Cerebrospinal fluid neurofilament light chain differentiates primary psychiatric disorders from rapidly progressive, Alzheimer's disease and frontotemporal disorders in clinical settings
Dhamidhu Eratne, Samantha M. Loi, Qiao‐Xin Li, Christiane Stehmann, Charles B. Malpas, Alexander Santillo, Shorena Janelidze, Claire J. Cadwallader, Nirbaanjot Walia, Blair Ney, Victoria Lewis, Matteo Senesi, Christopher Fowler, Amelia McGlade, Shiji Varghese, Parsa Ravanfar, Wendy Kelso, Sarah Farrand, Michael Keem, Matthew Kang, Anita Goh, Kunal Dhiman, Veer Bala Gupta, Rosie Watson, Nawaf Yassi, Cath Kaylor‐Hughes, Richard Kanaan, Piero Perucca, Hannah Dobson, Lucy Vivash, Rashida Ali, Terence J. O’Brien, Oskar Hansson, Henrik Zetterberg, Kaj Blennow, Mark Walterfang, Colin L. Masters, Samuel F. Berkovic, Steven Collins, Dennis Velakoulis, The MiND Study Group
Abstract
INTRODUCTION: Many patients with cognitive and neuropsychiatric symptoms face diagnostic delay and misdiagnosis. We investigated whether cerebrospinal fluid (CSF) neurofilament light (NfL) and total-tau (t-tau) could assist in the clinical scenario of differentiating neurodegenerative (ND) from psychiatric disorders (PSY), and rapidly progressive disorders. METHODS: Biomarkers were examined in patients from specialist services (ND and PSY) and a national Creutzfeldt-Jakob registry (Creutzfeldt-Jakob disease [CJD] and rapidly progressive dementias/atypically rapid variants of common ND, RapidND). RESULTS: A total of 498 participants were included: 197 ND, 67 PSY, 161 CJD, 48 RapidND, and 20 controls. NfL was elevated in ND compared to PSY and controls, with highest levels in CJD and RapidND. NfL distinguished ND from PSY with 95%/78% positive/negative predictive value, 92%/87% sensitivity/specificity, 91% accuracy. NfL outperformed t-tau in most real-life clinical diagnostic dilemma scenarios, except distinguishing CJD from RapidND. DISCUSSION: We demonstrated strong generalizable evidence for the diagnostic utility of CSF NfL in differentiating ND from psychiatric disorders, with high accuracy.