Premium
Could cerebrospinal fluid neurofilament light chain reduce misdiagnosis in neurodegenerative and neuropsychiatric disorders in a real‐world setting? A retrospective clinical and diagnostic utility study
Author(s) -
Kang Matthew,
Dobson Hannah,
Li QiaoXin,
Keem Michael,
Loi Samantha M,
Masters Colin L,
Collins Steven,
Velakoulis Dennis,
Eratne Dhamidhu
Publication year - 2021
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.058721
Subject(s) - neuropsychiatry , medical diagnosis , medicine , retrospective cohort study , dementia , pediatrics , diagnostic accuracy , psychiatry , disease , pathology
Background Patients presenting with neuropsychiatric symptoms often face significant diagnostic odyssey. Our recent research (Eratne et al, ANZJP, 2020) found neurofilament light (NfL) differentiated between neurodegenerative and psychiatric disorders, with high accuracy. Yet the clinical utility of NfL, as to whether it can aid clinicians in avoiding misdiagnosis in a real‐world clinical setting, is unknown. Our primary aim was to measure the rates of diagnostic change in patients with neuropsychiatric symptoms assessed at a tertiary multidisciplinary service, and determine whether baseline cerebrospinal (CSF) NfL level could have prevented misdiagnoses, by predicting the final diagnosis after follow up. Methods We conducted a retrospective file review of patients assessed at an Australian neuropsychiatry and young‐onset dementia service between 2009‐2020. NfL levels were measured from CSF collected at their baseline assessment. Blinded investigators (MK, HD, DE) extracted clinical data including diagnoses from discharge summaries and outpatient letters from the initial assessment and re‐assessment. Baseline and final diagnoses were categorised as neurodegenerative disorder [ND], or, other non‐neurodegenerative conditions including primary psychiatric disorder [Other/PPD]. We also obtained follow‐up information on patients that were subsequently seen at external services where available. Results From a preliminary analysis of those with follow‐up information for at least a year (N=32), six patients’ diagnostic categories (19%) were revised (ND to Other/PPD=5; Other/PPD to ND=1). In all six cases (figure 2), baseline CSF NfL levels, using our previously established cut‐off, would have predicted the final revised diagnosis. As this study is underway, findings for over 200 patients will be presented for the Conference. Conclusions In a real‐world tertiary clinical setting, baseline CSF NfL would have accurately predicted diagnostic change, showing promise to aid clinicians assessing patients with neuropsychiatric symptoms, and reduce misdiagnosis. An elevated level could help exclude primary psychiatric provisional or differential diagnoses, and prompt assertive investigations for neurological and neurodegenerative causes. Conversely, a low NfL, could reassure against a neurodegenerative disorder, preventing unnecessary assessments. Timely and accurate diagnosis will reduce uncertainty, enable early care planning, reduce patient and carer burden, thus improving outcomes and the diagnostic odyssey faced by patients and families.