Automated Diagnostic Testing for Detection of Alzheimer’s Disease

Conger, Tori BS
Julovich, David BS
Hall, James PhD
O'Bryant, Sid PhD
Journal Title
Journal ISSN
Volume Title

Purpose: Alzheimer’s disease (AD) historically has been detected with brain imaging scans by magnetic resonance imaging (MRI) or positron emission tomography (PET) and by measurement of protein levels in cerebrospinal fluid (CSF). These tests pose limitations due to high cost and invasiveness of the procedures. Our automated blood-based biomarker screening utilizes ELISA based detection and provides a realistic, cost-effective approach for detection of AD and mild cognitive impairment (MCI). Methods: MesoScale Discovery (MSD) multiplex biomarker platform, integrated with a Hamilton StarPlus liquid handler and Quanterix single molecule array (Simoa) HD-1 fully automated analyzer are the platforms used to conduct our multiplexed biomarker screening. Several experiments to assess the effectiveness of an automated system were conducted. Precision and accuracy of hand pipetting verses automated pipetting was tested using aliquots of control materials. Assays were conducted over a three-day period by separate individuals and by the automation. Plate washing and ECL addition was held constant and carried out by the automation for both groups. Covered verses uncovered incubation determined if evaporative loss caused differences on biomarker detection levels. Sample incubation times were tested by shortening the interval from the recommended maximum to determine if less time had effect on detection levels. Automation liquid classes were verified using a Mettler Toledo balance and Hamilton software. Gravimetric measurements were made under control conditions (room temperature, humidity and reagent temperature). Different sample volumes were tested to determine the effect of freeze thaw on detection levels. Results: Although interestingly similar the automation pipetting profile was slightly different than the hand pipetting; automation was more precise. Covered verses uncovered incubation showed virtually no difference in detection levels. Shorting sample incubations times had little effect on detection levels. Gravimetric measurements of liquid classes showed excellent precision and accuracy. Sample volume size did influence detection levels, lower volumes showed a decrease in biomarker levels. Conclusions: Although differences are seen screening with automation verses hand processing, the significance is negligible. The automated diagnostic screening for detection of AD would be a cost effect, easily implemented solution and could become the first in line test.