Description

Numerous biotechnologies have been integrated to develop personalized medicine. Some of these have been described in other MedLink Neurology articles. A few of them are described here briefly.

Neurogenomics.

Neurogenomics is improving our understanding of neurologic diseases, which is an important basis for the development of rational therapies in integrated healthcare of the future. Neurogenomics has contributed to neuropharmacology in the area of pharmacogenetics, the term applied to the influence of genetic factors on the action of drugs, ie, which drugs work best on which patients and the genetic basis of susceptibility to adverse reactions of drugs. This term should be distinguished from pharmacogenomics (an offshoot of genomics), which usually refers to the application of genomic technologies to drug discovery and development. Significant contributions of neurogenomics to personalized neurology include the following:

Epigenomics and epigenetics. The epigenome is a record of the chemical changes to the DNA and histone proteins of an organism, which can be inherited by an organism's offspring. The epigenome is involved in regulation of gene expression, development, and tissue differentiation. Unlike the underlying genome, which is largely static within an individual, the epigenome can be altered by environmental conditions. Changes in the epigenome can result in changes in the function of the genome.

Epigenetics refers to the study of changes in the regulation of gene activity and expression that are not dependent on gene DNA sequence. Whereas epigenetics often refers to the study of single genes or sets of genes, epigenomics refers to more global analyses of epigenetic changes across the entire genome.

Neurologic disorders are not only associated with genomic mutations and transcriptomic dysregulations, but with changes in the epigenome as well. Among the various types of epigenomic modifications, DNA methylation, histone modifications, and expression levels of microRNAs have been the most widely studied. DNA methylation is implicated in the development of the human brain as well as plasticity underlying learning and memory. Widespread reconfiguration occurs in the methylome and conserved non-CG methylation accumulates in the neuronal genome during development (38).

Targeting the complete mitochondrial exome provides a greater potential to identify rare variants that disrupt normal mitochondrial function, enabling an exact diagnosis in a large proportion of patients that remain undiagnosed by other methods. Over 95% of the target bases can be sequenced to an average coverage of 400x, providing highly accurate and sensitive results.

Neuroproteomics. Insight into the protein-based mechanisms of neurologic disorders will provide more relevant targets for drug discovery as well as development for central nervous system disorders. Pharmacoproteomics, the application of proteomics to pharmacology, has the following advantages for the development of personalized medicines (24):

Neuroproteomics studies show that neurodegenerative diseases share many common molecular mechanisms, including misfolding of proteins. Further understanding of these mechanisms may lead to strategies for the prevention of neurodegenerative diseases and the development of effective drugs. Identification of neurodegeneration-associated changes in protein expression will facilitate the identification of novel biomarkers for the early detection of neurodegenerative diseases and targets for therapeutic intervention. Applications of proteomics technologies will facilitate a more comprehensive analysis of novel therapeutic strategies for central nervous system disorders. This will also accelerate development of specific diagnostic and prognostic disease biomarkers for the personalized management of neurologic disorders.

Neurometabolomics. Metabolomics plays an important role in personalized medicine. Neurometabolomics is the study of metabolites in the nervous system, many of which are biomarkers of disease. Neurons in the human central nervous system perform a wide range of motor, sensory, regulatory, behavioral, and cognitive functions, which requires diverse neuronal as well as non-neuronal cell types. Metabolomics can help in assessing the specificity of metabolic traits and their alterations in the brain and in fluids such as cerebrospinal fluid and plasma. Current applications of metabolomics include the discovery of potential biomarkers of aging and neurodegenerative diseases (30). Neurometabolomics will improve our understanding of the physiology as well as pathology of the nervous system and facilitate the development of personalized neurology by providing biomarkers for disease monitoring as well as for targets for therapeutics.

Role of biomarkers in the development of personalized neurology. A biomarker—DNA, RNA, or protein—is defined as a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention (26). Biomarkers play an important role in the development of personalized medicine. An example of the usefulness of biomarkers in the development of personalized medicine is biomarkers for Huntington disease. Genome-wide gene expression profiles from blood samples of Huntington disease patients have identified changes in blood mRNAs that clearly distinguish Huntington disease patients from controls. The elevated mRNAs are significantly reduced in Huntington disease patients involved in a dose-finding study of the histone deacetylase inhibitor sodium phenylbutyrate. These alterations in mRNA expression correlate with disease progression and response to experimental treatment. Such biomarkers may provide clues to the state of Huntington disease and may be of predictive value in clinical trials.

Role of therapeutic drug monitoring in the development of personalized neurology. Therapeutic drug monitoring is the measurement of drug concentrations in blood to optimize pharmacological therapy and identify the high interindividual variability of pharmacokinetics of patients, which enables personalized pharmacotherapy (55).

Role of functional imaging of the brain in the development of personalized medicine. The combination of PET and fMRI in hybrid PET/MRI scanners enhances the significance of both modalities for research in biomarkers and molecular bases of neurologic disorders, which will facilitate stratification procedures for the development of personalized medicine (40).

Role of synthetic biology in the development of personalized medicine. Techniques of synthetic biology enable the growth of living systems using parts that are not derived from nature but designed and synthesized in the laboratory. Genes as well as proteins can be synthesized. Advances in sequencing can be used to combine synthetic biology with personalized medicine. Therapeutic systems can be based on a synthetic genome using an expanded genetic code and can be designed for specific personalized drug synthesis as well as delivery and activation by a pathological signal (28). The ability to control expression in target areas within a genome is important for the use of synthetic biology in designing personalized medicines.

Role of nanobiotechnology in the development of personalized neurology. Nanotechnology is the creation and utilization of materials, devices, and systems through the control of matter on the nanometer-length scale, ie, at the level of atoms, molecules, and supramolecular structures. It is the popular term for the construction and utilization of functional structures with at least 1 characteristic dimension measured in nanometers (a nanometer is 1 billionth of a meter (10^-9 m). Nanobiotechnology is the application of nanotechnology in the life sciences; applications in medicine are referred to as nanomedicine (27). Nanomedicine covers all specialties of medicine, including neurology, which is termed “nanoneurology.” Nanobiotechnology contributes to personalized medicine by refinement of molecular diagnosis, integration of diagnostics with therapeutics, and improvement of targeted drug delivery, eg, across the blood-brain barrier for cerebral lesions.

Neurotechnologies. Wearable devices are being used for monitoring of signs and biomarkers of neurologic disorders. These are useful, not only for diagnosis, but also for the integration of diagnosis with therapeutics for personalized neurology. These devices enable assessment of patients in daily situations. For example, tasks can be focused on examining the exact types of memory disabilities in the heterogeneous group of cognitive impairments by remote use of devices and enable a virtual reality approach to personalized neurorehabilitation (52). Use of remote monitoring devices into a patient-centric virtual clinical trial design enables collection of real-time information at each stage of the trial as well as monitoring of endpoints, which would be suitable for evaluation of personalized therapies.

Role of neuroinformatics in personalized neurology. Neuroinformatics is the integration of computational biology or bioinformatics with neurosciences. Enormous amounts of data from neuroimaging and structural as well as functional studies of the brain can be used to construct a “virtual brain” as a basis for disease-specific models, which can be applied to personalize therapy of neurologic disorders at the individual level (13).

Digital technologies. Digital technologies include mobile devices, neuroimaging, electronic medical records, artificial intelligence, and robotics. All of these can be used for personalized neurology.

Personalized biological therapies for neurologic disorders. Biological therapies particularly relevant to personalized treatment of neurologic disorders are uses of autologous stem cells for the treatment of neurologic disorders or immunotherapies of brain tumors involving tumor cells taken from a patient’s tumor. The patient’s own somatic cells, genetically engineered to release therapeutic substances, are also in clinical trials. This approach overlaps with gene therapy.

Ex vivo gene therapy involves the genetic modification of the patient’s cells in vitro, mostly by use of viral vectors, prior to reimplanting these cells into the tissues of the patient’s body. This is a form of individualized therapy.

Apart from infectious diseases involving the nervous system, therapeutic vaccines are in development for malignant tumors such as glioblastoma, autoimmune disorders such as multiple sclerosis, and degenerative disorders such as Alzheimer disease. Cancer vaccination involves attempts to activate immune responses against antigens to which the immune system has already been exposed. Most of the personalized cancer vaccines are cell-based, and these were the earliest forms of personalized medicine (21).

Monoclonal antibodies play an important role in personalized neurology, both as diagnostic and therapeutic agents. Designed to bind to specific receptors, monoclonal antibodies can be used to guide passage of nanomedicines across the blood-brain barrier to specific targets in the brain, such as glioblastoma multiforme.

Clinical applications

A personalized approach is being applied to the management of several neurologic disorders. Several personalized therapies are still undergoing investigations. Some examples are given here.

Alzheimer disease. One problem with the management of Alzheimer disease is that it is not a single clinical entity and is frequently associated with pathologies that require different therapeutic approaches, which may obscure small responsive subgroups. A comprehensive risk profile of the patient with Alzheimer disease reveals vast variation in biomarkers of risk that may serve to guide individualized treatment (34).

The interaction of different transcription factors with the regulatory region of the ApoE gene plays an important role in the neuroinflammatory process seen in Alzheimer disease and is a target for developing new therapeutics for the disease. Genotype-specific responses of Alzheimer disease patients to a certain drug or combination of drugs have been demonstrated, although several studies examining the role of ApoE produced conflicting results. The pharmacogenomics of Alzheimer disease may, in the future, contribute to optimizing drug development and therapeutics, increasing efficacy and safety, and reducing side effects.

Associations between the GAB2 gene and late-onset Alzheimer disease (LOAD) risk has been characterized in APOE epsilon4 carriers by a genome-wide survey of more than 300,000 single nucleotide polymorphisms (49). Discovery of this LOAD susceptibility gene, if replicated, provides new opportunities to investigate LOAD pathogenesis, predisposition, treatment, and prevention. Genome-wide studies using even higher density platforms and compound genetic analyses in sufficiently large samples of well-characterized cases and controls promise to play increasingly important roles in the scientific understanding, evaluation, personalized treatment, and prevention of Alzheimer disease.

Parkinson disease. The complex multi-neurotransmitter dysfunction in Parkinson disease can manifest clinically as nonmotor subtypes, which enable classification for a personalized approach to the disease (54).

Personalized pharmacotherapy of Parkinson disease. The most prescribed drug for Parkinson disease, levodopa, is a precursor of dopamine. Levodopa is used to titrate dopamine up to an optimal level for movement and some aspects of cognition. The response of individual patients and adverse reactions vary considerably. An understanding of the basis of these differential effects may enable modification of the drug dose, or the combination of levodopa with other drugs, to produce the best outcome for individual patients and to avoid such reactions. There is a trend now toward incorporating genetics into clinical studies of therapy for Parkinson disease to investigate how a person's genetic make-up influences the effect of drugs that work by neurochemical intervention. Approximately 50% of Parkinson disease patients treated with levodopa develop levodopa-induced dyskinesias in the long term, and the use of pharmacogenetics should be explored to reduce this complication.

Entacapone, a drug used for the treatment of Parkinson disease, inhibits catechol-O-methyltransferase (COMT) in a dose-dependent, reversible, and tight-binding manner, but does not affect other catechol-metabolizing enzymes. It enables the reduction of levodopa dose. Results of clinical studies, however, indicate that COMT genotype seems to be a minor factor in judging the beneficial effects of entacapone administration.

Diagnosis for Parkinson disease. Currently, there is no diagnostic test that can confirm Parkinson disease. Diagnosis is usually made by clinical observation and confirmed only by postmortem neuropathological studies. Because Parkinson disease is heterogeneous in its clinical presentation and prognosis, it is recommended that patients should be screened for mild cognitive impairment, depression, orthostatic hypotension, and rapid eye movement sleep behavior disorder to predict disease course and to design more efficient personalized treatment approaches (14). A system with wrist and ankle motion sensors can provide accurate measures of tremor, bradykinesia, and dyskinesia as patients complete routine daily activities (48). This technology can provide insight on motor fluctuations in daily life of Parkinson disease patients to guide clinical management and aid in development of new therapies.

Exome sequencing has now been applied to Parkinson disease research and has the potential for use as a screening method to identify pathogenic mutations in some Parkinson disease patients (05). A major challenge of exome sequencing is the amount of data generated and the rapid evolution of methods to evaluate these data.

Cytochrome P450 CYP2D6 enzyme, which metabolizes many drugs, is also involved in the metabolism of dopamine. In studies comparing Parkinson disease patients exhibiting side effects such as "on-off" phenomenon and dyskinesia (both suggesting favorable response to therapy) with a subgroup of patients showing no such response, only the prevalence of the CYP2D6 4 allele was found to differ significantly between the Parkinson disease patients and control group.

Gene-expression profiles. These are used to define the 4 major classes of dopaminergic and noradrenergic neurons in the brain. Molecular profiles provide a basis for understanding the common and population-specific properties of catecholaminergic neurons and will facilitate the development of selective drugs. One goal of such studies is to identify genes that may influence the selective vulnerability of catecholaminergic neurons in Parkinson disease. The substantia nigra is most susceptible to Parkinson disease pathology, whereas the adjacent ventral tegmental area dopaminergic neurons are less vulnerable and hypothalamic dopaminergic neurons are spared. The sparing of ventral tegmental area neurons could be mediated by selective expression of neuroprotective factors, including neurotrophic factors, detoxifying enzymes, lipoprotein lipase, etc. There is selective high expression of alpha-synuclein in neurons of the substantia nigra and in locus coeruleus noradrenergic neurons that degenerate in Parkinson disease, which may modify the toxic effects of the widely expressed alpha-synuclein protein. Likewise, selective expression of the Zn2+ transporter by the substantia nigra and ventral tegmental area may play a role in the pathophysiology of Parkinson disease. Low concentrations of Zn2+ can exert a cell-protective effect; however, excess of Zn 22+ is neurotoxic and has been shown to promote degeneration of midbrain dopaminergic neurons. Thus, the molecular signatures of the major classes of catecholaminergic neurons improve our understanding of the characteristic features and functions of these neurons and facilitate the discovery of subgroup-selective drug targets.

Role of biomarkers. With an increasing understanding of pathomechanism, it is possible to identify additional biological targets and pathways associated with Parkinson disease to generate validated, standardized, and clinically feasible biomarkers for segregating patients into biologically relevant disease subtypes. The imaging tool DAT (dopamine transporter) scan provides a significant resource to identify patients with dopaminergic dysfunction, which is particularly powerful for inclusion into clinical trials targeting early disease, but a wider range of imaging and biochemical markers are needed to capture the potential biological diversity underlying Parkinson disease as well as atypical forms of parkinsonism (51).

Ambroxol as personalized therapy of Parkinson disease. Mutations of the glucocerebrosidase gene, GBA1, are the most important risk factor for Parkinson disease. Ambroxol, a mucolytic agent used in cough syrups, increases beta-glucocerebrosidase enzyme activity and reduces alpha-synuclein levels, supporting its potential role in modifying a relevant pathogenetic pathway in Parkinson disease. Results of an open trial suggest that ambroxol therapy is well tolerated, the drug penetrates the blood-brain barrier, and CSF alpha-synuclein levels were increased (41). Upregulation of brain cytosolic/lysosomal GCase activity may reduce alpha-synuclein levels, mediating a neuroprotective effect in patients with Parkinson disease both with and without GBA1 mutations (42). This is the first clinical trial of personalized therapy for a stratified (genetically defined) subtype of Parkinson disease.

Personalized cell therapy for Parkinson disease. Various types of cell therapy have been used for Parkinson disease, but none of these has, so far, proven entirely satisfactory. Rejection of transplanted cells (iPSCs) is 1 problem. Use of induced pluripotential stem cells for personalized cell therapy of Parkinson disease is feasible, without the problem of tissue rejection.

Deep brain stimulation. A combination of unilateral subthalamic nucleus and contralateral globus pallidus interna stimulation has been used to maximize the clinical advantages of each target while inducing fewer adverse side effects in selected patients with Parkinson disease because each target has its own clinical effects and risk profiles (61). In this symptom-tailored approach, intake of levodopa was reduced with a 64% improvement in measures of gait and falls at 12-month follow-up.

Amyotrophic lateral sclerosis. Because of the complex multifactorial pathomechanism of amyotrophic lateral sclerosis, the conventional "1 target, 1 drug" approach has failed so far to provide effective therapeutic solutions. There is a need for further identification of biomarkers and therapeutic targets to assist in diagnosis as well as stratification of patients into different subgroups for personalized therapeutic approaches. Neurotrophic factors and histamine signaling, which are dysregulated in specific subgroups of patients with amyotrophic lateral sclerosis, have shown promising results in models of amyotrophic lateral sclerosis. Integration of various “omics” combined with network-based approaches will provide additional guidance for personalization of amyotrophic lateral sclerosis treatment (56).

Multiple sclerosis. Genetic factors are responsible for the increased frequency of the disease seen in the relatives of individuals affected with multiple sclerosis. Pharmacogenetic research models based on high-throughput single nucleotide polymorphism technology have been used to establish the correlation between drug responsiveness and genetic polymorphisms of multiple sclerosis patients, which may promote the development of personalized medicine for this disease. Spectratyping has shown that T cells bearing certain types of receptors are activated in multiple sclerosis patients. T-cell receptor–based immunotherapy can be applicable to multiple sclerosis patients if the T-cell receptor activation pattern of each patient is determined at different stages of the disease. BEST-PGx (Betaferon/Betaseron in Early relapsing-remitting MS Surveillance Trial-Pharmacogenomics) has investigated the value of RNA expression profiling and pharmacogenetics in predicting treatment response to interferon beta in patients with early relapsing multiple sclerosis (32).

Genome-wide expression studies in brain tissue and blood samples of multiple sclerosis patients are expected to reveal biomarkers that would help to determine disease course, outcome, or treatment response in early stages of the disease (18). An increasing number of genetic polymorphisms have been correlated with multiple sclerosis, but so far, their relevance to the diagnosis of multiple sclerosis is rather low. A large number of genes (including GSTM, IL1B, PD-1, CCR5, OPN, IL4, HLA-DRB1*1501, CD24, ESR1, CD59, CNTF, CRYAB, IFNγ, MEFV, APOE, TGFB1) have been associated with certain multiple sclerosis phenotypes, but these correlations are often controversial. Research on pharmacogenomics of multiple sclerosis is increasing, but, so far, has not produced a useful biomarker for clinical practice (09). An important advance is functional characterization of the multiple sclerosis risk variant on chromosome 12p13.31 containing the gene TNFRSF1A, which encodes the tumor necrosis factor receptor superfamily member 1A with apoptotic activity. TNFRSF1A shows association with multiple sclerosis risk and provides an insight into the pathophysiology, which can lead to novel therapeutic strategies (37).

Multiple sclerosis is multifactorial disease. Advanced neuroinformatic techniques will be required to model and analyze interactions between the numerous factors involved in the onset of the disease, its progression, and the response to treatment. Collection of individual clinical data in specific multiple sclerosis registries and use of intelligent software instruments, such as the Multiple Sclerosis Documentation System 3D, will be needed to process this complex information for enabling personalization of treatment for multiple sclerosis patients (62).

A study was done to develop predictive algorithms for individual treatment response using demographic, clinical, and paraclinical predictors in patients with multiple sclerosis to evaluate accuracy and internal as well as external validity of these algorithms (31). The most prominent modifiers of treatment response were age, disease duration, disease course, previous relapse activity, disability, predominant relapse phenotype, and previous therapy.

Baseline biomarkers of inflammation are the most important prognostic indicators of effectiveness of fingolimod for relapsing-remitting multiple sclerosis and provide hints for selection of therapies for personalized management of the disease (12).

Traumatic brain injury. There is considerable variation in the response of patients to traumatic brain injury. Expression of some genes such as APOE have been implicated in outcome following traumatic brain injury, but extensive review of the literature has revealed contradictory results that are attributable to the heterogeneity of studies. Further research is needed to assess the relationship between genetic traits and clinical outcome in traumatic brain injury (10). Biomarkers are useful as diagnostic, prognostic, and monitoring adjuncts. Changes in the expression profile of biomarkers, such as microRNAs in peripheral blood mononuclear cells, may reflect molecular alterations following brain injury that contribute to the sequelae (46). Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury should provide novel multidimensional approaches to its characterization and classification, which will enable clinical trials in restricted patient populations by taking into consideration differences in biology, care, and outcome for developing optimal personalized patient management (39). According to 1 hypothesis, transfusion of autologous white blood cells (from a person’s blood stored prior to trauma) in combination with conventional management can restore impairment of immune system in traumatic brain injury due to systemic complications and surgical procedures (16). Evidence from experimental animal and human studies supports this suggestion within the framework of personalized medicine.

Personalized approach to management of cerebral edema in traumatic brain injury. Because cerebral edema in traumatic brain injury is a diffuse, multifaceted, heterogeneous, and dynamic process, it may be prudent to complement these measures with other forms of multimodal monitoring, imaging studies and phenotypic and genetic biomarkers to provide a personalized approach (29). The underlying mechanisms contributing to cerebral edema and clinical intracranial pressure targets may differ between patients based on a variety of characteristics, including age, gender, injury characteristics, and genetics. A personalized approach should address these factors. Bedside neuromonitoring will enable rapid stratification of patients with severe traumatic brain injury based on individual pathophysiology for optimizing edema-targeting strategies.

Epilepsy. The primary criterion for the selection of antiepileptic drugs (AEDs) is the patient's seizure type. This practice derives largely from drug studies that assess AED effectiveness for specific seizure types rather than the defined causes of seizures. Despite restriction to partial seizures, the response to an investigational AED is variable. The reasons for this include (1) patient-to-patient variation in the metabolism of the AED; (2) variations in the ability of the AED to bind to the target; (3) variations in the amount of AED target produced by different individuals; and (4) different pathophysiological events accounting for the same seizure phenotype. Currently, a trial-and-error approach is employed to choose the most effective AED for a patient from numerous choices, but approximately 30% of all patients are resistant to AED therapy, which can be partially attributed to the presence of polymorphisms of genes encoding enzymes involved in AED metabolism.

Genetic epilepsies. These include over 30% of all epilepsy syndromes. Several genetic tests are now available, particularly those based on next-generation sequencing, which can identify gene mutations in up to a third of the patients with epilepsies that are potential targets for personalized medicine (45).

Ohtahara syndrome. It has an onset within the first 3 months of life with characteristic “burst suppression” on EEG. Ohtahara syndrome due to aberrations on KCNQ2 gene, revealed on sequencing, has an outcome that differs according to the location of the mutation on the gene. Mutations in 1 part can cause seizures without long-term sequelae in development, but mutation in another part of the gene causes persistent seizures with intellectual disability and cerebral palsy. This condition is refractory to phenobarbital, pyridoxine, pyridoxal phosphate, and levetiracetam, but it responds to Na channel blockers such as intravenous phenytoin, which is later switched to maintenance therapy with carbamazepine (11). This is an example of personalized approach.

Role of genomics in the personalized treatment of epilepsy. Considerable information is being generated by advances in genomic technologies. Integration of these techniques with functional biology and bioinformatics will improve our understanding of the genetic contribution to epilepsy, use of genetic testing for risk assessment, and personalized treatment (33). Personalized whole exome sequencing is already available, and whole genome sequencing is likely to be routinely available within the next few years.

One of the approaches to optimize AED therapy is pharmacogenomic testing to detect polymorphisms that may affect the efficacy, tolerability, and safety of AEDs, including variations in the genes encoding drug-metabolizing enzymes such as cytochrome P450, or drug transporters such as MDR1 and MRP2 (60). It is also becoming increasingly clear that single nucleotide polymorphisms play an integral role in variability in both the pharmacokinetics and pharmacodynamics of AEDs. Gene expression patterns of children on valproic acid monotherapy differ according to whether they have continuing seizures or remain free from seizures. This information can be used for personalizing AED therapy. The publication of the human genome and increasingly sophisticated and powerful genetic tools offer new methods for screening drugs and predicting serious idiosyncratic side effects.

Available evidence suggests that genetic variants from the ABCC2 transporter may be associated with an altered response to AEDs. Results of a metaanalysis indirectly suggest a possible role of the ABCC2 transporter at the blood-brain barrier in altered drug response in patients with epilepsy (17). The authors suggest further studies in different ethnic groups to investigate the effects of the ABCC2 haplotypic variants and performing stratified analysis based on different phenotypic covariates.

No AED treatment guidelines based on pharmacogenetic data are yet available. There is a need for, and an opportunity to, establish standards specific to the conduct of future AED studies to improve the management of epilepsy.

Control of epilepsy with phenytoin can be a difficult and lengthy process because of the wide range of doses required by different patients and the drug's narrow therapeutic index. Similarly, appropriate doses of carbamazepine take time to determine because of the drug's variable effects on patient metabolism and its potential neurologic side effects. People with epilepsy are genetically different from one another, and some of those differences affect their responses to drugs in a predictable manner. A variant of the SCN1A gene, which is found more frequently in patients on the highest doses of phenytoin, has been implicated in many inherited forms of epilepsy. Detection of these gene variants might identify in advance the patients who will need the higher dose and enable a more optimal dose schedule at the start. Otherwise, it could take months to get the seizures under control. These findings provide a direction for a dosing scheme that could be tested in a clinical trial to assess whether pharmacogenetic testing can improve dosing decisions. Transcranial magnetic stimulation is useful for investigating the effects of genetic variants on cortical excitability and response to drugs.

Role of biomarkers in the personalized management of epilepsy. Biomarkers would facilitate clinical trials as well as the routine management of patients with epilepsy. However, with so many different types of seizures and causes of epilepsy, there are no universal biomarkers except EEG measurements. Some biomarkers detect diseases that manifest in seizures. There are no characteristic biomarkers of idiopathic epilepsy except those for monitoring seizures and response to treatment. Development of reliable epilepsy biomarkers would be a major advance in the personalized management of epilepsy.

Results of a clinical study show that serum HSP70 levels have an inverse correlation with hippocampal volume after controlling for the effect of age in patients with temporal lobe epilepsy, and HSP70 is a biomarker for prediction of higher frequencies of seizures in these patients (08). HSP70 is a stress biomarker in temporal lobe epilepsy in that it correlates inversely with memory scores and hippocampal volume. In addition, the symmetric extratemporal atrophic patterns may be related to damage of neuronal networks and epileptogenesis in temporal lobe epilepsy. Quantitative measurements by MRI of overall brain volume (gray matter, white matter, and CSF) in temporal lobe epilepsy are clinically meaningful biomarkers that are associated with increased cognitive morbidity. Focal cortical dysplasia is a common cause of pharmacoresistant epilepsy that is amenable to treatment by surgical resection. The identification of structural focal cortical dysplasia by MRI can contribute to the detection of the epileptogenic zone and improve the outcome of epilepsy surgery. New magnetic resonance–based techniques, such as MR spectroscopy, fMRI, and fMRI/EEG, are more frequently being used to increase the yield of MRI in detecting abnormalities associated with epilepsy.

Noninvasive imaging of brain inflammation would be helpful in determining its role in epileptogenesis and serve as a biomarker for epilepsy. The current imaging toolbox is limited by the range of neuroinflammatory targets that can be visualized directly. Research in this area will further advance as highly specific ligands and reproducible as well as practical imaging approaches become available (02).

Algorithm for the personalized management of epilepsy. Several stratification approaches to address therapeutic challenges in epilepsy take into consideration several investigations, including pharmacogenomic and pharmacogenetic studies (58).

Future prospects for the personalized management of epilepsy. It is expected that several gene mutations, eg, those in ion channel genes, will be identified in epileptic subjects in the future using techniques such as DNA microarrays for gene expression and sequencing. Future drugs may be designed specifically according to the electrophysiological dysfunction as personalized medicines for epilepsy. There is ample scope for development of new products with a benign side effect profile or higher effectiveness. Several such drugs are in development, but there is still need for better drugs and strategies to overcome drug resistance. Initial studies have focused on genes whose products play a putatively important role in AED pharmacology, particularly drug transporter proteins, drug metabolizing enzymes, and ion channel subunits. However, there is a lack of good correspondence between results from different laboratories, and more recent findings are awaiting attempts at confirmation. Thus, there are currently no AED treatment guidelines that are based on pharmacogenetic data. Various suggestions that have been made to facilitate the development of the personalized treatment of epilepsy include the following:

Personalized management of stroke. The diagnosis of stroke, its etiology, recurrence, recovery, and therapeutics pose a series of highly individualized questions. The guiding principles of personalized medicine in stroke underscore the need to identify, evaluate, organize, and analyze the multitude of variables obtained from an individual to generate a precise approach to optimize cerebrovascular health (19). Novel technologies, bioinformatics, and practical clinical paradigms can be guided by the principles of stroke management to develop personalized approaches based on synthesis of data from clinical trials. Three-dimensional printing and direct patient data modeling have the potential to develop methodology for the individualized prediction of recurrence risk, medication effects, procedural outcomes, and recovery from stroke.

Personalization of stroke management should start at the stage of clinical trials of various therapies. Stroke treatments may be neuroprotective in the acute stage and neuroregenerative or neurorestorative in the subacute and chronic stages. Various biomarkers and brain imaging can be used to guide clinical trials. Antithrombotics and their efficacy as well as safety can be improved by using pharmacogenetics and pharmacogenomics.

Analyses of patient subgroups have enabled modeling of the risks and benefits of endarterectomy and stenting in the individual patient. The Italian Stroke Organization and Stroke Prevention and Awareness Diffusion group have revised their original methodology to follow the new Scottish Intercollegiate Guideline Network’s grade-like approach, which is to integrate it with personalized medicine (36). This guideline offers recommendations on personalized medicine for the single patient and can be followed in addition to the more standard stroke guidelines.

Brain imaging has also been used to measure volume of infarct for making a decision about thrombectomy in stroke. In a randomized trial on stroke patients who present 6 to 24 hours after the initial episode with mismatch between clinical deficit and infarct volume, outcomes for disability at 90 days were better with thrombectomy plus standard care than with standard care alone (43). Complications such as hemorrhagic infarction due to reperfusion are less in cases with smaller infarcts than in large infarcts.

Major depressive disorders. Although the mechanisms of action are not well understood, several antidepressants, including serotonin-selective reuptake inhibitors (SSRIs) and tricyclic antidepressants, have been used for the treatment of major depressive disorders.

According to multiple trials, approximately 85% of patients respond to antidepressant treatment. However, only 60% to 65% respond to any 1 drug, and response to treatment usually takes 4 to 8 weeks, if the drug works. A failed first treatment is the best predictor of treatment dropout, and treatment dropout is the best predictor of suicide. Although antidepressant response takes weeks, the effects of antidepressants on monoamine systems is very rapid. Therefore, it is possible that the therapeutic effects of all antidepressants are due to common expression of genes after chronic treatment. The first step toward answering this question is finding out which transcripts are increased or decreased by antidepressant treatment. Such research can be done using an animal model. If a system is found to be responsible for the therapeutic effects of antidepressants, a new antidepressant pharmacotherapy could be developed to activate that system more acutely.

Pharmacogenomic approaches could help in predicting some of these outcomes. A 5-HT6 receptor polymorphism (C267T) is associated with treatment response to antidepressant treatment in major depressive disorders. A pharmacogenomic approach to individualize antidepressant drug treatment is recommended to be based on 3 levels:

Biomarker-guided personalized therapy of major depressive disorders. The most promising biomarkers for response to antidepressant therapy include genetic variants and gene expression profiles, proteomic and metabolomic markers, neuroendocrine function tests, electrophysiology, and brain imaging. Incorporation of biomarkers in the treatment of major depressive disorders could help improve the efficiency of treatment trials and ultimately speed remission (06). Changes in brain activity prior to treatment with antidepressants can flag patient vulnerability. Quantitative EEG (qEEG) measures have revealed that changes in brain function in the prefrontal region during the 1-week placebo lead-in are related to side effects in subjects who received an antidepressant (20). This study is the first to link brain function and medication side effects, and to show a relationship between brain function changes during brief placebo treatment and later side effects during treatment with medication. The findings show the promise of new ways for assessing susceptibility to antidepressant side effects. The ability to identify individuals who are at greatest risk of side effects would greatly improve the success rate of antidepressant treatment. For example, physicians might select a medication with a lower side-effect profile, start medication at a lower dose, or choose psychotherapy alone when treating patients susceptible to antidepressant side effects.

A latent-space machine-learning algorithm tailored for analyzing EEG data, called Spare EEG Latent SpacE Regression (SELSER), was applied to data from the largest imaging-coupled, placebo-controlled antidepressant study known as “Establishing Moderators And Biosignatures Of Antidepressant Response In Clinic Care (EMBARC)” (47). SELSER reflected response to antidepressant therapy, eg, outcome of repetitive transcranial magnetic stimulation (57). These findings improve our understanding of the neurobiological basis of antidepressant treatment through an EEG-tailored computational model to enable personalized treatment of depression.

A qEEG biomarker, the Antidepressant Treatment Response (ATR) index, has been associated with outcomes of treatment with SSRIs in patients with major depressive disorders. In an open study of the norepinephrine reuptake inhibitor reboxetine, the ATR index predicted response with 70.6% sensitivity and 87.5% specificity, and remission with 87.5% sensitivity and 64.7% specificity (07). These results suggest that the ATR index may be a useful biomarker of clinical response during norepinephrine reuptake inhibitor treatment of adults with major depressive disorders. Future studies are warranted to further investigate the potential utility of the ATR index as a predictor of noradrenergic antidepressant treatment response.

Having a biomarker of likely treatment effectiveness to predict and guide clinicians' decisions would reduce the likelihood of unsuccessful treatments with antidepressants. The PRISE-MD (Personalized Indicators for Predicting Response to SSRI Treatment in Major Depression) study tested whether qEEG measures taken after 1 week of treatment can predict the effectiveness of a full treatment regimen with antidepressant medications. The study, conducted by the University of California, Los Angeles, in collaboration with the U.S. National Institute of Mental Health, was completed, but no results have been published as of the end of 2014.

Vilazodone, a dual SSRI and a 5HT1A partial agonist, is in a phase 3 clinical trial in parallel with genetic biomarkers to guide its use as an antidepressant. As approximately one-half of depressed patients do not achieve satisfactory results with current first-line treatment options, a product that combines a genetic test with vilazodone will assist physicians in matching patients with a drug that is more likely to be effective for each patient in the first instance. The primary and supportive secondary efficacy endpoints were met in a randomized, double-blind, placebo-controlled trial. In addition, the study separately identified candidate biomarkers for a potential companion pharmacogenetic test for response to vilazodone.

Individualization of SSRI treatment. The introduction of SSRIs has significantly transformed the pharmacological treatment of several neuropsychiatric disorders, particularly in individuals affected by depression, panic disorder, obsessive-compulsive disorder, and social phobia. Compared with the previous generation of psychotropic drugs, SSRIs offer an improved tolerability to therapy while maintaining a high level of efficacy. Nevertheless, despite these advantages, not all patients benefit from treatment; some do not respond adequately, whereas others may react adversely. This necessitates a review of the initial treatment choice, often involving extended periods of illness while a more suitable therapy is sought. Such a scenario could be avoided were it possible to determine the most suitable drug prior to treatment.

Pharmacogenetics of SSRIs. The influence of genetic factors on SSRI efficacy now represents a major focus of pharmacogenetics research. Current evidence emerging from the field suggests that gene variants within the serotonin transporter and cytochrome P450 drug-metabolizing enzymes are important. It also appears likely that further key participating genes remain to be identified. A study in progress at the Pharmacogenetics Research Network at the University of California, Los Angeles, is investigating the genetic basis of response to fluoxetine and desipramine among Mexican Americans, in part by identifying novel single nucleotide polymorphisms that may be relevant to differing responses to antidepressants. The most important areas of future research are exploration of known candidate systems and the discovery of new targets for antidepressants, as well as prediction of clinical outcomes. By comprehensively delineating these genetic components, it is envisaged that this will eventually facilitate the development of highly sensitive protocols for individualizing SSRI treatment.

The Mayo Clinic (Rochester, MN) is offering a new genetic test through Mayo Medical Laboratories to help U.S. physicians identify patients who are likely to have side effects from drugs commonly used to treat depression. Mayo has obtained a nonexclusive license from Pathway Diagnostics, Inc. to test for a key genetic biomarker, 5HTT-LPR. The 5HTT-LPR biomarker identifies people who respond differently to antidepressants, including SSRIs, which act specifically by binding to the serotonin transporter and increasing the concentration of the neurotransmitter serotonin in the synapse. These medications include fluoxetine, sertraline, paroxetine, citalopram, and escitalopram.

The 5HTT-LPR biomarker has the potential to improve the management of patients with major depression and others who benefit from SSRI treatment. It provides unique information relating to drug response: side effect and compliance. The serotonin transporter genotype assists the physician in making a better choice of antidepressant medications for their patients based on their serotonin transporter genotype used in conjunction with CYP450 genotyping. Depending on genotypes, some patients should respond well to SSRIs, some may respond to SSRIs but more slowly, and some may respond more effectively to non-SSRI antidepressants.

Usually, genetic profiles cannot predict a large percentage of variation in response to citalopram. Data available through the Sequenced Treatment Alternatives to Relieve Depression database accurately classified response to citalopram in 78% of cases and concluded that genetic biomarkers can be used to guide selection of citalopram (01). The rules identified in this study can help personalize the prescription of antidepressants.

International guidelines for rational therapeutic drug monitoring are recognized for personalized treatment with antidepressants and antipsychotics. Retrospective analysis of genotyping of patients with depression shows a correlation between the poor metabolizer and ultrarapid metabolizer genotypes, the therapeutic drug monitoring data, and clinical outcomes. Therapeutic drug monitoring combined with genotyping of CYP2D6 is particularly useful in verifying concentration-dependent adverse drug reactions due to poor metabolizers and diagnosing pharmacokinetic reasons for drug failure in ultrarapid metabolizers. This is important because adverse drug reactions may mimic the psychiatric illness itself, and therapeutic failure due to ultrarapid metabolizers may be mistaken for poor compliance with the prescription.

Genetic variation within the hypothalamic-pituitary-adrenal axis is associated with risk for depression as well as response to antidepressant therapy. Analysis of data from the International Study to Predict Optimized Treatment in Depression revealed that rs28365143, a variant within the corticotropin-releasing hormone binding protein gene, has a role in predicting which patients will improve with antidepressants and the type of antidepressant that may be most effective (44). These results will facilitate personalized antidepressant therapy.

Personalized neuromodulation for depression. Deep brain stimulation is a promising treatment for severe depression, but lack of efficacy in randomized trials raises questions regarding anatomical targeting. In a case study, effects of mild stimulation of several mood-related brain sites were mapped in a patient with severe treatment-resistant depression (50). Results showed that stimulation at different sites could alleviate distinct symptoms (reducing anxiety, boosting energy levels, or restoring pleasure in everyday activities), and the benefits of different stimulation sites depended on the patient's mental state at the time. Results provide proof-of-concept for a personalized, circuit-specific method of treatment in psychiatry.

Attention deficit hyperactivity disorder (ADHD). Several different medications are available to treat ADHD, yet little data exist to guide treatment choices, which often must be based on trial and error. Stimulant medications, such as methylphenidate, are the most common and effective treatment for ADHD. Methylphenidate acts primarily by inhibiting the dopamine transporter, a protein responsible for the reuptake of dopamine from the synapse into presynaptic terminals. However, it is often difficult to predict how patients will respond to ADHD medications.

A double-blinded, crossover trial found that children with a variant form of a dopamine transporter gene, 9/9-repeat DAT1 3'-UTR genotype, responded poorly to methylphenidate in contrast to those with the 10/10-repeat variant who showed excellent response (53). This study shows that testable genetic differences might be used to predict the effectiveness of methylphenidate in children with ADHD. Further research is needed to determine the mechanisms related to poor response in patients with the 9/9-repeat genotype and to determine if this group responds differentially to alternative treatments. A larger study is evaluating children with ADHD on 2 other medications to see if their genes predict who will respond to either or both drugs.

The “impaired vigilance” subgroup of ADHD with excess frontal theta or alpha activity on EEG responds well to stimulant medication, whereas, in depression, this subtype might be unresponsive to antidepressant treatments and respond better to stimulant medication (03). A slow individual alpha peak frequency is an endophenotype associated with treatment resistance in ADHD. Future studies should incorporate this endophenotype in clinical trials to further investigate the efficacy of new treatments in this substantial subgroup of patients.