The move toward secondary prevention trials for Alzheimer’s disease has brought new ethical challenges, such as how to inform cognitively healthy people they have a positive amyloid PET scan. To find better ways to do this, researchers led by Jason Karlawish at the University of Pennsylvania, Philadelphia, embedded the Study of Knowledge and Reactions to Amyloid Testing (SOKRATES) as a sub-study of the Anti-Amyloid Treatment in Asymptomatic Alzheimer’s (A4) trial. In the October 23 JAMA Neurology, Karlawish and colleagues reported the first results from SOKRATES. These initial data examined how well participants understood their scan results.

The highly educated A4 cohort grasped that a positive scan indicated a higher risk for AD but was not diagnostic. However, participants expressed frustration that they were not told how high their risk was. Researchers do not yet know how to translate tracer uptake values into risk estimates. The findings suggest that researchers need to better explain the limitations of amyloid scanning, Karlawish noted. He believes these data can help other prevention trials refine how they communicate results to participants. “We are developing a clinical protocol for how to inform people of amyloid scan results and are finding out what works. There’s a lot of unknown territory,” he told Alzforum. The researchers are still collecting data on whether the scan data affected people’s decisions or mood over the following year.

In an accompanying editorial, Winston Chiong at the University of California, San Francisco, noted that this is some of the first data on how cognitively healthy people react to news of a positive scan result. “Overall, the findings of [first author Jessica] Mozersky and colleagues are broadly reassuring regarding research participants’ ability to understand the prognostic uncertainty of amyloid imaging,” he wrote. However, Chiong noted that it remains to be seen whether the findings from this select group will hold in the broader clinical population.

Previous research in the AD field had assessed the effects of disclosing genetic risk to people, and found that most handled the information well (Jun 2007 webinarJul 2009 news). Nonetheless, some researchers worried that amyloid imaging findings are more distressing, since a positive result indicates an ongoing disease process in the brain (Feb 2012 conference newsAug 2012 conference news). The SOKRATES study investigates this question (Aug 2015 conference newsSep 2016 conference news).

Mozersky and colleagues interviewed 50 A4 participants, half of them women, by telephone one to three months after the scan. All had had positive scan results. Seventy percent of them were between 65 and 74 years of age, and the rest were older. All but one were white, and 60 percent of them had attended graduate school. Eighty percent had a family history of AD. The interviewers invited these participants to explain, in their own words, what their scan results meant, and asked if they had expected the results and if the findings clarified anything for them.

Most participants understood that a positive scan was not a diagnosis of AD, a fact emphasized in the A4 educational materials. However, about 12 percent believed they had an AD diagnosis or were at imminent risk of developing the disease. Richard Caselli at the Mayo Clinic in Phoenix noted that this latter group could represent a significant number of people given the size of many prevention trials. He added that the finding concerns him because these participants received extensive education about amyloid scanning beforehand. This type of education might not be available from a primary care physician, suggesting the potential for people to misinterpret results if scanning of cognitively healthy people ever advances to general clinical practice (see full comment below). Other commenters agreed. “This reinforces to me that release of amyloid PET results in cognitively normal persons is ‘not ready for prime time’—that is, for widespread clinical application,” wrote John Ringman at the University of Southern California, Los Angeles (see full comment below). Currently, appropriate use criteria recommend against scanning people who have no cognitive problems (Jan 2013 conference news). 

Just more than half the SOKRATES participants had expected a positive scan, mostly because they had a family history of the disease or had noticed memory problems. Only 16 percent had anticipated a negative finding.

Researchers presented amyloid scan results to the volunteers as a binary outcome—either “elevated” or “not elevated.” Notably, 40 percent of participants were unsatisfied with this. They wanted more information about how high their amyloid burden was, and what that meant for their AD risk. As one person put it, “I don’t know how elevated the risk is. It could be like right over the edge, and other people are right under the edge.” Another said, “I don’t know how positive I was … that was the one thing that I would have liked to have known.”

Why not share information about the PET signal strength, known as the standard uptake value ratio (SUVR), with participants? Karlawish thinks this information could do more harm than good because researchers cannot translate an SUVR value into a clinical outcome. “We don’t know what the numbers mean in terms of what a person’s risk is, or when they will develop problems,” he told Alzforum. He fears the information would confuse or mislead people. Other groups are working on prognostic models for AD patients based on PET and other biomarker data that might provide risk estimates (Oct 2017 news). 

To better explain the value and limitations of amyloid PET to the general public, Karlawish suggests making it clearer that the scans do not provide a numerical value for amyloid burden. Study investigators would then ask if people wanted more information about how researchers interpret scans. For those interested, researchers would explain the limitation in more detail, giving them a chance to opt out of the study if they were unsatisfied. “People want to be better informed,” said Karlawish. “My hunch is it would enhance recruitment and potentially retention if they had the opportunity to learn this knowledge,” he speculated.

More data about how people react to scan results will soon be available. SOKRATES researchers are finishing up the one-year follow-up interviews on the same 50 people, and plan to publish that data in the next year or so. Those findings will examine any changes in self-image, relationships, and behaviors as a result of the scan data. The researchers will compare the findings from the 50 A4 participants to another 30 A4 recruits who tested negative on amyloid PET and enrolled in the companion Longitudinal Evaluation of Amyloid Risk and Neurodegeneration (LEARN) study. This observational study follows amyloid-negative people with yearly cognitive assessments, to provide a control group for A4.—Madolyn Bowman Rogers


  1. Overall, I think this addresses an important issue given the rise of molecular imaging. Although the authors reported that the smallest number of responses indicated “imminent” risk or were “diagnostic of” Alzheimer’s, this constituted 12 percent of the sample, which is not insignificant, especially when one considers 1) that these individuals received insightful education (which is more than one would expect from a primary care provider should this ever become more widely used clinically), and 2) the burgeoning number of people entering clinical trials that require amyloid PET.

    I also felt the authors’ recommendation to explain “why a dimensional biomarker is converted to a categorical classification” was right on target. It is not intuitive that an amyloid PET scan would simply be “positive or negative.” Amyloid buildup is certainly not all or nothing, and the patient responses are telling us that people more likely understand quantities, or some quantitative modifier (e.g., mild, moderate, severe). I was not surprised at the uncertainties the authors uncovered in the participants and it is consistent with a related study we did using just a hypothetical scenario, very brief education, and a multiple-choice questionnaire. A question we asked that was not included in the current study was, what did the patients intend to do now that they had the results? Would they discuss them with their doctors? Would they buy long-term-care insurance? Did this make them anxious or depressed, etc.? 

  2. The only surprising thing to me was that, despite extensive education on the meaning of a positive and negative amyloid scan, only 62 percent of participants appeared to correctly grasp the significance of a positive scan as putting them at an increased risk for developing AD symptoms and signs. The education they received in the course of this study is undoubtedly more than one would get in general clinical practice, so this reinforces to me that release of amyloid PET results in cognitively normal persons is “not ready for prime time”—that is, for widespread clinical application.

Make a Comment

To make a comment you must login or register.


Webinar Citations

  1. Susceptibility Testing and Risk Assessment in Alzheimer Disease

News Citations

  1. Early ApoE4 Memory Effects, But Do You Really Want to Know?
  2. Miami: Scan and Tell? Amyloid Imaging Confronts Disclosure Dilemma
  3. To Reveal or Not to Reveal? New Data on the Question
  4. How Do You Communicate Alzheimer’s Risk in the Age of Prevention?
  5. To Know or Not to Know: Trial Participants Confront the Question
  6. HAI—Amyloid Imaging in the Clinic: New Guidelines and Data
  7. Can Biomarker Data Predict Individual Risk for Alzheimer’s?

Further Reading

Primary Papers

  1. . Comprehension of an Elevated Amyloid Positron Emission Tomography Biomarker Result by Cognitively Normal Older Adults. JAMA Neurol. 2018 Jan 1;75(1):44-50. PubMed.
  2. . Challenges in Communicating and Understanding Predictive Biomarker Imaging for Alzheimer Disease. JAMA Neurol. 2018 Jan 1;75(1):18-19. PubMed.