Measuring neurofilament light in human plasma and cerebrospinal fluid: a comparison of five analytical immunoassays

Objectives Neurofilament light (NfL) is an established biofluid marker of neuroaxonal injury for neurological diseases. Several high-throughput and sensitive immunoassays have been developed to quantify NfL in blood and cerebrospinal fluid (CSF), facilitating the use of NfL as a biomarker in research and clinical practice. However, because of the lack of rigorous comparisons of assays, it has been difficult to determine whether data are comparable and whether assay performance differs. Here, we compared the performance of five NfL immunoassays.

To assess the five NfL immunoassays (Fujirebio, ProteinSimple, Quanterix, Roche and Siemens), we used pooled plasma or pooled CSF, as well as unique samples from 20 healthy controls and 20 individuals with El Escorial defined probable or definite amyotrophic lateral sclerosis (ALS), to evaluate precision, parallelism and/or bias. We also examined correlations between plasma and CSF NfL concentrations within and across assays and evaluated their ability to differentiate healthy controls from individuals with ALS.

Four of the five assays demonstrated exemplary performance based on our analyses of precision and parallelism. Across the five assays, NfL concentrations were lower in plasma than in CSF, although they displayed a high degree of correlation. We noted bias across assays; plasma NfL concentrations were lowest for the Roche assay and highest for the ProteinSimple assay. In addition, all assays reliably distinguished healthy controls from individuals with ALS using plasma or CSF NfL.

Four NfL assays demonstrated similar analytic performance. Alongside performance, other factors such as costs, accessibility, useability, footprint, and intended use, should be considered. Competing Interest Statement Competing interests US, EGR, AB, RK, NB and TFG have no competing or conflicting interests. RH and EdD are employees of Astex Pharmaceuticals, which is an operationally independent wholly owned subsidiary of Otsuka Pharmaceutical Co., Ltd., which contributed to the funding of the work. KF and DG are employees of Biogen, for which they receive salary and company stock as compensation. MB reports consulting fees from Alaunos, Alector, Arrowhead, Biogen, Denali, Eli Lilly, Novartis, Roche, uniQure, and Woolsey. The University of Miami has licensed intellectual property to Biogen to support design of the ATLAS study. GAH is an employee of the Rainwater Charitable Foundation, which contributed funding to this work. OIK is an employee of Alector LLC and may have an equity interest in Alector, Inc. Alector LLC is a member of the consortium and has contributed funding for the work. YL is an employee of Otsuka Pharmaceutical Development & Commercialization (OPDC), and OPDC is part of the consortium and contributed funding to the work. HZ has served at scientific advisory boards and/or as a consultant for Abbvie, Acumen, Alector, Alzinova, ALZpath, Amylyx, Annexon, Apellis, Artery Therapeutics, AZTherapies, Cognito Therapeutics, CogRx, Denali, Eisai, Enigma, LabCorp, Merry Life, Nervgen, Novo Nordisk, Optoceutics, Passage Bio, Pinteon Therapeutics, Prothena, Quanterix, Red Abbey Labs, reMYND, Roche, Samumed, Siemens Healthineers, Triplet Therapeutics, and Wave, has given lectures sponsored by Alzecure, BioArctic, Biogen, Cellectricon, Fujirebio, Lilly, Novo Nordisk, Roche, and WebMD, and is a co-founder of Brain Biomarker Solutions in Gothenburg AB (BBS), which is a part of the GU Ventures Incubator Program (outside submitted work). LLM is an employee of the Bluefield Project to Cure Frontotemporal Dementia, which contributed funding to this work. List of abbreviations - ALS - amyotrophic lateral sclerosis - AUC - area under the receiver operating curve - CI - confidence interval - CSF - cerebrospinal fluid - %CV - percent coefficient of variance - EDTA - ethylene diamine tetra-acetic acid - FTD - frontotemporal dementia - MS - multiple sclerosis - NfL - neurofilament light - r - Spearman’s correlation coefficient - REML - restricted maximum likelihood model - SD - standard deviation - SOD1 - superoxide dismutase 1