Neurofilament light chain reference values in serum and cerebrospinal fluid: a bi-compartmental analysis in neurological diseases

Steffen Halbgebauer,Veronika Klose,Badrieh Fazeli,Paula Klassen,Christoforos Alexudis,Gabriele Nagel,Angela Rosenbohm,Dietrich Rothenbacher,Nerea Gomez de San Jose,Simon Witzel,Zeynep Elmas,Maximilian Wiesenfarth,Joachim Schuster,Johannes Dorst,Andre Huss,Franziska Bachhuber,Markus Otto,Georg Bernhard Landwehrmeyer,Albert Christian Ludolph,Hayrettin Tumani
DOI: https://doi.org/10.1101/2024.11.21.24317697
2024-11-22
Abstract:Background: Concentrations of neurofilament light chain (NfL), a neuroaxonal damage marker, increase with age. Therefore, age-dependent reference values are important in clinical practice. However, so far these have only been established with a bead based system and age-dependent z-scores for CSF are missing. In addition, we here propose how the combined analysis of CSF and serum NfL could help in the discrimination between central (CNS) and peripheral nervous system (PNS) axonal degeneration. Methods: For the calculation of age-reference values for serum and CSF 1514 NfL concentrations of control subjects determined with the microfluidic Ella system, were applied. Results: Age-dependent NfL levels were calculated with additive quantile regression and presented with percentiles and z-scores. We observed a non-linear increase of NfL in serum and CSF. The spearman r of the association with age was 0.81 (95% CI: 0.78-0.83), p<0.0001 and 0.82 (95% CI: 0.79-0.85), p<0.0001 for serum and CSF NfL, respectively. Serum and CSF NfL levels were also associated with each other (r=0.68 (95%CI: 0.62-0.73), p<0.0001). Furthermore, we used this association to establish a bi-compartmental CSF and serum NfL model allowing to differentiate between peripheral or central origin of neurodegeneration. Conclusions: The age-reference curves corroborate findings of an exponential elevation of NfL in serum and CSF with increasing age. As NfL values from different platforms are not interchangeable this is of additional high relevance. Moreover, the association between CSF and serum NfL values could be applied for clinical use regarding overlapping symptoms of CNS and PNS based neurological diseases.
What problem does this paper attempt to address?