Objective/Rationale:
The diagnosis of Parkinson disease is based on clinical criteria and is labor-intensive, expensive and often inconclusive. Unfortunately, there are currently no accepted biological markers to assist in the diagnosis of PD. Oxidative stress (free radical damage) has been implicated in the development of PD and affects both brain and peripheral tissues. The primary objective of this study is to determine whether measurement of oxidatively-modified blood constituents by near-infrared (NIR) spectroscopy distinguishes PD patients from normal subjects.
Project Description:
Blood samples will be procured at intervals over a two-year period from a meticulously characterized PD population of over 400 individuals (PostCEPT study cohort) and equal numbers of carefully matched control subjects. Blood plasma will be separated, frozen and shipped to Molecular Biometrics (MB) in Montreal. Plasma spectral signatures will be obtained using an NIR spectrometer and analyzed by a) hypothesis-based pre-selection of wavelength regions reflecting known oxidative substrate modifications and b) empirical computer-driven algorithms providing optimal diagnostic separation between the PostCEPT and control cohorts. We will determine the sensitivities, specificities, diagnostic accuracy, and positive/negative predictive values by which both models differentiate PD from normal aging. As secondary aims, plasma biospectra will be analyzed as a function of disease stage, age, gender, medications, cognitive-neuropsychological scores, concurrent chronic illness and neuroimaging.
Relevance to Diagnosis/Treatment of Parkinson’s Disease:
The advent of a biological marker that differentiates PD from normal aging and other movement disorders would represent a significant achievement in the management of this debilitating neurodegenerative condition. A successful NIR-based diagnostic blood test for PD would be novel, minimally-invasive, inexpensive, amenable to high-throughput sampling and easily implemented in most hospitals and clinics. Plasma NIR biospectroscopy may also prove useful in monitoring PD progression and the impact of neuroprotective and other disease-modifying interventions.
Anticipated Outcome:
Based on a sound theoretical platform and our initial data sets, we anticipate that plasma NIR spectroscopy will distinguish early PD from healthy individuals with 80% accuracy or greater. In the course of this study, we will learn whether the test additionally informs on the clinical stage of the disease (useful for clinical drug trial patient stratification, etc.) and whether anti-PD medications and concurrent illness impact the relevant biospectra.
Progress Report
The diagnostic evaluation of patients with Parkinson's disease (PD) is based entirely on clinical criteria and is labor-intensive, expensive and often inconclusive in the earlier stages of the illness. Unfortunately, there are currently no accepted biological markers in routine clinical use to assist in the diagnosis of PD. Molecular Biometrics is developing a diagnostic blood test for PD, based on the link between oxidative stress, PD and the unique analytical capacity of the company’s proprietary biospectroscopy based metabolomics (BSM) platform. Rather than assess single chemical constituents in isolation, this method utilizes a systems biology (metabolomic) approach to generate a biological 'fingerprint' by simultaneously detecting and integrating chemical modifications among various functional groups. In the current study, funded in part by MJFF, the company is aiming to validate the potential of blood biospectroscopy for the diagnosis of PD in a large, multi-center patient cohort. To date, over 300 blood plasma samples have been obtained from normal controls and the PostCEPT cohort, an extensively characterized population of PD subjects. Analysis of this sample pool is set to begin in Fall 2010.
Final Outcome
MB has investigated the possibility of developing a biomarker for PD in blood, based on the link between oxidative stress, PD and the unique measurement capacity of vibrational spectroscopy. The company has utilized its novel bioinformatics platform to develop a rapid, high-throughput metabolomic assay for PD using customized, inexpensive spectroscopes and mathematical models sensitive to oxidative substrate modifications. This biospectroscopic assay objectively assesses the status of the patient and eliminates the subjectivity and variability associated with physician interpretation of clinical signs.
The initial model developed from the first 100 samples showed an improved p-value ( 0.0047 vs 0.069 previously) with increased number of samples in the blind validation group. The permutation test showed that the chance of obtaining this value from ramdomized data is less than 1.7%. This highly significant separation indicates that there is an underlying signal within this cohort of PD samples that allows differentiation of normal and PD patients. The ROC curve however also shows that there is a significant overlap in scores between the two groups.