Predicting cognitive decline with non-clinical markers in Parkinson's disease (PRECODE-2).

Objectives: To investigate whether baseline [I]FP-CIT SPECT and CSF markers can predict cognitive impairment (CI) in PD patients, and provide a profile of those most at risk.

Methods: 262 de novo PD patients from the Parkinson's Progression Markers Initiative database were stratified into two CI groups at the 36-month follow-up: MoCA-defined diagnosis: PD patients who had a MoCA score < 26; neuropsychological test-defined diagnosis: PD patients with MoCA-defined diagnosis and at least two test scores (of six; irrespective of test domain) greater than 1.5 standard deviation below the mean score in healthy controls. Predictive variables of CI were divided into deciles, providing us with ideal cutoff values for each variable.

Results: At the 36-month follow-up, 108/262 (41.2%) PD patients had CI as defined by the MoCA, of which 40/108 (37.0%) had neuropsychological test-defined CI. Baseline CSF Aβ42 (hazard ratio [HR]: 0.996, confidence interval [CI]: 0.992-0.999, p = 0.025), CSF total tau ([HR]: 1.023, [CI]: 1.002-1.044, p = 0.031) and caudate [I]FP-CIT SPECT uptake ([HR]: 0.332, [CI]: 0.115-0.960, p = 0.042) were predictors of CI. Patients with reduced CSF Aβ42 (< 384.6 pg/mL), increased CSF total tau (> 45.0 pg/mL) and reduced caudate [I]FP-CIT SPECT uptake (< 1.82) had a 65% risk of developing CI at 36-month follow-up.

Conclusion: We report a characteristic profile (reduced CSF Aβ42, increased CSF total tau and reduced caudate [I]FP-CIT SPECT uptake) that enables identification of early PD patients at risk of developing CI. These findings confirm previous reports of low CSF Aβ42, elevated CSF total tau and reduced dopaminergic integrity being associated with cognitive decline in PD.