Concordance between upper and lower airway microbiota in children with cystic fibrosis

<b>Introduction.</b> Sputum is the most used sample type to monitor the lower respiratory tract microbiota in cystic fibrosis (CF), but young patients often cannot expectorate.<b>Hypothesis.</b> We hypothesized that throat swabs could reflect lower airway colonization and assessed the concordance of bacterial community composition between paired sputum and throat swab samples from children with CF.<b>Aim.</b> We aimed to compare bacterial community diversity and composition between sputum and throat swabs in the full cohort and in patients with paired samples from the same visit.<b>Methodology.</b> The prospective longitudinal multicentre MUCOVIB cohort included 379 samples from 61 CF children. Using V3-V4 16S rRNA amplicon metagenomics, we compared bacterial community diversity and composition between sputum and throat swabs in the full cohort and in 11 patients with paired samples from the same visit.<b>Results.</b> Sputum and throat swabs exhibited similar bacterial diversity, regardless of the exacerbation status, and presented a substantial agreement for detecting pathogens (Cohen's kappa: 0.6). Differences in bacterial abundance were observed (<i>P</i>=0.001), but not presence/absence (<i>P</i>=0.098). Community typing revealed three distinct community types, with 86% of paired samples falling into the same cluster, highlighting the homogeneity between sputum and throat swab microbiota. Network analysis demonstrated slight, non-random similarities in microbial interactions between sample types (adjusted Rand index=0.08 and 0.10). The average beta-diversity distances between samples collected from the same visit were shorter (0.505±0.056 95% confidence interval), compared with sputum (0.695±0.017) or throat swab (0.704±0.045) from the same patient collected during different visits.<b>Conclusion.</b> Throat swabs can provide representative information on lower respiratory microbiota. Clinicians should collect throat swabs rather than relying on sputum samples from previous visits to guide antibiotic prescriptions in CF children unable to expectorate.