A metabolomic signature of treated and drug-naïve patients with Parkinson’s disease: a pilot study

• Published in: Metabolomics 2019-06, Vol.15 (6), p.90-11, Article 90
• Main Authors: Troisi, Jacopo, Landolfi, Annamaria, Vitale, Carmine, Longo, Katia, Cozzolino, Autilia, Squillante, Massimo, Savanelli, Maria Cristina, Barone, Paolo, Amboni, Marianna
• Format: Article
• Language: English
• Subjects: Aged; Aged, 80 and over; Basal ganglia; Biochemistry; Biomarkers - blood; Biomarkers - metabolism; Biomedical and Life Sciences; Biomedicine; Case-Control Studies; Cell Biology; Central nervous system diseases; Developmental Biology; Discriminant analysis; Disease; Drug therapy; Female; Gas chromatography; Gas Chromatography-Mass Spectrometry; Humans; Life Sciences; Male; Mass spectroscopy; Mathematical models; Metabolic pathways; Metabolites; Metabolome - drug effects; Metabolomics; Middle Aged; Molecular Medicine; Movement disorders; Neurodegenerative diseases; Original Article; Parkinson Disease - blood; Parkinson Disease - drug therapy; Parkinson Disease - metabolism; Parkinson Disease - pathology; Parkinson's disease; Pilot Projects; Statistical analysis
• ISSN: 1573-3882; 1573-3890
• DOI: 10.1007/s11306-019-1554-x

Introduction About 90% of cases of Parkinson’s disease (PD) are idiopathic and attempts to understand pathogenesis typically assume a multifactorial origin. Multifactorial diseases can be studied using metabolomics, since the cellular metabolome reflects the interplay between genes and environment. Objective The aim of our case–control study is to compare metabolomic profiles of whole blood obtained from treated PD patients, de-novo PD patients and controls, and to study the perturbations correlated with disease duration, disease stage and motor impairment. Methods We collected blood samples from 16 drug naïve parkinsonian patients, 84 treated parkinsonian patients, and 42 age matched healthy controls. Metabolomic profiles have been obtained using gas chromatography coupled to mass spectrometry. Multivariate statistical analysis has been performed using supervised models; partial least square discriminant analysis and partial least square regression. Results This approach allowed separation between discrete classes and stratification of treated patients according to continuous variables (disease duration, disease stage, motor score). Analysis of single metabolites and their related metabolic pathways revealed unexpected possible perturbations related to PD and underscored existing mechanisms that correlated with disease onset, stage, duration, motor score and pharmacological treatment. Conclusion Metabolomics can be useful in pathogenetic studies and biomarker discovery. The latter needs large-scale validation and comparison with other neurodegenerative conditions.