according to an analysis published online ahead of print June 17 in JAMA Neurology. The biomarker has the potential to distinguish between frontotemporal dementia (FTD) and other dementia subtypes, as well as between Parkinson’s disease and atypical parkinsonian syndromes, said the investigators. It may be necessary to identify age- and sex-specific reference values for NfL, they added.
Neurologists have long understood CSF levels of NfL to be elevated in neurodegenerative conditions, but researchers previously had not compared these levels systematically among neurologic disorders. Similarly, the literature indicates a positive association between CSF NfL level and age in healthy controls, but this association has not been evaluated systematically in neurologic disorders. The resulting lack of clarity has impeded the use of NfL as a diagnostic biomarker.
A meta-analysis of CSF samples
Claire Bridel, MD, PhD, of the department of clinical chemistry at the VU University Medical Centre in Amsterdam and colleagues conducted a systematic review and meta-analysis to compare CSF levels of NfL among diagnoses, assess the associations of age and sex with NfL, and evaluate the potential of NfL as a diagnostic biomarker. The investigators searched PubMed for studies published between Jan. 1, 2006, and Jan. 1, 2016, that reported CSF levels of NfL in neurologic or psychiatric conditions or in healthy controls. They included only studies that used the same commercially available immunoassay that has been used in most studies since 2006. The literature indicates that this enzyme-linked immunosorbent assay is sensitive and robust. Dr. Bridel and colleagues contacted study authors and requested their individual-level data.
The investigators sorted the most common neurologic conditions into three groups of similar disorders. The first group included inflammatory conditions of the CNS, such as multiple sclerosis, clinically isolated syndrome (CIS), and optic neuritis. The second group included dementia syndromes (such as Alzheimer’s disease, FTD, vascular dementia, and dementia with Lewy bodies) and amyotrophic lateral sclerosis (ALS). The third category included parkinsonian syndromes such as Parkinson’s disease, Parkinson’s disease dementia, multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and corticobasal syndrome (CBS). The authors used generalized linear mixed-effects models to estimate the fixed effects of age, sex, and diagnosis on log-transformed NfL levels. They modeled cohort of origin as a random intercept.
NfL increased with age
Dr. Bridel and colleagues identified 153 relevant investigations, of which 44 met their inclusion criteria. The original investigators provided data sets for these studies, along with three previously unpublished data sets. The data sets included information from 10,059 participants (mean age, 59.7 years; 54.1% female). After excluding diagnostic categories with fewer than five observations per sex, Dr. Bridel and colleagues included data for 10,012 people in the analysis. In this population, the researchers identified 2,795 patients with inflammatory diseases of the CNS, 4,284 patients with dementia or predementia, 984 patients with parkinsonian disorders, and 1,332 healthy controls.
CSF level of NfL was elevated in most neurologic conditions, compared with healthy controls. The largest effect sizes were in cognitively impaired patients with HIV (21.36), patients with FTD/ALS (10.48), patients with ALS (7.58), and patients with Huntington’s disease (5.88).
In healthy controls, the level of NfL in CSF increased by 3.30% annually. The investigators also observed an association between age and CSF NfL level in people with subjective complaints, bipolar disorder, and most neurodegenerative conditions. They found no association, however, in patients with MS, HIV and cognitive impairment, and rapidly progressive neurodegenerative conditions (such as FTD, ALS, FTD/ALS, MSA, PSP, CBS, and Huntington’s disease). CSF level of NfL was 26.0% higher in men among healthy controls. This discrepancy also was observed in a minority of neurologic conditions, including MS, Alzheimer’s disease, vascular dementia, and Parkinson’s disease.
Mean CSF levels of NfL were similar between patients with inflammatory conditions of the CNS. Among dementias and related disorders, mean CNS level of NfL was significantly higher in FTD than in Alzheimer’s disease (2.08), vascular dementia (1.56), and dementia with Lewy bodies (2.50). Among parkinsonian syndromes, the mean CSF levels of NfL were higher in MSA, PSP, and CBS, compared with Parkinson’s disease.