Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Biomarkers for Alzheimer's disease: academic, industry and regulatory perspectives

Key Points

  • Recent research progress into the molecular pathogenesis of Alzheimer's disease has been translated into several promising drug candidates that have the potential to produce disease-modifying effects.

  • Biomarkers that reflect the central pathogenic processes in Alzheimer's disease have been developed and validated in numerous studies. These include cerebrospinal fluid (CSF) assays for tau and amyloid-β isoforms; magnetic resonance imaging (MRI) measurements of brain atrophy; and positron emission tomography (PET) techniques for brain metabolism and amyloid-β deposition.

  • Academic institutions, the pharmaceutical industry and regulatory organizations all agree that biomarkers have an important role in the drug development process.

  • Biomarkers have several potential uses in clinical trials. These include their use as diagnostic aids to enrich the patient sample with cases of Alzheimer's disease; as tools to identify and monitor the biochemical effect of the drug candidate; and as safety markers to detect potential side effects of the drug.

  • Evidence obtained from biomarker studies showing that a drug candidate affects the central disease processes in Alzheimer's disease will, together with a beneficial effect on cognition, be essential for the drug to be labelled as disease-modifying.

  • A catch-22-like situation exists in validating Alzheimer's disease biomarkers for use in drug development. Biomarker validation depends on effective drugs that target Alzheimer's disease pathogenesis, which are not currently available. At the same time, evidence from biomarker studies is needed for a new drug to be labelled as disease-modifying.

  • There are numerous ongoing clinical trials investigating disease-modifying drug candidates, which include biomarkers as end points. These trials will provide information on whether biomarkers will be valuable tools as surrogate end points to predict the clinical outcome and as the basis for a disease-modifying claim of the drug.

  • If disease-modifying drugs are approved for the treatment of Alzheimer's disease, biomarkers will facilitate the diagnosis of Alzheimer's disease very early on in the course of the disease before neurodegeneration is too severe and widespread.

Abstract

Advances in therapeutic strategies for Alzheimer's disease that lead to even small delays in onset and progression of the condition would significantly reduce the global burden of the disease. To effectively test compounds for Alzheimer's disease and bring therapy to individuals as early as possible there is an urgent need for collaboration between academic institutions, industry and regulatory organizations for the establishment of standards and networks for the identification and qualification of biological marker candidates. Biomarkers are needed to monitor drug safety, to identify individuals who are most likely to respond to specific treatments, to stratify presymptomatic patients and to quantify the benefits of treatments. Biomarkers that achieve these characteristics should enable objective business decisions in portfolio management and facilitate regulatory approval of new therapies.

Main

Basic and clinical research advances over the past decades have provided detailed knowledge of the molecular mechanisms and clinical course of Alzheimer's disease. Alzheimer's disease is a complex progressive condition that involves sequentially interacting pathological cascades, including the interaction of amyloid-β (Aβ) aggregation with plaque development, and the hyperphosphorylation and aggregation of tau protein with formation of tangles. Together with associated processes, such as inflammation and oxidative stress, these pathological cascades contribute to loss of synaptic integrity and progressive neurodegeneration1. These advances in research have been translated into several new drug candidates with disease-modifying potential, many of which are now being evaluated in clinical trials2. Studies in transgenic mouse models of Alzheimer's disease suggest that the majority of these new types of disease-modifying drugs may be most effective early on in the process of Aβ aggregation, and be less effective in later stages when there is severe plaque pathology and neurodegeneration3,4,5. However, the current diagnostic criteria (DSM-IV, ICD-10 and NINCDS-ADRDA) that are used to identify patients with Alzheimer's disease who have overt dementia correspond to neuropathologically advanced disease. Consequently, early recognition of the disease needs to be improved; indeed, it is estimated that interventions that could delay the clinical onset of dementia by 1 year would reduce the prevalence in 2050 by 9 million cases6.

The primary end points in current clinical trials for Alzheimer's disease measure the symptoms of the disease, such as cognitive and functional impairment. However, performance in these tests is determined by multiple factors other than key pathological events, which lead to neurodegeneration. As a result, drug trials that use clinical rating scales as outcome measures will fail to define the impact of new drugs on Alzheimer's disease pathology.

There is a growing need for biomarkers of Alzheimer's disease pathology to improve drug development related to the disorder. Animal models of Alzheimer's disease have low predictive power for determining the efficacy of treatments in patients with sporadic Alzheimer's disease1. Human biomarkers that facilitate the identification of the biochemical effects of a drug in short-term pilot studies may identify those drug candidates derived from animal models that affect the disease process in patients. In clinical academic research, when qualified for use (Box 1) in a specific patient population, biomarkers should aid in population selection and assessment of drug effects on disease progression. In industry-led drug discovery and development, biomarkers should facilitate selection of drug candidates, verify the mechanism of action, define dose effects and enable clinical trials to be shortened and run with reduced sample size. Biomarkers could serve as surrogate end points (discussed in more detail below) for clinical outcomes, and so could increase objectivity and efficiency in regulatory decision-making. These could include, for example, labelling decisions in which a drug could be labelled as 'disease modifying' rather than just providing symptomatic treatment7. In addition to this, biomarkers should serve as diagnostic tools in clinical practice to allow early and presymptomatic identification of patients with Alzheimer's disease, and to aid treatment decisions and monitoring in individualized care. Finally, biomarkers could serve as screening tools for disease prevention programmes.

To qualify for these purposes, a biomarker must be measured by reliable and validated methods, must be sensitive and specific when used as diagnostic markers, must be sensitive to the effects of a drug, and must be predictive of clinical outcomes8,9.

In clinical trials, biomarkers that are indicators of the central or downstream elements of Alzheimer's disease pathogenesis could serve at least three different purposes: as diagnostic biomarkers, to detect and monitor effects of drug candidates on the disease process, and as safety markers to detect and monitor potential side effects of drug candidates at an early stage. The last item may be of particular importance in clinical trials of novel drugs against Alzheimer's disease, given the risk of encephalitis associated with monoclonal antibodies10. Biomarkers can be compounds obtained from bodily fluids or tissue, such as cerebrospinal fluid (CSF) assays11, or technically derived correlates of Alzheimer's disease pathology, such as brain imaging markers. In particular, neuroimaging techniques12 have been developed that provide evidence for Aβ deposition, tau aggregation and neurodegeneration already at very early clinical disease stages13. However, the current generation of biomarkers are not yet robust against standard criteria that biomarkers in other branches of medicine fulfil. Also, the reporting of studies involving diagnostic biomarkers in Alzheimer's disease needs to follow strict quality criteria, such as the quality assessment of diagnostic accuracy studies (QUADAS) criteria14.

Here, we review the current status of multimodal core biomarker candidates derived from structural, functional and metabolic neuroimaging, and from neurochemistry and genetic studies of Alzheimer's disease. We also provide the converging perspectives of industry stakeholders and regulatory agencies on biomarker discovery and development. In our opinion, the integration of scientific knowledge and united international and interdisciplinary research efforts by academic institutions, the pharmaceutical industry and the regulatory agencies will accelerate the discovery and co-development of new and more informative biomarkers for broad clinical diagnostic use, as well as for the many trials of innovative treatments in neurodegenerative diseases and Alzheimer's disease.

Imaging as an end point in clinical trials

Four imaging modalities have been used as secondary end points in clinical trials on Alzheimer's disease: structural magnetic resonance imaging (MRI), functional MRI (fMRI), magnetic resonance spectroscopy (MRS) and positron emission tomography (PET). In structural MRI studies, there are correlations between MRI-based volume and neuron numbers in specific brain regions15. The blood oxygen-dependent level (BOLD) fMRI signal is primarily a measure of the input and processing of neuronal information within a brain region16. MRS represents changes in the biochemical composition of the brain tissue. PET using 18F-2-fluoro-2-deoxy-D-glucose (FDG) is thought to represent neuronal glucose consumption as the main determinant of neuronal metabolism17. PET using tracers for Aβ is thought to represent accumulation of the pathognomonic sign of Alzheimer's disease, that is, Aβ plaques18.

The validity of a biomarker with respect to a supposed neurobiological substrate will be relevant for the evaluation of disease-modifying treatments. The imaging techniques provide information on the regional distribution of changes on a macroscopic (fMRI, PET, MRS) or a mesoscopic (MRI) scale. Such knowledge on the spatial distribution and temporal dynamic changes of the brain in Alzheimer's disease is important. This is because a systematic brain disease that follows a specific pattern of progression through the brain may lead to differences in the extent of pathological changes depending on the stage of the disease. Additionally, imaging biomarkers have been used as predictive markers of the progression of dementia in defined risk groups of patients. For example, biomarkers can increase the accuracy of predicting the conversion of mild cognitive impairment — a clinically defined risk syndrome in which normal cognition and dementia cannot be sharply separated — to dementia19,20.

Structural MRI. In Alzheimer's disease, structural MRI typically shows a pattern of decreased grey matter in the parahippocampal gyrus, the hippocampus, the amygdala, the posterior association cortex and the subcortical nuclei including the cholinergic basal forebrain. In longitudinal studies MRI can be used as a potential marker to discriminate between disease-modifying and symptomatic treatment effects by determining the rate of atrophy of brain regions (Fig. 1). Table 1 and Supplementary information S1 (table) outline the characteristics of different MRI-based disease markers.

Figure 1: Detection of disease-modifying treatment effects.
figure1

a | The solid red line indicates the course of cognitive decline (assessed by cognitive testing) during disease-modifying or symptomatic treatment compared with the course of decline seen with placebo (green line). Only after withdrawal from drug treatment in the second half of the trial does the disease-modifying treatment effect (upper dashed red line) differentiate from the symptomatic treatment effect (lower dashed orange line). b | The upper solid blue line indicates the course of atrophy in the brain (as measured by magnetic resonance imaging) with a disease-modifying compound compared with a symptomatic treatment (lower pale blue line) and placebo (green line). This difference is maintained after withdrawal of the drug, but the withdrawal phase of the trial would not be required to differentiate the disease-modifying effect from the symptomatic treatment effect based on a volumetric outcome.

Table 1 Imaging biomarkers for Alzheimer's disease

The reliability of volumetric measures obtained from repeated MRI scans is generally high21, which is an important prerequisite for its use as a disease progression marker. Additionally, multicentre variability of manual and automated volumetric measures across 12 different MRI scanners was below 5% in one study22. To reduce the variability of MRI measures between different scanners in clinical trials, a phantom test is required to ensure that participating centres meet a set of minimal criteria for scanner quality. The most commonly MRI-derived measure is hippocampus volume, which is measured by visual inspection or manual drawing on MRI slices. However, more recent approaches use automated spatial transformation of MRI volumes in a common standard space and derive univariate or multivariate statistics from maps of cortical grey matter density, cortical thickness, white matter density or high-resolution spatial transformation maps. These approaches are less labour intensive than manual volumetric analyses and so far have been applied as secondary outcomes in few clinical trials. However, the multicentre reliability of these techniques has yet to be determined.

Pharmacological fMRI. Studies of memory in patients with Alzheimer's disease using fMRI discovered a pattern of altered activation in the medial temporal lobes and parietal lobes, which were consistent with structural MRI results. Based on initial single-centre studies, specific effects of treatment on regional brain activation could be detected in Alzheimer's disease23 (Fig. 2; Table 1; see Supplementary information S1 (table)). Multicentre and longitudinal studies using fMRI in patients with Alzheimer's disease have yet to be done, which will be a major step in the further development of fMRI-based biomarkers.

Figure 2: Detection of effects of cholinergic treatment on cortical activation in patients with Alzheimer's disease using functional magnetic resonance imaging.
figure2

Regions of decreased activation (shown in blue) during a visual perception task in a group of patients with Alzheimer's disease after 3 months of open-label treatment with an acetylcholinesterase inhibitor compared with activation before treatment. The visual perception task recruits parietal areas in healthy controls and the decrease in activation after treatment (blue areas) occurred in the cognitively relevant brain regions. Left side of image is left brain.

Proton-MRS. Proton-MRS (1H-MRS) provides quantitative biochemical measures of compounds in brain tissue. The best established 1H-MRS marker is the amino acid N-acetyl aspartate (NAA), which reflects the functional status of neuronal mitochondria24. A reduction of NAA levels independent of brain atrophy is a consistent finding in Alzheimer's disease. Preliminary single-centre studies have demonstrated that NAA levels are responsive to pharmacological treatment, and multicentre application of this technique has been demonstrated (Table 1; see Supplementary information S1 (table)). However, large-scale applications in clinical trials are still pending. In addition to NAA, several other metabolites — such as choline-containing compounds, creatine and phosophocreatine, myoinisitol, and gluatmate and glutamine — are detectable with 1H-MRS, but their potential as biomarker candidates in Alzheimer's disease is controversial and requires refined investigations25.

FDG-PET. FDG-PET measures local glucose metabolism as a proxy for neuronal activity at a resting state without the need for cognitive activation. In Alzheimer's disease, neuronal activity is impaired, as FDG uptake is reduced, predominantly in temporoparietal association areas including the precuneus and posterior cingulate cortex. These changes are closely related to cognitive impairment as demonstrated in cross-sectional and longitudinal studies. The changes in FDG uptake can be measured objectively with smaller coefficients of variance than standard neuropsychological measures, thus increasing the power of studies26,27. Alterations in FDG uptake are usually attributed to pharmacodynamic drug effects but will also reflect disease progression, especially when measured after several months of follow-up. This technique has been used as a secondary outcome parameter in some clinical trials28,29,30,31.

Amyloid-PET. Several studies have shown that selective binding of a11C-labelled thioflavin analogue known as Pittsburgh compound B (11C-PIB)32 to Aβ can be visualized in PET scans of patients with Alzheimer's disease (Table 1; see Supplementary information S1 (table)). These amyloid-PET scans have provided information regarding the Aβ plaque burden that is independent from structural changes in brain anatomy33. 11C-PIB binding sites in the brain are associated with Aβ sheets in classical plaques and diffuse plaques, as well as cerebrovascular amyloid angiopathy18,34. It does not bind to soluble and oligomeric amyloid. Increased cortical uptake of 11C-PIB is seen in 2 out of 3 of patients with mild cognitive impairment with probable progression to Alzheimer's disease35,36. However, a significant proportion of elderly subjects without mild cognitive impairment show increased 11C-PIB uptake with as yet unknown prognostic implications33,35. Currently, Phase II trials have been initiated with amyloid tracers that can be labelled with 18F isotopes, which have a longer half-life and wider availability compared with 11C-PIB37.

Cholinergic neurotransmission. Cholinergic projections from the basal forebrain to cortical areas degenerate in Alzheimer's disease, and treatment with choline esterase inhibitors has modest but proven efficacy. Ligands for choline esterase, cholinergic receptors and transporters have been used in human studies (Table 1; see Supplementary information S1 (table)), but have not yet been qualified as biomarkers.

Imaging biomarkers: future challenges. Presently, clinical trials include imaging end points on a project-by-project basis. The standardization of imaging protocols and analysis techniques is paramount for acceptance in multicentre trials. In addition, regulatory authorities require proof that imaging outcomes are of relevance to drug safety or efficacy. A biomarker would serve as a surrogate end point if it reflects clinically relevant outcomes such as cognitive decline or health economic outcomes such as institutionalization. For instance, hippocampal atrophy could serve as a surrogate end point if hippocampus atrophy reflects the clinically relevant outcome of memory loss. This association needs to be further established in future studies.

Nevertheless, an imaging marker could be used as a primary pre-specified outcome measure in a proof-of-concept study if the imaging marker reflects an underlying disease process. Hippocampus atrophy as determined by MRI correlates with hippocampal neuronal loss as shown in numerous clinical pathological studies15,38. Thus, hippocampus atrophy could be used to assess the mechanism of action of a new compound that claims to reduce neurodegeneration in Alzheimer's disease. Therefore, imaging markers may be used to assess the presumed mechanism of action of a new compound in proof-of-concept studies. For example, in the AN1792 vaccination trial39, as well as in the ALZHEMED Phase III trial40, there were higher rates of brain and hippocampus atrophy in the treated subjects than in the untreated controls. However, treated subjects showed no evidence of more rapid cognitive decline compared with the placebo groups. These examples illustrate that imaging end points can lead to unexpected findings that help to critically evaluate our primary assumptions on the mode of action of new compounds.

Genetic analysis in biomarker research

The use of all the biomarkers discussed in this paper — and indeed all biomarkers for any disease — are limited in part by variability between cases of disease and variability between controls. This variability increases the sample sizes that are needed to achieve statistical significance and is notably higher, for many reasons including genetic variability, in humans compared to animal studies. So, genetic analysis of individuals is beneficial in four broad ways.

First, some biomarkers (such as Aβ42) came directly from genetic analysis of kindreds with the condition. As other genetic risk factors for Alzheimer's disease are identified, it will be appropriate to assess their protein products, and other proteins, in the same pathway. Recent studies using large whole genome scans for Alzheimer's disease41,42 identified clusterin (CLU; also known as apolipoprotein J) and complement receptor 1 (CR1) as risk loci, suggesting that these proteins themselves may be useful biomarkers for Alzheimer's disease (although admittedly apolipoprotein E (APOE) has not proved useful in this regard). These results also suggest that the complement cascade, of which both CLU and CR1 are a part of, should be investigated.

Second, it is well known that those who are either amyloid precursor protein (APP) or presenilin (PSEN) mutation carriers, or who have Down's syndrome or are APOE ɛ4 allele homozygotes, have an extremely high risk of developing Alzheimer's disease at relatively well-defined ages. These defined populations offer the possibility of following the presymptomatic changes in biomarker levels and therefore correlating their behaviour with clinical state. This approach has been particularly informative and useful in the analysis of brain imaging biomarkers43,44,45, but also offers potential in the assessment of blood and CSF biomakers46.

Third, one of the problems of biomarker research is normal variability in the level of the biomarker between individuals. At least some of this variability is genetic and can be factored out through genetic analysis. For example, expression levels of microtubule-associated protein tau (MAPT) in the brain47 and tau levels in the CSF48,49 have been suggested to be influenced by the MAPT haplotype. This phenomenon is true of many proteins for which blood (and CSF) levels are directly influenced by genetic variability50.

Fourth, the incorporation of data relating to an individual's response to a drug and to how the drug is metabolized (which can vary between individuals) into biomarkers and in clinical studies will be helpful51.

Genetic tests and Alzheimer's risk. Genetic tests are available for individuals who have family members with early-onset familial Alzheimer's disease, and are offered using the Huntington's protocol. These genetic screens include testing for PSEN1, PSEN2, APP and APOE mutations; however, as APP or PSEN2 mutations are rare, screening for these mutations currently has to be organized on an ad hoc basis. Although genetic testing of APOE and predictive testing in typical late-onset disease had been discouraged52, the predictive value of APOE ɛ4 homozygosity is as high as for many other Mendelian diseases. Indeed, it seems that attitudes to APOE genotyping are changing53, and it is likely that the predictive potential of genetic analysis will be considerably increased. This is in part due to a growing appreciation of the increase in predictive value given by genotyping the whole APOE locus (including the TOMM40 region of the APOE promoter), together with hits from the recent genome-wide association studies41,42.

A likely outcome of these endeavours is that these predictive tests will classify individuals at risk of developing Alzheimer's disease into three groups. The first group includes individuals who are APOE ɛ4 homozygotes at high risk of developing Alzheimer's disease. This risk is modulated by other genes and the polymorphisms of the APOE promoter (about 3% of the population). The second group includes individuals who are APOE ɛ4 heterozygotes who are at modest risk of developing Alzheimer's disease. For this group, the precise risk is modulated by other APOE alleles (ɛ2 being less risky than ɛ3) and by the polymorphisms of the APOE promoter, as well as by the other genome-wide association study hits. The third group includes individuals who do not carry an APOE ɛ4 allele and are therefore at low risk of developing Alzheimer's disease. The interpretation of these data will need the development of a sophisticated risk chart, and their explanation to lay people will present a challenge. Although still to be investigated, it is expected that the adjunct use of these genetic tests will help considerably in the identification of individuals with mild cognitive impairment who will later be diagnosed with Alzheimer's disease54.

As our understanding of the genetic architecture of Alzheimer's risk increases through the identification of relatively common risk variants such as CLU, C1R and PICALM (phosphatidylinositol binding clathrin assembly protein), and the later identification of individually rare high-risk variants, we will be able to asymptotically construct a predictive algorithm of Alzheimer's disease risk. Although this will considerably improve on what can be currently achieved, it is unlikely that this algorithm will be amenable for screening large at-risk populations (as indicated in our proposed diagnostic flow model depicted in Fig. 3).

Figure 3: The four categories of biomarker: target, mechanism, pathophysiological and diagnostic.
figure3

Biomarkers can be categorized into four groups on the basis of their contribution to business, regulatory and clinical decision-making. Clinical decision-making can be further divided into clinical research and patient care diagnostic subcategories. The objective is to use biomarkers as early as possible in the drug development process. The initial step is to confirm that a test compound hits the target and to quantify the extent to which it does so. Next is to test three concepts in logical sequence. First, that hitting this target alters the pathophysiological mechanism. Second, that altering this mechanism affects the pathophysiology. Third, that affecting pathophysiology predictably improves the clinical status of the patients. Biomarkers qualified to confirm the presence of the target and or extent to which the drug candidate hits the target may be validated later as diagnostic tests for early detection or diagnosis of Alzheimer's disease (when that target is expressed differentially between healthy and diseased states). Biomarkers qualified for confirming and quantifying mechanistic effects may be validated later as diagnostic tests to inform choice of therapeutic regimen, either in choice of drug or initial dosing regimen. Biomarkers qualified for longitudinal quantification of patient response in terms of clinically relevant pathophysiology, may be validated later as diagnostic tests for monitoring and individualization of a therapeutic regimen. Biomarkers qualified for either monitoring or individualization of therapy on clinically relevant pathophysiology may also serve as surrogate end points to support regulatory decision-making. In addition, they can be used to ensure appropriateness of use, and as quantifiers of clinical outcomes to support reimbursement decisions.

Biochemical biomarkers for use in clinical trials

There are two main types of biochemical (that is, other than genomic or imaging) biomarkers. The first are core biomarkers that mirror fundamental pathogenic events in Alzheimer's disease; for example, the deregulated metabolism of APP and Aβ. The second are downstream biomarkers that reflect secondary phenomena, such as axonal degeneration. Compared with CSF biomarkers, efforts to discover reliable biomarkers for Alzheimer's disease in peripheral blood have been of limited success. Indeed, although there are many publications on potential candidate blood biomarkers, follow-up studies by other research groups are either lacking, or have failed to confirm a solid diagnostic value. Nevertheless, a recent pilot study reported diagnostically useful patterns of 18 different signalling, acute phase and inflammatory proteins in plasma that warrant further investigation55. Plasma Aβ40 and Aβ42 can be measured in peripheral blood, but do not contribute to the identification of Alzheimer's disease in a reproducible manner56, and are unlikely to reflect Aβ processing in the brain57. Instead, most research has focused on CSF biomarkers, which more directly reflects brain neurochemistry, and can be obtained through lumbar puncture without significant side effects58.

Established CSF biomarkers. Four CSF biomarkers have been evaluated in a large number of independent studies: Aβ40, Aβ42, total tau and phosphorylated tau59. Aβ40, Aβ42 and phosphorylated tau reflect the core elements of the disease process in Alzheimer's disease; that is, levels of brain amyloid60,61 and cortical neurofibrillary pathology62. Conversely, total tau is a nonspecific marker for axonal damage that mirrors the activity of the neurodegenerative process11. The combination of elevated levels of total tau and phosphorylated tau together with reduced levels of Aβ42 or reduced Aβ42/Aβ40 ratio in CSF is a consistent finding in biomarker studies of patients with different stages of Alzheimer's disease, including mild cognitive impairment11 (Table 2). A high diagnostic performance of these CSF biomarkers has been verified in three large multicentre studies, including the Alzheimer's disease Neuroimaging Initiative (ADNI) study63, the DESCRIPA study64, and the Swedish Brain Power project65.

Table 2 Core CSF candidate biomarkers for Alzheimer's disease

Potential CSF biomarkers. There are numerous other candidate biomarkers that reflect either elements of the primary pathogenic process in Alzheimer's disease or secondary events of the disease. Biomarkers that mirror the pathogenic process include, for example, Aβ oligomers, β-site amyloid precursor protein-cleaving enzyme 1 (BACE1) activity and concentration, secreted isoforms of APP, and Aβ degradation products66. Secondary events include oxidative stress responses, inflammation and gliosis67. The diagnostic potential of these biomarkers is less well studied. However, some biomarkers such as BACE1 activity68 and APP isoforms69 may also give important information on desired biochemical effects of certain drug candidates, such as BACE1 inhibitors.

Monitoring the biochemical effect of new drugs. Because there is slow and variable progression of symptoms in Alzheimer's disease, very large cohorts of patients and a treatment duration of several years will be needed to identify a change in the rate of cognitive decline in clinical trials of disease-modifying drug candidates. Small, short-term trials that provide biochemical evidence of an effect of the drug on the central pathogenic processes would therefore be of great value to make go/no-go decision before embarking on Phase II or III clinical trials.

Clinical treatment trials with acetylcholine esterase (AChE) inhibitors may serve as a proof of concept for the potential of CSF biomarkers to identify and monitor the biochemical effect of a drug in Alzheimer's disease. Such studies found a marked increase in CSF AChE activity during treatment, which was dose dependent and linked to the mechanism of action of the drugs and the clinical outcome70. The mechanism for the increase in CSF AChE is unknown, but may involve an increased gene expression and release of AChE to the CSF, which is mediated by a muscarinic acetylcholine receptor feedback loop71.

The intra-individual variability of CSF tau proteins and Aβ42 is very low over time, with coefficients of variation of 4–6% during 6 months and 7–9% during 24 months of follow-up72,73. These results indicate that even minor changes in biomarker levels can be monitored. Studies in transgenic mice and in guinea pigs show reduced cortical, CSF and plasma Aβ levels during γ-secretase inhibitor treatment74,75. The decrease in CSF Aβ also correlated with the decrease in cortical Aβ75. In a Phase II trial of the γ-secretase inhibitor LY450139, acute treatment resulted in a dose-dependent decrease in plasma Aβ levels, whereas after chronic treatment there was only a tendency for a reduction in CSF Aβ42 (Ref. 76). Treatment of monkeys with a BACE1 inhibitor resulted in lower CSF Aβ42, Aβ40 and soluble APPβ levels77. Furthermore, a clinical trial on PBT2, which is suggested to affect metal-induced Aβ aggregation78, showed a dose-dependent reduction in CSF Aβ42 levels79. The mechanism underlying the reduction in Aβ42 during treatment is unclear. Data from a small clinical study on the amyloid-targeting drug phenserine also suggest that CSF Aβ may be of value for evaluating treatment effects31.

Based on longitudinal studies of conditions that involve acute neuronal injury80,81 and data from the interrupted Phase IIa AN1792 trial82, total tau should decrease towards normal levels if a treatment is successful in inhibiting the neurodegenerative process. The same may be expected for phosphorylated tau 181, which is supported by one pilot study of memantine83.

It is currently unknown whether changes in levels of biomarkers, especially for CSF Aβ42, during treatment will be different for Alzheimer's disease cases who have low levels of CSF Aβ42 due to deposition in plaques at beginning of treatment. The degree and direction of change may also depend on the time-point of sampling after initiation of treatment. In summary, these findings support the proposal that CSF biomarkers could be valuable tools to monitor biochemical drug effects in clinical trials.

Industry perspective on biomarkers

Monitoring the biochemical effect of new drugs. Biomarkers that are qualified for use in clinical trials to facilitate business and regulatory decision-making should also be available as diagnostic agents to enable appropriate prescribing. Therefore, each biomarker will have utility at one stage or another stage of medical product development; that is, from discovery to adoption in clinical practice (Fig. 4). Moreover, the biomarkers could be used in our proposed diagnostic flow that incorporates risk assessment screening, diagnosis and prognosis, and monitoring treatment effect (Fig. 3). This diagnostic flow begins with tests that have high sensitivity but low specificity (and low cost) to those that have increasing specificity and potential for longitudinal quantification of treatment benefit.

Figure 4: Translation of biomarkers.
figure4

The translation of biomarkers from being research tools that can be used in clinical trials, to being aids in regulatory decision-making and to being commercial diagnostics that can aid clinical decision-making facilitates the efficient selection of different patient populations. When used as research tools, biomarkers can identify a large population of individuals (that is, the base of the pyramid) who may benefit from low-cost and safe primary prevention strategies. Diagnostic biomarkers can be used in smaller populations with early forms of the disease which, when proven by a definitive diagnosis, warrants a more expensive and riskier therapeutic intervention that is tailored to their individual pathophysiology and monitored for efficacy (that is, the pinnacle of the pyramid). This role of biomarkers for diagnostics in individualization of therapy is also enabled by information technology, known as 'clinical decision support'. Different biomarkers such as genomics and molecular imaging will each be useful in a diagnostic flow. This begins with risk assessment (enabling primary prevention) to screening (for early detection and early intervention), to diagnosis and prognosis (for staging and best choice of therapy) and finally to monitoring treatment effect (for true individualization of treatment). This flow starts from tests with high sensitivity but low specificity and low cost to those with increasing specificity and value with potential for longitudinal quantification. Patients will enter at different points in this diagnostic flow (from the bottom to the top of the pyramid) according to the manner in which they present; the two main options being with or without symptoms.

What an Alzheimer's disease biomarker could achieve in drug development. The determination of complete pharmacokinetic and pharmacodynamic relationships in relevant preclinical models shortens and aids objectivity in the selection of drug candidates and doses for clinical development. Given the costs associated with clinical development, the long duration of clinical trials relative to patent life and the difficulty in recruiting participants for clinical trials — which is partly due to the multiplicity of drugs being developed and other factors such as patient awareness — no drug candidate should enter the clinic without a species-independent biomarker as the central element of translational medicine (Fig. 4).

In addition to measuring pharmacological and toxicological effects during clinical trials, as well as efficacy and safety issues that arise post-approval, biomarkers facilitate stratification of the patients and dose optimization based on phenotype or genotype. Moreover, enrichment strategies for recruitment to the trial (that is, choosing only those most likely to gain benefit or least likely to suffer an adverse event) can shorten trials and improve the response rate. A successful example of this approach is HER2/neu (also known as ERBB2) testing in breast cancer for trastuzumab (Herceptin; Genentech) treatment84. Furthermore, biomarkers can be used for within-patient dose titration during clinical trials, thereby effectively individualizing the therapeutic index. For example, a patient without amyloid plaques in the brain is less likely to respond to a drug that targets Aβ, whereas patients screened into the study with a positive plaque biomarker at baseline may not respond to the low starting dose, therefore warranting escalation to the next dose.

Identifying biomarkers at Phase 0. Currently, the efficacy of drugs for Alzheimer's disease is assessed in patients clinically diagnosed with symptomatic Alzheimer's disease. Although refinements in neuropsychological testing are facilitating earlier (presumptive) diagnoses, it is hoped that in the future Alzheimer's disease can be diagnosed definitively before symptoms become apparent. It is at this stage that the drug has a better chance of efficacy, before the pathological changes in the brain are too advanced for therapeutic intervention. This is particularly important because of the long duration of pathophysiology changes before Alzheimer's disease manifests in cognitive loss, and the shortcomings in clinical diagnosis85. For this purpose, biomarkers of early stages of Alzheimer's disease are needed86.

There is currently no regulatory framework for which to approve drugs that could treat Alzheimer's disease in its presymptomatic stages, so in anticipation of efficacy trials eventually being performed in early-stage disease, and in hope of drugs being approved for asymptomatic Alzheimer's disease, Phase 0 trials of potential biomarkers should be started immediately so that the data are available to inform design of the Phase I–III trials in early disease87,88. Such a biomarker of early-stage pathophysiolgy would also enable treatment effects to be distinguished from symptomatic improvements, so that the drug can be labelled and prescribed appropriately (and priced accordingly). Such labelling could include disease modification claims that were based on the drug's modification of the pathophysiological mechanism89. Conversely, if it is demonstrated that a drug inhibits the target without affecting downstream biomarkers, the drug class and target need reconsideration (Table 3). Molecular imaging biomarkers are particularly valuable because they provide non-invasive anatomical specificity for Aβ plaques, tau and neurofibrillary tangles in animals and in humans.

Table 3 Recommended biomarkers in clinical research and development

Clinical use of Alzheimer's disease biomarkers. The most valuable role for biomarkers has been identified by the US Food and Drug Administration (FDA), and the National Cancer Institute90 as a clinical diagnostic (rather than a surrogate end point). A convergence is occurring between the requirements of biomarkers for quantification of drug effects in research and development, which are analysed as population means with standard errors, and the requirements of diagnostics in clinical practice, which are assessed on a per-patient basis. The common element in this convergence is longitudinal quantification; both analyses require pretreatment and post-treatment effects of the drug to be measured. As diagnostics, biomarkers are of interest to payers and purchasers of health care for parallel applications. As an example, clinical evidence from the National Oncologic PET Registry resulted in expanded coverage by Medicare for FDG-PET/CT. In addition to diagnosis and staging, a new 'subsequent' category was covered, which includes the monitoring of therapeutic benefit. Furthermore, earlier detection of disease facilitates earlier intervention, which, when followed by effective, individualized treatment, will yield better patient outcomes and reduced institutionalization. This is not unique to Alzheimer's disease, but is particularly important because of the development of significant pathophysiology before the appearance of symptoms in Alzheimer's disease, the inaccessibility of the lesions in the brain, and the need to distinguish between causes of dementia and then discern symptomatic improvement from disease modification. Once a biomarker is proved to be 'qualified for use' (a phrase to be defined by regulatory guidance), within-patient dose titration can be incorporated into design of clinical trials and the resultant biomarker data could provide the information necessary for prescribing information to optimize outcomes for individuals by dose adjustments in clinical practice (Table 3). The biomarkers that are most likely to be utilized for the individualization of therapy are those that are useful in the longitudinal quantification of drug effects during research and development.

Who should pay for biomarker development? The costs and risks involved in qualifying biomarkers to satisfy the regulatory authorities, whether as surrogate end points or as a standard biomarker, may be prohibitive for a single company to invest in, but the alternatives of collaboration or public–private partnerships are complex. These costs, risks and complexities could be addressed by the issuance by the FDA, the European Medicines Agency (EMA) and other regulatory bodies, of harmonized guidelines for biomarker qualification that were more practical than unattainably pure statistical criteria that require capture of both efficacy and safety effects. Such guidelines would clearly outline the linkage between the degree of qualification required and the utility of the biomarker (that is, the biomarker would be qualified for use).

What creates disincentives? The qualification of a biomarker for disease modification requires that the biomarker is shown to capture the treatment benefit of a drug in development, that is, the drug candidate is a positive control. Once this correlation between treatment benefit and biomarker has been achieved for two drugs with the same mechanism of action it might be construed as a surrogate end point for drugs with that mechanism of action. A biomarker that correlated with clinical benefit across drugs with differing mechanisms of action might be construed as a surrogate end point irrespective of mechanism of action.

Although companies may collaborate on Phase 0 trials, such as the ADNI (see Further information), trials that aim to measure treatment effect are inherently linked to specific compounds. The first company to succeed may enter the market earliest but will have taken all the risks, thereby effectively underwriting the qualification of that biomarker for competitors to use in reducing development costs, reducing regulatory risk, and increasing speed to market. When a second company successfully deploys that same biomarker, it might elevate it to surrogate end point status, but with the benefits again accruing to competitors and, ultimately, to the regulators and to patients.

An exclusivity provision analogous to that developed for paediatric trials ( Guidance for Industry Qualifying for Pediatric Exclusivity Under Section 505A of the Federal Food, Drug, and Cosmetic Act ) might compensate for these disincentives; however, more immediately, regulatory guidance on the qualification of biomarkers would be helpful, perhaps rendered less complex by making it specific to particular research platforms (for example, imaging as distinct from in vitro tests). Meanwhile, the qualification of a biomarker as a surrogate end point might be achieved after marketing. Guidance on co-development of therapeutics and diagnostics would also be helpful.

Regulatory perspective on biomarkers

For regulatory purposes, the efficacy and safety of treatments are typically determined in at least two separate randomized, double-blind, placebo-controlled trials that are each of at least 6 months duration. As a further requirement, efficacy must be established in a cognitive domain and a functional (for example, activities of daily living) or global domain (for example, clinical global impression of improvement). Significant differences must be demonstrated in both end points. These co-primary outcome measures are required for symptomatic and disease-modifying approaches in pivotal Phase III studies. However, because of the pattern of symptom progression in Alzheimer's disease and depending on the proposed mechanism of action of the drug, attempts to establish that a drug leads to a disease-modifying effect are likely to require a much longer study duration and alternative trial designs91 compared with trials that measure symptomatic improvements.

As a consequence, the use of biomarkers for disease-modifying approaches and the search for adequate surrogate end points is encouraged by all stakeholders involved in drug development in Europe (for example, the Qualification of novel methodologies for drug development — Guidance to applicants by the EMA, and the Improved Predictivity of Efficacy Evaluation — Brain Disorders by the Innovative Medicines Initiative) and the United States92 (for example, the Critical Path Initiative). However, it is necessary to consider specific regulatory requirements when biomarkers are used as outcome measures to assess therapeutic agents — particularly those that lead to disease modification — for Alzheimer's disease.

Biomarkers as surrogate end points. Although the use of biomarkers in earlier phases of drug development is well established, their value in pivotal efficacy studies for Alzheimer's disease is still limited. A biomarker is a surrogate end point when it can be considered as a substitution for a clinically relevant end point that is a direct measure of how a patient feels, functions or survives, and can be expected to predict the effect of and allow measurement of the specific therapy93,94 (Box 2). However, even a perfect correlation between the level of marker with stage of disease in the untreated state is not sufficient to accept a biomarker as a surrogate for a clinical end point95,96. None of the imaging or biochemical markers has been sufficiently qualified as a surrogate end point in Alzheimer's disease. There should be a link between a treatment-induced change in the biomarker and the desired clinical outcome measure, as well as a link between the treatment-induced change in the biomarker and change of disease process95,96. Additionally, based on the assumed mechanism of action of a given compound, there should be high plausibility (for example, based on preclinical models) that the disease process will be modified.

So, although the use of biomarkers as primary outcome measures (that is, surrogate end points) in pivotal efficacy trials is unlikely at this time, they can currently be used to allow decision-making on other aspects of drug development. For instance, they can be used in Phase II studies (for example, proof of concept, dose finding) as primary pre-specified outcome measures or to better define patient populations at risk (for example, enriching populations likely to respond to therapy) for efficacy trials (Box 3). The use of biomarkers in these ways may allow Phase II studies to be shorter or smaller in size, and if potential biomarkers can be qualified as surrogate end points, definitive effectiveness trials may also be considerably shorter and/or smaller than studies that use traditional clinical outcomes as primary end points.

Problems and concerns with using biomarkers as surrogate end points. As noted above, the reliance on the drug's pharmacological effect to modulate levels of a biomarker that has not been qualified for use as a surrogate end point has interpretative problems. For example, it is assumed that a treatment that is efficacious in Alzheimer's disease will slow the progression of medial temporal lobe atrophy as measured by MRI. However, in the trial with the vaccine AN1792 the extent of brain atrophy seemed to increase in patients that had increased antibodies to Aβ and that showed clinical improvement97. This was an entirely unexpected outcome, and provides evidence that the effects of a treatment (even a potentially beneficial treatment) on a biomarker can be unpredictable.

Of more concern though are potential cases in which the desired effect on a non-validated biomarker is achieved. If the clinical effects of a medicinal product are unknown, then concluding that the drug has a beneficial effect for the patients would be based on the assumption that the desired effect seen on the biomarker will translate into the desired beneficial clinical effect. If this assumption is wrong, a drug that has no beneficial effect (or even, perhaps, a deleterious effect) might be approved. It is for this reason that approval of a treatment for Alzheimer's disease based on an effect on a biomarker alone (that is, a surrogate end point) is unlikely at this time. However, as outlined in other sections, there is progress with some biomarkers towards validation for this purpose. As a general principle, if it is possible to ascertain effects of a drug on clinical outcomes in trials of reasonable size and duration, it is likely that approval of the drug will not be granted if the study has examined biomarkers as primary outcomes. That is, a clinical outcome is preferable to a biomarker. The use of biomarkers as surrogate end points is likely to be considered in settings in which clinical outcomes cannot be practically assessed; for example, in studies that examine prevention of Alzheimer's disease in which the clinical outcomes (such as the time of onset of clinical Alzheimer's disease) may not occur for many years after treatment initiation.

Neurochemical markers from CSF that have high sensitivity and specificity for Alzheimer's disease may be considered helpful for diagnostic purposes, whereas addition of a neuroimaging tool such as MRI or amyloid-PET may also offer the possibility to assess the effects of a particular treatment in these populations. Multicentre and worldwide activities such as those carried out by the ADNI — which use several pre-specified neurochemical and neuroimaging biomarkers, and standardization across sites and platforms98,99 — should help in further qualification of at least some of these biomarkers.

Creating innovation in biomarker development. Regulatory bodies such as the FDA and the EMA have identified development of biomarkers as a high priority in general and particularly in dementia. To foster innovation in this field, consortia like the Biomarker Consortium, the Critical Path Initiative and the Innovative Medicines Initiative, which involve collaboration of the different stakeholders, are highly welcomed and proactively supported. We think that such public–private partnerships (Box 4) encourage detection and development of new biomarkers and improve consensus on the requirements of their validation process. This will be supported by the ongoing dialogue on harmonization of regulatory practices between the EMA and the FDA, and both organizations will consider the experience with biomarkers in other fields — for example, oncology or cardiovascular medicine — for Alzheimer's disease. Therefore regulators encourage the different stakeholders to involve regulatory bodies as early on as possible in biomarker development to accomplish better coordination of regulatory requirements and ongoing research in the different stages of drug development.

Discussion and perspectives

There is a consensus between academic institutions, industry and regulatory authorities that there are three potential uses of biomarkers in clinical trials. One use is as diagnostic tools to enrich the sample of patients with Alzheimer's disease and to exclude other causes of dementia for which symptoms resemble Alzheimer's disease. However, this could restrict the clinical indication of the investigational compound in the general population. Furthermore, as drug candidates with a disease-modifying effect can be expected to be most effective in the earlier stages of the disease, biomarkers would be essential for early disease detection.

A second use of biomarkers is to identify and monitor the biochemical effect of a drug to facilitate drug development. This would increase the predictive power of non-imaging biomarkers when translating drug effects from mouse models of Alzheimer's disease to the clinical setting. Of the more than 40 molecules that reduce plaque burden in these animal models1 several lack preventive or clinical effect in treating patients with Alzheimer's disease. In addition, small, short-term clinical trials using biomarkers can verify a biochemical effect of disease-modifying drug candidates, before the expensive and time-consuming step is taken to larger Phase II or III clinical trials. Indeed, biomarkers have already been implemented in Phase I and II studies in patients with Alzheimer's disease, and are non-controversial for regulators. But before a biomarker can be used as surrogate end points in Phase III studies, a link between the treatment-induced change in the biomarker and the desired clinical outcome has to be established. Large-scale worldwide multicentre initiatives, such as the ADNI100,101, will support validation and qualification efforts in this respect.

A third application of biomarkers in clinical trials is to enable early and specific detection of side effects of the drug. As an example, in Aβ immunotherapy trials there is a risk of encephalitis10 and vasogenic oedema102, which may be difficult to identify clinically, but are easily detected by MRI and CSF analyses. Furthermore, if longitudinal MRI and CSF measures are part of a clinical trial, this can provide definitive information that a drug candidate is not associated with this type of adverse events.

Although several MRI-, PET- and CSF-based biomarkers have proved useful as diagnostic markers in clinical trials, there are only some reports suggesting that biomarkers may identify and monitor the effect of the drug on the neurodegenerative process70,79,82, and there are no qualified surrogates for the clinical end point. It is hoped that the involvement of regulatory authorities as early on as possible in the biomarker discovery process will accomplish better coordination of regulatory requirements for the qualification of biomarkers, and that biomarker consortia, with strong collaboration of the different stakeholders, will foster the ongoing dialogue on harmonization of regulatory practices between the EMA and the FDA.

Another challenge is the lack of drugs with an established disease-modifying effect that can be used to validate that a certain biomarker will correlate with the biochemical effect of the drug and therefore predict the clinical outcome. At the same time, evidence that a drug candidate affects the central disease process and the characteristic neuropathology will be required to label the drug as disease-modifying. This dual dependency introduces a catch-22-like situation in Alzheimer's disease drug development and biomarker research. Nevertheless, there are numerous ongoing clinical trials on disease-modifying drug candidates that include biomarkers as end points. These trials will successively provide accumulating evidence on whether biomarkers will be useful as proof-of-concept tools for the intended mechanism of action of a compound, as surrogate end points to predict the clinical outcome, and as the basis for a disease-modifying claim of the drug.

References

  1. 1

    Blennow, K., de Leon, M. J. & Zetterberg, H. Alzheimer's disease. Lancet 368, 387–403 (2006).

    CAS  PubMed  Google Scholar 

  2. 2

    Klafki, H. W., Staufenbiel, M., Kornhuber, J. & Wiltfang, J. Therapeutic approaches to Alzheimer's disease. Brain 129, 2840–2855, (2006).

    Article  PubMed  Google Scholar 

  3. 3

    Garcia-Alloza, M. et al. Existing plaques and neuritic abnormalities in APP:PS1 mice are not affected by administration of the gamma-secretase inhibitor LY-411575. Mol. Neurodegener. 4, 19 (2009).

    PubMed  PubMed Central  Google Scholar 

  4. 4

    Das, P., Murphy, M. P., Younkin, L. H., Younkin, S. G. & Golde, T. E. Reduced effectiveness of Aβ1–42 immunization in APP transgenic mice with significant amyloid deposition. Neurobiol. Aging 22, 721–727 (2001).

    CAS  PubMed  Google Scholar 

  5. 5

    Levites, Y. et al. Anti-Aβ42- and anti-Aβ40-specific mAbs attenuate amyloid deposition in an Alzheimer disease mouse model. J. Clin. Invest. 116, 193–201 (2006).

    CAS  PubMed  Google Scholar 

  6. 6

    Brookmeyer, R., Johnson, E., Ziegler-Graham, K., Arrighi, H. M. Forecasting the global burden of Alzheimer's disease. Alzheimers Dement. 3, 186–191 (2007).

    PubMed  Google Scholar 

  7. 7

    Siemers, E. R. How can we recognize “disease modification” effects? J. Nutr. Health Aging 13, 341–343 (2009).

    CAS  PubMed  Google Scholar 

  8. 8

    Frank, R. & Hargreaves, R. Clinical biomarkers in drug discovery and development. Nature Rev. Drug Discov. 2, 566–580, (2003).

    CAS  Google Scholar 

  9. 9

    Frank, R. A. et al. Biological markers for therapeutic trials in Alzheimer's disease. Proceedings of the biological markers working group; NIA initiative on neuroimaging in Alzheimer's disease. Neurobiol. Aging 24, 521–536, (2003).

    PubMed  Google Scholar 

  10. 10

    Orgogozo, J. M. et al. Subacute meningoencephalitis in a subset of patients with AD after Aβ42 immunization. Neurology 61, 46–54 (2003).

    CAS  PubMed  Google Scholar 

  11. 11

    Blennow, K. & Hampel, H. CSF markers for incipient Alzheimer's disease. Lancet Neurol. 2, 605–613 (2003).

    CAS  PubMed  Google Scholar 

  12. 12

    Teipel, S. J., Meindl, T., Grinberg, L., Heinsen, H. & Hampel, H. Novel MRI techniques in the assessment of dementia. Eur. J. Nucl. Med. Mol. Imaging 35 (Suppl. 1), 58–69 (2008).

    Google Scholar 

  13. 13

    Hampel, H. et al. Core candidate neurochemical and imaging biomarkers of Alzheimer's disease. Alzheimers Dement. 4, 38–48 (2008).

    CAS  PubMed  Google Scholar 

  14. 14

    Whiting, P., Rutjes, A. W., Reitsma, J. B., Bossuyt, P. M. & Kleijnen, J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med. Res. Methodol. 3, 25 (2003).

    PubMed  PubMed Central  Google Scholar 

  15. 15

    Nagy, Z. et al. Hippocampal pathology reflects memory deficit and brain imaging measurements in Alzheimer's disease: clinicopathologic correlations using three sets of pathologic diagnostic criteria. Dementia 7, 76–81 (1996).

    CAS  PubMed  Google Scholar 

  16. 16

    Logothetis, N. K. The neural basis of the blood-oxygen-level-dependent functional magnetic resonance imaging signal. Philos. Trans. R. Soc. Lond. B Biol. Sci. 357, 1003–1037 (2002).

    PubMed  PubMed Central  Google Scholar 

  17. 17

    Reivich, M. et al. The [18F]fluorodeoxyglucose method for the measurement of local cerebral glucose utilization in man. Circ. Res. 44, 127–137 (1979).

    CAS  PubMed  Google Scholar 

  18. 18

    Lockhart, A. et al. PIB is a non-specific imaging marker of amyloid-β (Aβ) peptide-related cerebral amyloidosis. Brain 130, 2607–2615 (2007).

    CAS  PubMed  Google Scholar 

  19. 19

    Allegri, R. F., Glaser, F. B., Taragano, F. E. & Buschke, H. Mild cognitive impairment: believe it or not? Int. Rev. Psychiatry 20, 357–363 (2008).

    PubMed  Google Scholar 

  20. 20

    Fennema-Notestine, C. et al. Structural neuroimaging in the detection and prognosis of pre-clinical and early AD. Behav. Neurol. 21, 3–12 (2009).

    PubMed  PubMed Central  Google Scholar 

  21. 21

    Giedd, J. N. et al. Reliability of cerebral measures in repeated examinations with magnetic resonance imaging. Psychiatry Res. 61, 113–119 (1995).

    CAS  PubMed  Google Scholar 

  22. 22

    Ewers, M. et al. Multicenter assessment of reliability of cranial MRI. Neurobiol. Aging 27, 1051–1059 (2006).

    CAS  PubMed  Google Scholar 

  23. 23

    Bokde, A. L. et al. Decreased activation along the dorsal visual pathway after a 3-month treatment with galantamine in mild Alzheimer disease: a functional magnetic resonance imaging study. J. Clin. Psychopharmacol. 29, 147–156 (2009).

    CAS  PubMed  Google Scholar 

  24. 24

    Moffett, J. R., Ross, B., Arun, P., Madhavarao, C. N. & Namboodiri, A. M. N-Acetylaspartate in the CNS: from neurodiagnostics to neurobiology. Prog. Neurobiol. 81, 89–131 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  25. 25

    Kantarci, K. 1H magnetic resonance spectroscopy in dementia. Br. J. Radiol. 80, S146–S152 (2007).

    CAS  PubMed  Google Scholar 

  26. 26

    Alexander, G. E., Chen, K., Pietrini, P., Rapoport, S. I. & Reiman, E. M. Longitudinal PET evaluation of cerebral metabolic decline in dementia: a potential outcome measure in Alzheimer's disease treatment studies. Am. J. Psychiatry 159, 738–745 (2002).

    PubMed  Google Scholar 

  27. 27

    Hirono, N., Hashimoto, M., Ishii, K., Kazui, H. & Mori, E. One-year change in cerebral glucose metabolism in patients with Alzheimer's disease. J. Neuropsychiatry Clin. Neurosci. 16, 488–492 (2004).

    CAS  PubMed  Google Scholar 

  28. 28

    Heiss, W. D., Kessler, J., Mielke, R., Szelies, B. & Herholz, K. Long-term effects of phosphatidylserine, pyritinol, and cognitive training in Alzheimer's disease. A neuropsychological, EEG, and PET investigation. Dementia 5, 88–98 (1994).

    CAS  PubMed  Google Scholar 

  29. 29

    Mega, M. S. et al. Metabolic patterns associated with the clinical response to galantamine therapy: a fludeoxyglucose F18 positron emission tomographic study. Arch. Neurol. 62, 721–728 (2005).

    PubMed  Google Scholar 

  30. 30

    Stefanova, E. et al. Longitudinal PET evaluation of cerebral glucose metabolism in rivastigmine treated patients with mild Alzheimer's disease. J. Neural Transm. 113, 205–218 (2006).

    CAS  PubMed  Google Scholar 

  31. 31

    Kadir, A. et al. Effect of phenserine treatment on brain functional activity and amyloid in Alzheimer's disease. Ann. Neurol. 63, 621–631 (2008).

    CAS  PubMed  Google Scholar 

  32. 32

    Klunk, W. E. et al. Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B. Ann. Neurol. 55, 306–319 (2004).

    CAS  PubMed  Google Scholar 

  33. 33

    Jack, C. R. Jr et al. 11C PiB and structural MRI provide complementary information in imaging of Alzheimer's disease and amnestic mild cognitive impairment. Brain 131, 665–680 (2008).

    PubMed  PubMed Central  Google Scholar 

  34. 34

    Maeda, J. et al. Longitudinal, quantitative assessment of amyloid, neuroinflammation, and anti-amyloid treatment in a living mouse model of Alzheimer's disease enabled by positron emission tomography. J. Neurosci. 27, 10957–10968 (2007).

    CAS  PubMed  Google Scholar 

  35. 35

    Villemagne, V. L. et al. Aβ deposits in older non-demented individuals with cognitive decline are indicative of preclinical Alzheimer's disease. Neuropsychologia 46, 1688–1697 (2008).

    CAS  PubMed  Google Scholar 

  36. 36

    Forsberg, A. et al. PET imaging of amyloid deposition in patients with mild cognitive impairment. Neurobiol. Aging 29, 1456–1465, (2008).

    CAS  PubMed  Google Scholar 

  37. 37

    Rowe, C. C. et al. Imaging of amyloid beta in Alzheimer's disease with 18F-BAY94-9172, a novel PET tracer: proof of mechanism. Lancet Neurol. 7, 129–135 (2008).

    CAS  PubMed  Google Scholar 

  38. 38

    Bobinski, M. et al. The histological validation of post mortem magnetic resonance imaging-determined hippocampal volume in Alzheimer's disease. Neuroscience 95, 721–725 (2000).

    CAS  PubMed  Google Scholar 

  39. 39

    Vellas, B. et al. Long-term follow-up of patients immunized with AN1792: reduced functional decline in antibody responders. Curr. Alzheimer Res. 6, 144–151 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  40. 40

    Saumier, D. et al. Lessons learned in the use of volumetric MRI in therapeutic trials in Alzheimer's disease: the ALZHEMED (Tramiprosate) experience. J. Nutr. Health Aging 13, 370–372 (2009).

    CAS  PubMed  Google Scholar 

  41. 41

    Harold, D. et al. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nature Genet. 41, 1088–1093 (2009).

    CAS  PubMed  Google Scholar 

  42. 42

    Lambert, J. C. et al. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease. Nature Genet. 41, 1094–1099 (2009).

    CAS  PubMed  Google Scholar 

  43. 43

    Klunk, W. E. et al. Imaging the pathology of Alzheimer's disease: amyloid-imaging with positron emission tomography. Neuroimaging Clin. N. Am. 13, 781–789 (2003).

    PubMed  Google Scholar 

  44. 44

    Chen, K. et al. Correlations between apolipoprotein E ɛ4 gene dose and whole brain atrophy rates. Am. J. Psychiatry 164, 916–921 (2007).

    PubMed  Google Scholar 

  45. 45

    Scahill, R. I., Schott, J. M., Stevens, J. M., Rossor, M. N. & Fox, N. C. Mapping the evolution of regional atrophy in Alzheimer's disease: unbiased analysis of fluid-registered serial MRI. Proc. Natl Acad. Sci. USA 99, 4703–4707 (2002).

    CAS  PubMed  Google Scholar 

  46. 46

    Scheuner, D. et al. Secreted amyloid β-protein similar to that in the senile plaques of Alzheimer's disease is increased in vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer's disease. Nature Med. 2, 864–870 (1996).

    CAS  PubMed  Google Scholar 

  47. 47

    Myers, A. J. et al. A survey of genetic human cortical gene expression. Nature Genet. 39, 1494–1499 (2007).

    CAS  PubMed  Google Scholar 

  48. 48

    Kauwe, J. S. et al. Alzheimer's disease risk variants show association with cerebrospinal fluid amyloid β. Neurogenetics 10, 13–17 (2009).

    CAS  PubMed  Google Scholar 

  49. 49

    Kauwe, J. S. et al. Variation in MAPT is associated with cerebrospinal fluid tau levels in the presence of amyloid-β deposition. Proc. Natl Acad. Sci. USA 105, 8050–8054 (2008).

    CAS  PubMed  Google Scholar 

  50. 50

    Melzer, D. et al. A genome-wide association study identifies protein quantitative trait loci (pQTLs). PLoS Genet. 4, e1000072 (2008).

    PubMed  PubMed Central  Google Scholar 

  51. 51

    Makeeva, O., Stepanov, V., Puzyrev, V., Goldstein, D. B. & Grossman, I. Global pharmacogenetics: genetic substructure of Eurasian populations and its effect on variants of drug-metabolizing enzymes. Pharmacogenomics 9, 847–868 (2008).

    CAS  PubMed  Google Scholar 

  52. 52

    Farrer, L. A. et al. Statement on use of apolipoprotein E testing for Alzheimer disease. American College of Medical Genetics/American Society of Human Genetics Working Group on ApoE and Alzheimer disease. JAMA 274, 1627–1629 (1995).

    Google Scholar 

  53. 53

    Green, R. C. et al. Disclosure of APOE genotype for risk of Alzheimer's disease. N. Engl. J. Med. 361, 245–254 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  54. 54

    Herukka, S. K. et al. CSF Aβ42, Tau and phosphorylated Tau, APOE ɛ4 allele and MCI type in progressive MCI. Neurobiol. Aging 28, 507–514 (2007).

    CAS  PubMed  Google Scholar 

  55. 55

    Ray, S. et al. Classification and prediction of clinical Alzheimer's diagnosis based on plasma signaling proteins. Nature Med. 13, 1359–1362 (2007).

    CAS  PubMed  Google Scholar 

  56. 56

    Irizarry, M. C. Biomarkers of Alzheimer disease in plasma. NeuroRx 1, 226–234 (2004).

    PubMed  PubMed Central  Google Scholar 

  57. 57

    Vanderstichele, H. et al. Standardization of measurement of beta-amyloid(1–42) in cerebrospinal fluid and plasma. Amyloid 7, 245–258 (2000).

    CAS  PubMed  Google Scholar 

  58. 58

    Peskind, E. R. et al. Safety and acceptability of the research lumbar puncture. Alzheimer Dis. Assoc. Disord. 19, 220–225 (2005).

    PubMed  Google Scholar 

  59. 59

    Blennow, K. Cerebrospinal fluid protein biomarkers for Alzheimer's disease. NeuroRx 1, 213–225 (2004).

    PubMed  PubMed Central  Google Scholar 

  60. 60

    Strozyk, D., Blennow, K., White, L. R. & Launer, L. J. CSF Aβ 42 levels correlate with amyloid-neuropathology in a population-based autopsy study. Neurology 60, 652–656 (2003).

    CAS  PubMed  Google Scholar 

  61. 61

    Fagan, A. M. et al. Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Aβ42 in humans. Ann. Neurol. 59, 512–519 (2006).

    CAS  PubMed  Google Scholar 

  62. 62

    Buerger, K. et al. CSF phosphorylated tau protein correlates with neocortical neurofibrillary pathology in Alzheimer's disease. Brain 129, 3035–3041 (2006).

    PubMed  Google Scholar 

  63. 63

    Shaw, L. M. et al. Cerebrospinal fluid biomarker signature in Alzheimer's disease neuroimaging initiative subjects. Ann. Neurol. 65, 403–413, (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  64. 64

    Visser, P. J. et al. Prevalence and prognostic value of CSF markers of Alzheimer's disease pathology in patients with subjective cognitive impairment or mild cognitive impairment in the DESCRIPA study: a prospective cohort study. Lancet Neurol. 8, 619–627, (2009).

    PubMed  Google Scholar 

  65. 65

    Mattsson, N. et al. CSF biomarkers and incipient Alzheimer disease in patients with mild cognitive impairment. JAMA 302, 385–393, (2009).

    CAS  PubMed  Google Scholar 

  66. 66

    Andreasson, U., Portelius, E., Andersson, M. E., Blennow, K. & Zetterberg, H. Aspects of β-amyloid as a biomarker for Alzheimer's disease. Biomarkers Med. 1, 59–78 (2007).

    CAS  Google Scholar 

  67. 67

    de Jong, D., Kremer, B. P., Olde Rikkert, M. G. & Verbeek, M. M. Current state and future directions of neurochemical biomarkers for Alzheimer's disease. Clin. Chem. Lab. Med. 45, 1421–1434 (2007).

    CAS  PubMed  Google Scholar 

  68. 68

    Zhong, Z. et al. Levels of β-secretase (BACE1) in cerebrospinal fluid as a predictor of risk in mild cognitive impairment. Arch. Gen. Psychiatry 64, 718–726 (2007).

    CAS  PubMed  Google Scholar 

  69. 69

    Zetterberg, H. et al. Elevated cerebrospinal fluid BACE1 activity in incipient Alzheimer disease. Arch. Neurol. 65, 1102–1107 (2008).

    PubMed  Google Scholar 

  70. 70

    Davidsson, P. et al. Differential increase in cerebrospinal fluid-acetylcholinesterase after treatment with acetylcholinesterase inhibitors in patients with Alzheimer's disease. Neurosci. Lett. 300, 157–160 (2001).

    CAS  PubMed  Google Scholar 

  71. 71

    von der Kammer, H. et al. Muscarinic acetylcholine receptors activate expression of the EGR gene family of transcription factors. J. Biol. Chem. 273, 14538–14544 (1998).

    CAS  PubMed  Google Scholar 

  72. 72

    Blennow, K. et al. Longitudinal stability of CSF biomarkers in Alzheimer's disease. Neurosci. Lett. 419, 18–22 (2007).

    CAS  PubMed  Google Scholar 

  73. 73

    Zetterberg, H. et al. Intra-individual stability of CSF biomarkers for Alzheimer's disease over two years. J. Alzheimers Dis. 12, 255–260 (2007).

    CAS  PubMed  Google Scholar 

  74. 74

    Lanz, T. A., Hosley, J. D., Adams, W. J. & Merchant, K. M. Studies of Aβ pharmacodynamics in the brain, cerebrospinal fluid, and plasma in young (plaque-free) Tg2576 mice using the γ-secretase inhibitor N2-[(2S)-2-(3,5-difluorophenyl)-2-hydroxyethanoyl]-N1-[(7S)-5-methyl-6-oxo-6,7-dihydro-5H-dibenzo[b, d]azepin-7-yl]-L-alaninamide (LY-411575). J. Pharmacol. Exp. Ther. 309, 49–55 (2004).

    CAS  PubMed  Google Scholar 

  75. 75

    Anderson, J. J. et al. Reductions in β-amyloid concentrations in vivo by the γ-secretase inhibitors BMS-289948 and BMS-299897. Biochem. Pharmacol. 69, 689–698 (2005).

    CAS  PubMed  Google Scholar 

  76. 76

    Fleisher, A. S. et al. Phase 2 safety trial targeting amyloid beta production with a gamma-secretase inhibitor in Alzheimer disease. Arch. Neurol. 65, 1031–1038 (2008).

    PubMed  PubMed Central  Google Scholar 

  77. 77

    Sankaranarayanan, S. et al. First demonstration of cerebrospinal fluid and plasma Aβ lowering with oral administration of a β-site amyloid precursor protein-cleaving enzyme 1 inhibitor in nonhuman primates. J. Pharmacol. Exp. Ther. 328, 131–140 (2009).

    CAS  PubMed  Google Scholar 

  78. 78

    Adlard, P. A. et al. Rapid restoration of cognition in Alzheimer's transgenic mice with 8-hydroxy quinoline analogs is associated with decreased interstitial Aβ. Neuron 59, 43–55 (2008).

    CAS  PubMed  Google Scholar 

  79. 79

    Lannfelt, L. et al. Safety, efficacy, and biomarker findings of PBT2 in targeting Aβ as a modifying therapy for Alzheimer's disease: a phase IIa, double-blind, randomised, placebo-controlled trial. Lancet Neurol. 7, 779–786 (2008).

    CAS  PubMed  Google Scholar 

  80. 80

    Hesse, C. et al. Transient increase in total tau but not phospho-tau in human cerebrospinal fluid after acute stroke. Neurosci. Lett. 297, 187–190 (2001).

    CAS  PubMed  Google Scholar 

  81. 81

    Zetterberg, H. et al. Neurochemical aftermath of amateur boxing. Arch. Neurol. 63, 1277–1280 (2006).

    PubMed  Google Scholar 

  82. 82

    Gilman, S. et al. Clinical effects of Aβ immunization (AN1792) in patients with AD in an interrupted trial. Neurology 64, 1553–1562 (2005).

    CAS  PubMed  Google Scholar 

  83. 83

    Degerman Gunnarsson, M., Kilander, L., Basun, H. & Lannfelt, L. Reduction of phosphorylated tau during memantine treatment of Alzheimer's disease. Dement. Geriatr. Cogn. Disord. 24, 247–252 (2007).

    CAS  PubMed  Google Scholar 

  84. 84

    Slamon, D. J. et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 244, 707–712 (1989).

    CAS  PubMed  Google Scholar 

  85. 85

    Dubois, B. et al. Research criteria for the diagnosis of Alzheimer's disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 6, 734–746 (2007).

    PubMed  Google Scholar 

  86. 86

    Carrillo, M. C. et al. Early risk assessment for Alzheimer's disease. Alzheimers Dement. 5, 182–196 (2009).

    CAS  PubMed  Google Scholar 

  87. 87

    Coats, M. & Morris, J. C. Antecedent biomarkers of Alzheimer's disease: the adult children study. J. Geriatr. Psychiatry Neurol. 18, 242–244 (2005).

    PubMed  Google Scholar 

  88. 88

    Mintun, M. A. et al. [11C]PIB in a nondemented population: potential antecedent marker of Alzheimer disease. Neurology 67, 446–452 (2006).

    CAS  PubMed  Google Scholar 

  89. 89

    Mani, R. B. The evaluation of disease modifying therapies in Alzheimer's disease: a regulatory viewpoint. Stat. Med. 23, 305–314 (2004).

    PubMed  Google Scholar 

  90. 90

    Pepe, M. S. et al. Phases of biomarker development for early detection of cancer. J. Natl Cancer Inst. 93, 1054–1061 (2001).

    CAS  PubMed  Google Scholar 

  91. 91

    Leber, P. Slowing the progression of Alzheimer disease: methodologic issues. Alzheimer Dis. Assoc. Disord. 11 (Suppl. 5), 10–21; discussion 37–19 (1997).

    Google Scholar 

  92. 92

    Woodcock, J. & Woosley, R. The FDA critical path initiative and its influence on new drug development. Annu. Rev. Med. 59, 1–12 (2008).

    CAS  PubMed  Google Scholar 

  93. 93

    Katz, R. Biomarkers and surrogate markers: an FDA perspective. NeuroRx 1, 189–195 (2004).

    PubMed  PubMed Central  Google Scholar 

  94. 94

    Temple, R. Are surrogate markers adequate to assess cardiovascular disease drugs? JAMA 282, 790–795 (1999).

    CAS  PubMed  Google Scholar 

  95. 95

    Baker, S. G. & Kramer, B. S. A perfect correlate does not a surrogate make. BMC Med. Res. Methodol. 3, 16 (2003).

    PubMed  PubMed Central  Google Scholar 

  96. 96

    Fleming, T. R. & DeMets, D. L. Surrogate end points in clinical trials: are we being misled? Ann. Intern. Med. 125, 605–613 (1996).

    CAS  PubMed  Google Scholar 

  97. 97

    Fox, N. C. et al. Effects of Aβeta immunization (AN1792) on MRI measures of cerebral volume in Alzheimer disease. Neurology 64, 1563–1572 (2005).

    CAS  PubMed  Google Scholar 

  98. 98

    Jack, C. R. Jr et al. The Alzheimer's Disease Neuroimaging Initiative (ADNI): MRI methods. J. Magn. Reson. Imaging 27, 685–691 (2008).

    PubMed  PubMed Central  Google Scholar 

  99. 99

    Mueller, S. G. et al. Ways toward an early diagnosis in Alzheimer's disease: The Alzheimer's Disease Neuroimaging Initiative (ADNI). Alzheimers Dement. 1, 55–66 (2005).

    PubMed  PubMed Central  Google Scholar 

  100. 100

    Williams, S. A., Slavin, D. E., Wagner, J. A. & Webster, C. J. A cost-effectiveness approach to the qualification and acceptance of biomarkers. Nature Rev. Drug Discov. 5, 897–902 (2006).

    CAS  Google Scholar 

  101. 101

    Gutman, S. & Kessler, L. G. The US Food and Drug Administration perspective on cancer biomarker development. Nature Rev. Cancer 6, 565–571 (2006).

    CAS  Google Scholar 

  102. 102

    Grundman, M. & Black, R. O3-04-05: clinical trials of bapineuzumab, a β-amyloid-targeted immunotherapy in patients with mild to moderate Alzheimer's disease. Alzheimers Dement. 4, T166 (2008).

    Google Scholar 

  103. 103

    Sanhai, W. R., Spiegel, J. & Ferrari, M. A critical path approach to advance nanoengineered medical products. Drug Discov. Today Technol. 4, 35–41 (2007).

    PubMed  Google Scholar 

  104. 104

    Ridha, B. H. et al. Application of automated medial temporal lobe atrophy scale to Alzheimer disease. Arch. Neurol. 64, 849–854 (2007).

    PubMed  Google Scholar 

  105. 105

    Byrum, C. E. et al. Accuracy and reproducibility of brain and tissue volumes using a magnetic resonance segmentation method. Psychiatry Res. 67, 215–234 (1996).

    CAS  PubMed  Google Scholar 

  106. 106

    Jack, C. R. Jr et al. Medial temporal atrophy on MRI in normal aging and very mild Alzheimer's disease. Neurology 49, 786–794 (1997).

    PubMed  PubMed Central  Google Scholar 

  107. 107

    Wang, P. N., Lirng, J. F., Lin, K. N., Chang, F. C. & Liu, H. C. Prediction of Alzheimer's disease in mild cognitive impairment: a prospective study in Taiwan. Neurobiol. Aging 27, 1797–1806, (2006).

    CAS  PubMed  Google Scholar 

  108. 108

    Fox, N. C., Cousens, S., Scahill, R., Harvey, R. J. & Rossor, M. N. Using serial registered brain magnetic resonance imaging to measure disease progression in Alzheimer disease: power calculations and estimates of sample size to detect treatment effects. Arch. Neurol. 57, 339–344 (2000).

    CAS  PubMed  Google Scholar 

  109. 109

    Loubinoux, I. et al. Within-session and between-session reproducibility of cerebral sensorimotor activation: a test–retest effect evidenced with functional magnetic resonance imaging. J. Cereb. Blood Flow Metab. 21, 592–607 (2001).

    CAS  PubMed  Google Scholar 

  110. 110

    Kircher, T. T., Erb, M., Grodd, W. & Leube, D. T. Cortical activation during cholinesterase-inhibitor treatment in Alzheimer disease: preliminary findings from a pharmaco-fMRI study. Am. J. Geriatr. Psychiatry 13, 1006–1013 (2005).

    PubMed  Google Scholar 

  111. 111

    Bokde, A. L. W. et al. Decreased activation along the dorsal visual pathway after a 3-month treatment with galantamine in mild Alzheimer disease. J. Clin. Psychopharmacol. 29, 147–156 (2009).

    CAS  PubMed  Google Scholar 

  112. 112

    Goekoop, R., Scheltens, P., Barkhof, F. & Rombouts, S. A. Cholinergic challenge in Alzheimer patients and mild cognitive impairment differentially affects hippocampal activation — a pharmacological fMRI study. Brain 129, 141–157 (2006).

    PubMed  Google Scholar 

  113. 113

    Chao, L. L. et al. Reduced medial temporal lobe N-acetylaspartate in cognitively impaired but nondemented patients. Neurology 64, 282–289 (2005).

    CAS  PubMed  PubMed Central  Google Scholar 

  114. 114

    Jessen, F. et al. Treatment monitoring and response prediction with proton MR spectroscopy in AD. Neurology 67, 528–530 (2006).

    CAS  PubMed  Google Scholar 

  115. 115

    Herholz, K. et al. Comparability of FDG PET studies in probable Alzheimer's disease. J. Nucl. Med. 34, 1460–1466 (1993).

    CAS  PubMed  Google Scholar 

  116. 116

    Edison, P. et al. Amyloid, hypometabolism, and cognition in Alzheimer disease: an [11C]PIB and [18F]FDG PET study. Neurology 68, 501–508 (2007).

    CAS  PubMed  Google Scholar 

  117. 117

    Heiss, W. D. et al. Effect of piracetam on cerebral glucose metabolism in Alzheimer's disease as measured by positron emission tomography. J. Cereb. Blood Flow Metab. 8, 613–617 (1988).

    CAS  PubMed  Google Scholar 

  118. 118

    Teipel, S. J. et al. Effects of donepezil on cortical metabolic response to activation during 18FDG-PET in Alzheimer's disease: a double-blind cross-over trial. Psychopharmacology (Berl.) 187, 86–94 (2006).

    CAS  Google Scholar 

  119. 119

    Engler, H. et al. Two-year follow-up of amyloid deposition in patients with Alzheimer's disease. Brain 129, 2856–2866 (2006).

    PubMed  Google Scholar 

  120. 120

    Bohnen, N. I. et al. Degree of inhibition of cortical acetylcholinesterase activity and cognitive effects by donepezil treatment in Alzheimer's disease. J. Neurol. Neurosurg. Psychiatry 76, 315–319 (2005).

    CAS  PubMed  PubMed Central  Google Scholar 

  121. 121

    Andreasen, N. et al. Cerebrospinal fluid β-amyloid(1–42) in Alzheimer disease: differences between early- and late-onset Alzheimer disease and stability during the course of disease. Arch. Neurol. 56, 673–680 (1999).

    CAS  PubMed  Google Scholar 

  122. 122

    Olsson, A. et al. Simultaneous measurement of β-amyloid(1–42), total tau, and phosphorylated tau (Thr181) in cerebrospinal fluid by the xMAP technology. Clin. Chem. 51, 336–345 (2005).

    CAS  PubMed  Google Scholar 

  123. 123

    Fukuyama, R. et al. Age-dependent change in the levels of Aβ40 and Aβ42 in cerebrospinal fluid from control subjects, and a decrease in the ratio of Aβ42 to Aβ40 level in cerebrospinal fluid from Alzheimer's disease patients. Eur. Neurol. 43, 155–160 (2000).

    CAS  PubMed  Google Scholar 

  124. 124

    Kanai, M. et al. Longitudinal study of cerebrospinal fluid levels of tau, Aβ1–40, and Aβeta1–42(43) in Alzheimer's disease: a study in Japan. Ann. Neurol. 44, 17–26 (1998).

    CAS  PubMed  Google Scholar 

  125. 125

    Lewczuk, P. et al. Neurochemical diagnosis of Alzheimer's dementia by CSF Aβ42, Aβ42/Aβ40 ratio and total tau. Neurobiol. Aging 25, 273–281 (2004).

    CAS  PubMed  Google Scholar 

  126. 126

    Hansson, O. et al. Prediction of Alzheimer's disease using the CSF Aβ42/Aβ40 ratio in patients with mild cognitive impairment. Dement. Geriatr. Cogn. Disord. 23, 316–320 (2007).

    CAS  PubMed  Google Scholar 

  127. 127

    Shoji, M. et al. Combination assay of CSF tau, Aβ1–40 and Aβ1–42(43) as a biochemical marker of Alzheimer's disease. J. Neurol. Sci. 158, 134–140 (1998).

    CAS  PubMed  Google Scholar 

  128. 128

    de Leon, M. J. et al. Longitudinal cerebrospinal fluid tau load increases in mild cognitive impairment. Neurosci. Lett. 333, 183–186 (2002).

    CAS  PubMed  Google Scholar 

  129. 129

    de Leon, M. J. et al. Longitudinal CSF and MRI biomarkers improve the diagnosis of mild cognitive impairment. Neurobiol. Aging 27, 394–401 (2006).

    CAS  PubMed  Google Scholar 

  130. 130

    Olsson, A. et al. Measurement of α- and β-secretase cleaved amyloid precursor protein in cerebrospinal fluid from Alzheimer patients. Exp. Neurol. 183, 74–80 (2003).

    CAS  PubMed  Google Scholar 

  131. 131

    Verheijen, J. H. et al. Detection of a soluble form of BACE-1 in human cerebrospinal fluid by a sensitive activity assay. Clin. Chem. 52, 1168–1174 (2006).

    CAS  PubMed  Google Scholar 

  132. 132

    Holsinger, R. M., Lee, J. S., Boyd, A., Masters, C. L. & Collins, S. J. CSF BACE1 activity is increased in CJD and Alzheimer disease versus [corrected] other dementias. Neurology 67, 710–712 (2006).

    CAS  PubMed  Google Scholar 

  133. 133

    Holsinger, R. M., McLean, C. A., Collins, S. J., Masters, C. L. & Evin, G. Increased beta-secretase activity in cerebrospinal fluid of Alzheimer's disease subjects. Ann. Neurol. 55, 898–899 (2004).

    PubMed  Google Scholar 

  134. 134

    Georganopoulou, D. G. et al. Nanoparticle-based detection in cerebral spinal fluid of a soluble pathogenic biomarker for Alzheimer's disease. Proc. Natl Acad. Sci. USA 102, 2273–2276 (2005).

    CAS  PubMed  Google Scholar 

  135. 135

    Vanderstichele, H. et al. Analytical performance and clinical utility of the INNOTEST PHOSPHO-TAU(181P) assay for discrimination between Alzheimer's disease and dementia with Lewy bodies. Clin. Chem. Lab. Med. 44, 1472–1480 (2006).

    CAS  PubMed  Google Scholar 

  136. 136

    Vanmechelen, E. et al. Quantification of tau phosphorylated at threonine 181 in human cerebrospinal fluid: a sandwich ELISA with a synthetic phosphopeptide for standardization. Neurosci. Lett. 285, 49–52, (2000).

    CAS  PubMed  Google Scholar 

  137. 137

    Kohnken, R. et al. Detection of tau phosphorylated at threonine 231 in cerebrospinal fluid of Alzheimer's disease patients. Neurosci. Lett. 287, 187–190 (2000).

    CAS  PubMed  Google Scholar 

  138. 138

    Hampel, H. et al. Measurement of phosphorylated tau epitopes in the differential diagnosis of Alzheimer disease: a comparative cerebrospinal fluid study. Arch. Gen. Psychiatry 61, 95–102 (2004).

    CAS  PubMed  Google Scholar 

Download references

Acknowledgements

The authors would like to thank K. Duggan for technical assistance. H.H. acknowledges support by the Science Foundation Ireland Investigator Neuroimaging Grant Programme (08/IN.1/B1846).

Author information

Affiliations

Authors

Corresponding author

Correspondence to Harald Hampel.

Ethics declarations

Competing interests

Harald Hampel has participated on the Advisory Board of BRAHMS, Henningsdorf, Germany, and has received research support for his institutions from the same company.

Richard Frank is an employee of GE Healthcare.

John Hardy is a consultant for Eisai and MerckSerono.

Kaj Blennow has participated on the Advisory Board for Innogenetics NV, Ghent, Belgium, and has received research support to his institution from the same company.

Karl Broich, Stefan J. Teipel, Russel G. Katz, Karl Herholz, Arun L.W. Bokde, Frank Jessen, Yvonne C. Hoessler, Wendy R. Sanhai, Henrik Zetterberg and Janet Woodcock declare no competing financial interests.

Supplementary information

Supplementary information S1 (table)

Imaging biomarkers for Alzheimer's disease (PDF 565 kb)

Related links

Related links

DATABASES

OMIM

Alzheimer's disease

FURTHER INFORMATION

ADNI

ADNI clinical trial

Alzheimer's Association Research Roundtable

Guidance for Industry Qualifying for Pediatric Exclusivity Under Section 505A of the Federal Food, Drug, and Cosmetic Act

Improved Predictivity of Efficacy Evaluation — Brain Disorders

Improved Predictivity of Efficacy Evaluation — Brain Disorders

The Biomarkers Consortium

The Critical Path Initiative

Glossary

Amyloid-β

(Aβ). An aggregation-prone peptide derived from the amyloid precursor protein. The 42 amino acid isoform of the peptide is the main component of plaques in Alzeimer's disease.

Tau protein

A microtubule-associated protein located in the neuronal axons. Hyperphosphorylated tau is the main component of neurofibrillary tangles in Alzheimer's disease.

DSM-IV

(Diagnostic and Statistical Manual of Mental Disorders, Fourth edition). This manual outlines diagnostic criteria for psychiatric disorders and is published by the American Psychiatric Association.

ICD-10

(International Statistical Classification of Diseases, Tenth edition). This book outlines diagnostic criteria for human diseases and is published by the World Health Organization.

NINCDS-ADRDA

(The National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association). This outlines criteria for diagnosing Alzheimer's disease.

Primary end point

A primary end point is defined as the single main question to be answered in a given clinical trial.

Biomarker

An objective measure of a biological or pathogenic process that can be used to evaluate disease risk or prognosis. It can be used to guide clinical diagnosis or to monitor therapeutic interventions.

Surrogate end point

A substitute for a clinical end point in a clinical trial.

Secondary end point

An end point in a trial that provides additional characterization of treatment effect, but is not sufficient by itself to fully characterize the benefit or to support a claim for a treatment effect.

Longitudinal study

A research study with repeated observations of the same patients over long periods of time.

Phantom test

A plastic cylinder with standardized measures and density used for calibration of magnetic resonance imaging devices.

Transformation map

A spatially extended vector field that describes the spatial warps that are needed to align a three-dimensional image of the brain into a common standard space.

Classical plaque

A dense aggregation of amyloid-β protein with classical amyloid characteristics, which is surrounded by swollen neuritis and reactive glial cells.

Diffuse plaque

An aggregation of amyloid-β protein that can only be detected using immunohistochemistry.

AN1792 vaccination trial

The first clinical trial on active amyloid-β immunotherapy in Alzheimer's disease.

Huntington's protocol

Individuals who have family members with Huntington's disease are offered genetic counselling, which includes the possibility of presymptomatic DNA testing. However, this is only offered after a series of careful counselling sessions to ensure that the individual understands the issues involved in knowing their genetic status.

TOMM40

A gene encoding a mitochondrial protein that is located on chromosome 19 adjacent to the apolipoprotein E gene.

Risk chart

A graph illustrating the risk of developing Alzheimer's disease by age. This risk is modified by genetic status, especially by apolipoprotein E genotype status. Other genes will have a smaller effect on this chart (except in families with amyloid precursor protein and presenilin mutations).

Phase 0 trial

An exploratory first-in-human trial with single subtherapeutic drug doses and small numbers of subjects to provide first data on drug pharmacokinetics and pharmacodynamics.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hampel, H., Frank, R., Broich, K. et al. Biomarkers for Alzheimer's disease: academic, industry and regulatory perspectives. Nat Rev Drug Discov 9, 560–574 (2010). https://doi.org/10.1038/nrd3115

Download citation

Further reading

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing