Predicting Cerebrospinal Fluid Alpha-Synuclein Seed Amplification Assay Status from Demographics and Clinical Data

Area of Science:

• Neurology
• Biomarker Discovery
• Clinical Prediction Modeling

Background:

• Alpha-synuclein (a-syn) is a key protein implicated in Parkinson's disease (PD) pathogenesis.
• Accurate prediction of a-syn status can aid in early diagnosis and disease management.
• Current methods for a-syn detection often require invasive procedures.

Purpose of the Study:

• To develop and validate predictive models for alpha-synuclein (a-syn) status.
• To utilize easily accessible clinical predictors for in vivo a-syn status prediction.
• To assess model performance in populations with and without Parkinson's disease.

Main Methods:

• Logistic regression models were developed using data from the Parkinson Progression Marker Initiative (PPMI) study.
• Models were trained to predict cerebrospinal fluid (CSF) a-syn status measured by seeding amplification assay (SAA).
• External validation was performed on the Systemic Synuclein Sampling Study (S4) cohort.

Main Results:

• The multivariable model achieved high internal performance (AUROC 0.920) and external validation performance (AUROC 0.976).
• Key predictors included age- and sex-specific University of Pennsylvania Smell Identification Test (UPSIT) percentile values, sex, constipation, LRRK2, and GBA status.
• Models relying solely on UPSIT percentile also demonstrated strong predictive accuracy.

Conclusions:

• Data-driven models incorporating non-invasive clinical features can accurately predict CSF a-syn SAA status.
• The University of Pennsylvania Smell Identification Test (UPSIT) scores were highly significant predictors of a-syn SAA status.
• These findings support the use of accessible clinical data for predicting a-syn status in PD research.