Self-sampling of throat swabs at home for respiratory tract microbiome analysis

Longitudinal studies of the respiratory tract microbiome (RTM) would benefit from repeated sampling, but it is unclear if patients can reliably self-collect throat swabs at home. We compared RTM composition in 76 stable COPD patients (median age: 63, FEV1%pred: 56, 61% male) using swabs collected by medical professionals (PRO), self-samples collected simultaneously (SELF0), and self-samples collected one week later (SELF1). Microbiomes were profiled via 16–23s rDNA analysis (Molecular Culture, Inbiome, Amsterdam). Different patient identity explained 74% of overall compositional variation of Bray-Curtis dissimilarities (p=0.001), whereas sampling method explained <1% (p=0.028) (PERMANOVA). Pairwise tests revealed no significant differences in beta-diversity between PRO vs. SELF0 (p=0.662) or PRO vs. SELF1 (p=0.216). Alpha diversity also didn't differ: mean richness was 23.6±8.1 (PRO) vs. 22.7±7.4 (SELF0, p >0.99) and 24.9±7.2 (SELF1, p=0.6); Shannon diversity 2.5±0.4 (PRO) vs. 2.5±0.4 (SELF0, p=0.808) and 2.6±0.3 (SELF1, p=0.646); Evenness 0.8±0.04 (PRO) vs. 0.8±0.04 (SELF0, p=0.135) and 0.8±0.04 (SELF1, p >0.99). At the phylum level, relative abundance of Firmicutes/Actinobacteria/Fusobacteria/Verrucobacteria (FAFV) was 0.39±0.18 (PRO) vs. 0.39±0.16 (SELF0, p >0.99) and 0.37±0.13 (SELF1, p=0.424). Bacteroidetes was 0.23±0.11 (PRO) vs. 0.23±0.11 (SELF0, p >0.99) and 0.25±0.12 (SELF1, p >0.99), and Proteobacteria 0.38±0.14 (PRO) vs. 0.38±0.13 (SELF0, p >0.99) and 0.38±0.10 (SELF1, p >0.99). Of 31 detected genera, only Granulicatella was significantly lower in SELF1 compared to PRO (p=0.041). We conclude that self-sampling of throat swabs at home by COPD patients is a valid approach for RTM analysis.