Even as Alzheimer’s disease biomarkers are wending their way through a lengthy standardization and qualification process, individual labs are experimenting with deploying them toward the next challenge, that is, to improve the differential diagnoses of related dementias. In the November Archives of Neurology, scientists led by Oskar Hansson, Skåne University Hospital, Malmo, Sweden, suggest that adding measures of cerebrospinal fluid (CSF) α-synuclein and neurofilament light chain to established markers of AD can help clinicians distinguish among disorders on the dementia-parkinsonism spectrum. In the October 17 Journal of Alzheimer’s Disease, members of that same team, this time led by Kaj Blennow of the University of Gothenburg in Molndal, Sweden, propose a technique that uses mass spectrometry to skirt the matrix effect that confounds current methods of estimating Aβ. Both papers highlight the rising importance of fluid biomarkers, which researchers hope will offer a means of diagnosing and tracking neurodegenerative disease progression.

A Multiple Biomarker Assay
While CSF Aβ42, phosphorylated tau, and total tau are different between people with Alzheimer’s disease (AD) and controls, those markers alone do not distinguish AD from other disorders with similar pathology. For instance, dementia with Lewy bodies (DLB) and Parkinson’s disease with dementia (PDD) tend to come with amyloid plaques, though the distribution, time course, and overall pathology differ from AD. To test if additional markers could sharpen differential diagnoses, Hansson and colleagues examined CSF from 453 people with various neurodegenerative disorders. They found that taking into account the concentrations of α-synuclein and neurofilament light chain—a component of the neuronal cytoskeleton—might make for a better diagnostic. Researchers find α-synuclein pathology in about half of AD patients, where it appears as a component of amyloid plaques. Previous research also reported more soluble α-synuclein in AD brain tissue (see ARF related news story). In addition, neurofilament light chain has been shown to be a marker of neuronal injury (see ARF related news story).

The researchers used a novel bead-based immunoassay to test four biomarkers simultaneously—Aβ42, total tau, phosphorylated tau, and α-synuclein. They measured neurofilament light chain separately by ELISA. Not only were CSF tau levels higher and Aβ42 levels lower in AD patients relative to those who had PDD or DLB, but AD patients had higher CSF α-synuclein levels as well, possibly because it leaked from degenerating neurons, the authors wrote. These combined biomarkers separated disease categories with 90 percent sensitivity and 81 percent specificity.

Because clinical manifestations and pathologies overlap among these dementias, it is hard to tell them apart based solely on clinical tests or limited biomarker panels. Correct diagnosis is important now, both to give a prognosis and to avoid medication errors. For instance, people with DLB are prone to hallucinations and should avoid certain antipsychotics prescribed for AD. Differential diagnosis will be critical in the future. “When we have disease-modifying treatments that target individual pathophysiologies, it will be important to correctly diagnose patients early on,” said Annika Öhrfelt, University of Gothenburg, co-first author with Sara Hall, Lund University, Sweden.

The research team further found that higher levels of CSF neurofilament light chain distinguished people with atypical parkinsonian disorders such as multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and corticobasal degeneration (CBD) from those who had idiopathic Parkinson’s disease (PD). Neurofilament light chain could reflect more severe neurodegeneration inflicted by these atypical disorders, suggested the authors. The protein may eventually make a good biomarker to exclude PD, something that the field “urgently” seeks, said Brit Mollenhauer, Paracelsus-Elena-Klinik, Kassel, Germany. “Even movement disorder experts incorrectly diagnose 20 percent of patients at the earliest stages of Parkinson’s,” she told Alzforum. A solid biomarker would help diminish that margin of error, Mollenhauer said, though she agreed that neurofilament light chain itself needs more study and validation.

The particular four-biomarker test used in this study is no longer available from Innogenetics in Gent, Belgium, which developed the assay and was recently sold to the Japanese diagnostics company Fujirebio. Even so, this study provides proof of concept that may lead to the development of other assays that look at these biomarkers simultaneously, said Öhrfelt.

Antibody-Free Aβ Detection
The second paper by Blennow and colleagues proposes a way to measure CSF Aβ that avoids matrix effects. This is where concentration-dependent binding of the peptide to different proteins in the CSF compromises analysis, as sequential dilution does not proportionally free up Aβ (see ARF related news story). The new method, involving mass spectrometry, does not depend on antibody binding. It measures Aβ directly, even in the presence of other proteins, and can distinguish among Aβ38, Aβ40, and Aβ42.

First author Josef Pannee, University of Gothenburg, and colleagues froze CSF samples from two separate sets of 15 AD patients and 15 healthy controls. Just before analysis, the team thawed the specimens and denatured the proteins with guanidine hydrochloride. The researchers then added known quantities of “heavy” standards for Aβ38, Aβ40, and Aβ42, all of which incorporated a heavier isotope of nitrogen, as internal controls before mass spectrometry analysis.

The method detected about twice the Aβ42 as an ELISA done in parallel, and it more accurately identified AD patients and controls based on the Aβ42/Aβ40 ratio. However, the method is laborious and time-consuming, meaning high-throughput use on a par with ELISA is not yet feasible, Pannee told Alzforum. “The next generation of mass spectrometers coming out now are much faster and have much better sensitivity,” said Pannee. He plans to use this method to detect other Aβ peptides, and expand its use to other analytes. However, this method will not distinguish different oligomeric forms of Aβ, which are widely believed to be the most toxic species. Organizations, including the Institute for Reference Materials and Measurements and the International Federation of Clinical Chemistry, are taking steps to standardize CSF assays of AD biomarkers (see ARF related news story).

“We have a golden opportunity to once and for all measure CSF Aβ directly, free from matrix effects and with adequate sensitivity,” wrote Les Shaw, University of Pennsylvania Medical Center, Philadelphia, to Alzforum in an e-mail. Shaw agreed that these methods are not yet ready to be widely used in the clinic, but suggested that if companies decide to invest in this technology, more efficient mass spectrometry methods could be on the horizon (see full comment below).—Gwyneth Dickey Zakaib


  1. This paper describes development of a direct selected reaction monitoring (SRM) mass spectrometric method for measuring Aβ42, Aβ40, and Aβ38. This technique avoids using antibodies to capture Aβ peptides and their possible bias due to affinity and epitope binding properties. The authors chose cerebrospinal fluid (CSF) as the calibrator matrix and a guanidine hydrochloride sample treatment protocol followed by a mixed-bed ion exchange “cleanup” step, as recently described (Lame et al., 2011). Keep in mind that the majority of proteomic mass spectrometry analyses have to use proteolytic digestion of the sample prior to the SRM mass spectrometry. However, since Aβ peptides have mass-to-charge ratios that are within the limits of mass spectrometers, the analysis can avoid the proteolytic digestions. Thus, we have a golden opportunity to once and for all measure Aβ in the CSF directly, free from matrix effects, and with adequate sensitivity, as demonstrated in this paper. And, as the authors note, there is a joint project among four labs that is well underway to develop both a reference method and reference material for Aβ42 measurements. We can enjoy the data presented in this paper, and even more to come from participants in that group.

    This method improves on other CSF Aβ measures by its freedom from matrix effects, its ability to determine total Aβ42 as compared to the free form that immunoassays measure, and its accuracy-based approach. However, fully qualifying an analytical method for use in research studies and treatment trials requires lots of repetitions to ensure calibrator linearity, quality control reproducibility, and sufficient comparisons between samples from AD patients and various control groups. In addition, from the perspective of one whose lab does lots of mass spectrometry method development and implementation, there is a lot of room for further assessments of what columns work best and how to sustain stable sensitivity of the mass spectrometer. It usually takes thousands of injections of sample preparations to nail this down.

    No question—this is a labor-intensive procedure. We have been working with our version of this type of approach for approximately one and a half years in our lab at UPenn. I am sure that the mass spectrometry companies will become engaged in this field if they see a market. Then they will put more resources into it, and with efforts from investigators in the field, that could lead to improvements in the efficiency of this type of methodology. As it stands now, based on my experience in both the research lab and the clinical lab arena, this will be primarily a research tool, useful for tracking Aβ42 and other Aβ peptides in treatment trials. Immunoassays are likely the mainstay methodology for the foreseeable future in this field as far as the clinical lab is concerned. That we don’t have candidate mass spectrometry methods for other important CSF biomarkers such as tau, α-synuclein, and TDP-43, etc., is all the more reason for the predominant use of immunoassays for the CSF AD biomarkers.

    I agree with the authors that follow-up to this pilot study is needed. Larger numbers of study subjects, better calibrations, additional testing in patient cohorts, and inter-laboratory study (which appears well along in the planning stages), are all needed. This is a very demanding type of methodology that does require a high level of skilled personnel.

    It is interesting to see, albeit on a very small number of study subjects, some improvement in the separation between AD patients and controls using the Aβ42/Aβ40 ratio. This will need to be confirmed by the authors and others, but if this can be replicated, it could provide somewhat better—more specific—AD detection. Also, and importantly, these pilot data do lead to the preliminary impression that the immunoassay (ELISA) provides comparable receiver operator characteristics and area under the curves for the AD versus normal controls, and if replicated that would provide important support for the clinical utility of Aβ measurement using currently available research immunoassays.


    . Quantitation of amyloid beta peptides Aβ(1-38), Aβ(1-40), and Aβ(1-42) in human cerebrospinal fluid by ultra-performance liquid chromatography-tandem mass spectrometry. Anal Biochem. 2011 Dec 15;419(2):133-9. PubMed.

    View all comments by Leslie Shaw
  2. This new report from Kaj Blennow and Oskar Hansson's group appears to support our own findings that soluble α-synuclein levels are increased in brains of subjects with AD (Larson et al., 2012). Whether this elevation corresponds to a "leakage of α-synuclein from degenerating neurons" or to an influx of α-synuclein in the CSF from the periphery remains to be demonstrated. We would favor an alternate hypothesis, i.e., that brain expression of α-synuclein is profoundly increased in AD, resulting in enhanced secretion of the protein into the interstitial fluid (ISF). Another key question relates to the nature of the soluble species of α-synuclein accumulating in AD, whether it is in the ISF, CSF, or neuronal cells. More work is needed to address this question at this time. Overall, this new study and ours highlight the possible involvement of soluble α-synuclein as an important modulator of AD pathophysiology.


    . Soluble α-synuclein is a novel modulator of Alzheimer's disease pathophysiology. J Neurosci. 2012 Jul 25;32(30):10253-66. PubMed. Correction.

    View all comments by Sylvain Lesne

Make a Comment

To make a comment you must login or register.


News Citations

  1. Is α-Synuclein the Third Musketeer in AD Pathophysiology?
  2. Keystone: Diagnosis and Model Treatments for Traumatic Brain Injury
  3. CSF Markers: Goodbye, Research Use Only; Hello, Clinical
  4. Metrology, Certification Heavies Take CSF Tests Under Their Wings

Further Reading


  1. . Systematic review of the body of evidence for the use of biomarkers in the diagnosis of dementia. Alzheimers Dement. 2012 Oct 27; PubMed.
  2. . Combination of 18F-FDG PET and cerebrospinal fluid biomarkers as a better predictor of the progression to Alzheimer's disease in mild cognitive impairment patients. J Alzheimers Dis. 2013;33(4):929-39. PubMed.
  3. . Novel One-step Immunoassays to Quantify α-Synuclein: APPLICATIONS FOR BIOMARKER DEVELOPMENT AND HIGH-THROUGHPUT SCREENING. J Biol Chem. 2012 Sep 28;287(40):33691-705. PubMed.
  4. . Phosphorylated α-synuclein as a potential biomarker for Parkinson's disease and related disorders. Expert Rev Mol Diagn. 2012 Mar;12(2):115-7. PubMed.
  5. . Identification of biomarkers in Lewy-body disorders. Q J Nucl Med Mol Imaging. 2012 Feb;56(1):39-54. PubMed.
  6. . CSF α-synuclein does not differentiate between parkinsonian disorders. Neurobiol Aging. 2012 Feb;33(2):430.e1-3. PubMed.

Primary Papers

  1. . Accuracy of a panel of 5 cerebrospinal fluid biomarkers in the differential diagnosis of patients with dementia and/or parkinsonian disorders. Arch Neurol. 2012 Nov;69(11):1445-52. PubMed.
  2. . A selected reaction monitoring (SRM)-based method for absolute quantification of Aβ38, Aβ40, and Aβ42 in cerebrospinal fluid of Alzheimer's disease patients and healthy controls. J Alzheimers Dis. 2013;33(4):1021-32. PubMed.