Biomarkers could change the landscape of Alzheimer’s disease (AD) forever, but we are only just scratching the surface of their potential, the audience heard during the prestigious Charles-Édouard Brown-Séquard Lecture at the European Academy of Neurology (EAN) congress in Lisbon. Delivered by Philip Scheltens from Amsterdam University Medical Center, the Netherlands, the lecture examined the past, present and future impact of biomarkers on AD and dementia.
Professor Scheltens began with a brief historical timeline of AD diagnosis. In 1984, the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) proposed several criteria for diagnosis, including memory impairment, disease progression and impacted quality of life, as well as a strategy of exclusion to rule out other possible diagnoses (e.g. tumours).
These basic symptom evaluations were supported with clinical assessments of dysfunction such as drawing a simple clockface, but Professor Scheltens cautioned that these methods did not help in understanding the biological basis of the disease.
In the same decade, biomarkers emerged with the potential of improving diagnostic capability in AD. As Professor Scheltens described, it started with fludeoxyglucose (FDG) – a radiopharmaceutical utilised as far back as 1980 in positron emission tomography (PET) scanning of AD patients. In the ‘90s, serial MRI and hippocampal atrophy markers were developed, and as the new millennium hit, amyloid- and tau-protein markers in PET took shape.
In the cerebrospinal fluid (CSF), biomarkers can be employed to investigate the neurodegenerative process. In this arena, established biomarkers including amyloid and tau sit alongside a wealth of emerging markers: “Neurofilament light [chain] is probably the most important and most promising,” said Professor Scheltens, adding that neurogranin – a synaptic marker – is also likely to play a significant role going forward.
Professor Scheltens went on to emphasise that a key step forward is to harness all of the knowledge surrounding specific biomarkers, and then incorporate them into clinical criteria. For instance, a clinical diagnosis of episodic memory impairment can be reinforced by abnormal CSF biomarkers, PET markers of glucose abnormalities or detection of amyloid. In this way, the so-called clinical-biological concept is able to disentangle AD from dementia syndrome. “So this is really one step forward to diagnose an underling disease before there are symptoms and signs of dementia,” he said.
Recent refinement of the concept has expanded its utility to different clinical phenotypes and preclinical states, but the idea is the same: “It doesn’t matter what the variant is. As long as you have the biomarkers, you can call it Alzheimer’s disease,” commented Professor Scheltens.
In April of this year, Professor Scheltens and colleagues launched the National Institute on Aging and Alzheimer’s Association (NIA-AA) Research Framework1, which underlines that AD should be diagnosed based on the biological definition of amyloid and tau. “Now we say, it doesn’t matter what the phenotype looks like, those biomarkers decide whether we call it AD or not,” he explained, adding: “This helps to identify the patients that we really should include in clinical trials.”
Biomarkers in the clinic
Professor Scheltens stressed that a crucial consideration is how biomarkers can impact clinical practice. A key advantage of biomarkers, he noted, is that they allow the distinction of AD subtypes that have the same underlying pathology. Indeed, tau deposition on brain scans correlates well with clinical phenotype, for example being found primarily in the posterior cortical area in patients with posterior cortical atrophy.
In a 2015 study2, secondary diagnosis utilising biomarkers led to a 7% change in diagnosis from a clinical-measure led primary diagnosis. What’s more, diagnostic confidence was increased by 5%, and in 13% of patients there were consequential changes to their management. The wider literature saw even greater impact, as Professor Scheltens described: “Of all the studies that have been done in the last five years in selected cohorts … the overall change in diagnosis ranges from 10% to even 55%.”
He continued: “But these are selected cohorts: we have been intrigued as to how the real impact is measured in unselected cohorts. Can you do [a] scan in every patient? And will that actually help you?”
To that end, Professor Scheltens and colleagues initiated the Dutch ABIDE study (Alzheimer’s Biomarkers In Daily Practice) which, similar to the IDEAS study in the United States (n = 18,000), has been evaluating the benefit of routine biomarker testing for every patient. In ABIDE, every patient admitted to a memory clinic was offered an amyloid PET scan, regardless of their initial diagnosis.
In total, 507 patients across three groups – dementia, mild cognitive impairment (MCI) and subjective cognitive decline – were characterised as having AD or non-AD. Similar to the previous studies, utilising biomarker testing changed the diagnostic aetiology in 19-63% of cases. The overall improvement in diagnostic confidence was 9%. Paraphrasing another editorial on the impact of the results, Professor Scheltens commented: “These results clearly demonstrated that this test is ready now for use in clinical practice, for patients with MCI and dementia.”
Could biomarkers be used to offer a prognosis over time in AD? “Based on the literature that we have, we know that 50% of patients will progress. But is this helpful for the particular patient sitting in front of you? No!” said Professor Scheltens.
In reality, he added, harnessing biomarker data allows a clinician to start predicting risk over time. For example, if the patient’s MRI shows only moderate atrophy, and if the CSF is completely normal, then the risk is reduced to closer to 15%. The converse is also true: with higher risks of dementia development at three years being seen when certain biomarkers come back positive.
A particular promising marker for disease prediction is the presence of amyloid, which is now know to precede dementia syndrome by at least 20 years. “This is good news for [drug] developers. If you want to prevent further decline you have to act very, very early, and amyloid can be a helpful tool in that sense.”
Professor Scheltens stressed that many studies have now pinpointed that amyloid presence in the brain predicts neuropsychological progression. Memory, language and executive functions will decline as soon as amyloid develops in the brain.
In terms of treatment, compounds such as aducanumab – perhaps the most advanced monoclonal antibody against amyloid beta – have been shown to target amyloid in the brain. “If you increase the dose you will see on scans that the brain is almost devoid of amyloid plaques. So it is possible to ‘clean’ the brain,” said Professor Scheltens.
Blood-based biomarkers for amyloid are also emerging, which could be incredibly useful in screening the first stages of amyloid development, and guiding earlier intervention.
“We need to stop Alzheimer’s before it starts,” said Professor Scheltens. “If patients come to the memory clinic, we are already too late. We have to find individuals before that.
“We have developed our own website, www.hersenonderzoek.nl, which has been a really dramatic success. We opened in September , and now there are more than 12,000 people already registered saying, ‘Yes: I have a family member with AD/Parkinson’s/MS – I want to participate in research’.”
Furthermore, Professor Scheltens is co-developing an app, Adappt, that allows clinicians to fill in various details about a patient and construct a predictive model for their disease. While still at the prototype stage, the potential is staggering.
Despite the progress that has been made, there is still an enormous amount of work to be done in the field of AD biomarkers, said Professor Scheltens. “There are still analytical issues, especially with CSF, there are ethical issues and reimbursement issues,” he said.
But as a whole, the biomarker future is very bright, he said, adding: “In my vision, we can actually diagnose AD, frontotemporal dementia and other diseases in the future by using biomarkers. It’s not that we don’t need clinicians anymore: clinicians are extremely important in defining the phenotype, and of course to work with these biomarkers. But for the development of new drugs and new therapies, we need to actually get rid of all of these [old] labels from very important people like Alzheimer and Arnold Pick.”
In closing, he summarised that biomarkers will almost certainly one day facilitate a personalised medicine approach, including individual prediction of disease progression, targeting of underlying proteinopathy, guiding treatment selection, and ultimately, preventing AD in its entirety.
- Jack CR Jr, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement. 2018;14(4):535-562.
- Duits FH, et al. Diagnostic impact of CSF biomarkers for Alzheimer’s disease in a tertiary memory clinic. Alzheimers Dement. 2015;11(5):523-32.