Biomarkers in Brain Disease: Biological and Regulatory Challenges

Web Sites

Alzheimer's Association
The U.S. Alzheimer's Association Web site offers information for researchers about grant opportunities, its research membership society, and access to the journal Alzheimer's and Dementia.

Alzheimer's Disease Neuroimaging Initiative
Study protocols and information about applying for sample access can be found at the Alzheimer's disease Neuroimaging Initiative's scientific home page.

Alzheimer's Society
The Alzheimer's Society provides general information about Alzheimer's disease in the UK, descriptions of research programs and grant opportunities and access to the Dementia Catalogue of the Alzheimer's Society library.

Cambridge Centre for Neuropsychiatric Research
The Cambridge Centre for Neuropsychiatric Research describes research in the Bahn and Lowe laboratories in biomarkers for diseases such as schizophrenia and bipolar disorder, including recent publications and an overview of research.

CANTAB
A thorough description of CANTAB, the cognitive test battery for Alzheimer's, is offered by test developer Cambridge Cognition.

European Alzheimer's Disease Consortium
The EADC is a fully functional network of European centres of excellence working in the field of Alzheimer's disease. It provides a setting in which to increase the basic scientific understanding of and to develop ways to prevent, slow, or ameliorate the primary and secondary symptoms of Alzheimer's disease. This is done by facilitating large Europe wide research studies. The EADC is funded by the European Commission and as such enjoys the privilege of complete independence and autonomy from the pharmaceutical industry while maintaining close working links with it.

Global Medical Excellence Cluster, UK (GMEC)
GMEC, the largest Global Medical Excellence Cluster in Europe, brings together in a collaborative effort, leading universities, pharmaceutical and medical device companies, and hospitals in the UK. The goal is to improve innovation, product development, and leading-edge research activity; create employment opportunities; and enhance patient outcomes.

Innovative Medicines Initiative
The Innovative Medicines Initiative, a partnership between the European Community and the European Federation of Pharmaceutical Industries and Associations, supports AddNeuroMed, a collaborative effort to identify diagnostic and predictive biomarkers for Alzheimer's disease.

National Cancer Institute
In oncology, the REMARK guidelines were written to improve biomarker studies, offering standards for study design, methodology, and statistical analyses.

Articles

Keynote Address: William Potter

Bieck PR, Potter WZ. 2005. Biomarkers in psychotropic drug development: integration of data across multiple domains. Annu. Rev. Pharmacol. Toxicol. 45: 227-246.

Gelenberg AJ, Thase ME, Meyer RE, et al. 2008. The history and current state of antidepressant clinical trial design: a call to action for proof-of-concept studies. J. Clin. Psychiatry 69: 1513-1528.

Orest Hurko

Pangalos MN, Schechter LE, Hurko O. 2007. Drug development for CNS disorders: strategies for balancing risk and reducing attrition. Nat. Rev. Drug Discov. 6: 521-532.

Tobert JA. 2003. Lovastatin and beyond: the history of the HMG-CoA reductase inhibitors. Nat. Rev. Drug Disc. 2: 517-526.

Carol Brayne

Brayne C, Fox C, Boustani M. 2007. Dementia screening in primary care: is it time? JAMA 298: 2409-2411.

Brayne C. 2007. The elephant in the room — healthy brains in later life, epidemiology and public health. Nat. Rev. Neurosci. 8: 233-239.

Jeremy Nicholson

Holmes E, Wilson ID, Nicholson JK. 2008. Metabolic phenotyping in health and disease. Cell 134: 714-717.

Jia W, Li H, Zhao L, Nicholson JK. 2008. Gut microbiota: a potential new territory for drug targeting. Nat. Rev. Drug Discov. 7: 123-129.

Nicholson JK, Lindon JC. 2008. Systems biology: Metabonomics. Nature 455: 1054-1056.

Nick Fox

Ridha BH, Anderson VM, Barnes J, et al. 2008. Volumetric MRI and cognitive measures in Alzheimer disease: Comparison of markers of progression. J. Neur. 255: 567-574.

Scahill RI, Fox NC. 2007. Longitudinal imaging in dementia. Br. J. Radiol. 80 Spec No 2: S92-S98.

Henrik Zetterberg

Portelius E, Brinkmalm G, Tran AJ, et al. 2009. Identification of Novel APP/Abeta isoforms in human cerebrospinal fluid. Neurodegener. Dis. Feb 20. [Epub ahead of print]

Simon Lovestone

Lovestone S, Francis P, Strandgaard K. 2007. Biomarkers for disease modification trials — the innovative medicines initiative and AddNeuroMed. J. Nutr. Health Aging 11 : 359-361.

Thambisetty M, Hye A, Foy C, et al. 2008. Proteome-based identification of plasma proteins associated with hippocampal metabolism in early Alzheimer's disease. J. Neurol. 255: 1712-1720.

Holly Soares

Ray S, Britschgi M, Herbert C, et al. 2007. Classification and prediction of clinical Alzheimer's diagnosis based on plasma signaling proteins. Nat. Med. 13: 1359-1362.

Roher AE, Esh CL, Kokjohn TA, et al. 2009. Amyloid beta peptides in human plasma and tissues and their significance for Alzheimer's disease. Alzheimers Dement. 5: 18-29.

Barbara Sahakian

Beddington J, Cooper CL, Field J, et al. 2008. The mental wealth of nations. Nature 455: 1057-1060.

Clark L, Sahakian BJ. 2008. Cognitive neuroscience and brain imaging in bipolar disorder. Dialogues Clin. Neurosci. 10: 153-163.

De Jager C, Blackwell AD, Budge MM, Sahakian BJ. 2005. Predicting cognitive decline in healthy older adults. Am. J. Geriatr. Psychiatry 13: 735-740.

Richard Mayeux

Dubois B, Feldman HH, Jacova C, et al. 2007. Research criteria for the diagnosis of Alzheimer's disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 6: 734-746.

Lee JH, Cheng R, Graff-Radford N, et al., National Institute on Aging Late-Onset Alzheimer's Disease Family Study Group. 2008. Analyses of the National Institute on Aging Late-Onset Alzheimer's Disease Family Study: implication of additional loci. Arch. Neurol. 65: 1518-1526.

Sabine Bahn

Huang JT, Leweke FM, et al. 2007. CSF metabolic and proteomic profiles in patients prodromal for psychosis. PLoS ONE 2: e756. Full Text

Schwarz E, Bahn S. 2008. Biomarker discovery in psychiatric disorders. Electrophoresis 29: 2884-2890.

Roger Barker

Antoniades CA, Barker RA. 2008. The search for biomarkers in Parkinson's disease: a critical review. Expert Rev. Neurother. 8: 1841-1852.

Michell AW, Goodman AO, Silva AH, et al. 2008. Hand tapping: a simple, reproducible, objective marker of motor dysfunction in Huntington's disease. J. Neurol. 255: 1145-1152.

Wayne Drevets

Drevets WC, Price JL, Furey ML. 2008. Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression. Brain Struct. Funct. 213: 93-118. Full Text

Drevets WC, Gadde KM, Krishnan, KRR. 2004. The neurobiological foundation of mental illness. In Charney D, Nestler E, eds., Neurobiology of Mental Illness. Oxford Press, Oxford.

Gavin Giovannoni

CAMMS223 Trial Investigators, Coles AJ, Compston DA, et al. 2008 Alemtuzumab vs. interferon beta-1a in early multiple sclerosis. N. Engl. J. Med. 359: 1786-1801.

Giovannoni, G. 2004. The yin and yang of inflammation in multiple sclerosis. In G. Comi, ed., Early indicators, early treatment and neuroprotection in multiple sclerosis. Springer, Rome.

Giovannoni G. 2006. Multiple sclerosis cerebrospinal fluid biomarkers. Dis. Markers 22: 187-19.

Cristina Sampaio

Sampaio C. 2007. Clinical relevance on Alzheimer's disease endpoints. J. Nutr. Health Aging 11: 316-317.

Maria Carrillo

Ivinson AJ, Lane R, May PC, et al. 2008. Partnership between academia and industry for drug discovery in Alzheimer's disease. Alzheimers Dement. 4: 80-88.

Sheila Taube

Drake TA, Braun J, Marchevsky A, et al. 2007. A system for sharing routine surgical pathology specimens across institutions: the Shared Pathology Informatics Network. Hum. Pathol. 38: 1212-1225.

Harris L, Fritsche H, Mennel R, et al. 2007. American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer. J. Clin. Oncol. 25: 5287-5312.

Taube SE, Jacobson JW, Lively TG. 2005. Cancer diagnostics: decision criteria for marker utilization in the clinic. Am. J. Pharmacogenomics 5: 357-364.

David Brooks

Brooks DJ. 2007. Assessment of Parkinson's disease with imaging. Parkinsonism Relat. Disord. 13 Suppl 3: S268-S275.

Leslie Shaw

Schuff N, Woerner N, Boreta L, et al., the Alzheimer's Disease Neuroimaging Initiative. 2009. MRI of hippocampal volume loss in early Alzheimer's disease in relation to ApoE genotype and biomarkers. Brain Feb 27. [Epub ahead of print.] Full Text

Arthur Toga

Morra JH, Tu Z, Apostolova LG, et al. 2009. Automated 3D mapping of hippocampal atrophy and its clinical correlates in 400 subjects with Alzheimer's disease, mild cognitive impairment, and elderly controls. Hum. Brain Mapp. 2009 Jan 26. [Epub ahead of print]

Bruno Vellas

Vellas B, Andrieu S, Sampaio C, Coley N, Wilcock G; European Task Force Group. 2008. Endpoints for trials in Alzheimer's disease: a European task force consensus. Lancet Neurol. 7: 436-450.

Vellas B, Coley N, Andrieu S. 2008. Disease modifying trials in Alzheimer's disease: perspectives for the future. J. Alzheimers Dis. 15: 289-301.

Winblad B, Frisoni GB, Frolich L, et al. 2008. Recommendations for future Alzheimer disease research in Europe. J. Nutr. Health Aging 12: 683-684.

Gordon Wilcock and Bruno Vellas

Andrieu S, Coley N, Aisen P, et al. 2009. Methodological issues in primary prevention trials for neurodegenerative dementia. Alzheimers Dis. 16: 235-270.

Sponsors

Silver

Bronze

• Imperial College
• Johnson & Johnson
• King's College London
• Oxford University
• The Ellison Medical Foundation
• University College London
• Wellcome Trust

Academy Friends

• Alzheimer's Association
• BIOBASE GmbH
• Eisai
• Foresight Programme, Government Office for Science
• Rules-Based Medicine, Inc.
• Schering-Plough Research Institute
• Scout Diagnostics
• The Alzheimer's Research Trust

Media Sponsors

Are there plasma biomarkers that can diagnose or track progression of AD?

How should the enormous international Alzheimer's research community standardize information sharing?

What is the best way to organize and promote international collaboration on validation trials?

Will it be possible to prevent relapses in multiple sclerosis?

How should biomarkers be combined?

Will it be possible to establish a surrogate biomarker for a brain disease?

How can we use biomarkers to reduce the failure rate of clinical trials?

Will neuroprotective agents prove to be the best way to prevent Alzheimer's disease?

Until relatively recently, biomarkers were not a popular area of investigation. By the 1980s, many believed that by individualizing treatment, subdividing disease, and explicating pathophysiology, genetics would make biomarkers unnecessary. The essential role of biomarkers as tools to understanding how drugs work and disease progresses was eclipsed by the rising star of the human genome.

Yet the need for biomarkers—simple, accessible indices of complex biological phenomena—only grew. Now their absence is crippling pharmaceutical research; drug development pipelines are drying up. Meanwhile, record numbers of experimental compounds are failing in increasingly expensive late-stage clinical trials.

Neurological and psychiatric conditions, in particular, need good biomarkers. The inaccessibility of the brain, the lack of knowledge about pathophysiology, and the chronic degenerative course of many of these diseases make it difficult to judge who has the disease, how best to treat, and whether or not experimental treatments are successful. The need for clinical biomarkers may soon become acute: the most promising candidate treatments for conditions such as Alzheimer's disease (AD) are likely to be most effective for patients in the earliest stages of disease—possibly even before the onset of symptoms.

How to move forward collectively with identifying and validating such brain-based biomarkers was the focus of Biomarkers in Brain Disease, a conference sponsored by the New York Academy of Sciences and the Global Medical Excellence Cluster of South East England and held from January 26–28, 2009, at Oxford University. In general, the meeting focused on biomarkers as used in drug development and clinical trials.

In the last few decades, researchers have proposed increasingly sophisticated hypotheses of disease for conditions ranging from depression to Parkinson's, but their ability to test them has not kept up, pointed out keynote speaker William Potter of Merck Research Laboratories. Drug development for depression is a good example: Of the more than 75 potential targets in depression, the mechanisms by which they might affect the condition have been established for only a handful. "That's the reality, and that's a field that we thought was way ahead," he cautioned.

Biomarkers are shining a light on the brain processes involved in these diseases.

But there are signs of progress. Disease-related profiles in cerebrospinal fluid (CSF) have been established for Alzheimer's disease, and much attention is focused on doing the same in plasma. Imaging techniques such as positron emission tomography (PET), magnetic resonance imaging (MRI), and, more recently, diffusion tensor imaging (DTI) are becoming useful adjuncts for the diagnosis and monitoring of neurodegenerative disease and for the testing of antidepressant candidate drugs. Proteomics and metabolomics will no doubt identify characteristic biochemical signatures of physiological processes. None of these techniques are yet ready to answer the biggest questions, said Potter, but we're getting closer. "It's not too strong to say that biomarkers are shining a light on the brain processes involved in these diseases," he said. After years of neglect, we can now look forward to an era in which the biomarker finally comes of age.

Biomarkers: a typology

Most broadly, a biomarker is a quantifiable measure correlated with or predictive of a physiological process involved in health or disease. "Biomarker" means different things in different contexts, so one of the most important challenges in biomarker development is conceptual rather than scientific: researchers must be absolutely clear about what they're looking for and what they plan to do with the information.

Even at this conference, speakers did not agree upon one typology to categorize biomarkers, but rather proposed different categories based on different criteria. Potter distinguished two types of biomarkers. For neurologists, a biomarker is a quantifiable difference in brain tissue or CSF associated with the course or severity of symptoms, he said. But for a drug developer, a biomarker could be any measure of drug action that is proximal to its clinical effect.

Cristina Sampaio of the University of Lisbon proposed three categories:

• Disease-associated biomarkers of risk, diagnosis, and progression
• drug-related biomarkers that relate to pharmacogenomics and drug response
• patient-associated biomarkers that reflect compliance or relate to adverse events

Focusing solely on biomarkers as used in drug development, Orest Hurko of Wyeth Research identified four subtypes:

• Biomarkers reflecting dosing and receptor occupancy
• those identifying patients most likely to suffer toxic effects from the drug
• those predicting which patients will have the most robust response
• those offering an early indicator of efficacy

"It is meaningless to speak of a biomarker unless you specify which of these four questions it is to address," said Hurko.

Biomarker development should be considered an "extremely high-risk undertaking."

The time and expense of identifying and validating a biomarker can be just as burdensome as developing a drug, so biomarker development should be considered an "extremely high-risk undertaking," cautioned Hurko. And the Holy Grail—a biomarker that can serve as a surrogate measure of disease for regulatory purposes—is vanishingly rare. A few have held up in other fields of medicine, such as cholesterol levels for risk of heart disease or T-cell count for AIDS progression. But so far, none stand alone in neurological or psychiatric disease.

Expensive as they are, however, biomarkers can be useful in clinical trials by improving power, rationalizing dosing, and saving money by preventing even costlier research. Candidate drugs for CNS diseases have one of the highest attrition rates in the pharmaceutical industry. "We fail late, which is not a good place to fail," said Holly Soares of Pfizer. A biomarker can tell you when it's time to call it quits. A biomarker can also convert a failure into a learning opportunity, by testing the hypothesis that engaging the new target has an effect. Without being able to monitor what's going on inside the brain, drug development is just a shot in the dark.

As Carol Brayne of the University of Cambridge pointed out, if biomarkers can be developed for risk, diagnosis, and progression, it is also important to keep a public health perspective when deciding whether they should be used in clinical practice. In the case of biomarkers for predicting risk, for example, there is the potential for misdiagnosis and expensive overtreatment of disease, particularly for brain diseases like dementia, which is actually a spectrum of disorders. She advocated for a system that would operate in parallel with biomarker development to assess the public health, ethical, social, and legal implications of potential biomarker-based interventions.

Alzheimer's: The state of the art

Biomarkers are sorely needed for Alzheimer's disease: As many as 150 targets have been suggested to be relevant to AD, Potter pointed out, but few have been validated; that is, proven to influence the pathophysiology of the disease. In terms of treatment, because the disease is slow to progress, assessing whether or not a therapy is effective is a challenge. Clinical trials for potential AD drugs will by necessity be very large and very long, and a biomarker that reflects pathophysiology could give an early signal of success or failure.

The best AD treatments may be neuroprotective agents that must be given before clinical symptoms become evident; thus a biomarker would be required to identify who is likely to benefit. Some of the most promising potential treatments are disease-modifying agents such as β-secretase inhibitors, anti-inflammatories, or immunotherapy, but a disease-modifying effect can't be proven on clinical outcomes alone; physiological changes in the brain must be documented.

A few good AD biomarkers have been established, although none have yet been fully validated. Three peptides, amyloid-beta 42 (Aβ-42), total tau, and phosphorylated tau (phosphotau), have been most thoroughly studied, with the confirmed finding that tau levels increase and Aβ-42 levels decrease in the CSF of people with AD. By monitoring all three, Henrik Zetterberg's group at the Sahlgrenska Academy at the University of Gothenburg can predict conversion to AD in a heterogeneous population. However, peptide levels have no relationship to individual disease state and cannot be used to track progression. Zetterberg mentioned another promising approach: analyzing the pattern of Aβ fragments in CSF with mass spectrometry. He noted that Aβ-42 levels in plasma have no relationship to those in CSF and cannot be used to diagnose disease.

Some types of neuroimaging are coming into their own, such as volumetric structural images that capture hippocampal atrophy, which can diagnose AD at specificity and sensitivity above 90%. Pittsburgh-B is a relatively new ligand that can be paired with PET to image overall amyloid in the brain. It has low specificity: many people have a significant amyloid load and normal cognition.

Alzheimer's: Where we must go

So far, the many efforts to identify plasma biomarkers have not succeeded, but there is good evidence that they ultimately will be found, said Simon Lovestone of King's College London. Lovestone has used 2-D gel electrophoresis and mass spectrometry-based proteomics to identify two plasma-based candidates, complement factor H and α-2 macroglobulin, that correlate with disease activity in the brain. An "in silico" search of existing literature also pointed to C-reactive protein as a potential predictor of disease progression.

Holly Soares described Pfizer's project with the biomarker-testing firm Rules-Based Medicine to develop a panel of up to 151 analytes in blood that could identify presymptomatic patients. She and Lovestone were two of the many speakers who emphasized that multiple biomarkers across several modalities will ultimately be necessary. "I don't think we're in the game of any one protein or gene or anything else being the answer," Lovestone said.

A technique that precisely lines up sequential images can monitor decline in brain volume.

Imaging, although expensive, could reduce the number of people needed for a successful clinical trial, argued Nick Fox of University College London. In one recent study, an estimated 320 human volunteers were needed to show a 50% effect on progression in one year, using a standard behavioral measure. With MRI structural imaging of hippocampal volume, the number could have been as low as 21. Structural imaging can also be used to monitor disease onset or progression in individual patients: Fox's group has pioneered a technique that allows sequential structural images of the brain in people at risk of AD to be precisely compared. Other imaging technologies that are more preliminary include diffusion tensor imaging, which visualizes white matter, functional MRI, which might profitably be combined with vMRI, and FDG-PET, which can distinguish between AD and other dementias by pinpointing activity changes.

Multiple biomarkers across several modalities will ultimately be necessary.

Genetic biomarkers may be most practical in AD to identify who is at high risk for the disease or to predict prognosis, suggested Richard Mayeux of Columbia University. The unique power of genetic biomarkers is that they are present at birth, long before disease onset. His group has focused on variants of SORL1 in a population in the Dominican Republic; other risk variants include the well-known ApoE4 allele, and possibly LRP6 and GAB2.

Cognitive biomarkers may not be as flashy, but refinements in these could yield cheap and widely applicable tools. Barbara Sahakian of the University of Cambridge described success with CANTAB, a cognitive battery that probes the function of the hippocampus, a brain region experiencing early decline in AD. This test, particularly when paired with tests of semantic memory, can distinguish elderly adults with AD from those with other cognitive problems and predict decline in a nonclinical sample of healthy older adults.

Other diseases: lessons learned

Alzheimer's is the 800-pound gorilla of neurodegenerative disease, but biomarkers are being pursued for many other brain diseases. Schizophrenia offers a very different challenge: Effective therapies are available for this disease, but the major difficulty is in timely diagnosis. The hope is that because the disorder appears to have a long prodromal phase, early intervention could alleviate or even prevent psychotic episodes.

Beginning with postmortem brain samples and moving to CSF and serum, Sabine Bahn at the University of Cambridge has found evidence of dysregulation in the periphery as well—which hints of an accessible schizophrenia biomarker. Working with Rules-Based Medicine, her group has identified a panel of 54 biomarkers that predicts schizophrenia with more than 90% specificity and sensitivity, and differentiates from depression, MS, and bipolar disorder.

Making the argument for systems thinking and the power of metabolomics, Jeremy Nicholson of Imperial College London presented provocative data linking autism to variation in gut enzymes and gut microbial flora. Metabolomics can explicate disease even where gene-association studies fail, he argued, by identifying biomarker clusters that reflect both environmental and genetic variation. At the individual level, metabolomic profiling can predict therapeutic outcomes. At the population level, the approach can be used for biomarker discovery to generate hypotheses that are mechanism-based and physiologically testable. The 1.5 kg of gut bacteria make a highly significant contribution to human physiology, Nicholson pointed out: collectively, the human microbiome may have 20 times as many druggable targets as does the human genome.

In Parkinson's, one of the simplest biomarkers is measuring a patient's ability to tap his or her hand.

In Huntington's disease, the D-2 dopamine receptor antagonist raclopride can be used as a ligand in combination with PET imaging to visualize the loss of brain tissue, a powerful but expensive technique. One of the simplest biomarkers turns out to be surprisingly effective, Roger Barker of the University of Cambridge explained: measuring the patient's ability to tap his or her hand, which deteriorates as the disease progresses. Barker also described a slightly more sophisticated version of this movement-measuring technique: a "saccadometer" built of head-mounted lasers to track eye movements. The patient's ability to follow a moving laser light is a sensitive and accurate index of disease progression. For Huntington's disease, an autosomal-dominant genetic disorder with highly variable disease onset, the central questions are whether a patient is beginning to suffer symptoms, and whether treatment is working. A major European study, Track-HD, is now evaluating biomarkers in a large population of pre-manifest carriers.

In multiple sclerosis, a demyelinating disorder, visualization of brain lesions via MRI is now accepted as part of the clinical criteria for diagnosis by the European drug-regulatory agency. Gavin Giovannoni showed a striking example of the neurological damage that occurs in MS patients, highlighting the need for early diagnosis and treatment. MRI data helped get Avonex approved for patients who had had a single demyelinating episode and were at a very high risk of developing definitive MS, "the only real approval based on biomarkers" so far in neurological disease, said Cristina Sampaio.

In MS, autoimmune antibodies attack axons and demyelinate them, but the hope is that sodium-channel blockers and anti-inflammatory agents such as lamotrigine or riluzole may protect axons from death. Clinical trials suggest that if used early on, treatments such as alemtuzumab (CAMPATH), a monoclonal antibody approved for B-cell chronic lymphocytic leukemia, may prevent relapses and disability over the long term. These drugs are promising, but because they are neuroprotective agents, studies are difficult to power. Giovannoni described using heavy-chain neurofilament (NF), a nonspecific marker of axonal damage, as a biomarker. In the CSF, baseline levels of hypophosphorylated neurofilament predict disability in three years' time. "We think this is a prognostic marker for disease progression, or degree of damage due to MS attack," he said. NF could "enrich" clinical trials by identifying the patients who are most likely to benefit.

A collective future

Relying on surrogate biomarker is risky, cautioned Sampaio, telling the story of glycemia in diabetes: For a long time, the FDA accepted glycemic control as a valid surrogate biomarker of disease improvement. In 2007, rosiglitazone, which reduces blood glucose, was shown to increase raise cardiac risk. Some diabetes researchers have suggested that it should be abandoned as a primary endpoint surrogate; at the end of 2008, the FDA concluded that while this biomarker is still an "acceptable" surrogate for new drug approval, developers must now also prove that the candidate drug does not raise cardiovascular risk, a substantial additional burden.

Even validation of "non-surrogate" or "presurrogate" biomarkers will require unprecedented collaboration, cooperation, and standardization, because the process will likely demand large studies that can prove a putative biomarker is robust and reliable among a heterogeneous population. Many talks focused on how best to carry that out. Issues that seem trivial, such as which samples to collect, what readouts to measure, and how these should be curated and documented, are in fact major hurdles to overcome. "Something as seemingly innocuous as file formats can have a profound impact on how effectively we share things," said Arthur Toga of the University of California, Los Angeles School of Medicine.

Validating biomarker will require unprecedented collaboration and standardization.

Successful large-scale collaborations will likely require public-private partnerships, free data sharing, and extensive involvement of regulatory agencies. One such model is the Alzheimer's Disease Neuroimaging Initiative, a $60 million project funded by the National Institute on Aging, private foundations, and the pharmaceutical industry and led by Michael Weiner at University of California, San Francisco. European and Japanese projects have been incorporated into this study, which now includes 25 centers and more than 800 volunteers, and all the data is freely available. "This is an example of how a small amount of money can make a difference, expand collaboration and open up dialog," said Maria Carrillo of the Alzheimer's Association USA, which has taken on the job of maintaining international information flow.

AddNeuroMed, a cross-European effort that links clinical and preclinical research, is another such example of collaboration. Here, the effort is focused on identifying a key target of progression that can substitute for clinical measures in trials of disease-modifying agents. Richard Mayeux also described his data-sharing successes with the NIA-LOAD study, which has made available data from 3698 people either with AD or in an AD family. "My personal experience is that making data and samples available to everyone from the beginning actually works," he said. "You find you have more collaborators than you would have dreamed of."

Despite the obvious challenges, Alzheimer's disease may present a unique opportunity for massive international collaboration in biomarker development to succeed. All the ingredients are there: A disease with a major impact, plenty of funding, a strong public mandate for new treatments, and hundreds of labs around the world focused on the problem. These collaborations could set the pace for biomarker development in other diseases.

Bruno Vellas, head of the European Task Force on Disease-Modifying Alzheimer's Trials reminded the audience of the critical need for better research and clinical tools as he reviewed a number of failed clinical trials for drugs to treat Alzheimer's disease. In the hope of improving future outcomes, the task force has focused its third meeting on the use of biomarkers.