At the second annual Clinical Trials on Alzheimer’s Disease meeting, 29-30 October 2009 in Las Vegas, prevention—how, when, with what interventions—dominated the discussions. Everyone seems for it in principle, but views diverged from there. Here are some of the main points.

Proving that prevention can work is going to require lengthy and large-scale clinical trials, and but one challenge for such trials will be figuring out whom to include. One option—to study only people who are highly susceptible to AD—raises the question of how to identify such high-risk groups. Speakers offered several suggestions for screening tools, ranging from bio- and genetic markers to memory testing to non-invasive techniques such as a risk index.

ApoE4 to date is the strongest genetic risk factors for late-onset Alzheimer disease, but it is not predictive. The recent discovery of polymorphisms in the Tomm40 gene that lies near ApoE on chromosome 19 might eventually fine-tune the predictive power of ApoE genotyping (see ARF related CTAD news). A different approach to genetic risk is to study adult children of affected individuals. Mark Sager, University of Wisconsin, Madison, described the Wisconsin Registry for Alzheimer’s Prevention. WRAP is following middle-aged people who have at least one parent with AD. Begun 8 years ago, WRAP has recruited participants from all over the country. They are highly motivated, Sager noted, adding that that WRAP has been overwhelmed with volunteers and experienced little attrition. WRAP tests individuals every four years for cognition and also collects serum, plasma, and DNA samples. In collaboration with Sager, Sterling Johnson, from the William S. Middleton Memorial Veterans Hospital, Madison, has included members of the WRAP cohort in brain imaging analyses. He recently reported that family history affects functional MRI signals in specific areas of the brain, including the medial temporal lobe, during memory tasks (see Xu et al., 2009). At CTAD, Johnson also reported that a variety of brain areas show differences in diffusion tensor imaging (DTI) signals between people with and without a family history of AD. Curiously, the scientists found that DTI showed greater atrophy in people with a maternal family history than with a paternal one (see also Mosconi et al., 2007; Mosconi et al., 2009).

Johnson is not sure why fMRI or DTI differences emerge in people with a family history of AD who are themselves cognitively healthy at present, but said he plans to image participants with PiB PET, as well, to look for the presence of amyloid as a possible link. The scientists also found that the additional presence of ApoE4 alleles lead to even greater DTI differences from controls with no family history. The work suggests that family history confers risk over and above that explained by ApoE4 genotype along. Sager said he plans to work with Roses to investigate the possible genetic connection, which could be influenced by genetic variation in the nearby Tomm40 gene (see ARF related CTAD news).

There is currently no robust blood test for AD. However, blood samples collected from cohorts such as WRAP might eventually prove useful for biomarker analysis. One novel idea presented by Rick Einstein from the biotech company ExonHit Therapeutics in Paris, France, is to use a transcript microarray approach to test biological fluids. Genotyping delivers a static assessment of risk, whereas transcription profiling informs about dynamic changes that may be useful for diagnosis or to monitor intervention, Einstein said. The company performed a proof of concept with blood samples from AD patients and controls. From 177 samples in a training set, Einstein and colleagues generated transcript signatures from 120 samples chosen at random. They then used the 57 remaining samples to re-iterate the test and improve the transcript signature. This was then applied to 110 AD patient and 101 control samples, yielding a sensitivity of 74 percent and specificity of 67 percent. When applied to people with Mini-Mental State Exam (MMSE) scores of 26-30 specificity rose to 80 percent. (CSF Aβ/tau tests exceed that percentage – they and combinations of CSF with brain imaging have set a high bar for sensitivity and specificity of AD detection.) Einstein did not describe the profile but noted that ranking transcripts in the array against disease severity highlighted the kinase JAK2 as strongly linked to disease. This kinase has been implicated in Aβ toxicity (Chiba et al., 2008; see also Ray et al., 2007).

Another area of urgency for prevention trials concerns cognitive instruments that pick up subtle memory effects not detected by the MMSE, a crude screen for memory impairment. Herman Buschke, Albert Einstein College of Medicine, New York, summarized his Memory Capacity Test (MCT). Even though the MCT is a simple cued recall test, it appears to detect subtle and early cognitive decline, Buschke claimed.

The test uses two 16-category word lists that people are asked to remember. The test analysis measures “associative binding” of words on the second list to words from the same category in the first list. In other words, it calculates to what extent a person remembers words on the second list when they remember words from the same category on the first. Buschke found that some people with normal recall on the first list have lower-than-normal binding scores. “The test identifies early presymptomatic memory impairment when memory is otherwise within normal limits,” said Buschke. People who are demented or have MCI also have lower-than-normal binding scores on the MCT, and people who score below normal on many common tests of cognition also score below normal on the MCI, giving it some face validity.

Poor MCT scores also correlate with the presence of amyloid in the brain as measured by PiB PET, said Buschke. Slated for publication in Neurology, this work, by Reisa Sperling, Brigham and Women’s Hospital, and Keith Johnson at Massachusetts General Hospital, both in Boston, may help address the significance of PiB retention in supposedly normal individuals (see ARF related news story). Since PiB imaging began, researchers have been surprised to find substantial binding in cognitively normal people. The MCT analysis suggests that these people may actually have underlying memory problems, making Buschke’s memory test potentially useful for identifying truly normal controls in prevention or treatment trial settings. In addition, the MCT is a pencil-and-paper test that only takes six minutes, making it suitable for rapid screening of normal subjects for prevention trials a primary care setting.

A related methodology for detecting early cognitive decline is the CogState test battery of working memory. This is a commercial computer-based set of tests that uses playing card images to query memory (has the person seen the card presented earlier) and reaction time (how quickly do they push a button to record whether they have seen the card or not). In the Melbourne Healthy Aging Study in Australia, researchers found that people with MCI perform poorly compared to controls on this test, and also show faster decline over a twelve month period (see Maruff et al., 2004). David Darby from CogState, Melbourne, and the Florey Neuroscience Institute in Parkville, Australia, plans to use the test in a novel trial design that seemed well-received by the audience at CTAD. As with the MCT, Darby used PiB binding correlation to show that this CogState test has some validity. In a community screening study, Darby found that people who decline fastest in quarterly tests also test positive for PiB. Darby now plans to enroll 10,000 normal elders in a screening program and to offer subsequent clinical trials to them. Part of the novelty is that because Darby will have CogState baseline and six months of trajectory data, the patients will serve as their own controls. After this initial run-in, Darby envisages a delayed-start or a withdrawal trial design to assess disease modification. Darby said that looking very early in the disease process enhances one’s chance of seeing a drug effect. Lon Schneider, University of Southern California, Los Angeles, expressed concern about the ethics of such a design, but Darby replied that volunteers enrolled in the screening program would sign consent forms before any baseline data would be taken.

Overall, a writer took away the impression that the field is moving toward studying people earlier in the disease process. One of the themes that repeated itself was the idea of distinguishing between early (eMCI ) and late (lMCI) phases. Attendees also grappled with whether MCI should be labeled preAD, or prodromal AD. Ron Petersen, Mayo Clinic, Rochester, Minnesota, cautioned that clinicians may be uncomfortable telling patients they have “pre-AD” because some patients who do not end up with AD would be labeled erroneously. On that note, Schneider tempered the audience’s collective prevention enthusiasm with a reminder that the underlying biology of AD is still not understood and that the etiology is likely highly heterogeneous. He noted that it is hard to prevent something when the etiology is unclear. “We should be very careful about the drug we use. We should have a really good reason to believe it will work because these trials are time-consuming exercises,” Schneider said, adding that he strongly advocated efforts to model and simulate prevention trials before embarking on real trials.—Tom Fagan.


No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Las Vegas: AD, Risk, ApoE—Tomm40 No Tomfoolery
  2. St. Louis: Imaging Preclinical AD—Can You See it Coming in the Brain?

Paper Citations

  1. . The influence of parental history of Alzheimer's disease and apolipoprotein E epsilon4 on the BOLD signal during recognition memory. Brain. 2009 Feb;132(Pt 2):383-91. PubMed.
  2. . Maternal family history of Alzheimer's disease predisposes to reduced brain glucose metabolism. Proc Natl Acad Sci U S A. 2007 Nov 27;104(48):19067-72. PubMed.
  3. . Declining brain glucose metabolism in normal individuals with a maternal history of Alzheimer disease. Neurology. 2009 Feb 10;72(6):513-20. PubMed.
  4. . Amyloid-beta causes memory impairment by disturbing the JAK2/STAT3 axis in hippocampal neurons. Mol Psychiatry. 2009 Feb;14(2):206-22. PubMed.
  5. . Classification and prediction of clinical Alzheimer's diagnosis based on plasma signaling proteins. Nat Med. 2007 Nov;13(11):1359-62. PubMed.
  6. . Subtle memory decline over 12 months in mild cognitive impairment. Dement Geriatr Cogn Disord. 2004;18(3-4):342-8. PubMed.

External Citations

  1. Clinical Trials on Alzheimer’s Disease
  2. ExonHit Therapeutics
  3. CogState

Further Reading