Biomarkers in the CSF and Blood as Predictors of Alzheimer's Dementia — Are They Ready for Clinical Use?

Abstract & Commentary

By Matthew E. Fink, MD, Vice Chairman, Professor of Clinical Neurology, Weill Medical College, Chief of Division of Stroke and Critical Care Neurology, NewYork-Presbyterian Hospital. Mr. Fink reports no consultant, stockholder, speaker's bureau, research, or other relationships related to this field of study.

Synopsis: A variety of peptide markers in the blood and CSF have been identified by proteomic techniques and should increase our ability to identify, at an early stage, those patients at risk for developing Alzheimer's dementia (AD).

Sources: Finehout EJ, et al. Cerebrospinal Fluid Proteomic Biomarkers for Alzheimer's Disease. Ann Neurol. 2007; 61:120-129.Fagan AM, et al. Cerebrospinal Fluid tau/ -Amyloid42 Ratio as a Prediction of Cognitive Decline in Non demented Older Adults. Arch Neurol 2007; 64:343-349. Graff-Radford NR, et al. Association of Low Plasma A 42/A 40 Ratios With Increased Imminent Risk for Mild Cognitive Impairment and Alzheimer Disease. Arch Neurol. 2007;64:354-362. Simonsen AH, et al. Novel Panel of Cerebrospinal Fluid Biomarkers for the Prediction of Progression to Alzheimer Dementia in Patients With Mild Cognitive Impairment. Arch Neurol. 2007;64:366-370.

In recent months, a series of studies using advanced techniques of proteomics have shed further light on potential biomarkers that will help us identify patients at risk for AD. Since AD is the most common cause of dementia, and its prevalence doubles every 5 years from the age of 65, it will be extremely important to be able to identify such patients before they are symptomatic or at a very early stage in order to intervene with new therapies that prevent progression. Minimal cognitive impairment (MCI) has a conversion rate of 8% to 15% per year, but we are currently unable to identify which of those patients will be in the progressive group. Previous biomarker studies have documented the high sensitivity of CSF measurements of β-Amyloid42 , total tau, and phosphorylated tau in patients with MCI and AD, but these tests do not discriminate between MCI, early AD, or patients who will progress. Therefore, the recent studies that I will briefly review are a welcome addition to our current diagnostic capabilities.

The work by Finehour et al studied an antemortem cohort of CSF proteins from 34 AD and 34 non-AD patients and separated the samples using 2-dimensional gel electrophoresis. The protein patterns were analyzed with a multivariate statistical analysis. The gels revealed an average of 1,188 spots, and a Student's t-test analysis comparing spots from AD and non-AD gels identified 252 spots with a significant change in expression level. Further analysis identified 23 spots that could be used to differentiate AD and non-AD gels with a sensitivity of 94%, and a specificity of 94%, and a predicted classification error rate of only 5.9%. The proteins in the 23 spots were identified and fell into 4 classes of compounds related to (1) transport of β-amyloid, (2) the inflammatory response, (3) proteolytic inhibition, and (4) the neuronal membrane proteins.

The work by Fagan et al looked at CSF β-amyloid40 , Aβ42 , tau, and phospholylated tau, as well as plasma Aβ40 and Aβ42 , with longitudinal follow-up studies from 1 to 8 years. They studied a group of 139 community-dwelling volunteers aged 60-91 who were judged as cognitively normal, or having very mild AD, and showed that the very mild AD patients had reduced levels of CSF Aβ42 and increased levels of CSF tau and phosphorylated tau. The levels of Aβ42 corresponded with the presence or absence of brain amyloid (imaged with Pittsburgh Compound B) in both demented and non-demented individuals. The ratio of CSF tau/Aβ42 (hazard ratio = 5.21; 95% CI, 1.58-17.22) predicted the conversion from normal, to demented over the follow-up period. The investigators concluded that even the very mildest degree of cognitive impairment exhibits the same CSF biomarker phenotype as more advanced AD and this corresponds to PET imaging. In addition, the ratio of tau/Aβ shows promise as a marker to identify normal older adults who will go on to develop dementia.

The third study by Graff-Radford et al looked at plasma measurements and ratios of Aβ40 and Aβ42 to see if they could identify elderly adults at increased risk for mild cognitive impairment and AD. They measured plasma Aβ40 and Aβ42 in 563 cognitively normal volunteers (median age, 78; 62% female) and followed them for 2 to 12 years (median, 3.7 years) to detect the incident cases of MCI and AD. During follow-up, 53 subjects developed MCI or AD. Subjects with Aβ42/Aβ40 ratios in the lower quartiles showed a significantly greater risk of developing MCI or AD. Comparing those patients with plasma ratios from the lowest to highest quartile gave a relative risk of 3.1(95% CI, 1.1-8.3) for the development of MCI or AD. The authors concluded that measurements of Aβ40 and Aβ42 ratios may be useful as a predictor of future development of dementia.

And finally, the study authored by Simonsen et al analyzed CSF proteins and peptides with a novel technique known as surface-enhanced laser desorption/ionization time-of-flight mass spectroscopy (SEL-TOF-MS) in 113 patients with MCI, 56 who remained stable and 57 who progressed to AD over a 4 to 6 year follow-up period. They also studied 28 healthy controls. The investigators identified 17 potential biomarkers that could distinguish between patients with stable MCI and patients with MCI that progressed to AD. Five of the biomarkers are relevant to the pathogenesis of AD and are involved in the metabolic pathways of Aβ and tau proteins.


At a time when we are developing clinical trials for prevention or slowing of progression of AD, it is critical that we have biomarkers that can identify high-risk groups of elderly normals or patients with MCI who will go on to develop AD. The above investigations have gone a long way to accomplish this goal, by demonstrating the ability to use proteomic techniques to identify an array of proteins in the CSF that can distinguish progressive AD from stable MCI syndromes. In addition, reliable measurements of Aβ in the plasma may also provide supportive evidence for high-risk disease, along with CSF studies.

However, these analyses are still in the realm of investigative tests that are difficult to perform outside of research laboratories, and they are not currently applicable for routine clinical testing. Until we have a proven treatment for AD that slows or halts progression, there is no clinical utility for these tests outside of research studies. To facilitate the completion of clinical trials of new therapeutic agents, we would recommend that all neurologists refer appropriate patients to centers where they can be enrolled in clinical trials, thereby speeding up the time it takes to get answers regarding possible treatments of this epidemic disease.