LAUSR

Background Reflux aspiration (RA) occurs when gastric fluid is introduced into the lower respiratory tract, and has the potential to injure the lung. However in a cohort of 41 adult CF patients, we were unable to demonstrate a relationship between oesophageal combined pH impedance (pH-MII) derived reflux measures and traditional measures of lung disease severity, including change in FEV1% predicted and pulmonary exacerbation frequency. To extend these studies, we present initial data from a sputum proteomics approach, the purpose of which is to identify differences in the sputum proteome between those with pH-MII derived reflux measures presumed ‘high-risk’ for RA (total, proximal and supine), and to identify RA protein biomarkers. Methods Sputum samples were collected from 20 CF patients and compared to 20 healthy volunteers (HV), who were asymptomatic for gastro-oesophageal reflux. CF patients were selected from our original cohort to provide a spread of pH-MII measures. The CF patients were divided into the highest (n=10) and lowest (n=10) for each pH-MII measure. To explore the proteomic profiles, and to uncover protein biomarkers for RA, samples were analysed by bottom-up shotgun proteomic analysis using the Thermo QExactive HF mass spectrometer. Results Shotgun proteomic analysis highlighted clear differences between CF and HV sputum proteome using ANOVA and principal component analysis, especially with large differences in neutrophil derived proteins in CF samples. Equivalent significant changes in specific proteins were not evident when repeated for CF subjects with high and low pH-MII reflux measures and no proteins were identified of gastric origin. Conclusion Our initial data suggest that the sputum proteome is not correlated with pH-MII measures denoting potential ‘high-risk for RA’. This accords with our previous findings, which did not expose any relationship to standard respiratory endpoints. No gastric proteins were detected in CF sputum using this approach. Such proteins may be below the limit of detection, and more advanced proteome approaches may be required to increase their detection and quantification. We intend to analyse the rest of the cohort (n=41) and use more advanced proteomic methods in the continued search for RA protein biomarkers.