Comparative analysis of feature annotation methods for SESI-HRMS in exhaled breath analysis

Secondary electrospray ionization coupled to high-resolution mass spectrometry (SESI-HRMS) is a powerful method for the analysis of exhaled breath in real time. However, feature annotation is challenging due to the flow-injection nature of the technique. To evaluate alternative methods for enhancing feature annotation, a study was conducted where the exhaled breath of sixteen subjects was condensed and analyzed using dynamic headspace vacuum in-trap extraction gas chromatography-mass spectrometry (DHS-V-ITEX-GC-MS) and liquid chromatography coupled to mass spectrometry (LC-MS) using polar and reverse-phase conditions along with a data-independent MS²-acquisition method based on multiple injections. The annotation results obtained from these methods were compared to those from SESI-HRMS. The use of these techniques on breath condensate is unprecedented. The GC-MS method primarily detected compounds of exogenous origin, particularly additives in oral hygiene products like menthol. On the other hand, LC-MS detected a vast number of features, especially with the utilized data-independent acquisition method. Chemical classes to these features were assigned in-silico. In positive ion mode, mostly amino acids and amines were detected, while the largest group in negative ion mode consisted of carboxylic acids. Approximately 25% and 5% of SESI features had a corresponding match with LC-MS and GC-MS. While both GC-MS and LC-MS methods partially overlapped with the SESI features, there was limited overlap of both in the mass-to-charge range from 150 to 200. In conclusion, both GC-MS and LC-MS analysis of breath condensate can serve as supplementary tools for annotating features obtained from SESI-MS. However, to increase confidence in the annotation results, combining these methods with additional on-line fragmentation techniques is recommended.