Back

Sub-phenotypes of pneumonia defined by pulmonary histopathological features

Shastry, A.; Hiller, B. E.; Sanders, N. L.; Tseng, A. E.; Kothari, J.; O'Connell, A. K.; Gertje, H. P.; Ha, C. T.; Murzin, E.; Traber, K. E.; Lederer, J. A.; Jones, M. R.; Douam, F.; Beach, T. G.; Remick, D. G.; Crossland, N. A.; Monti, S.; Campbell, J. D.; Mizgerd, J. P.

2025-05-20 immunology
10.1101/2025.05.18.654716 bioRxiv
Show abstract

Establishing sub-phenotypes of pneumonia based on distinct host processes will be a step towards using host-directed therapies (to complement microbe-directed therapies) more rationally and precisely. Although pneumonia is a pulmonary pathophysiology, histological changes within the lungs have not been leveraged for sub-phenotyping. We addressed this by scoring 18 histopathology features (e.g., type 2 cell hyperplasia or necrosis) across rapid autopsy lung samples from 276 elderly subjects with pneumonia. Machine learning algorithms segregated subjects into seven different sub-phenotypes of pneumonia with distinct histopathology signatures. Quantitative immunofluorescence demonstrated associations of macrophages, neutrophils, T cells, and B cells with select histology features and pulmonary pathology sub-phenotypes. Mouse models revealed corollary sub-phenotypes, although some histology features observed in human lungs were never observed in mice. By illuminating this spectrum of histopathologies and discriminating discrete sub-phenotypes of pneumonia, a foundational framework emerges for developing and using host-directed therapies for subsets of pneumonia patients.

Published in American Journal of Respiratory and Critical Care Medicine (predicted rank #1) · training set

Matching journals

The top 8 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.