Triple Threat: Alzheimer's Biomarkers Occur in Sequence

Pauline Anderson

February 06, 2018

Alzheimer's disease (AD) biomarkers follow a sequential pattern in the brain that starts with amyloid β (Aβ) deposition more than 20 years before the expected onset of dementia and is followed by a decline in glucose metabolism and then structural brain atrophy, new research shows.

These findings could have a major impact on research, including drug trials, lead author, Brian A. Gordon, PhD, Mallinckrodt Institute of Radiology, and Knight Alzheimer's Research Center, Washington University School of Medicine, St Louis, Missouri, told Medscape Medical News.

"We hope that this knowledge can help the design of clinical trials and also help formulate new research questions to better understand this pathological cascade."

The study was published online January 31 in Lancet Neurology.

Bolstered Efforts

The long preclinical phase of AD has bolstered efforts to identify in vivo biomarkers to aid disease diagnosis and prognosis. Positron emission tomography (PET) or MRI are used to asses the amount and location of Aβ plaques, altered glucose metabolism, and structural brain changes. Imaging is also now being used to detect tauopathy, including tau-containing neurofibrillary tangles.

The new study included 88 symptomatic and 141 asymptomatic carriers of autosomal dominant mutations in genes linked to Alzheimer's disease. These include amyloid precursor protein (APP), presenilin 1 (PSEN1), and PSEN2. The study also included 70 unaffected relatives.

Participants were enrolled in the Dominantly Inherited Alzheimer's Network (DIAN). Researchers assessed dementia status by using the Clinical Dementia Rating scale.

Participants recruited for the study had to have a family history of the disease. Through structured interviews, researchers learned when a participant's parent first showed signs of dementia.

This information, along with the patient's age, allowed investigators to estimate the number of years to expected symptom onset.

Most patients had the PSEN1 mutation. Data on this and the other mutation types were analyzed together because the sample size for single mutations was small.

To assess biomarker trajectories, researchers accessed three imaging techniques. They used 11C-Pittsburgh compound B (11C-PiB) PET for Aβ aggregation, 18F-fluorodeoxyglucose (18F-FDG) PET for metabolism, and structural MRI to mainly measure cortical thickness.

The researchers did not assess tauopathy.

"The tracer used to measure tau is a relatively recent addition and has just been added to the project in the last year," explained Dr Gordon.

The research team has now collected tau-related data on about 50 carriers, but Dr Gordon noted that it's too early to determine the pattern and timing of tau-related pathology.

In the current paper, 11C-PiB PET was available for 346 individuals (162 with longitudinal imaging), 18F-FDG PET was available for 352 individuals (175 with longitudinal imaging), and MRI data were available for 377 individuals (201 with longitudinal imaging).

Individuals were followed on average for 2 to 3 years and had a mean of 2.4 visits. However, some had up to 6 visits.

"When people come close to when they're expected to become demented, we start sampling them more often," said Dr Gordon.

Triple Pattern

The researchers estimated rates of biomarker change as a function of estimated years to symptom onset at baseline using linear mixed-effects models and determined the earliest point at which biomarker trajectories differed between mutation carriers and noncarriers.

The analysis showed that the rate of Aβ accumulation was significantly higher in mutation carriers than in noncarriers, starting more than two decades before the expected age of dementia onset.

Overall, in regions with a significant difference between mutation carriers and noncarriers, rates of Aβ deposition were significantly higher in carriers, at a mean of 18.9 years before expected symptom onset.

Metabolism began to differ between carriers and noncarriers at a mean of 14.1 years before expected symptom onset. Finally, MRI structural measures differed between carriers and controls at a mean of 4.7 years before expected symptom onset.

The researchers characterized these trajectories throughout the brain. Aβ aggregation started in the precuneus, followed shortly in the medial orbitofrontal cortex, before spreading to neighboring areas.

In the precuneus, there's this "triple pattern" with first amyloid and then hypometabolism followed by structural atrophy, said Dr Gordon.

"It's a very consistent pattern," he said. "It's the area of the brain that is always affected earliest and it also shows the highest degree of abnormality."

The researchers examined the consistency of this pattern across other brain regions to determine whether every brain region shows the same triple pattern.

"We found that the majority of brain regions do, but not all of them," said Dr Gordon. "Some regions won't show hypometabolism, the drop in metabolism, but show both amyloid and volumetric loss or cortical thinning."

Brain areas that exhibit this pattern include the occipital and temporal regions.

In some regions, for example, the medial and lateral temporal areas, atrophy starts before Aβ aggregation and glucose hypometabolism.

Implications for Future Drug Trials

Rather than being homogeneous, the same biomarker often shows different longitudinal trajectories across brain regions, noted the authors.

Although the numbers were small, the researchers looked at the AD genes individually and "didn't see any obvious patterns that differed" from the findings of the genes analyzed together, said Dr Gordon.

"But we don't have quite enough people to definitively say that they don't differ; there may be some subtle differences that we can't detect with the number of people."

The authors believe this is the first study to examine both the longitudinal temporal trajectories and the spatial patterns of AD pathology in autosomal dominant AD cohorts using neuroimaging.

This information could be useful for selecting participants for clinical trials, as well as for measuring outcomes.

"From a practical standpoint, if you're going to design a drug trial, you need to know the way that the different biomarkers we can measure become abnormal, and how they behave over time," said Dr Gordon.

A drug that stops amyloid accumulation might also protect metabolism and brain structure. So with an early intervention with an antiamyloid agent, researchers would want to not only measure amyloid accumulation but also look at whether the agent preserves metabolism or brain structure.

Trials now being carried out will evaluate alterations in longitudinal biomarker trajectories. These trials include the Alzheimer's Prevention Initiative (API) and the Anti-Amyloid Treatment in Asymptotic Alzheimer's Study (A-4) in addition to DIAN.

The information from this new study is also important to further inform how the disease affects the brain, said Dr Gordon.

"We know there's something about amyloid accumulation that's leading to a drop in metabolism, likely due to cellular death and dysfunction, which is then leading to structural atrophy. Knowing this temporal pattern can point us toward research questions that people might want to look at, in animal models or cell lines, to figure out how the disease actually works."

Generalizable Findings?

In an accompanying editorial, Betty M. Tijms and Pieter Jelle Visser from the Alzheimer Center and Department of Neurology, VU University Medical Center, Amsterdam, the Netherlands, note that only about 1% of patients with AD have an autosomal dominant mutation. The new study raises the question of whether the results also apply to sporadic AD.

Dr Gordon said the temporal sequencing pattern in patients with sporadic AD "roughly parallels" that uncovered in gene carriers.

As for the spatial patterns, which regions are most affected, "these overlap but are not identical between the two forms of AD," he said.

Keith Fargo, PhD, director of scientific programs and outreach, medical and scientific relations, Alzheimer's Association, pointed out that the large ongoing Alzheimer's Disease Neuroimaging Initiative (ADNI) trial is measuring biomarkers in people with the sporadic form of AD and should provide additional related information.

Another study, the Longitudinal Early-onset AD Study (LEADS), which was launched last year, will investigate patients who develop dementia symptoms before age 65 years. Many such patients don't carry known mutations for the disease. 

With results of all three studies (DIAN, ADNI, and LEADS), researchers will be able to answer several questions, including whether the pattern is the same in different patient populations and how long the disease takes to develop in the various groups, said Dr Fargo.

"These are important questions to answer because we want medications and other interventions that will work for everyone at risk for AD."

Dr Fargo described the current study as "an excellent addition to the literature" that was carried out by a "top-notch research group."

"It's really showing how these biomarkers relate to one another over time and helps us understand the development of the disease," said Dr Fargo.

Earlier smaller studies had investigated this area but didn't look at all three biomarkers or look at them longitudinally, said Dr Fargo.

"This was the first time that it was really all put together, in more than just a theoretical model with just a little bit of evidence here and there."

Confirming that amyloid begins to build up essentially two decades before dementia onset "really opens a wide window of possible preventive therapeutics," said Dr Fargo.

However, not including a tau biomarker was "a missing piece" in the study, said Dr Fargo. Having that information "will add even more to the picture," and the scientific community "will be looking forward to seeing that" information when available, he said.

The new results underline the need for trials to have stage-specific biomarker endpoints because "the start, rate of decline, and location of changes vary during the disease," the editorialists write.

In addition, they note that the study calls for refinement of present research classification schemes for disease staging, including categories for Aβ pathology, tau pathology, and neuronal injury.

"Hippocampal atrophy and glucose hypometabolism are often used interchangeably as markers for neuronal injury, but given their different trajectories, they should be considered separate entities."

Dr Gordon reports participating in a clinical trial of AV-1451 sponsored by Avid Radiopharmaceuticals. Dr Tijms reports grants from ZonMw and Biogen. Dr Visser reports grants from the Innovative Medicine Initiative, ZonMw, and Biogen; nonfinancial support (in-kind supply of an amyloid tracer) from GE Healthcare; and fees for consultancy from Eli-Lilly and Janssen Pharmaceutica. 

Lancet Neurol. Published online January 31, 2018. Abstract, Editorial

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.