Combined CSF α-SYN RT-QuIC, CSF NFL and midbrain-pons planimetry in degenerative parkinsonisms: From bedside to bench, and back again.

Abstract

Introduction Differential diagnosis between Parkinson's disease (PD) and atypical parkinsonisms (APs: multiple system atrophy[MSA], progressive supranuclear palsy[PSP], corticobasal degeneration[CBD]) remains challenging. Lately, cerebrospinal fluid (CSF) studies of neurofilament light-chain (NFL) and RT-QuIC of alpha-synuclein (α-SYN) have shown promise, but data on their combination with MRI measures is lacking. Objective (1) to assess the combined diagnostic ability of CSF RT-QuIC α-SYN, CSF NFL and midbrain/pons MRI planimetry in degenerative parkinsonisms; (2) to evaluate if biomarker-signatures relate to clinical diagnoses and whether or not unexpected findings can guide diagnostic revision. Methods We collected demographic and clinical data and set up α-SYN RT-QuIC at our lab in a cross-sectional cohort of 112 participants: 19 control subjects (CSs), 20PD, 37MSA, 23PSP, and 13CBD cases. We also determined CSF NFL by ELISA and, in 74 participants (10CSs, 9PD, 26MSA, 19PSP, 10CBD), automatized planimetric midbrain/pons areas from 3T-MRI. Results Sensitivity of α-SYN RT-QuIC for PD was 75% increasing to 81% after revisiting clinical diagnoses with aid of biomarkers. Sensitivity for MSA was 12% but decreased to 9% with diagnostic revision. Specificities were 100% against CSs, and 89% against tauopathies raising to 91% with diagnostic revision. CSF NFL was significantly higher in APs. The combination of biomarkers yielded high diagnostic accuracy (PD vs. non-PD AUC = 0.983; MSA vs. non-MSA AUC = 0.933; tauopathies vs. non-tauopathies AUC = 0.924). Biomarkers-signatures fitted in most cases with clinical classification. Conclusions The combination of CSF NFL, CSF RT-QuIC α-SYN and midbrain/pons MRI measures showed high discriminant ability across all groups. Results opposite to expected can assist diagnostic reclassification.