CSF detection of the 14-3-3 protein in unselected patients with dementia

OBJECTIVE: To determine the usefulness of the 14-3-3 test in patients with dementia of various causes. BACKGROUND: Recent reports have suggested that the detection of the 14-3-3 protein in the CSF of patients with Creutzfeldt--Jakob disease is a highly sensitive and specific marker of the disease that might be used as a diagnostic criterion. We examined the validity of this test when applied to a cohort of unselected patients prospectively examined for an ongoing dementing process. METHODS: One hundred patients underwent an extensive neurologic examination for dementia, including a CSF 14-3-3 protein immunoblotting assay. Final clinical diagnoses were compared with the qualitative results of the test, and statistical measures of test validity were carried out. RESULTS: We found a positive test in 14 of 100 patients, only two of whom had definite Creutzfeldt--Jakob disease. Positive results were found in patients with various degenerative dementias, including AD (4), frontotemporal dementia (2), and dementia with Lewy body (1), and in patients with vascular dementia (1), carcinomatous meningitis (1), and anoxic encephalopathy (1). In two other positive patients, the dementia could not be confidently classified. Sensitivity, specificity, and negative predictive value were fairly good, but positive predictive value was poor. Similar results were found independently of the disease duration. There was no correlation between intensity nor pattern of the 14-3-3 protein expression and diagnosis. CONCLUSIONS: The 14-3-3 test is not valid for discriminating between Creutzfeldt--Jakob disease and non-Creutzfeldt--Jakob disease in unselected patients with dementia. Positive results are found in various degenerative and secondary, prion-unrelated dementias.