Application of volatile organic compound analysis to differentiate between bacterial and viral respiratory infections

Abstract

Respiratory infections are a leading cause of morbidity and mortality. Bacteria and viruses precipitate the majority of respiratory tract infections and also induce exacerbations in patients. Infection aetiology is particularly difficult to distinguish clinically and this represents a major unmet clinical need. Accurate identification of infection aetiology of respiratory infections would facilitate initiation of the most appropriate prompt management. There is evidence from in vitro and in vivo models that Volatile Organic Compounds (VOCs) can be measured through multiple analytical mass spectrometry platforms to differentiate between pathogens. The aim of this PhD is to investigate the utility of VOC measurements as a method of identification of infection aetiology and to determine whether such tests could be useful in differentiation of bacterial and viral exacerbations. Gas Chromatography- Mass Spectrometry (GC-MS) will be the primary method of analysis. Hypothesis Interaction between bacteria or viruses with their host results in different biochemical reactions. Metabolic degradation products including VOCs can be detected in exhaled breath and can be utilised to identify aetiology of exacerbations. Sub hypothesis • Pathogenic bacteria produce distinct VOCs that can be measured in the headspace following culture on agar medium. • Infection of airway epithelial cell cultures in vitro with pathogenic bacteria and respiratory viruses will lead to production of virus and bacteria specific VOCs that can be detected in the headspace. • Exhaled breath sampling following experimental infection with rhinovirus 16 in healthy subjects will identify VOCs that are upregulated following virus infection in vivo. • Targeted breath analysis in COPD patients during acute exacerbation will allow identification of pathogen-specific VOCs identified in sub-hypotheses 1, 2 and 3 Results In vitro agar plate growth of Streptococcus pneumoniae (SP) and Haemophilus influenzae (HI) showed different VOC profiles which allowed for differentiation of the two bacteria species. The higher the concentration of streptococcus plated, there was noted to be a dose dependent increase in VOCs detected. 23 and 13 VOCs respectively were produced by SP and HI culture on agar plates. 8/23 VOCs showed dose dependent increase with bacterial concentration plated. Similarly, VOCs specific to viral and bacterial infection were identified from bronchial epithelial cells stimulated with rhinovirus (RV) 16, respiratory syncytial virus (RSV), SP and HI. Bacterial markers in vitro were able to be separated with an area under curve (AUC) of 0.89. Long chain alkanes, such as decane were found in the exhaled breath of healthy subjects following RV infection. Findings from breath collection from COPD patients identified VOCs to differentiate between subjects who were having exacerbation compared to baseline state; subjects with either viral or bacterial positive sputum in exacerbations compared to those at baseline with negative microbiology could also be differentiated (later AUC 0.74). Conclusion Specific VOCs are associated with viral and bacterial infection. VOCs could provide an exciting tool to accurately determine aetiology, guiding judicious antimicrobial use.