A committee convened by the National Institute on Aging (NIA) and Alzheimer’s Association (AA) has revised the postmortem assessment of Alzheimer’s disease neuropathology. In essence, postmortem brains will receive a composite "ABC" score of A) measurements of amyloid-β (Aβ) deposits determined by Thal et al., 2002; B) the locations of neurofibrillary tangles according to Braak and Braak, 1991; and C) the number and types of neuritic plaques according to scoring established by the Consortium to Establish a Registry for AD (CERAD). The biggest change from the current NIH-Reagan guidelines is that the neuropathological assessment of AD will no longer require a prior diagnosis of dementia (see Part 1).
Last July’s Alzheimer's Association International Conference (AAIC) in Paris, France, featured a special session dedicated to the topic, in which committee members introduced the new guidelines and then conducted a question-and-answer period with the audience. Questions were answered by Thomas Montine of the University of Washington, Seattle, and Bradley Hyman of the Massachusetts General Hospital, Boston, who co-chaired the committee. The summary of this discussion, below reflects some of the unanswered questions in the AD field. Since then, the committee has fielded comments from the research community and will present an updated draft for approval at a meeting of the Alzheimer’s Disease Research Centers on 23-24 September 2011 in San Diego, California. If all goes according to plan, the new guidelines will be published in early 2012. (Note that questions and answers below were edited for clarity; in some cases similar responses were combined.)
Is the Diagnosis of AD Based on Pathological or Clinical Criteria?
Some AD researchers believe that the neuropathologist should make the diagnosis of AD, whereby “one plaque, one tangle” is sufficient for such a conclusion. Others, however, think that the neuropathologist can only make the conclusion of AD-related changes, and the actual diagnosis is based on clinical findings. The second position prevailed in the draft document presented by the NIA-AA committee.
Montine: What do we do with someone who has high levels of plaques and tangles but was cognitively intact at death? The majority opinion was that the surest ground is to report our observations, the neuropathologic changes that we see, independent of clinical standing. If we were to adopt this approach, the clearest way to answer this question is to say we observe neuropathologic changes of AD.
Hyman: I agree. We try to disambiguate the term AD, which historically connotes both a neuropathologic and a clinical entity. The science has changed to the point where we now think of AD as a long disease process with a long preclinical stage that has the neuropathologic characteristics of AD but no overt symptoms. We can draw an analogy with atherosclerosis and angina; you see the former, but the patient does not yet have the latter. Asking the neuropathologist to determine what a person’s MMSE [Mini-Mental State Exam] score would have been is like asking a pathologist who sees atherosclerosis under the microscope if that patient had heart pain walking up the steps. There are certainly individuals in whom there is a clinicopathologic disconnect; we have to understand that in terms of the spectrum of disease rather than be confused by the nomenclature.
Does CERAD Scoring Need to Go Into the Neuropathologic Assessment of AD?
The two defining characteristics of AD are presence of amyloid deposits and of neurofibrillary tangles. One view in the field is that measuring those two parameters is sufficient to characterize AD-related pathology without any need for the CERAD score, which is a semi-quantitative measure of neuritic plaques. The NIA-AA committee recommended using CERAD as part of the neuropathologic examination.
Hyman: This was a matter of a good deal of discussion. We agreed that we do not know exactly how much each individual criterion contributes to the assessment of AD-related pathology. We all felt that molecular-defined lesions, the Aβ, and tau deposits are well established. But we do not yet understand as well the alterations that occur around plaques and which lesions damage the brain most. So we are still uncertain which of these lesions are most informative with regard to neural dysfunction; we decided to record them all until research has sorted that out.
Montine: The idea of having both the Thal phases and CERAD came from our realizing that we knew less than we thought we knew. The first measure is topographically based and the second is density-based. Humility dictated that we include both until we understand their significance to brain function better.
Why Not Specify Techniques for Detecting Plaques and Tangles?
The draft document recommended certain techniques for detecting AD-associated pathology, but the committee decided to leave it up to neuropathologists to choose their method. Several members of the audience thought that more guidance was needed, charging that “liberalism” might lead to inconsistencies. It may also lead to differences in the protocols used in the U.S. compared to Europe and Japan. In Europe, for example, BrainNet has spent several years comparing different methods. Their findings suggest that immunohistochemistry yields the most consistent results among labs.
Montine: In the report, we recommend a preferred stain for each of the elements in the ABC score. There were strongly held opinions on what is the optimal method. The danger is each has its own nuances and collects slightly different data. We are not agnostic on this point. We do recommend certain stains over others, but we stop short of saying everyone must do it this way. We do prefer certain sets of stains, and they are largely antibody-based.
Our panel had robust discussions on whether to recommend particular reagents, particular antibodies. We know research questions would benefit from that. But it would not help all audiences. Researchers will find those antibodies on their own. It is not really the role of a national committee like ours to proscribe particular reagents. We welcome comments from researchers in Europe and Japan.
What Are the Implications for the NACC and Other Databases?
Audience members suggested that the committee should recommend a system, either National Alzheimer’s Coordinating Center (NACC) or Alzheimer’s Disease Cooperative Study (ADCS), where neuropathological and clinical data are collected in a database. Ideally, that would be a database other international groups could contribute to.
Montine: We plan to immediately work with Budd Kukull [at the University of Washington, Seattle], who runs the NACC to revise its database to reflect these changes in neuropathologic guidelines. That will happen once the new criteria are finalized. We will likely proscribe methods that researchers at all Alzheimer’s Disease Research Centers would use.
Nigel Cairns (Washington University in St. Louis, Missouri, and a member of the NIA-AA committee): I am the neuropathology coordinator for the Alzheimer’s Disease Neuroimaging Initiative (ADNI). I also do neuropath for the Dominantly Inherited Alzheimer Network (DIAN). For each of these large projects, we are using standard neuropathologic protocols, standard stains, which are not mandated by these new guidelines. The data from different sites are then submitted to a central database.
Montine: Is that assessment compatible or incompatible with these new criteria?
Cairns: It is entirely compatible. We are using three sets of criteria: we do Khachaturian (Khachaturian, 1985), CERAD, and the NIH-Reagan guidelines.
What to Make of Tangles in Braak Stages I to IV When There Is No Amyloid?
Montine: We do not know exactly what these cases represent. They may represent AD in evolution. But pathologists should be warned that if they see tangle-only, they should evaluate other tauopathies before settling on AD. We say it is not AD.
Hyman: After a lot of discussion, the committee’s consensus was that it’s best to be humble and go with describing what we see and not overreaching on conclusions.
What Is the Recommendation for Other Characteristics of AD?
Other features of AD pathology include synapse loss, neuron loss, atrophy, gliosis, and other neuronal lesions like TDP-43 immunoreactive inclusions, granulovacuolar degeneration, and Hirano bodies, as well as congophilic amyloid angiopathy.
Montine: The report is at 28 pages and going…. TDP-43 and white matter are in there. In our discussions about this, we discovered that we each held strong views on what these other characteristics mean and how to evaluate them, but there are almost no head-to-head comparisons.
Is the Morphology of Plaques Important?
There is a lot of history, literature, and debate about the relative meaning of diffuse, dense, compacted, neuritic, cotton-wool, and other shapes of plaques to disease mechanism.
Hyman: It is safe to say that the issue of morphology had plenty of airtime in the committee’s discussions. We did not find consensus on what the different morphologies mean for neural or cognitive function. There are many opinions. We hit barriers there. We could not even reach a consensus on what to call many types of plaques. The committee’s view is that this should be excluded from the guidelines.—Gabrielle Strobel.
This concludes a two-part series. See also Part 1.
- Thal DR, Rüb U, Orantes M, Braak H. Phases of A beta-deposition in the human brain and its relevance for the development of AD. Neurology. 2002 Jun 25;58(12):1791-800. PubMed.
- Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol. 1991;82(4):239-59. PubMed.
- Khachaturian ZS. Diagnosis of Alzheimer's disease. Arch Neurol. 1985 Nov;42(11):1097-105. PubMed.
No Available Further Reading