Diagnostic Performance Of CSF Protein 14-3-3 For Prion Disease In The Modern Era
Anas Elgenidi1, Yoav Piura1, Nihal Satyadev1, Dror Shir1, Jody G. Lavrich2, Tracy Haldiman2, Xiaoqin Liu2, Christian Lachner1, Neill Graff-Radford1, Brian Appleby2, Gregory Day1
1Neurology, Mayo Clinic in Florida, 2National Prion Disease Pathology Surveillance Center, Case Western Reserve University School of Medicine
Objective:
To evaluate the real-world diagnostic utility of cerebrospinal fluid (CSF) protein 14-3-3 for prion disease in the modern diagnostic era.
Background:
CSF protein 14-3-3 has been incorporated into diagnostic criteria for Creutzfeldt–Jakob disease for decades. Advances in MRI and RT-QuIC enable earlier detection of prion disease, when neuronal injury biomarkers such as 14-3-3 may be below detection thresholds. To improve sensitivity, the US National Prion Disease Pathology Surveillance Center (NPDPSC) lowered the positivity threshold from ≥4,000 to ≥2,000 AU/mL in May 2022. The impact of this change on diagnostic performance remains uncertain.
Design/Methods:
We analyzed CSF biomarker data from 629 participants evaluated at academic centers (Washington University in St. Louis and Mayo Clinic; 2016–2025) for prion disease (definite, n=9; probable, n=37), non-prion rapidly progressive dementia (RPD; n=147), and typically progressive dementia (n=436). Biomarkers were measured using standardized NPDPSC protocols, including RT-QuIC and ELISA assays of 14-3-3 and total tau. Brain MRIs were assessed for prion-associated changes. Diagnostic performance was summarized using sensitivity, specificity, accuracy, and likelihood ratios. Survival was assessed using Cox models.
Results:
At the revised threshold (≥2,000 AU/mL), 14-3-3 showed high sensitivity (100%) but very low specificity (≤6.1%) and low accuracy (≤28.5%), with positive likelihood ratios approximating 1. Elevated 14-3-3 levels were common in non-prion RPD (93.9%) and typically progressive dementias (95.6%). In contrast, total tau (≥1,150 pg/mL), RT-QuIC, and MRI demonstrated high diagnostic accuracy. Higher 14-3-3 concentrations were modestly associated with shorter survival (HR 1.02 per 1,000 AU/mL), comparable to total tau but weaker than RT-QuIC and MRI. Using the prior cutoff (≥4,000 AU/mL), 14-3-3 showed limited specificity and accuracy; an optimized threshold (10,400 AU/mL) improved accuracy but remained inferior to other markers.
Conclusions:
Despite its historically outsized role, CSF protein 14-3-3 provides no meaningful incremental diagnostic value and may no longer be justified as a diagnostic test for prion disease.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.