On 9 July 2008, Alzforum held a Virtual Town Hall on a subject of tremendous strategic importance in the fight against Alzheimer disease: early detection. Alzforum was honored to welcome our guest moderator, Harold Varmus, Nobel laureate, former director of the National Institutes of Health, and president and CEO of Memorial Sloan-Kettering Cancer Center. Dr. Varmus is a member of the Alzheimer’s Study Group (ASG), an independent working group mandated by the United States Congress to develop a national strategic plan for Alzheimer disease (see ARF related news story). We thank the Geoffrey Beene Foundation Alzheimer's Initiative for its generous sponsorship of this town hall. Alzforum will be providing a transcription of this Webinar to the ASG.

The Alzheimer Disease Early Detection Surveys were designed to gauge perceptions and knowledge of early detection of Alzheimer disease. The surveys were developed by the editors of the Alzheimer Research Forum website in collaboration with the Geoffrey Beene Foundation as a follow-up to this live discussion.

There is a consensus among AD researchers that therapies will be most effective if given early. But how can treatments be developed when there are not yet accepted diagnostics or clinical trial designs for such early-stage AD trials? Even if such therapies are developed, how will the public be educated and motivated to ask for early diagnosis? How will physicians respond? The past decade has seen encouraging progress in both the tools for early diagnosis and in candidate drugs. The Alzforum invited a panel of thought leaders to give brief presentations about the most promising advances and frustrating bottlenecks in their areas of expertise.

View Webinar Recording

Audience Q&A

Q: Regarding the Lockhart graph presented by Gilman, what percentage of MCI progresses to AD versus what percentage of MCI does not? Is MCI a prodromal form of AD or a distinct disease?

Q: Which is better for early AD (not MCI) detection, biomarkers or cognition?

Q: How confident can we be that when we are measuring candidate biomarkers in "early stage AD patients" that these patients in fact have AD and not some other form of dementia?

Reply by Douglas GalaskoPosted 26 July 2008
Confidence that early-stage AD patients indeed have AD is high if those patients underwent a gold standard clinical evaluation by an expert clinician that was detailed, included an informant interview if there were cognitive problems, and extensive testing, particularly neuropsychological tests. Assurance can be enhanced by following the patients over time, to chart their course. Studies have shown that the expert clinical diagnosis of early-stage AD predicts AD pathology in 90 percent or more of cases.

Reply by Anne FaganPosted 20 August 2008
This is indeed a challenge when we use clinical diagnosis (which can be inaccurate at early stages of the disease) as the gold standard for identifying patient groups in biomarker studies. The other problem is that an appreciable number of cognitively normal elderly individuals will actually have AD pathology, but prior to any symptoms. Our strategy has been to identify candidate biomarkers in late-stage disease, when we can be fairly confident of the diagnosis (or even from individuals who have died and for whom we have a neuropathological diagnosis of AD), and then look at these markers in presumed cases of early-stage AD. It’s also important to assess “specificity” of biomarkers by evaluating people believed to have non-AD dementias (based on specific clinical characteristics and, if possible, autopsy verified non-AD).

Q: Does circulating Aβ have different flavors (Aβ40, Aβ42, among others)? If so, which ones correlate with AD?

Reply by Douglas GalaskoPosted 26 July 2008
Plasma Aβ has many different forms, such as Aβ1-38, 1-40, and 1-42, as well as shorter forms that begin at different parts of the Aβ molecule. So far, none of these correlates with Alzheimer disease. This probably relates to the Aβ in blood originating from many different cells in the body, not just those in the brain.

Reply by Anne FaganPosted 20 August 2008
Aβ indeed comes in various forms based on the number of amino acids it possesses. Aβ that is 40 amino acids long (Aβ40) is the most abundant in normal human fluids (CSF, blood, urine), whereas the form with 42 amino acids (Aβ42) is typically found in levels that are 10 percent that of Aβ40. Amyloid plaques in AD are made up of both Aβ40 and Aβ42, but we’ve discovered that it’s the level of Aβ42 (but not Aβ40) in the CSF (but not other fluids) that tells whether there are amyloid plaques in the brain or not. Overall, people with AD plaque pathology have lower levels of CSF Aβ42 in the brain compared to people without plaques. We believe plaques in the brain act as a “sink,” keeping Aβ from being transported out to the CSF (and eventually the blood) as it usually does so levels in the CSF drop in the presence of plaques. Other shorter forms of Aβ exist in various fluids but they have not been carefully analyzed for their potential correlation with AD dementia or AD pathology.

Q: What about measuring cytokines in serum as a potential fluid biomarker?

Reply by Douglas GalaskoPosted 26 July 2008
Cytokines have been measured in serum in AD, typically as single markers or in small panels of tests. In general, they do not distinguish patients with AD from controls. Some studies suggest that increases in certain cytokines in cognitively normal individuals may be associated with a higher risk of progressing to AD.

Reply by Anne FaganPosted 20 August 2008
A recent study by Wyss-Coray and colleagues has identified a panel of “signaling proteins” (including cytokines, growth factors, etc.) in the blood that can discriminate patients with AD dementia from those without dementia with good sensitivity and specificity. The validity of their results awaits replication by other groups. In addition, it is not known whether these signaling proteins correlate with AD pathology in the brain or whether they can predict which cognitively normal individuals will go on to develop future dementia. One of the caveats in measuring inflammatory markers is that, while AD certainly has a large inflammatory component, so do many other neurodegenerative diseases and other types of neurological disorders.

Q: What is the relation between glutamate excitotoxicity and stages of AD? Glutamate is recognized as a causative biological factor, and how does one determine the titer in the serum?

Reply by Douglas GalaskoPosted 26 July 2008
There is no relationship between glutamate excitotoxicity and stages of Alzheimer disease—glutamate may play a role in nerve cell damage in the brain, in which case it could contribute to AD at any stage. Measuring levels of glutamate is not useful because glutamate levels in serum do not reflect levels in the brain exclusively; many other cells in the body produce glutamate.

Q: What is the panelists' view on true surrogate parameters? Are any of the imaging or fluid biomarkers a candidate for this?

Reply by Michael WeinerPosted 19 July 2008
Yes almost everything that is being discussed is a candidate for a true surrogate, but much work needs to be done to validate such a surrogate. This is a major goal of ADNI.

Reply by Douglas GalaskoPosted 26 July 2008
No imaging or fluid biomarkers at present are candidates for surrogate parameters. This is because there are extremely high standards for defining and establishing surrogates. To validate a surrogate requires a positive clinical trial in which meaningful clinical outcome measures were used. The surrogate needs to map onto those clinical outcome measures and onto the biology of the underlying disease. Further clinical trials will continue to include imaging or fluid biomarkers, but only once there are drugs that work to alter the clinical course of AD, can the dialogue about “true surrogates” begin.

Comment by Milan FialaPosted 4 September 2008
We have developed the first flow cytometric blood test "amyloid-β stress test" for Alzheimer disease that will be presented 28 July, 5-7 p.m., in Chicago and is undergoing testing for approval.

Q: For Anne Fagan: using CSF tau/Aβ42 ratio, what percent of the population tested are outliers relative to prediction of MCI to AD, and what do these outliers tell us?

Reply by Anne FaganPosted 20 August 2008
In our 2007 study published in Archives of Neurology, we looked at predictors of conversion from cognitively normal (CDR 0) to cognitively abnormal (CDR >0), not MCI to AD. The issue of outliers is an important one but we haven’t analyzed enough people to be able to make any meaningful statements about outliers. Since we continue to clinically evaluate our research participants on an annual basis, we will be able to investigate such questions directly as our subject numbers increase.

Comment for Anne Fagan: regarding reluctance of non-demented elderly to volunteer for lumbar puncture, it is very understandable if taking into consideration that any invasive procedure can trigger health problems in a senior who may have existing health problems, already dealing with chronic pain due to arthritis, etc.

Reply by Anne FaganPosted 20 August 2008
In our experience (with over 500 lumbar punctures (LP) performed to date), we have not had any cases in which the LP has triggered or exacerbated any health problem. Some individuals who have severe scoliosis (spinal curvature) or who have had significant back surgery in the lumbar region may not meet eligibility criteria for getting an LP due to their back-related problems. We do not even attempt to do the procedure on such folks.

Q: Are there stages of MCI itself? Mild, medium, and advanced?

Reply by Michael WeinerPosted 19 July 2008
MCI is a continuum with many stages progressing in some to dementia.

Q: Mike Weiner’s slide of PIB labeling shows three different levels of binding. Does this correlate with other measures?

Reply by Michael WeinerPosted 19 July 2008
Thus far there is poor correlation of PIB binding with severity of impairment, except for the general observation that subjects with dementia due to ad have high PIB binding and many normals have low PIB binding, but some normals have high PIB binding with little impairments.

Q: For Mike Weiner: are there any significant changes in anterior cingulate [in early AD]?

Reply by Michael WeinerPosted 19 July 2008
The changes are largely in posterior cingulate.

Q: Does Dr. Weiner think that we could use PET in longitudinal, epidemiologically based populations?

Reply by Michael WeinerPosted 19 July 2008
Definitely. Dr Jagust has published PET in the SALSA study which is epidemiologically based and longitudinal.

Q: Is there any discussion about the neuroimaging of retina and optic nerve fiber layer as an early biomarker?

Reply by Michael WeinerPosted 19 July 2008
Don't know about this.

Q: We are three brothers, Caucasian, 57, 59, and 61, cognitively normal. Onset of AD in our father was at age 70, in mother 80, and in her father 80. What is our risk? Before the appearance of any symptoms, would knowledge of our ApoE genotypes give us a better understanding of our risk?

Reply by Anne FaganPosted 20 August 2008
I certainly understand your interest in learning your risk for AD given your strong family history. However, while family history is indeed a risk factor for AD, as is APOE genotype, we are not at the point in our understanding of the disease to be able to provide individual risk assessments. That is our ultimate goal, but the research has not yet progressed far enough to be able to do that. The problem with APOE genotype is that it is a “risk factor” for AD, not a cause of AD. Although people who have an APOE4 allele are more likely to get AD (if you compare large groups of people with and without this allele), there are many people without an E4 allele who get AD and others with an E4 allele who do not get the disease. Therefore, knowing one’s APOE genotype isn’t very informative for an individual’s risk assessment (let alone a diagnosis). Our group (and others) is performing studies looking specifically at middle-aged individuals who have a positive or negative family history of AD and for whom we have their APOE genotype data. We hope that studying these individuals as they age will give us information as to the timing of cognitive change (or no cognitive change) and its relationship to various AD-related biomarkers, such as fluid measures, neuroimaging, personality and psychometric test performance, etc.

Q: If a particular patient cohort is enriched by specifying ApoE status, wouldn't that raise concerns with respect to potential differential response to therapies regarding ApoE status? Won't we need both classes in clinical trials, as is being done with bapineuzumab?

Reply by Douglas GalaskoPosted 26 July 2008
Enriching for APOE status in clinical trials can be done for various reasons. It does run the risk of making the trial less generalizable to all patients who have Alzheimer disease. However, enriching for APOE may be important in studies of MCI or prevention and may result in the trials gaining statistical power.

Comment by Walter KukullPosted 4 September 2008
APOE genotype is not likely to be of great value for enrichment since perhaps 40 percent of AD cases are E3/E3; then we would be stuck with the problem of defining and studying AD among those non-E4 subjects. Probably better to focus on biomarkers to identify early pathological features of disease.

Q: Does anyone believe PET imaging, perhaps with F-18 amyloid tracers, will be clinically practicable if it becomes available and has been documented in a research setting and in clinical trials?

Reply by Michael WeinerPosted 19 July 2008
Yes, it will likely be clinically practical in two to three years.

Reply by Douglas GalaskoPosted 26 July 2008
PET imaging of amyloid could be clinically practicable if the health system decides to pay for it and if the investment is worthwhile. Some diseases have relatively expensive screening tests, for example, colonoscopy for colon cancer, because a positive finding has been shown to be highly predictive of the disease, and there is effective therapy once early detection is found. At present, there is no particular advantage to know that one has a positive amyloid PET scan.

Reply by Richard MohsPosted 7 August 2008
Yes, but with restrictions. The current diagnostic criteria work well enough to enable clinicians to make decisions for persons with clinical symptoms. The major value for the new tracers is for cases where there is uncertainty about etiology or symptoms are so mild as to be ambiguous. The disadvantage of the new tracers is that they will be costly and not available to all patients. The motivation to use this technology will go up substantially if there are treatments that are effective only in patients with amyloid pathology, so the decision about whether to treat or not is dependent upon knowing the patient's amyloid burden.

Q: I would like to know more about how "early detection" gets to determination of true causes of the disease.

Reply by Michael WeinerPosted 19 July 2008
Everyone is interested in this topic. No one knows the true cause of the disease, but if we identify early changes, it helps point to the true cause.

Q: To achieve really effective therapy for AD, will we have to switch to a preventative mindset and treat at-risk people before any symptoms are present? If so, can we adopt the long-term mindset that will be required, and how do we make the first steps?

Reply by Michael WeinerPosted 19 July 2008
The answer to these questions is very uncertain: it is possible that a therapy for AD dementia could be developed which completely halts progression and possibly even allows for some recovery, but many people think that to be really effective, we need a prevention approach.

I think that the long-term mindset can be adopted. The steps are as follows:

1. Identify some effective treatments.
2. Identify normal populations, without symptoms, who are at risk (for example have ApoE4, or high brain amyloid on amyloid scanning).
3. Perform a prevention trial on these normal but “at risk” subjects.

Reply by Douglas GalaskoPosted 26 July 2008
For really effective therapy for AD, prevention will be a critical component. By analogy, people who are at risk of other diseases can be treated before symptoms are present; for example, cholesterol can be lowered, blood pressure can be lowered, and colon polyps can be removed, in people who are otherwise asymptomatic. At some point, intervening to treat people at high risk of developing AD may prove to be valuable. This will be a long-term research enterprise, with many steps required. These include continuing to fund longitudinal studies that characterize cognitively healthy people, collect data on risk factors, and collect biomarkers to establish their predictive value, using cognitive change or dementia as outcomes. We also need to fund primary prevention trials and incorporate biomarkers into those trials. However, we also need to continue the current approach to treatment trials of AD, namely initiating treatment once symptoms have begun, because designing such studies is more efficient and feasible.

Reply by Richard MohsPosted 7 August 2008
Many people would agree that prevention of disease or, at least, delay in the onset of disease symptoms is the most valuable way to approach therapy for AD. Some of the drugs currently being tested for their effects to slow the progression of AD symptoms might also be tested for their ability to delay symptom onset. The major impediment to testing drugs for their effects to delay symptom onset is the lack of accepted diagnostic criteria for early diagnosis. Pharmaceutical companies are reluctant to test drugs in groups that have not been accepted by the medical community or regulatory bodies because the data collected in such groups might not be recognized by regulators or clinicians.

Q: Diagnosis should start in young adulthood, be minimally invasive, inexpensive, and initiated by any physician. It would be worthwhile to discuss the strategy that would establish the validity of such a diagnostic.

Reply by Michael WeinerPosted 19 July 2008
Everyone agrees with the goal. The field is working on methods that ultimately will make this happen.

Reply by Douglas GalaskoPosted 26 July 2008
“Diagnosis” in young adulthood is not possible, given current definitions of Alzheimer disease: the disease depends on the presence of symptoms. If we were to diagnose preclinical Alzheimer pathology in adulthood, the age at which this should be undertaken will depend on the outcome of many different studies. For example, colonoscopy is considered a useful screen after the age of 50 and not earlier—unless there is a strongly positive family history. Similarly, there are guidelines about when to screen with mammography. These both remain controversial (e.g., who should be screened, what age, how often to repeat screening) after undergoing large-scale clinical trials. To validate the significance of a diagnosis of preclinical AD pathology or of clear-cut AD risk in adults will require large-scale prospective studies.

Reply by Richard MohsPosted 7 August 2008
Standards for establishing such a diagnostic are being discussed. However, no clear path for validating such a diagnostic has been established.

Q: How do we get "pre-symptomatic" subjects (who have the AD pathology) for study? They are perceived as "normal."

Reply by Michael WeinerPosted 19 July 2008
They could be defined by having "high risk," meaning they might have ApoE4, high brain amyloid, reduced size of brain volumes, PET scans which "look like" AD or MCI.

Reply by Douglas GalaskoPosted 26 July 2008
Presymptomatic subjects may be interested in research studies for a number of reasons even though they perceive themselves as being normal. Some may be mutation bearers of early-onset AD genes, who have not yet developed symptoms; e.g., the DIAN study will enroll such individuals. Another example is people who are carriers of the APOE4 allele, who have been enrolled in a number of biomarker, psychometric, and other studies. Another strategy is to enlist people who have a positive family history who may have some self-interest in participating in such a study. Baby boomers may wish to participate in studies to find out more about their own risk. Finally, altruism is another reason for normal volunteers to participate in a variety of biomarker studies.

Reply by Richard MohsPosted 7 August 2008
There should probably be a concerted effort by government funding agencies to identify such persons. Potential therapeutic candidates could be selected from those available for test so that controlled treatment trials of the most promising agents could be tested in these persons.

Q: What constitutes "early" treatment?

Reply by Richard MohsPosted 7 August 2008
Early treatment would be when a person has evidence of biological changes typical of early AD, but does not yet have symptoms severe enough to warrant a diagnosis by currently accepted criteria.

Q: How much do we currently know about preventing AD?

Reply by Douglas GalaskoPosted 26 July 2008
Current knowledge about preventing Alzheimer disease is still not clear. Clues from epidemiologic studies have led to risk factors being identified, which generally have relatively low risk increasing or risk reducing hazard ratios. While general measures such as a heart-healthy diet, treatment of cardiovascular risk factors, physical activity, and mental activity are reasonable, it is likely that pharmacologic interventions targeting the earliest critical biochemical steps in Alzheimer disease will need to be added. Preventive studies need to continue, in order to define who is at risk and how we measure risk.

Reply by Richard MohsPosted 7 August 2008
There are some data to suggest that physical fitness, mental activity, and better education can diminish the risk of AD. However, these effects do not prevent most cases of AD, and age, which cannot be prevented, is still the most potent risk factor. Further reduction in the incidence of AD cases is likely to be achieved only through medical intervention, such as drug therapy for delay of symptoms.

Q: How can early detection be helpful for families with a PS1 gene mutation?

Q: My wife, just turned 41, has been having symptoms since she was 37 (PSEN1 mutation). We are eager to participate in research for both her and our children's benefit. I would like to know how we can open up research to those who have EOAD.

Q: Under all conditions and genotypes, women are more likely to develop AD. Sixty-eight percent of all AD victims are women. Is AD a national women's health issue, and if so, is there a national plan to address it (Brookmeyer et al., 1998)?

Q: For Drs. Aisen and Doody: your view about the often voiced skepticism on amyloid targeting strategies? Skeptics contend that tau accumulation over time maps with the progression of symptoms, while amyloid does not.

Reply by Rachelle Smith DoodyPosted 19 July 2008
I have never thought that we should advocate one treatment theory over another. A patient could potentially benefit by alleviating multiple aspects of any disease, through multiple mechanisms. So we should judge the preclinical data regarding therapies for AD in terms of whether or not the idea is reasonable, and whether or not the animal safety data justifies a trial. Multiple different amyloid targeting strategies meet my criteria for reasonable and with an adequate risk-benefit ratio to justify further study.

Q: Clinical-level AD remains poorly identified and undermanaged. How can we spur development of dementia health services research as a national priority?

Q: I am taking part in an experimental study for Aging and Memory. Why can't the researchers disclose the results of tests if any positive findings are found? If the patient is informed of a potential risk, he or she can take preventive measures.

Q: Giving a diagnosis of early-stage AD can have profound effects on the patient and family members. Is there a need for more psychosocial research on these impacts before we advocate for screening?

Q: Are there lessons from the "war on cancer" that could inform the current discussion about AD?

Reply by Michael WeinerPosted 19 July 2008
Yes, early detection is critical!

Comment by Margaret Noel, M.D.Posted 4 September 2008
Early diagnosis is of benefit to future caregivers. They can then begin an education process that will prepare them for their role. In our practice, many of the conflicts that occur in early stage disease regarding transitioning independence can be avoided with caregiver education.

Comment by Jay Smith, AdvocatePosted 4 September 2008
One of the greatest bottlenecks to progress is language, and the need for putting this information into language that the public, and even elected officials, can understand. Let's find a way to summarize this information in accessible language—this should help generate common understanding and effective action.


  1. It Is Time to Make Alzheimer’s Prevention a National Priority

    Alzheimer disease afflicts one in 10 people over 65 and almost half over 85. By the time today’s young adults become senior citizens, the number of Alzheimer’s patients and caregivers is projected to create an overwhelming societal and financial problem. But we now have an opportunity to create a different future. Former House Speaker Newt Gingrich and former Senator Bob Kerrey last year established a non-partisan Alzheimer’s Study Group consisting of national leaders. Its charge is to increase public awareness about Alzheimer disease and develop a strategic plan to address it in the most comprehensive way. This is a timely initiative: a bold plan is critically needed to find effective risk-reducing and prevention therapies.

    Researchers have developed a growing number of investigational medication and immunization treatments to slow and prevent the disease, some of which are entering clinical trials in AD patients. They have also raised the possibility that some already available medications, dietary supplements, and healthy lifestyle interventions might reduce a person’s risk. An even modestly effective risk-reducing treatment would have an extraordinary public health impact—and some treatments may be most effective if started before the development of symptoms or the brain pathology that is already apparent by that time.

    In my opinion, the hope of finding effective Alzheimer’s prevention therapies will never be realized unless we are able to develop a rapid and rigorous way to test them. It currently takes too many healthy research subjects, too much money, and too much time—longer than the life of a drug company’s patent—to test a suggested prevention therapy in a clinical trial of a typical primary prevention design. It’s just not practical to wait so long to determine which healthy subjects develop memory and thinking problems.

    My colleagues and I have been using imaging techniques to track Alzheimer’s-like brain changes in healthy people carrying the common Alzheimer’s susceptibility gene ApoE4. We have suggested how these markers of disease progression could be used in these and other at-risk people to test promising risk-reducing and prevention therapies without having to wait many years to determine whether healthy subjects develop symptoms. While this strategy and others like it offer hope, more needs to be done.

    What can Congress do to find effective Alzheimer’s prevention therapies in the shortest possible time? Here is what I recommend:

    1. Encourage pharmaceutical companies to include brain imaging and other biomarkers of disease progression in clinical trials of every disease-slowing treatment. If regulatory agencies like the FDA are confident a treatment’s effect on biological markers of the disease predicts a good clinical outcome in already ill patients, they will be more likely to approve a prevention therapy showing the same disease-slowing effects without waiting to see if healthy research subjects develop symptoms.

    2. Give pharmaceutical companies a compelling incentive to evaluate risk-reducing and prevention therapies. By offering to extend a drug's marketing exclusivity, awarding research grants, and providing tax breaks, Congress has promoted the development of “orphan drug” treatments for rare disorders. Imagine what similar incentives could do to promote the identification of an effective Alzheimer’s prevention therapy.

    3. Provide federal funding to evaluate the risk-reducing effects of available medications, dietary supplements, and otherwise healthy lifestyle interventions for which industry lacks a patent incentive.

    4. Make the scientific understanding, treatment, and prevention of Alzheimer disease a national priority.

    In 1961, President Kennedy called on the nation to send a man to the moon and return him safely to Earth before the end of the decade. With fewer scientific resources to address that problem than are now available to the problem of Alzheimer disease, the nation completed the job ahead of schedule.

    When the new administration takes office, the Alzheimer’s Study Group will call on the nation to address the problem of Alzheimer disease in a transformative way. Let’s stop and end Alzheimer’s without losing another generation.

    Dr. Reiman is Executive Director of the Banner Alzheimer’s Institute and Director of the Arizona Alzheimer’s Consortium.

    A similar version of this column ran in The Arizona Republic on 23 February 2008.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Projections of Alzheimer's disease in the United States and the public health impact of delaying disease onset. Am J Public Health. 1998 Sep;88(9):1337-42. PubMed.

Other Citations

  1. ARF related news story

External Citations

  1. Alzheimer’s Study Group (ASG)

Further Reading


  1. . Estrogen replacement therapy for treatment of mild to moderate Alzheimer disease: a randomized controlled trial. Alzheimer's Disease Cooperative Study. JAMA. 2000 Feb 23;283(8):1007-15. PubMed.