(626) Lower Respiratory Tract Microbiome Changes in Intubated Infants With RSV and REV Bronchiolitis

Introduction: The leading cause of infant hospitalizations in the United States is severe viral bronchiolitis (VB) most often caused by respiratory syncytial virus (RSV) or rhinovirus/enterovirus (REV). Bacterial lower respiratory tract infection (LRTI) is a major contributor to mortality and morbidity in VB. In this study, we endeavored to determine the impact antibiotics have on the microbiome of infants with VB and the correlation with clinical outcomes.

Methods: Tracheal samples were collected daily from recruited intubated patients in the Helen DeVos Pediatric ICU from February 1, 2020 to January 31, 2024 (SH IRB# 2017-198). DNA was stored at -80C, extracted, and 16S rRNA gene amplicons were sequenced. Raw sequences were processed through Qiita using QIIME2-based algorithms to identity the assigned sequence variants and taxonomy was through the Silva database. Data was analyzed using phyloseq and ggplot in RStudio. We used linear regressions (Spearman’s rho) of the pathogen:commensal log-ratio with each of the continuous variables measured above to determine if there is a positive correlation between this ratio and outcomes. Categorical variables were compared with the ratio using Wilcoxon-rank sum tests.

Results: A total of 26 patients were recruited (RSV=18; REV=6, Controls=2) with 146 samples. Differences in alpha diversity between REV and RSV (p= 0.0036; Shannon; p= 0.00034; Chao1) and at different sampling time points were found; none in beta diversity. RSV alpha diversity (Shannon score) was different between intubation and final day (p= 0.015), as well as the first and second time points (p= 0.018) and first and last (p= 0.034) (Chao1 score), which is the diversity within an area or ecosystem. All but 2 patients were exposed early to antibiotics.

Conclusions: Initial samples showed different bacterial predominance and diversity between RSV and REV. Notably, pathogenic organisms such as Hemophilus and Moraxella predominated in specimens the first 24hrs, especially in RSV, and by the second sample, this pathogenic predominance (as evidenced by a rising Shannon index) changed with early antibiotic exposure. Further study may optimize the length of antibiotic treatment in VB with LRTI, as well as the relationship of the microbiome to clinical outcomes.