Biomarkers and Alzheimer spectrum


ALZHEIMER'S DISEASE (AD) is a complex disorder with pathological hallmarks of senile plaques, neurofibrillary tangles, and neuronal loss. Amyloid beta peptide is aggregated into amyloid fibril and it is deposited in the core of senile plaques, diffuse plaques and in cerebral blood vessels of amyloid angiopathy. Hyper-phosphorylated tau protein is transformed to paired helical filaments of neurofibrillary tangles inside and outside of neuronal cell body and dystrophic neurites. The precise mechanism of neuronal loss in the brain with AD is yet to be clarified, but it is speculated that various pathological processes are involved, such as oxidative stress, inflammation, intracellular signal transduction aberration and others.

Drugs for AD are now widely used to treat AD patients, including acetylcholine esterase inhibitors (donepezil, galantamine, rivastigmine), and NMDA receptor antagonist (memantine). These drugs are useful to AD patients: improving cognitive function,1 reducing symptoms,2 improving the activities of daily living and quality of life of the patients,3 reducing the burden of caregivers,4 and delaying institutional care,5 even though there is a significant number of AD patients who do not respond to these drugs.6 However, they are symptomatic drugs, not modifying the de novo pathological course of AD. In other words, the pathological process of AD is still ongoing under the treatment of these drugs.

Based on the understanding of the pathological process of AD, disease-modifying drugs have been under development, including γ-secretase inhibitors, γ-secretase modifiers, β-secretase inhibitors, amyloid vaccines, inhibitors of amyloid aggregation, inhibitors of tau phosphorylation, inhibitors of tau aggregation, and many others. The development of disease-modifying drugs is generating the possibility of early intervention to AD pathology, even before the development of the full-blown clinical symptoms of AD.

In clinical settings and for research purposes, diagnostic criteria for AD are in use. Among them, the frequently used criteria are the Diagnostic and Statistical Manual of Mental Disorders by the American Psychiatric Association (DSM-IV), the International Classification of Diseases (ICD-10) by the World Health Organization and the National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria. As all of these diagnostic criteria are based on clinical signs and symptoms, it is theoretically impossible to apply these criteria to subjects in the earlier stage of AD.

In recent years, the earlier stage of AD has attracted more attention because earlier intervention is believed to be beneficial for a better outcome for possible AD patients. The concepts of mild cognitive impairment (MCI), or age-related cognitive decline (ARCD) are proposed to capture the subjects in the earlier stage of AD, in which subjects are showing cognitive impairment compared with the age-matched population but not showing clinical symptoms of AD. Early detection and intervention to subjects with MCI or ARCD is expected to prevent transition into AD. Prevention of AD is believed to be the possible target of pharmacological intervention by the disease-modifying drugs of AD.

In clinical trials of drugs for AD, two pivotal clinical studies are usually required for administrative approval. Any candidate drug will be evaluated by the following two outcomes: improvement in cognitive functions measured by the Alzheimer's Disease Assessment Scale-Cognitive Subscale and other measures of cognitive evaluation and improvement in clinical symptoms evaluated with the Clinician's Interview-Based Impression of Change (CIBIC) and other measures.

However, cognitive functions and clinical symptoms of AD are influenced by various factors of the patients, including genetic factors,7–10 aging,11 hormones,12–14 comorbidities,15 lifestyle,16 and environmental factors, including education.17,18 There is a certain gap between pathological features of AD and clinical symptoms, but the use of biomarkers is expected to be useful in clinical trials as the surrogate end-point of the trial.19


Biomarkers of AD are parameters of the biological response caused by AD pathology, which is useful to evaluate the natural course and effect of intervention to AD. Biomarkers of AD should satisfy the following requirements: (i) it can be measured objectively; (ii) it should be specific and sensitive to AD; (iii) it can be sensitive enough to reflect the progress of the natural course of AD; and (iv) it should be able to predict the effectiveness of the intervention.

In reality, there is no single biomarker that can fulfill all of the requirements above, and different biomarkers will be sought for different purposes, such as those for screening, differential diagnosis, clinical evaluation of the natural course and evaluation of the intervention. In clinical settings, brain-imaging techniques, genetic screening, and biochemical measures are widely used for biomarkers of AD.

1. Imaging markers

AD patients are evaluated by structural and functional brain imaging. Magnetic resonance imaging (MRI) is used to evaluate the structural change of the brain. The unique atrophic pattern is documented in vulnerable areas of AD brain, such as the parahippocampal area, hippocampus, amygdale, posterior association area, and basal ganglia. The gray matter volume in these areas is shown to be decreased along with the progress of AD pathology. For quantitative evaluation of the grey matter atrophy around the parahippocampal area, the voxel-based specific regional analysis system for Alzheimer's disease (VSRAD) is used as an aid for early diagnosis of AD.20,21 The brain atrophy in certain areas reflects the neuronal loss in those areas.22

Magnetic resonance spectroscopy (MRS) complements MRI as a non-invasive means for the characterization of the brain tissue component. While MRI uses the signal from hydrogen protons to form anatomic images, proton MRS uses this information to determine the concentration of brain metabolites, such as N-acetyl aspartate (NAA), choline (Cho), creatine (Cr) and lactate in the tissue examined. MRS has its limitations and is not always specific but, with good technique and in combination with clinical information and conventional MRI, can be very helpful in diagnosing central nervous system disorders, including AD.

Single photon emission computed tomography (SPECT) is used to evaluate cerebral blood flow with 99m-tecnethium hexamethyl propylene oxim (99mTc-HMPAO) and other tracers. 99mTc transported into the brain through blood flow is absorbed in the brain tissue and is measured as the index of brain blood flow in the particular area.23 Meta-analysis has shown that the sensitivity of 99mTc-HMPAO SPECT for AD diagnosis is 74%, which is lower than that of cognitive measures. However, due to the introduction of quantitative analysis of SPECT data, the sensitivity has been improved to 88% in recent studies. SPECT is a relatively inexpensive and versatile tool for brain blood flow because it uses traces with the longer half-life than those for PET study.

Functional MRI (fMRI) measures the responsive blood flow to the assigned task in the particular area of the brain. The blood oxygen-dependent level (BOLD) signal of fMRI is regarded as reflecting the neuronal activity of the brain area studied.24

PET study is further classified into two categories: activation study and ligand study. In activation study, 18F-2-fluoro-2-deoxy-D-glucose (FDG) is used as the tracer reflecting glucose utilization of the particular area as the indicator of brain metabolism, which can be used for AD diagnosis from other neurodegenerative dementia.25 In addition to the tracers of glucose metabolism, there are various kinds of isotope-labeled ligands specifically binding to neurotransmitter receptors and transporters in dopamine, serotonin, benzodiazepine, GABA and other transmitter systems.

In recent years, several ligands specific to amyloid beta peptide (Aβ) have been developed, which reflect the level of Aβ accumulation in the brain.2611C-labeled ligand, Pittsburg compound B (11C-PIB) specifically binds to amyloid deposition in senile plaques, diffuse plaques and amyloid angiopathy in the affected blood vessels, but it does not bind to soluble or oligomeric Aβ.27 A high signal of PIB PET is regarded as reflecting the amyloid burden in the brain of the subjects. MCI subjects with higher PIB signals are reported to convert to AD development.28 However, it is also true that some cognitively normal elderly subjects show high signals by PIB PET imaging.29 The 18F-labled ligand for amyloid, which has a longer half-life of radioactivity than 11C-PIB, has recently been developed and is under clinical trial for its usefulness in a clinical setting.30

2. Genetic markers

About 10–20% of AD is familial, caused by mutations of amyloid precursor protein (APP), presenilin-1 (PSEN1) or presenilin-2 (PREN2) gene. Apolipoprotein E (APOE) is the most powerful risk gene for AD and age-related cognitive dysfunction. A carrier of one apoE4 allele has a six-times higher risk for AD and a carrier of two apoE4 alleles has a 20-times higher risk than a non-carrier of apoE4. A carrier of ApoE4 shows a generally worse profile in imaging markers31 as well as in biochemical marker profiles.32

Recent genome-wide association studies (GWAS) for AD have revealed new risk genes of AD: clusterin (CLU, also called apolipoprotein J), complement receptor1 (CR1), and phosphatidylinositol-binding clathrin assembly protein (PICALM).33,34 These proteins are possible biochemical markers in the brain, cerebrospinal fluid (CSF) and serum, indicating the involvement of compliment cascade in AD pathology. There are several papers reporting new risk genes for AD, such as MTHFD1 (methylenetetrahydrofolate dehydrogenase [NADP+ dependent] 1-like),35 SORL1 (sortilin-related receptor 1),36 dynamin 237 and PPP2R2B.38 These new findings may indicate possible biomarkers in the process of axonal transport, trafficking, endocytosis, apoptosis, intracellular signal transduction, and other relevant processes in which risk genes are involved.

3. Biochemical markers

CSF biomarkers reflecting the amyloid cascade (Aβ40, Aβ42) and cytoskeletal degeneration (total tau, and phosphorylated-tau [p-tau]) have been studied extensively and these biochemical markers have been established as reliable biomarkers for AD pathology. Changes in Aβ40 and Aβ42 levels in CSF of AD patients correlate well with amyloid burden in the brain, and the higher level of p-tau in CSF indicates more neurofibrillary tangles in the brain. The level of total tau may reflect the axonal injury and degeneration due to various pathological conditions, and an increase in total tau may not be specific to AD. Increase in total tau and p-tau, and decrease in Aβ42 level and Aβ42/Aβ40 ratio have been documented in CSF from AD patients and from MCI subjects. The usefulness of these four parameters has been confirmed with large prospective studies, including the Alzheimer's Disease Neuroimaging Initiative (ADNI) project,41 the DESCRIPA study, and the Swedish Brain Power Project.

CSF Aβ42

CSF Aβ42 is decreased with AD and MCI patients. When CSF Aβ42 measured with the ELISA kit developed by Innogenetics is compared between AD patients (n = 600) and healthy control subjects (n = 450), the CSF Aβ42 level shows a sensitivity of 80% and a specificity of 90%.39 Decrease in CSF Aβ42 is observed in other pathological conditions, such as diffuse Lewy body disease, amyotrophic lateral sclerosis, multiple systemic atrophy, and Creutzfeldt–Jakob disease. The data of CSF Aβ42 level from a multicenter study with AD patients (n = 150), healthy controls subjects (n = 150), and patients with other dementia (n = 79) show 81% specificity against healthy control groups but only 59% against other dementia.40 From the ADNI project, CSF Aβ42 level was reported as 205.6 ± 55.1 pg/mL for the healthy group, 162.8 ± 56.0 pg/mL for the MCI group, and 143.0 ± 40.8 pg/ml for the AD group, with significant difference among these three groups.41 MCI subjects with lower CSF Aβ42 were shown to convert to AD in 12 months.


CSF total tau is increased with AD patients, but it can be the marker of axonal injury or degeneration due to various diseases and conditions. In fact, higher CSF total tau has been demonstrated in many diseases, including neurodegenerative disorders and injuries, such as frontotemporal dementia, vascular dementia, Creutzfeldt–Jakob disease and acute stroke. When CSF total tau is compared between AD and healthy subjects, it shows a sensitivity of 80% and a specificity of 90%. In the ADNI study,41 CSF total tau was 69.7 ± 30.4 pg/mL for the healthy control group, 101.4 ± 62.2 pg/ml for the MCI group, and 119.1 ± 59.6 pg/mL for the AD group, with significant difference among the three groups.


Tau is highly phosphorylated in the AD brain and p-tau at Thr181, Ser199, Thr231 or Ser235 have been measured with the sandwich ELISA method.42 In particular, p-tau at 181 threonine (181p-tau) has been extensively studied. The Thr181p-tau level was followed in the ADNI study,41 showing 24.9 ± 14.6 pg/mL for the healthy group, 35.5 ± 18.0 pg/mL for the MCI group, and 41.6 ± 19.8 pg/mL for the AD group, respectively. Thr181p-tau is specifically increased with AD, and it can discriminate from non-AD pathological conditions, such as frontotemporal dementia, diffuse Lewy body disease, vascular dementia, Parkinson's disease, amyotrophic lateral sclerosis, schizophrenia, and depression.


The combined index of CSF Aβ42 level and CSF p-tau level can be used for better diagnostic value in differential diagnosis of AD with specificity and sensitivity of 80–90%, which is high enough in clinical use. Higher CSF tau and lower CSF Aβ42 is correlated with the presence of apo E4 allele.43 It has been demonstrated that MCI or subjective cognitive impairment (SCI) subjects with higher CSF tau and lower Aβ42 in CSF have more tendency to develop AD in the near future. Furthermore, a higher index of Aβ42/ tau was significantly higher in SCI subjects (52%) than in the healthy controls (31%).44


There are many other biomarkers under study even though they are waiting for confirmation. β-site amyloid precursor protein-cleaving enzyme 1 (BACE1) activity45 is shown to be increased in CSF of AD patients. Products reflecting γ-secretase action are strong candidate biomarkers that reflect the γ-secretase activity modified in the AD brain.46–48 Biomarkers related with secondary process, such as oxidative stress, inflammation, glial activation and complement activation are also promising as possible biochemical markers for AD.


The present diagnostic criteria have two major shortcomings: (i) the criteria can give only possible AD before it can be confirmed as definite AD by autopsied brain examination; and (ii) a clinical diagnosis can only be made after excluding other diseases. After accumulation of data of biomarkers specific to AD diagnosis, more active use of biomarkers should be considered to improve the diagnostic criteria of AD.

If we aim for the earlier diagnosis of AD for possible intervention, it is advisable to adopt the new diagnostic criteria that enable us to give an early diagnosis of AD in the stage where clinical signs and symptoms of AD are still in the minimum stage.49

At the International Conference on Alzheimer's Disease in Hawaii in 2010, the National Institute on Aging (NIA) and the Alzheimer's Association jointly proposed a new understanding and new criteria of AD in which the significance of AD biomarkers are fully installed, and the concept of the AD spectrum was proposed.

This proposal can be validated by the understanding that biomarkers are available that are specific and sensitive enough for AD diagnosis. If we have identified the suitable biomarkers in clinical use, we will be able to give the diagnosis of the AD spectrum at an even earlier stage. In the proposal, the AD spectrum is divided into the three stages: (i) the preclinical stage; (ii) mild cognitive impairment; and (iii) clinical AD.

In the preclinical stage, only changes in a specific biomarker are observed with neither cognitive impairment nor clinical signs of AD. The mild cognitive impairment stage may include those showing biomarker changes as well as mild cognitive decline but no clinical signs and symptoms of AD. AD is diagnosed in patients with biomarker changes and clinical signs and symptoms of AD.

This concept will promote understanding of the continuous transition from preclinical stage to AD via mild cognitive impairment, in which biomarkers can be utilized to discriminate and clearly define each stage of the AD spectrum. The new criteria will promote earlier detection of the subjects who will develop AD in later life, and also to initiate intervention aiming for the prevention of AD.