Drawing on Blood for Biomarkers of Alzheimer Disease

The new funding initiatives and innovative programs to expedite research efforts could hasten the search for biomarkers of Alzheimer disease (AD) in the blood and, ultimately, open the door to easier, quicker, and more affordable diagnostic methods. These initiatives have focused mainly on finding traces of proteins often associated with AD in the blood, such as tau, beta-amyloid (Aß), and neurofilament light (NFL), though some have turned their sights toward identifying new targets. One such project, the Alzheimer’s Drug Discovery Foundation’s (ADDF) Diagnostic Accelerator program, which was launched in July 2018 and seeks to develop novel biomarkers for early detection of AD and related dementias, and will provide more than $30 million in grants to investigators over the next 3 years.

Howard Fillit, MD, the ADDF’s founding executive director and chief science officer, described the importance of the programs in a January announcement of the Foundation’s January 2019 $2.5 million commitment to the initiative, matched by the Association for Frontotemporal Degeneration, saying that “to develop treatments for dementia, we need to be able to diagnose it as early as possible and, just as important, determine specifically what type of dementia we’re dealing with. That requires biomarkers that are both sensitive and specific.”

Fillit explained the priority of identifying biomarkers in blood in an interview with NeurologyLive. “The advantage of the blood test is that it’s noninvasive and, hopefully, will be much cheaper than the neuroimaging technologies that are out there right now and the spinal fluid tests that require, obviously, a spinal tap, but are also quite expensive,” he said.

Regarding research directions that might prove most productive, Fillit described both the targets with demonstrated links to the disease and others that fit into intriguing hypotheses.

“Certainly, at this point of research progress to date, we know that [Aβ] and tau are pretty validated biomarkers for AD—both on neuroimaging and in spinal fluid—and the field is in the process of evaluating whether a blood test for Aß and tau can be made. Those define the basic pathology of AD are also drug targets for the disease,” Fillit explained.

“I think the next generation of biomarkers that are in development now target not just these pathways—Aß and tau and these pathologies—but are going to be directed to new biomarkers, toward other pathologies seen in the disease, such as α-synuclein, TDP-43 [transactive response DNA-binding protein 43] and others that represent pathology,” he added.

Fillit noted that it now appears that most affected individuals have mixed pathologies. He said that most patients do not have pure Aß plaques or pure tau tangles. He clarified that 30% to 35% of individuals might have α-synuclein whereas 10% to 15% might have TDP-43, and each is occurring in a mix in each patient. “And these all represent misfolded proteins in the brain and are potential drug targets,” he said.

Tempering Enthusiasm

Results from new reports of potential biomarkers for AD in the blood suggest progress in the search for signals of the disease in a more accessible medium than cerebrospinal fluid (CSF). However, challenges exist, and the studies could mark only tentative progress in the view of experts recently invited to assess developments in the field.

Thirty-five review articles were collected in an issue of the Journal of Alzheimer’s Disease, and additional papers anticipating future trends were presented in a supplement issue in 2018. The progress and challenges in developing a clinically viable blood test were described in the collection, and recommendations were offered on directions and methods of future research.

In one paper, Steven Kiddle, PhD, of the Medical Research Council of the Social, Genetic and Developmental Psychiatry Centre in the Institute of Psychiatry, Psychology & Neuroscience at King’s College London in the United Kingdom, and colleagues pointed out that although there has been some degree of consis­tency and replication across independent studies, there has also been a failure to replicate several high-profile findings.¹

“Our reviews have shown that more than 1404 studies have been done on blood markers of Alzheimer’s disease, and despite that, no blood test is currently being used in the clinic,” Kiddle told NeurologyLive. “This suggests that progress is a lot slower than press releases would suggest.”

Fillit agreed that not all research will be fruitful but emphasized the importance of supporting both exploratory and clinical research of promising blood test candidates. Facilitating development from early research to robust assays that can ulti­mately be commercialized is the main purpose of the Diagnostics Accelerator program, he indicated.

“The really critical thing about [early biomarker candidates] will be that they’ll be studied both in exploratory, phase 1 studies where novel biomarkers will be tested, and phase 2 validation studies where biomarkers that have some preliminary data assays will then be tested in large cohorts of patients,” Fillit said. “Assays will be refined with the goal of moving these blood tests into commercialization, and we’ll also be funding some of the efforts in commercialization of these tests.”

Finding Focus for Discoveries

One factor that could contribute to the profusion of findings with potential but uncertain clinical application and perhaps overly optimistic press releases is the absence of an explicit context of use in much of the research on potential biomarker substances.

In the most recent and first update in 12 years, the consensus paper of the Task Force on Biological Markers in Psychiatry from the World Federation of Societies of Biological Psychiatry calls on investigators to incorporate context of use (COU) to increase the relevance of their findings.² The term is taken from the FDA/ National Institutes of Health BEST (Biomarkers, Endpoints, and Other Tools) Resource, which defines the COU as “a statement that fully and clearly describes the way the medical product develop­ment tool is to be used and the medical product development-re­lated purpose of the use.”³

The task force urges incorporation of COU into research objec­tives and admonishes research efforts that have proceeded without this lodestar guide. “Explicitly defining the COU with the end goal in mind guides the entire development program of the biomarker itself; however, a key limitation to the research field has been the lack of explicit outlining of these potential COUs, which results in continuous discovery studies that rarely move beyond initial clin­ical replication,” the task force advised.²

Kiddle and colleagues concurred with the need for research to be guided by COU and with the priority of targeting a blood test that could be helpful in the diagnosis of AD. They pointed out that none of the blood sampling has been drawn from the primary health care population in which it would have the greatest utility. This, they suggested, reflects the challenges of recruiting research partici­pants and the priority given to large-scale recruitment rather than representatives of populations relative to anticipated COU.

An additional problem related to an unrepresentative sample population, they indicated, is that it may not reflect the prevalence of AD phenotypes in populations appropriate to anticipated COU, which could inflate the positive and negative predictive values.

Kiddle and colleagues also faulted the system of incentives in academic research, which left investigators reluctant to publish negative findings and to allocate resources to the replication of the work of others. They recommended that raw data be shared as widely as possible, allowing independent statisticians to verify reproducibility. They cited several programs that have been successful in facilitating data sharing, particularly highlighting the Alzheimer’s Disease Neuroimaging Initiative.

“More funding needs to go toward rigorous assessment and stan­dardization of blood tests, and in my opinion, less [should go] to the development of new markers in small studies,” Kiddle commented to NeurologyLive.

Fillit acknowledged that the challenges can be daunting and that major funding and collaborative initiatives, like the Diagnostics Accelerator program, can facilitate the identification and stan­dardization of biomarkers in blood. He took issue, however, with the suggestion that the path to identifying viable biomarkers for AD should necessarily be more direct than the circuitous and often disappointing processes that can precede medical research breakthroughs.

“That is the nature of research,” Fillit said. “The challenge is [that] we do need new ideas. We need to investigate those new ideas to see whether these [putative] biomarkers have any kind of signal in them. These are the early stages. We need assays to detect the biomarkers, and then as the process goes, once we get a possible signal, we need to refine the assays, we need to make them repro­ducible, and we need to make them standardized. Ultimately, we need to make them standardized to the point where they can be commercialized for large populations.”

Realizing Reproducible Results

In another of the collected reviews, Simon Lovestone, PhD, MRCPsych, of the Department of Psychiatry at the University of Oxford in the United Kingdom, and colleagues suggested that blood biomarker research often fails to be repli­cated because of the heterogeneity of the disease itself as well as the complexity of blood.⁴

They noted that the many fractions of blood influence the concentration of proteins. Each of the cellular and noncellular compartments of blood—eg, red and white blood cells, platelets, serum, and plasma—might be the source of biomarkers, and each has been used in Alzheimer biomarker studies.

“Indeed, the protein levels in blood span 10 orders of magnitude, making the investigation of lower abundant proteins extremely challenging,” Lovestone and colleagues indicated.

Even as clues to disease pathophysiology in the CSF guide the search in the blood, that more accessible medium has remained more impenetrable for detecting biomarkers, according to a review by Kaj Blennow, MD, PhD, and Henrik Zetterberg, MD, PhD, from the Clinical Neurochemistry Laboratory at the Institute of Neuroscience and Physiology in the Sahlgrenska Academy at the University of Gothenburg in Mölndal, Sweden.⁵

“[Whereas] the CSF is continuous with the brain extracellular fluid, with a free exchange of molecules from the brain to the CSF, only a fraction of brain proteins enter the bloodstream,” Blennow and Zetterberg explained.

Although they could foresee the difficulty in measuring minute amounts of brain proteins within a matrix of very high levels of plasma proteins, ultimately yielding to devel­opments in ultrasensitive immunoassays and mass spectrometry, Blennow and Zetterberg remained cautious about prospects for identifying substances in blood that reflect the pathology in brain. They pointed out, for example, that though there have been numerous papers associating CSF Aβ42 peptide with senile plaque deposits, studies on plasma Aβ42 have yielded contradictory results and no significant differentiation between patients and controls. Blennow and Zetterberg suspected the lack of association could be the result of the contribution of Aβ in plasma from peripheral tissue, as well as shortcomings in the bioanalytical methods.

Other leading candidates of biomarkers in blood suggested from CSF studies include tau protein and NFL. Blennow and Zetterberg found that current data indicate a minor increase of plasma tau in AD but that there is, again, too large of an overlap with controls to be considered diagnostically useful. Although NFL is the second-most AD-associated biomarker identified in the AlzBiomarker database,6 they noted that it is not AD-specific but linked to various neurodegener­ative disorders.

“Thus, a possible future application for plasma NFL is as a screening test at the first clinical eval­uation of patients with cognitive disturbances,” Blennow and Zetterberg suggested. “Here plasma NFL might serve as a simple, noninvasive, and cheap screening tool, primarily to rule out neurodegeneration.”

Researching Targets From Known to Novel

Research on potential AD biomarkers in blood presented at the Alzheimer’s Association International Conference, held in Chicago, Illinois, from July 22 to 26, 2018, provided new data on known candidates and explored novel pathways and even tenuous associ­ations. Select presentations are described henceforth.

Sebastian Palmqvist, MD, PhD, of the Department of Neurology at Skåne University Hospital in Lund, Sweden, and the Clinical Memory Research unit at Lund University in Malmö, Sweden, and colleagues examined multiple putative and potential biomarkers in CSF and plasma, as well as with magnetic resonance imaging and positron emission tomography (PET) at various stages of fibrillar Aβ deposit and cognitive health. They sought to explore how these may change from the initial onset of Aß to symptomatic AD.⁷

Cross-sectional data for 386 subjects were obtained from the Swedish BioFINDER study, including 136 who were cognitively healthy, 113 with subjective cognitive decline, and 137 with mild cognitive impairment (MCI). The substances in plasma were Aβ42, Aβ40, tau, and NFL. Palmqvist reported that in both plasma and CSF, Aβ42 was the first to change with deterioration, followed by the ratio of Aβ42 to Aβ40. Tau measured in CSF was next to change, but there was no inflection point for tau in plasma. Change in NFL in both CSF and plasma followed when the Aβ PET signal became abnormal.

“The results highlight the onsets of different pathophysiological mechanisms in AD. A combi­nation of several biomarkers could thus improve the accuracy of staging the AD process,” Palmqvist said. “The slopes of the biomarkers also indicate that repeated measures might further improve this accuracy.”

Akinori Nakamura, MD, PhD, of the National Center for Geriatrics and Gerontology (NCGG) in Obu, Japan, and colleagues reported on testing a composite plasma biomarker, which they had previously determined to have an association with brain status of Aβ, as a means to screen for AD in a general population.⁸ They accessed 2 independent and large cohorts—one in Japan, the NCGG (n = 121), and the other in Australia, the Australian Imaging, Biomarker & Lifestyle Flagship Study of Ageing (n = 111). The cohort included 125 cognitively normal individuals, 63 with MCI, and 44 with AD.

Pittsburgh Compound B imaging tracer—PET (PiB-PET) was used to identify brain amyloid deposits, and immunoprecipitation and mass spectrometry established plasma ratios of amyloid precursor protein (APP669-711) to Aβ1-42 and Aβ1-40 to Aβ1-42. Receiver operating characteristic analysis revealed concordance between PiB-PET and the composite plasma biomarker of 87.5%.

“The plasma biomarker could reliably detect individuals within the AD continuum and therefore [be] expected to be efficacious for population screening,” Nakamura concluded.

Another study seeking to relate the change in puta­tive biomarkers to increasing brain amyloid burden and AD progression examined 33 distinct proteins in plasma. This targeted plasma proteomic anal­ysis was conducted in amyloid-positive and -negative cognitively healthy elderly individuals (n = 400), in individuals with MCI (n = 400), and in patients with mild AD dementia (n = 200), iden­tified in the European Medical Information Framework database.

Sarah Westwood, PhD, of the University of Oxford in the United Kingdom, and colleagues indicated success in identifying a panel of plasma proteins that could significantly discriminate between amyloid-positive and -negative individuals, as well as relating patterns in proteins to cognitive decline and brain atrophy.⁹

“A minimally invasive and cost-effective blood biomarker of AD pathology could facilitate clinical trials by contributing to rapid and effective selection of research participants most likely to have neocortical Aβ burden and hence improve prospective participants’ recruiting experience, reduce screen failure rates, as well as significantly reduce cost and time of trial start-up,” Westwood said.

Shannon Risacher, PhD, an assistant professor in the Department of Radiology and Imaging Sciences at the Indiana University School of Medicine in Indianapolis, and colleagues evaluated the association of plasma Aβ with both cerebral amyloid and tau deposition by using amyloid and tau PET imaging and assays of free and total plasma Aβ42 and Aβ40 in 47 elderly adults who were cognitively normal (n = 20), had MCI (n = 11), or had been given a diagnosis of AD (n = 16).¹⁰

Risacher et al reported lower ratios of free plasma Aβ42 to Aβ40 and of total plasma Aβ42 to Aβ40 associated with increased cerebral amyloid and tau disposition. Specifically, the lower free component ratio was associated with amyloid deposition in the bilateral frontal lobe and tau deposition in the bilateral temporal lobe; the lower total component ratio was associated with amyloid deposition in the frontal, parietal, and temporal lobes and tau deposition in the bilateral temporal lobes, the left anterior cingulate, and the frontal lobe.

“The plasma Aß showed good association with amyloid and tau in the cortex,” Risacher said. “Given these results, these analytes show promise for development as blood-based biomarkers of cerebral amyloid and tau.”

The possible utility of serum NFL was considered in several investigations. Raquel Sánchez-Valle, MD, PhD, of the Alzheimer’s Disease and Other Cognitive Disorders Unit at the Hospital Clinic of Institut d’Investigacions Biomèdiques August Pi i Sunyer in Barcelona, Spain, and colleagues examined its relation to severity measures and neurodegeneration markers in the CSF of patients with autosomal dominant AD (ADAD).¹¹

The serum NFL was measured in 60 individuals from families with ADAD with the ultrasensitive immunoassay on the single-molecule array platform. Forty-two participants were mutation carriers, including 22 who were demonstrating symptoms. Those who were symptomatic showed statistically significantly higher serum NFL levels than those who were asymptomatic, and there was no significant difference between asymptomatic participants and individuals without ADAD.

“Serum NFL might be a feasible noninvasive biomarker to track disease onset and progression in ADAD,” Sánchez-Valle said.

Another examination of serum NFL considered its relation to established AD neuroimaging biomarkers. Stephanie Schultz, BSc, of the Washington University School of Medicine in St Louis in Missouri, presented results from a multimodal, longitudinal study in familial mutation carriers and noncarriers from the Dominantly Inherited Alzheimer Network, including an ADAD cohort.¹²

Schultz reported that the cross-sectional relationship between serum NFL and cortical thickness, glucose metabolism, and Aβ burden was different between mutation carriers and noncarriers. Among the mutation carriers, higher baseline serum NFL levels were associated with lower cortical thickness, hypometabolism, and higher Aβ burden. In longitudinal analysis, the increased rate of change in serum NFL was also associated with increased cortical thinning, hypometabolism, and Aβ accumulation.

“Importantly, this suggests that serum NFL has immense potential to monitor disease progression in neurodegenerative diseases such as AD and should be considered in future clinical trials,” Schultz said.

Presented studies that went further afield in search of candidate substances in plasma and serum included examinations of fibrinogen, C-reactive protein, oxysterols, lysophosphatidic acids, and primary fatty amides. With the increasing capabilities of assays and analyses, entire groups of substances were also assessed for possible patterns that could be linked to AD, including serum lipids and plasma-free amino acids and amines.

In an introduction to a report on plasma primary fatty amides, Min Kim, a student in a research group that included Zetterberg and Lovestone, offered what could serve as the rationale for these wide-ranging investigations.

“A critical and as yet unmet need in AD is the discovery of peripheral molecules that can act as biomarkers in the early stages of AD. Given that brain pathology precedes clinical symptom onset, small molecule markers associated [with] pathology could provide new drug targets and molecules that are relatively easy to measure,” Kim indicated.¹³

Fillit anticipates that the most promising of these early findings will be subject to validation in large population studies. “They need to be validated in large population studies where the patients who are under study are very well clinically phenotyped with neuroimaging and spinal fluid so we can correlate what we’re looking at in the blood with what is actually going on in the patient’s brain,” Fillit told NeurologyLive.

“Therapy will then, ultimately, be tailored to the patient’s biomarker phenotype in well-validated assays in primary or specialty care,” Fillit predicted.

REFERENCES

1. Kiddle SJ, Voyle N, Dobson R. A blood test for Alzheimer’s disease: progress, challenges and recom­mendations. J Alzheimers Dis. 2018;64(suppl 1):S289-S297. doi: 10.3233/JAD-179904.

2. Lewczuk P, Riederer P, O’Bryant SE, et al. Cerebrospinal fluid and blood biomarkers for neurodegen­erative dementias: an update of the consensus of the Task Force on Biological Markers in Psychiatry of the World Federation of Societies of Biological Psychiatry. World J Biol Psychiatry. 2018;19(4):244- 328. doi: 10.1080/15622975.2017.1375556.

3. FDA-NIH Biomarker Working Group. BEST (Biomarkers, Endpoints, and Other Tools) Resource. Silver Springs, MD: US Food and Drug Administration; Bethesda, MD: National Institutes of Health; 2016.

4. Shi L, Baird AL, Westwood S, et al. A decade of blood biomarkers for Alzheimer’s disease research: an evolving field, improving study designs, and the challenge of replication. J Alzheimers Dis. 2018;62(3):1181-1198. doi: 10.3233/JAD-170531.

5. Blennow K, Zetterberg H. The past and future of Alzheimer’s disease fluid biomarkers. J Alzheimers Dis. 2018;62(3):1125-1140. doi: 10.3233/JAD-170773.

6. Olsson B, Lautner R, Andreasson U, et al. CSF and blood biomarkers for the diagnosis of Alzheimer’s disease: a systematic review and meta-analysis. Lancet Neurol. 2016;15(7):673-684. doi: 10.1016/ S1474-4422(16)00070-3.

7. Palmqvist S, Mattsson N, Strandberg O, et al. CSF, plasma and MRI biomarker trajectories during the development of Alzheimer’s disease. Oral presentation at: Alzheimer’s Association International Conference; July 25, 2018; Chicago, IL. alzheimersanddementia.com/article/S1552- 5260(18)32860-7/pdf. Accessed December 27, 2018.

8. Nakamura A, Kaneko N, Villemagne VL, et al. Plasma biomarker with high accuracy in predicting brain amyloid-ß burden: initial results across two independent large cohorts—NCGG (Japan) and AIBL (Australia). Paper presented at: Alzheimer’s Association International Conference; July 24, 2018; Chicago IL. alzheimersanddementia.com/article/S1552-5260(18)31773-4/pdf. Accessed December 28, 2019.

9. Westwood S, Baird AL, Anand SN, et al. Discovery, replication and extension study of plasma proteomic biomarkers relating to brain amyloid burden and Alzheimer’s disease and progression. Paper presented at: Alzheimer’s Association International Conference; July 22, 2018; Chicago, IL. alzheimersanddementia.com/article/S1552-5260(18)32480-4/pdf. Accessed December 27, 2018.

10. Risacher SL, Fandos N, Romero J, Sherriff I, Saykin AJ, Apostolova LG. Association between plasma Aβ levels and cerebral amyloid and tau deposition. Paper presented at: Alzheimer’s Association International Conference; July 25, 2018; Chicago, IL. alzheimersanddementia.com/article/S1552- 5260(18)33356-9/pdf. Accessed December 26, 2018.

11. Sánchez-Valle R, Heslegrave A, Foiani MS. Serum neurofilament light levels correlate with severity measures and neurodegeneration markers in autosomal dominant Alzheimer’s disease. Paper presented at: Alzheimer’s Association International Conference; July 24, 2018; Chicago, IL. alzhei­mersanddementia.com/article/S1552-5260(18)32993-5/pdf. Accessed December 28, 2018.

12. Schultz SA, Apel A, Preische O, et al. Serum neurofilament light chain levels are associated with cortical thickness, beta-amyloid burden, and cerebral glucose metabolism in autosomal dominant Alzheimer disease. Paper presented at: Alzheimer’s Association International Conference; July 21, 2018; Chicago, IL. alzheimersanddementia.com/article/S1552-5260(18)32222-2/fulltext. Accessed December 28, 2018.

13. Kim M, Snowden SG, Ahmad T, et al. Plasma primary fatty amides associate to CSF amyloid levels and Alzheimer’s disease progression in the EMIF-AD biomarker discovery cohort. Paper presented at: Alzheimer’s Association International Conference; July 24, 2018; Chicago, IL. alzheimersandde­mentia.com/article/S1552-5260(18)31764-3/pdf. Accessed December 28, 2018