Plasma neurofilament light outperforms glial fibrillary acidic protein in differentiating behavioural variant frontotemporal dementia from primary psychiatric disorders
(2024) In Journal of the Neurological Sciences 467.- Abstract
Objective: Timely, accurate distinction between behavioural variant frontotemporal dementia (bvFTD) and primary psychiatric disorders (PPD) is a clinical challenge. Blood biomarkers such as neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) have shown promise. Prior work has shown NfL helps distinguish FTD from PPD. Few studies have assessed NfL together with GFAP. Methods: We investigated plasma GFAP and NfL levels in participants with bvFTD, bipolar affective disorder (BPAD), major depressive disorder (MDD), treatment-resistant schizophrenia (TRS), healthy controls (HC), adjusting for age and sex. We compared ability of GFAP and NfL to distinguish bvFTD from PPD. Results: Plasma GFAP levels were significantly... (More)
Objective: Timely, accurate distinction between behavioural variant frontotemporal dementia (bvFTD) and primary psychiatric disorders (PPD) is a clinical challenge. Blood biomarkers such as neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) have shown promise. Prior work has shown NfL helps distinguish FTD from PPD. Few studies have assessed NfL together with GFAP. Methods: We investigated plasma GFAP and NfL levels in participants with bvFTD, bipolar affective disorder (BPAD), major depressive disorder (MDD), treatment-resistant schizophrenia (TRS), healthy controls (HC), adjusting for age and sex. We compared ability of GFAP and NfL to distinguish bvFTD from PPD. Results: Plasma GFAP levels were significantly (all p < 0.001) elevated in bvFTD (n = 22, mean (M) = 273 pg/mL) compared to BPAD (n = 121, M = 96 pg/mL), MDD (n = 42, M = 105 pg/mL), TRS (n = 82, M = 67.9 pg/mL), and HC (n = 120, M = 76.8 pg/mL). GFAP distinguished bvFTD from all PPD with an area under the curve (AUC) of 0.85, 95 % confidence interval [0.76, 0.95]. The optimal cut-off of 105 pg/mL was associated with 73 % specificity and 86 % sensitivity. NfL had AUC 0.95 [0.91, 0.99], 13.3 pg/mL cut-off, 88 % specificity, 86 % sensitivity, and was superior to GFAP (p = 0.02863) and combination of GFAP and NfL (p = 0.04726). Conclusions: This study found elevated GFAP levels in bvFTD compared to a large cohort of PPD, but NfL levels exhibited better performance in this distinction. These findings extend the literature on GFAP in bvFTD and build evidence for plasma NfL as a useful biomarker to assist with distinguishing bvFTD from PPD. Utilisation of NfL may improve timely and accurate diagnosis of bvFTD.
(Less)
- author
- author collaboration
- organization
- publishing date
- 2024-12-15
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Behavioural variant frontotemporal dementia, Biomarkers, Dementia, Diagnosis, Psychiatry
- in
- Journal of the Neurological Sciences
- volume
- 467
- article number
- 123291
- publisher
- Elsevier
- external identifiers
-
- pmid:39577322
- scopus:85209351795
- ISSN
- 0022-510X
- DOI
- 10.1016/j.jns.2024.123291
- language
- English
- LU publication?
- yes
- additional info
- Publisher Copyright: © 2024 The Authors
- id
- 495d1005-9f25-43aa-9802-81d4bf36eeab
- date added to LUP
- 2024-12-09 06:46:47
- date last changed
- 2025-07-07 23:33:11
@article{495d1005-9f25-43aa-9802-81d4bf36eeab, abstract = {{<p>Objective: Timely, accurate distinction between behavioural variant frontotemporal dementia (bvFTD) and primary psychiatric disorders (PPD) is a clinical challenge. Blood biomarkers such as neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) have shown promise. Prior work has shown NfL helps distinguish FTD from PPD. Few studies have assessed NfL together with GFAP. Methods: We investigated plasma GFAP and NfL levels in participants with bvFTD, bipolar affective disorder (BPAD), major depressive disorder (MDD), treatment-resistant schizophrenia (TRS), healthy controls (HC), adjusting for age and sex. We compared ability of GFAP and NfL to distinguish bvFTD from PPD. Results: Plasma GFAP levels were significantly (all p < 0.001) elevated in bvFTD (n = 22, mean (M) = 273 pg/mL) compared to BPAD (n = 121, M = 96 pg/mL), MDD (n = 42, M = 105 pg/mL), TRS (n = 82, M = 67.9 pg/mL), and HC (n = 120, M = 76.8 pg/mL). GFAP distinguished bvFTD from all PPD with an area under the curve (AUC) of 0.85, 95 % confidence interval [0.76, 0.95]. The optimal cut-off of 105 pg/mL was associated with 73 % specificity and 86 % sensitivity. NfL had AUC 0.95 [0.91, 0.99], 13.3 pg/mL cut-off, 88 % specificity, 86 % sensitivity, and was superior to GFAP (p = 0.02863) and combination of GFAP and NfL (p = 0.04726). Conclusions: This study found elevated GFAP levels in bvFTD compared to a large cohort of PPD, but NfL levels exhibited better performance in this distinction. These findings extend the literature on GFAP in bvFTD and build evidence for plasma NfL as a useful biomarker to assist with distinguishing bvFTD from PPD. Utilisation of NfL may improve timely and accurate diagnosis of bvFTD.</p>}}, author = {{Eratne, Dhamidhu and Kang, Matthew J.Y. and Lewis, Courtney and Dang, Christa and Malpas, Charles and Ooi, Suyi and Brodtmann, Amy and Darby, David and Zetterberg, Henrik and Blennow, Kaj and Berk, Michael and Dean, Olivia and Bousman, Chad and Thomas, Naveen and Everall, Ian and Pantelis, Chris and Wannan, Cassandra and Cicognola, Claudia and Hansson, Oskar and Janelidze, Shorena and Santillo, Alexander F. and Velakoulis, Dennis}}, issn = {{0022-510X}}, keywords = {{Behavioural variant frontotemporal dementia; Biomarkers; Dementia; Diagnosis; Psychiatry}}, language = {{eng}}, month = {{12}}, publisher = {{Elsevier}}, series = {{Journal of the Neurological Sciences}}, title = {{Plasma neurofilament light outperforms glial fibrillary acidic protein in differentiating behavioural variant frontotemporal dementia from primary psychiatric disorders}}, url = {{http://dx.doi.org/10.1016/j.jns.2024.123291}}, doi = {{10.1016/j.jns.2024.123291}}, volume = {{467}}, year = {{2024}}, }