Airway microbiome dynamics and relationship to ventilator-associated infection in intubated pediatric patients

Background: Little is known about the airway microbiome in intubated mechanically ventilated children. We sought to characterize the airway microbiome longitudinally and in association with clinical variables and possible ventilator-associated infection (VAI). Methods: Serial tracheal aspirate samples were prospectively obtained from mechanically ventilated subjects under 3 years old from eight pediatric intensive care units in the United States from June 2017 to July 2018. Changes in the tracheal microbiome were analyzed by sequencing bacterial 16S ribosomal RNA gene relative to subject demographics, diagnoses, clinical parameters, outcomes, antibiotic treatment, and the Ventilator-Associated InfectioN (VAIN) score. Results: A total of 221 samples from 58 patients were processed and 197 samples met the >1000 reads criteria (89%), with an average of 43,000 reads per sample. The median number of samples per subject was 3 (interquartile range [IQR]: 2–5), with a median VAIN score of 2 (IQR: 1–3). Proteobacteria was the highest observed phyla throughout the intubation period, followed by Firmicutes and Actinobacteria. Alpha diversity was negatively associated with days of intubation (p =.032) and VAIN score (p =.016). High VAIN scores were associated with a decrease of Mycobacterium obuense, and an increase of Streptococcus peroris, Porphyromonadaceae family (unclassified species), Veillonella atypica, and several other taxa. No specific pattern of microbiome composition related to clinically diagnosed VAIs was observed. Conclusions: Our data demonstrate decreasing alpha diversity with increasing VAIN score and days of intubation. No specific microbiome pattern was associated with clinically diagnosed VAI.