Epilepsy has often been described as the great imitator with protean paroxysmal manifestations from sudden arousals out of sleep, to confusional states, to stiffening only when the individual arises and turns to one side and not the other. However, each of the behaviors listed previously (and many, many others) can be equally well explained in many individuals by nonepileptic mechanisms. According to the International League Against Epilepsy, imitators of epileptic seizures are defined not by the presentation but by the absence of abnormal and excessive neuronal discharges. The International League Against Epilepsy (ILAE) subdivides nonepileptic events into physiological disturbances with a nonepileptic mechanism such as syncope, sleep disorders, paroxysmal movements, transient global amnesia, and migraine as well as nonepileptic events of psychogenic origin (which may occur in the same patient with documented epileptic seizures).1
There is an expression in medicine that “all that wheezes is not asthma.” Yet many parents and primary care physicians still assume that all paroxysmal events are epileptic in origin. This becomes even more problematic in high-risk individuals such as developmentally delayed toddlers with nocturnal arousals, academically challenged children who “zone out,” or adolescents with syncope. It often comes as a shock to many observers that so many individuals admitted to an epilepsy-monitoring unit (EMU), even those with undisputed epilepsy, have a nonepileptic basis for the event in question. A recent series of 223 children referred to a specialized epilepsy center in Denmark found that 39% did not have epilepsy.2 Even in a selected group from the Danish study where the referring physician was certain about the diagnosis of seizure disorder, fully 35% did not have epilepsy. It has also been demonstrated that there were occasional children diagnosed with medically refractory epilepsy who had a nonepileptic basis for their entire clinical syndrome.
The value of electroencephalographic (EEG) is indisputable, but nonspecific abnormalities or unrelated epileptiform features like rolandic spikes may be wrongly interpreted to explain confusional episodes, headaches, or attention deficit disorder. A positive response to clinical trials or antiepileptic medication may not always provide a definitive answer to the challenge of the etiology of paroxysmal behavior such as aggression, headache, or movement disorders. Certainly there are many antiepileptic drugs such as lamotrigine, carbamazepine, and oxcarbazepine that are used (off-label) for mood stabilization, and topiramate and valproate have been approved for migraine prophylaxis. While it has long been recognized that carbamazepine can induce tics, there are recent reports demonstrating the value of topiramate and levetiracetam for controlling symptoms of Tourette syndrome.
It is difficult to organize the imitators of epilepsy into a pathophysiologically based or clinically relevant schema. One could choose syndromes that occur primarily in sleep versus waking, disorders of early childhood or older childhood into adolescence, or paroxysmal disorders with presenting symptoms such as unusual movements or altered awareness, respiratory abnormalities, perceptual changes, or sudden behavioral alterations. This chapter will review the major imitators of epileptic seizures in children and try to combine and synthesize all of these approaches starting with sleep disturbances at different ages followed by movement disorders during the day, breathing abnormalities, and acute behavioral changes.
Sleep starts, also called massive hypnogogic myoclonus, is a normal phenomenon at the transition from wakefulness to sleep.3 This is almost always unrecognized by the individual but commonly reported by parents or bed partners. It is unusual to have single hypnic jerks come to medical attention, but occasionally one sees repetitive, prominent jerking, or rarely injury from falling out of bed. The myoclonic jerks are of subcortical origin, and can be brought out by marked fatigue, stress, interaction with caffeine, or other stimulants. Reassurance that it is a normal phenomenon is all that is necessary, and no specific treatment is indicated beyond avoidance of known precipitants.
More persistent migratory myoclonus is also reported during light stages of non-rapid eye movement (non-REM) sleep, and it typically abates after the first hour as sleep staging descends into N3 (deep non-REM sleep). Asymmetric and asynchronous movements of the facial muscles, trunk, or extremities are more common with narcolepsy and periodic limb-movement disorder.
Restless legs (RLs) syndrome is an unusual phenomenon in children probably because it requires subjective reporting, but the associated objective finding of periodic limb movements (PLMs) can be found in children. PLMs have actually become a fairly common finding now that surface limb electromyography (EMG) electrodes have become a standard feature of pediatric polysomnography.
RLs syndrome is a common complaint in adults and is characterized by an uncomfortable, sometimes painful sensation in the extremities, usually more pronounced in the legs; it typically gets worse as the day progresses and peaks at night. Affected individuals report the inability to get comfortable and may walk around in an effort to relax. Often, the uncomfortable sensation and need to move around interferes with the ability to fall asleep or to achieve consolidated sleep. While uncommon in children, a detailed history can sometimes elicit the same complaint. Often it is attributed to “growing pains.”
Most affected individuals have associated PLMs, especially children and adolescents.4 PLMs are clusters of brief (0.5–5 seconds) movements of the extremities recurring every 10–60 seconds. These are usually subtle in children and rarely produce full awakening, but brief minor arousals are common and may contribute to sleep fragmentation with resultant daytime irritability and impaired concentration mimicking attention deficit hyperactivity disorder (ADHD).
While the pathophysiology of PLMs and RLs are incompletely understood, they may relate to dopaminergic dysfunction. Treatment usually consistent of repletion of low iron stores if found (dopamine hydroxylase requires an iron-containing cofactor). Dopaminergic agonists have become the first-line treatment otherwise, but benzodiazepines at bedtime can be used symptomatically, presumably working by decreasing arousability.
Rhythmic movements of sleep used be called jactatio capitis nocturna in recognition of the prominence of head banging, but legs or trunk are involved in many cases. It is felt to be a form of self-soothing that is most common in infants and typically disappears by 2–4 years of age. Typically starting in drowsiness, it may continue into light non-REM sleep.5
Rhythmic movement disorder of sleep is rarely confused with epilepsy. However, it becomes a problem when the motor symptoms are primarily described in the arousals from sleep rather than the typical self-soothing behaviors noted during drowsiness. It can also be quite challenging when rhythmic movements are found in children with developmental delay. Not only is the behavior more common in the disabled population but also it is often more severe and persists longer in the cognitively impaired population where the index of suspicion is already higher for epilepsy. Only in these situations have there been reports of serious injuries including subdural hematomas and even blindness.
Most non-REM parasomnias occur during the first third of the night predominately, but not exclusively, during transition from the first or second cycle of deep non-REM (N3) sleep, and cluster between 1 and 3 hours after sleep onset. Typical presentations include sleep talking (somniloquoy), sleep walking (somnambulism), agitated arousals, and full-blown night terrors (pavor nocturnus).6 Sleep talking is rarely confused with epilepsy, but sleep walking with complex automatic activity such as fidgeting and shuffling gait may be misinterpreted as a complex partial seizure, especially when it ends with return to sleep that mimics postictal unresponsiveness. Night terrors with agitated thrashing and screaming can look just like a partial seizure of frontal lobe origin, especially when there are no definable surface EEG abnormalities due to movement and muscle artifact or deep subcortical localization. All of the nonepileptic paradoxical arousals share a misperception and relative unresponsiveness to the environment, automatic behavior, and a variable degree of amnesia for the event. The EEG is characterized by muscle and motor artifact with persistence of slow (delta) activity of deep sleep or by a mixture of frequencies without features of full arousal.
Occasional confusional arousals are seen in the majority of normal preschoolers. Night terrors affect up to 6% of prepubertal children with a peak incidence of 5–7 years, while sleepwalking peaks at an older age (8–12 years).7 Many children have a family history of similar events and support a hypothesis of genetic predisposition, although environmental factors from febrile illness, to emotional stress, sleep deprivation, or medications are often reported.8 Anything that produces arousal at a vulnerable point in the sleep cycle can trigger typical motor patterns from forced awakenings to obstructive sleep, and conversely treating the underlying sleep disorder can make parasomnias disappear.9 As with any behavior that can be caused by different etiologies, it is important to consider the rare possibility of refractory night terrors proven to have an epileptic mechanism.10
REM behavior disorder (RBD) is an unusual parasomnia in childhood, except in the few recognized cases associated with narcolepsy or brain stem pathology such as pontine gliomas.11 In this condition, the usual muscular paralysis of REM is lost, and patients literally act out their dreams. RBD may be misinterpreted as hypermotor epileptic seizures. Pointing toward RBD is the presence of dream content, the lack of stereotyped movements, and any report of typical generalized tonic–clonic seizures.
Because seizures are always in the differential diagnosis, it is important to recognize the differences and similarities between seizures and parasomnias.12
While video-EEG is clearly the only way to definitively distinguish epileptic seizures from other “things that go bump in the night,” there are certain features that can be helpful. Nocturnal seizures are usually stereotypic, are not limited to the first third of the night, occur most often out of lighter stages of non-REM sleep, and are usually associated with epileptiform EEG abnormalities (Fig. 22–1).
Nonepileptic paroxysmal movement disorders during wakefulness are a diagnostic challenge that can mimic seizures, and conversely there are epileptic seizures that are misinterpreted as movement’s disorders. Abrupt paroxysmal, involuntary movements including chorea, athetosis, dystonia, and tics can be confused with seizures as well as disorders of impaired coordination such as tremor and paroxysmal ataxia.
Affected children have movement-induced attacks of brief dystonia, chorea, or ballismus, often brought out under stressful conditions.13,14 Paroxysmal dystonic choreoathetosis (Mount Reback syndrome) often starts very early in infancy; it is usually an autosomal dominant disorder with relative brief attacks lasting a few minutes that can be brought out by fatigue, hunger, stress, caffeine, or alcohol. Paroxysmal kinesigenic choreoathetosis typically presents in childhood and often includes a brief sensory prodrome with paresthesias and dizziness. Attempts to swing a bat or to write on the blackboard in front of the class can cause the bat or the chalk to fly out of the child’s hand. Events occur frequently, sometimes many times per day. Most cases are sporadic although occasionally it occurs in families or in the context of a prior neurological insult. These very brief events, usually lasting seconds, typically respond to sodium channel drugs like carbamazepine or phenytoin even though the EEG is always normal during the event. Recently, there have been genetic breakthroughs.
Paroxysmal nonkinesigenic dystonia may look similar, except that the episodes come on without apparent precipitation by movement and events last much longer (usually 10 minutes to 2 hours); most are autosomal dominant. Both types present more often with fatigue, stress, alcohol, and fatigue, Paroxysmal nonkinesigenic dystonia is harder to treat. It does not respond to the sodium channel drugs and there is only minimal success with benzodiazepines.
Myoclonus is characterized by rapid, forceful, usually isolated, and nonrhythmic jerking movements. It is considered nonepileptic when the paroxysmal movements occur in the absence of an ictal EEG pattern and the patient has no other neurological signs and follows a benign course.
Nonepileptic myoclonus in older children can be a diagnostic challenge since it may occur in situations where epilepsy is also found. This includes following acute encephalopathies and in progressive degenerative disorders. When the movements are of cortical origin they can be considered myoclonic seizures, and the EEG will demonstrate spike or polyspike slow activity. Both cortical and subcortical myoclonus may respond to GABAergic drugs such as benzodiazepines or valproate.
Motor tics are stereotyped, repetitive movements of one or more muscle groups usually involving head, eyes, face, or neck.15 Most children are fully alert and many are aware of the mannerisms. Unlike seizures, it is usually possibly to voluntarily inhibit tics, at least for brief periods of time. They are not often confused unless the mannerisms overlap with features more typical for seizures such as eye blinking with apparent loss of awareness to suggest nonconvulsive status epilepticus or prolonged unilateral twitching of an extremity to be confused with epilepsia partialis continua. A pediatric epilepsy monitoring unit will see the occasional child referred for such repetitive, stereotyped movements that “seize” the child and leave him without the ability to suppress the rapid blinking, motor activity, or “automatisms” and leave him exhausted as in a postictal stupor.
A history of waxing and waning pleomorphic mannerisms that evolve over time, associated ADHD symptoms that typically precede the tics (seen in more than 60%), anxiety disorders including obsessive compulsive disorder and a family history of tics and related neuropsychiatric disorders point strongly away from epilepsy and toward a developmental disorder of basal ganglia and its connections. However, there are children who will have both typical tics and evidence for epilepsy. For example, both tics and seizures are frequent in autistic spectrum disorders. Under these circumstances, one might well consider treatment with topiramate or levetiracetam, both of which have some support for efficacy in treating tics as well as being broad-spectrum antiseizure medications (Fig. 22–2).
