Preclinical Alzheimer's disease

In May 2011, three consensus groups organized by the National Institute on Aging (NIA) and the Alzheimer's Association (AA) published a new set of diagnostic guidelines for Alzheimer's diseases (ADs) — the first revision since 1984. In those intervening 27 years, findings about the prevalence of AD and its clinical-pathophysiologic relationships have refined and altered the medical field's concept of the disease.

AD is now recognized as a complex disease for which a single therapeutic target may not be sufficient. Longitudinal studies of age-related neurological and cognitive changes have confirmed a marked temporal lag between the initiation of neuropathologic characteristics and symptom appearance. In response, the new guidelines divide AD into three phases:

  • A dementia phase, in which clinical symptoms are present
  • A symptomatic, predementia phase, in which mild cognitive impairment (MCI) consistent with AD is present
  • An asymptomatic preclinical phase, intended as a platform from which to develop primary prevention therapies

"Amyloid deposition appears to begin as early as 20 years prior to symptom development," explains David S. Knopman, M.D. Dr. Knopman, along with his Mayo colleagues Clifford R. Jack Jr., M.D., and Ronald C. Petersen, M.D., Ph.D., was a member of the NIA-AA consensus group that developed the guidelines. For the first time, the guidelines include biomarkers — such as fMRI, FDG-PET, amyloid imaging, and cerebrospinal fluid analysis — specifically for the MCI and preclinical phases.

Preclinical guidelines are intended to identify people who are asymptomatic for AD but have biomarker risk. Given the less-than-hoped-for effect of symptomatic therapies and the fact that cerebral amyloidosis begins before symptoms appear, researchers hope that a clear definition of preclinical AD will aid in the development of preventive therapies. As of 2012, however, detection of preclinical AD is strictly a research initiative.

Pathophysiologic model of AD

Findings from the Mayo Clinic Study of Aging (MCSA), a prospective, population-based research program in Rochester, Minn., and from the Mayo Clinic Alzheimer's Disease Research Center in Florida helped inform the pathophysiologic model used to generate the preclinical guidelines.

Dr. Petersen leads the MCSA and also directs the research center with the center's co-directors, Neill R. Graff-Radford, M.D., and Steven G. Younkin, M.D., Ph.D. The model, developed by Dr. Jack in collaboration with Dr. Knopman, Dr. Petersen and others at Mayo Clinic in Rochester, Minn., posits an ordered pattern of progression in AD biomarkers in which amyloid deposition reaches a peak before symptom onset. The amyloid peak is followed by neuronal injury and synaptic degeneration (tau pathology) and the onset of overt symptoms that worsen as neuronal tissue is lost.

Stages of preclinical AD

The NIA-AA work group identified three stages of preclinical AD:

  • Stage 1 is characterized by asymptomatic amyloidosis.
  • Stage 2 adds evidence of synaptic dysfunction and neurodegeneration.
  • Stage 3 marks the addition of subtle cognitive changes not overtly evident in day-to-day behavior.

Dr. Knopman points out that with more than 2,000 study participants, 500 of whom have been tested for AD biomarkers in the MCSA, "Mayo is in a unique position to evaluate the success of NIA-AA criteria in classifying people with normal cognitive function and those with mild cognitive impairment and observing changes over time."

A recent Mayo study operationalized the NIA-AA guidelines criteria to test the validity and frequency of the preclinical stages of AD (Jack, and others. Annals of Neurology. 2012;71[6]:765-75). The investigators' results confirmed the criteria for stages 1 through 3. Two additional categories were needed to classify all of the participants in the study sample:

  • Stage 0 includes cognitively normal (CN) individuals with neither biomarker evidence of AD pathophysiology nor cognitive impairment.
  • The stand-alone category, suspected non-AD pathophysiology (SNAP), accounts for individuals with normal amyloid levels but biomarker evidence of neuronal injury consistent with non-AD dementia.

The research team was able to classify 97 percent of the sample and found that:

  • 43 percent were at stage 0
  • 16 percent were at stage 1
  • 12 percent were at stage 2
  • 3 percent at stage 3
  • 23 percent were classified as having SNAP

Using the operationalized criteria in a 15-month follow-up study, Dr. Knopman and colleagues confirmed the utility of the criteria in predicting progression of cognitive impairment (Neurology. 2012;78[20]:1576-82).

Molecular mechanisms of AD

The risk of AD from the APOE ℇ4 allele is well established. The allele is a gene variant present in 20 percent of the population and is thought to contribute to amyloidosis.

In 1994, Richard J. Caselli, M.D., a neurologist at Mayo Clinic in Arizona, began prospective studies of cognitive changes across the life span, using longitudinal modeling and in-depth neuropsychological testing.

In 2009, grouping study participants as young as 21 years and as old as 97 years by their APOE ℇ4 status and comparing them to noncarriers, Dr. Caselli and his colleagues found that memory decline diverged between the groups by age 60 years despite ongoing normal clinical status (The New England Journal of Medicine. 2009;361[3]:255-63). Their findings helped consolidate the concept of preclinical AD.

In that same year, he and Eric M. Reiman, M.D., director of the Arizona Alzheimer's Consortium, of which Mayo is a participating member, found that amyloid burden in CN individuals is associated with APOE ℇ4 gene dose. They also found that amyloid deposition is particularly concentrated in the frontal and posterior cingulate-precuneus and in the temporal, parietal and basal ganglia areas (Proceedings of the National Academy of Sciences. 2009;106[16]:6820-5).

At Mayo Clinic in Florida, Dr. Younkin, Dr. Graff-Radford, Dennis W. Dickson, M.D., Nilufer Taner, M.D., Ph.D., and their colleagues are investigating other candidate genes associated with late-onset AD. In 2009, Dr. Younkin conducted one of the first genome-wide association studies (GWASs) of AD (Carrasquillo and others. Nature Genetics. 2009;41[2]:192-8). GWASs are hypothesis-free studies that require many thousands of cases and controls to generate the statistical power needed to find individual and combination gene effects.

The study by Dr Younkin and his co-investigators has been included in studies by international, consortium-based GWAS groups, in which Mayo Clinic participates. Aside from APOE ℇ4, gene risk can be considered cumulative. Numerous AD-associated genes have been identified and will undergo further validation through in vivo and in vitro functional assessment to determine their impact on risk.

Establishing heritable risk is only one part of the GWAS effort. As Dr. Graff-Radford states, "A major goal of molecular genetics is to uncover the mechanism and pathways of disease." For example, genes that contribute to inflammation and oxidative stress appear to be particularly important in neurodegenerative diseases, including AD.

Cerebrovascular (CV) risk may also contribute to AD risk. Dr. Caselli and colleagues recently found, for example, that CV risk plays a role in age-related memory decline in APOE ℇ4 homozygotes (Neurology. 2011;76[12]:1078-84). The prevalence of CV-related dementia is thought to be higher in African-Americans than in Caucasians, and Dr. Graff-Radford and colleagues are comparing amyloid burden between these two populations in an ongoing study.

In addition to participating in consortium-based GWAS studies, Mayo Clinic in Florida is one of 11 institutions participating in the NIH-funded Dominantly Inherited Alzheimer Network (DIAN) study. The goal of DIAN is to discover the molecular mechanisms of a rare form of familial AD, in an effort to decode the pathophysiology of AD and other dementias. As is true of other AD types, in this rare heritable form, amyloidosis can begin to develop years before symptoms appear.

Toward preventive intervention

Although overt AD symptoms are preceded by amyloid deposition, amyloid alone does not interfere greatly in cognitive function. Neurofibrillary tangles (NFTs), generated by tau pathology, appear to be the critical biologic requirement for symptomatic AD and other forms of dementia.

Some non-AD dementias, such as progressive supranuclear palsy, are characterized by tau pathology in the absence of amyloid deposition. Also of note, there appears to be a mismatch between areas of amyloid burden, such as the frontal lobes and cingulate gyrus, and anatomical areas critical to memory loss, such as the hippocampus and mesial temporal lobe, where NFTs first develop and the greatest neuronal death occurs. Even in APOE ℇ4 carriers, memory declines much more quickly and earlier than frontal lobe-mediated function (Caselli and others. Neurology. 2011;76[16]:1383-8).

The trigger or triggers for amyloid deposition and the pathophysiologic relationship between amyloidosis and tau pathology are among the important features of AD yet to be uncovered. Across Mayo Clinic, neuroscientists are collaborating to address these and other potential mechanisms through a coordinated effort in molecular genetics and longitudinal population studies of AD biomarkers and normal and abnormal cognitive function. Defining the features of preclinical AD marks a critical step in moving the field toward the development of preventive therapies.