Metabolomic breath landscape analysis unravels lipid biomarker candidates in patients with monogenic and idiopathic Parkinson's disease

Parkinsons disease (PD) is the fastest growing neurodegenerative disorder. Lack of efficient early diagnostic tools highlights the critical necessity for novel approaches in biomarker discovery. We propose an untargeted metabolomics approach using non-invasive exhaled breath analysis. Breath samples, collected from 73 PD patients, encompassing both genetic (LRRK2: n=12, GBA1: n=35, PRKN: n=6) and idiopathic PD (n=20), 4 unaffected LRRK2 carriers and 90 controls underwent extreme-resolution FT-ICR-MS analysis. Findings were compared with metabolomics data from blood plasma. Biostatistical analyses identified discernible metabolic patterns in both biofluids, enabling differentiation of PD patients from healthy controls (OOB error < 1%). Metabolomic breath profiling of PD patients yielded 10 significant hits putatively identified as tetracosanoic acid, tricosanoic acid, HMVA, docosanamide, eicosanoic acid, nonadecanoic acid, homophytanic acid, nonadecyl-MG, stearic acid and palmitic acid in PD patients, irrespective of the genetic status. Most of the proposed structures are intermediates in fatty acid metabolism, introducing new candidate biomarkers for breath analysis in PD. Seven of these metabolites were also found in unaffected carriers of pathogenic variants in LRRK2 when compared to controls. Breath analysis effectively distinguishes between PD patients and healthy controls and nominates metabolites that could serve as noninvasive biomarkers for PD, potentially including its presymptomatic stage.