥ 1 ARDS risk factor; and iii) underwent research or clinical bronchoscopy with BAL within 7 days of ICU admission. We measured 25 host-response biomarkers related to immune dysregulation and tissue injury-repair, then evaluated their 1) differences in concentration between patients with and without Berlin ARDS at BAL sampling; and 2) associations with hospital mortality, overall and stratified by ARDS, using interaction terms to test for effect modification. Results Among 306 adults (58% ARDS, 20% mortality), 22 of 25 BAL biomarkers were elevated in ARDS with Benjamini-Hochberg (BH)-adjusted P 0.05, adjusting for age, sex, infection, and bronchoscopy timing (Figure 1A-B). Amphiregulin, a growth factor involved in epithelial tissue repair; monocyte chemokines (CCL2, CCL13); and interleukin-6 (IL-6) were 2-3-fold higher in patients with ARDS and differed by key clinical severity measures (PaO2/FIO2, compliance) (Figure 1C). Many of these biomarkers were not associated with mortality in the overall cohort, and exhibited varying associations across patients with and without ARDS (Figure 1D). Notably, a doubling of CCL17, a T-cell chemokine, was associated with a nominally higher risk of death in non-ARDS, but a 19% lower risk of death in ARDS that was significant after BH-adjustment (BH-adjusted P for interaction = 0.03). In post-hoc analyses, we found patients with low (below median) BAL CCL17, compared to high BAL CCL17, exhibited features previously linked to T-cell exhaustion, including older age; male sex; and lower plasma IL-6 and IL-4 (which induce CCL17 expression). Conclusions ARDS diagnosis and severity was associated with elevations in numerous alveolar host-response biomarkers, including growth factors, pro-inflammatory mediators, and immunomodulatory proteins. A subset of these signals, notably elevated CCL17, were paradoxically associated with a lower risk of death in ARDS. This may reflect an adaptive, reparative response to lung injury, rather than injurious—highlighting the importance of distinguishing these processes in future work. This abstract is funded by: NHLBI
Sathe et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: