Share this article on:

How to Diagnose and Define Alzheimer's: New Criteria Spur Debate in the Field

Talan, Jamie

doi: 10.1097/01.NT.0000452268.87868.75
Back to Top | Article Outline


A new set of criteria for diagnosing Alzheimer's diverges from other standards set by an earlier working group of experts, inviting discussion about the need to re-examine and test criteria for diagnosis.

Figure. No caption a...

With a more in-depth understanding of the clinical and pathophysiological features of Alzheimer's disease (AD), a team of international investigators — the International Working Group (IWG) — has refined research diagnostic criteria that they first proposed in 2007.

The latest IWG revisions, which were published in the June issue of Lancet Neurology, rely on specific biomarkers to make a definitive diagnosis, but are intended for research purposes only. There are almost 60 different clinical studies underway and such diagnostic information is crucial to identify patients with AD, especially in its earliest forms.

Recent studies have shown that 20 to 30 percent of AD patients in the pre-biomarker era were later tested and found not to have pathological signs of AD in their brains, the IWG authors noted. This suggested that they were suffering from other conditions or that amyloid is only one component of clinical dementia and does not have to be present.

Bruno Dubois, MD, director of the l'Institut de la Mémoire et de la Maladie d'Alzheimer at the Hôpital Pitié Salpêtrière in Paris, who led the IWG guideline effort, said that he never felt comfortable with the low accuracy for the diagnosis for AD. In the mid-2000s, Dr. Dubois gathered a team of AD experts to begin thinking about a new definition of AD. During most of his tenure as a neurologist, the diagnosis was purely a clinical one.

“I never saw why Alzheimer's should be the only disease where it must reach a threshold of severity for people to receive a diagnosis,” said Dr. Dubois. “Normally, we try to diagnose diseases when they start. The argument was that no one really needed to make a diagnosis for a disease for which there were no treatments. That just wasn't a good argument. I never understood why we should wait for dementia. Many things can cause dementia. It is better to characterize what is behind the syndrome.”

So he set out to improve the accuracy of the diagnosis and allow researchers to arrive at a diagnosis early in the course of the disease. The pathophysiological biomarkers developed in the last decade fit nicely into his vision.

Back to Top | Article Outline


Dr. Dubois said the IWG proposed a simplified algorithm that encompasses AD at any stage of the disease. The evidence continues to support the presence of an amnestic syndrome of the hippocampal type as the clinical core criterion for typical AD, he said, which could best be identified with a list-learning memory test such as the free and cued selective reminding test as well as other delayed recall tests.

Along with behavioral and cognitive changes, the IWG criteria depend on evidence of a positive biomarker in CSF — low amyloid beta 42 and high tau or high phosphorylated tau. The committee felt that MRI biomarkers to identify hippocampal atrophy or findings on FDG PET are not necessary to make a diagnosis; but that these were more downstream effects. [See “IWG Criteria for Typical AD.”]

The IWG also included diagnostic information for other atypical forms of AD, including posterior cortical atrophy, logopenic aphasia variant, and frontal variant. The biomarkers needed to confirm a diagnosis include CSF (low amyloid beta 42 and high tau or high phosphorylated tau) or a positive amyloid PET scan. [See “IWG Criteria for Atypical AD.”]

“These new IWG criteria are about the present and future,” said Dennis Selkoe, MD, co-director of the Center for Neurologic Diseases in the department of neurology at Brigham and Women's Hospital and the Vincent and Stella Coates professor of neurologic diseases at Harvard Medical School. “This is a more simplified document, and I feel comfortable advocating this set of criteria.”

Randall J. Bateman, MD, the Charles F. and Joanne Knight Distinguished professor of Neurology at Washington University School of Medicine, agrees. “We know more now than we ever have in the history of AD,” said Dr. Bateman, who was part of the IWG. “Our rationale was to be more comprehensive. The old criteria took into account the amnesic forms of AD but there are other presentations that are important to document. These research criteria are more accurate than what we have been using in the field. But remember, this is not for clinical use.”

Back to Top | Article Outline


The IWG guidelines differ in several ways from the criteria developed by a workgroup convened by the National Institute on Aging and Alzheimer's Association (NIA-AA) and published in a series of papers in 2011. [See Neurology Today's report, “New Guidelines for Diagnosing Alzheimer Disease: First Update Since 1984,”] Neurologists say that testing the competing diagnostic reports in the field will be necessary to fine-tune each set of criteria.

The NIA-AA guidelines were designed to enable clinicians who may not have access to the latest biomarkers to arrive at a diagnosis. They also included preclinical guidelines that were based on clinical and biomarker findings.

The NIA-AA breaks the diagnoses down into three different components: preclinical, mild cognitive impairment (MCI), and AD dementia. All three stages would be corroborated by biomarker evidence of underlying AD pathophysiology.

But this is one area that has been subject to debate by AD experts who worked on both sets of guidelines. In the NIA-AA guidelines, if a patient has a positive AD biomarker, he or she “would meet research criteria for preclinical AD.

In contrast, the IWG group puts the patient into a category of ‘asymptomatic at risk for AD,’ which I feel is a more accurate and appropriate term,” said Yaakov Stern, PhD, a professor of clinical neuropsychology in the departments of neurology, psychiatry, and psychology at Columbia University Medical Center, who served on the IWG and NIA-AA committees. The IWG report does not include the diagnosis of MCI.

For his part, Dr. Dubois said he does not think that MCI due to AD is necessary in diagnostic guidelines because the “diagnosis of AD relies on a clinical phenotype and at least one pathophysiological biomarker. “In my view,” he said, “MCI is a syndrome that may be due to different etiologies. Because it is now possible to identify AD before the dementia threshold, I think that it is more adequate to speak about ‘prodomal AD.’ MCI should now be restricted to patients for whom there is no disease clearly identified.”

Ronald C. Petersen, MD, PhD, director of the Mayo Alzheimer's Disease Research Center, coined the term MCI and led studies to confirm that many of these patients go on to develop AD. Dr. Petersen, who served on the NIA-AA committee, said: “MCI is by definition a heterogeneous condition. It is no more specific for AD than dementia. But there are subtypes that are highly specific for AD. We call this amnestic MCI. In our guidelines, we enrich this diagnosis with biomarkers. If they are positive we can call this AD at the MCI stage of clinical impairment.”

Back to Top | Article Outline


But can AD be defined when there are no symptoms? This question is at the heart of the heated discussion that surrounds these two sets of guidelines. The NIA-AA data suggest that biomarkers that measure amyloid and tau in the brain and CSF are a solid indication of disease. But the field has known for decades that there are people who die with the classic hallmark signs of AD who had no evidence of memory problems in life.

“I just don't think there are enough data either way to make a case that positive biomarkers without clinical signs is AD,” said Dr. Bateman. And a diagnosis does have implications for the person who is handed a diagnosis years, even decades, before clinical signs even exist, as well as those investigators designing studies to test medicines in the earliest stages of disease, he added.

“This is about how you use words,” said Dr. Bateman. “We shouldn't tell people that they have AD if they don't have clinical signs of disease.” Dr. Petersen agrees that clinicians “can't afford to label people with AD prematurely. AD is a fatal disease and we as clinicians have to be correct.”

But, said Reisa Sperling, MD, “AD is not just a clinical disease. It begins with pathology in the brain a decade or more before clinical signs.” Dr. Sperling, a professor of neurology at Harvard Medical School's Brigham and Women's Hospital and Massachusetts General Hospital, led the NIA-AA's work group on assessing preclinical AD.

“The NIA-AA working group felt it was important that the disease can begin before symptoms so that we can start treatment as early as possible. We don't wait for any other disease to start showing symptoms before we begin treatment.

“For me it is important to call it a disease so we can do something about it,” she added. “We will have to see if treating before symptoms will change the course of the disease.” Dr. Sperling is heading the A4 study to identify older people with amyloid positive markers and no clinical signs and randomize them into treatment with an anti-amyloid agent.

“It would be ideal if we could develop a standardized set of imaging or biomarkers for identifying people at the highest risk of memory decline,” she said.

Dr. Sperling said that the NIA-AA working group is discussing re-evaluating its criteria. “All of these criteria will evolve over time. We need to see whether the biomarkers will be able to help predict clinical outcome on an individual basis.”

As the field learns more about AD, it is certain that both of these definitions, however subtle in their differences at the non-dementing state, will change. “What is it that kicks someone over to AD? Eventually we will be able to say who is at risk and who is not,” added Dr. Stern.

Ultimately, biomarkers may be critical with the new generations of experimental AD medicines that target specific pathologies. “If you have an amyloid agent you better be sure the patient has amyloid accumulating in their brain,” said Dr. Petersen.

“Amyloid is an important component but should it be the end-all? There are patients who are impaired who have other pathological issues contributing to the clinical condition. I wonder if we wouldn't be better off just using clinical descriptors and then identifying pathological states contributing to that clinical picture, amyloid being just one of them?” said Dr. Petersen. “The term AD sometimes gets us into trouble.”

Back to Top | Article Outline


“I wish the term ‘Alzheimer's disease’ would just go away,” said Clifford Jack, MD, a professor of radiology at the Mayo Clinic in Rochester, MN, who also was on the NIA-AA committee. “There is a fundamental problem in our field. There is a clinical definition of AD and a pathological one. A clinical diagnosis of Alzheimer's dementia is defined by the level of functional impairment,” Dr. Jack explained. “It doesn't matter if it is the same disease. We need terms to distinguish between different stages of the disease.

“It is no different than diagnosing someone with pre-hypertension or pre-diabetes,” Dr. Jack added. “Our field has not properly come to grips with differences between pathological signs and symptoms.”

Dr. Petersen agreed. He said that both sets of diagnostic tools will be tested to see how sensitive and specific they are in identifying AD.

Back to Top | Article Outline


Back to Top | Article Outline


•. Dubois B, Feldman HH, Jacova C, et al. Advancing research diagnostic criteria for Alzheimer's disease: The IWG-2 criteria. Lancet Neurol 2014; 13:614–629.
•. Jack CR Jr., Albert MS, Phelps CH, et al. Introduction to the recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement 2011;7(3):257–262.
•. McKhann GM, Knopman DS, Phelps CH, et al. The diagnosis of dementia due to Alzheimer's disease: Recommendations from the National Institute on Aging and the Alzheimer's Association workgroup. Alzheimers Dement 2011;7(3):263–269
•. Albert MS, DeKosky ST, Phelps CH, et al. The diagnosis of mild cognitive impairment due to Alzheimer's disease: Recommendations from the National Institute on Aging and Alzheimer's Association workgroup. Alzheimers Dement 2011;7(3):270–279.
•. Sperling RA, Aisen PS, Phelps CH, et al. Toward defining the preclinical stages of Alzheimer's disease: Recommendations from the National Institute on Aging and the Alzheimer's Association workgroup. Alzheimers Dement 2011;7(3):280–292.
© 2014 American Academy of Neurology