Plasma neurofilament light chain as a predictor of multiple system atrophy in idiopathic REM sleep behavior disorder

Abstract

Background: The predictive value of plasma biomarkers for phenoconversion in idiopathic REM sleep behavior disorder (iRBD) remains uncertain.

Objective: To evaluate the utility of plasma neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), and DOPA decarboxylase (DDC) levels for predicting phenoconversion in iRBD.

Methods: Fifty patients with polysomnography-confirmed iRBD and 38 matched healthy controls (HCs) were prospectively enrolled. Plasma NfL and GFAP were measured using SIMOA at baseline and after phenoconversion, and DDC levels via ELISA. Clinical assessments included UPDRS III, MMSE, MoCA, Geriatric Depression Scale, and SCOPA-AUT, and prodromal markers predictive of phenoconversion.

Results: Over 3.7 years, 22 patients phenoconverted (7 multiple system atrophy [MSA]; 15 Lewy body diseases [LBDs]). Baseline NfL levels were higher in converters (median, 14.00 pg/mL), especially MSA converters (16.10 pg/mL), than in non-converters (11.15 pg/mL) and HCs (11.55 pg/mL) (p = 0.025 and p = 0.009, respectively). Baseline GFAP and DDC levels did not differ. In MSA converters, NfL correlated with SCOPA-AUT score (r = 0.925, p = 0.008), and a cutoff of 12.60 pg/mL predicted MSA phenoconversion (sensitivity 100%, specificity 62.8%, AUC = 0.83, p = 0.003). Post-phenoconversion NfL levels were elevated in converters (18.80 pg/mL) than in non-converters (13.05 pg/mL) (p < 0.001), highest in MSA converters (24.45 pg/mL; p < 0.001). GFAP showed a marginal increase (p = 0.051), while DDC showed no significant differences.

Conclusions: Plasma NfL is a promising biomarker for predicting phenoconversion in iRBD, particularly to MSA, and may reflect underlying autonomic dysfunction.

Keywords: DOPA decarboxylase; Dementia with Lewy bodies; Glial fibrillary acidic protein; Multiple system atrophy; Neurofilament light chain; Parkinson’s disease; Synucleinopathy.