Background: Idiopathic pulmonary fibrosis (IPF) is a progressive fibrotic lung disease of unknown cause that leads to respiratory failure and death within five years of diagnosis. It is thought to arise in genetically susceptible individuals as a consequence of an aberrant wound-healing response following repetitive alveolar injury. The exact triggers, which initiate the fibrotic process, remain unknown. Overt respiratory infection and immunosuppression carry a high mortality, while polymorphisms in genes related to epithelial integrity and host defence predispose to IPF. There has however, been no systematic search for potential infective sources of alveolar injury in IPF.

Methods: Patients diagnosed with IPF were prospectively enrolled, together with matched controls, and followed up over a two-year period. Subjects underwent bronchoalveolar lavage (BAL) at recruitment and during any exacerbations allowing quantification of bacterial load and identification of communities by 16S rRNA qPCR and pyrosequencing (Roche 454). Longitudinal peripheral whole blood gene expression profiles were generated (Affymetrix 1.1ST Arrays) and MUC5B SNP genotyping performed.

Results: IPF patients have double the burden of bacteria in BAL compared to controls. Baseline bacterial burden predicted the rate of decline in lung volume, risk of death and associated independently with the rs35705950 polymorphism of the MUC5B gene. Haemophilus, Streptococcus, Neisseria and Veillonella spp. were more abundant in cases than controls. During an exacerbation of IPF there is an increase in BAL bacterial burden and changes in the microbiome from baseline. Peripheral whole blood gene expression profiling has given potential mechanistic insights in to the pathogenesis of IPF. A total of 1,358 transcript clusters, differentially expressed between cases and controls, were identified. Apoptosis, oxidative stress, angiogenesis, cellular inflammation and antimicrobial processes were all implicated. Novel potential interactions with well established putative fibrotic pathways were identified and for the first time longitudinal expression changes in IPF subjects were demonstrated.

Conclusions: IPF is characterized by an increased bacterial burden in BAL that predicts decline in lung function and death. Bacterial burden and the microbiota both change during an exacerbation, despite current guidelines dictating they are non-infective events. The association demonstrated between bacterial load and MUC5B genotype suggests there may be a direct relationship between host immunity and the respiratory microbiome. IPF patients have a characteristic peripheral blood transcriptome. Longitudinal expression and WCGNA analysis not only identified genes associated with IPF and survival but also confirmed previously suggested potential prognostic biomarkers including Lymphocyte-specific protein tyrosine kinase (LCK).