Epilepsiu, 32(3):322-328, 1991 Raven Press, Ltd., New York 0 International League Against Epilepsy
Nonepileptic Events in Childhood *tM. E. Metrick, *tF. J. Ritter, *tJ. R. Gates, *M. P. Jacobs, "S. S. Skare, and *R.B. Loewenson *Epilepsy Research Center, Department of Neurology, University of Minnesota; and tMINCEP Epilepsy Cure, P.A., Minneapolis, Minnesota, U . S . A .
Summary: The medical records of 27 children admitted to the MINCEP Epilepsy Program for evaluation of intractable epilepsy but later shown to have nonepileptic events by EEG with simultaneous video monitoring were reviewed. Four groups were identified: pure psychogenic events ( 5 patients), psychogenic events plus epileptic seizures (3 patients), pure nonepileptic physiologic events (5 patients), and nonepileptic physiologic events plus seizures (14 patients). Historical data, physical examinations, and neurodiagnostic evaluations (including previous EEGs, neuroradiologic evaluations, and neuropsychologic testing) were reviewed. Children in all groups, except for those with pure psychogenic seizures, had a history of multiple seizure types identified by parents or caretakers. A history of status epilepticus was obtained in 64% (of 22 patients), including 1 1 of 14 patients with phys-
iologic events plus seizures. Abnormal findings on neurologic examination were common, especially in children with nonepileptic physiologic events. All but two patients had a history of interictal epileptiform abnormalities on previous routine EEGs. Based on identification of nonepileptic events, antiepileptic drugs (AEDs) were discontinued completely in eight patients (30%) and the total number of AEDs was reduced in nine others (33%). A diagnosis of nonepileptic events should be considered in all children with refractory seizures or multiple seizure types. Abnormal findings on routine (interictal) EEG may actually confound the diagnosis. Intensive neurodiagnostic EEG-video recording is the preferred method for distinguishing nonepileptic from epileptic seizures. Key Words: Epilepsy-Children-Nonepileptic-EEG monitoring.
Between 10 and 20% of patients referred to epilepsy centers for seizure control are later shown to have conditions other than epilepsy (Holmes et al., 1980; Scott, 1982; Duchowny et al., 1988). An epileptic seizure is defined as the paroxysmal change in behavior associated with a burst of epileptiform electrical activity in the brain. On the other hand, a nonepileptic seizure can be considered a paroxysmal change in behavior unassociated with bursts of epileptiform electrical brain activity. The nonepileptic events can be categorized as either psychogenic or physiologic. In adults, snd less often in children, psychogenic events mimicking true seizures have been recognized in both epileptic and nonepileptic populations. Scott (1982) estimated the prevalence of psychogenic seizures at 5% of an outpatient epilepsy population. In a study of 53 pediatric patients ad-
mitted to an epilepsy unit, Holmes et al. (1980) documented psychogenic seizures in 1 1 patients (21%), eight of whom also had epileptic seizures. Previous studies have identified several risk factors associated with an increased tendency for psychogenic seizures; in particular, a history of sexual abuse (Roy, 1977; LaBarbera and Dozier, 1980). In children, undetected learning disabilities have also been identified as precipitating factors (Silver, 1982). In addition to psychogenic events, some physiologic conditions in infancy and childhood resemble epilepsy. Many of the more common types of paroxysmal nonepileptic events are identified by history and physical examination; therefore, individuals who experience them are rarely referred to epilepsy centers for treatment or clarification. Common categories include breath-holding spells, pallid infantile syncope, gastroesophageal reflux associated with laryngospasm and apnea, night terrors and other parasomnias, and complicated migraine (Pedley, 1983). In addition, staring episodes and various repetitive behaviors that may not be clini-
Received August 1989; revision accepted May 1990. Address correspondence and reprint requests to Dr. M. E. Metrick at Lutheran General Children's Medical Center, 1775 Dempster St., Park Ridge, IL 60068, U.S.A.
322
323
NONEPILEPTIC EVENTS IN CHILDHOOD cally distinguishable from epileptic seizures are common in mentally retarded persons, who are also at increased risk for epilepsy. Several reports have described physiologic nonepileptic events that have been identified during inpatient evaluations at epilepsy centers. In a retrospective study by Duchowny et al. (1988), of 60 children admitted to an epilepsy unit, 40% of the children identified as having nonepileptic events had either staring episodes or rhythmic movements such as rocking, shaking, and arm waving. A study by Neil1 and Alvarez (1986) of nonepileptic events in 124 mentally retarded patients showed that the most frequently identified nonepileptic events (as determined by radiotelemetry video-EEG recording) included staring episodes, cessation of movement, and a variety of repetitive movements. Diagnosis of children with nonepileptic events may be difficult. The history given by the patient’s caretaker often fits the description of an epileptic seizure. Many of the children with nonepileptic events, especially those who also have seizure disorders, have epileptiform abnormalities on their routine (interictal) EEG. The incidence of epileptiform abnormalities, defined as paroxysmal discharges containing spikes or sharp-waves, has been estimated to occur in 2.2% of the nonepileptic population (Zivin and Ajmone-Marsan, 1968). Failure to recognize these patients can result in unnecessary use of antiepileptic drugs (AEDs), with the subsequent risk of long-term side effects, behavioral and cognitive disturbances, and systemic toxicity. In the extreme example, misdiagnosis of status epilepticus has led to treatment with general anesthesia and mechanical ventilation (Krumholz and Niedermeyer, 1983). In addition, use of AEDs may actually increase the incidence of nonepileptic events in patients with psychogenic seizures by exacerbating their mood disturbance (Trimble, 1983). At the same time, these patients are not receiving appropriate medical or psychiatric intervention. Use of simultaneous EEG-video monitoring to correlate behaviors with temporally related EEG changes may clarify the nature of the events and thus lead to appropriate treatment plans (Ramani et al., 1980; Desai et al., 1982; King et al., 1982). In this retrospective study, we reviewed the records of 222 children referred to an epilepsy center for treatment of intractable epilepsy, 27 of whom were later identified as having nonepileptic events. METHODS The records of all children aged < 16 years admitted to the MINCEP Epilepsy Program for Children
at Gillette Children’s Hospital in St. Paul, Minnesota, between August 1986 and August 1988 were reviewed. All patients had been referred for evaluation and treatment of seizures that had been refractory to standard medical therapy. The patient population included children with both normal and abnormal intelligence and patients with suspected neurodegenerative conditions. Both preadmission data and information obtained during hospitalization were reviewed. These data included historical information, physical examinations, and neurodiagnostic evaluations (radiologic studies, routine and video-EEGs, and neuropsychologic testing). While at Gillette Children’s Hospital, all children had a minimum of 24 h of consecutive EEG recordings with simultaneous video monitoring in both the awake, drowsy, and sleep states on 16-channel cable telemetry using the International 10-20 System. Spontaneous events were recorded both before and after AED dosages were reduced. Photic stimulation was performed on all patients, and hyperventilation was performed when possible. Sphenoidal electrodes were used in selected individuals. Both parents and trained observers, including nurses and nurses aides, assisted in documenting events. The presence of epileptic seizures was based on identification of paroxysmal discharges on EEG that were temporally linked to the documented episodes recorded on video tape. Events were classified as nonepileptic when there was a lack of epileptiform correlation to the episodes and when the events also were not stereotyped and did not progress into other seizure types as medications were decreased or eliminated. An ability to interrupt events by verbal or physical intervention also aided in differentiating nonepileptic from epileptic seizures. After nonepileptic events were identified, therapeutic regimens were examined to determine if proper diagnosis resulted in simplification of AED regimens or in redirection into appropriate therapies. Historical factors and findings on neurodiagnostic testing and physical examination, which differed between groups, were identified. When possible, these were assessed by statistical analysis (Fisher’s exact test); however, statistical analysis was limited by the small sample sizes. RESULTS Two hundred twenty individuals were admitted to the center during the study period. Of these, 27 patients (12%) were identified as having nonepileptic events by prolonged video-EEG monitoring (9 males and 18 females aged 7 months to 16 years; median age 8.4 years). The patients represented Epilepsia, Vol. 32, N o . 3, 1991
M . E . METRICK ET AL.
324
four groups: pure psychogenic seizures (5 patients), psychogenic events plus epileptic seizures (3 patients), pure physiologic events (5 patients), and physiologic events plus epileptic seizures (14 patients). The most frequent nonepileptic physiologic events in our population were a child's unusual reaction to environmental stimuli that had been misinterpreted by parents as seizures. These reactions included nonepileptic staring spells, repetitive stereotypic behaviors, and abnormal movements secondary to abnormalities in muscle tone. In two patients, the nonepileptic events consisted of paroxysmal choreoathetosis which in one patient was medication related. Differences among the four groups were identified. Patients with psychogenic nonepileptic events, with or without seizures (n = 8) had a median age of 12.7 years; patients with physiologic nonepileptic events, with or without seizures (n = 9), had a median age of 6.3 years. The children with physiologic nonepileptic events, especially those who also had seizures, were most likely to have movement disorders or major abnormalities on neurologic examination (Table 1). All five children with pure psychogenic events had normal intellectual function as determined by IQ testing and/or school placement, whereas all but one of the children in the other groups were mentally retarded (Table 1). Intelligence testing was compared between groups, using standardized IQ scores [Stanford-Binet or Wechsler Intelligence Scale for Children-Revised (WISC-R)]
when available. In the pure psychogenic subgroup, Full-Scale IQ scores on WISC-R ranged from 95 to 144. An IQ score was unavailable for one child with pure psychogenic seizures who was in a regular classroom setting. In the psychogenic plus seizure subgroup, two children had a WISC-R Full-scale IQ of 40 and 64, respectively. The third child, whose IQ scores were unavailable, was in a remedial educational program. In the pure physiologic subgroup, one child, aged 9 months, tested in the normal range for age on Bayley Scales. The other children evidenced mild to profound mental retardation, either as determined by formal IQ testing or by remedial school placement. In the mixed physiologic plus seizures subgroup, seven patients had no formal IQ testing because of profound mental retardation. The other children scored in the mild to severely mentally retarded range on standardized tests. The children with pure psychogenic events were female and, as evidenced by social services evaluation and psychologic testing, belonged to dysfunctional family units. None of the children with pure psychogenic events had a history of a first-degree relative with seizures. All children in the pure psychogenic group had only one seizure type identified by parents on admission, whereas multiple seizure types (range two to seven types) were identified in the other nonepileptic groups (Table 1). This difference was greatest between the pure psychogenic and the physiologic plus seizures groups (p =
TABLE 1. Results of neurologic examinations and tests of intellectual function, frequency of seizure types per patient identified at admission, and history of previous EEGs Psychogenic (n = 5) Major abnormalities on neurologic examination Dysmorphic features Movement disorder Normal intellectual function" Abnormal intellectual function" No. of seizure types 1 L
3 4 5
6 7 Mean no. of seizure types per person History of status epilepticus History of all EEGs normal History of all epileptiform EEGs History of both normal and epileptiform EEGs
Psychogenic + seizures (n = 3) 1
0 0 0 5 0
0 0 0 0 3
3 2 I 4
5 0 0 0 0 0 0 I .o 1 ( I unknown) 2 1
0 0 2 0 1 0 0 3.7 1 ( I unknown) 0 1
0 1 0 2 1 1 0 4.2 I (1 unknown) 0 2
2
l ( 1 unknown)
3
" Intellectual function was determined by IQ testing and/or school placement. Epilepsia, Vol. 32, No. 3, 1991
Physiologic (n = 5)
Physiologic + seizures (n = 14) 7 4 5 0 14 0 2
5 4 0 2 1 3.9 I 1 (2 unknown) 0 8 6
325
NONEPILEPTIC EVENTS IN CHILDHOOD O.OOOl), but was also significant between the pure psychogenic and the pure physiologic groups (p = 0.004) and the pure psychogenic and the psychogenic plus seizures groups (p = 0.018). In the pure psychogenic group, nonepileptic events resembled complex partial seizures, with a frequency of events ranging from 12 to 120 episodes per month. All these events had paroxysmal onset. In four of the 5 patients in this group the event began with a brief period of staring; the fifth patient usually cried out, or the event began with automatisms of the upper extremities. During these events, all 5 of the patients were unresponsive to verbal command, and 3 of the 5 resisted physical intervention. They all reported amnesia for the event and feeling tired after the nonepileptic episode. All 5 patients had movements described as automatisms, and during the event 4 of the 5 had intermittent stiffening for 3-10 s. Three of the 5 patients had episodes that included intermittent hyperventilation. In the other groups, the events resembled a variety of seizure types including generalized tonicclonic convulsions, complex partial seizures, absence seizures, and tonic, atonic, and myoclonic seizures. In the mixed psychogenic plus epileptic seizure group, episodes were quite variable. One patient had dysrhythmic jerking of the right upper extremity, associated with decreased responsiveness. Another complained of feeling warm, then became unresponsive to verbal intervention and complained of headache after the episode. The third patient in the psychogenic plus epileptic group had nonepileptic episodes of multiple types, including staring and decreased responsiveness that either did or did not progress into generalized stiffening, followed by shaking. The patients in the pure physiologic group all demonstrated repeated, paroxysmal nonepileptic events, each with motor activity accompanied by a decrease in responsiveness. These episodes varied from staring with apnea, followed by stiffening of all extremities, to nonepileptic myoclonus. The physiologic plus epileptic seizure group included eight children with movement disorders. The most common movement disorders observed were paroxysmal choreoathetosis, paroxysmal dystonia, and nonepileptic myoclonus. In this group, nonepileptic events with automatic, repetitive, self-stimulatory behaviors were also frequent. These behaviors always began abruptly, interrupting ongoing activity. More than half of these nonepileptic events began with staring and then rapidly progressed to the automatic repetitive behavior. An example was a teenage girl who-suddenly stopped all activity, stared, had a frightened look on her
face, turned her head toward the left, and repeatedly made chewing motions. She was unresponsive to verbal commands. At times, these events could be interrupted by physical intervention. A history of status epilepticus was obtained in 64% of the 22 patients for whom data were available and was particularly common in patients with physiologic events plus seizures ( 1 1 of 14 patients). Fourteen of the 27 patients, representing all groups except those with psychogenic events and seizures, had a history of abnormal computed tomography (CT) scans or magnetic resonance imaging (MRI) studies. Four patients had diffuse abnormalities, 7 patients had focal abnormalities, and 3 patients had both diffuse and focal abnormalities. Most patients in all groups had a history of epileptiform abnormalities on EEG, and at least 1 patient in each group had a history of both normal EEGs and EEGs with interictal epileptiform discharges (Table 1). Of the 10 patients with pure nonepileptic events, 7 had a history of interictal epileptiform abnormalities based on outside records (outside EEGs were not available for review). Five of the 10 patients had interictal epileptiform abnormalities recorded during their hospitalization at Gillette Children’s Hospital. Based on identification of nonepileptic events (in all groups), AEDs were completely discontinued in 8 patients (30%)and the total number of AEDs was reduced in 9 others (33%). All patients in the pure psychogenic and pure physiologic groups were no longer taking AEDs at the time of discharge. DISCUSSION
Most studies of nonepileptic events are based on the adult population in which psychogenic seizures appear to be the most common nonepileptic condition misdiagnosed as epilepsy. Relatively few studies have been made of nonepileptic events in childhood. In our study we observed that, in contrast to adults, children misdiagnosed with seizure disorders were more likely to have physiologic (19 patients) than psychogenic (8 patients) nonepileptic events. The basis for misdiagnosis was usually parental misinterpretation of behaviors, especially staring episodes, abnormal reactions to environmental stimuli, and repetitive movements. In addition, unusual movements based on abnormal muscle tone were often misinterpreted as seizures. Parents’ or caretakers’ descriptions of these movements and behaviors were often indistinguishable from those of epileptic seizures. Occasionally, a trained professional could identify the event as nonepileptic by Epitepsia. Vol. 32, N o . 3, I991
M . E . METRICK ET AL. observation alone, but often the distinction between epileptic and nonepileptic episodes could be made only after careful review of simultaneous EEGvideo monitoring. Other neurodiagnostic methods were often misleading. In particular, in 96% of these children, interictal EEG showed epileptiform abnormalities that might have led to an erroneous diagnosis. In our study, almost all patients in every group, including patients with pure psychogenic or pure physiologic events, had EEGs with interictal epileptiform abnormalities (Table 1). Epileptiform abnormalities on interictal recordings are particularly common in children with both epileptic and nonepileptic events, which emphasizes the need to observe and record all of the types of seizures in question. Other findings such as focal abnormalities on CT or MRI scan, observed in 37% of our patients, could further confound the diagnosis. The events in our study resembled a variety of seizure types. Pure psychogenic events were commonly misinterpreted as complex partial seizures, similar to the findings by Holmes et al. (1980). Events in the other groups resembled a variety of seizure types, including generalized tonic-clonic convulsions, absence seizures, and myoclonic, tonic, and atonic seizures. These differences may be related to the different etiologies of the nonepileptic events; e.g., staring, misdiagnosed as absence seizures, may reflect the decreased attention span common in children with varying degrees of mental subnormality. Head nods, misdiagnosed as atonic or tonic seizures, may reflect hypotonia. Our findings are similar to those of Duchowny et al. (1988) and Neil1 and Alvarez (1986), who reported that repetitive behaviors were commonly misinterpreted as seizures, particularly in the more severely mentally retarded. The children with pure psychogenic seizures may represent a patient population different from the other groups. Although our patient population was too small for statistical analysis of most variables, certain features in this group appear to differentiate them from the others. In particular, all these children were of normal intelligence, whereas all but one of the patients in the other groups had some degree of intellectual impairment. This may reflect the patient selection process of our center, but may also characterize children with intractable epilepsy. In addition, patients with pure psychogenic seizures generally had only one seizure type (mean number of types per group = 1 . l ) , usually resembling complex partial seizures (Table 1). Patients in the other groups had multiple seizure types identified by parents on admission; e.g., in the pure physiologic group, one patient was originally reported to have Epilepsia, Val. 32, N o . 3, 1991
six seizure types (mean for the group = 4.2). The children with pure psychogenic seizures were also older than those with pure physiologic events, with median ages of 12.8 and 2.6 years, respectively. Previous studies have identified some common stressors as precipitating psychogenic seizures, especially a history of sexual abuse o r undetected learning disabilities. Although our patients with pure psychogenic seizures had no documented history of sexual abuse, a previous history of abuse was suspected in several cases. All family units were identified as dysfunctional by social services personnel and/or by psychological evaluation. One of the patients, though of normal intelligence, had a Verbal and Performance IQ split of 24 IQ points (on the WISC-R) and was described as having difficulty integrating information on neuropsychological testing. These findings suggest that this patient, who was in a regular classroom setting, may have learning disabilities that contribute to her nonepileptic events. Although AEDs were successfully discontinued in all three patients in the psychogenic group who were admitted while receiving seizure medication, long-term follow-up will be needed to assess the impact of identifying these events properly. Certain features of the nonepileptic events, in addition to findings on video monitoring, supported the diagnosis of nonepileptic conditions. All nonepileptic events were refractory to conventional AEDs. Our patient population was selective in that most patients referred to our epilepsy program are refractory to AEDs. In addition, the number of patients who had been treated for status epilepticus because of lack of response to AEDs was striking; I 1 of 14 patients in the group of physiologic nonepileptic events plus seizures had a previous history of status epilepticus, an unusually high incidence which suggests that at least some of these episodes may have been nonepileptic in nature. Another characteristic of nonepileptic seizures that contrasted with those of epileptic seizures was the variability of seizure expression. Many of the patients with pure physiologic events had multiple seizure types identified by parents. Although patients with pure psychogenic seizures had only one seizure type, review of the video tapes disclosed a lack of stereotypy from one event to the next. This contrasts to the usually stereotyped epileptic seizures. Furthermore, these events almost always could be interrupted by either verbal and/or physical intervention, whereas true epileptic seizures could not be interrupted. The most difficult distinction was differentiating certain nonepileptic seizures from simple partial seizures, which may also show little or no surface
327
NONEPILEPTIC EVENTS IN CHILDHOOD TABLE 2. Reductions in AEDs at time of discharge
Status
Psychogenic (n = 5 )
Psychogenic + seizures (n = 3)
Physiologic + seizures (n = 14)
-.
~~
Medication reductions Complete discontinuation No. of patients whose AEDs were reduced Decreased dosage Total No AEDs on admission and discharge No change in AEDs Change to appropriate medication a
Physiologic (n = 5 )
0
1
1
2 0
8“ 2h 11 0
I
2
0
1
Ih
Five patients in this group had their medication changed to the appropriate AED.
* The patient’s medication was changed to the appropriate AED. AED, antiepileptic drug.
EEG expression. In addition to the above distinguishing features, certain additional clues helped us properly distinguish between simple partial seizures and nonepileptic events. The true epileptic seizures were more likely to progress to CPS or GTC convulsions as medications were discontinued and to respond to AEDs to some degree. On occasion, simple partial seizures that showed no EEG expression while the patients were receiving AEDs would show some corresponding EEG changes after AEDs were decreased. By taking into account these differences, we usually could distinguish between the two groups. In our patients, the nonepileptic events resulted in admission for refractory seizures, whereas epileptic seizures were adequately controlled. Proper identification of nonepileptic events resulted in a marked reduction in AED therapy. AEDs were completely discontinued in 30% of our patient population (8 patients), and the total number of AEDs was reduced in an additional 33% (9 patients) (Table 2). The patients in the pure psychogenic and pure physiologic groups had discontinued all AEDs at the time of discharge. If indicated, these patients were redirected into other treatment modalities. Thus, recognition of nonepileptic events had significant impact on the treatment plans for these children. Abnormal findings on EEG, neurologic examination, and neuroradiologic studies were common in our patients and actually confounded the appropriate diagnosis. Only intensive EEG with simultaneous video monitoring reliably led to the appropriate diagnosis. Thus, intensive neurodiagnostic EEG-video monitoring is the preferable method for identifying nonepileptic events. Its use should be considered in all children with multiple seizure types or refractory seizures. Acknowledgment: This work was supported in part by NIH-NINDS Grant No. P50 16308 and by a John Hugh-
lings Jackson Clinical Research Fellowship from the Epilepsy Foundation of America.
REFERENCES Desai BT, Porter RJ, Penry JK. Psychogenic seizures: a study of 42 attacks in six patients, with intensive monitoring. Arch Neurol 1982;39:202-9. Duchowny MS, Resnic TJ, Deray MJ, Alvarez LA. Video EEG diagnosis of repetitive behavior in early childhood and its relationship to seizures. Pediatr Neurol 1988;4:1 6 2 4 . Holmes GL, Sackellares JC, McKiernon J, et al. Evaluation of childhood pseudoseizures using EEG telemetry and video tape monitoring. J Pediatr 1980;97:554-8. King DW, Gallagher BB, Murvin AJ, et al. Pseudoseizures: diagnostic evaluation. Neurology 1982;32:18-23. Krumholz A, Niedermeyer E. Psychogenic seizures: a clinical study with follow-up data. Neurology 1983;33:498-502. LaBarbera JD, Dozier JE. Hysterical seizures: the role of sexual exploitation. Psychosomatics 1980;21:897-903. Neil1 JC, Alvarez N. Differential diagnosis of epileptic versus pseudoepileptic seizures in disabled persons. Appl Res Ment Retard 1986;7:285-98. Pedley TA. Differential diagnosis of episodic symptoms. Epilepsia 1983;24(suppl 1):S3144. Rarnani SV, Quesney LF, Olson D, Gumnit RJ. Diagnosis of hysterical seizures in epileptic patients. Am J Psychiatry 1980;137:705-9. Roy A. Hysterical fits previously diagnosed as epilepsy. Psycho/ Med 1977;7:271-3. Scott DF. Recognition and diagnostic aspects of nonepileptic seizures. In: Riley TL, Roy A, eds. Pseudoseizures. Baltimore: Williams & Wilkins, 1982:21-33. Silver LB. Conversion disorder with pseudoseizures in adolescence: a stress reaction to unrecognized and untreated learning disabilities. A m Acad Child Psychiatry 1982;21:50&12. Trimble MR. Pseudoproblems, pseudoseizures. Br J Hosp Med 1983;29:326-30. Zivin L , Ajmone-Marsan C. Incidence and prognostic significance of “epileptiform” activity in the EEG of nonepileptic subjects. Brain 1968;91 :75 1-78,
RESUME Les dossiers de 27 enfants hospitalises pour bilan d’une epilepsie intratable, mais diagnostiquee ulterieurement comme presentant des evenements non epileptiques par monitorage EEG-video, ont CtC re-examines. Les auteurs ont identifie 4 groupes: evenements purement psychogeniques (5 patients); tvenements psychogeniques associes a des crises epileptiques (3 patients), evenements physiologiques non epileptiques purs (5
Epilepsia, Vol. 32, N o . 3 , 1991
328
M . E . METRICK ET AL.
patients); evenements physiologiques non epileptiques associts a des crises (14 patients). Les antecedents, I’examen clinique et I’evaluation neuro-diagnostique (incluant les EEG anterieurs, les examens neuroradiologiques et les tests neuropsychologiques) ont ete re-evaluks. Les enfants de tous les groupes, a l’exception de ceux presentant des crises purement psychogeniques, avient des antecedents de crises epileptiques de plusieurs types, identifies par les parents ou par le personnel qui les avait en charge. Les antecedents d’etat de ma1 epileptique ont ete observes chez 64% des case (sur 22 patients), y compris chez 1 1 patients sur 14 presentant des evtnements physiologiques associes a des crises. Les anomalies a I’examen neurologique Ctaient frequentes, tout particulitrement chez des enfants presentant des evenements physiologiques non epileptiques. Tous les patients saut 2 avaient des antecedents d’anomalies intercritiques d’allure Cpileptique sur des EEG de routine anterieurs. L’identification des Cvenements non epileptiques a permis d’arriter completement le traitement antiepileptique chez 8 patients (30%), et le nombre total de medicaments antiepileptiques a etC reduit chez 9 patients supplementaires (3%). Le diagnostic d’evenements non epileptiques doit i t r e consider6 chez tous les enfants presentant des crises rebelles ou plusieurs types d e crises associes. Des constatations anormales sur les EEG intercritiques de routine peuvent rendre ce diagnostic difficile. Une surveillance neurodiagnostique EEG at video intensive represente la modalit6 de choix pour le diagnostic differentiel des evenements non Cpileptiques et des crises epileptiques. (P. Genton, Marseille)
RESUMEN Se han revisado las historias medicas de 27 ninos ingresados en el MINCEP, programa de Epilepsia, para valoracion de epilepsia incontrolable per0 que, mas tarde, se encontro que existian episodios no-epilepticos segun el EEG con monotorizaci6n simultanea con video. Se identificaron cuatro grupos: acontecimientos psicongenicos puros ( 5 pacientes); acontecimientos psicogenicos mas ataques epilepticos (3 pacientes); acontecimientos puros fisiologicos no-epilepticos ( 5 pacientes) y acontecimientos fisiologicos no-epilepticos mas ataques (I4 pacientes). Se revisaron 10s datos historicos, las exploraciones fisicas y las evaluaciones neurodiagnosticas (incluyendo previos EEGs, evaluaciones neuroradiologicas y estudios neuropsicoIogicos). En todos 10s grupos, 10s ninos, con la excepcion de 10s que tenian ataques psicogenicos puros, tenian una historia de mutiples tipos de ataques identificados por 10s padres. Se obtuvo una historia de “status epilepticus” en 64% de 22 enfermos incluyendo 1 I de 10s 14 enfermos con acontecimientos fisiologicos mas ataques. Los hallazgos anormales en la exploracion neuro-
Epilepsia, Vol. 32, N O . 3 , 1991
Iogica fueron comunes, especialmente en ninos con aconb cimientos fisiologicos no epilepticos. Todos 10s pacientes, mew 2, tenian una historia de anomalias epileptiformes interictalesa 10s EEG previos rutinarios. La medication antiepilktica se inta rumpio completamente en 8 enfermos (30%) al identificarh acontecimientos no epilepticos y el numero total de mediw ciones antiepilepticas se redujo en 9 casos (33%). El diagndstim de acontecimientos no epilepticos debe ser considerado en todas 10s ninos con ataques refractarios o con ataques de multiples tipos. Las anomalias en EEG interictales rutinarios pueden COB. fundir el diagnostico. El estudio intensivo neurodiagndstico con EEG y registros con video es la modalidad de elecci6n pan distinguir ataques epilepticos d e 10s no-epileticos. (A. Portera-Sanchez, Madrid
ZUSAMMENFASSUNG Von 27 Kindern wurden die Krankengeschichten nochmals gesichtet. Sie waren zur Behandlung einer therapieresistenten Epilepsie zugewiesen worden -durch EEG und simultane Videoaufnahmen konnte das Vorliegen nicht epileptischer Anfae bewiesen werden. 4 Gruppen wurden identifiziert: 1. reine psy. chogene Anfalle (N = S ) , psychogene Anfalle und epileptische Anfalle (N = 3); reine nicht-epileptische physiologische Ereignisse (N = 5 ) and nicht-epileptische physiologische Ereignisse und Anfalle (N = 14). Vorgeschichte, korperliche Untersuchung and Neurodiagnostik (einschlieplich vorangegangener EEG. Untersuchungen, neuroradiologischer Abklarung und neuropsychologischen Testen) wurden gesichtet. Die Kinder aller Grup pen ausgenommen mit reinen psychogenen Anfallen, hoten eine Krankengeschichte mit zahlreichen Anfallen, die durch Eltern oder Betreuer identifiziert wurden. In 64% (22 Patienten) konnte ein Status epilepticus gefunden werden einschlieplich 1 1 von 14 Patienten mit physiologischen Ereingnissen und Anfallen. Auffallige neurologische Befunde waren haufig, besonders bei Kindern mit nicht-epileptischen physiologischen Ereignissen. In den vorangegangenen EEGs hatten alle, auper 2 Patienten interiktal epileptische Entladungen. Aufgrund der Diagnose nichtepileptische Anfalle, konnten bei 8 Patienten (30%) die Antiepileptika vollstandig abgesetzt werden und die Gesamtmenge antiepileptischer Medikation bei 9 weiteren (33%) ebenfalls reduziert werden. Bei allen Kindern mit therapieresistenten oder unterschiedlichen Anfallen sollten nicht-epileptische Anfalle differntialdiagnostisch in Betracht gezogen werden. Auffallige Befunde im interiktalen EEG konnen die Diagnose verschleiern. Intensive neurodiagnostische EEG-Videoanalyse stellt das Mittel der Wahl dar, um nicht-epileptische von epileptischen Anfallen zu unterscheiden. (C. G.Lipinski, HeidelberglNeckargemiind)
Nonepileptic Events in Childhood *tM. E. Metrick, *tF. J. Ritter, *tJ. R. Gates, *M. P. Jacobs, "S. S. Skare, and *R.B. Loewenson *Epilepsy Research Center, Department of Neurology, University of Minnesota; and tMINCEP Epilepsy Cure, P.A., Minneapolis, Minnesota, U . S . A .
Summary: The medical records of 27 children admitted to the MINCEP Epilepsy Program for evaluation of intractable epilepsy but later shown to have nonepileptic events by EEG with simultaneous video monitoring were reviewed. Four groups were identified: pure psychogenic events ( 5 patients), psychogenic events plus epileptic seizures (3 patients), pure nonepileptic physiologic events (5 patients), and nonepileptic physiologic events plus seizures (14 patients). Historical data, physical examinations, and neurodiagnostic evaluations (including previous EEGs, neuroradiologic evaluations, and neuropsychologic testing) were reviewed. Children in all groups, except for those with pure psychogenic seizures, had a history of multiple seizure types identified by parents or caretakers. A history of status epilepticus was obtained in 64% (of 22 patients), including 1 1 of 14 patients with phys-
iologic events plus seizures. Abnormal findings on neurologic examination were common, especially in children with nonepileptic physiologic events. All but two patients had a history of interictal epileptiform abnormalities on previous routine EEGs. Based on identification of nonepileptic events, antiepileptic drugs (AEDs) were discontinued completely in eight patients (30%) and the total number of AEDs was reduced in nine others (33%). A diagnosis of nonepileptic events should be considered in all children with refractory seizures or multiple seizure types. Abnormal findings on routine (interictal) EEG may actually confound the diagnosis. Intensive neurodiagnostic EEG-video recording is the preferred method for distinguishing nonepileptic from epileptic seizures. Key Words: Epilepsy-Children-Nonepileptic-EEG monitoring.
Between 10 and 20% of patients referred to epilepsy centers for seizure control are later shown to have conditions other than epilepsy (Holmes et al., 1980; Scott, 1982; Duchowny et al., 1988). An epileptic seizure is defined as the paroxysmal change in behavior associated with a burst of epileptiform electrical activity in the brain. On the other hand, a nonepileptic seizure can be considered a paroxysmal change in behavior unassociated with bursts of epileptiform electrical brain activity. The nonepileptic events can be categorized as either psychogenic or physiologic. In adults, snd less often in children, psychogenic events mimicking true seizures have been recognized in both epileptic and nonepileptic populations. Scott (1982) estimated the prevalence of psychogenic seizures at 5% of an outpatient epilepsy population. In a study of 53 pediatric patients ad-
mitted to an epilepsy unit, Holmes et al. (1980) documented psychogenic seizures in 1 1 patients (21%), eight of whom also had epileptic seizures. Previous studies have identified several risk factors associated with an increased tendency for psychogenic seizures; in particular, a history of sexual abuse (Roy, 1977; LaBarbera and Dozier, 1980). In children, undetected learning disabilities have also been identified as precipitating factors (Silver, 1982). In addition to psychogenic events, some physiologic conditions in infancy and childhood resemble epilepsy. Many of the more common types of paroxysmal nonepileptic events are identified by history and physical examination; therefore, individuals who experience them are rarely referred to epilepsy centers for treatment or clarification. Common categories include breath-holding spells, pallid infantile syncope, gastroesophageal reflux associated with laryngospasm and apnea, night terrors and other parasomnias, and complicated migraine (Pedley, 1983). In addition, staring episodes and various repetitive behaviors that may not be clini-
Received August 1989; revision accepted May 1990. Address correspondence and reprint requests to Dr. M. E. Metrick at Lutheran General Children's Medical Center, 1775 Dempster St., Park Ridge, IL 60068, U.S.A.
322
323
NONEPILEPTIC EVENTS IN CHILDHOOD cally distinguishable from epileptic seizures are common in mentally retarded persons, who are also at increased risk for epilepsy. Several reports have described physiologic nonepileptic events that have been identified during inpatient evaluations at epilepsy centers. In a retrospective study by Duchowny et al. (1988), of 60 children admitted to an epilepsy unit, 40% of the children identified as having nonepileptic events had either staring episodes or rhythmic movements such as rocking, shaking, and arm waving. A study by Neil1 and Alvarez (1986) of nonepileptic events in 124 mentally retarded patients showed that the most frequently identified nonepileptic events (as determined by radiotelemetry video-EEG recording) included staring episodes, cessation of movement, and a variety of repetitive movements. Diagnosis of children with nonepileptic events may be difficult. The history given by the patient’s caretaker often fits the description of an epileptic seizure. Many of the children with nonepileptic events, especially those who also have seizure disorders, have epileptiform abnormalities on their routine (interictal) EEG. The incidence of epileptiform abnormalities, defined as paroxysmal discharges containing spikes or sharp-waves, has been estimated to occur in 2.2% of the nonepileptic population (Zivin and Ajmone-Marsan, 1968). Failure to recognize these patients can result in unnecessary use of antiepileptic drugs (AEDs), with the subsequent risk of long-term side effects, behavioral and cognitive disturbances, and systemic toxicity. In the extreme example, misdiagnosis of status epilepticus has led to treatment with general anesthesia and mechanical ventilation (Krumholz and Niedermeyer, 1983). In addition, use of AEDs may actually increase the incidence of nonepileptic events in patients with psychogenic seizures by exacerbating their mood disturbance (Trimble, 1983). At the same time, these patients are not receiving appropriate medical or psychiatric intervention. Use of simultaneous EEG-video monitoring to correlate behaviors with temporally related EEG changes may clarify the nature of the events and thus lead to appropriate treatment plans (Ramani et al., 1980; Desai et al., 1982; King et al., 1982). In this retrospective study, we reviewed the records of 222 children referred to an epilepsy center for treatment of intractable epilepsy, 27 of whom were later identified as having nonepileptic events. METHODS The records of all children aged < 16 years admitted to the MINCEP Epilepsy Program for Children
at Gillette Children’s Hospital in St. Paul, Minnesota, between August 1986 and August 1988 were reviewed. All patients had been referred for evaluation and treatment of seizures that had been refractory to standard medical therapy. The patient population included children with both normal and abnormal intelligence and patients with suspected neurodegenerative conditions. Both preadmission data and information obtained during hospitalization were reviewed. These data included historical information, physical examinations, and neurodiagnostic evaluations (radiologic studies, routine and video-EEGs, and neuropsychologic testing). While at Gillette Children’s Hospital, all children had a minimum of 24 h of consecutive EEG recordings with simultaneous video monitoring in both the awake, drowsy, and sleep states on 16-channel cable telemetry using the International 10-20 System. Spontaneous events were recorded both before and after AED dosages were reduced. Photic stimulation was performed on all patients, and hyperventilation was performed when possible. Sphenoidal electrodes were used in selected individuals. Both parents and trained observers, including nurses and nurses aides, assisted in documenting events. The presence of epileptic seizures was based on identification of paroxysmal discharges on EEG that were temporally linked to the documented episodes recorded on video tape. Events were classified as nonepileptic when there was a lack of epileptiform correlation to the episodes and when the events also were not stereotyped and did not progress into other seizure types as medications were decreased or eliminated. An ability to interrupt events by verbal or physical intervention also aided in differentiating nonepileptic from epileptic seizures. After nonepileptic events were identified, therapeutic regimens were examined to determine if proper diagnosis resulted in simplification of AED regimens or in redirection into appropriate therapies. Historical factors and findings on neurodiagnostic testing and physical examination, which differed between groups, were identified. When possible, these were assessed by statistical analysis (Fisher’s exact test); however, statistical analysis was limited by the small sample sizes. RESULTS Two hundred twenty individuals were admitted to the center during the study period. Of these, 27 patients (12%) were identified as having nonepileptic events by prolonged video-EEG monitoring (9 males and 18 females aged 7 months to 16 years; median age 8.4 years). The patients represented Epilepsia, Vol. 32, N o . 3, 1991
M . E . METRICK ET AL.
324
four groups: pure psychogenic seizures (5 patients), psychogenic events plus epileptic seizures (3 patients), pure physiologic events (5 patients), and physiologic events plus epileptic seizures (14 patients). The most frequent nonepileptic physiologic events in our population were a child's unusual reaction to environmental stimuli that had been misinterpreted by parents as seizures. These reactions included nonepileptic staring spells, repetitive stereotypic behaviors, and abnormal movements secondary to abnormalities in muscle tone. In two patients, the nonepileptic events consisted of paroxysmal choreoathetosis which in one patient was medication related. Differences among the four groups were identified. Patients with psychogenic nonepileptic events, with or without seizures (n = 8) had a median age of 12.7 years; patients with physiologic nonepileptic events, with or without seizures (n = 9), had a median age of 6.3 years. The children with physiologic nonepileptic events, especially those who also had seizures, were most likely to have movement disorders or major abnormalities on neurologic examination (Table 1). All five children with pure psychogenic events had normal intellectual function as determined by IQ testing and/or school placement, whereas all but one of the children in the other groups were mentally retarded (Table 1). Intelligence testing was compared between groups, using standardized IQ scores [Stanford-Binet or Wechsler Intelligence Scale for Children-Revised (WISC-R)]
when available. In the pure psychogenic subgroup, Full-Scale IQ scores on WISC-R ranged from 95 to 144. An IQ score was unavailable for one child with pure psychogenic seizures who was in a regular classroom setting. In the psychogenic plus seizure subgroup, two children had a WISC-R Full-scale IQ of 40 and 64, respectively. The third child, whose IQ scores were unavailable, was in a remedial educational program. In the pure physiologic subgroup, one child, aged 9 months, tested in the normal range for age on Bayley Scales. The other children evidenced mild to profound mental retardation, either as determined by formal IQ testing or by remedial school placement. In the mixed physiologic plus seizures subgroup, seven patients had no formal IQ testing because of profound mental retardation. The other children scored in the mild to severely mentally retarded range on standardized tests. The children with pure psychogenic events were female and, as evidenced by social services evaluation and psychologic testing, belonged to dysfunctional family units. None of the children with pure psychogenic events had a history of a first-degree relative with seizures. All children in the pure psychogenic group had only one seizure type identified by parents on admission, whereas multiple seizure types (range two to seven types) were identified in the other nonepileptic groups (Table 1). This difference was greatest between the pure psychogenic and the physiologic plus seizures groups (p =
TABLE 1. Results of neurologic examinations and tests of intellectual function, frequency of seizure types per patient identified at admission, and history of previous EEGs Psychogenic (n = 5) Major abnormalities on neurologic examination Dysmorphic features Movement disorder Normal intellectual function" Abnormal intellectual function" No. of seizure types 1 L
3 4 5
6 7 Mean no. of seizure types per person History of status epilepticus History of all EEGs normal History of all epileptiform EEGs History of both normal and epileptiform EEGs
Psychogenic + seizures (n = 3) 1
0 0 0 5 0
0 0 0 0 3
3 2 I 4
5 0 0 0 0 0 0 I .o 1 ( I unknown) 2 1
0 0 2 0 1 0 0 3.7 1 ( I unknown) 0 1
0 1 0 2 1 1 0 4.2 I (1 unknown) 0 2
2
l ( 1 unknown)
3
" Intellectual function was determined by IQ testing and/or school placement. Epilepsia, Vol. 32, No. 3, 1991
Physiologic (n = 5)
Physiologic + seizures (n = 14) 7 4 5 0 14 0 2
5 4 0 2 1 3.9 I 1 (2 unknown) 0 8 6
325
NONEPILEPTIC EVENTS IN CHILDHOOD O.OOOl), but was also significant between the pure psychogenic and the pure physiologic groups (p = 0.004) and the pure psychogenic and the psychogenic plus seizures groups (p = 0.018). In the pure psychogenic group, nonepileptic events resembled complex partial seizures, with a frequency of events ranging from 12 to 120 episodes per month. All these events had paroxysmal onset. In four of the 5 patients in this group the event began with a brief period of staring; the fifth patient usually cried out, or the event began with automatisms of the upper extremities. During these events, all 5 of the patients were unresponsive to verbal command, and 3 of the 5 resisted physical intervention. They all reported amnesia for the event and feeling tired after the nonepileptic episode. All 5 patients had movements described as automatisms, and during the event 4 of the 5 had intermittent stiffening for 3-10 s. Three of the 5 patients had episodes that included intermittent hyperventilation. In the other groups, the events resembled a variety of seizure types including generalized tonicclonic convulsions, complex partial seizures, absence seizures, and tonic, atonic, and myoclonic seizures. In the mixed psychogenic plus epileptic seizure group, episodes were quite variable. One patient had dysrhythmic jerking of the right upper extremity, associated with decreased responsiveness. Another complained of feeling warm, then became unresponsive to verbal intervention and complained of headache after the episode. The third patient in the psychogenic plus epileptic group had nonepileptic episodes of multiple types, including staring and decreased responsiveness that either did or did not progress into generalized stiffening, followed by shaking. The patients in the pure physiologic group all demonstrated repeated, paroxysmal nonepileptic events, each with motor activity accompanied by a decrease in responsiveness. These episodes varied from staring with apnea, followed by stiffening of all extremities, to nonepileptic myoclonus. The physiologic plus epileptic seizure group included eight children with movement disorders. The most common movement disorders observed were paroxysmal choreoathetosis, paroxysmal dystonia, and nonepileptic myoclonus. In this group, nonepileptic events with automatic, repetitive, self-stimulatory behaviors were also frequent. These behaviors always began abruptly, interrupting ongoing activity. More than half of these nonepileptic events began with staring and then rapidly progressed to the automatic repetitive behavior. An example was a teenage girl who-suddenly stopped all activity, stared, had a frightened look on her
face, turned her head toward the left, and repeatedly made chewing motions. She was unresponsive to verbal commands. At times, these events could be interrupted by physical intervention. A history of status epilepticus was obtained in 64% of the 22 patients for whom data were available and was particularly common in patients with physiologic events plus seizures ( 1 1 of 14 patients). Fourteen of the 27 patients, representing all groups except those with psychogenic events and seizures, had a history of abnormal computed tomography (CT) scans or magnetic resonance imaging (MRI) studies. Four patients had diffuse abnormalities, 7 patients had focal abnormalities, and 3 patients had both diffuse and focal abnormalities. Most patients in all groups had a history of epileptiform abnormalities on EEG, and at least 1 patient in each group had a history of both normal EEGs and EEGs with interictal epileptiform discharges (Table 1). Of the 10 patients with pure nonepileptic events, 7 had a history of interictal epileptiform abnormalities based on outside records (outside EEGs were not available for review). Five of the 10 patients had interictal epileptiform abnormalities recorded during their hospitalization at Gillette Children’s Hospital. Based on identification of nonepileptic events (in all groups), AEDs were completely discontinued in 8 patients (30%)and the total number of AEDs was reduced in 9 others (33%). All patients in the pure psychogenic and pure physiologic groups were no longer taking AEDs at the time of discharge. DISCUSSION
Most studies of nonepileptic events are based on the adult population in which psychogenic seizures appear to be the most common nonepileptic condition misdiagnosed as epilepsy. Relatively few studies have been made of nonepileptic events in childhood. In our study we observed that, in contrast to adults, children misdiagnosed with seizure disorders were more likely to have physiologic (19 patients) than psychogenic (8 patients) nonepileptic events. The basis for misdiagnosis was usually parental misinterpretation of behaviors, especially staring episodes, abnormal reactions to environmental stimuli, and repetitive movements. In addition, unusual movements based on abnormal muscle tone were often misinterpreted as seizures. Parents’ or caretakers’ descriptions of these movements and behaviors were often indistinguishable from those of epileptic seizures. Occasionally, a trained professional could identify the event as nonepileptic by Epitepsia. Vol. 32, N o . 3, I991
M . E . METRICK ET AL. observation alone, but often the distinction between epileptic and nonepileptic episodes could be made only after careful review of simultaneous EEGvideo monitoring. Other neurodiagnostic methods were often misleading. In particular, in 96% of these children, interictal EEG showed epileptiform abnormalities that might have led to an erroneous diagnosis. In our study, almost all patients in every group, including patients with pure psychogenic or pure physiologic events, had EEGs with interictal epileptiform abnormalities (Table 1). Epileptiform abnormalities on interictal recordings are particularly common in children with both epileptic and nonepileptic events, which emphasizes the need to observe and record all of the types of seizures in question. Other findings such as focal abnormalities on CT or MRI scan, observed in 37% of our patients, could further confound the diagnosis. The events in our study resembled a variety of seizure types. Pure psychogenic events were commonly misinterpreted as complex partial seizures, similar to the findings by Holmes et al. (1980). Events in the other groups resembled a variety of seizure types, including generalized tonic-clonic convulsions, absence seizures, and myoclonic, tonic, and atonic seizures. These differences may be related to the different etiologies of the nonepileptic events; e.g., staring, misdiagnosed as absence seizures, may reflect the decreased attention span common in children with varying degrees of mental subnormality. Head nods, misdiagnosed as atonic or tonic seizures, may reflect hypotonia. Our findings are similar to those of Duchowny et al. (1988) and Neil1 and Alvarez (1986), who reported that repetitive behaviors were commonly misinterpreted as seizures, particularly in the more severely mentally retarded. The children with pure psychogenic seizures may represent a patient population different from the other groups. Although our patient population was too small for statistical analysis of most variables, certain features in this group appear to differentiate them from the others. In particular, all these children were of normal intelligence, whereas all but one of the patients in the other groups had some degree of intellectual impairment. This may reflect the patient selection process of our center, but may also characterize children with intractable epilepsy. In addition, patients with pure psychogenic seizures generally had only one seizure type (mean number of types per group = 1 . l ) , usually resembling complex partial seizures (Table 1). Patients in the other groups had multiple seizure types identified by parents on admission; e.g., in the pure physiologic group, one patient was originally reported to have Epilepsia, Val. 32, N o . 3, 1991
six seizure types (mean for the group = 4.2). The children with pure psychogenic seizures were also older than those with pure physiologic events, with median ages of 12.8 and 2.6 years, respectively. Previous studies have identified some common stressors as precipitating psychogenic seizures, especially a history of sexual abuse o r undetected learning disabilities. Although our patients with pure psychogenic seizures had no documented history of sexual abuse, a previous history of abuse was suspected in several cases. All family units were identified as dysfunctional by social services personnel and/or by psychological evaluation. One of the patients, though of normal intelligence, had a Verbal and Performance IQ split of 24 IQ points (on the WISC-R) and was described as having difficulty integrating information on neuropsychological testing. These findings suggest that this patient, who was in a regular classroom setting, may have learning disabilities that contribute to her nonepileptic events. Although AEDs were successfully discontinued in all three patients in the psychogenic group who were admitted while receiving seizure medication, long-term follow-up will be needed to assess the impact of identifying these events properly. Certain features of the nonepileptic events, in addition to findings on video monitoring, supported the diagnosis of nonepileptic conditions. All nonepileptic events were refractory to conventional AEDs. Our patient population was selective in that most patients referred to our epilepsy program are refractory to AEDs. In addition, the number of patients who had been treated for status epilepticus because of lack of response to AEDs was striking; I 1 of 14 patients in the group of physiologic nonepileptic events plus seizures had a previous history of status epilepticus, an unusually high incidence which suggests that at least some of these episodes may have been nonepileptic in nature. Another characteristic of nonepileptic seizures that contrasted with those of epileptic seizures was the variability of seizure expression. Many of the patients with pure physiologic events had multiple seizure types identified by parents. Although patients with pure psychogenic seizures had only one seizure type, review of the video tapes disclosed a lack of stereotypy from one event to the next. This contrasts to the usually stereotyped epileptic seizures. Furthermore, these events almost always could be interrupted by either verbal and/or physical intervention, whereas true epileptic seizures could not be interrupted. The most difficult distinction was differentiating certain nonepileptic seizures from simple partial seizures, which may also show little or no surface
327
NONEPILEPTIC EVENTS IN CHILDHOOD TABLE 2. Reductions in AEDs at time of discharge
Status
Psychogenic (n = 5 )
Psychogenic + seizures (n = 3)
Physiologic + seizures (n = 14)
-.
~~
Medication reductions Complete discontinuation No. of patients whose AEDs were reduced Decreased dosage Total No AEDs on admission and discharge No change in AEDs Change to appropriate medication a
Physiologic (n = 5 )
0
1
1
2 0
8“ 2h 11 0
I
2
0
1
Ih
Five patients in this group had their medication changed to the appropriate AED.
* The patient’s medication was changed to the appropriate AED. AED, antiepileptic drug.
EEG expression. In addition to the above distinguishing features, certain additional clues helped us properly distinguish between simple partial seizures and nonepileptic events. The true epileptic seizures were more likely to progress to CPS or GTC convulsions as medications were discontinued and to respond to AEDs to some degree. On occasion, simple partial seizures that showed no EEG expression while the patients were receiving AEDs would show some corresponding EEG changes after AEDs were decreased. By taking into account these differences, we usually could distinguish between the two groups. In our patients, the nonepileptic events resulted in admission for refractory seizures, whereas epileptic seizures were adequately controlled. Proper identification of nonepileptic events resulted in a marked reduction in AED therapy. AEDs were completely discontinued in 30% of our patient population (8 patients), and the total number of AEDs was reduced in an additional 33% (9 patients) (Table 2). The patients in the pure psychogenic and pure physiologic groups had discontinued all AEDs at the time of discharge. If indicated, these patients were redirected into other treatment modalities. Thus, recognition of nonepileptic events had significant impact on the treatment plans for these children. Abnormal findings on EEG, neurologic examination, and neuroradiologic studies were common in our patients and actually confounded the appropriate diagnosis. Only intensive EEG with simultaneous video monitoring reliably led to the appropriate diagnosis. Thus, intensive neurodiagnostic EEG-video monitoring is the preferable method for identifying nonepileptic events. Its use should be considered in all children with multiple seizure types or refractory seizures. Acknowledgment: This work was supported in part by NIH-NINDS Grant No. P50 16308 and by a John Hugh-
lings Jackson Clinical Research Fellowship from the Epilepsy Foundation of America.
REFERENCES Desai BT, Porter RJ, Penry JK. Psychogenic seizures: a study of 42 attacks in six patients, with intensive monitoring. Arch Neurol 1982;39:202-9. Duchowny MS, Resnic TJ, Deray MJ, Alvarez LA. Video EEG diagnosis of repetitive behavior in early childhood and its relationship to seizures. Pediatr Neurol 1988;4:1 6 2 4 . Holmes GL, Sackellares JC, McKiernon J, et al. Evaluation of childhood pseudoseizures using EEG telemetry and video tape monitoring. J Pediatr 1980;97:554-8. King DW, Gallagher BB, Murvin AJ, et al. Pseudoseizures: diagnostic evaluation. Neurology 1982;32:18-23. Krumholz A, Niedermeyer E. Psychogenic seizures: a clinical study with follow-up data. Neurology 1983;33:498-502. LaBarbera JD, Dozier JE. Hysterical seizures: the role of sexual exploitation. Psychosomatics 1980;21:897-903. Neil1 JC, Alvarez N. Differential diagnosis of epileptic versus pseudoepileptic seizures in disabled persons. Appl Res Ment Retard 1986;7:285-98. Pedley TA. Differential diagnosis of episodic symptoms. Epilepsia 1983;24(suppl 1):S3144. Rarnani SV, Quesney LF, Olson D, Gumnit RJ. Diagnosis of hysterical seizures in epileptic patients. Am J Psychiatry 1980;137:705-9. Roy A. Hysterical fits previously diagnosed as epilepsy. Psycho/ Med 1977;7:271-3. Scott DF. Recognition and diagnostic aspects of nonepileptic seizures. In: Riley TL, Roy A, eds. Pseudoseizures. Baltimore: Williams & Wilkins, 1982:21-33. Silver LB. Conversion disorder with pseudoseizures in adolescence: a stress reaction to unrecognized and untreated learning disabilities. A m Acad Child Psychiatry 1982;21:50&12. Trimble MR. Pseudoproblems, pseudoseizures. Br J Hosp Med 1983;29:326-30. Zivin L , Ajmone-Marsan C. Incidence and prognostic significance of “epileptiform” activity in the EEG of nonepileptic subjects. Brain 1968;91 :75 1-78,
RESUME Les dossiers de 27 enfants hospitalises pour bilan d’une epilepsie intratable, mais diagnostiquee ulterieurement comme presentant des evenements non epileptiques par monitorage EEG-video, ont CtC re-examines. Les auteurs ont identifie 4 groupes: evenements purement psychogeniques (5 patients); tvenements psychogeniques associes a des crises epileptiques (3 patients), evenements physiologiques non epileptiques purs (5
Epilepsia, Vol. 32, N o . 3 , 1991
328
M . E . METRICK ET AL.
patients); evenements physiologiques non epileptiques associts a des crises (14 patients). Les antecedents, I’examen clinique et I’evaluation neuro-diagnostique (incluant les EEG anterieurs, les examens neuroradiologiques et les tests neuropsychologiques) ont ete re-evaluks. Les enfants de tous les groupes, a l’exception de ceux presentant des crises purement psychogeniques, avient des antecedents de crises epileptiques de plusieurs types, identifies par les parents ou par le personnel qui les avait en charge. Les antecedents d’etat de ma1 epileptique ont ete observes chez 64% des case (sur 22 patients), y compris chez 1 1 patients sur 14 presentant des evtnements physiologiques associes a des crises. Les anomalies a I’examen neurologique Ctaient frequentes, tout particulitrement chez des enfants presentant des evenements physiologiques non epileptiques. Tous les patients saut 2 avaient des antecedents d’anomalies intercritiques d’allure Cpileptique sur des EEG de routine anterieurs. L’identification des Cvenements non epileptiques a permis d’arriter completement le traitement antiepileptique chez 8 patients (30%), et le nombre total de medicaments antiepileptiques a etC reduit chez 9 patients supplementaires (3%). Le diagnostic d’evenements non epileptiques doit i t r e consider6 chez tous les enfants presentant des crises rebelles ou plusieurs types d e crises associes. Des constatations anormales sur les EEG intercritiques de routine peuvent rendre ce diagnostic difficile. Une surveillance neurodiagnostique EEG at video intensive represente la modalit6 de choix pour le diagnostic differentiel des evenements non Cpileptiques et des crises epileptiques. (P. Genton, Marseille)
RESUMEN Se han revisado las historias medicas de 27 ninos ingresados en el MINCEP, programa de Epilepsia, para valoracion de epilepsia incontrolable per0 que, mas tarde, se encontro que existian episodios no-epilepticos segun el EEG con monotorizaci6n simultanea con video. Se identificaron cuatro grupos: acontecimientos psicongenicos puros ( 5 pacientes); acontecimientos psicogenicos mas ataques epilepticos (3 pacientes); acontecimientos puros fisiologicos no-epilepticos ( 5 pacientes) y acontecimientos fisiologicos no-epilepticos mas ataques (I4 pacientes). Se revisaron 10s datos historicos, las exploraciones fisicas y las evaluaciones neurodiagnosticas (incluyendo previos EEGs, evaluaciones neuroradiologicas y estudios neuropsicoIogicos). En todos 10s grupos, 10s ninos, con la excepcion de 10s que tenian ataques psicogenicos puros, tenian una historia de mutiples tipos de ataques identificados por 10s padres. Se obtuvo una historia de “status epilepticus” en 64% de 22 enfermos incluyendo 1 I de 10s 14 enfermos con acontecimientos fisiologicos mas ataques. Los hallazgos anormales en la exploracion neuro-
Epilepsia, Vol. 32, N O . 3 , 1991
Iogica fueron comunes, especialmente en ninos con aconb cimientos fisiologicos no epilepticos. Todos 10s pacientes, mew 2, tenian una historia de anomalias epileptiformes interictalesa 10s EEG previos rutinarios. La medication antiepilktica se inta rumpio completamente en 8 enfermos (30%) al identificarh acontecimientos no epilepticos y el numero total de mediw ciones antiepilepticas se redujo en 9 casos (33%). El diagndstim de acontecimientos no epilepticos debe ser considerado en todas 10s ninos con ataques refractarios o con ataques de multiples tipos. Las anomalias en EEG interictales rutinarios pueden COB. fundir el diagnostico. El estudio intensivo neurodiagndstico con EEG y registros con video es la modalidad de elecci6n pan distinguir ataques epilepticos d e 10s no-epileticos. (A. Portera-Sanchez, Madrid
ZUSAMMENFASSUNG Von 27 Kindern wurden die Krankengeschichten nochmals gesichtet. Sie waren zur Behandlung einer therapieresistenten Epilepsie zugewiesen worden -durch EEG und simultane Videoaufnahmen konnte das Vorliegen nicht epileptischer Anfae bewiesen werden. 4 Gruppen wurden identifiziert: 1. reine psy. chogene Anfalle (N = S ) , psychogene Anfalle und epileptische Anfalle (N = 3); reine nicht-epileptische physiologische Ereignisse (N = 5 ) and nicht-epileptische physiologische Ereignisse und Anfalle (N = 14). Vorgeschichte, korperliche Untersuchung and Neurodiagnostik (einschlieplich vorangegangener EEG. Untersuchungen, neuroradiologischer Abklarung und neuropsychologischen Testen) wurden gesichtet. Die Kinder aller Grup pen ausgenommen mit reinen psychogenen Anfallen, hoten eine Krankengeschichte mit zahlreichen Anfallen, die durch Eltern oder Betreuer identifiziert wurden. In 64% (22 Patienten) konnte ein Status epilepticus gefunden werden einschlieplich 1 1 von 14 Patienten mit physiologischen Ereingnissen und Anfallen. Auffallige neurologische Befunde waren haufig, besonders bei Kindern mit nicht-epileptischen physiologischen Ereignissen. In den vorangegangenen EEGs hatten alle, auper 2 Patienten interiktal epileptische Entladungen. Aufgrund der Diagnose nichtepileptische Anfalle, konnten bei 8 Patienten (30%) die Antiepileptika vollstandig abgesetzt werden und die Gesamtmenge antiepileptischer Medikation bei 9 weiteren (33%) ebenfalls reduziert werden. Bei allen Kindern mit therapieresistenten oder unterschiedlichen Anfallen sollten nicht-epileptische Anfalle differntialdiagnostisch in Betracht gezogen werden. Auffallige Befunde im interiktalen EEG konnen die Diagnose verschleiern. Intensive neurodiagnostische EEG-Videoanalyse stellt das Mittel der Wahl dar, um nicht-epileptische von epileptischen Anfallen zu unterscheiden. (C. G.Lipinski, HeidelberglNeckargemiind)
