Study Rationale:
Autoantibodies are antibodies that mistakenly target and react with a person's own tissues or organs. Autoantibodies can be detected many years before symptoms develop in many diseases. Accumulating evidence suggests that Parkinson’s disease (PD) begins years before clinical motor symptoms are detectable. We will use a laboratory tool called a microarray containing 1600+ immune-related human antigens (molecules capable of stimulating an immune response) to identify autoantibody-based biomarkers in high-risk Parkinson’s patients from the Parkinson Associated Risk Syndrome (PARS) study. We will use machine learning — a type of artificial intelligence that enables computers to learn without being explicitly programmed — to stratify the spectrum of Parkinson’s based on an autoantibody signature. With each exposure to new data, a machine learning algorithm grows increasingly better at recognizing patterns over time. We will then conduct a longitudinal study to determine the velocity of change of the autoantibody biomarker with progression of Parkinson’s, using a PARS custom chip designed based on the biomarkers identified from the initial phase.
Hypothesis:
We hypothesize that machine learning approaches will enable us to combine autoantibody biomarkers measured in body fluids with detailed clinical annotation provided for the PARS cohort. This will identify discrete clusters of high-risk patients with pre- and non-motor symptoms, and will stratify ‘no neurological disease’ patients in the PARS cohort according to subsequent disease trajectory.
Study Design:
In the initial phase, we will acquire plasma samples from patients in the PARS cohort with no neurological disorder and will conduct an autoantibody test using the microarray tool to identify biomarkers that differentiate these two groups, as well as sub-clusters within each group, using a machine learning approach. The identified biomarkers will be used to develop a custom PARS chip, which will be validated using samples from the PARS cohort collected at three different time points through disease progression and will determine the velocity of change of the autoantibody signature with progression of PD.
Impact on Diagnosis/Treatment of Parkinson’s Disease:
Autoantibodies are detectable many years before clinical symptoms and have been observed in an ever-widening range of diseases including PD, suggesting that novel autoantibodies might be potential biomarkers for early diagnosis in Parkinson’s. Such autoantibody biomarkers of PD measured in liquid biopsies have the potential to revolutionize approaches to early detection and progression monitoring.
Next Steps for Development:
We will determine the clinical performance and likely clinical impact of our custom PARS chip as a diagnostic tool in a larger Parkinson’s study cohort. Based on the outcome, we will submit documentation to the FDA for approval as a clinical diagnostic. In addition, we will use retrospective clinical trial samples to determine whether our autoantibody signatures can distinguish Parkinson’s patients who will respond to a particular treatment from those who will not.