Sputum proteomics uncovers the heterogeneity of Pseudomonas aeruginosa infections in bronchiectasis

<italic>Pseudomonas aeruginosa</italic> (PA) infections in bronchiectasis (BE) are common but poorly characterised. We hypothesise inflammatory endotypes are present in these infections, which interact with the microbiota and influence patients outcomes. Sputum was collected from patients with chronic PA infection in the ORBIT trials of inhaled liposomal ciprofloxacin (ARD-3150 n=196, placebo n=85). Sputum proteomics was performed by using a label-free proteomics workflow (n=164). Olink target 48 (n=116) measured sputum cytokines. 16S rRNA sequencing (LoopSeq) measured microbial abundances (n=155). PA bacterial load was determined by culture (n=140). PCA analysis showed heterogenous inflammatory profiles; influential proteins included MPO, ELANE and CAT. Significant changes were observed in 81 proteins based on alpha diversity, 92 proteins by PA relative abundance, and 30 proteins by PA bacterial load. 18 proteins were consistently altered across all three analyses, dominated by a neutrophil signature (ELANE, MMP8, CAT). Low alpha diversity, but not changes in PA abundance, was associated with platelet (PFN1), neutrophil (PYGL) and eosinophil (EPX) degranulation. High alpha diversity increased patients time-to-first exacerbation (PE) (p=0.017, HR=0.82, 95%CI=0.70-0.96). LSP1 was significantly upregulated in patients >1 PE in 12-months (p<0.001, ≤1 PE=17.6±1.93, >1 PE=18.7±1.28 [mean±SD]) but did not predict time-to-first PE (p=0.304). IL7 was decreased in patients with >1 PE (p=0.002, ≤1 PE=0.32±0.91, >1 PE=0.076±0.13) and predicted time-to-first PE (p=0.023, HR=0.57, 95%CI=0.35-0.92). PA infections in BE have a heterogeneous inflammatory profile altered by PA abundance and microbiota diversity.