The Overlap Between ARDS and Sepsis – or Not?

Sepsis and ARDS are complex syndromes where point-of-care testing and biomarker-guided, personalised treatments offer growing promise. With high mortality, prolonged recovery, and emerging therapies like anti-fibrotics, even small gains may be cost-effective.

Sepsis and Acute Respiratory Distress Syndrome (ARDS) are both clinical syndromes with a high mortality and morbidity burden on the intensive care unit (ICU) and beyond; they may occur together or apart, and other than high quality supportive care, seem to have few magic bullets over the years. This is likely to reflect the heterogeneity of the hosts and the underlying cause, where in the time-course of disease process we intercede, and in the case of ARDS, how host lung architecture heals after injury.

Biomarkers

The traditional biomarkers to avail oneself of for hyperinflammatory conditions are IL-6, TNFa and IL-1. The actions of these can be significantly modified by presence of anti-inflammatory cytokines such as IL-10. These are well established phenotypes in both sepsis and ARDS (Wilson and Calfee 2022; Sinha et al. 2023). It is probable that such baseline characteristics may influence steroid responsiveness – secondary analyses of VANISH suggest so (Antcliffe et al. 2019). Inflammatory phenotype also had a significant treatment interaction with activated protein C in PROWESS-SHOCK, p = 0.0043 – it is therefore essential we start identifying phenotypes as early as possible, both within trials and the real world (Sinha et al. 2023).

It is also important to look at endotheliopathy in both sepsis and ARDS – early markers of endothelial damage reflect disease severity – and predict which patients with sepsis are at risk of developing ARDS (Hendrickson and Matthay 2018). An excellent review describes the predictive effect of endothelial markers such as soluble tyrosine receptor kinase, for mortality, being equal to that of lactate (0.74) or SOFA score (0.73); unclear if combined models improve prognostication further. In general biomarkers alone do not outperform multivariate regression models in predicting mortality.

However there is an increasing interest in sub-group analysis, where statistically robust, which hints at more point-of-care role for biomarkers e.g. in sub analyses of FACTT and ProCESS, there is a signal that fluid liberal treatment strategies are superior only in hypoinflammatory patients – that is, those with low IL-6/TNF. This may well reflect less vulnerability to accumulating lung water, a pathology important to ARDS (Wiedemann et al. 2006).

Moreover it is becoming increasingly clear that metabolomics has a role to play with any inflammatory condition, and reflects the impact of obesity, metabolic syndrome and diabetes on outcomes in infection and inflammation (Alanazi et al. 2024). The impact of diabetes is widespread; it adversely affects chemotaxic, phagocytic and functional performance of neutrophils (Alba-Loureiro et al. 2007), primes a more reactive NLRP3 inflammasome (intracellular sensor of pathogens) (Meier et al. 2025), and creates a prolonged inflammatory trajectory, with a TNF profile that lingers past the point of health controls (Naguib et al. 2004). Of note, newer diabetic drugs may target some of these pathways in their non-diabetic survival enhancing effect (Kim et al. 2020).

Similarly, although not elucidated, there are strong signals for the role of lipid signalling in the sepsis-multi organ failure pathway (Guirgis et al. 2021).

Genetics

With respect to sepsis associated ARDS, genes pertinent to the cholesterol metabolism pathway have been recently demonstrated (Guillen-Guio et al. 2025), and one putative mechanism is the way in which lipids interact with the endothelium to recruit or ameliorate inflammatory cells and ligands/complement.

In the meantime, despite extensive forays into genome wide association studies, there is little overt impact of any single gene candidates in sepsis – for a variety of reasons. One is that any gene with enough penetrance to cause disease would quickly be selected against. Moreover in adult ICU sepsis is almost invariably a disease of the frail/elderly with the exception of particular organism virulence/niche host vulnerabilities.

Mortality

Both conditions have similar devastating mortalities. However it is important to note that death from multi organ failure is common to sepsis, and hyperinflammatory ARDS – hypoinflammatory ARDS is more likely to progress to death via respiratory failure (Evrard et al. 2024). Interestingly 12 month mortality for ARDS is high – approaching 60% - with both hypo/hyper phenotypes averaging out by 12 months, the burden of mortality from hyperinflammatory illness being in the first 90 days (Hashem et al. 2022).

Meanwhile, 12 month mortality for sepsis is lower but still significant – 36% in a large German cohort (Oliveros et al. 2024), with 45% within 3 years, and new functional dependence in ¼ of survivors (Fleischmann-Struzek et al. 2024).

Treatment

The most effective strategies for both appear to be good quality intensive care, minimising excess fluid, treating infection, avoiding ventilator induced lung injury, avoiding hyperoxia and second hits.

Steroids seem to have a tentatively widening evidence base, more so in ARDS than sepsis and this may reflect contribution of a pro inflammatory phenotype, the likely higher oxygen exposure (direct oxidative lung insult, and modification of the fibrotic stage of tissue injury.

Therapies to be trialled include lipid-modifying drugs, various mesenchymal stem cell treatments (Wang et al. 2023) and IL-6 antagonists for sepsis identified as hyperinflammatory (Lee et al. 2025) – traditionally IL-6 antagonists have actually increased mortality in sepsis, though they have proved efficacious in COVID-ARDS, and there is almost certainly a requirement to select patients for treatment with a biomarker approach. This feels like an exciting step forwards for personalised medicine already available in other fields such as oncology and the results of the PHIND study will be available to us this year.

One of the innovations in ARDS is targeted consideration of anti-fibrotic agents – we know this late healing stage can contribute to the longer term morbidity in ICU survivors and can be the final determinant of weaning. There are some hypothesis generating trials in COVID-19 ARDS that anti fibrinolytics reduce mortality (Umemura et al. 2021). Naturally this is a discrete subset of ARDS with less variability in patient presentation. However we also have cautious signals for monoclonal antibody therapies for LPS induced ARDS in early studied, with respect to fibrosis (Umemura et al. 2021).

Conclusion

Sepsis and ARDS are clinical syndromes with immunopathology, the details of which will require point of care testing to enable individualised treatment. Patients with pro-inflammatory baselines are at increased risk of both. Trials will also require biomarker based enrichment as subphenotypes are becoming clear, and like other fields, personalised medicine can make a difference. With long lengths of stay, mortality and ongoing low quality survivorship, many tailored treatments will become cost effective, even with small immediate survival advantages. Lingering morbidity from ARDS may be solved with judicious use of novel therapies such as anti-fibrotics. It is likely to be a decade of innovation and advantage for both syndromes; an exciting time to be an intensivist.

Conflict of Interest

None.