Investigating evolution of the paediatric cystic fibrosis lung microbiota using induced sputum sampling and culture-independent techniques

Children with cystic fibrosis are frequently non-productive of sputum even during exacerbation. Current routine sampling methods are either invasive (Bronchoalveolar lavage; BAL) or insensitive (cough swab) which makes pathogen surveillance challenging. We investigated induced sputum (IS) as a promising complementary sampling technique, looking at both its comparison to BAL and the evolution of the lung microbiota across children aged 0.5–18 years.

BAL and IS samples were collected as part of the CF-SpIT study (UKCRN14615; ISRCTNR12473810). DNA was extracted from samples and Illumina NextSeq sequencing of the 16S rRNA gene V4 region was performed. Bioinformatics data processing was carried out using Mothur. Microsoft Excel and R statistical software were used for downstream analyses.

Comparison of the microbiota of 30 BAL-IS matched patient samples indicated that at a presence/absence level, IS captured >80 % of the pathogens observed in BAL samples. These findings validated the expansion of the study to 136 un-matched IS sputum samples, and we found that as patient age increased, bacterial diversity decreased, and changes in the abundance of key genera occurred over time. Both Neisseria and Haemophilus decreased with age, whilst Pseudomonas and a sub-group of Prevotella increased.

Here we uniquely demonstrate using culture-independent techniques that IS captures the majority of the bacterial diversity observed in BAL samples. The age associated decline in lung microbiota diversity, previously documented using BAL samples, can also be captured using IS samples. This suggests that surveillance of microbiota evolution may be possible using this method.