Subphenotypes of pneumonia defined by pulmonary histopathological features

Establishing subphenotypes of pneumonia based on host biology will be a step toward using host-directed therapies (to complement microbe-directed therapies) more rationally and precisely. Pneumonia is a pulmonary pathophysiology, but histopathological changes within the lungs have not been leveraged for subphenotyping.

To determine whether pneumonia includes distinct subphenotypes based on pulmonary histopathology.

We scored 20 histopathology features (eg, type 2 cell hyperplasia or necrosis) across rapid autopsy lung samples from 276 elderly participants with pneumonia. Statistical analyses were used to define associations among histopathological features, and clustering segregated participants into groups of similar histopathologies. Lung leukocytes were quantified using multispectral immunofluorescence. Lung samples from mice with varied infections were similarly examined to assess generalizability of findings and utility of experimental models.

Participants with pneumonia clustered into 7 subphenotypes with distinct histopathology signatures. Some histopathological features tended to associate with others, and each of the leukocytes measured (including macrophages, neutrophils, T cells, and B cells) associated with select histology features and pulmonary pathology subphenotypes. Corollary subphenotypes were observed in mouse models, although some histology features observed in human lungs were never observed in mice.

By illuminating a spectrum of histopathologies and discriminating discrete subphenotypes of pneumonia, these studies provide a framework that may prove useful for developing and testing host-directed therapies for subsets of pneumonia patients.At a Glance CommentaryScientific Knowledge on the Subject: Distinguishing subphenotypes of critical illnesses including pneumonia, ARDS, and sepsis is being pursued as a promising step towards optimizing application of host-directed therapies. Pneumonia is a pulmonary pathophysiology, making lung-specific data especially meaningful for establishing subphenotypes. While pneumonia associates with multiple different pathologic features in the lung, whether pneumonia subphenotypes can be distinguished by segregating sets of pulmonary histopathology features has yet to be directly addressed.What This Study Adds to the Field: Using unbiased machine learning to interrogate pathological assessments of lung tissues from people who died with pneumonia, seven different patterns of pulmonary pathology are discriminated. These pathological subphenotypes involve different types of immune cells in the lungs, and they can be partially captured with mouse models of lower -respiratory infection. Recognition of distinct pulmonary pathology subphenotypes of pneumonia suggests new opportunities for associating discriminatory biomarkers, defining underlying mechanisms, and testing responses to therapies.