By Anna S. Nordvig, MD, MS
Synopsis: Behavioral-variant frontotemporal dementia (bvFTD) can be difficult to distinguish from a primary mood disorder at the early stages of disease. Biological markers, such as serum or cerebrospinal fluid neurofilament light chain, can help to make a distinction between the disorders, but the test carries a high false-negative rate.
Source: Davydow DS,Brasfield M, Morrow CB, et al. Neurofilament light chain and differentiation of behavioral variant frontotemporal dementia from psychiatric disorders: A systematic review. JAMA Psychiatry. 2025; Sep 10. doi: 10.1001/jamapsychiatry.2025.2429. [Online ahead of print].
Frontotemporal dementia (FTD) is a type of neocortical neurodegenerative disease with three established subtypes: two with language disturbances (aphasia) and one with prominent progressive behavioral and personality changes (behavioral variant FTD [bvFTD]), including apathy, disinhibition, loss of interest in socialization and self-care, and socially inappropriate behaviors. Patients first presenting with psychiatric symptoms such as major depression and obsessive-compulsive disorder rather than cognitive symptoms may experience a delay in FTD diagnosis. Existing diagnostic imaging may lack sensitivity and specificity. This study aims to determine whether serum or cerebrospinal fluid (CSF) neurofilament light (NfL), generally known as a sensitive but nonspecific marker of neuroaxonal damage, is a useful biomarker for distinguishing bvFTD from primary psychiatric disorders (PPD).
A systematic review searched PubMed, Embase, Web of Science, PsycINFO, and the Cochrane Collaborative according to Preferred Reporting Items for a Systematic Review and Meta-Analysis of Diagnostic Test Accuracy (PRISMA-DTA) guidelines. Eligible articles included cohorts of patients with bvFTD and psychiatric disorders ascertained using validated methods and conducted analyses to evaluate NfL from CSF and/or blood in distinguishing bvFTD from psychiatric disorders. The gold standard for diagnosis of FTD in these studies included consensus conferences and use of current clinical criteria.
Of 3,828 titles and abstracts and 434 full-text articles, 12 articles met eligibility criteria. The studies included 694 unique patients with bvFTD and 1,594 unique patients with a range of psychiatric disorders. The results of this paper were based on four studies measuring NfL levels in CSF, six studies measuring NfL levels in blood, and two studies measuring NfL levels in both. Heterogenous methods of biomarker determination were used, including enzyme-linked immunosorbent assay (ELISA) and Simoa. Overall, levels of NfL in CSF and blood were significantly higher among patients with bvFTD than among patients with psychiatric disorders. The area under the curve (AUC) for NfL for bvFTD vs. psychiatric disorders ranged from 0.86 to 0.95 (sensitivities: 63% to 96%, specificities: 81% to 100%) in CSF and from 0.79 to 0.98 (sensitivities: 65% to 100%, specificities: 69% to 96%) in blood. This suggests a false-negative rate of up to 35% for CSF and 37% for blood.
Commentary
Psychiatric clinical symptoms in the spectrum of neocortical neurodegenerative disorders are widespread and increasingly recognized as prominent and sometimes presenting features. BvFTD particularly is known for causing mood and behavior changes. These symptoms often are phenotypically indistinguishable from primary isolated psychiatric disorders. Serum and CSF NfL testing is a promising diagnostic test to distinguish bvFTD from psychiatric disorders, but concerns about validity and utility remain.
In planning diagnostic testing, including measurements of NfL in blood or CSF, clinicians also should be aware of the overlap of symptoms between bvFTD and frontal-variant Alzheimer’s disease (AD), as well as frontal and limbic presentations of other cortical conditions and other neurodegenerative disease, vascular and microvascular disease, focal lesions, infection, and brain injury. Many of these disorders also may show increases in NfL. Also, when evaluating for neurodegenerative disease in a patient with psychiatric presentations, biomarkers such as ptau217 and AB42/40 ratio testing are helpful in identifying cases caused by frontal variant AD. Also, NfL may increase with age, so results should be interpreted with more caution in older patients.
Genetic testing may be another useful diagnostic approach for patients when bvFTD is suspected. Genetic subtypes continue to be identified and are present in 15% to 20% of FTD patients. Some of these also express different serum biomarkers, such as progranulin levels in GRN gene mutation carriers.
Anna S. Nordvig, MD, MS, is Assistant Professor of Neurology and Neuroscience, Memory Disorders Clinic, Weill Cornell Medicine.
