Jayanth Kumar Narayana, Tavleen Kaur Jaggi, Katerina Dimakou, Fransiskus Xaverius Ivan, Micheál Mac Aogáin, Mau Ern Poh, Rebecca C Hull, Chandani Hennayake, Merete B Long, Emma D Johnson, Holly Lind, Pieter C Goeminne, Michal Shteinberg, Anthony De Soyza, Stefano Aliberti, Josje Altenburg, Charles S Haworth, Oriol Sibila, Eva Polverino, Michael R Loebinger, Amelia Shoemark, Felix C Ringshausen, Natalie Lorent, Francesco Blasi, James D Chalmers, Sanjay H Chotirmall
PA-abundotypes and interactotypes represent independent traits capturing different dimensions of PA ecology and potentially contributing to observed clinical and treatment heterogeneity in bronchiectasis.
INTRODUCTION: Pseudomonas aeruginosa (PA) associates with poor clinical outcomes, however, exhibits inter-individual clinical and treatment heterogeneity in bronchiectasis.
METHODS: Sputum metagenomes from n = 600 individuals with bronchiectasis from 10 countries were assessed to derive PA subgroups based on abundance (abundotypes) and/or microbial interactions (interactotypes). Longitudinal dynamics of these subgroups were assessed in two common clinical scenarios: acute exacerbation and PA eradication. Transition probabilities for each subgroup were estimated in these scenarios.
RESULTS: Four PA-abundotypes with comparable exacerbation risk were identified: PA-dominant (PD), Moderate PA-dominant (MPD), Other Microbe Dominant (OMD) and High Diversity (HD). Four separate PA-interactotypes were characterized: BC1 (PA-Neisseria interactions); BC2 (PA-S. aureus & E. coli interactions); BC3 (PA-Streptococcus interactions) and BC4 (mixed interactions). Core (conserved) interactions with commensal taxa were shared across all interactotypes while ancillary (variable) interactions remain interactotype-specific. Abundotypes correlate with bronchiectasis severity, symptoms and lung function while interactotypes stratify exacerbation risk with BC3 the lowest-risk group. BC1 and BC4 exhibit 1.45- and 1.44-higher fold rates (p = 0.037 and p = 0.048 respectively) for exacerbation while BC2 demonstrates a non-significant increase in exacerbation risk (1.32-fold; p = 0.18). In hypothesis-generating analyses using a small longitudinal cohort, PA-abundotypes remain stable through exacerbations (n = 6) while PA-interactotypes shift substantially suggesting dynamic responsiveness to acute clinical states. During PA eradication (n = 11), abundotypes exhibit structured transition with standard therapy while interactotypes demonstrate no consistent patterns.
CONCLUSION: PA-abundotypes and interactotypes represent independent traits capturing different dimensions of PA ecology and potentially contributing to observed clinical and treatment heterogeneity in bronchiectasis.