Detection of Tuberculosis by The Analysis of Exhaled Breath Particles with High-resolution Mass Spectrometry

Abstract

Tuberculosis remains a global health threat killing over 1 million people per year. Current sputum-based diagnostics are specific but lack sensitivity resulting in treatment of many sputum negative cases. In this proof-of-concept study, we used high-resolution mass spectrometry to identify specific lipids in peripheral lung fluid samples of TB patients and controls, captured using a novel non-invasive sampling system. Exhaled respiratory particles were collected in liquid and after concentration and lipid extraction directly infused into a high-resolution mass spectrometer. High-resolution mass spectrometric data collection was conducted in a dual ion mode and chemical compositions were constructed using accurate mass measurement. Over 400 features with high segregating capacity were extracted and optimized using feature selection algorithm and machine learning, from which the accuracy of detection of positive tuberculosis patients was estimated. This current strategy provides sensitivity offered by high-resolution mass spectrometry and can be readily susceptible for developing a novel clinical assay exploring peripheral lung fluid for the detection of active TB cases