NEW YORK (GenomeWeb) – A team led by researchers at Japan's Center for Development of Advanced Medicine for Dementia has identified a series of plasma biomarkers that could help detect and monitor Alzheimer's disease.
In a study published this week in Nature, the researchers found that ratios of amyloid-β precursor protein (APP)669–711, amyloid-β (Aβ)42, and, amyloid-β (Aβ)40 predicted brain Aβ burden, which is thought to be a key factor in Alzheimer's disease.
The results bolster previous findings that similarly demonstrated the potential of plasma markers for assessing brain Aβ levels and suggest that blood-based markers could prove useful for screening people for the disease, particularly in the context of trials looking at Alzheimer's drugs and prevention, said Katsuhiko Yanagisawa, director general of Japan's National Center for Geriatrics and Gerontology Research Institute and senior author on the paper.
Yanagisawa and his colleagues used immunoprecipitation followed by MALDI mass spec to measure levels of the three markers, APP669-711, Aβ42, and Aβ40 in two cohorts, a discovery set consisting of 121 Japanese patients and a validation set consisting of 252 Australian patients. Both sets included a mix of cognitively normal individuals, subjects with mild cognitive impairment, and subjects with clinically diagnosed Alzheimer's disease with dementia. All patients also had Aβ- positron-emission tomography (PET) imaging data, providing an assessment of their brain Aβ burdens. A subset of the Australian patients also had measurements taken of their cerebrospinal fluid Aβ levels.
The researchers tested ratios of the three markers to develop a composite blood-based marker corresponding to brain Aβ levels as assessed by PET and CSF measurements. They used ratios as opposed to absolute measurements to account for individual variation in blood Aβ42 levels across individuals, Yanagisawa said.
"Aβ42 is unique in terms of the tendency to assemble into plaques in the brain, showing reduced levels in the plasma of individuals with amyloid deposition," he said. "However, the level of Aβ42 in the plasma can be influenced by the presence of various individual-specific factors."
While Aβ40 and APP669-711 are different from Aβ42 "in terms of their assembly in the brain, they likely show the same metabolism rate, binding tendency to other protein and/or lipids, and penetration capability through the blood-brain-barrier," he added. "Collectively, plasma levels of Aβ40 and APP669-711 work as a good reference for Aß1-42 when you look at amyloid status in the brain."
A composite of the three plasma protein ratios corresponded well with both patient Aβ42 CSF levels and brain Aβ42 levels as determined by PET. Using PET measurements as the gold standard, the researchers found their plasma measurements had comparable performance to the CSF measurements, with both showing accuracy of 80.4 percent and areas under the curve of 83.8 and 87.4 for the blood and CSF measurements, respectively.
Using the CSF measurements as the gold standard, the composite plasma markers performed with an accuracy of 80.4 percent and an AUC of 87.6.
The results, the authors wrote, "demonstrate that the three different types of Aβ-related biomarker (plasma and CSF Aβ and PET imaging), are highly correlated with each other, clearly indicating that plasma Aβ biomarkers are strongly linked with the Aβ status of the CNS, but less affected by the Aβ known to be produced in peripheral tissues."
This finding reinforces other recent studies that determined plasma-based measurements of Aβ variants can identify patients with heightened brain amyloid burdens, said Yanagisawa. For instance, last year, researchers at Washington University in St. Louis published a study in Alzheimer's and Dementia that found that the kinetics of three Aβ isoforms, Aβ38, Aβ40, and Aβ42, could distinguish between patients with and without brain amyloidosis with an area under the curve of .89.
The Wash U researchers used immunoprecipitation combined with parallel-reaction monitoring mass spectrometry on a Thermo Fisher Scientific Orbitrap Fusion Tribrid instrument. Yanagisawa said, however, that he and his colleagues were able to manage more sensitive Aβ measurements using MALDI mass spec than they were LC-MS/MS.
He suggested that one possible reason for this might have been sample loss during the LC process.
"In MALDI-MS, Aβ [proteins] after the IP process are directly applied onto a MALDI sample plate without [prior] LC, which can minimize sample loss," he said.
He added that being able to use MALDI was advantageous in that, because it does not use an upfront LC separation, run-times are considerably shorter than for LC-MS/MS, which increases throughput — an important factor for assays intended for clinical environments.
Yanagisawa said he and his colleagues hoped to develop the assay as a tool for screening patients for Alzheimer's drug trials. This is a major goal of Alzheimer's biomarker development work due to the challenges of identifying patients at risk for the disease who may benefit from treatment.
Brain amyloidosis doesn't guarantee a person will develop Alzheimer's, but it is a significant risk factor for the disease and many drugs in development for Alzheimer's target amyloid plaques.
Alzheimer's drug development has a poor track record, however, with the last US Food and Drug Administration approval for an Alzheimer's therapy coming in 2003. With regard to amyloid-targeting drugs, one theory for the high failure rate is that researchers have traditionally tested these agents in patients whose disease is too far progressed for the drugs to make a difference.
This makes identification of patients with the early stages of amyloidosis a key challenge for putting together clinical trial cohorts. Imaging and CSF-based tests for screening asymptomatic patients exist, but they are expensive or invasive and difficult to do on the scale needed to recruit tens of thousands of patients for trials.
Additionally, asymptomatic patients are often reluctant to undergo a brain scan or spinal tap in order to join a clinical trial. A blood-based test could give drug researchers an easy way to do an initial scan of large numbers of patients with candidates identified by this test then passed on to imaging or CSF studies.
Moving forward Yanagisawa said he and his colleagues plan to test the markers in larger trials, including in a longitudinal study. He added that they are also looking into questions around site-to-site measurement variability.
He said they have no specific plans around commercializing the markers.