Back

Pulmonary inflammation in severe pneumonia is characterised by compartmentalised and mechanistically distinct sub-phenotypes.

Jeffrey, M.; Bartholdson Scott, J.; White, R. J.; Higginson, E.; Maes, M.; Forrest, S.; Pereira Dias, J.; Parmar, S.; Heasman-Hunt, E.; Curran, M. D.; Polgarova, P.; Herre, J.; Davenport, E. E.; Baker, S.; Dougan, G.; Navapurkar, V.; Conway Morris, A.

2024-09-03 intensive care and critical care medicine
10.1101/2024.09.02.24312971 medRxiv
Show abstract

Pneumonia is the leading infectious disease killer worldwide and commonly requires admission to critical care. Despite its prevalence, the underpinning biology of severe pneumonia remains incompletely understood. We performed multifaceted assessments of bronchoalveolar transcriptome, cytokines, microbiology, and clinical features to biologically dissect a cohort of patients with suspected severe pneumonia. Our data revealed three lung-restricted transcriptionally defined severe pneumonia endotypes (termed Pneumotypes (Pn)). All three Pneumotypes had comparable clinical presentations and severity of respiratory failure but critically had divergent outcomes. Pn1, the most common, was characterised by low alveolar cytokines, expanded tolerogenic macrophages and epithelial damage. Pn3 was characterised by neutrophil-monocyte infiltration, IL-6-STAT3 activation and longer duration of mechanical ventilation. Pn2 displayed the fastest resolution, exhibiting a balanced immune response and epithelial-endothelial repair signatures. Our work has identified mechanistically distinct phenotypes in the lungs of patients with suspected pneumonia and acute lung injury, providing new targets for personalised therapy.

Published in Nature Communications (predicted rank #3) · training set

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.