Better Biomarkers for AD R&D
Priming diagnostics for the future of Alzheimer’s disease research and treatment
Carlo Medici |
At a Glance
- Thus far, drug development for Alzheimer’s disease has not seen great success
- Clinical trial failures suggest we must catch and treat early – but at the moment, we have no effective way to detect pre-symptomatic disease
- We need valid biomarkers for Alzheimer’s disease, not only to detect it, but also to stratify patients and monitor treatment success
- Even with fluid biomarkers showing early promise, brain imaging of disease and treatment response will remain essential
Alzheimer’s disease has been a minefield for recent drug development, marked by a series of clinical trials whose failures suggest that therapeutic interventions will need to be directed earlier in the disease process. How much earlier? Well, before the onset of dementia – in fact, ideally before any symptoms manifest at all. Complicating the search for drugs to treat the disease is the fact that, in most instances, the symptomatic decline in affected individuals is slow. For a clinical trial to show a definite effect on cognition, it may need to run for many years. At this point, though, such clinical trials are some way in the future. First, to identify and ultimately treat the disease at a pre-symptomatic stage, we need valid biomarkers for Alzheimer’s disease. Biomarkers are critical to stratifying and monitoring patients for clinical trials in drug development. Then, once treatments are available, biomarkers are needed to screen patients and confirm diagnoses in the clinic – and also to properly direct the use of drugs and track responses to therapy.
The current situation in Alzheimer’s disease research is similar to one that, until recently, existed in the field of multiple sclerosis (MS) – another disease where the clinical diagnosis is frequently neither sensitive nor specific for the pathology. In MS, magnetic resonance imaging (MRI) is now invaluable for detecting and monitoring brain lesions associated with demyelination, which often present far earlier than clinical symptoms. Once the detection of such lesions via MRI of the brain became well-accepted as a biomarker for disease diagnosis and monitoring, drug research advanced quickly to the point where, today, there are 16 approved disease-modifying drugs for MS (not to mention many more that are used off-label to manage severe relapses or to treat specific symptoms).
These twin factors – realization of the need for early therapeutic intervention in Alzheimer’s disease and the complexities of drug development under a scenario where clinical diagnosis is inadequate – have led the US Food and Drug Administration (FDA) to propose a biomarker-based approach to defining the illness that could guide new drug development efforts. In February 2018, the agency issued a new set of draft guidelines (1) that indicate its openness to a drug approval pathway based on surrogate biomarker measurements – signals that indicate a drug candidate is working as intended, even before cognitive benefit can be measured. The FDA believes that such a biomarker-driven approach could provide a new foundation for studies to find drugs that help treat, or even prevent, the onset of symptoms in people who are unknowingly in the early stages of Alzheimer’s disease. Once a drug is approved, the new draft guidelines suggest the developer would then need to conduct further studies to confirm that it offers a benefit – in particular, an effect on cognition. A biomarker-based approach like this is not a radical idea; statins were developed based on cholesterol lowering as a biomarker even as research continued to establish the drugs’ long-term benefit on cardiovascular disease and heart attacks.
Experts today recognize that the pathophysiological changes that ultimately result in Alzheimer’s dementia are the result of a multifaceted process that begins many years before symptoms appear. Affected individuals advance along a seamless continuum from asymptomatic to severely impaired. Although the pathology of Alzheimer’s disease is still not completely understood, there are three general groups of biomarkers with recognized associations to the disease:
- Biomarkers related to beta-amyloid plaques - Amyloid plaques form in the brain when beta-amyloid protein fragments (Aβ) clump together and build up between cells. Some experts believe the most damaging form of Aβ may be smaller aggregates of a few pieces, rather than the larger plaques themselves. Such small aggregates may block cell-to-cell signaling or activate immune cells that trigger inflammation in the brain. A number of experimental drugs targeting the formation of amyloid plaques have failed in clinical trials, causing some to question the role of amyloid as an essential part of the disease process. However, extensive human genetic evidence exists to support the importance of amyloid in Alzheimer’s disease pathology. Familial Alzheimer’s disease and Down’s syndrome (in which affected individuals have a high risk of developing a type of dementia that closely resembles Alzheimer’s) are both associated with genetic abnormalities that increase the formation and deposition of Aβ. In sporadic Alzheimer’s disease, it is well known that ApoE4, associated with a higher risk of the disease, decreases the clearance of Aβ. Moreover, the A673 mutation in the amyloid precursor protein (APP) is known to reduce the cleavage of APP by beta secretase into Aβ42, thus decreasing the risk of Alzheimer’s (2)(3).
- Biomarkers related to pathologic tau protein and aggregated tau “tangles” - Tau proteins regulate the assembly and structural stability of microtubules in the neurons. In Alzheimer’s disease, tau becomes abnormally phosphorylated, which causes the pathologic tau molecules to twist and aggregate into “tangles.” These tau tangles disrupt the microtubules’ transport function so that nutrients and other essential materials can no longer move through the cells, which eventually die.
- Biomarkers of neuronal degeneration or neuronal injury - Although not specific to Alzheimer’s disease, a number of potential biomarkers related to neurodegeneration or brain inflammation are under study as potentially useful markers for Alzheimer’s research and drug development. These biomarkers are useful for helping to understand disease progression and severity, but must be used with caution because they are also associated with other brain pathologies. Only Aβ and pathologic tau are specific indicators of Alzheimer’s disease and could be considered as potential biomarker definitions of the disease. For this reason, the National Institute on Aging and the Alzheimer’s Association have proposed a biomarker-based definition of Alzheimer’s involving these indicators for research purposes (4), rather than defining the disease based on clinical symptoms. Additionally, although many interesting potential biomarkers are under study, only Aβ and tau are either available or close to available for practical use in clinical trials of potential disease-modifying drugs for Alzheimer’s disease – or for potential future use in clinical medicine. Given the likely need to intervene in the pathologic processes that result in Alzheimer’s dementia well before symptoms emerge, and the failure to date of potential treatments focused solely on Aβ and amyloid plaque formation, it is likely researchers will need to develop a multiple-biomarker approach to Alzheimer’s diagnosis, drug development, and the measurement of response to potential treatments.
Brain imaging and Alzheimer’s
Positron emission tomography (PET) is an imaging modality that uses radioactive tracers to examine targeted tissue and organs. The FDA has approved three PET tracers for the imaging of Aβ in the setting of Alzheimer’s disease: florbetapir, flutemetamol, and florbetaben. PET imaging of brain amyloid deposits can successfully detect Alzheimer’s pathology even before cognitive symptoms emerge – and, conversely, the lack of such deposits enables Alzheimer’s disease to be ruled out as a likely cause of cognitive problems, speeding the search for other reasons for a patient’s symptoms.
Amyloid PET scans are already playing an important role in clinical research. A positive cortical amyloid PET scan is now required prior to patient enrollment in many drug trials, thanks to several studies showing that over 60 percent of patients diagnosed with Alzheimer’s based on clinical assessments were actually amyloid-negative on PET scan (5–10). The appropriate role of amyloid PET scans in clinical medicine is less clear-cut, and is the subject of the ongoing Imaging Dementia – Evidence for Amyloid Scanning (IDEAS) study to assess the clinical usefulness and impact on patient-oriented outcomes of brain amyloid PET scanning in patients with mild cognitive impairment or dementia of uncertain cause (11). The study, which has enrolled over 18,000 patients, is expected to generate sufficient data to assess whether amyloid imaging has a positive impact on patient outcomes and thus should be reimbursed by Medicare and other third-party payers. The IDEAS study is also expected to lay the groundwork for the type of information regulators and payers will need when considering coverage for future Alzheimer’s biomarkers.
Despite its current utility in clinical research, amyloid PET has significant drawbacks that are likely to limit its widespread use in clinical medicine. PET scans employ radioactive tracers that expose patients to the equivalent of approximately 40 to 70 chest X-rays during a single scan. As a result, patients can only safely undergo this procedure a limited number of times, making PET unsuitable for tracking disease progression or response to treatments over time. PET imaging is also an extremely expensive procedure, with an average cost of US$6,000 or more, which makes it prohibitively costly for patient screening. Moreover, PET scans require the use of cyclotrons (unavailable in the majority of medical centers worldwide), making them a difficult procedure to access.
MRI is widely used for the imaging of soft tissues, and is currently the standard imaging test for brain disorders. MRI offers several advantages over PET in that does not expose patients to radiation, offers higher resolution scans, and is significantly lower-cost (less than 20 percent of the per-scan cost of PET). Additionally, MRI equipment is widely available throughout the world, with 10 to 20 times as many scanners available as PET cyclotrons. Structural MRI has long been used to examine brain atrophy in the course of diagnosing and monitoring the progression of Alzheimer’s disease – but structural changes have not historically been useful in clinical diagnosis. Why? Chiefly because the structural changes (in particular, brain atrophy) noted on conventional MRI scans are not exclusive to Alzheimer’s disease. Such structural changes can be used to distinguish clusters of patients, but are not necessarily useful for individual patient diagnosis. Additionally, significant loss of brain tissue on conventional MRI would be detected late in the course of the disease, when treatment would be less likely to help.
One industry/academic partnership is working to develop intravenous contrast agents for the MRI imaging of Aβ and tau. The technology is based on liposomal nanoparticles, carrying a ligand and an MRI contrast agent, that bind precisely to their targeted brain protein. The researchers expect to initiate clinical trials of the Aβ MRI agent in 2019. Because the platform involves no radiation, it could permit frequent longitudinal imaging of patients, enabling the monitoring of disease response and progression over time – particularly critical in the clinical assessment of new therapies. In addition, such an MRI-based diagnostic agent would make widespread screening for early disease feasible.
A fluid transition
The first fluid biomarkers reflecting Alzheimer’s disease pathology were identified in cerebrospinal fluid (CSF) in the 1990s: Aβ42, total tau, and phosphorylated tau. These biomarkers are reported to have greater than 95 percent sensitivity and greater than 85 percent specificity to the disease. CSF analysis also provides a measurement of the equilibrium level of a given protein at a single time point, because CSF content reflects the net result of rates of both protein production and clearance. Unfortunately, CSF biomarkers have several downsides. For one, they provide indirect information; their analysis implies, rather than specifically measures, brain pathology. For another, the lumbar punctures required to sample the fluid are painful and invasive. These difficulties have spurred considerable research aimed at finding suitable biomarkers in blood or plasma for a simple, low-cost, minimally invasive Alzheimer’s disease screen.
Advances toward a simple blood test have been reported on several fronts. Research published in early 2018 by scientists from Japan’s National Center for Geriatrics used mass spectrometry to ionize and scan blood plasma from more than 300 people for a particular peptide or amino acid linked to Aβ (12). The researchers found that the amount of amyloid they detected in blood predicted – with over 90 percent accuracy – the degree of cognitive problems faced by each of the subjects tested. In addition, the blood test results correlated to amyloid measurements in the subjects’ CSF and to PET brain scans of amyloid deposition. Even more recently, a German research group found that changes in the structure of plasma Aβ might foretell Alzheimer’s disease. The researchers used infrared spectroscopy to measure the ratio of β-sheet to α-helical forms of Aβ in plasma, and found that they could distinguish prodromal Alzheimer’s disease from healthy individuals (13). In longitudinal studies using this approach, healthy people who tested positive for β-sheet forms were nearly eight times more likely to be diagnosed with Alzheimer’s disease within the next eight years. Although research looks positive, we’ll need further studies before we are able to establish Aβ as a stable blood-based biomarker.
A protein related to neurodegeneration, neurofilament light (NfL), is also showing strong potential as a blood-based biomarker for Alzheimer’s disease and other conditions (such as traumatic brain injury) marked by similar tissue damage (14). Dying neurons release proteins into the brain, some of which – including NfL – can be found in trace amounts in blood. Researchers have found high levels of NfL in the blood of people with Alzheimer’s disease and mild cognitive impairment (8). Rising plasma NfL levels over time have also been associated with worsening cognitive scores and brain atrophy. Although NfL is neither sensitive nor specific enough to be employed on its own as a marker of Alzheimer’s disease, it has been able to distinguish Alzeimer’s from mild cognitive impairment and from healthy controls with a performance equal to that of Aβ or tau in CSF. Furthermore, blood NfL measurements have accurately predicted disease progression, making it a potentially useful biomarker for patient enrichment in clinical trials, longitudinal studies of drugs aimed at slowing cognitive deterioration, or clinical trials of drugs that target neurodegeneration.
Biomarker diagnostics in the real world
Brain imaging modalities are already validated as biomarkers of Alzheimer’s disease. Tests based on biomarkers found in blood and CSF have not yet reached that status, and no blood- or plasma-based tests are commercially available. However, the availability of such blood tests could be a big step forward for drug development, making it much easier to recruit, screen, and track patients’ responses to experimental drugs. Ultimately, the utility of such technologies will depend on a number of factors: the setting in which they are used, the required test accuracy, the tests’ cost, and their general availability. Once approved treatments for Alzheimer’s disease are available, a blood or plasma test could also enable the widespread screening of older individuals, allowing those with positive biomarker tests to be sent for a more specific confirmatory diagnosis via imaging.
Eventually, the field of Alzheimer’s disease research and clinical medicine may come to resemble that of oncology, where the diagnosis of “cancer” is recognized as including a variety of disparate malignancies, and drug development and prescription is increasingly based on the specific biomarkers expressed by an individual patient’s tumor. The ability to discern the cause of a patient’s cognitive decline based on biomarkers of specific pathologies could change our understanding of Alzheimer’s disease in the same way, moving us away from the current blanket diagnosis and toward a spectrum of pathologies that lead to cognitive impairment. The availability of such a precision medicine approach to Alzheimer’s disease diagnosis and treatment could only lead to earlier diagnosis and intervention and ultimately better outcomes.
Carlo Medici is chief executive officer of Alzeca Biosciences, Houston, USA, a company developing novel advanced imaging agents for the early diagnosis of neurodegenerative diseases, including Alzheimer’s disease.
- US Food and Drug Administration, “Early Alzheimer’s Disease: Developing Drugs for Treatment” (2018). Available at: bit.ly/2o96fnl. Accessed August 17, 2018.
- T Jonsson et al., “A mutation in APP protects against Alzheimer’s disease and age-related cognitive decline”, Nature, 488, 96–99 (2012). PMID: 22801501.
- H Martiskainen et al., “Decreased plasma β-amyloid in the Alzheimer’s disease APP A673T variant carriers”, Ann Neurol, 82, 128–132 (2017). PMID: 28556232.
- CR Jack Jr et al., “NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease”, Alzheimers Dement, 14, 535–562 (2018). PMID: 29653606.
- ER Fantoni et al., “A systematic review and aggregated analysis on the impact of amyloid PET brain imaging on the diagnosis, diagnostic confidence, and management of patients being evaluated for Alzheimer’s disease”, J Alzheimers Dis, 63, 783–796 (2018). PMID: 29689725.
- R Ossenkoppele et al., “Impact of molecular imaging on the diagnostic process in a memory clinic”, Alzheimers Dement, 9, 414–421 (2013). PMID: 23164552.
- P Sánchez-Juan et al., “Practical utility of amyloid and FDG-PET in an academic dementia center”, Neurology, 82, 230–238 (2014). PMID: 24353340.
- M Grundman et al., “Potential impact of amyloid imaging on diagnosis and intended management in patients with progressive cognitive decline”, Alzheimer Dis Assoc Disord, 27, 4–15 (2013). PMID: 23203162.
- AS Zannas et al., “Impact of ¹⁸F-florbetapir PET imaging of β-amyloid neuritic plaque density on clinical decision-making”, Neurocase, 20, 466–473 (2014). PMID: 23672654.
- EM Mitsis et al., “A consecutive case series experience with [18 F] florbetapir PET imaging in an urban dementia center: impact on quality of life, decision making, and disposition”, Mol Neurodegener, 9, 10 (2014). PMID: 24484858.
- iDEAS, “Imaging Dementia – Evidence for Amyloid Scanning” (2018). Available at: www.ideas-study.org. Accessed August 17, 2018.
- A Nakamura et al., “High performance plasma amyloid-β biomarkers for Alzheimer’s disease”, Nature, 554, 249–254 (2018). PMID: 29420472.
- A Nabers et al., “Amyloid blood biomarker detects Alzheimer’s disease”, EMBO Mol Med, 10, e8763 (2018). PMID: 29626112.
- N Mattsson et al., “Association of plasma neurofilament light with neurodegeneration in patients with Alzheimer disease”, JAMA Neurol, 74, 557–566 (2017). PMID: 28346578.