What is the optimal sampling technique for assessing the microbiota with bronchoscopy?

Abstract

Introduction: We examined the degree of contamination of the airway microbiome with different bronchoscopic sampling techniques. Methods: We collected oral wash (OW), protected specimen brushes (PSB), protected broncho-alveolar lavage (PBAL), and small-volume lavage (SVL) using the suction canal on the bronchoscope, in 16 COPD patients and 8 controls. Together with negative control samples (NCS) we performed paired-end sequencing of the V3-V4 region of 16S rRNA on a Illumina MiSeq. We performed analyses of α- and β-diversity after removing all operational taxonomic units (OTUs) in the NCS, using QIIME. Results: NCS OTUs accounted for 80% of all sequences, and we retained 2.5 million sequences from 144 samples distributed in 803 OTUs for analyses. Bonferroni-corrected, Wilcoxon matched-pairs test indicated that OW samples were more α-diverse (phylogenetic distance) than the PBAL and PSB samples, but not compared to the SVL. Also the first fraction of the PBAL was more α-diverse than the second (p In principle coordinates analysis of β-diversity (UniFrac distance), all respiratory tract samples clustered differently from the OW samples, the overlap was larger for SVL and PBAL samples than for PSB (Figure 1). Thus, SVL samples were more similar to OW than the protected samples, indicating contamination. Conclusions: We need protected sampling techniques (PSB and PBAL) to avoid contaminated samples of the airway microbiome.