Validation and generation of age-specific reference intervals for a new blood neurofilament light chain assay
Ronald A. Booth, Daniel R. Beriault, Raphaël Schneider, Pascal Benkert, Eline A.J. Willemse, Aleksandra Maleska Maceski, David Leppert, Mark S. Freedman, Jens Kühle
Abstract
BACKGROUND: Neurofilament light chain (NfL) is the first blood biomarker for multiple sclerosis (MS) with utility for prognosis and disease monitoring. The Simoa NF-light is a research-based assay and is the most highly cited NfL method available, however development of assays specifically designed for routine clinical laboratories are necessary for clinical adoption. The Roche Elecsys NfL assay was developed specifically for routine laboratory use. METHODS: Specimens from 718 MS patients, paired serum-plasma and plasma-CSF specimens were used to compare the Simoa and Elecsys assays. Elecsys NfL precision, linearity, parallelism, limit-of-quantitation and interferences were assessed. RESULTS: Elecsys precision ranged from 1.8 % to 7.3 % coefficient of variation. Serum samples showed a high correlation (Pearson's r = 0.991) with a mean bias of -85.1 % compared to Simoa. Plasma-CSF correlations were high (Pearson's r = 0.993), with CSF values 54.8-fold higher than plasma. Age-specific 97.5th percentile reference limits were generated by transforming Elecsys to Simoa NfL values. The previously developed Simoa NfL reference app was extended to the Elecsys platform using the transformed Simoa reference values to provide age and BMI-adjusted Z scores and percentiles. CONCLUSIONS: The Elecsys NfL assay demonstrated precision and accuracy acceptable for use in clinical laboratories. The Elecsys assay generates NfL concentrations numerically lower than Simoa, however the results are highly correlated, requiring the need for Elecsys-specific reference values.