Objective: Identify Parkinson’s disease (PD) subpopulation clusters present at disease onset using a number of motor and non-motor phenotypes.  

Background: PD is considered a syndrome, which presents clinically as variable impairment in different motor and non-motor features. It is a common clinical observation that PD subjects not only show different rates of progression, but also great variability in motor and non-motor features.

Methods: Motor and non-motor scores in 407 Parkinson’s Progression Markers Initiative (PPMI) subjects were identified using an ordinal variable specific factor analysis. We then used these to identify patient clusters which a similar presentation at disease onset using a non-supervised, data driven clustering approach.

Results: Twenty-one factors were identified among the evaluated clinical scales: 3 for UPDRS-I, 1 for UPDRS-II, 4 for UPDRS-III, 2 for the cognitive scores, 4 for anxiety (STAI), 1 for sleep (ESS), 3 for depression (GDS), 1 for impulse control disorders (ICD), 4 for autonomic symptoms (SCOPA). In the analysis four different clusters were identified. Main differences among groups were in 1) anxiety PC of STAI and UPDRS I scales, with clusters 2 and 4 having greater symptoms, 2) memory PC with cluster 4 being more amnestic, 3) lower scores in perspiration, salivation and swallowing, constipation, sleep disturbance, postural and kinetic tremor in cluster 3, 4) greater depression was present in cluster 4, 5) greater rigidity and bradykinesia was seen in cluster 4. Differences were observed in longitudinal progression.

Conclusions: Our analysis in early stage medication-naïve PD patients identified 4 clusters of patient presentations, differing in the severity of motor and non-motor aspects. These patients showed differences in disease progression during follow-up. Indicating that individualized treatment for non-motor symptoms and cognitive impairment might be needed in the different PD groups.

« Back to 2018 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/early-stage-parkinsons-disease-phenotypes/