Differential airway resistome and its correlations with clinical characteristics in Haemophilus- or Pseudomonas-predominant microbial subtypes of bronchiectasis.

ABSTRACT

The prevalence and clinical correlates of antibiotic resistance genes (ARGs) in bronchiectasis are not entirely clear. We aimed to profile the ARGs in sputum from adults with bronchiectasis, and explore the association with airway microbiome and disease severity and subtypes. In this longitudinal study, we prospectively collected 118 sputum samples from stable and exacerbation visits of 82 bronchiectasis patients and 19 healthy subjects. We profiled ARGs with shotgun metagenomic sequencing, and linked these to sputum microbiome and clinical characteristics, followed by validation in an international cohort. We compared ARG profiles in bronchiectasis according to disease severity, blood and sputum inflammatory subtypes. Unsupervised clustering revealed a Pseudomonas predominant subgroup (n = 16), Haemophilus predominant subgroup (n = 48), and balanced microbiome subgroup (N = 54). ARGs of multi-drug resistance were over-dominant in the Pseudomonas-predominant subgroup, while ARGs of beta-lactam resistance were most abundant in the Haemophilus-predominant subgroup. Pseudomonas-predominant subgroup yielded the highest ARG diversity and total abundance, while Haemophilus-predominant subgroup and balanced microbiota subgroup were lowest in ARG diversity and total abundance. PBP-1A, ksgA and emrB (multidrug) were most significantly enriched in Haemophilus-predominant subtype. ARGs generally correlated positively with Bronchiectasis Severity Index, fluoroquinolone use, and modified Reiff score. 68.6% of the ARG-clinical correlations could be validated in an independent international cohort. In conclusion, ARGs are differentially associated with the dominant microbiome and clinical characteristics in bronchiectasis.