Development of a targeted proteomic assay for the quantitation of neurofilament light in cerebrospinal fluid: Correlation with Simoa‐based quantitation across neurodegenerative diseases and centres

Abstract

AbstractBackgroundCerebrospinal fluid (CSF) neurofilament light (NfL) is a dynamic biomarker of neurodegeneration that is used for supporting clinical diagnosis and monitoring disease‐modifying therapies. Development of new analytical tests has been limited, with quantitation of NfL in biomarker studies largely restricted to immunoassay methods. Targeted mass spectrometry is a highly specific technique that can be easily translated into existing proteomic and clinical pathology laboratories. In this multi‐centre study, we have developed two targeted mass spectrometry assays which were compared to Simoa, the gold standard method for NfL quantitation. The developed methods included immunocapture of intact NfL protein or specific peptides (immune‐peptide capture) prior to targeted proteomic analyses. We discuss the development, validation and merits of each technology to quantitate NfL in CSF from healthy controls and several neurodegenerative diseases.MethodCSF was collected from patients with a range of neurological diseases at the National Hospital for Neurology and Neurosurgery (London, UK). Clinical diagnosis was determined by their treating clinician according to consensus clinical criteria. NfL concentrations were measured by Simoa (Quanterix, USA) and two immunoprecipitation‐targeted mass spectrometry (IP‐MS) assays developed by a collaboration between UCL and Washington University in St Louis. To evaluate replication of results across centres and mass spectrometry approaches, results were compared using Spearman’s correlation analysis and Bland‐Altman testing.ResultsA high correlation was observed for CSF NfL levels between Simoa and both IP‐MS assays (immune‐peptide capture, r = 0.90, P<0.0001 and immune‐protein capture, r = 0.88, P<0.0001). Relative differences in mean NfL concentration between clinical groups remained the same across the analytical methods, with the highest and lowest NfL concentrations observed in the Huntington’s disease and the healthy control groups respectively. Interestingly, differences in the extent of correlation between Simoa and IP‐MS were observed between the different clinical groups (r = 0.37 ‐ 0.94).ConclusionImmunoprecipitation‐targeted mass spectrometry quantitation of NfL correlates highly with NfL Simoa measures across a range of neurodegenerative and neuroinflammatory diseases, and is replicable between laboratories using different immunoprecipitation‐targeted mass spectrometry approaches.