Chapter 100 Poisoning and Drug-Induced Neurologic Diseases
Introduction
Poisonings are common. In 2003, nearly 2.4 million cases of poisoning were reported to the USA’s poison control centers, 93 percent of which occurred at home [Watson et al., 2004]. Of these, 1.58 million involved persons younger than 20 years of age, and 1.24 million (or 52 percent of all poisoning victims) were children younger than 5 years [Watson et al., 2004]. Intoxication is the second most common injury (behind falls) in children younger than 48 months [Agran et al., 2003]. Data from other countries suggest similar demographics [Andiran and Sarikayalar, 2004].
Neurotoxins are among the most commonly encountered toxins, although they are the culprits less commonly in children younger than 6 years [Watson et al., 2004]. Poisonings and drug-induced neurologic disease may mimic infection, trauma, neoplasm, psychiatric illness, or metabolic disorders. Intoxications should be considered in the differential diagnosis of a child with an unexplained change in sensorium, seizures, ataxia, involuntary movements, muscle weakness, or autonomic dysfunction. Prompt recognition and management of intoxication reduce both mortality and morbidity in most cases.
Neurotoxins may have more long-term cognitive sequelae. Unfortunately, studies of low-level environmental exposures often suffer from imperfect control of confounding factors [Mink et al., 2004; Weiss et al., 2004]; this situation may limit the conclusions that may be drawn from these studies.
Most childhood poisonings are accidental [Nhachi and Kasilo, 1994; Singh et al., 1995; Watson et al., 2004]. The most common agents implicated in children younger than 6 years are cosmetics, personal care products, and household cleaning solutions. Important factors contributing to accidental poisonings include parental factors, such as improper medicinal or chemical storage, lapses in child monitoring, and ignorance of poison control methods. Likewise, children may be noncompliant, may mistake a potential toxin for food, may imitate parental medication-taking behavior, or simply may be curious. One unusual example was that of an 11-year-old male who ingested 35 percent hydrogen peroxide solution. Although H2O2 exposure may result in irreversible cerebral injury, his magnetic resonance imaging (MRI) test demonstrated bilateral posterior reversible diffusion restriction; he later made a full recovery [Cannon et al., 2003]. Failure of “childproof” containers was cited in 18 percent of poisonings in one study [Brayden et al., 1993].
Adolescents and preadolescents are at risk for recreational drug and solvent abuse. The rate of substance abuse by preteens in the United Kingdom approaches 5 percent [McArdle, 2004]. Aggressive attempts to identify substances in such intoxications may reveal unsuspected agents [Dresen et al., 2007]. Poisoning occurs deliberately in suicide attempts, child abuse, and other attempts to harm a child. In a 2002 Polish study, 14 percent of poisonings in children younger than 15 years were intentional (including self-inflicted) [Kotwica and Czerczak, 2002]. Two-thirds of suicide victims younger than 19 years in the United Kingdom use self-poisoning [Camidge et al., 2003]. In Hong Kong, poisoning is implicated in 15–20 percent of murders and suicides in the general population [Chan et al., 2003]. These high percentages of nonaccidental poisoning deaths, compared with the United States, may reflect differences in firearm regulation. Among the pediatric population, those at highest risk for abuse are infants and preschoolers; teenagers have the highest risk for suicide [McClure, 1994; McClure et al., 1996]. Abuse of alcohol and other recreational drugs may predict suicide attempts, although the method may involve neither [Hawton and Fagg, 1992; Hawton et al., 1993]. Common suicide methods identified recently in teenagers include hanging, exhaust inhalation, and overdose on medications, including paracetamol, benzodiazepines, and tricyclic antidepressants [Andiran and Sarikayalar, 2004; Ghazi-Khansari and Oreizi, 1995]. Pesticides, herbicides, and caustic agents may also be used; fatalities caused by ingestion of such agents may be the result of suicidal intent rather than accidental exposure more often in adolescents than in adults [Andiran and Sarikayalar, 2004; Klein-Schwartz and Smith, 1997; Thompson et al., 1995].
In the United States, with a preponderance of firearm-related deaths, a fatal outcome of Russian roulette is often classified as suicide. This may be a forensic determination rather than a reflection of intent. Although cause of death is a gunshot wound, a recent South Carolina report notes alcohol and or marijuana use in 6 of 8 victims. In these cases, the recreational agents likely are potentiating adolescent risk-taking behaviors with accidentally fatal outcomes [Collins, 2010].
Poisoning as a result of child abuse usually occurs in conjunction with a background of other, separate injuries. Still, poisoning may occur in isolation or within a more cryptic history of unexplained illnesses; this is especially evident in Munchausen’s syndrome by proxy [Chadwick, 1997]. A diagnosis of Munchausen’s syndrome by proxy suggests the secondary gain of medical contact as motive; it is sometimes difficult to differentiate Munchausen’s syndrome by proxy from other causes of nonaccidental injury. The incidence of nonaccidental poisoning in the British Isles is more than 2.8 per 100,000 children younger than 1 year each year, and more than 0.5 per 100,000 children younger than 16 years each year [McClure et al., 1996]. Methods reported include forced ingestion of antiepileptic drugs, opioids, and caustics, as well as various other agents [Gotschlich and Beltran, 1995; McClure et al., 1996]. Unfortunately, escalation of Munchausen’s syndrome by proxy often occurs, with suffocation a frequent terminal event [Chadwick, 1997; McClure et al., 1996]. One instructive case of Munchausen’s syndrome by proxy was that of an 11-year-old female with a history of cyclic vomiting. Investigation eventually revealed arsenic poisoning by her mother [Embry, 1987].
Conversely, some children and adults present with complaints of symptoms that are believed to be the result of a toxic exposure, but for which no toxin is identified. In these cases, symptoms often are vague; litigation is common (30 percent), and the patients or families do not easily accept refutation of the alleged exposure [Leikin et al., 2004].
Finally, chemical agents have been used as weapons against groups of people. Although their use in battle is not new, recent events have raised concern over covert use of chemical agents to attack civilian populations. In the 1994 Matsumoto nerve agent terrorist attack using sarin, 58 of 600 exposed people required hospital admission, and 7 died. The most common symptom was miosis, as the victims were exposed via vapors rather than direct contact with liquid nerve agent [Newmark, 2008]. Central nervous system (CNS) effects and transient cardiomyopathy also occurred [Okudera, 2002]. The U.S. Department of Health and Human Services has published a brief guide to toxidromes associated with likely chemical weapons used covertly [Patel et al., 2003].
Clinical trials, preclinical studies, anecdotal reports, and epidemiologic data have contributed to our understanding of neurotoxins [Erinoff, 1995; Fray and Robbins, 1996; Indulski and Lutz, 1996; Kurz et al., 1995]. Detailed discussion of antiepileptic drugs and immunization side effects is omitted here; these topics are covered in other chapters.
Emergency Evaluation and Management
Management of the poisoned child requires skilled immediate stabilization of the patient and appropriate corrective and supportive therapy. It requires also that the physician review the history and examine the child carefully for clues that may suggest poisoning or drug effects. Discovery of such evidence is not always easy, and a high index of suspicion is necessary. Three-quarters of all poisonings are by ingestion. Although cosmetics and household cleaning solutions are ingested, medications account for most deaths [Watson et al., 2004; Bronstein, et al., 2008]. Careful physical examination helps to establish the cause of the child’s distress and guide therapy. Neurologic findings may result from the drug or toxic agent itself, or from CNS hypoxia or ischemia caused by a generalized disturbance of circulation and respiration. Systemic abnormalities that occur after various types of intoxication may include cardiac dysrhythmias, gastrointestinal disturbances, and varying degrees of metabolic acidosis. Dysrhythmias, including marked degrees of bradycardia with cyanide or physostigmine intoxication, may be clues to the identity of the toxic agent and signal the need for emergency treatment. Cardiac conduction abnormalities, such as prolonged QT interval with phenothiazine overdose and widened QRS interval with tricyclic antidepressants, quinine, or quinidine overdose, may be present. Gastrointestinal complications, including severe diarrhea and vomiting, may occur with lithium, mercury, phosphorus, arsenic, mushroom, and organophosphate poisoning. Metabolic acidosis with a large anion gap may result from intoxication with cyanide, methyl alcohol, ethylene glycol, propylene glycol, and salicylates [Gardner et al., 2004]. Finally, steps should be taken to protect the child from future exposures. Appropriate general management steps are outlined in Box 100-1; details may be found in several references [Arena, 1985; Banner et al., 1994; Chan et al., 1993; Goetz, 1985; Gosselin et al., 1984; Haddad and Winchester, 1983; Osterhoudt et al., 2004; Leikin and Paloucek, 1995; POISINDEX, 2010; TOXNET, 2010; Zimmerman, 2003, Newmark, 2008]. Therapy for specific toxidromes is discussed with the individual agents.
Box 100-1 Suggested General Management of Suspected Intoxications and Poisoning
Testing
The value of routine toxicologic testing in cases of possible intoxications is debated in the literature. The low sensitivity of specific tests and the inability to test for all possible agents, medicolegal concerns, and cost-effectiveness have fueled arguments that routine testing should be limited [Bond, 1995]. Ideally, history and physical examination narrow the list of possible exposures and direct specific toxicologic testing. However, young children may not provide adequate history, and poisoned children sometimes present with unfamiliar or atypical clinical features [Lifshitz et al., 1997]. These situations may require a more extensive search for the toxic agent.
Toxicologic screening methods have other limitations as well. Most are not designed specifically for use in children, although modified panels or pediatric protocols are available [Badcock and Zoanetti, 1996]. Many physicians are unaware of the agents actually identified by the blood and urine panels available at their institutions. A standard screen for drugs of abuse may detect barbiturates, benzodiazepines, opioids, amphetamines, cocaine metabolites (benzoylecgonine), phencyclidine palmitate, and marijuana metabolites (tetrahydrocannabinol) by immunoassay, but other drugs, such as lysergic acid diethylamide (LSD), may remain undetected [Bond, 1995]. Other broad drug screens may detect phenothiazines, tricyclic antidepressants, ethanol or other volatile substances, and sympathomimetic amines. Urine screens for drugs of abuse may detect 2–9 agents (e.g., tetrahydrocannabinol, cocaine, opioids, amphetamines, barbiturates, benzodiazepines, methamphetamines, phencyclidine palmitate, tricyclic antidepressants in Diagnostix kits). Other commercially available screens include EMIT (Behring Diagnostics, San Jose, California), Abuscreen (Roche Diagnostic Systems, Basel, Switzerland), and others. Mach et al. compared on-site screening to gas chromatography/mass spectroscopy (GC/MS) and found roughly 20 percent discordance between the screening test and GC/MS [Mach, et al., 2007]. Thin-layer chromatography and ultraviolet spectroscopy are also popular, but older tests, including crude spot tests, are still available. The sensitivity of the screens and the need for subsequent confirmation (i.e., by high-performance liquid chromatography) depend on the methods used. Likewise, routine urine screens for heavy metals may identify only arsenic, mercury, and lead. Results from some screens may be obtained in less than 1 hour, but broad screens may require 24 hours for results; specific tests may require a week [Pathology and Laboratory Medicine, 1994]. Hair sample-based toxicologic testing has become increasingly popular and overcomes many limitations of other modalities. The test is minimally invasive, does not rely on dilution methods, and gives historical information. Use of a wide variety of substances can be determined from hair samples, including drugs of abuse, haloperidol, antidepressants, sympathomimetics, and heavy metals, including mercury [Hoffman and Nelson, 2001; Leikin and Paloucek, 1995; Schoeman et al., 2009, 2010]. The primary cocaine metabolite, benzoylecgonine, and acidic, polar drugs are difficult to identify reliably in hair samples, and the test is probably not appropriate for acute toxicity testing situations, although, in the case of cocaine, qualitative and semiquantitative analysis of more chronic use can be demonstrated [Katikaneni et al., 2002; Stephens et al., 2004].
In a prospective pediatric emergency department study, using urine GC/MS drug screens, the best clinical predictors of poisoning were odor on the child’s breath, symptoms and signs consistent with poisoning, and poison actually on the child’s clothing [Hwang et al., 2003]. Positive predictive values for these three variables were 100, 92, and 86 percent, respectively.
Other Ancillary Testing
Ancillary testing is directed by specific clinical findings. Electroencephalography (EEG) is important to exclude subtle or subclinical seizures in intoxicated patients who display altered mental status. Neuroimaging studies have at least three roles. First, in the acutely intoxicated child or teenager, a cranial computed tomography (CT) scan may be required if there is suspicion of concomitant trauma or hemorrhage. Second, brain MRI findings may explain a patient’s neurologic findings when clinical history and other laboratory studies do not. As neuroimaging of an increasing number of poisoned children and adults has occurred, a couple of recurring MRI patterns have become evident. Agents that cause hypoxic injury, including cellular respiratory poisons, may result in an MRI picture of bilateral symmetric cortical and subcortical gray-matter injury, although white-matter changes occur as well [Halavaara et al., 2002; Kim et al., 2003; Rachinger et al., 2002]. This includes methanol poisoning [Karayel et al., 2010]. Medications used for immunosuppression and those that cause hypertension or changes in blood–brain barrier permeability may contribute to a characteristic neuroradiographic pattern of bilateral posterior reversible leukoencephalopathy (so-called posterior reversible leukoencephalopathy syndrome) [Renard et al., 2004]. Third, the increasing use of diffusion-weighted MRI sequences and apparent diffusion coefficient calculations (apparent diffusion coefficient mapping) often provides a method to delineate early cerebral injury.
Neurologic Examination
Examination of a poisoned child includes initial and serial assessment of neurologic status; this assessment guides clinical management, predicts prognosis, and often identifies the offending agent. Assessment of mental status is important because several poisons, medications, and recreational drugs may be associated with affective symptoms, and may also cause acute changes in sensorium, such as irritability, other affective symptoms, delirium, coma, lethargy, and seizures. Although psychotropic or analgesic medications have the greatest potential to cause these effects, antihistamines and other routinely used pediatric medications are common culprits [Bassett et al., 1996]. Environmental and biologic toxins likewise may cause change in sensorium, although more often in younger children. More chronic effects include dementia, subtle learning difficulties, and apparent psychiatric illness. Agents that cause seizures or changes in sensorium are listed in Box 100-2.
Box 100-2 Selected Agents that Cause Changes in Sensorium or Seizures
Medications
Cranial nerve examination may reveal specific toxin-induced syndromes. Decreased visual acuity or changes in color perception may suggest anticholinergic or cardiac glycoside toxicity, respectively. Papilledema reflects increased intracranial pressure and may suggest pseudotumor cerebri, or benign intracranial hypertension, caused by systemically administered steroids, excessive vitamin A intake, or use of outdated tetracyclines. Visual impairment and a macular cherry-red spot has been reported in association with dapsone poisoning. The patient had concomitant peripheral neuropathy, and the macular changes were considered secondary to toxic retinal damage [Abhayambika et al., 1990]. Pupillary dysfunction may be caused by medication use, abuse of street drugs, and exposure to various environmental or biologic toxins [Leikin and Paloucek, 1995; Slamovits and Glaser, 1990]. Nystagmus is common in intoxications, especially with antiepileptic agents; bradykinetic extraocular movements follow exposure to antidopaminergic medications, and extraocular muscle paresis is an important indicator of botulism poisoning [Glaser and Bachynski, 1990]. Loss of the blink reflex and facial paresis may occur as part of a generalized encephalopathy, such as in profound narcotic intoxications, or occur in isolation, such as in botulism or vincristine-induced neuropathy. Acute hearing loss resulting from poisoning is uncommon but may be caused by aminoglycoside use, toxicity, or overdose of nicotine or lithium. Vestibular dysfunction is one of the most common presenting symptoms of drug-induced neurologic syndromes, including those caused by exposure to antiepileptic drugs, antibiotics, and metals [Mount et al., 1995; Wood et al., 1996]. Vestibular dysfunction should be differentiated from vagally mediated “dizziness,” although both may result from intoxications. Isolated toxin-induced disorders of the pharyngeal and neck musculature (e.g., loss of ability to swallow in clostridial toxidromes) are rare, but dysgeusia is not; the latter occurs commonly with use of lithium-containing preparations. Table 100-1 lists toxins that cause cranial nerve deficits.
Motor weakness may be due to poisoning of the anterior horn cells, peripheral nerves, neuromuscular junction, or muscles. Proximal myopathies (e.g., caused by steroid use) or distal motor neuropathies (e.g., caused by metal exposure) should be differentiated from infectious, inflammatory, metabolic, and degenerative processes. Profound weakness may occur with use of depolarizing agents and neuromuscular blockers. Rigidity may result from either central or peripheral disinhibition (Boxes 100-3 to 100-5) [Katz et al., 1996; Reeves et al., 1996]. Extrapyramidal disorders may be caused by exposure to environmental toxins but also are well-recognized side effects of many medications (Box 100-6). Drugs implicated most frequently in movement disorders are antipsychotics and related agents, calcium channel antagonists, CNS stimulants, antidepressants, antiepileptic drugs, antiparkinsonian drugs, and lithium [Jimenez-Jimenez et al., 1997; Kerrick et al., 1995]. With severe intoxications, the cardiovascular effects of many of these medications overshadow the acute neurologic symptoms. Antibiotics, antidepressants, and some antineoplastic drugs may cause myoclonus (Box 100-7) [Chow et al., 2003].
Box 100-4 Selected Agents Causing Peripheral Neuropathy
(Additional references: Le Quintrec and Le Quintrec [1991]; Dyck and Thomas [1993]; Leikin and Paloucek [1995]; Patel et al. [2003]; REPROTOX [2010].)
Box 100-5 Selected Agents Associated with Paralysis and Muscular Rigidity
Box 100-6 Selected Agents Associated with Parkinsonism and Other Acute Extrapyramidal Reactions
