Strong Predictive Algorithm of Pathogenesis-Based Biomarkers Improves Parkinson’s Disease Diagnosis

Easily accessible and accurate biomarkers can aid Parkinson’s disease diagnosis. We investigated whether combining plasma levels of α-synuclein, anti-α-synuclein, and/or their ratios to amyloid beta-40 correlated with clinical diagnosis. The inclusion of amyloid beta-40 (Aβ40) is novel. Plasma levels of biomarkers were quantified with ELISA. Using receiver operating characteristic (ROC) curve analysis, levels of α-synuclein, anti-α-synuclein, and their ratios with Aβ40 were analyzed in an initial training set of cases and controls. Promising biomarkers were then used to build a diagnostic algorithm. Verification of the results of biomarkers and the algorithm was performed in an independent set. The training set consisted of 50 cases (age 65.2±9.3, range 44–83, female:male=21:29) with 50 age- and gender-matched controls (67.1±10.0, range 45–96 years; female:male=21:29). ROC curve analysis yielded the following area under the curve results: anti-α-synuclein=0.835, α-synuclein=0.738, anti-α-synuclein/Aβ40=0.737, and α-synuclein/Aβ40=0.663. A 2-step diagnostic algorithm was built: either α-synuclein or anti-α-synuclein was ≥2 times the means of controls (step-1), resulting in 74% sensitivity; and adding α-synuclein/Aβ40 or anti-α-synuclein/Aβ40 (step-2) yielded better sensitivity (82%) while using step-2 alone yielded good specificity in controls (98%). The results were verified in an independent sample of 46 cases and 126 controls, with sensitivity reaching 91.3% and specificity 90.5%. The algorithm was equally sensitive in Parkinson’s disease of ≤5-year duration with 92.6% correctly identified in the training set and 90% in the verification set. With two independent samples totaling 272 subjects, our study showed that combination of biomarkers of α-synuclein, anti-α-synuclein, and their ratios to Aβ40 showed promising sensitivity and specificity.