Prospective clinical validation of the use of protein biomarkers from newly drawn patient sera for diagnosis of Alzheimer's disease

The recent widespread use of over the counter and prescription medications introduces additional variables that can confuse the interpretation of profiles of protein biomarkers for molecular diagnostics. This increases risk when transitioning a diagnostic test from discovery to clinical validation. Prospectively collected newly drawn serum samples, 39 AD, 62 PD, and 70 age-matched normal controls from 2 clinical sites were analyzed by 2D gel electrophoresis, fluorescent staining, quantitative digital image analysis, and individual and multivariate biostatistics (validated %CV≤20%; LOD ≥0.5ng/spot, 300μg/ml of blood serum). The concentrations of fifty-seven protein biomarkers of neurodegenerative disease, discovered in banked samples, were monitored. The protein biomarkers included 19 proteins of cellular degeneration, 7 Haptoglobin isoforms, 13 auto-immune and innate inflammatory proteins, 11 transfer proteins, and 8 proteins of unknown function. The profiles of the patients with AD, PD and age-matched normal controls were significantly different. Step wise multivariate discriminant analysis shows a combination of 33 of the proteins as optimal to provide for discrimination between AD and age-matched normal controls(100% sensitivity for patients mild and moderate AD symptoms; and 90.3% specificity). A different combination of 27 of the proteins, were optimal for discrimination between patients with Alzheimer's Disease and Parkinson's Disease (100% sensitivity and 100% specificity). Yet a different combination of 21of the proteins were optimal to provide discrimination between patients with Parkinson's Disease and age-match controls (92.9% sensitivity and 93.3% specificity). When all 57 of the proteins were employed, a sensitivity of 92.3% for detection of Alzheimer's disease and a specificity of 90.2% for detection of not Alzheimer's disease (Not AD = PD + Ctrl) were obtained using a combination of discriminant functions. Concentration differences of the 57 serum protein biomarkers can provide useful tools for diagnosis of Alzheimer's disease using newly drawn serum from patients who are not drug naïve.