Relationships among cerebrospinal fluid biomarkers in dementia of the Alzheimer type.”
Csernansky JG, Miller JP, McKeel D and Morris JC;
Commented by , 25 Nov 2002
Aim of the study
Establish if levels of amyloid and tau- related proteins in the cerebrospinal fluid (CSF) are altered in Alzheimer’s disease (AD), and potentially useful as biological makers of disease severity.
Method
Levels of proteins related to the pathophysiology of AD (soluble amyloid precursor protein (sAPP), Abeta total, Abeta 1-42, tau, apolipoprotein E (ApoE)) were measured using enzyme-linked immunosorbent assays in the CSF obtained from lumbar punctures performed on subjects with probable or definite AD (N: 33), and cognitively healthy subjects (N: 11).
AD severity was measured using the sum-of-boxes from the Clinical Dementia Rating Scale. Student t tests were used to compare CSF concentrations of the various proteins between groups of subjects, followed by logistic regression and receiver operating curves. Relationship between CSF levels of proteins and disease severity was explored using Pearson product-moment correlations.
Results
CSF tau levels were higher in subjects with AD relative to controls (p<0.006), but did not correlate with CDR sum-of-boxes (p= 0.67), suggesting that elevations of tau concentration were found regardless of disease severity.
Although the lower levels of Abeta total and sAPP in AD subjects relative to controls did not reach statistical significance, there were inverse correlations of these protein levels with CDR sum-of-boxes (p = 0.06 and p = 0.02 respectively). Using logistic regression the ratio of tau to Abeta 1-42 gave the best discrimination between subjects with AD and controls (likelihood ratio chi-square = 18.1; df = 1; p< 0.0001).
These results confirm that CSF tau levels are abnormally high in subjects with AD but do not change further with increasing disease severity. The levels of Abeta total are lower in early stages of AD and decrease further with disease progression. Previously reported changes in the levels of Abeta 1-42 as well as sAPP and ApoE could not be found.
Discussion
There have been many attempts at measuring selected proteins in CSF from subjects with AD in order to establish if specific changes could be used towards early diagnosis (possibly pre-clinical or in mild cognitive impairment stages) or towards disease-modification clinical trials.
The ratio of tau to Abeta 1-42 is the most robust finding so far to allow discrimination between AD subjects and controls, but the authors do not think that this biomarker is specific and sensitive enough to be used to establish the very early diagnosis of AD in individual cases.
Nevertheless, tau and Abeta 1-42 levels in CSF remain useful research tools in ongoing phase II short-term randomized clinical trials aiming at modification of amyloid aggregation, processing or clearance, as indicators of dose-related biological effects, with the hope that a positive biological signal in phase II would correlate with disease-stabilization in long-term phase III studies.