Large-scale plasma proteomics uncovers preclinical molecular signatures of Parkinson disease and overlap with other neurodegenerative disorders

BackgroundParkinsons disease (PD) remains incurable, with a long prediagnostic phase currently undetectable by existing methods. Identifying individuals at high risk of PD would enhance our understanding of the underlying pathophysiology and will open up new preventive or early therapeutic avenues. MethodsIn the largest proteomic study in neurodegenerative diseases to date, we analyzed blood samples from [~]74,000 individuals across discovery and validation cohorts. In the discovery phase, we leveraged data from the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort with up to 28 years of follow-up to identify incident PD cases across five European countries. Prediagnostic plasma samples from initially healthy participants underwent high-throughput proteomic profiling (7,285 protein markers) using the SomaScan platform. Cox proportional hazards models based on 4,538 participants (including 574 incident PD cases) were used to identify protein biomarkers associated with a future PD diagnosis. In the validation phase, we tested three cohorts with incident PD cases (AGES, ARIC, UK Biobank, (n=64,856; 1,034 incident cases), a case-control dataset with prevalent cases (GNPC, n=2,592), and the longitudinal Tracking PD cohort with data on PD progression (n=794). FindingsIn the EPIC4PD discovery case-cohort, 17 proteins that predict PD up to 28 years before diagnosis were detected (FDR=0.05). Additional proteins revealed sex-specific and time-varying effects, capturing disease progression before symptom onset. Replication in three prospective cohorts confirmed at least 12 key prediagnostic biomarkers with strong evidence, including TPPP2, HPGDS, ALPL, MFAP5, OGFR, ACAD8, TCL1A, GPC4, GSTA3, LCN2, KRAS, and GJA1. Furthermore, in the longitudinal Tracking PD cohort, HPGDS and MFAP5 also predicted cognitive decline. Notably, several of the identified PD biomarkers overlapped with those for incident Alzheimers disease and amyotrophic lateral sclerosis, indicating partially shared molecular signatures. A machine learning-derived PD protein risk score improved PD but not Alzheimers disease risk prediction in the independent AGES cohort. ConclusionOur extensive proteomics effort examining PD from the prediagnostic to the progression phase identified novel, actionable biomarkers opening new avenues for early PD risk stratification and precision medicine. FundingMichael J Fox Foundation, Cure Alzheimers Fund, Clinical Research in ALS and Related Disorders for Therapeutic Development (CreATe) Consortium, Michael J Fox Foundation, Interdisciplinary Centre for Clinical Research, University Munster