Biomarkers in neurodegenerative diseases

Drug development and the identification of biomarkers in neurological diseases remains impeded by many obstacles and has to overcome a number of challenge barriers. Among them is the availability of tissue at the site of pathology, poor clinical diagnostics and the complexity of the brain, lack of functional endpoints and models for validations, not the least of which is the inaccessibility of the brain itself, when evaluating the consequences of drug-target interactions. Identifying the optimal biomarker may depend on establishing which cellular components drive the dys-regulation of the immune response in neurodegenerative diseases.

Improving drug discovery

There is a clear necessity to fundamentally change drug discovery paradigms to meet today’s health care needs. The past 20 years, which includes the decade of the brain (the 1990s), have seen the development of few novel therapeutic agents for CNS (central nervous system) disorders. Indeed, many of the drugs coming to market have been derived from the pharmacology of older products with proven clinical utility with the idea of improving potency, tolerability or easy administration.

Drug development in the CNS domain has to overcome a number of challenge barriers not the least of which is the availability of tissue at the site of pathology; poor clinical diagnostics and extend of disease progression at the time of the diagnosis; inaccessibility of the brain itself, when evaluating the consequences of drug-target interactions; complexity of brain and tissue heterogeneity; lack of functional end points and models for validation. The lack of facile surrogate endpoints for potential CNS therapeutics means that making the right choice of targets and molecules for development is crucial. The other aspect of the problem for CNS is confounded by the paucity of process and technologies that can provide a true translational bridge between preclinical studies and subsequent clinical testing and evaluation. These challenges are likely to be underestimated by the fact that translational research advances in the field of neuroscience is further slowed by the ability of preclinical bench work to predict efficacy in humans, and is more limited than in other therapeutic areas. Additional caution is signaled by recent failures of the therapeutic translation in setting where the fidelity of rodent models might a priori be expected to be above average. A quick perusal of the literature yields myriad proposed approaches for altering, for instance the course of Alzheimer’s disease (AD) or Multiple Sclerosis (MS). However, only a few of these promising ideas can possibly undergo adequately full clinical testing involving years of work and thousands of patients, yet it is unclear that the preclinical experiments can provide a rational basis for culling the pack. Thus, the current best way to refine target selection may be to utilise the only powerful model system for human nervous system diseases available – humans – prioritising the rapid, efficient, yet safe exploration of early candidate approaches in both healthy volunteers and patients with targeted diseases. Today, new methodologies such as genomics, proteomics and imaging have begun to permit the detection of intermediate molecular, biochemical and physiological consequences of drug action within the intact, functioning human nervous system.

Measuring with biomarkers

Biomarkers are useful characteristics to evaluate disease process and targets of therapeutic agents. They are objectively measured and obtained by non-invasive or minimally invasive procedures, such as extraction of blood or spinal fluid.

Biomarkers should also be easy to measure in a simple assay at a practical level. More importantly, when detected early in the course of multiple sclerosis (MS) or other neurological disease they should be potentially effective in predicting long term clinical response.

Neuroimaging techniques have helped overcome some of the problems associated with clinical diagnosis of neurological disorders. Today, imaging is well respected as early diagnostic tools in CNS. Used in the past, primarily, as a technique to exclude structural lesions, neuroimaging has today become increasingly utilised as a tool for diagnosis and to test the effectiveness of treatment of neurological diseases.

Many earlier studies showing that quantitative measures of hippocampal atrophy distinguish Alzheimer’s disease (AD) from non-demented elderly individuals. These are different ways to measure atrophy on brain images ranging from brief to highly detailed. Positron Emission Tomography (PET) in combination with appropriate designed emitter ligands, is now regularly used to measure the fraction of a target brain receptor occupied by a given compound, reducing the guesswork involved in the selection of drug dosage for Phase II testing. Functional imaging with PET and magnetic resonance imaging (MRI), alone or in combination with electrophysiological monitoring, and biochemical detection thresholds in blood and CSF (cerebrospinal fluid) permit pharmacodynamic assessment and may ultimately prove useful in establishing drug efficacy. To gain maximal advantage from these measurements, changes in the protocol may need to be operated to enhance flexibility and feedback to discovery scientists. The goal should be to provide a smooth transition from bench to bedside, such that the process of drug discovery and concept refinement continues during clinical testing.

The identification of specific biomarkers to detect, assess and predict final infarct, in the case of acute neuron degenerative diseases, such as stroke or head injury size, can provide potential endpoints. Comparing, for instance, perfusion weighted and diffusion weighted MRI scanning, shortly after stroke onset, may identify brain that is at risk for infarction, but potentially salvageable. In contrast, to acute neurodegenerative process the picture gets more complicated in the case chronic neurodegenerative diseases like MS, Parkinson’s disease and AD that generally remain asymptomatic for many years after the neurodegenerative process has begun.

So the ultimate goal is to identify biomarkers that quantitatively correlate to disease state and that should represent the goal to be achieved in the next few years. However, the intricacy of the nervous system has limited the availability of such quantitative biomarkers in plasma or CSF, but imaging techniques are promising. Another factor associated with the inherent brain complexity that further confounds the identification of biomarkers in neurological diseases is the herogeneity of the representative neuropathologies.

The progression of chronic neurodegenerative diseases might also be accessible to imaging biomarkers, such as hippocampal atrophy or deposition of amyloid plaques in AD, as well as loss of presynaptic dopaminergic markers in Parkinson’s disease. Specific patterns of regional brain activation, detected with fMRI or PET imaging, may provide an objective measure of depression, though further studies are required to prove that these methods can provide equal or greater reliability than clinical examination.

A wide range of imaging based biomarkers is currently being studied for AD and other neurodegenerative diseases. Multiple biochemical analyses in blood, urine and cerebrospinal fluid have also been proposed. Markers such as Beta amyloid and Tau proteins seem to be intimately involved in the pathology of AD. It has been shown that CSF levels of A-beta 42 are reduced in AD vs. control. Moreover, levels of Tau and A-beta 42 do not change with aging or dementia severity. High levels of Tau and low levels of A-beta42 are also seen in some other dementias and neurologic disorders, limiting the specificity of these markers. CSF Tau reflects neuronal damage and is acutely elevated after a stroke; phospo-tau may more reflect NFTs (neurofibrillary tangles) and levels do not rise after stroke. Although levels of phosphor-tau and total tau in CSF are correlated, phosphor-tau appears to be less prone to non specific increases in conditions such as CNS inflammation and stroke.

Several studies measured an increased level of A-beta in plasma in AD or other chronic neurodegenerative diseases. However, findings were inconsistent and there appears to be much overlap between patients with sporadic AD (Alzheimer’s disease) compared to controls. Thus, plasma and CSF A-beta levels do not correlate, and it is not clear to what extent plasma A-beta reflects events in the brain. In fact, none of the biomarkers studied seem to be linearly related to stages of disease throughout the full course of disease. The picture is further complicated by the many therapeutic approaches currently used on different pathophysiological hypotheses that might require different mechanistic markers. It is then likely that a multimodal biomarkers approach stratified for ease of use, sensitivity and specificity will be needed in AD and other neurodegenerative diseases. New biomarkers of pre-symptomatic disease will be important for population enrichment strategies and confirmation of efficacy during the assessment of novel AD modifying therapies. These markers appear promising but will require further study in other dementias. Further research on these biomarkers will need to compare how different techniques and analytic methods perform relative to each other. Larger numbers of patients studies are needed to establish the extent to which these biomarkers add additional diagnostic value to careful clinical evaluation.

An excellent example of potential biomarkers application in neurodegenerative diseases is represented by multiple sclerosis (MS), an immune-mediated process directed at structural components of the CNS. Both the inflammatory response and neurodegenerative events provide a rich source of potential factors that can be monitored and treated in an attempt to halt the underlying pathophysiologic process. Although visual evidence of disease activity on MRI has not been an accurate predictor of sympthomatology, biomarkers have the potential to yield evidence of ongoing disease activity particularly if their up regulated activity is isolated to areas of active disease such as myelin sheaths. The list of biomarkers with the potential to provide information about disease activity in MS include immunomodulatory factors, such as tumour necrosis factor alpha (TNFa) and various interleukins (IL), interferons, integrins, eicosanoids and proteases as well as markers of neuron degeneration such as up-regulation of nitric oxide (NO). Identifying the optimal biomarker may depend on establishing which cellular components drive the dysregulation of the immune response in MS. However, there is still controversy whether MS is still better characterised as a T or B cell mediated disease.

Biomarkers promise to reveal the pathophysiology of MS or other chronic neurodegenerative diseases, but the large number of factors and inconsistencies between experimental and clinical studies suggest heterogeneity among patients. Some biomarkers that have been demonstrated promising as prognostic indicators and potential targets for therapy in single center studies, such as IL-12, have yet to demonstrate viability in large multicenter trials. Even if the cascade of molecular events in the development of lesions is similar among patients with MS, lesion location may be important for prognosis. As a result, the ability to predict symptomatic episodes with biomarkers may require a methodology to identify symptom-producing lesions and then perform localised quantification of changes in target biomarkers. Just as an increase in the number or volume of lesions, as measured by MRI, correlates imperfectly with symptom expression. An up-regulation of immunologic mediators may not be meaningful unless they can be isolated to sites where their activity is likely to predict risk of events. In addition, biomarkers may have different significance at different stages of the disease.

Despite the heterogeneity in expression of biomarkers in MS or other chronic neurodegenerative diseases, achievements toward isolating the molecular events that characterise these diseases has enormous promise for understanding the pathophysiologic process and opportunities to intervene. It is also plausible that the importance of the biomarkers will vary across patients’ groups, changing the utility of specific makers both as predictors and targets for therapeutic intervention at the different stages of many diseases.

Advancing clinical diagnostics and therapeutics

The identification of new biomarkers for neurological disease is essential for the continued advancement of clinical diagnostics and therapeutics. In order for the biomarkers to be utilised to their full potential, the key components of the different available technologies should be combined. Thus, standardisation of the diagnostic protocols to include the addition of neuroimaging and individual genotyping for the identification of genetic/genomic biomarkers to the customary clinical protocols of today is essential to achieving better precision in disease diagnostics. In addition, the constant advances in neuroimaging will eventually lead to the use of more reliable diagnostic tools. The increased diagnostic efficacy should allow therapeutic intervention, before more significant pathologies develop and, thus, result in more efficacious treatment for these patients which have, as of yet, had little relief from these devastating diseases.

In order for the biomarkers to be utilised to their full potential, the key components of the different available technologies should be combined. Moreover, the constant advances in neuromaging will eventually lead to their use as a more reliable diagnostic tool. Neuroimaging data will likely be used to identify specific biomarkers in the future. The development of specific ligands that target defined pathologies is of utmost importance and will revolutionise biomarkers identification. The next five years will bring to the forefront mass spectrometry protocols, which are still in their infancy at present. Clinical diagnoses will be increasingly refined and sub-classifications of common diseases will most certainly occur as we increase our ability to discover genetic and histologic variants. In addition, some diseases that are thought to be distinct may actually contain overlapping pathological features. The future of functional validation lies in the hands of novel high-throughput assays that can precisely and reproducibly analyse the huge amounts of data that are being generated by current and future experimental methodologies.