Skip to main content
 

Adjusting neurofilament light levels using an MRI-based estimate of total axonal bulk impvroves its prediction of MS progression

Friday, September 14, 2018 — Poster Session V

12:00 p.m. – 1:30 p.m.
FAES Terrace
NIAID
NEURO-18

Authors

  • J Phillips
  • R Masvekar
  • P Kosa
  • K Jackson
  • A Weideman
  • B Bielekova

Abstract

Neurofilament Light Chain (NfL) is a scaffolding protein of the neural cytoskeleton with roles in axonal growth. Increased NfL levels in the cerebrospinal fluid (CSF) have been implicated as a biomarker of neurodegeneration in Multiple Sclerosis (MS) and other neurological diseases. Although increased NfL CSF levels have been associated with progression of neurological disability, the predictive power of this association is low. Because NfL is measured in CSF of HVs, often with values overlapping with neurological patients, we hypothesized that there are two pools of CSF NfL: 1. Physiological pool, associated with processes such as axonal maintenance and 2. Pathological pool, associated with axonal transections. The extension of this hypothesis is that the physiological pool of NfL will depend on the proportion of the axonal bulk (i.e., number of axons, which represents NfL source) to CSF volume (which represents the dilution factor). We tested this hypothesis by estimating axonal bulk and CSF volume from quantitative volumetric MRI imaging of the brain tissue and combined these measurements to derive NfL-adjustment “weight”, used to multiply measured CSF NfL concentrations. We assessed clinical utility of this adjustment by evaluating correlations of raw, or adjusted NfL values with MS severity, defined as speed of accumulation of neurological disability and measured by the Multiple Sclerosis Disease Severity Scale (MS-DSS). We observed that prior to adjustment, baseline NfL offers nearly zero predictive value (⍴=0.010, p=0.864, n=285), while MRI-adjusted baseline NfL has significantly higher predictive utility related to subsequent disability progression (⍴=0.261, p

Category: Neuroscience