S50 Profiling the microbiological and proteomic heterogeneity in patients with bronchiectasis and chronic Pseudomonas aeruginosa infection

<h3>Introduction</h3> <i>Pseudomonas aeruginosa</i> (PA) infection is associated with worse clinical outcomes and more severe disease in bronchiectasis. Despite this, treatments such as inhaled antibiotics targeting <i>P. aeruginosa</i> have shown mixed results. Recent data suggests that even within apparently similar clinical cohorts, different inflammatory and microbial endotypes may be evident. We hypothesised that we would be able to identify endotypes within the PA infected bronchiectasis population. <h3>Methods</h3> Sputum samples from patients with bronchiectasis were obtained from two replicate international randomized studies. Samples from baseline (before intervention) were used for analysis to exclude any treatment effects. Sputum protein profiling was performed by LC/MS. 16s sequencing and targeted qPCR were used for microbiome analysis and bacterial load respectively. Targeted measurement of neutrophil elastase (NE) activity was performed. <h3>Results</h3> Proteomics identified a high level of heterogeneity within the PA infected population including two previously identified clusters, one with an intense neutrophil extracellular trap (NET) dominated response and another with higher levels of antiproteinases and epithelial proteins. Bacterial load varied from 0 to 9.34 log units/ml using qPCR and higher bacterial load was associated with a more neutrophilic proteome and a higher NE activity. Patients with a higher NET mediated response had worse symptoms (p=0.003) and FEV1 (p&lt;0.001). Major drivers of the more severe cluster included LCP1, Azurocidin-1, calprotectin, CAT and MMP9 while the second cluster was defined by higher levels of SLPI, CST4, PIP and MUC5B. Partial least squares discriminant analysis confirmed this proteome profile was associated with higher bacterial load. In the randomised ORBIT 4 trial higher bacterial load was associated with enhanced treatment response (prolonged time to first exacerbation (TTFE) (HR 0.34 [0.19–0.62], p&lt;0.001) and reduced frequency of exacerbations (RR 0.38 [0.21–0.70], p=0.002). Reduced neutrophilic inflammation, reflected by a reduction during the trial in NE activity, was also associated with a significant benefit in prolonging TTFE (HR 0.48 [0.24–0.96], p=0.037). <h3>Conclusion</h3> Patients with PA infection represent a heterogeneous group of patients with varied bacterial load and host inflammatory response. Our data suggest that biomarkers such as qPCR or host inflammatory biomarkers may be required to select for treatment responders in future RCTs.