Historical note and nomenclature

The term "drug-induced neurologic disorders" refers to unintended or undesirable effects on the nervous system caused by drugs or associated with drug use. It is an iatrogenic illness that also includes neurologic disorders due to other therapies such as diagnostic and surgical procedures. The use of the word "induced" within the term "drug-induced neurologic disorders" does not necessarily imply a proven causal relationship of the drug to the disorder. The drug may affect the nervous system directly (primary neurotoxicity) or indirectly by other systemic disturbances caused by the drug (secondary neurotoxicity). Drug-induced neurologic disorders can include disorders caused by inappropriate use of a drug, overdose of a drug, or harmful drug interactions, but environmental and industrial toxins are excluded.

Terms that are commonly used to refer to adverse effects of drugs are defined by the World Health Organization as follows:

Adverse event or adverse experience. This is any untoward medical occurrence that may present during treatment with a pharmaceutical product, but that does not necessarily have a causal relationship with this treatment.

Adverse drug reaction. This is an event that is related, or suspected to be related, to the trial medicine. An adverse drug reaction is a response to a drug that is noxious, unintended, and occurs at doses normally used in man for prophylaxis, diagnosis, therapy of disease, or for modification of physiological function.

Side effect. This is any unintended effect of a pharmaceutical occurring at doses normally used in humans that is related to the pharmacological properties of the drug.

The concept of harm resulting from medical treatment is more than 3500 years old. In the 7th century BC, the Code of Hammurabi prescribed penalties for physician errors that resulted in harm. Similarly, in the 1st century AD, the Roman law included penalties for such harms. Great medical writers throughout the centuries have described complications of medical treatment. Around the middle of the 20th century, with the development of pharmacotherapy, increasing attention was given to adverse drug reactions. They were not viewed merely as iatrogenic phenomena, but rather as inevitable consequences of medical progress and introduction of new drugs. Among the earliest adverse effects of therapies to be recognized were those of the nervous system. Polyneuropathies as complications of vaccinations were recognized in 1934 (05). Encephalopathies due to vaccines were not described until 1949 (12). Payk was the first to separate complications of treatment of neurologic disorders from neurologic complications resulting from treatment of diseases of other systems (24). Around the same time, pathogenesis of iatrogenic neurologic disorders was discussed under a dozen categories, 9 of which concerned adverse drug effects (33). Most of the complications related to antibiotic and hormone therapy. The first work to focus on iatrogenic pathology in neurology was published in 1975 (01); however, there was concern regarding the increasing number of adverse effects. Surveillance systems have been developed during the past 20 years for collection and analysis of adverse drug reactions. Drug-induced neurologic disorders are well recognized now, and form the topic of a book (16). In view of the increasing occurrence of drugs and environmentally induced neurotoxicity, the National Toxicology Program of the National Institute of Environmental Health Sciences of the United States has modified its protocol to include a more extensive histopathological examination of the nervous system (26). Clinical significance, epidemiology, pathophysiology, and principles of management of drug-induced neurologic disorders will be discussed in this article.

Clinical significance

Drug-induced neurologic disorders can mimic neurologic disorders due to other causes; however, when a reasonable suspicion exists of association with a drug, or a known or plausible pathomechanism for drug-induced neurologic disorders, the drug should be considered in the differential diagnosis. Questionnaires, algorithms, and computer-based approaches are methods used to assess the causal relationship of the drugs to the adverse drug reactions. None of these methods have found universal application because they are tedious, time consuming, and expensive. In practice, causality is usually assessed by global judgment (ie, opinion of the causal relation of the drug to the event after taking into consideration the available relevant information, as shown in the questionnaire in Table 1).

Table 1. A Questionnaire for Global Evaluation of Causal Relationship of a Drug to an Adverse Event

(1) Is there a biological explanation (pathomechanism) for the adverse event?

(2) Is the adverse event temporally related to the drug?

(3) Do the symptoms subside after discontinuation of the drug? (dechallenge positive)

(4) Do the symptoms recur after resumption of the drug? (rechallenge positive)

(5) Is the event already known and documented (in literature or in the database of the manufacturer)?

(6) Is the adverse event known to occur in the course of the natural disease?

(7) Is the adverse event known with the concomitant therapy?

The answers are taken into consideration to allocate the adverse drug reactions to the following categories: (1) probable, (2) possible, and (3) unlikely. These are approximate terms, and the term "definite" is rarely used. "Possible" means that such a reaction can take place with the drug, and sometimes this term simply is used because the possibility cannot be excluded. Sufficient information for assessment may not be available in some cases.

Assessment of a drug-induced neurologic disorder in a patient receiving a multimodal treatment is difficult. Patients undergoing organ transplants are liable to neurologic complications from the primary disease, the procedure, and the drugs (that include immunosuppressants and antibiotics). Critical decisions need to be made in the case of liver and heart transplants. For this, an understanding of the pathomechanism of neurologic disorders and their natural history is important. An immunosuppressant may not need to be discontinued if the evidence for causal relationship is weak and the neurologic complications are transient.

Anecdotal reports of suspected adverse drug reactions involving the nervous system are common in the medical literature and play an important role in providing early warning alerts. They are, however, of limited value because confirmatory investigations are not often carried out, and this information is not systematically incorporated into commonly used drug information sources.

Epidemiology

The exact incidence of drug-induced neurologic disorders is unknown. Reported adverse effects of drugs on the nervous system form only a small proportion of all neurologic disorders, but they are under-recognized and under-reported. Adverse drug reactions, both from clinical trials and postmarketing surveillance, are usually reported to the manufacturers of the product involved. The manufacturers make the initial assessment of these reports and file the adverse drug reactions with the health authorities of the countries involved according to the regulatory requirements (Food and Drug Administration in the United States). The World Health Organization also maintains a database for adverse drug reactions.

The incidence of adverse drug reactions in hospitalized patients was determined by a meta-analysis of 39 prospective studies from American hospitals. The overall incidence in this analysis of serious adverse drug reactions was 6.7%, and fatal adverse drug reactions was 0.32%, making these the fourth leading cause of death in the United States after heart disease, cancer, and stroke (20). The incidence is much higher than that assumed from the spontaneous reporting system, which identifies only about 1 in 20 adverse drug reactions. In a study of 18,820 hospital admissions in the United Kingdom, 1225 were related to an adverse drug reaction, giving a prevalence of 6.5%, with the adverse drug reaction directly leading to the admission in 80% of cases (25). Assessment of 1263 consecutive admissions to the neurology unit of a Swiss university hospital over 1 year revealed drug-related problems in 29% of cases, and these were the cause of admission in 0.8% (31).

For determining incidence of adverse drug reactions, a study systematically reviewed all major electronic databases for studies published from 1990 to the end of 2018 and found that the body system associated with the most adverse reactions reported was the central nervous system (18). The most frequent adverse drug reactions were fatigue (55%) followed by dizziness (18.4%) and tremor (15.8%).

Only a small fraction of the adverse drug reactions are published as case reports or as a part of the results of clinical trials of new drugs. Most of the information received by the pharmaceutical companies is inadequate for establishing the diagnosis of drug-induced neurologic disorders, but it is used for answering questions from physicians and for making decisions for inclusions of adverse reactions in basic drug information and package inserts.

Most of the products listed in the Physicians' Desk Reference include a mention of at least 1 untoward effect that relates to the nervous system. Entries in the list of adverse reactions are not always based on medical judgment but may be a measure to protect the manufacturer against legal liability by having declared that such an adverse reaction has been reported. The adverse drug reaction may not be causally related to the drug, and this information is not of significant practical value for a neurologist. Currently, the Food and Drug Administration is proposing to reduce this bulky list of all adverse events to meaningful adverse drug reactions related to the drug.

The frequency of occurrence of adverse events in clinical trials can be calculated and compared with that in the placebo group. An adverse event is an adverse drug reaction only if the relation to the drug is proven or suspected. The size of clinical trials is limited and seldom involves more than 500 patients and, therefore, rare adverse drug reactions cannot be expected to show up in these trials. Postmarketing surveillance continues for the lifetime of a drug to detect such adverse drug reactions. The frequency of occurrence of adverse drug reactions in this phase is difficult to determine because of poor reporting rate and lack of knowledge of the denominator. The number of patients exposed to the drug is sometimes estimated from the quantity of the drug sold and the standard dose for a patient. These figures are not reliable because the amount of drug used by individual patients varies a great deal, and all the drugs sold are not administered to the patients.

A study from Sweden showed that fatal adverse drug reactions account for approximately 3% of all deaths in the general population, and two thirds of these were due to hemorrhages, of which 29% involved the nervous system (32). Antithrombotic agents were implicated in more than half of the fatal adverse drug reactions.

Chimeric antigen receptor T cell (CAR-T-cell) therapy has become an important tool in the treatment of relapsed and refractory malignancy; however, it is associated with significant neurologic toxicity. A study has revealed that neurologic side effects were extremely prevalent for individuals undergoing CAR-T-cell therapy, with 77% of participants displaying at least 1 manifestation of the following (27): encephalopathy (57%), headache (42%), tremor (38%), aphasia (35%), and focal weakness (11%).

Pathophysiology

Adverse drug reactions are categorized into 3 groups: drug related, allergic reactions, and idiosyncratic reactions. Various pathomechanisms of drug-induced neurologic disorders will be discussed under the following 3 categories:

(1) Direct neurotoxicity or primary neurotoxicity
(2) Indirect mechanisms (ie, neurotoxicity due to drug-induced disturbances of other organs)
(3) Predisposing or risk factors for drug-induced neurologic disorders (ie, related to the patient or related to the drug)

Direct neurotoxicity. Because some of the drugs used for treatment of neurologic disorders target receptors in the nervous system, there remains the possibility of direct neurotoxic effect. For example, dopamine receptors, which are profusely expressed in the caudate-putamen of the brain, represent the molecular target of several drugs used in the treatment of neurologic disorders such as Parkinson disease. Although such drugs are effective in alleviating the symptoms of the disease, their long-term use could lead to the development of severe side effects. Neurotoxicant-induced changes in protein level, function, or regulation could have a detrimental effect on neuronal viability. Direct oxidative or covalent modifications of individual proteins by various drugs are likely to lead to disturbance of tertiary structure and a loss of function of neurons. The proteome and the functional determinants of its individual protein components are, therefore, likely targets of neurotoxicant action, and resulting characteristic disruptions could be critically involved in corresponding mechanisms of neurotoxicity. Neuroproteomic studies can provide an overview of cell proteins, and in the case of neurotoxicant exposure, can provide quantitative data regarding changes in corresponding expression levels and/or posttranslational modifications that might be associated with neuron injury.

Two important considerations for direct neurotoxicity are: (1) breech of the blood-brain barrier, which usually prevents the access of drugs to the fluid spaces of the brain and (2) retrograde axonal transport. For direct neurotoxic effects, the drugs must cross the blood-brain barrier. Despite this barrier, lipid-soluble molecules as benzodiazepines readily enter the brain. Damage to the blood-brain barrier facilitates the passage of drugs that normally do not cross the blood-brain barrier. Diseases in which the blood-brain barrier is damaged, such as multiple sclerosis, malignant brain tumors, and meningitis, would facilitate the direct neurotoxic effect of drugs. There is a risk of neurotoxicity of nanomedicines based on metallic nanoparticles as they can easily cross the blood-brain barrier and exert toxic effect via several possible mechanisms, such as oxidative stress, lysosome dysfunction, and activation of certain signaling pathways (11). This can be avoided by use of relatively nontoxic nanoparticles such as those made of polymers and using targeted delivery to the lesion.

Drugs may bypass the blood-brain barrier by retrograde intra-axonal transport. In the case of peripheral nerves, the blood-brain barrier may be deficient in posterior root ganglia and perineurium, making them susceptible to peripheral neuritis. Damage to the blood-brain barrier can occur in several diseases, which can predispose the patient to drug-induced neurologic disorders when a neurotoxic drug is administered.

It is a well-established fact that neurons have the capacity to incorporate substances at the periphery of the axons and that material can be transported within the axons to the perikaryon by means of retrograde axonal transport. When toxic substances are picked up by axons and transported to perikaryon, death or degeneration of the neuron results, a phenomenon called "suicidal axonal transport."

Various direct mechanisms of neurotoxicity are listed in Table 2.

Table 2. Direct Mechanisms of Neurotoxicity

Disturbances of brain energy metabolism
	• ATP synthetase inhibition, eg, oligomycin • Uncoupling, ie, dissociation of oxidative phosphorylation, eg, by barbiturates • Disturbances of oxygen consumption • Enzymatic dysfunction, eg, theophylline leading to convulsions by inhibiting adenosine • Selective vulnerability of the central nervous system
Sequelae of disturbances of brain energy metabolism
	• Ca2+ ion entry in the cell • Oxygen free radical formation • Excitatory amino acids
Ion channel disturbances
	• Mitochondrial dysfunction, eg, zidovudine-induced mitochondrial myopathy • Neurotransmitter disturbances, eg, atropine produces memory impairment by reducing acetylcholine • Metabolite-mediated toxins, eg, MPTP neurotoxicity to dopamine cell in the substantia nigra
Drug-induced selective cell death
	• Unknown mechanisms
Disturbance of the proteome
	• Alterations of individual protein structure by drugs leading to loss of neuronal function

Secondary drug-induced neurologic disorders. The central nervous system is affected by changes in other organs. Many of the neurologic disorders are secondary to diseases of other organs. Similarly, the nervous system is affected by adverse drug reactions affecting other body systems. Such drug-induced neurologic disorders are referred to as secondary to distinguish them from primary drug-induced neurologic disorders for discussion of pathomechanisms. Examples of secondary drug-induced neurologic disorders according to the primary system involved are shown in Table 3.

Table 3. Examples of Secondary Drug-Induced Neurologic Disorders

Drug-induced cardiovascular disorders
	• Drug-induced cardiac arrhythmias can lead to dizziness, syncope, and cerebral ischemia. • Severe drug-induced hypotension may lead to cerebral ischemia. • Drug-induced hypertension may lead to hypertensive encephalopathy and facial nerve palsy.
Drug-induced hematological disorders
	• Coagulation disorders can lead to cerebral hemorrhage or thrombosis.
Drug-induced respiratory disorders
	• Respiratory disorders can lead to decreased ventilation and cerebral hypoxia.
Drug-induced renal disorders
	• Renal failure can lead to uremia and uremic encephalopathy.
Drug-induced hepatic disorders
	• Hepatic disorders can lead to hepatic encephalopathy.
Drug-induced electrolyte and metabolic disorders
	• Drug-induced hyponatremia and hypoglycemia can lead to convulsions. • Hyponatremia can produce cerebral edema.
Drug-induced vitamin deficiency
	• Drug-induced vitamin deficiency can lead to neurologic disorders such as peripheral neuropathy.
Drug-induced endocrine disorders
	• Drug-induced hypothyroidism can lead to decline of mental function, ataxia, and seizures. • Hyperthyroidism can produce tremor and myopathy.

Risk factors that predispose a patient to the development of drug-induced neurologic disorders are shown in Table 4.

Table 4. Conditions That Predispose a Patient to Drug-Induced Neurologic Disorders

	• Genetic predisposition: pharmacogenetic factors • Old age • Patients with a history of neurologic disorders • Use by normal subjects of drugs indicated for neurologic disorders • Compromised brain function:
		- degenerative brain disease - intracranial space-occupying lesions - brain damage, eg, stroke or head injury
	• Systemic diseases, eg, those involving the kidneys and the liver, AIDS • Noncompliance • Pregnancy: risk to the offspring from prenatal exposure to drugs

Multiple mechanisms of neurotoxicity from drugs. An example of this is neurotoxicity of amphetamine. Mechanisms that mediate this damage involve oxidative stress, excitotoxic mechanisms, neuroinflammation, the ubiquitin proteasome system, as well as mitochondrial and neurotrophic factor dysfunction (34). These mechanisms are also involved in the toxicity associated with chronic stress and HIV infection, both of which have been shown to enhance the toxicity to methamphetamine. Repetitive use of methamphetamine causes sustained elevations of central monoamines that can manifest as a variety of neurologic disorders such as cognitive disorders, drug-induced psychosis, and movements disorders (29).

Drug-induced perturbations of cholesterol homeostasis cause mitochondrial DNA disorganization, whereas mitochondrial DNA aggregation in the genetic cholesterol trafficking disorder Niemann-Pick type C disease further corroborates the interdependence of mitochondrial DNA organization and cholesterol with involvement of the ATAD3 gene as well (09). Thus, the dual problem of perturbed cholesterol metabolism and mitochondrial dysfunction may play an important role in the causation of neurologic disorders.

Another example is that of highly active antiretroviral therapy (HAART), which has increased survival rates of patients infected by HIV, but HIV-associated neurocognitive disorders (HAND) may persist. An experimental study to examine the effect of highly active antiretroviral therapy on biomarkers of senescence in primary cultures of human has shown that it induces inhibition or arrest of cell cycle, senescence-associated beta-galactosidase, reactive oxygen species, as well as upregulation of mitochondrial oxygen consumption. These changes in mitochondria correlate with increased glycolysis in highly active antiretroviral therapy drug treated astrocytes (06). Taken together, these results indicate that highly active antiretroviral therapy drugs induce the senescence program in human astrocytes and that highly active antiretroviral therapy may play a role in the neurocognitive impairment observed in HIV-infected patients.

Pharmacogenetics. This concerns the study of influence of genetic factors on the action of drugs and is relevant to adverse drug reactions. One reason for the high incidence of serious and fatal adverse drug reactions is that existing drug development does not incorporate genetic variability in pharmacokinetics and pharmacodynamics of new drug candidates. Polymorphisms in the genes that code for drug-metabolizing enzymes, drug transporters, drug receptors, and ion channels can affect an individual's risk of having an adverse drug reaction, or they can alter the efficacy of drug treatment in that individual. Genetic aberrations associated with adverse reactions are of 2 types. Most of these arise from classical polymorphism in which the abnormal gene has a prevalence of more than 1% in the general population. Toxicity is likely to be related to blood drug concentration and, by implication, to target organ concentration as a result of impaired metabolism. The other type is rare, and only 1 in 10,000 to 1 in 100,000 persons may be affected. Most idiosyncratic drug reactions fall into the latter category. Examples of adverse reactions with a pharmacogenetic basis are malignant hyperthermia and extrapyramidal movement disorders in psychiatric patients on neuroleptic therapy.

Enzymes controlling the drug metabolism are relevant to pharmacogenetics. The cytochrome P450 enzyme system consists of a large family of proteins that are involved in the synthesis and/or degradation of a vast number of endogenous compounds as well as the metabolism of exogenous toxins. P450 enzymes can alter, abolish, or enhance drug metabolism. Most of the clinically used drugs are cleared through the action of P450 enzymes: CYP2D6, CYP3A4, and CYP2C19. Drugs used in neurology and psychiatry practice, which are metabolized by CYP2D6, include the following:

• Amitriptyline • Clomipramine • Clozapine • Codeine • Fluoxetine • Fluvoxamine • Haloperidol • Imipramine • Mexiletine • Nortriptyline • Olanzapine • Paroxetine • Phenacetin • Risperidone • Sertraline • Tramadol • Venlafaxine

CYP2C19 is the gene encoding S-mephenytoin hydroxylase, and its mutations lead to poor metabolism of various drugs that include amitriptyline, citalopram, clomipramine, diazepam, imipramine, and mephenytoin. AmpliChip CYP450 microarray enables clinical diagnostic laboratories to identify polymorphisms in CYP2D6 and CYP2C19, which play a major role in drug metabolism (15).

Expression of ATP-dependent efflux transporter, P-glycoprotein, at the level of the blood-brain barrier limits the entry of many drugs into the central nervous system. Single nucleotide polymorphisms in P-glycoprotein are a potential determinant of interindividual variability in efficacy and toxicity of drugs on the central nervous system.

Genotyping may be able to identify individuals at risk of reactions to certain drugs, but multiple genes involved in drug reactions may be difficult to detect. A patient may have a life-threatening reaction due to other factors that influence the onset of drug reactions, such as drug-disease interaction and drug-drug interaction.

Although genetic associations with individual idiosyncratic adverse drug reactions have been identified, further studies are needed to provide insights into the mechanism of drug interaction with the gene variant that leads to a phenotype, which can manifest in several clinical forms with variable severity.

Peripheral neuropathy is a side effect of proteasome-inhibitor bortezomib used for the treatment of multiple myeloma, which may limit its use in some patients. A genome-wide association study on multiple myeloma patients treated with bortezomib has revealed 4 new single nucleotide polymorphisms (SNPs) for bortezomib-induced peripheral neuropathy at 4q34.3 (rs6552496), 5q14.1 (rs12521798), 16q23.3 (rs8060632), and 18q21.2 (rs17748074), suggesting a genetic basis for neurotoxicity (04). Further research is needed to clarify the mechanism of action of the identified single nucleotide polymorphisms in the development of drug-induced peripheral neuropathy.

Epigenetic factors. Gene regulation by genetics involves a change in the DNA sequence whereas epigenetic regulation involves alteration in chromatin structure and methylation of the promoter region, representing a means of inheritance without associated DNA sequence alterations (08). Epigenetic factors may explain changes in gene expression that persist long after drug exposure has ceased. One hypothesis proposes that some iatrogenic diseases, such as tardive dyskinesia, are epigenetic in nature.

Drug-drug interaction. Drug-to-drug interactions contribute significantly to adverse reactions to drugs. There may be unpredictable potentiation of synergistic effect resulting in neurotoxicity if 2 drugs with similar mechanisms of action are used concomitantly. The interaction of a drug acting on the nervous system with a drug acting on another system that interferes with the metabolism of the CNS drug may lead to neurotoxicity. Anticonvulsants have clinically significant interactions with other drugs used in brain tumor patients. A drug interaction resulted in elevated phenytoin levels after initiation of erlotinib therapy in a patient who was receiving phenytoin (13).

In polypharmacy there may be a combination of drug-drug and drug-disease interactions. An example is serotonin syndrome that developed after the use of tramadol and ziprasidone in a patient with an indwelling deep brain stimulator for Parkinson disease (10).

Patients with a history of neurologic disorders. Patients who have had neurologic signs and symptoms in the past are more likely to have a recurrence of these as adverse drug reactions.

Use by normal subjects of drugs indicated for neurologic disorders. Drugs given to modulate brain function counteract an abnormal function and aim to achieve a balanced neurologic status. Use of such drugs by volunteers in studies and misuse by persons without neurologic disease may produce disturbances of neurologic function. Neurologic patients who have been taking certain medications for long periods usually develop tolerance to minor adverse reactions seen at the start of the therapy. Another mechanism, seen for example in epileptic patients on antiepileptic drugs, is enhanced metabolism by hepatic p450 enzyme. Nonepileptic patients taking these drugs for the first time may show adverse effects after doses that are well tolerated by patients on chronic antiepileptic drug therapy.

Patients with compromised brain function. The human brain has a considerable reserve capacity and plasticity, which enables it to function without obvious deficits even after parts of it are rendered inactive; however, any further insult, such as drug toxicity, may lead to decompensation and manifestations of neurologic disorders.

Systemic diseases. Patients suffering from systemic diseases are more prone to develop adverse reactions to drugs. Drug-induced neurologic disorders are more likely to develop in patients with diseases of the kidneys and liver. Several cases of neurotoxicity resulting from treatment of herpes zoster with acyclovir or valacyclovir have been reported in patients with renal failure, whereas this is a rare side effect of these drugs in patients with normal renal function (28).

Patients with AIDS have multiple diseases and impairment of the immune system. They have a disproportionate share of adverse drug reactions including those involving the nervous system. These patients do not tolerate tricyclic antidepressants well because of the anticholinergic effects of these drugs. They are more prone to develop drug-induced peripheral neuropathy with zidovudine.

Risk factors related to drugs. Factors related to the drugs that influence neurotoxicity are:

Dose. Some of the neurotoxic effects are dose-related, and some occur only as an overdose effect. Management of non-neurotoxic side effects of drugs enables higher doses of these drugs to be given, leading to neurotoxicity. An example of this is high-dose chemotherapy enabled by the addition of granulocyte colony stimulating factor (filgrastim), which counteracts neutropenia. Eventually, neurotoxicity becomes the limiting factor with this approach.

Rate of drug delivery. Faster rates of intravenous delivery of some drugs may produce transient neurologic disturbances.

Route of administration. Some drugs, which usually do not cross the blood-brain barrier, have neurotoxic effects only when applied directly to the CNS such as by intrathecal or intraventricular routes.

Drug interactions. Interactions may occur with concomitant medications.

Drug-disease interaction. Drugs may unfavorably influence the course of the neurologic condition being treated.

Drug withdrawal. Withdrawal of drugs after prolonged use may produce neurologic disorders.

Principles of management of drug-induced neurologic disorders

Only general principles are stated here, and specific management of various disorders is described in other articles and in more detail elsewhere (16).

Clinical diagnosis. A careful history of drug use and a high index of suspicion are important factors in the diagnosis of drug-induced neurologic disorders. The physician should obtain a complete list of current medications, medications the patient has taken previously, as well as other possible offending medications that might be available from family members and other sources. Drugs most likely to cause such disorders should be considered first. For example, among drug-induced movement disorders, antipsychotic drugs and other dopamine-receptor blocking agents should be high on the list of suspected drugs.

Drug-induced neurologic disorders should be considered in assessment of neuropsychiatric problems in cancer patients. In a prospective study, akathisia, a drug-induced movement disorder, was found in 4.8% of cancer patients and was caused by an antiemetic drug, prochlorperazine (17).

Diagnostic procedures. Laboratory studies of patients with neurologic disorders induced by drugs are like those used for nondrug-induced disorders. For example, brain imaging techniques can directly or indirectly assess many mechanisms of anticancer drug-induced neurotoxicity. Functional MRI, diffusion tensor imaging, and MR spectroscopy are noninvasive techniques that can yield important complementary data regarding the nature of neural changes after chemotherapy. A non-invasive nucleic acid-based MRI technique that enables real-time assessment of gene transcription profiles in the living brain could provide useful information, not only about naturally occurring diseases, but about drug-induced neurologic disorders as well (22).

MRI lesions of the splenium can develop in some epileptic patients after rapid anticonvulsant withdrawal. CAR-T-cell therapy-induced neurologic deficits are not always easily observable clinically and MRI is mostly normal, but diagnostic studies that more directly evaluate neuronal functioning, like EEG and PET scan, can reliably detect and predict neurologic dysfunction (27).

In drug-induced neurodegeneration, reactive glial cells at the site of degeneration are characterized by an abundance of peripheral benzodiazepine receptors as compared to surrounding neurons and provide an indirect biomarker for reliable estimation of central nervous system damage. Increased binding of ligands to peripheral benzodiazepine receptors can be measured by in vivo positron emission tomography.

Single photon emission computed tomography using (123)I-ioflupane or (123)I-iodobenzamide or both, can be used to differentiate between idiopathic Parkinson disease and drug-induced parkinsonism.

Detection of biomarkers of neurotoxicity may be useful for detection of toxic effects of drugs in the absence of clinical features or any clues in routine laboratory studies and imaging procedures. Detection of genetic biomarkers associated with neurotoxicity induced by single-agent and combination platinum chemotherapy can be useful in reducing the adverse effects of chemotherapy for cancer (23).

Cognitive safety assessment is important at all stages of drug development, particularly during clinical trials. The Clinical Trials Information System Adverse Effects device, an iPad-tablet product on cloud computing, is available for measurement of cognitive impairment in phase I clinical trials. This provides fast, sensitive, and reproducible data that enables early informed decisions to be made about the cognitive safety of drugs, reducing the risk of costly failure in late-stage clinical trials.

Prevention. This is the most important approach to reduce the incidence of adverse reactions. When several therapeutic alternatives are available, the one with the least neurotoxicity should be selected, particularly in susceptible patients with risk factors. Dose-related adverse effects may be reduced by stepwise increase dosage over a longer period, allowing the patient to tolerate the dose.

Epidemiological studies indicate that antithrombotic agent-associated cerebral hemorrhages constitute a significant proportion of adverse drug reactions. Preventive measures should be considered to reduce these fatalities.

The availability of an in vitro model of stem cell-derived human neural cell lines enables investigators to study molecular mechanisms of various drugs that produce early developmental apoptosis of neurons in humans (03). Use of this model will enable identification of drugs that are less likely to cause death of neurons. Adverse drug reactions may be reduced by newer methods of targeted controlled drug delivery, particularly genetically engineered cell-based therapy. Genotyping may identify patients who would develop adverse reactions to certain drugs; those drugs should be avoided in these subsets of patients. Application of pharmacogenetics, by individualizing treatment to patients for whom it is safe, provides a rational framework to minimize the uncertainty in outcome of drug therapy and should significantly reduce the risk of adverse reactions to drugs.

Improved understanding of the molecular mechanisms of drug-induced neurologic disorders can lead to strategies for prevention by devising better therapies. One example is antipsychotic-induced extrapyramidal disorders, which are due to blocking of dopamine receptors involved in drug action. Preclinical studies in the search for new drugs with fewer extrapyramidal side effects suggest the role of 5-hydroxytryptamine (serotonin)-1A and 2A/2C receptors in the modulation of dopaminergic neurotransmission that will improve therapeutic strategies for schizophrenia with safer drugs (30).

Neuroprotective agents can prevent neurotoxicity induced by chemotherapeutic agents used for the treatment of cancer. An experimental study in rats has shown that dl-3-n-butylphthalide has a therapeutic potential against neurotoxicity induced by doxorubicin, an anticancer agent, by attenuating oxidative stress, inflammatory reaction, and neural apoptosis (21).

Treatment. Usually, the first step is discontinuation of the responsible drug in case of severe adverse drug reactions. Symptomatic treatment may be given for persisting symptoms. The specific treatment is based on knowledge of pathomechanism of the drug-induced neurologic disorder. For example, management of drug-induced movement disorders in the older patient requires careful consideration of the contraindications imposed by such agents as anticholinergics, which can induce dementia in this patient population.

In the case of secondary drug-induced disorders, the primary cause should be corrected. A boy who developed hypertension due to atomoxetine therapy for ADHD presented with facial palsy but recovered following discontinuation of the drug and normalization of blood pressure (19).

Special considerations

Patients with preexisting neurologic disorders, renal insufficiency, and advanced age may be particularly susceptible to neurotoxicity of antibiotics, and this can be reduced by dosage adjustments in high-risk populations (14).

Geriatric age group. The elderly who are frequently on multidrug therapy and are 2 to 3 times more likely to experience adverse reactions to drugs than young people. The brain is an especially sensitive drug target in old age. Elderly patients are particularly prone to adverse effects of psychotropic agents, including benzodiazepines, because of the altered pharmacokinetics. Memory and psychomotor impairment are more pronounced in the elderly on benzodiazepine treatment as compared with younger controls. The elderly are also more susceptible to antimuscarinic effects of some antidepressants and neuroleptic drugs and may manifest agitation, confusion, and delirium. These drugs should be used sparingly in elderly patients. If their use is essential, the dosage should be titrated to a clearly defined clinical or biochemical therapeutic goal starting from a low initial dose.

Pregnancy. The potential of drugs used by the pregnant mother to harm the fetus is well known, with risk of teratogenicity. This is considered for each of the drugs used for neurologic disorders that are described in MedLink Neurology. Antiepileptic therapy in pregnancy is most often implicated in the etiology of malformations of the neural tube. There is a controversy about the causal relationship of drugs to congenital malformations. However, prenatal antiepileptic drug exposure in the setting of maternal epilepsy is associated with developmental delay and later childhood morbidity in addition to congenital malformations. Valproic acid use as an antiepileptic in pregnant women is associated with neural tube defects in the offspring. Underlying genetic susceptibility is a factor in the etiology of this adverse effect.

Drugs used for other diseases can also have adverse effects and produce neurologic disorders in the offspring that may not manifest clinically until later in infancy or childhood. Prenatal exposure to drugs has also been implicated in the development of neurologic disorders in later life, and epigenetic factors may be involved. Analysis of transcriptome data from in vitro models has enabled prediction of toxicity patterns of drugs in humans (02). Changes in the histone methylation pattern could explain how drug-induced changes are perpetuated over time even after washout of the drug, enabling transformation of an early insult to an adverse effect later in life.

In the case of psychotropic medications such as opioids, antidepressants, antiepileptics and antipsychotics, a withdrawal syndrome or “neonatal abstinence syndrome” is caused by the discontinuation of the drug transfer from the mother to the newborn after delivery, with symptoms that may include tremors, irritability, hypotonia/hypertonia, vomiting, and persistent crying, which can occur a few hours to 1 month after delivery (07). Neonatal neurologic and behavioral effects can also be caused by residual drug in the blood and tissues of the newborn.