< back to all PPMI Presentations

Identifying drug-naïve Parkinsonism with Non-motor features & CSF biomarkers: A case-control study

Jain, S; Young Park, S; Comer, D; Rudoy, D; Ivanco, L
AAN: Washington, DC
Download Presentation: Jain-poster-classification-AAN-2015.pdf
OBJECTIVE: We hypothesized non-motor features or CSF biomarkers would aid in identifying early Parkinsonism.

BACKGROUND: Given that non-motor features can predate the onset of Parkinsonian motor features, non-motor features or CSF biomarkers associated with neurodegenerative processes may be useful in identifying those likely to have PD diagnosis earlier. This would allow earlier treatment of non-motor features and testing of potential neuroprotective agents. To develop a model identifying those likely to have PD without using motor features, we classified recently diagnosed, medication naïve Parkinsonism with confirmed striatal dopaminergic deficits and healthy controls using non-motor features and potential biomarkers. We then externally validated our findings in an independent cohort.

METHODS: Non-motor features (cognitive, olfactory, autonomic, psychiatric and sleep disturbances) and CSF biomarkers (?-synuclein, tau phosphorylated at threonine 181, total tau, and ?-amyloid 1-42) were obtained from early, medication naïve Parkinsonian participants and controls as part of the Parkinson?s Progression Marker Initiative (PPMI). Only participants with confirmed dopamine transport deficits by DaTScan were included in the Parkinsonism group. Scans without evidence of dopaminergic deficits were excluded from analyses. Group comparisons of Parkinsonism and controls were done with t tests, Wilcoxon rank sum test, and chi-square tests. Classification performance was evaluated by receiver operating characteristic (ROC) curves for each predictor and a multivariable logistic model (Parkinsonism vs. controls). Results were internally validated with 10-fold cross-validation, and externally validated in the Parkinson Autonomic and other Non-motor features Study (PANS).

KEY FINDINGS: Participants with early Parkinsonism had more severe non-motor features and lower concentrations of all CSF biomarkers than controls. A model with olfactory function (University of Pennsylvania Smell Inventory (UPSIT)) and a composite score of other non-motor features (Movement Disorders Society Unified Parkinson Disease Rating Scale Part 1 (MDS-UPDRS Part 1)) classified early Parkinsonism with dopaminergic deficiency in the PPMI with 89% sensitivity and 79% specificity, and more advanced PD in PANS with 100% sensitivity and 90% specificity.

CONCLUSIONS: To our knowledge, this is the first report of classifying Parkinsonism and controls using several non-motor features and potential CSF biomarkers with both internal and external validation. The MDS-UPDRS Part 1 used with UPSIT identified those with Parkinsonism with dopaminergic deficiency. CSF biomarkers as measured in the PPMI were not helpful. Previous case-control studies have reported similar results for olfactory testing (UPSIT). To our knowledge the findings of the MDS-UPDRS Part 1 are novel. Study is needed in longitudinal cohorts with incident PD cases where non-motor features may pre-date motor findings, and may aid in identifying likely PD.