Towards quantifying axonal damage in blood samples from patients with neurological diseases.

Abstract

Reliable biomarkers of axonal damage are urgently needed in neurological diseases. Neurofilaments (Nf) are specific structural elements of neurons composed of at least three subunits: Nf light chain (NfL), Nf medium and Nf heavy chain (NfH). This PhD aimed to characterise NfL levels and their correlation with clinical features in patients with neurological diseases with a different rate of progression and following and under different treatment regimes. An important aim was also to develop a bioassay for NfL measurements in blood. Cerebrospinal fluid (CSF) NfL levels discriminated patients with a clinically isolated syndrome (CIS) (p=0.001) or multiple sclerosis (MS) (p=0.035) from healthy controls more efficiently, and was more sensitive to change after natalizumab therapy (p<0.0001) than CSF NfH (p=0.002). Further, CSF NfL levels decreased in fingolimodtreated MS patients (p=0.001), but not in those receiving placebo (p=0.433). Based on these findings, a sensitive method for the detection of NfL in serum was developed and validated. Patients with neurological diseases had higher serum NfL values than controls. In acute spinal cord injury (SCI), serum NfL levels correlated with injury severity and long-term motor outcome, and Minocycline treatment was associated with decreased NfL levels in complete SCI patients compared to placebo. Finally, I found that serum NfL levels were higher in CIS patients than in healthy controls but did not predict conversion to clinically definite MS (CDMS). Independent predictors of CDMS were instead oligoclonal bands, number of T2 lesions and age at CIS. Lower 25-OHvitamin D levels were associated with CDMS in univariate analysis, but this was attenuated in the multivariate model. In conclusion, NfL proved to be an analytically stable protein which is an important prerequisite for biomarkers. The role of NfL quantification as a surrogate measure of neuroaxonal damage is corroborated by my findings and further supports the usefulness of NfL as a putative biomarker of axonal damage in various neurological diseases.